jbba
the
table of contents
Editorial
7
Editorial Board
9
Testimonials from Authors & Readers
12
Peer-Reviewed Research
Transitioning to a Hyperledger Fabric Hybrid Quantum Resistant-Classical
Public Key Infrastr ucture
15
The Contractual Cr yptoeconomy:
An Ar row of Time for Economics
27
Singapore’s Open Digital Token Offering Embrace:
Context & Consequences
39
Cr yptocur rency Investing Examined
51
Blockchain Investigations: Beyond the ‘Money’
65
A Blockchain Infrastr ucture for Transpor tation in
Low Income Countr y Cities, and Beyond
75
analytical essay
A Review of fast-g rowing Blockchain Hubs in Asia
83
commentary
Decentralisation is Coming: The Future of Blockchain
101
perspective
Is Blockchain Par t of the Future of Ar t?
109
for authors
113
The JBBA | Volume 2 | Issue 2 | October 2019
3
jbba
the
editorial board
Editor-In-Chief:
Dr. Naseem Naqvi
FBBA FRCP FHEA MAcadMEd MSc
(Blockchain & Cryptocurrency)
Centre for Evidence Based Blockchain, UK
Professor Dr. Shada Alsalamah PhD
(Healthcare Informatics & Blockchain)
Massachusetts Institute of Technology, USA
Professor Dr. Sushmita Ruj PhD
(Applied Cryptography, Security)
Indian Statistical Institute, India
Associate Editor-In-Chief:
Professor Dr. Kevin Curran PhD FBBA
(Cybersecurity)
Ulster University, UK
Professor Dr. Marc Pilkington PhD FBBA
(Cryptocurrencies/ Digital Tech)
University of Burgundy, France
Professor Dr. John Domingue PhD FBBA
(Artificial Intelligence/ Education)
The Open University, UK
Professor Dr. David Lee K Chuen PhD FBBA
(Applied Blockchain)
Singapore University of Social Sciences, Singapore
Professor Dr. Bill Buchanan PhD FBBA
(Cryptography/ Cybersecurity)
Edinburgh Napier University, UK
Contributing Editors & Reviewers:
Professor Dr. Mary Lacity PhD
(Blockchain/ Information Systems)
University of Arkansas, USA
Professor Dr Sandeep Shukla PhD
(Blockchain & Cybersecurity)
Indian institute of Technology, India
Professor Dr. Wulf Kaal PhD
(Blockchain & Law)
University of St. Thomas, USA
Professor Dr. Jason Potts PhD FBBA
(Applied Blockchain)
RMIT University, Australia
Professor Dr. Chris Sier PhD
(DLT in Finance / Capital Markets)
University of Newcastle, UK
Professor Dr. Anne Mention PhD
(Blockchain & Economics)
RMIT University, USA
The JBBA | Volume 2 | Issue 2 | October 2019
Professor Dr. Jim KS Liew PhD FBBA
(Blockchain, Finance, AI)
Johns Hopkins University, USA
Professor Dr. Eric Vermeulen PhD FBBA
(Financial Law, Business, Economics)
Tilburg University, The Netherlands
Professor Dr. Jeff Daniels PhD
(Cybersecurity, Cloud Computing)
University of Maryland, USA
Professor Dr. Mark Lennon PhD
(Cryptocurrencies, Finance, Business)
California University of Pennsylvania, USA
Professor Dr. Walter Blocher PhD
(Blockchain, Law, Smart Contracts)
University of Kassel, Germany
Professor Dr. Clare Sullivan PhD
(Cybersecurity / Digital Identity)
Georgetown University, USA
Professor Dr. Andrew Mangle PhD
(Cryptocurrency, Smart contracts)
Bowie State University, USA
Professor Dr. Isabelle C Wattiau PhD
(Information Systems, Smart Data)
ESSEC Business School, France
Professor Dr. Lee McKnight PhD
(IoT & Blockchain)
Syracuse University, USA
Professor Dr. Chen Liu PhD
(Fintech, Tokenomics)
Trinity Western University, Canada
Professor Dr. Markus Bick PhD
(Business Information Systems)
ESCP Business School, Germany
Professor Dr. Sandip Chakraborty PhD
(Blockchain, Distributed Networks)
Indian Institute of Technology, India
9
jbba
the
Dr. Stefan Meyer PhD
(Blockchain in Food Supply Chain)
University of Leeds, UK
Dr. Anish Mohammed MSc
(Crypto-economics, Security)
Institute of Information Systems, Germany
Dr. Marcella Atzori PhD FBBA
(GovTech/ Smart Cities)
University College London, UK
Demelza Hays MSc
(Cryptocurrencies)
University of Liechtenstein, Liechtenstein
Dr. Mureed Hussain FBBA MD MSc
(Blockchain Governance)
The British Blockchain Association, UK
Alastair Marke FRSA MSc
(Blockchain & Climate Finance)
Blockchain Climate Institute, UK
Dr. Maria Letizia Perugini PhD
(Digital Forensics & Smart Contracts)
University of Bologna, Italy
Adam Hayes MA BS CFA
(Blockchain & Political Sociology)
University of Wisconsin-Madison, USA
Dr. Stylianos Kampakis PhD
(ICOs, Big Data, Token Economics)
University College London, UK
Jared Franka BSc
(Cryptocurrency / Network Security)
Dakota State University, USA
Dr. Phil Godsiff PhD
(Cryptocurrencies)
University of Surrey, UK
Navroop K Sahdev MSc
(Innovation / Applied Blockchain)
Massachusetts Institute of Technology, USA
Dr. Sean Manion PhD FBBA
(Blockchain in Health Sciences)
Uniformed Services University, USA
Raf Ganseman
(DLT in Trade & Music Industry)
KU Leuven University, Belgium
Dr. Duane Wilson PhD
(Cybersecurity/ Computer Science)
The Johns Hopkins University, USA
Sebastian Cochinescu MSc
(Blockchain in Culture Industry)
University of Bucharest, Romania
Dr. Darcy Allen PhD
(Economics/Innovation)
RMIT University, Australia
Jared Polites MSc
(ICOs & Cryptocurrencies)
Blockteam Ventures, USA
Dr. Christian Jaag PhD
(Crypto-economics, Law)
University of Zurich, Switzerland
Managing Editor:
Dr. Larissa Lee JD
(Blockchain & Law)
University of Utah, USA
Dr. Jeremy Kronick PhD
(Blockchain & Finance/ Economics)
C.D Howe Institute, Canada
Saba Arshad MSc
(Machine Learning)
Chungbuk National University, South Korea
Publishing Consultant:
John Bond
(Riverwinds Consulting, USA)
Marketing & Public Relations Assistant:
Dr. Hossein Sharif PhD
(Blockchain, AI, Cryptocurrencies)
University of Newcastle, UK
Dr. Wajid Khan PhD
(Big Data, E-Commerce)
University of Hertfordshire, UK
Tracy Smith
Type-setting, Design & Publishing:
Zeshan Mahmood
Institute Pavoniano Artigianelli, Italy
Dr. Ifigenia Georgiou PhD
(Crypto-economics)
University of Nicosia, Cyprus
The JBBA | Volume 2 | Issue 2 | October 2019
10
jbba
the
The JBBA | Volume 2 | Issue 2 | October 2019
11
jbba
the
Testimonials from Authors and Readers
The JBBA has an outstandingly streamlined submissions process, the reviewers
comments have been constructive and valuable, and it is outstandingly well produced,
presented and promulgated. It is in my opinion the leading journal for blockchain
research and I expect it to maintain that distinction under the direction of its forwardlooking leadership team.
Dr Brendan Markey-Towler PhD, University of Queensland, Australia
It is really important for a future world to be built around peer-review and publishing
in the JBBA is one good way of getting your view-points out there and to be shared
by experts.
Professor Dr. Bill Buchanan OBE PhD, Edinburgh Napier University, Scotland
The JBBA has my appreciation and respect for having a technical understanding
and the fortitude for publishing an article addressing a controversial and poorly
understood topic. I say without hesitation that JBBA has no equal in the world of
scientific Peer-Review Blockchain Research.
Professor Rob Campbell, Capitol Technology University, USA
Within an impressively short time since its launch, the JBBA has developed a strong
reputation for publishing interesting research and commentary on blockchain
technology. As a reader, I find the articles uniformly engaging and the presentation
of the journal impeccable. As an author, I have found the review process to be
consistently constructive.
Dr. Prateek Goorha PhD, Blockchain Researcher and Economist
We live in times where the pace of change is accelerating. Blockchain is an emerging
technology. The JBBA’s swift review process is key for publishing peer-reviewed
academic papers, that are relevant at the point they appear in the journal and beyond.
Professor Daniel Liebau, Visiting Professor, IE Business School, Spain
The JBBA submission process was efficient and trouble free. It was a pleasure to
participate in the first edition of the journal.
Dr. Delton B. Chen PhD, Global4C, USA
The JBBA | Volume 2 | Issue 2 | October 2019
“
“
“
“
“
“
“
“
“
“
“
“
12
jbba
the
This is a very professionally presented journal.
Peter Robinson, Blockchain Researcher & Applied Cryptographer, PegaSys, ConsenSys
“
“
“
Very professional and efficient handling of the process, including a well-designed
hard copy of the journal. Highly recommend its content to the new scientific field
blockchain is creating as a combination of CS, Math and Law. Great work!
“
“
Simon Schwerin MSc, BigChain DB and Xain Foundation, Germany
JBBA has quickly become the leading peer-reviewed journal about the fastest growing
area of research today. The journal will continue to play a central role in advancing
blockchain and distributed ledger technologies.
I had the honour of being an author in the JBBA. It is one of the best efforts
promoting serious blockchain research, worldwide. If you are a researcher, you
should definitely consider submitting your blockchain research to the JBBA.
Dr. Stylianos Kampakis PhD, UCL Centre for Blockchain Technologies, UK
I would like to think of the JBBA as an engine of knowledge and innovation,
supporting blockchain industry, innovation and stimulate debate.
Dr. Marcella Atzori PhD, EU Parliament & EU Commission Blockchain Expert, Italy
“
“
“
“
“
“
John Bond, Senior Publishing Consultant, Riverwinds Consulting, USA
The overarching mission of the JBBA is to advance the common monologue within
the Blockchain technology community. JBBA is a leading practitioners journal for
blockchain technology experts.
“
“
Professor Dr. Kevin Curran PhD, Ulster University, Northern Ireland
The articles in the JBBA explain how blockchain has the potential to help solve
economic, social, cultural and humanitarian issues. If you want to be prepared for the
digital age, you need to read the JBBA. Its articles allowed me to identify problems,
find solutions and come up with opportunities regarding blockchain and smart
contracts.
The JBBA | Volume 2 | Issue 2 | October 2019
“
Professor Dr. Eric Vermeulen, Tilburg University, The Netherlands
13
jbba
the
Peer-reviewed Research
OPEN ACCESS
ISSN Print: 2516-3949
https://doi.org/10.31585/jbba-2-2-(4)2019
Transitioning to a Hyperledger Fabric
Hybrid Quantum Resistant-Classical
Public Key Infrastructure
Robert E. Campbell Sr.
Capitol Technology University, Laurel, USA
Correspondence: rc@medcybersecurity.com
Received: 13 June 2019 Accepted: 26 July 2019 Published: 31 July 2019
Abstract
Hyperledger Fabric (HLF) is a permissioned, blockchain designed by IBM and uses Public Key Infrastructure
(PKI), for digital signatures, and digital identities (X.509 certificates), which are critical to the operational
security of its network. On 24 January 2019, Aetna, Anthem, Health Care Service Corporation, PNC Bank,
and IBM announced a collaboration to establish a blockchain-based ecosystem for the healthcare industry
[1]. Quantum computing poses a devastating impact on PKI and estimates of its large-scale commercial
arrival should not be underestimated and cannot be predicted. The HIPAA (Health Insurance Portability and
Accountability Act) and General Data Protection Regulation (GDPR), requires “reasonable” measures to be
taken to protect Protected Health Information (PHI), and Personally Identifiable Information (PII). However,
HLF’s ecosystem is not post-quantum resistant, and all data that is transmitted over its network is vulnerable
to immediate or later decryption by large scale quantum computers. This research presents independent
evaluation and testing of the National Institute of Standards and Technology (NIST), based Second Round
Candidate Post-Quantum Cryptography (PQC), lattice-based digital signature scheme qTESLA. The secondround submission is much improved, however; its algorithm characteristics and parameters are such that it
is unlikely to be a quantum-resistant “as is,” pure “plug-and-play” function and replacement for HLF’s PKI.
This work also proposes that qTESLA’s public keys be used to create a quantum-resistant-classical hybrid
PKI near-term replacement.
Keywords: Hyperledger Fabric, PKI, HIPAA, GDPR, distributed ledger, post-quantum cryptography, qTESLA, Ring
Learning with Errors, cybersecurity, enterprise blockchains
JEL Classifications: D02, D71, H11, P16, P48, P50
1. Introduction
An X.509 PKI is a security architecture that uses
cryptographic mechanisms to support functions
such as email protection, web server authentication,
signature generation, and validation. It is a specification
upon which applications like Secure Multipurpose
Internet Mail Extensions (S/MIME) and Transport
Layer Security (TLS) are based. It also can be defined
as a collection of methods, rules, policies, and roles
that are required to generate, manage, provide, employ,
and revoke digital certificates; it is also responsible
for the management of public-key encryption.
A PKI ensures the secure transfer of data over
various network infrastructures, such as Intranet and
Internet architectures. HLF’s Enterprise Blockchain,
and in general the secure communications, critical
infrastructure, banking, and Internet commerce
The JBBA | Volume 2 | Issue 2 | October 2019
depends upon the security and reliability of PKI
cryptography. Cryptographic encryption and signature
algorithms are used to ensure confidentially, integrity,
and authenticity of messages, data, and information.
PKI is used to bind identities, and public-keys and
Fabric uses Certificate Authorities (CA), as the primary
trusted party that uses digital signature algorithms to
sign certificates of trust. The architecture, deployment,
and operation of HLF impact the blockchain network’s
cybersecurity risks and determine the controls best
able to mitigate those risks. Key considerations
include the ability of untrusted or unauthorized
persons to participate in the network; and the strength
of the encryption protocols. Advances in quantum
computing are threatening today’s global encryption
standards, including PKI [2]. There is an immediate
need to develop, deploy, and migrate the consortium’s
blockchain ecosystem to a hybrid safe PQC. PQC is
15
jbba
the
cryptosystems which run on classical computers and
are considered to resistant to quantum computing
attacks. There are significant uncertainties associated
with PQC, such as, the possibility of new quantum
algorithms being developed which would cause new
attacks. Also, new PQC algorithms are not thoroughly
tested and analyzed. It takes years to understand
their security in a classical computing environment.
This work evaluates HLF’s blockchain post-quantum
computing vulnerabilities and threats given global
regulatory requirements and provides valuable secondround qTESLA independent testing and evaluation
data and aids in the NIST Post-Quantum Cryptography
Standardization Process [3]. Further, the author
encourages additional independent testing, verification,
and validation of qTESLA as one of the most practical
hybrid quantum-resistant PKI systems.
2. Implications in this Work
Without plans for quantum-resistant cryptography and
security, all data and information, including encrypted,
that is transmitted today, and tomorrow is vulnerable.
This would violate all known regulatory requirements
for data privacy and security. HIPAA enacted in 1996
and is United States legislation that provides security
and data protection for medical information [4].
GDPR requires in the case of a personal data breach
notification not later than 72 hours after having become
aware of it [5]. Both GDPR and HIPAA levies hefty
fines and penalties due to non-compliance. GDPR noncompliance with various provisions of the GDPR shall
be fined according to the gravest infringement, which
can be Up to €20 million, or 4% of the worldwide
annual revenue of the prior financial year, whichever is
higher [6]. HIPAA violations of penalties and fines for
noncompliance are also based on the level of perceived
negligence. These fines can range from $100 to
$50,000 per violation (or per record), with a maximum
penalty of $1.5 million per year for each violation
[7]. It takes years of study and analysis of quantumresistant cryptography algorithms before governments
and industry can trust their security. Given the nature
and the far-reaching implications of the legal and
financial obligations of both these laws, it is essential
to have plans and strategies to address and mitigate
vulnerabilities and threats that may lead to data breaches
and non-compliance. Permissioned blockchains are not
immune to cyber-attacks, and further exploration of
the quantum-resistant cryptography is a necessity, and,
a consensus between industry and regulators regarding
the appropriate cybersecurity standards to apply to
blockchain solutions in the healthcare, financial and
GDPR covered services industry. An honest discussion
and principles approach to cybersecurity regulation
all in mitigating cybersecurity risk in permissioned
blockchains while allowing the technology to continue
to evolve through innovation.
The JBBA | Volume 2 | Issue 2 | October 2019
Failure to comply with HIPAA, GDRP, and other
regulating authorities can result in stiff penalties. Fines
will increase with the volume of data or the number
of records exposed or breached, and the amount of
neglect. The lowest fines begin with a breach when
the rules are not known, and by exercising reasonable
diligence, would not have known the provisions were
violated. At the other end of the spectrum are fines
levied where a breach is due to negligence and not
corrected appropriately.
We need a coordinated strategy and approach with
specific recommendations and policies for academia,
policymakers, and industry participants regarding and
promoting the development of secure blockchain
technologies and applications through viable
cybersecurity standards. The enterprise blockchain
cybersecurity risks must be understood, and risk
management plans along with policies for HLF and
enterprise blockchain, in general, must have policies
that are by regulating authorities.
3. Significance of the Findings
IBM simultaneously is a leading developer of
enterprise-grade blockchains and quantum computers.
In 2018, Harriet Green, chairman, and CEO of IBM
Asia Pacific, stated: “IBM sees quantum computing
going mainstream within five years” [8]. Currently,
there is not a specific strategy to mitigate the threat
of quantum computers, and as such, all known data
security and privacy laws will be violated. There
are significant regulatory responsibilities of its
participants that own, create, modify, store, or
transmit regulated data and information. Enterprisegrade blockchains must enact holistic approaches to
cybersecurity across applications, infrastructure, and
processes. Cybersecurity must defend against attacks,
but also maintain control of data content. This
research illuminates the need for new policies to be
developed for those entities whose data is regulated.
To the author’s knowledge, no cybersecurity policy
addresses regulated data on enterprise blockchains.
A cybersecurity policy outlines the assets that need
protection and the threats to those assets and the rules
and controls for protecting them. The policy should
inform all approved users of their responsibilities
to protect information about those assets. Policy
management, reporting, and administration will be
essential for organisations inputting their data on
blockchains. Participants will need to be able to report
enterprise-wide on everything users have done with
regulated content to satisfy compliance requirements.
HLF’s PKI system of trust is broken with the arrival
of large-scale quantum computing, and all PII and PHI
are at risk with no known plans to mitigate. HIPAA,
GDPR, FINRA, and all known data and privacy laws
that will be violated. The author has independently
16
jbba
the
tested, verified, and validated qTESLA’s much improved
Second Round Submission to NIST Post-Quantum
Cryptography Standardization Process and has
proposed a hybrid quantum-resistant PKI system for
replacement in HLF. The test result yields smaller key
sizes; however; given today’s standards and applications
in use only qTESLA’s public key is recommended for
use in a hybrid PKI solution. qTESLA’s public-key is an
adequate replacement for the current ECDSA publickey. In HLF’s PKI, it is the public key that is used
most often and qTESLA’s second submission offers an
acceptable size that could reinforce a mix of the most
practical quantum-resistant digital signature scheme
with current ECDSA algorithms.
Given what is at risk for the blockchain implementors
and its users, reasonable measures must be taken
to mitigate the threat of data privacy and security.
To safeguard data on a blockchain platform, the
participants must be able to control who has access to
their data and under what circumstances. Blockchain
networks must be able to provide reasonable measures
and safeguards that adhere to privacy regulations such
as HIPAA, FINRA, and GDPR.
4. HLF and PKI and Membership Services
Technology
IBM offers Cryptographic PKI Services that allow
users to establish a PKI infrastructure and serve
as a certificate authority for internal and external
users, issuing and administering digital certificates.
It supports the delivery of certificates through the
Secure Sockets Layer (SSL) for use with applications
that are accessed from a web browser or web server.
It includes delivery of certificates that support the
Internet Protocol Security standard (IPSEC) for use
with VPN applications and delivery of certificates that
support Secure Multipurpose Internet Mail Extensions
(S/MIME), for use with email applications. All these
functions are essential but critically vulnerable.
Fabric is a private, blockchain technology that uses
smart contracts, and participants or members manage
its transactions. The members of the network enroll
through a “trusted” Membership Service Provider
(MSP) [9]. The blockchain is advertised as an
implementation of distributed ledger technology
(DLT) that delivers enterprise-ready network security,
scalability, confidentiality, and performance, in modular
blockchain architecture.
The MSP issues, cryptography, protocols, encryption,
signature keys and issues and validates certificates and
user authentication to clients and peers. HLF’s PKI
consists of Digital Certificates, Public and Private
Keys, and Certificate Authorities (CA) which issues
digital certificates to parties, who then use them to
authenticate messages. A CA’s Certificate Revocation
The JBBA | Volume 2 | Issue 2 | October 2019
List (CRL) is a reference for the certificates that are no
longer valid. PKI is used to generate certificates which
are tied to organizations, network components, and endusers or client applications. The MSP dispenses X.509
certificates that can be used to identify components
as belonging to an organization. Certificates issued by
CAs can also be used to sign transactions to indicate
that an organization endorses the transaction result and
is a necessary precondition of it being accepted onto
the ledger. These X.509 certificates are used in client
application transaction proposals and smart contract
transaction responses to digitally sign transactions. Its
digital certificate is compliant with the X.509 standard
and holds the attributes relating to the holder of the
certificate. The holder’s public key is distributed within
the certificate, and the private signing key is not.
The public-keys and private-keys are made available and
act as an authentication “anchor,” and the private keys
are used to produce digital signatures. Recipients of
digitally signed messages can validate and authenticate
the received message by checking that the attached
signature is valid with the use of the public key.
Digital identities are cryptographically validated digital
certificates that comply with X.509 standard and are
issued by a Certificate Authority (CA). HLF uses a list
of self-signed (X.509) certificates to constitute the root
of trust and a list of self-signed (X.509) certificates to
form the root of trust. A CA dispenses certificates that
are digitally signed by the CA and bind together the
actor with the actor’s public key. The above services
are critical to the operation of a secure enterprise
blockchain, and there must be plans and strategies in
place that provide reasonable measures to adhere to
regulatory policies.
5. Post-Quantum Computing Impact on HLF PKI
PQC algorithms must provide security against both
classical and quantum computing attacks. Their
performance is measured on classical computers and
considerations are made for the potential of “dropin replacements,” which infers compatibility and
interoperability with existing systems. Also, essential
requirements must include resistance to side-channel
attacks and misuse.
Cryptography in HLF is used in many applications
where secure communication is needed. The primary
use and role are signature generation, verification, and
authentication where algorithms are used to establish
confidentiality, integrity, and authenticity of messages
sent during communication. Public-key cryptography
is used where each participant has a private key and a
public key. In a public-key signature cryptosystem, the
signer has a private signing key that can be used to sign
messages and must keep this key secure. The public
key, which is visible to anyone, can be used to verify
that the signature is authentic and, if the signature
17
jbba
the
scheme is secure, then repudiation is achieved and only
the signer could have generated the signature. PKIs
are used to bind identities to the public keys, where
Certificate Authorities (CAs) play an essential role.
A CA is a commonly trusted party that uses digital
signature algorithms to author certificates consist of a
public key and information of its owner. The security
of public-key cryptography and ultimately, the private
key is based on cryptography that can no longer be
considered safe because of the emerging quantum
computing threat. HLF relies on a PKI, which is
based upon Elliptic Curve Cryptography (ECC),
and it is critically vulnerable to quantum computing
[10]. Specifically, the cryptography that secures web
browsers (TLS), certificates, software updates, virtual
private networks (IPsec), secure email (S/MIME)
and many other applications are no longer safe in
the PQC era [11]. Reasonable blockchain enterprise
cybersecurity measures require extensive planning and
testing for transition and migration to post-quantum
resistant cryptography.
It is unlikely that the current PQC algorithms under
review will function “as is” and will require modifications
such as hybrid quantum resistant-classical PKI systems.
Hybrid systems will likely be the way forward in the
near term, given the uncertainties and complexities
of the current crop of PQC algorithms. Current
cryptographic libraries will provide support for postquantum digital signature algorithms in PKI but will
require some modifications and testing in large-scale
scenarios.
In this paper, the author investigates the use of hybrid
digital signature schemes, specifically qTESLA. Much
testing needs to be done in real-world scenarios
involving digital signatures and PKI. Protecting
against quantum attacks will require changes that
designers and implementers will have to accommodate.
Cryptographic primitives may need to be a replaced,
and protocol-level modifications may be necessary to
provide new primitives. It is a complex and lengthy
undertaking to migrate to a new quantum-resistant PKI.
Other issues such as constrained devices, compatibility,
performance characteristics, and Internet of Things
(IoT) must also be considered. Currently, HLF uses
the Elliptic Curve Digital Signature Algorithm, which
is used for many functions such as digital signatures
and TLS protocol handshakes.
6. Elliptic Curve Cryptography in HLF
Elliptic curve cryptography is a class of publickey cryptosystem which assumes that finding the
elliptic curve discrete algorithm is not possible in a
“reasonable” amount of time. Public key cryptography
does not require any shared secret between the
communicating parties. The security of elliptic curve
or asymmetric cryptographic schemes relies on the
believed hardness of solving “hard problems,” such
as integer factorization and the computation of
discrete logarithms in finite fields or groups of points
on an elliptic curve. The ECDSA algorithm relies
critically on generating a random private key used
for signing messages and a corresponding public key
used for checking the signature. The bit security of
this algorithm depends on the ability to compute a
point multiplication and the inability to calculate the
multiplicand given the original and product points.
Decades ago, these were “hard problems,” due to
several factors such as the current state of computing
power, and the time it would take for a classical
computer to solve these problems. Other factors come
into play, such as the length of cryptanalysis and the
lack of known techniques that ensured the problems
remained hard. However, the technology of computing
power, cryptanalysis, and side-channel analysis always
threaten the existing cryptographic standards given
enough time. It can be noted that many real-world
cryptographic vulnerabilities do not stem from solely a
weakness in the underlying algorithms, but often from
implementation flaws such as side-channel attacks,
errors in software or code design flaws. An example is
the vulnerabilities ECDSA signature implementation,
is the property of weak randomness used during
signature generation, which can compromise the longterm signing key.
The HLF CA provides features such as, registration
of identities, or connects to Lightweight Directory
Access Protocol (LDAP) as the user registry, issuance
of Enrollment Certificates (ECerts), certificate renewal
and revocation. HLF’s ECDSA offers the following
key size options:
Table 1. Algorithms used to generate X.509 certificates
and keys are not secure [12]
are not secure [12]
Size
256
384
521
ASN1 OID
prime256v1
secp384r1
secp521r1
Signature Algorithm
ecdsa-with-SHA256
ecdsa-with-SHA384
ecdsa-with-SHA512
The approved security strengths for U.S. federal
applications are 128, 192, and 256 bits. Note that a
security strength of fewer than 128 bits is no longer
approved because quantum algorithms reduce the bit
security to 64 bits (see table 2). NIST Special Publication
800-57 Part 1 Revision 4: Recommended for Key
Management, as shown in Table 2 [13]. Table 2 shows
that Rivest, Shamir, and Adleman (RSA) and ECC
based PKI have zero bits of security and AES requires
larger keys. This table illustrates the vulnerability and
single point failure, of the fully trusted CA and X509
standard based on ECC. The quantum computing
threat collapses the RSA, ECC, and HLF’s PKI.
The JBBA | Volume 2 | Issue 2 | October 2019
18
Qua
jbba
the
Table 2. Comparison of conventional and quantum security levels
of typical ciphers [14]
of typical ciphers [14]
Effective Key Strength / Security Level
Algorithm Key Length
RSA-1024
1024 bits
Conventional
Computing
80 bits
Quantum
Computing
0 bits
RSA-2048
2048 bits
112 bits
0 bits
ECC-256
256 bits
128 bits
0 bits
ECC-384
384 bits
256 bits
0 bits
AES-128
128 bits
128 bits
64 bits
AES-256
256 bits
256 bits
128 bits
7. Evaluation of qTESLA’s Second Round
Submission to NIST
The National Institute of Standards and Technology
(NIST) is in the process of selecting one or more
public-key cryptographic algorithms through a
public competition-like process. The new publickey cryptography standards will specify one or more
additional digital signature, public-key encryption
algorithms. It is intended that these algorithms will
be capable of protecting sensitive information well
into the foreseeable future, including after the advent
of quantum computers. The author tracked with
NIST in identifying three broad aspects of evaluation
criteria that would be used to compare candidate
algorithms throughout the NIST PQC Standardization
Process. The three elements are 1) security, 2) cost and
performance, and 3) algorithm and implementation
characteristics. Security is the most crucial factor when
evaluating candidate post-quantum algorithms. Cost as
the second-most important criterion when assessing
candidate algorithms. In this case, cost includes
computational efficiency and memory requirements.
After security, the performance was the next most
important criterion in selecting the second-round
candidates [3].
qTESLA is a lattice-based signature scheme which
uses the assumption that RLWE distributions are
indistinguishable from random. The public key in
qTESLA is, roughly speaking, a sample of an RLWE
distribution. The signer keeps secret information
about this sample and uses that information along
with a hash function to produce signatures. Signature
verification involves some simple arithmetic within the
chosen ring, and then the recomputation of a hash
function. qTESLA has reasonably good performance
parameters that are comparable to the other latticebased signature schemes. The submitters of qTESLA
have claimed a tight security proof for the schemes in
the quantum random oracle model. It was noticed that
a bug in the security proof requires an adjustment of
the parameters (which reduces the efficiency of the
The JBBA | Volume 2 | Issue 2 | October 2019
scheme). Furthermore, the security argument assumes
(among other things) conjecture about the distribution
of random elements in the ring. Considering that the
conjecture does do not seem to fit the form of a typical
security assumption, and more analysis will need to be
conducted in the second round.
This section, tests, evaluates and analyzes qTESLA’s
second-round submission modifications in the latticebased digital signature scheme category to NIST’s postquantum standardization project. This second-round
submission is based on the hardness of the decisional
Ring Learning With Errors (R- LWE) problem.
qTESLA utilizes two approaches for parameter
generation that includes heuristic and provably- secure.
The heuristic approach is optimized for efficiency and
key size, and the provably- secure is targeted to highly
sensitive or classified transactions. A new feature added
in the second-round submission is a key compression
technique that produces a noticeable reduction in the
public key size. The vendor refers to this technique
as “public key splitting,” and is significant because
it is the public key that is used most often in typical
transactions. qTESLA has submitted twelve parameter
sets targeting various security levels. However, this
work focuses on submissions that include public-key
reduction and the most efficient submissions as the
most practical hybrid (classical and quantum-resistant)
PKI near-term algorithm solution [14].
8. Basic signature scheme
Informal descriptions of the algorithms that give
rise to the signature scheme qTESLA are shown in
Algorithms 1, 2, and 3. These algorithms require two
basic terms, namely, B-short and well-rounded, which
are defined below.
Let q, LE , LS , E, S, B, and d be system parameters
that denote the modulus, the bound constant for
error polynomials, the bound constant for the secret
polynomial, two rejection bounds used during signing
and verification that are related to LE and LS , the
bound for the random polynomial at signing, and the
rounding value, respectively. An integer polynomial y
is B-short if each coefficient is at most B in absolute
value. An integer polynomial w well-rounded if w is
(lq/2J − E)-short and [w]L is (2 d−1 − E)-short.
In Algorithms 1-3, the hash oracle H(•) maps to H,
where H denotes the set of polynomials c ∈ R with
coefficients in {−1, 0, 1} with exactly h nonzero entries.
Algorithm 2 is described as a non-deterministic
algorithm. This property implies that different
randomness is required for each signature. This
design feature is proposed as added to prevent some
implementation attacks and protect against some fault
attacks [13].
19
jbba
the
Algorithm 1 Informal description of the key generation
Require: Ensure: Secret key sk = (s, e1, ..., ek, a1, ..., ak), and public
key pk = (a1, ..., ak, t1, ..., tk)
•
•
•
•
•
a1, ..., ak ← Rq ring elements.
Choose s ∈ R with entries from Dσ. Repeat
step if the h largest entries of s sum to at
least LS .
For i = 1, ..., k: Choose ei ∈ R with entries
from Dσ. Repeat step at iteration i if the h
largest entries of ei sum to at least LE.
For i = 1, ..., k: Compute ti ← ai s + ei ∈ Rq.
Return sk = (s, e1, ..., ek, a1, ..., ak) and pk =
(a1, ..., ak, t1, ..., tk)
Algorithm 2 Informal description of the signature
generation
Require: Message m, secret key sk = (s, e1, ..., ek, a1, ...,
a k)
Ensure: Signature (z, c)
•
•
•
•
•
•
Choose y uniformly at random among
B-short polynomials in Rq.
c ← H([a1y]M , ..., [aky]M , m).
Compute z ← y + sc.
If z is not (B − S)-short then retry at step 1.
For i = 1, ..., k: If ai y − ei c is not wellrounded then retry at step 1.
Return (z, c).
Algorithm 3 Informal description of the signature
verification
Require: Message m, public key pk = (a1, ..., ak, t1, ...,
tk), and signature (z, c)
Ensure: “accept” or “reject” signature
•
•
•
•
If z is not (B − S)-short then return reject.
For i = 1, ..., k: Compute wi ← aiz − tic ∈
Rq.
If c /= H([w1]M , ..., [wk]M , m) then return
reject.
Return accept.
the implementation attacks as research shows the
vulnerabilities of lattice-based signature schemes such
as qTESLA [16]. The second and third new feature is the
AVX2-optimized implementations for the parameter
sets qTESLA-I, qTESLA-III, and qTESLA-V, and
their variants with smaller public keys, called “public
key splitting,” for qTESLA-I-s, qTESLA-III-s, and
qTESLA-V-s respectively. qTESLA’s AVX2-optimized
implementations submission included an Intel
Advanced Vector Extensions 2 (AVX2) submission
which significantly improved performance. The
author performed experiments with qTESLA’s AVX2
optimized implementation, and the results are included
in this paper. The public key splitting submission
is a variant that addresses public key size, which is
significant because the public key size is regarded as
more important than the secret key size because the
former needs to be transmitted more frequently [14].
10. Mitigation of implementation attacks
Side-channel cryptanalysis considers attackers trying
to take advantage of the physical interactions of
cryptographic devices to achieve recovery of the
secret key. In some cases, computational faults are
intentionally inserted to obtain faulty values for
the key recovery. Fault injections or attacks are also
used to obtain information leakage under the faulty
environment. These implementations-specific attacks
are more efficient than the best-known cryptanalytic
attacks. They are therefore generally more powerful than
classical cryptanalysis and are a serious class of attacks
that must be addressed. These attacks exploit timing or
power consumption, electromagnetic emanation, that
is correlated to some secret information during the
execution of a cryptographic scheme and protection
against this attack is a minimum-security requirement
for standardized cryptographic implementation.
qTESLA attempts to address the exploit timing
leakage, power consumption, electromagnetic
emanation, and cache attacks by adding constanttime execution to secure against side-channel analysis.
qTESLA ‘s approach indicates that it is in every signing
operation, it injects “fresh randomness,” that will make
it resilient to a catastrophic failure of the Random
Number Generator (RNG) protecting against fault
analysis attacks [14]. The verification and validity of
the previous statements are not in the scope of this
paper and will most likely require more independent
tests and analysis.
9. New features
qTESLA utilizes two approaches for parameter
generation, the first approach, referred to as “heuristic
qTESLA,” follows a heuristic parameter generation
and the second approach, referred to as “provably
secure qTESLA,” follows a provably secure parameter
generation according to existing security reductions.
New in this submission is mitigation steps to address
The JBBA | Volume 2 | Issue 2 | October 2019
11. Performance of
algorithms analysis
second-round
qTESLA
To evaluate the performance of the provided
implementations written in portable C, the author ran
benchmarking suite on one machine powered by an
Intel® Core™ i7-6500 CPU @ 2.50 GHz x 4 (Skylake)
processor, 16 GB of RAM, 500 GB hard drive,
20
jbba
the
GNOME:3.28.2, running Ubuntu 18.04.2 LTS. For
compilation, GCC version 7.3.0 was used in all tests.
The vendor proposed twelve parameter sets which
were derived according to two approaches (i) following
a “heuristic” parameter generation, and (ii) following
a “provably-secure” parameter generation according
to a security reduction. The proposed parameter sets
are displayed in Table 3, together with their targeted
security category.
The results for the optimized implementations are
summarized in Tables 4, and 5, respectively. The
results for AVX2 implementations are given in Tables
6, and 7, respectively. Additionally, the reference
implementations are summarized in Tables 8, and 9,
respectively. Results for the median and average (in
sign
verify
qTESLA-I
903.2
(940.9)
206.4
(268.2)
55.1
(55.8)
total
(sign +
verify)
261.5
(324)
qTesla-I-s
928.5
(952.4)
214.9
(276.6)
54.8
(55.9)
269.7
(332.2)
qTESLA-III
2373.5
(2677.0)
273.5
(343.5)
110.4
(111.3)
383.9
(454.8)
qTESLA-IIIs
2366.8
(2713.6)
291.4
(374.2)
110.0
(112.4)
401.4
(486.6)
qTESLA-V
12577.2
(14472.8)
734.1
(951.3)
254.9
(256.0)
989.0
(1207.3)
qTESLA-V-s
12593.2
815.3
256.1
1071.4
Table 7. Second Round AVX2 Implementation Key Sizes in7
Bytes
Secret Key
Signature
Security category
qTESLA-I
1504
1216
1376
NIST’s category 1
qTesla-I-s
480
2240
1568
-
NIST’s category 2
qTESLA-III
qTESLA-III-s
3104
1056
2368
4416
2848
3232
qTESLA-p-III
NIST’s category 3
qTESLA-V
qTesla-V-s
6432
1952
4672
6592
5920
5216
-
NIST’s category 5
-
NIST’s category 5
qTESLA-p-I
Table 4. Second Round Optimized Implementation tests
for 5000 iterations.
iterations.
Scheme
keygen
Public Key
Provably secure
qTESLA-I,
qTESLA-I-s
qTESLA-II,
qTESLA-II-s
qTESLA-III,
qTESLA-III-s
qTESLA-V,
qTESLA-V-s
qTESLA-V-size,
qTESLA-V-size-s
Scheme
Scheme
Table 3. Parameter sets and their targeted security [14]
Heuristic
Table 6. Second Round AVX2 Implementation
keygen
sign
verify
total
(sign +
verify)
qTESLAII
4410.7
(4963.6)
931.7
(1226.1)
232.8
(236.5)
1164.5
(1462.6)
qTESLAII-s
4004.0
(4818.7)
981.5
(1281.4)
232.7
(235.1)
1214.2
(1516.5)
qTESLAV-size
17177.0
(20416.5)
2161.4
(2812.1)
511.6
(514.2)
2673.0
(3326.3)
qTesla-Vsize-s
17201.1
(20340.2)
2341.4
(2972.4)
516.8
(523.1)
2858.2
(3495.5)
Table 5. Second Round Optimized Implementation Key Sizes in
Bytes
ey
ey
e
Table 8. Second Round Reference Implementation
Scheme
keygen
sign
verify
total
(sign + verify)
qTESLA-I
920.3
(971.5)
314.4
(425.6)
71.5
(72.6)
385.9
(498.2)
qTESLA-Is
926.4
(968.5)
334.2
(438.1)
73.3
(74.2)
481.7
(512.3)
qTESLA-pI
4130.2
(4316.4)
1990.4
(2605.6)
561.2
(567.9)
2551.6
(3173.5)
qTESLA-II
4466.0
(5047.9)
1536.6
(2027.2)
372.3
(375.7)
1908.9
(2402.9)
qTESLAII-s
4452.1
(5047.0)
1647.3
(2213.9)
385.5
(386.5)
2032.8
(2600.4)
qTESLAIII
2395.5
(2669.8)
433.9
(580.0)
143.0
(145.2)
576.9
(725.2)
qTESLAIII-s
2410.5
(2735.2)
471.9
(610.8)
150.9
(153.6)
622.8
(764.4)
6932.0
(8776.4)
Scheme
ey
Public Key
ey
Secret Key
Signaturee
qTESLA-pIII
21043.7
(21569.7)
5414.6
(7247.6)
qTESLA-II
qTESLA-II-s
2336
800
931.7
3136
232.8
2432
qTESLA-V
qTESLA-V-size
5024
3520
4640
12224.6
(14221.3)
1349.6
(1775.1)
1517.4
(1529.
4)
325.9
(329.1)
qTesla-V-size-s
1952
6592
5216
qTESLAV-s
12644.5
(14433.8)
1439.4
(1856.3)
335.4
(336.8)
1774.8
(2193.1)
parenthesis) are rounded to the nearest 102 cycles.
Signing is performed on a message of 59 bytes.
qTESLAV-size
17357.1
(20838.9)
3653.8
(4769.2)
825.2
(830.5)
4479.0
5599.7
This work is a follow-on to qTESLA’s NIST first-round
submission, and the evaluation focuses on the “new”
and improved features submitted in its second-round
NIST submission. This second-round submission
includes an expanded category of parameters in which
the author examined the most practical based on
qTESLAV-size-s
17859.4
(21204.1)
3824.2
(5044.1)
851.3
(847.3)
4675.5
(5891.4)
1675.5
(2104.2)
performance
significant
(in parenthesis) improvements.
are rounded to the The
nearestmost
102 cycles.
Signing
enhancements noted, is in the speed of key generation
and the size of the public keys. Techniques, such as
the AVX2 and Public key splitting, yields a dramatic
--
The JBBA | Volume 2 | Issue 2 | October 2019
s
21
jbba
the
improvement over the previous submissions. The public
key splitting offers acceptable sizes for various NIST
security category levels, While, these implementations
are not provably secure as defined by NIST, meaning
the algorithms may not be approved for top secret
information and operations, however; they may prove
useful for less critical data and processes.
Table 9: Second Round Reference Implementation Key
Sizes in Bytes.
Scheme
Public Key
Secret Key
Signature
qTESLA-I
1504
1216
1376
qTESLA-I-s
480
2240
1568
qTESLA-p-I
14880
5184
2592
qTESLA-II
2336
1600
2144
qTESLA-II-s
800
3136
2432
qTESLA-III
3104
2368
2848
qTESLA-III-s
1056
4416
3232
qTESLA-V
6432
4672
5920
qTESLA-V-s
2336
8768
6688
qTESLA-V-size
5024
3520
4640
qTesla-V-size-s
1952
6592
5216
12. Optimized implementations
All comparisons are made about qTESLA’s first-round
NIST submission where possible, due to the fact there
are new submissions and comparisons cannot be made.
The optimized implementation for key sizes shows
qTESLA-II vs. qTESLA-II-s shows 78.5% public-key
reduction; however; there is an increase in the secret key
and signature size of 236.5 % and 944.6 % respectively.
Submissions for qTESLA-V-size vs. qTESLA-V-size-s
shows 61.1 % public-key reduction, while there is an
increase in the secret key and signature size of 87.2 %
and 12.4 % respectively. (See Table 5).
there is an increase in the secret key and signature size
of 86.5 % and 41.0 % respectively, See Table 7.
12.2. Reference implementation
The last category examined is Reference
implementation, which has 12 parameters. Since many
of these parameters are new, direct comparison to the
previous submission cannot be made. However; the
author notes overall, there is a significant reduction
in key generation, signing, and verification times
compared to the first-round submission. The following
is a comparison of the first-round submission to the
second-round submission. For example, for key
generation, signing, and verification CPU cycles
qTESLA-I reduced key generation cycle time by
26.4 % but increased 5.7 % signing, decreased 12.1
% verification respectively. qTESLA-p-I showed key
generation cycle reduction of 23.0 %, but the 152 %
increase in signing, an increase of 34.1 % verification.
qTESLA-p-III showed a decrease of 16.3 % key
generation, but increase signing 71.6 %, and a reduction
of 28.3 % verification time (See Table 8 and [2]). The
test results of the Reference implementation key sizes
in bytes are in Table 9. The following observations
can be made from a comparison of the first-round
submission with the second-round submission; The
most dramatic improvement comes with the public
key splitting function, while test results show there is a
corresponding increase in secret key size and signature.
For example, for the public key of qTESLA-I-s vs.
qTESLA-I decreased by 68.0%, but the secret key
increased by 84.2 %, and the signature increased by 13.9
%. qTESLA-III-s vs. qTESLA-III show a reduction of
65.9 %, but an increase in the secret key size of 86.4 %,
and an increase in the signature size by 13.4 %. Please
see Table 9 for further comparisons.
13. Recommendations for Blockchain Implementors
12.1. AVX2 implementation
The AVX2 implementation for key generation, signing,
and verification is shown in Table 6 and is compared
to the new AVX2 and public-key reduction. The tests
show that there is a slight increase in key generation
time, signature and verification time for all categories
of submission when using the public-key reduction
techniques, however; these improvements are dramatic
compared to the respective timing in all categories in
qTESLA’s first submission [2]. (See Table 6). The
AVX2 implementation for key sizes shows qTESLA-I
vs. qTESLA-I-s shows 68.1 % public-key reduction;
however; there is an increase in the secret key and
signature size of 84.2 % and 13.9 % respectively.
Submissions for qTESLA-III vs. qTESLA-III-s shows
65.9 % public-key reduction, while there is an increase
in the secret key and signature size of 86.5 % and 13.4
% respectively. Finally, in this category, qTESLA-V vs.
qTESLA-V-s shows 69.6 % public-key reduction, while
The JBBA | Volume 2 | Issue 2 | October 2019
HLF implementors should develop and provide a
strategy or roadmap for maintaining the confidentiality,
integrity, and availability of private keys and stringent
cybersecurity controls to combat the quantum
computing threat. Also, implementers should review
their current cryptographic standards to make sure
they are up to date, and that infrastructure and
support exist to update when new NIST standards
become available rapidly. Immediate work should
begin to test and benchmark the most promising PQC
candidates that could be integrated into its blockchain
with interoperability and compatibility in mind. The
X.509v3 standard allows for algorithm flexibility in
that the Object Identifier (OID) defines the formats
of public keys. Adding a new cipher OID is needed to
extend X.509, but what is also required is for software
will be able to comprehend and process the new OID.
Currently, there are no known CAs issuing certificates
for quantum-safe public keys exist, and no CAs is
22
jbba
the
signing their certificates with a quantum-safe signature
algorithm.
for an ecosystem-level solution to protect
organizations and maintain contractual
obligations.
Determine how to implement best the
GDPR principle of “the right to be
forgotten.”
What is the ability to detect, correct
fraudulent, malicious, or erroneous records?
It is unclear which organization will be
considered as the data controller and
processor within the Fabric and enterprise
blockchains, especially when they cross
international borders.
Create new quantum-proof policies,
methods, and procedures aligned to use
cases/requirements. Update asset inventory
with newly implemented cryptographic
details.
Strong blockchain network security requires the roles
and responsibilities of each type of participant to
be clearly defined and enforced following regulatory
guidelines. It is essential to qualify, quantify, and
document cybersecurity risks posed by each type
of participant. It is also essential to anticipate and
understand the security consequences of participants
leaving and entering the network over time. Blockchain
developers should anticipate and understand these
threats resulting before committing regulated data to
the blockchain. There should be plans for penetration
testing that are similar to traditional networks using
various attack scenarios and vectors, document the
development process and obtain independent audits of
the design and development process.
•
Therefore, there is an urgent requirement to develop
and deploy plans to accommodate the most practical
hybrid PQC algorithms that are working towards
global standardization. The successful transition and
migration to PQC will require significant time and effort
given the complexities involved. Further, researchers
should examine hybrid solutions where both classical
cryptography algorithms and PQC algorithms working
together to mitigate the uncertainties in the pace
and development of quantum computers and the
reliability of candidate PQC under the global standards
community.
Healthcare, GDPR, and financial entities must not
think that there are no risks associated with blockchain
enterprise blockchain networks and must ask for
documented risk management strategies to protect
regulated data. As the HLF blockchain ecosystem
becomes more diverse and grows in popularity, vendors,
users, and implementors must be aware of possible
cyber-attack. While blockchains offer unique structures
and provide cybersecurity capabilities that are not
present in today’s networks, reasonable measures must
be taken. The cybersecurity risk must be evaluated,
documented, and its implications considered when
regulated, businesses policymakers, and institutions
commit protected data to any enterprise blockchain.
13.1. Recommendations for Healthcare and GDPR
Covered Entities
•
•
•
14. Conclusions and Future Work
HLF and other permissioned blockchains present
unique opportunities and vulnerabilities in managing
cybersecurity risks. As the healthcare industry, financial
services, and GDPR covered industry begin to
experiment with and commit to pilots, these entities
need to understand that the risks are appropriately
identified, and this is a risk management plan. This
risk management plan is required for regulated data,
and there must be one for enterprise blockchains.
Therefore, beyond the hype of any new technology,
a thorough cybersecurity program remains vital, and
all parties need to conduct due diligence to protecting
the network and participating organizations from cyber
threats. Also, the participation of multiple entities, each
with their on-ramps into the enterprise blockchain, is a
potential source of vulnerability.
Ask blockchain vendors about their quantum-safe
features to protect data that is under regulatory
guidance
•
Query software-as-a-service or thirdparty platform providers about their
embedded cryptographic methods and plans
The JBBA | Volume 2 | Issue 2 | October 2019
This work has shown that HLF, enterprise blockchains,
and current global PKI that relies on the PKI X.509
standard to ensure secure communication between
various network participants are utterly vulnerable to
the quantum computing threat. Falsified certificates
destroy the trust, integrity, confidentiality, and nonrepudiation in the entire blockchain and can have
enormous consequences if measurements are not
taken. It has been shown that quantum computers
break ECC on which PKI depends and therefore
exposes its implementers and users to potentially
massive fines for non-compliance and security
incidents with GDPR, FINRA and HIPAA laws.
Enterprise Blockchains such as HLF are being adopted
in many industries that have regulatory controls over
the data. For example; GDPR regulates European
Union citizens’ data with the potential of massive
fines irrespective of the location or headquarters of
the blockchain implementation location. Financial and
PII data privacy and information is becoming more
heavily regulated, especially on Wall Street and in the
state of New York and California. In the United States,
healthcare data privacy is a significant issue with the
23
jbba
the
increase in cyber-attacks, and the resulting lawsuits,
fines, and penalties levied on violators.
The author argues that blockchain technology has the
potential to address the documented issues of legacy
health and financial information technology systems,
such as interoperability, data access, speed, and privacy
and the ability to adapt to changing programs. However;
out-of-date cryptographic standards will be broken
and will not forestall any adversaries from breaking
their encryption and gaining access to highly regulated
data and information. Development and deployment
plans need to be developed to accommodate the most
practical hybrid PQC algorithms that are working
towards global standardization. Also, blockchain
cybersecurity policy is required to govern acceptable
use and should include standards, procedures, and
guidelines.
Cybersecurity should begin with an assessment that
includes current security policies, identification of
objectives, review of requirements, and determination
of existing vulnerabilities. It is imperative to begin
the development of “Policy Recommendations for
Enterprise Blockchains” because covered entities
must know that placing their data on permissioned
blockchains does not and cannot negate risks and
obligations. All must understand the risks before
committing regulated data, because it is required, and
it is also prudent in protecting PHI, PII, GDPR, and
FINRA regulated data and information. An evidencebased approach is needed to mitigate and adhere to
cybersecurity regulation. All aspects must be considered
such as geographic boundaries, jurisdictions and a
thorough understanding of the impact of widespread
governance of global regulators
As cyber threats to the HIPAA and GDPR and covered
financial entities continue to grow in dedication and
sophistication permissioned blockchains can contribute
to add “new and advanced cybersecurity techniques”
and can be a valuable tool in mitigating those threats if
the risks are understood and mitigated. Permissioned
blockchains offer significant cybersecurity capabilities,
share some of the same cyber risks that affect other IT
systems, and have unique characteristics, all of which
merit further evaluation by regulators and industry.
The author encourages new conversations about the
cybersecurity benefits of blockchain systems and ways
to promote appropriate government policies.
Finally, this research does not indicate any of NIST
Second Round candidate algorithms will be a simple
“drop-in replacement,” and it may require additional
NIST rounds and years of follow-on research, analysis
and testing for a suitable “drop-in replacement,” can
be identified or developed. Therefore, the author
believes that qTESLA offers a possible near-term
“Hybrid Quantum Resistant-Classical Public Key
Infrastructure,” a solution with a significant reduction
The JBBA | Volume 2 | Issue 2 | October 2019
in its public key size. As discussed, it is the public
key that is exposed and used the most in today’s
PKI systems, and it is possible to modify the X.509
certificate standard to accommodate this new PQC
algorithm that would only provide the public key that
would be much more resistant to implementation
and quantum computing attacks. Additional work
and testing are needed in large scale real-world
scenarios to ensure there are no significant issues
with incorporating PQC PKI X.509 certificates on
an industrial scale. Potential problems that need to
be examined are latency, overhead, and the ability for
software, hardware, and other constrained devices to
interoperate such as, smartphones, smart cards, and
IoT. Regardless of the estimated time of arrival of
large-scale quantum computers, cybersecurity should
be a primary concern to enterprises and healthcare
organizations because they cannot afford to have their
private communications and data decrypted even if it
is ten years away.
References:
[1] J. Emond, "IBM Newsroom," 24 January 2019. [Online].
Available: https://newsroom.ibm.com/2019-01-24-AetnaAnthem-Health-Care-Service-Corporation-PNC-Bank-andIBM-announce-collaboration-to-establish-blockchain-basedecosystem-for-the-healthcare-industry. [Accessed 16 May 2019].
[2] R. Campbell, "Evaluation of Post-Quantum Distributed
Ledger Cryptography," The Journal of the British Blockchain
Association, vol. 2, no. 1, pp. 17-24, 2019.
[3] NIST, "Second PQC Standardization Conference," 22
August 2019. [Online]. Available: https://www.nist.gov/
news-events/events/2019/08/second-pqc-standardizationconference. [Accessed 16 May 2019].
[4] NIST, "Second PQC Standardization Conference," 22
August 2019. [Online]. Available: https://www.nist.gov/
news-events/events/2019/08/second-pqc-standardizationconference. [Accessed 16 May 2019].
[5] E. Commission, "2018 reform of EU data protection
rules,"
[Online].
Available:
https://ec.europa.eu/
commission/priorities/justice-and-fundamental-rights/dataprotection/2018-reform-eu-data-protection-rules_en. [Accessed
16 May 2019].
[6] G. EU.org, "Fines and Penalties," [Online]. Available:
https://www.gdpreu.org/compliance/fines-and-penalties/.
[Accessed 16 May 2019]
[7] D. o. H. a. H. S. Office for Civil Rights, "Federal
Registry," [Online]. Available: https://www.federalregister.
gov/documents/2013/01/25/2013-01073/modifications-tothe-hipaa-privacy-security-enforcement-and-breach-notificationrules-under-the#h-95. [Accessed 16 May 2019].
[8] "CNBC interview: Harriet Green, Chairman and CEO
24
jbba
the
of IBM Asia Pacific," , . [Online]. Available: https://www.
cnbc.com/2018/03/30/ibm-sees-quantum-computing-goingmainstream-within-five-years.html. [Accessed 11 7 2019].
[9] A. M. V. V. K., Z. M. Jøsang, "The Impact of Quantum
Computing on Present Cryptography," Arxiv, 31 March 2018.
[Online]. Available: https://arxiv.org/pdf/1804.00200.
[Accessed 16 May 2019].
[10] H. U. M. M. S. D. Bindel Nina, "Transitioning to a
Quantum-Resistant Public Key Infrastructure," PQCryptoBHMS17, 2017. [Online]. Available: https://s3.amazonaws.
com/files.douglas.stebila.ca/files/research/papers/PQCryptoBHMS17.pdf. [Accessed 16 May 2019].
[11] Hyperledger, "Fabric CA User’s Guide," [Online].
Available:
https://hyperledger-fabric-ca.readthedocs.io/en/
latest/users-guide.html#table-of-contents. [Accessed 11 6
2019].
[12] "NIST Special Publications - NIST Computer Security
..., ”,. [Online]. Available: http://csrc.nist.gov/publications/
PubsSPs.html. [Accessed 7 1 2019].
[13] L.. Chen, S. P. Jordan, Y.-K.. Liu, D.. Moody, R. C.
Peralta, R. A. Perlner and D. C. Smith-Tone, "Report on
Post-Quantum Cryptography | NIST," , 2016. [Online].
Available: https://nist.gov/publications/report-post-quantumcryptography. [Accessed 30 12 2018].
[14] N. Bindel, "Submission to NIST's post-quantum project
(2nd round)," 2019. [Online]. Available: https://csrc.nist.gov/
Projects/Post-Quantum-Cryptography/Round-2-Submissions.
[Accessed 11 6 2019].
Competing Interests:
None declared.
Ethical approval:
Not applicable.
Author’s contribution:
RC designed and coordinated this research and prepared the manuscript
in entirety.
Funding:
None declared.
Acknowledgements:
RC wants to thank his PhD supervisor Dr. Ian McAndrew, Dean of
doctoral programs, Capitol Technology University, for his dedication,
encouragement and expert guidenace in this research.
The JBBA | Volume 2 | Issue 2 | October 2019
25
jbba
the
Peer-reviewed Research
The Contractual Cryptoeconomy:
An Arrow of Time for Economics
OPEN ACCESS
ISSN Print: 2516-3949
https://doi.org/10.31585/jbba-2-2-(1)2019
Prateek Goorha
Independent Scholar, Greater Boston, USA
Correspondence: goorha@sent.com
Received: 11 March 2019 Accepted: 18 April 2019 Published: 2 May 2019
Abstract
We consider the potential blockchains have for building a framework for all manner of contracts that
can characterize an economy using the unifying idea of control over their duration. Such a contractual
cryptoeconomy (CCE) would accommodate a broader variety of contracts than smart contracts, which are
suitable for a relatively small portion of the set of all feasible contracts. We proceed by examining the
idea of a contract’s natural life as a common feature shared across all contracts, be they incomplete or
complete. This simplifying idea suggests why providing flexibility over a contract’s duration on a blockchain
– through innovations such as HTLCs — is necessary to increasing the variety of contracts that can be
feasibly represented. We also assess participation in a CCE that features blockchains with differing degrees of
security. We do so by focusing on how the value of a contract is related directly to its natural life for both its
immediate participants and, through externalities across the CCE, to a wider set of users. A key idea provides
the overall impetus: When contracts rely on third-party intermediation, at least some contractual surplus is
dissipated in arbiter rent, making the quality of third-party arbitration as important as its scale. By contrast,
blockchains create contractual mechanisms that act as Coasian exchanges that can internalize this arbiter rent.
However, crucially, the degree to which their use requires forgoing contractual complexity and absorbing the
cost of externalities can determine the relative benefits provided by a CCE.
Keywords: Arbiter Rent; Contracts; Duration; HTLCs; Blockchains; Thomas Jefferson; Economic Arrow of Time;
Coasian exchange; Contractual Cryptoeconomy
1. Introduction
‘The Earth belongs, in usufruct, to the living.’ - Thomas
Jefferson [1].
While its specific focus is blockchains, the impetus for
this article came from Thomas Jefferson’s observation
cited above. It is extracted from a letter he wrote to
James Madison in 1789, impelled by his belief that a
contract’s length should be set at a fixed period.
Jefferson’s actuarial skills had enabled him to calculate
that – owing to the average life span of individuals
then – by the end of that period one of the parties
would likely have died. Contracts, he proposed, should
be rescinded every 19 years. The clock should, in
other words, be reset so that the usufruct of contracts
can more correctly reflect their true creators and
beneficiaries.
It is on the nature of this link
between the usufruct of a contract on the one hand
and its duration on the other that we shall focus our
attention on in this paper; it is, we shall see, key to the
The JBBA | Volume 2 | Issue 2 | October 2019
class of contracts that can be feasibly represented on a
blockchain.
Yet there is also a second aspect of Jefferson’s thought
process that is worth appreciating: The idea that this
usufruct is at risk of being delimited and squandered,
and that a mandated reset of some kind is the only
tool at hand to prevent this undesirable eventuality. Is
a resetting of the clock necessary to realign usufruct
across blockchains too, and can such a tool feasibly
even exist for blockchains without necessarily violating
its immutability characteristics? In relation to this idea
we shall also consider a particular source of risk to
a contractual cryptoeconomy (CCE) that emanates
from the externalities between its different blockchain
instantiations, and even between the CCE and the
traditional economy based on legacy contractual
mechanisms.
It is clear that Jefferson believed that successors to a
contract should not be forcibly shackled to the actions
of its predecessors, and that events from a time in
the past should not take hostage those who create
27
jbba
the
events at a time in the future. Given the linear and
immutable nature of blockchains, does a CCE not
meet this standard? We shall see how an appreciation
of contractual variety, and developing mechanisms
for a CCE to accommodate them, suggests quite the
opposite.
Jefferson understood moral hazard all too well. On
placing a hard duration on contracts, he wrote in his
letter:
contracting – and actualize a CCE – would seem then
to depend not merely on their ability to serve as the
proverbially dispassionate ‘arrow of time’, but also to
enable guiding such an arrow’s direction tractably when
a contractual application requires it. This is to say that
a CCE needs an ‘economic arrow of time’ that appeals
to time as the arbiter, but in a manner better suited to
maximizing contractual usufruct.
2. The Contractual Cryptoeconomy
‘This would put lenders, and the borrowers also, on
their guard. By reducing too, the faculty of borrowing
within its natural limits, it would bridle the spirit of
war, to which too free a course has been procured by
the inattention of money lenders to this law of nature,
that succeeding generations are not responsible for the
preceding.’ [1]
His thinking inspires considering the following
broader question: Do all contracts have some notion
of a natural life in common? Perhaps more generally:
What is the foundational role of time in transactions
and contracts? Is it to provide an absolute and final
verdict, like some digital super-precise photo-finish
line in a race? Or is to serve as the permissive referee
who taps an unseen wristwatch significantly, merely to
encourage a dawdling participant to adopt a somewhat
swifter pace of progress?
These are sweeping questions, but here we shall
examine these issues more narrowly in the context of
blockchains, for which a key characteristic is precisely
that of the inherent immutability of transactions they
enable alongside an impartial adherence to a linear
process that features time-stamping as a tool to appeal
to time as the ultimate impartial arbiter.
When time is connected with a sequence of
transactions – say as with an uncomplicated supply
chain or assembly line – the linearity of a secure
blockchain can trivially be used to reliably and usefully
bolster the operations with verifiability. However, a
large swath of research in economics examines the
myriad of situations where such linearity isn’t quite
so obvious. Often sequential investments are not
fully specifiable ex ante, which is to say that there is
no obvious chain to follow for contracting parties. In
several cases such incompleteness is actually desirable
to both parties in a contract, for example when the
nature of incompleteness is itself a basis for setting
expectations yet leaving room for creativity around
a shared goal. And, frequently, the sequence should
become terminable ex post to protect the value of an
investment, as in cases where recontracting becomes
necessary; in such cases, the prospect of recontracting
limits the ability of the inefficient ex post allocation to
endure.i
For blockchains to be a genuinely useful tool for
The JBBA | Volume 2 | Issue 2 | October 2019
While constitutions, transnational pacts, purchase
agreements and employment contracts can all be seen
as forms of ‘contracts’, they have several obvious and
several subtle differences that justify their examination
within the purlieus of separate fields of study.
Indeed, whether a constitution can be considered a
(social) contract in any real and useful sense is hardly
an uncontentious idea. [2] provides several useful
references and a general discussion, and, interestingly,
also considers their applicability within the context of
piratical constitutional contracts. See, also, [3].
Economists, for example, have long studied the
difference between a complete contract and an
incomplete contract. The incompleteness stems from
the fact that a vast majority of contracts in the real
world cannot be made fully contingent on a specifiable
state of the world. Smart contracts, by contrast, are
premised on fully specifiable states of the world and
are, in this respect, an interesting example of complete
contracts. For incomplete contracts, moral hazard is a
prime motivator. In other words, incomplete contracts
focus on ownership of productive assets because their
use can often not be fully specified ex ante, nor can
it frequently be monitored. It has been argued that
such incomplete contracts could, in theory, be made
equivalent to complete contracts provided only that
the parties are averse to risk and we assume that they
can at least provide a probability distribution for future
outcomes, even if they cannot predict exact features
of the possible future states of the world. This works,
provided we have access to an incentive compatible
mechanism that motivates the parties to declare the
state that does eventuate truthfully. [4]
It is not, therefore, hard to understand why incomplete
contracts are ubiquitous in the real world. For a
discussion of the difference between complete and
incomplete contracts in the context of blockchains, see
[5], [6] and [7].
Abstracting from differences between the variety of
applications of contracts and their broad types, here
we wish to focus thought on an essential similarity
observed by the third president: the idea of a natural
life. Time – its duration; its ability to be reset; its
impending horizon – is central to all contracts, and
it is this shared basis of a ‘progression across a series
28
jbba
the
of transactions’, each linked in some direct or indirect
manner to time, that makes their association with
blockchains an interesting subject to consider.
2.1 Internalizing Arbiter Rent
Blockchains operate on the essential principle of
time-stamping a batch of transactions and permit
the possibility of doing so immutably, verifiably and
in a decentralized manner; crucially, depending on
features of their particular instantiation, the degree to
which these features are secured from sabotage varies.
This lends them to be particularly useful for at least
two functions: providing a reliable infrastructure for
broadly accessible capital markets and serving as a basis
for reifying and securing property rights.
It has long been recognized in the development
literature that a government’s ability has to credibly
secure property rights and encourage well-functioning
financial markets are key to its capacity to signal its
commitment to private-sector investment, especially of
the variety that is accretive to longer-term growth. (See,
as examples, [8] and [9].)
Between these two functions, there is little doubt that
weak property rights do more insidious damage to
growth prospects than weak financial markets. [10]
However, it has been shown time and again that the
temptation for governments to spurn this advice and
turn to rapacious rent-seeking activities remains a real
threat to stunting economic growth and development
prospects. On this point, [11] is particularly convincing.
This broader observation is important for the context
of contracts, since third-party arbitration is key not
just to a contract’s enforceability but to the overall set
of contracts that can eventuate in an economy. This
function of arbitration, enforcement and verification
that governments provide – primarily through their
legal code and system of courts – yields them valuable
economic rent, which we can see as ‘arbiter rent’.
For a contractual space based on blockchains, however,
the economic value that is represented by the arbiter
rent is internalized within the same system that employs
actors on the decentralized network to function as
independent and neutral verifiers. Traditional arbiter
rent, in this broad sense at least, is reimagined by
blockchains. It is retained within the transactional
parameters defined by the contractual space a
blockchain’s design implements. It is not, however,
retained entirely within a given contract directly.
To see this point, contrast the contracts that rely on
third-party verification provided by institutions with
those contracts that entirely dispose of them, operating
purely on the basis of trust between parties.
When third-party arbitration is essential, the general
The JBBA | Volume 2 | Issue 2 | October 2019
institutional quality (see, for example, [12]) and the
reliability and efficiency of courts (as argued in [13])
becomes paramount to the extent that a contract
can generate surplus. The potential for regulatory
distortions resulting in higher arbiter rent and lower
contractual surplus for the participants looms large
over the market.
Since institutions also provide the broader context to
societal trust, or ‘social capital’ between contracting
agents, it is hard to separate the effects of each.
However, it has been shown that, even controlling for
such endogeneity, social capital still plays a very strong
role in enabling beneficial contracts; [14] provides a
discussion on the relative role that social capital plays
in financial contracts in the context of southern versus
northern Italy. Frequently such trust-based contracts
are used by those who would otherwise be priced out
of any feasible arbiter-enforced contract for a service
that entails some form of direct or indirect arbiter
rent. As such, ‘trust’ provides a useful social benefit
for contracting.
More generally, the ability to remove the extractive
influence of arbiter rent reduces the inframarginal cost
and enables greater contractual surplus.
The trouble, of course, is that contracts that are strongly
reliant on trust can only operate within the narrow
swath of applications where prosocial behaviours and
norms among the participants are socially embedded,
which is to say, ordinarily only within extended families
and smaller communities. [15] proposes a modelling
framework to see the role of social capital for informal
contractual enforcement in a network. The network
connections themselves serve as a collateral that can
be used for borrowing between participants in the
network.
Contracts that are enabled by blockchains derive their
basis from a third source. Neither do they directly rely
on social capital – derived from interpersonal trust –
nor do they need institutions that provide third-party
verification and arbitration – premised upon state
sanction. Instead, they replace both with a system based
on a consensus protocol for their users that requires no
intrinsic trust among its participants, but that creates
a reliable contractual space where transactions can be
made strongly verifiable.
Contracts operating on a blockchain are designed to
internalize the arbiter rent, thereby creating a dedicated
economic space – the ‘contractual cryptoeconomy’ –
which is more broadly accessible than those contracts
that rely entirely on social capital and less costly than
those that rely on third-party arbitration.
Naturally, this is the macro-view for a theoretical
motivation for the CCE. In practice, there are
significant problems that make it unclear whether a
29
jbba
the
CCE can indeed satisfactorily accommodate all other
forms of contracts.
Consider, for example, that competing blockchain
applications can be built ad nauseam without any
costless manner to distinguish between their relative
quality of implementation ex ante. Centralized thirdparty arbitration mechanisms, on the other hand,
are usually maintained under a system that grants
monopoly power over the arbiter’s function to
the state that defines the contract’s jurisdiction. In
theory, such proliferation can curtail the extent of the
internalization of arbiter rent. Contractual surplus
faces the risk of being dissipated when contracts are
allocated inefficiently between the legacy contractual
environment based on courts as the ultimate third-party
arbiter and the contractual blockchain economy. On the
other hand, proliferation might also generate positive
externalities for the CCE. Much depends on whether
we can make variegated blockchain implementations
compatible and convergent to theoretical ideals of
a contracting platform: interoperability between
blockchains certainly permits such compatibility in a
technical sense, though it only characterizes a fraction
of all feasible implementations of blockchains. We
shall develop these points further with the help of a
simple model later on.
2.2 Flexibility in Contractual Time
Since the prior description of all relevant states that
may affect a contract is either infeasible or impractically
expensive, contracts are routinely left incomplete,
without fully state-contingent clauses. Incompleteness
in contracts may exist for other reasons as well, some
of which are unavoidable and some deliberate.
Consider the case of a bilateral externality, for example,
where the parties engage in a contract without prior
information on the size of the externality that might be
generated by the scale of the primary activity that one
of the relevant parties engages in, and can therefore
not effectively set appropriate terms. [16] Conversely,
consider the case of crafting a contract to optimize on
the choice of providing contractual flexibility in the
terms of the contract ex ante as opposed to making
them more rigid. With flexible terms established ex
ante, the parties have more freedom to adjust their
behavior ex post, once they have better information on
how to make the division of surplus more agreeable
to both parties. At the cost of some loss of control,
flexibility in contractual terms can incentivize creativity,
make individual initiative more likely to affect surplus,
motivate the selection of more suitable projects, and
so forth. This suggests that there may be a strong role
for deliberate incompleteness in contracts as a tool
to set the expectations for the parties involved. [17]
Smart contracts, in such cases, would obviously be
suboptimal.
The JBBA | Volume 2 | Issue 2 | October 2019
Given the large variety of contracts in the real world
that are best described as incomplete, it is worth
considering the Jeffersonian idea of deliberate
recontracting (in other words, the proviso of a horizon
for contracts) for the particular context of contractual
implementations on a blockchain.
Blockchains have potential as a theoretical construct for
recreating consensual outcomes across a decentralized
market structure to leverage the value that is inherent
in aggregating distributed information efficiently. For
economics this is nothing short of revolutionary, for
the very obvious reason that we can now imagine a
third alternative to the dichotomy that underpins the
‘market versus organization’ dilemma (or firms versus
institutions) that [18] outlined. Blockchains permit
market orderings for value-generation that suspend
both the invisible hand of the price mechanism of
markets and the direct guiding hand of hierarchies
in organizations; [19] terms this third mechanism a
‘cryptographic stigmergy’.
The fact that they are immutable, time-dependent
databases that can be made exceedingly censorship
resistant makes the market and social orderings that
public blockchains enable especially durable. However,
blockchains are not amenable to providing nuanced
consideration of incentives and are, as a consequence,
less suitable for tackling contractual complexity that
such orderings must routinely grapple with. In this
respect, scaling solutions for blockchains that introduce
layers upon a foundational blockchain consensus
protocol, and then erect a network upon it that can
flexibly represent nuance that contractual incentives
contain are noteworthy.
Consider the idea of a hashed time-locked contract
(HTLC), which illustrates the connection between
providing some degree of control over time and the
types of contracts that it makes possible. An HTLC
is a particular kind of smart contract that has been
developed for the scalable transactional layer – the
Lightning Network – built on top of the underlying
Bitcoin blockchainii. [20] The Lightning Network
enables the creation of task-specific payment
channels off-chain that permit the aggregation of
several transactions that can be mapped onto fewer
transactions on the base layer, thereby lowering the
average transactional cost. In the limit, only two
transactions on the more expensive and slower base
layer suffice for a multitude of transactions on any
given payment channel: the initial transaction that funds
the payment channel shared by two or more agents in a
‘multisig’ account, and a final transaction that updates
the status of accounts after the payment channel is
closed off. This effectively loosens the dependency of
a multi-transactional contractual relationship on the
immutable time-stamping feature of the underlying
Bitcoin blockchain. Transactions proceed by a process
30
jbba
the
of sequential consensus over mutually preferred states
that, once agreed to, simultaneously also invalidate
deprecated states by instituting a penalty comprising
the loss of all staked funds should the previous state be
surreptitiously used to close off the payment channel
and published to the blockchain.
to the time it is completed. Therefore, HTLCs can be
seen as an organic and dynamic method to define a
nexus of contracts that determines the boundary of
a traditional firm, and it uses the underlying Bitcoin
blockchain as the third-party arbiter for a wider set of
contracts that inhere to traditional firms.
The network aspect of the Lightning Network permits
several ‘hops’ across any of its nodes with open
payment channels. This allows any participant to effect
payments to anyone else on the network much more
swiftly and cost-effectively than is possible with the
base Bitcoin layer. Moreover, the open nature of the
network creates a contestable market for transactions.
This is important since it ensures that competitive
market pressures influence the terms of all new
contracts, and the terms that pertain to the division of
the surplus that the contract can entail.
While this setup seems to have effectively created the
precursors to decentralizing a firm on a scaling solution
for blockchains, it remains far from certain that it rings
in the demise of traditional firms. Issues pertaining to
residual control over productive and complementary
assets, management of teams, the assumption of risks,
the delegation of authority across agents, and so forth
are complex contractual issues that will require further
developments, very likely relying on a suite of suitable
technologies working seamlessly to integrate not just
blockchains, but other types of ledger technologies as
well.
For our context, these developments are significant for
two compelling reasons.
First, more specifically, HTLCs make the significance
of a natural expiry for a contract in eliciting efficient
contractual investments clearer to apprehend. An
HTLC operates by first creating the hash of a secret.
The secret must be revealed by the recipient in order to
access some funds at stake. If the hash is kept private,
we have a more constrained and state-contingent
contract between a buyer and a seller. If the hash
is made public, we can then imagine a tournament
between a buyer and a pool of sellers who competitively
exert efforts to discover the secret. An HTLC also
involves an interplay between a definite time at which
the contract expires and the ability to adjust the terms
of the contract to the demands of a specific context
by decrementing this duration sequentially. An HTLC,
therefore, places emphasis on publicly specifying a
‘fixed duration’ before the contract’s outcome becomes
inviolably published to the Bitcoin blockchain, thereby
ending the contract and forcing a reset.
While this reset afforded by the base layer is
Jeffersonian in spirit, the HTLC permits context to
provide variability in the duration itself. This is because
an HTLC also features a method to introduce a
‘flexible horizon’ as a method to motivate and negotiate
efforts that help generate contractual usufruct in the
shadow of the Bitcoin blockchain. As such, HTLCs
are designed and can be developed further to capture a
broader swath of contracts in practice.
Second, and more broadly, note that the Lightning
Network could, in theory, permit defining any arbitrary
architecture for some given contractual mechanism as a
subgraph of its overall network structure. In particular,
it becomes feasible to specify not just any set of nodes
that are involved within a transaction, but also the order
in which they are involved from the time it is initiated
The JBBA | Volume 2 | Issue 2 | October 2019
3. The Economic Arrow of Time
The prospective role of time-stamping processes
that can then be marched immutably through time
looms large over applications that are considered
for blockchains. There is something attractive about
relying on time as an arbiter.
Of course, this view is rather limiting in its capacity for
the nuanced insight needed for dealing with contractual
variety in the real world. It is indeed true that some
physical processes feature an ‘arrow of time’: closed
systems with increasing entropy concretely indicate an
irreversible and directional arrow through time. Most
famous among these is the thermodynamic arrow of
time implied by the Second Law of Thermodynamics.
Other processes, however, are characterized by a ‘timereversal invariance’, in that they do permit possibilities
for a reversal of the process. [21] It is, therefore,
even at a rather general level, infeasible to rely on the
inviolability of some implied arrow of time as the
essential shared foundation for real world applications.
In the context of blockchains, while the law is
routinely taken to unleash the value of transactional
immutability, it can very well also be taken to suggest
the level of difficulty required to successfully sabotage
precisely that feature.
For instance, a supply chain, from initial input to
final output, may appear to represent a process very
conducive to the arrow of time analogy. Yet, the value
of any such arrow shrinks markedly when we are
interested in more than merely describing the process of
sequentially linking units into a chain. By concentrating
emphasis on the curation of information, a supply
chain on a blockchain sets aside several interesting and
important contractual issues, implicitly assuming that
they can all be considered complete.iii This delimits
the usefulness of blockchains by relegating a host of
31
jbba
the
incomplete contractual transformations that affect the
potential usufruct of the supply chain. By contrast,
when we begin to consider aspects of the various
contracts that exist between entities on a supply chain,
the emphasis shifts from one of an inexorable and rigid
arrow of time, to one that can be guided – perhaps
better seen to be a distinct ‘economic arrow of time’.
In a standard contract in economics (where the
principal is risk neutral and the agent is risk averse) the
prospect of renegotiating a contract serves to give the
contract precisely this characteristic of time-reversal
invariance. When an agent must select costly effort
that is unobservable by the principal over the course
of a contract and, simultaneously, must also commit
to not renegotiating, she exposes herself to a degree
of risk. To elicit the optimal level of effort through
any form of assurance of a payoff that corresponds
with the higher-level of effort, the principal would
need to distinguish between agents who would select
suboptimal levels of effort from those who select
the optimal level; instantly, we shift the focus of the
problem to one of resolving adverse selection rather
than a strict sequential progression through the
contractual parameters.
3.2 Aspects of time
Contracts that feature degrees of state dependency
and propensities for renegotiation underscore the
relevance of two aspects of time that are related but
subtly different in their effects: ‘timing’ and ‘duration’.
It is broadly understood that timing is integral to the
very rationale for a range of contracts. The sequence
and ordering of investment decisions that are stipulated
by a contract can determine the amount of contractual
surplus generated. One of the key messages of
transaction cost economics is that timing is key to
ameliorating a variety of opportunistic behaviors
that are inspired by appropriable quasi-rents; timing
is, indeed, central to motivating efficient investments,
reducing a range of social externalities and, of course,
in setting the overall boundaries of a firm with respect
to the market. A key difference between a simple statedependent smart contract and an HTLC is that the
latter permits a method to algorithmically delimit the
appropriable quasi-rents involved in a contract.
Contracts can also vary widely in their duration.
Constitutions usually have far more enduring lives
while several securities contracts can have extremely
short lives. Thus, a provision for flexibility over both
aspects of time that affect the contractual horizon
is both necessary and appropriate for any generic
contractual template.
The idea of a contractual duration has been examined
at some length in the literature. [22] and [23], for
example, suggest that, broadly, contract length depends
The JBBA | Volume 2 | Issue 2 | October 2019
on the level of uncertainty the investment represents
and the cost of renegotiation. Short-term contracts
with the option of renegotiation have been contrasted
with longer-term contracts. For example, [24] suggests
that, in the absence of a commitment to refrain from
renegotiation, a buyer and seller will prefer engaging
in a sequence of short-term contracts. (See also [25],
which contains useful references.) [26] demonstrates
the efficiency of short-term contracts over the long
run and [27] suggests that even spot contracts can be
efficient when inter-temporal smoothing concerns are
not a consideration.
Concerns with sequential short-term contracts arise
when pertinent information over incentives and
behaviour is revealed asymmetrically and in a manner
that is correlated over time so that bargaining power
shifts squarely towards one party to the detriment
of the other. Here, smart contracts that also strongly
guarantee anonymity of the participants ex ante would
incentivize undertaking a sequence of shorter-horizon
contracts, thereby avoiding introducing undesirable
divisions of surplus owing to the asymmetric
revelation of private knowledge. [28] develops a class
of contracts for the Lightning Network, called ‘discreet
log contracts’, that provide anonymity as a feature
while also reducing the scope of malfeasance by the
third-party nodes that act as intermediaries.
3.2 Phases in a contract’s natural life
Regardless of the nuance over aspects of time within
a contract’s natural life, most contracts are usually seen
dichotomously – a contract either exists or it does not,
whether in prospect or in fact, and whether it is tacit
or explicit. However, consider that most contracts exist
within contextual environments that impinge upon
them and lead them through ‘states’ of validities over
the duration of their existence. Generally, we can call
these states of a contract over its natural life its ‘phases’
and enumerate at least three: acceptability, vulnerability
and termination.
Quite simply, when an extant contract accords with
the intention of its participants it can be said to have
acceptable validity; when, over its life, it is susceptible
to being either terminated or unacceptable (at risk of
renegotiation) then it can be said to have a vulnerable
validity. The contract’s natural life can thus be parsed
into phases that describe stages of its existence, and
we can subsequently consider the transitions of the
contract through these phases over its duration.
While fluid transitions between phases that might exist
within a contract are not explicitly considered in the
literature, the general issue is recognized as one that
is significant in its social welfare implications. For
instance, [29], which focuses on contrasting ex ante
dispute resolution arrangements with ex post dispute
32
jbba
the
resolution; while ex ante arrangements enhance joint
surplus, they tend not to be legally enforced.
Our consideration of a contract’s natural life here
is not meant as a sensationalist departure from the
literature on contracts, but to draw attention to the
fact that several aspects of a contract, such as its
prospect for renegotiation, uncertainty, moral hazard,
and adverse selection, can usefully be seen as being
internal to the contract and manifested as transitions
across its phases. HTLCs provide a very promising
first step towards resolving such issues for contracts
on the blockchain, but they are hardly flexible enough
to accommodate complex transactions, multi-layered
contracts, complicated property rights, and a host of
other issues.
Contracts are often generic templates. They might
be drawn up to be applicable across a multitude
of transactions, with only limited consideration for
specific circumstances, or they might be drawn up and
made inviolable through the passage of time or across
its applications in a given period. Several examples
can be offered in support of this observation of a
social, political and economic nature: primogeniture,
constitutions and union-negotiated employment
contracts, for instance, are contracts that, perforce,
do not specify all feasible states explicitly, but their
incompleteness for a particular context or contingency
(intentionally or not) completely defines their phases.
This restates the result in [25], but for a different
reason: there the observation is that incompleteness
on account of transaction costs need not be relevant
so long as payoffs are known. Here, incompleteness
can never be entirely eradicated even if payoffs are
known so long as parties to a contract ‘care’ about
the transitions of the contract over its phases in its
duration, and that the phases are finite and foreseen. It
is, of course, feasible that the phases in the duration are
a mechanism relevant to the contracting parties since it
retrieves information relevant for payoffs.
4. Externalities in a CCE
Recall that [18] argued that there is an inherent ‘cost to
discover market prices,’ and that firms are motivated
by the ability to suspend using the price mechanism
of the market to coordinate production, permitting
the firm’s manager instead to direct the coordination
of resources. Similarly, a blockchain can be seen as a
‘Coasian exchange’: Participants are brought together
through an ecosystem that acts as a mechanism for
the coordination of activities organically, and which
is motivated by the ‘cost of discovery for the market
value of consensus’.
Arguably, the Lightning Network, as a second-layer
scaling solution for Bitcoin, can be seen as an effort
to encourage the Coasian exchange dynamics of the
The JBBA | Volume 2 | Issue 2 | October 2019
underlying layer by undertaking an ‘intervention’ to
ameliorate the negative externalities from congestion
on the base layer.
Intrinsic to these relative costs of discovery (those
for the market prices versus those for the market
value of consensus) are several externalities, positive
and negative, that a contractual blockchain economy
represents relative to the traditional economy.iv These
externalities may inhere in the social resource costs
for securing a blockchain implementation’s consensus
protocol. They may arise from the information costs
imposed by implementations of blockchains with less
desirable characteristics or the lack of interoperability
between the more desirable ones.v They may even
pertain to the developments upon it that alter its value
proposition.
There is a broad source of externalities that the
regulation of cryptocurrencies imposes upon this
relative cost consideration. Broadly, this source inheres
to the difference between the market for ideas as
opposed to the market for goods. Externalities are a
common basis for excessive regulatory intervention in
the market for goods, especially when contrasted with
a reluctance to apply similar regulatory predispositions
in the market of ideas. It was Coase again ([30] and
[31]) who articulated why a definitive treatment of
this issue was essential to any real consideration of
externalities affecting production in markets. The
notion, frequently heard, that software ought to be
treated by the government as speech makes this point
quite clear.
4.1 A traditional modeling framework on realigning
externalities
Let us briefly consider this issue of externalities as they
pertain to participation in the CCE. We use a simple
framework that should be instantly familiar to students
of public economics.
We might imagine that the economy comprises some
secure blockchain υ with a market price of pν, and
other blockchain instantiations conducive to hosting
contracts. We can think of this ecosystem collectively
as our contractual blockchain economy, Υ.
The point is to imagine a scenario where participation
in υ provides a net external value to other participants
across Υ, and that it is only partially accounted for
by the participants within the secure blockchain.
To capture the idea that other participants in the
blockchain economy experience varying degrees of
externality effects from υ, the nature of which can also
be multidimensional, we only need assume that the
joint probability distribution P(V,E) is known to all who
participate in Υ, where participation in Υ yields a private
benefit of V to the individual and, simultaneously, it
33
e of
νν
jbba
the
̌ν
̌ν ̌ν
Υ a net positive externality
Υ of value, E. In terms ̌νcomprises
the group of participants in υ who create
inspires
̌ν
ΥΥΥ
ΥΥΥ
υ
and the latter
of our Jeffersonian premise,
E can be seen to represent feweřνnet positive Υ-wide externalities
υ
Υ
υ who would be
that part of the contractual usufructs in υ that are not group Υwould have been participants
Υ to generate such positive externalities to Υ.
more likely
directly internalized by its participants.
Υ
υ
Υ
υ υυ
Υ
It is useful to see why this joint probability distribution It is quite obvious that any ability to price discriminate
would make sense for Υ. Information is inherently between these groups would be an immediate source
distributed, andΥso the secured and decentralized for an increase in the net social externality gain from
ΥΥ
the secured
by υoutcomes.
Υand decentralized
economic
orderings
enabledΥbyeconomic
υ entailsorderings
more of enabled
a Υ market
the
secured
and
decentralized
economic
orderings
enabled
υ υυ
the
economic
orderings
thesecured
secured
anddecentralized
decentralized
economic
orderings
enabled
by
gain toand
those
who are more
marginalized
by anyenabled
of
thebyby
Υ
Υ
υ toweconcern
themselves
withonincreasing
In our context
can imagine
higher layers
the securethe
distributively inefficient economic orderings that are blockchain
blockchain υ to concern themselves
with increasing
the
υ’s
base
settlement
layer. This
vi
blockchain
υ
to
concern
themselves
with
increasing
more centralized and less secure than υ.υ
blockchain υ to concern themselves
with increasing
the
υ’s base settlement
layer. This
υ υ υυ
the transaction throughput
base settlement
υ’s of
baseυ’ssettlement
layer. This
To fix ideas further, let us capture the social marginal layer.value
Thisofnaturally
as a screening
the dataserves
on υ through
time andmechanism
those who are
υ
cost that the security
of
υ
entails
on
the
Υ
ecosystem
between
those
participants
who
are
interested
theare
value
of
the
data
on
υ
through
time
and thoseinwho
υ
Υ
value
of immutability
the data on υ of
through
time of
andthe
those
who
υ υυpermits us toΥdefine
with s. This
and
the value
data
onare
ΥΥ a net socialΥ gain in the security
the secured
and
by υare interested, more
blockchain-enabled economic
system; for
individual ς economic
υυ through orderings
time and enabled
those who
Υ andecentralized
ς proximately, with securing frequent transactions at low
ΥΥ−
υ
in Υ,Υ participation inκυ yields a net
social
gain
of
υ
Υ
υ
the secured and
decentralized
enabled by υ ς cost, which we can capture with the variable ς.
−−orderings
κ=(V
+E-s). κ κκ economic
−
Υν
ς
υ
ς
ς
υ
This̃latter group
would then have a joint probability
ς
̃ς
υ
ς
(V,E),
whereas
and
the
former
group
distribution
of
P
ς
̃
υ
ν
All new entrants
to Υ face
νν p
ΥΥΥ
ν for access to the most
υ if pν exceeds the entrant’s
νν
secure blockchain. Naturally,
ν
reservation price she does
not υ
participate
υ in υ. As
υ υυ
such, a recognition of the presence of the net positive
makes itΥ advantageous for Υ to institute a
υΥ
ΥΥΥexternality
υ a social subsidy for all entrants to υ.
method
to
provide
κ
υ υυ
In the case
of the secure blockchain,
the magnitude
of
Υ
υ
υ
this ‘subsidy’
can
be
seen
as
the
social
resource
cost,
R,
κ
υ υυ −
Υ
of securing υ, and
it can be written as
ν
ν
would have (Pς ) ̃(V,E). The social resource cost, R, of
Υ
securing υ, now becomes
∫κ ς
∫ Υ∫ κ̃ς
υ ∫
ς
̃
∫
κ ς
∫
∫
κ
ν ∫
0 κP
0 κ(P ς (V, E))dV dE
̃ς (V, E). dV dE + p∫
∫
R=
− p∫νν ∫
ν
ν
ν
∫pν ∫0 κ
Υ
R = ∫νp∫ν∫∫
κ(P(V, E))dV dE ,
ν0∫∫ κκ
ppν 00
Υ
υ
Υ receives begins at pν
where the value that a participant
νν
ν
υ
without an upper bound whereas the externality
from
a given participant ranges from zero without an upper
υ
υ
bound.
p 0
pν 0
With the cost of access to the higher transactional
l
layer as l, the efficient price for participants solely in
υ
the settlement layer abides the same condition:
̌ν −
̌ν
α
̌ν ) − s
̌ν = α(E/p
p
̌ν −
̌ν
α
whereas, for the groups on the transactional layer, the
price abides:
̌ν + α ( E/p
̌ν ) .
s + l = p
layer is lower than that for the group on the base layer
and the net positive externalities are higher through
ν
discrimination. Specifically, the ability to sort the
00
0
participants in this manner permits participation in
ν
υ
the base layer to exclude those for whom E was lower
υ υ
This suggests that the efficient price υ
for υ is
but V was higher, and include them in the transactional
Υ
were
to
efficiently
select
a
priceinstead.
for υ we would have:
layer
̌ν
∂ κ∂ υ
−∫ κ
∫pν ∫
∂ ∂∂R/
∂ ∂∂pυ
υ0 υ −
0
=−∫
−∫∫κ κκ(P(V,
E))dE = 0.
α ( ⁄̌ν ) −
̌νp
ν ν=α α
(α(⁄(Ě
) )−)−−s,
̌
̌
ν̌
⁄
νwe
υ the number of transactional
ν would have:
p
Υ were to efficiently select a price for ⁄
υ̌
There is a technical limit for
υ
∂
∂
υ
−
∫
κ
layers
that
are
likely
to
be
υbuilt on υ as well as a practical
re α (E⁄̌ν ) rrepresents the average externalities 0 limit on the need for such layers. At a general level, this
where
α α(α(⁄(̌
) )) p
ν
ν νat the
⁄̌
∂e⁄efficient
υ̌
−∫
κ price; thus, the social marginal causes a degree of pooling of the participants across
efficient
pri
υ ∂applicable
ς
0 social marginal gain.
υ
υ υΥυ
cost is equal to the
the two groups
and creates limits to the ratcheting
ς
ΥΥΥ
ain υ with a market price of
cour
cour
cour
ς
Υ were to efficiently select a price for υ we would have:
̌νςς
̌ν
ας
ν
ν
υ
ΥΥwere
to
efficiently
select
a
price
for
υ
we
would
have:
̌
̌the
select
aaprice
have:
α
If
Υ to
were
to efficiently
a for
price
for
υwould
we ∫would
The
price
for
the
group
participating
in
transactional
ς
ς
Υwere
were
toefficiently
efficiently
selectselect
price
forυυwe
wewould
have:
∫ κ
pν 0
have:
ν
ν
p 0
pν 0
With
With
With
effect that curators of such layers might developςmerely
In words, even an efficient υprice consideration
discriminate on the basis of ς more and more
̌ν forυα (to⁄price
̌ν ) −See [32], who initially developed this idea in
υ υυ in perfectly.
υ can do no better than lump in relevant nuances
average externalities.
in Υ for whom the private the context of a two-period incentive contract with
̌
m υ, the nature of which can
α ( ⁄Those
Υν
̌ν ) −
ethe
nature
of
which
can
𝛶𝛶𝛶𝛶
benefit
and
net
externality
is
below the
social marginal asymmetric information on observed performance.
Υ
the
nature
of
which
can
ΥΥ
nature of which can
α (E ̌ν ) υ
cost participate in υ (V is higher than⁄p
υ υυ); those for
whom it is higher doυnot participateicient
(V is lower
5. Concluding Remarks
pri than
α (E⁄̌ν )
of
course,
because
the
former
p . This is undesirable,
Υ
icient pri
υ
Υ JBBA | Volume 2 | Issue 2 | October 2019
υ
The
34
ν
of externality effects from υ, the nature of which can
Υ υ
[2] P.T
[2] P.T
M
[2] P.T
M
M
[3] D.
[3] D.
C
[3] D.
C
C
jbba
the
With Jefferson’s observation as the overarching
impetus, we have examined the issue of a natural life for
contracts as a feature they all share. Contracts do not,
however, last forever, and the notion of their stability is
only relevant when seen from the perspective of their
vulnerability to partial failure; in other words, how
contracts behave over the course of their entire life
deserves attention. Blockchains draw attention to this
overarching fact. They hold the potential to develop
a platform, with features of a Coasian exchange, that
permits the use of an economic arrow of time that
can accommodate a genuine contractual blockchain
economy.
The Jeffersonian standpoint of favouring the living
is an acknowledgmentvii that the contractual enabling
of the usufruct is premised upon a period that comes
to a close. Logically, this period can be examined as a
duration with a definite commencement and expiration,
but with varying states of validity as economic rent
from a relationship varies over the course of the
duration of the contract; the contract then can be seen
to have conditional probabilities for these validities
over its duration. When contractual usufruct is lost
through the course of a contract’s natural life, the
Jeffersonian solution of recontracting makes patent
sense. However, when an economic arrow of time
can be appealed to that can service complete as well
as incomplete contracts, recontracting does not have
to be the default solution. The linear transformations
that blockchains accommodate so well provide a
strong basis for contractual mechanism design; the
organic networks that fluidly emerge from the evolving
patterns of contractual usufructs that higher-layer
scaling solutions provide suggest that a much wider
variety of incomplete contracts can be accommodated
as well. Together this gives us a strong basis for a
contractual blockchain economy.
Constitutional Contracts”, American Journal of Political
Science, vol. 31, no. 1, pp. 142-168, 1987.
[4] E. Maskin and J. Tirole, “Unforeseen Contingencies and
Incomplete Contracts”, Review of Economic Studies, vol. 66, no.
1 (special issue on contracts), pp. 83-114, 1999.
[5] P. Goorha, “Blockchains as Implementable Mechanisms:
Crypto-Ricardian Rent and a Crypto-Coase Theorem”, Journal
of the British Blockchain Association, Volume 1, Issue 2,
Available: 10.31585/jbba-1-2-(4)2018., 2018a.
[6] P. Goorha, “A Comprehensive Contracting Solution Using
Blockchains,” SSRN Working Paper, Available: http://
dx.doi.org/10.2139/ssrn.3237076 , 2018b.
[7] J.S. Gans (2019) ‘The Fine Print in Smart Contracts,’
SSRN Working Paper, Available: http://dx.doi.
org/10.2139/ssrn.3309709, 2019.
[8] D. North, Institutions, Institutional Change and Economic
Performance, Cambridge University Press, 1990.
[9] D. Rodrik, “Promises, Promises: Credible Policy Reform
via Signalling”, The Economic Journal, vol. 99, no. 397, pp.
756–772, 1989.
[10] S. Johnson, J. McMillan and C. Woodruff, “Property
Rights and Finance”, American Economic Review, vol. 92, no.
5, pp. 1335–1356, 2002.
[11] A. Shleifer and R. Vishny, The Grabbing Hand:
Government Pathologies and Their Cures, Harvard University
Press, 2002.
[12] D. Acemoglu, D. and S. Johnson “Unbundling
Institutions”, The Journal of Political Economy, vol., no. 5, pp.
949-995, 2005.
Admittedly there is a long way to go before the
contractual blockchain economy can be seen as a real
alternative – indeed, one that is to be preferred in an
era of technologies that favor distributed information
– to the traditional economy. However, the fact that
several of the necessary components exist in theory
and practice even today is a real source for optimism.
[13] A. Shleifer, “Efficient Regulation”, Regulation vs.
Litigation, ed. Daniel Kessler, NBER and University of
Chicago Press, pp. 27-43, 2010.
References:
[15] D. Karlan, M. Mobius, T. Rosenblat and A. Szeidl,
“Trust and Social Collateral”, Quarterly Journal of Economics,
vol. 124, no. 3, pp. 1307-361, 2009.
[1] T. Jefferson “Letter to James Madison”, The Founders’
Constitution, vol. 1, chapter 2, document 23, University of
Chicago Press. http://press-pubs.uchicago.edu/founders/
documents/v1ch2s23. Html, 1789.
[2] P.T. Leeson, “The Calculus of Piratical Consent: The Myth
of the Myth of Social Contract”, Public Choice, vol. 139, no.
3/4, pp. 443-459, 2009.
[14] L. Guiso, P. Sapienza and L. Zingales, “The Role of
Social Capital in Financial Development”, The American
Economic Review, vol. 94, no. 3, pp. 526-556, 2004.
[16] R. Pitchford and C.M. Snyder, “Coming to the Nuisance:
An Economic Analysis from an Incomplete Contracts
Perspective”, Journal of Law, Economics and Organization,
vol. 19, pp. 491-516, 2003.
[17] O. Hart and J. Moore, “Contracts as Reference Points”,
Quarterly Journal of Economics, vol. 123, pp. 1-48, 2008.
[3] D.D. Heckathorn and S.M. Maser “Bargaining and
The JBBA | Volume 2 | Issue 2 | October 2019
35
jbba
the
[18] R.H. Coase, “The Nature of the Firm”, Economica, vol.
4, pp. 386-405, 1937.
[19] P. Goorha, “The Return of ‘The Nature of the Firm’:
The Role of the Blockchain”, Journal of the British Blockchain
Association, volume 1, no. 1, Available: 10.31585/jbba-1-1(2)2018, 2018c.
[20] J. Poon and T. Dryja “The Bitcoin Lightning Network:
Scalable Off-Chain Instant Payments”, Available: https://
lightning.network/lightning-network-paper.pdf, 2016.
[21] B. Roberts, “When We Do (and Do Not) Have a
Classical Arrow of Time”, Philosophy of Science, vol. 80, no.
5, pp. 1112-1124, 2013.
[22] J. Gray, “On Indexation and Contract Length”, Journal of
Political Economy, vol. 86, no. 1, pp. 1-18, 1978.
[23] S. B. Vroman, “Inflation Uncertainty and Contract
Duration”, Review of Economics and Statistics, vol. 71, no. 4,
pp. 677-681, 1989.
[24] O. Hart and J. Tirole “Contract Renegotiation and
Coasian Dynamics”, Review of Economic Studies, vol. 55, no.
4, pp. 509-540, 1988.
[25] M. Dewatripont, “Renegotiation and Information
Revelation over Time: The Case of Optimal Labor Contracts”,
Quarterly Journal of Economics, vol. 104, no. 3, pp. 589–619,
1989.
[26] P. Rey and B. Salanie, “Long-Term, Short-Term and
Renegotiation: On the Value of Commitment in Contracting”,
Econometrica, vol. 58, pp. 597-619, 1990.
[27] D. Fudenberg, B. Holmstrom, and P. Milgrom “Shortterm Contracts and Long-Term Agency Relationships”, Journal
of Economic Theory, vol. 51, pp. 1-31, 1990.
Note that, when such inefficiencies are the source of
rent for one of the parties in a contract, recontracting
is undesirable to her, even if recontracting may lead to
a Pareto improvement for the contract.
i
Recall that the base layer of Bitcoin was the first
blockchain application and was created with the
intention to serve as a digital payment system for
networks that obviated the need for third-party
intermediation. Bitcoin secures its transactions through
the use of a consensus algorithm based on the idea of
incontestable proof of work done; it is operationalized
by nodes on the network called miners who must invest
in costly dedicated computer hardware and energy to
competitively solve cryptographic puzzles in order
to earn the right to batch transactions into a block
that then gets appended to the Bitcoin blockchain.
This provides the miner a payoff comprising a fixed
number of bitcoins and a smaller variable transaction
fee, while enabling all participants on the network to
verify the accuracy of the overall ledger of transactions
independently.
ii
iii
For instance, along each stage of a supply chain that
features a typical two-sided market, incentives provided
by the reference platform linking both sides of the
market may well change.
iv
Naturally, there are several externalities that pertain to
the mechanisms of a given blockchain implementation
as well. These may include externalities imposed by
the activities of a single node that affects the entire
network, such as when it engages in transactions that
increase the latency across the entire network and ties
up a disproportionate share of resources. However, we
are more interested here in considering externalities
directly relevant to the broader contractual blockchain
economy.
A key benefit of contracts on interoperable blockchains
is in reducing the costs of complexity in describing
outcomes that pertain to a contract. For example, the
nature of investments that parties make at time 2, once
the contract has been put into operation at time 1, is
often seen as being sufficiently complex to make them
effectively beyond being independently verified by any
third-party, such as a court. Blockchain interoperability
can assuage this concern by folding in more and more
aspects of an incomplete contract within the ambit
of what can be feasibly verified publicly by a ‘trusted
third-party’. Such aspects can pertain to the nature of
the investments, but also to the realized state of the
world ex post.
v
[28] T. Dryja, “Discreet Log Contracts”, undated working
paper, MIT Digital Currency Initiative, last accessed: February,
2019.
[29] S. Shavell, “Alternative Dispute Resolution: An Economic
Analysis”, Journal of Legal Studies, vol. 24, no. 1, pp. 1–28,
1995.
[30] R.H. Coase, “The Market for Goods and the Market for
Ideas”, American Economic Review, vol.64, no. 2, pp. 384–
391, 1974.
[31] R.H. Coase, “Advertising and Free Speech”, The Journal
of Legal Studies, vol. 6, no. 1, pp. 1-34, 1977.
vi
[32] J. Laffont and J. Tirole “The Dynamics of Incentive
Contracts”, Econometrica, vol. 56, no. 5, pp. 1153–1175,
1988.
Mutatis mutandis, this can be seen to extend beyond
the contractual blockchain economy to the traditional
economy as well.
vii
The JBBA | Volume 2 | Issue 2 | October 2019
The author readily admits that Jefferson’s observation
36
jbba
the
was more profound than what is made of it for the
purpose of this paper!
Competing Interests:
None declared.
Ethical approval:
Not applicable.
Author’s contribution:
PG designed and coordinated this research and prepared the manuscript
in entirety.
Funding:
None declared.
Acknowledgements:
None declared.
The JBBA | Volume 2 | Issue 2 | October 2019
37
jbba
the
Peer-reviewed Research
OPEN ACCESS
ISSN Print: 2516-3949
https://doi.org/10.31585/jbba-2-2-(3)2019
Singapore’s Open Digital Token Offering Embrace:
Context & Consequences
Robert W. Greene1, David Lee Kuo Chuen1,2
1
Singapore University of Social Sciences, Singapore
2
Stanford University Distributed Trust Initiative, USA
Correspondence: rwg1819@gmail.com
Received: 17 May 2019 Accepted: 4 June 2019 Published: 28 June 2019
Abstract
The overall global public’s ability to purchase some portion of a digital token project’s initial batch of
tokens is the defining feature of an open digital token offering. Using a dataset that differentiates this token
distribution model from other varieties – a distinction often underemphasised in regional analyses of digital
token sale trends – this research estimates 2017-18 open digital token offering activity by jurisdiction, finding
that Singapore-registered projects accounted for 21 percent of Q3/Q4 2018 dollar-volume, more than
any other country. Conversely, by late 2018, previous hubs of this distribution model represented a much
smaller share. Reasons for Singapore’s rise as a global hub of the open digital token offering are explored,
with a particular focus on examining contrasting regulatory approaches to distinguishing between this token
distribution model and traditional securities offerings. Notably, 11 percent of Singapore-registered Q3/Q4
2018 token offering dollar-volume was purely-private, versus 94 percent in the U.S. Policy considerations
related to this distribution method and the open digital token offering are presented, as are contrasting
outcomes: this research estimates that over 70 percent of Singapore’s one-to-two-year-old open token
offerings resulted in operational networks or minimum-viable-products, versus fewer than 40 percent of U.S.
private sales. Also, about 40 percent of smart contract platform projects that conducted 2017-18 token sales
were Singapore-registered – many more than in any other country. For reasons explored in this research,
these findings support the view that open digital token offerings benefit projects aiming to concurrently raise
funds, build up a user-base, and incentivise technologists to contribute to project development. Moreover,
risks to retail participants posed by this distribution method are manageable. Singapore’s policy approach
towards open digital token offerings has benefited the Lion City, which was likely home to more digital token
projects that conducted 2018 token sales than any other city in the world.
Keywords: Arbiter Rent; Contracts; Duration; HTLCs; Blockchains; Thomas Jefferson; Economic Arrow of Time;
Coasian exchange; Contractual Cryptoeconomy
JEL Classifications: G18, G28, F39, K20, K22, K23, O16, O38
1. Introduction
Last year, the dollar-volume of digital token
distributions eclipsed the value of initial public
offerings within a developed economy with robust
capital markets infrastructure. In Singapore, the value
of 2018 initial public offerings was $730 million [1], yet
as this research finds, Singapore-registered 2018 digital
token sales raised over $1.6 billion [2].i Understanding
Singapore’s important role within the digital token
economy first necessitates understanding digital
tokens. For purposes of this research, “digital tokens”
The JBBA | Volume 2 | Issue 2 | October 2019
are defined as “transferable units generated within a
distributed network that tracks ownership of the units
through the application of blockchain technology” [3].
Unlike traditional financial assets, a digital token serves
as “a cryptographically-secured representation of a
token-holder's rights” to perform certain functions
within or receive benefits from a token network [4] [5].
In the case of virtual currencies, a type of digital token,
these rights include the ability to store and exchange
value within a distributed peer-to-peer payments
network [4].
39
jbba
the
The initial batch of a project’s digital tokens can be
distributed through various approaches.ii For the last
two years, the most popular approach, by far, has been
the open digital token offering. This article defines
an “open digital token offering” as occurring when
a software project or business provides purchasing
access to some portion of the initial supply of digital
tokens associated with a project to most of the global
public (some barriers to access may existiii). Conversely,
“private initial token sales” – an alternative form of
initial token distribution –restrict outside purchases of
any share of a project’s first batch of tokens to only
a relatively small number of participants, generally
high-net-worth or institutional buyers. Funds raised via
these two distribution approaches are commonly used
to finance the development of a digital token project’s
network, platform, or services.
Section 2 presents estimates of 2017-18 regional open
digital token offering trends, finding that the Cayman
Islands, Singapore, Switzerland, and the U.S. were the
four major hubs of “successful” 2017 open digital
token offerings,iv but by Q3/Q4 2018, only Singapore
remained a leading home to this distribution model.
During the second half of 2018, purely-private digital
token sales accounted for nearly all digital token
offering dollar-volume in the U.S., but just 11 percent
in Singapore. This contrast stems from differing
regulatory approaches examined in Section 2, which
help explain Singapore’s role as a dominant hub of the
open digital token offering.
Section 3 assesses the outcomes of Singapore’s
open digital token offering embrace, finding that a
significantly greater share of one-to-two-year-old
Singapore-registered open token offerings relative to
U.S. private initial token sales resulted in operational
associated networks or services. Singapore-registered
token offerings also accounted for a disproportionately
large share of 2018 “smart contract platform”
projects (defined below). For reasons explored in
Section 3, these outcomes provide support to the
view that open digital token offerings are well-suited
for projects aiming to use a token distribution event
to concurrently fundraise, build up a project’s userbase, and incentivise contributions by developer
communities. Of course, operational projects are not
inherently successful projects, and many may fail, so
the scope and management of risks facing open digital
token offering retail participants is examined with a
focus on Singapore. The extent to which token projects
registered in Singapore are primarily physically-based
in the country is also estimated.
Section 4 concludes that the consequences of
Singapore’s open digital token offering embrace
highlight beneficial features of this distribution model,
which is well-suited for the swift development and
deployment of new distributed services and networks.
Photo
by Andrew
Kow on Unsplash
The JBBA
| Volume
2 | Issue
2 | October 2019
Singapore, likely home to more digital token projects
that conducted successful 2018 token sales than any
other city, stands to benefit in the years to come from
its open digital token offering embrace.
2. How Policy Influenced Regional Trends in
2017-18 Open Digital Token Offerings
While the first open digital token offering took place in
2013 [5], overall token sale volume did not dramatically
accelerate until 2016 and 2017 [3], after Ethereum’s
2015 release. Ethereum is an open-source, decentralised
platform for executing and recording “smart contracts”
(“set[s] of promises, specified in digital form, including
protocols within which the parties perform on these
promises” [6]) [7], and as a “smart contract platform,”
it allows programs to be transparently appended to and
run on its blockchain [8]. The late 2015 development
of an open-source standard for Ethereum smart
contracts [9], the “ERC-20 standard,” provided best
practices for coding applications that generate new
types of tokens recorded on the Ethereum blockchain
(tokens “run on top of Ethereum”) [8]. This drove a
huge increase in token offering volume [10] – roughly
$12 million was raised via 2015 digital token sales; in
2016 and 2017, that figure grew to over $100 million
and over $7.5 billion, respectively [3]. While a token
project may eventually swap tokens running on top of
Ethereum for tokens recorded and transmitted within
a new network it launches [5], at least 60 percent of
digital tokens with active secondary markets run on
top of Ethereum [11], and many of these may be used
within applications designed to permanently run on
the Ethereum blockchain.
By 2017, hundreds of token projects were utilising
smart contract platforms so that project supporters
across the world could receive some of a project’s
initial batch of tokens in exchange for providing funds
to the project to support its team’s efforts to either
build out an application or launch a new network – a
process that some policymakers consider to be, under
certain circumstances, an unregistered public securities
offering. The disclosure, reporting, and structural
requirements of a registered public securities offering,
however, are quite costly [12]. Moreover, while
regulators may exempt small-sized securities offerings
or sales exclusively available to wealthy persons from
certain public offering requirements, exemptions can
lead to regulatory complications for digital token
projects, as explained later. Indeed, widely-distributed
digital tokens are often quite different than the equity
securities historically issued via these public and
private channels, which generally entitle holders to a
share of distributed profits and the value of a firm,
and can provide ownership rights [13].v One analysis of
253 digital tokens distributed from 2014 through late
2017 finds that three dominant uses are: 1) access to
platform services (68 percent); 2) project governance
40
jbba
the
decisions (25 percent); and 3) payments (21 percent)
[14]. Other research finds that over 75 percent of
tokens distributed by projects from 2013 through early
2017 provide access to platform services and about
half enable payments [5]. Given the stark differences
between traditional securities and most digital tokens,
applying traditional securities regulations to small
projects focused on developing digital token networks
can make those projects unworkable [15].
The analysis below estimates 2017-18 successful
open digital token offering trends by jurisdiction
using data primarily obtained through collaboration
with Smith+Crown, a research and advisory
consultancy. Policy factors that influenced 2017-18
trends, particularly those related to securities law, are
concurrently examined, revealing external and internal
forces behind Singapore’s role as a hub of the open
digital token offering. The Appendix sets forth the
methodology used to construct this study’s dataset –
unlike other datasets used to analyse regional token
offering trends, it distinguishes between private initial
token sales and open digital token offerings as well as a
token project’s physical location versus the jurisdiction
of legal registration for its token sale.
2.1. Switzerland, the Cayman Islands, Singapore,
and the U.S.: 2017 Open Token Offering Hubs
As Figure 1vi shows, in 2017, Switzerland was the
jurisdictional home to a larger dollar-volume share of
successful open digital token offerings (24 percent)
existing definition of a security [16] (which in the
digital token context, largely hinges on a determination
of whether ownership or a security interest over the
token issuer’s assets exists [4] [17]) enabled many open
digital token offerings to not be classified as securities
offerings throughout 2017. Singapore’s emergence as
a hub of this distribution model was further enabled
by its technologist and legal communities’ proactive
engagement with the Monetary Authority of Singapore
(“MAS”) [18] – the country’s chief financial markets
regulator. By August 2017, the MAS clarified that
many open digital token offerings are not securities
distributions [4]. In November 2017, it released
guidelines providing clear examples of what token sale
activities do and do not constitute a securities offering,
as well as regulatory responsibilities of a digital token
project [19].vii
In 2017, the regulatory posture towards open token
offerings in the U.S., Switzerland, and the Cayman
Islands was relatively less proactive. U.S. Securities and
Exchange Commission (“SEC”) 2017 enforcement
actions provided some insights into circumstances
under which the agency will, by applying an ambiguous
multi-pronged legal test,viii view open digital token
offerings to constitute securities distributions, but
activity to clarify the regulatory status of particular
offering approaches was minimal [3]. Switzerland’s
top securities market regulator announced in late 2017
that it was investigating some previous open digital
token offerings for regulatory breaches [20], but that
depending on the circumstances, open digital token
offerings may not be considered securities distributions
[21]. In the Cayman Islands, regulators made no
statements regarding the applicability of securities
law to open digital token offerings, although its legal
definition of a security is quite narrow [22].
2.2. Singapore Remained an Open Token Offering
Hub as Policies Elsewhere and Market Trends
Shifted
Figure 1. Unlike other jurisdictions, Singapore was a leading
home of open digital token offerings during both 2017 and
2018
than any other jurisdiction in the world, followed by the
Cayman Islands (19 percent, of which over 75 percent
was U.S.-located EOS’s token sale) [2]. Singapore and
the U.S. accounted for 14 and 11 percent of total 2017
dollar-volume, respectively, and no other jurisdiction
made up more than five percent [2].
In Singapore, the Securities and Futures Act’s preThe JBBA | Volume 2 | Issue 2 | October 2019
Figure 1 illustrates how by the second-half of 2018,
negative digital token market conditions contributed
to a sharp dollar-volume decline in open token
offerings relative to early 2018. Yet these conditions
were global, and do not explain the disparate shifts in
jurisdictional shares of dollar-volume illustrated above.
By the second-half of 2018, Cayman, Swiss, and U.S.
open digital token offerings accounted for just seven,
seven, and two percent of total global dollar-volume,
respectively [2]. Alternatively, 21 percent occurred in
Singapore, 42 percent took place in smaller jurisdictions
(each accounting for less than five percent of total
2018 volume), and 20 percent was in the U.K. [2].ix
Several factors help explain these outcomes. For
starters, some Asian jurisdictions banned forms
of open digital token offerings in Q3 2017 [23].
41
jbba
the
Singapore’s location, regulatory approach towards open
digital token offerings, and rules on foreign investment
and visitors – some of the most open in the world,
and less-restrictive than those in Switzerland, the
U.K., and the U.S. [24] – drew Asia-based projects to
Singapore the following year amidst these unfavourable
regulatory shifts. Indeed, data indicate that half of
non-Singapore-based digital token project teams that
conducted successful 2018 Singapore-registered token
offerings were primarily physically-located elsewhere
in Asia (excluding Russia) [2]. Also in 2018, policy
changes drove Swiss banks to close accounts for
digital token projects in large volumes and reportedly
dramatically increased the relative cost of certain
compliance processes [25] [26]. As some countries’
regulatory approaches towards open token offerings
became stricter, relatively more accommodative policy
frameworks in the U.K. and smaller countries [27]
attracted a few sizable open digital token offerings [2],x
helping explain the larger role of these jurisdictions
in 2018 as compared to 2017. Conversely, after the
enormous EOS sale ended, Cayman-registered projects
accounted for a much smaller share of global open
token offering dollar-volume. Perhaps most notably,
2018 U.S. securities regulation trends drove an embrace
of the private initial token sale over the open digital
token offering for digital token projects seeking sale
participants from the U.S.
$1,000,000 [33]. Many Regulation D safe-harbour sales
used the U.S. accredited investor threshold as a sole
determinant for sale participation regardless of the
country where those seeking to purchase tokens were
legally-domiciled.xiii Several U.S. token projects utilised
the Regulation Crowdfunding (“CF”) exemption to
conduct open digital token offerings exempted from
public securities offering requirements, but these
capped sales likely accounted for just 1 percent of
overall 2018 token sale dollar-volume [2] [34].xiv
As Figure 2xv shows, by Q3/Q4 2018, purely-private
sales accounted for 94 percent of the dollar-volume of
successful U.S. token offerings, versus just 11 percent
in Singapore – which continued to embrace the open
digital token offering [2]. Indeed, in October 2018, the
MAS’s Managing Director Ravi Menon stated that the
MAS had “seen quite a lot of [digital token offering]
activity that is not security related” [35]. In only one
instance in 2018 did the MAS announce that it directed
2.3. Singapore Continued Embracing Open Token
Offerings as Private Initial Token Sales Dominated
in the U.S.
In February 2018, U.S. SEC Chairman Jay Clayton
notoriously remarked: “every [initial coin offering] I’ve
seen is a security” [28]. If a token project markets to
the general public securities not registered with the
SEC or issued under certain SEC exemptions, then
the issuer can be subject to serious penalties, as well
as costly class-action lawsuits [29]. Moreover, non-U.S.
persons can be subject to enforcement actions for
offering unregistered securities to U.S. persons [30] [31].
Chairman Clayton’s sweeping remarks were followed
by about twenty enforcement actions related to digital
tokens [32], and perhaps as many as 100 subpoenas
of token projects.xi By year-end, no open digital token
offering was affirmatively classified by name by the
SEC as not being a securities distribution.xii
Accordingly, throughout 2018, token projects
increasingly banned U.S. persons from participating
in open token offerings and relied upon private
initial token sales involving a “Regulation D”
securities offering to access U.S. buyers. Regulation
D allows fundraising events to avoid expensive public
securities offerings requirements if sales are generally
restricted only to “accredited investors” – primarily
defined as individuals/households making over
$200,000/$300,000 annually or with a net worth over
The JBBA | Volume 2 | Issue 2 | October 2019
Figure 2. Purely-private token sales accounted for almost all
Q3/Q4 2018 U.S. token sale volume, but were relatively
minimal in Singapore
a project to cease offering tokens to Singapore-based
persons because it considered the project’s sale of
tokens to be an unregistered public securities offering
[36] – evidence of a clearly-understood regulatory
distinction between open digital token offerings and
traditional securities distributions.
Surely, some 2018 private initial token sales took place
without involving Regulation D. The vast majority of
private initial token sale events, however, involved a
Regulation D offering [2]. Overall, approximately 75
percent of 2018 digital token offering dollar-volume
was open, rather than purely-private [2].
3. Exploring the Implications of Singapore’s Open
Digital Token Offering Embrace
Clearly, a number of external and internal policy factors
contributed to Singapore’s emergence as a global
open digital token offering hub. This section explores
outcomes of Singapore’s embrace of this token
distribution model related to: 1) the operational status
42
jbba
the
and focus of Singapore-registered token projects; 2)
open token offering retail participant risks; and 3)
the extent to which Singapore-registered projects are
physically-based in the country.
3.1. Open Digital Token Offerings Offer Unique
Benefits Related to Widespread Token Distribution
As a recent study helps illustrate, open digital
token offerings can enable digital token networks
to concurrently raise funds and build up an active
community of users and project contributors [37].
Indeed, research finds that higher community
engagement is associated with a token project’s success
[38]. As one analysis explains, despite the growing
relevance of institutional investors in open digital token
offerings (about 37 percent of 2018 token offerings
through mid-Q3 reportedly conducted private sale
stages [39]), “putting a token into the hands of 50,000
people who actually went through the process of
research and purchase is the best form of mass-market
engagement available that will increase the likelihood
of project success” [40].
Figure 3xvi suggests that an open token offering
model may indeed accelerate the pace at which token
networks and applications become operational relative
to purely-private sales. It shows that by mid-June 2019,
over 70 percent of Singapore-registered projects that
conducted successful open digital token offerings from
Q3 2017 through Q2 2018 launched “operational”
investors of digital tokens distributed via a Regulation
D safe-harbour offering [41] [42]. This impedes the
ability of projects that conduct token offerings using
the Regulation D safe-harbour to leverage primary or
secondary digital token markets to facilitate widespread
token ownership by a globally-dispersed community of
developers. As the founder of a project that conducted
one of the largest private initial token sales to date
remarked after apologising that his project’s token
offering would be purely-private: “[the accredited
investor threshold] excludes some of the groups
most capable of investing in these kinds of projects,
for example, cryptography and game theory PhD
students” [43].
Indeed, Ethereum sale data and subsequent survey
data suggests 50 to 75 percent of Ethereum’s open
digital token offering participants contributed less
than $1,000 [44] [10], and the network’s early attraction
of a large community of well-informed retail tokenholders played a critical role in its success [10]. Open
digital token offerings facilitate participation in opensource software development and create a sense
of empowerment and ownership, thus mobilising
programmers to test and improve underlying software
[14]. This open-source ethos is particularly important
for the development of smart contract platforms such
Figure 4. Nearly 40 percent of smart contract platform
projects that conducted successful 2017-18 token sales are
registered in Singapore
Figure 3. A greater share of one-to-two-year-old Singapore
open digital token offerings resulted in operational networks
and products relative to Q3 2017 - Q2 2018 U.S. private
initial token sales
products or networks related to the token sale, versus
37 percent of U.S.-registered projects that successfully
conducted a Regulation D safe-harbour private initial
token sale during that time. “Operational” is defined
as the publicly-available release of: 1) a token network’s
open-source and live testnet or mainnet; and/or 2)
a minimum-viable-product usable by the project’s
targeted customer base.
One driver of the discrepancy in Figure 3 is that
regulations restrict the re-sale to non-accredited
The JBBA | Volume 2 | Issue 2 | October 2019
as Ethereum – it is difficult to imagine developers
building applications or engaging with strangers on
a platform that they do not understand and cannot
test [45]. Accordingly, as Figure 4xvii illustrates, a
disproportionate share of smart contract platform
projects that conducted 2017-18 token offerings were
Singapore-registered, likely due in part to Singapore’s
embrace of the open digital token offering. These
projects largely aim to increase the range of economic
and social contexts in which open blockchain solutions
can be applied by building platforms that overcome
some of Ethereum’s scalability challenges.
43
jbba
the
3.2. Risks to Open Digital Token Offering Retail
Participants are Manageable
Open digital token offerings can result in inexperienced
persons purchasing tokens from digital token projects
that are not long-term viable – many projects have
failed or probably will fail [46] [47]. Yet inexperienced
retail exposure to these tokens is much more likely to
be facilitated by online accounts easily-opened with
secondary market trading venues rather than directly via
open digital token offerings. Moreover, few Singaporeregistered digital token offerings involve substantial
direct Singapore-based retail purchases, although this
is reportedly in part because some projects restrict
Singapore persons’ participation in token offerings
[48]. Research also suggests, however, that most digital
token offering participants contribute modest-size
dollar-amounts, and that these contributors largely have
a technology background or meaningful investment
experience [10]. Indeed, participation in open digital
token offerings usually necessitates a moderate level
of technological acumen and market awareness – a
purchaser often must understand how to operate an
ERC-20 “wallet,” and sale participation may require
first signing up via a whitelist.
Surely, despite these barriers, the low cost of structuring
an open digital token offering can allow fraudsters to
solicit funds with relative ease. As much as ten percent
of pre-mid-2018 digital token sale dollar-volume were
scams [49], although some research suggests that the
degree of fraud is much lower [50] and that “investors
are shrewd enough to spot [scams]” [46]. Moreover, in
Singapore, fraud can result in lengthy jail sentences [51],
and while some uncertainty surrounds the applicability
of criminal law to matters involving digital tokens
[17], two foreigners recently charged for promoting
a fraudulent digital token project may face up to five
years in jail [52]. Furthermore, the Singapore-registered
entity responsible for a token sale must have at least one
Singapore citizen or permanent resident on the board,
as well as a local secretary [53]. These gatekeepers,
as well as Singapore’s legal community (which drafts
token offering documents) and the Accounting and
Corporate Regulatory Authority, further minimise the
likelihood of fraudulent open digital token offerings.
While Singapore’s open token offering embrace has
not made it a safe-haven for fraudulent projects,
markets for some tokens generated via Singaporeregistered offerings have been nefariously manipulated.
Bad actors can create false optimism and spikes in a
token’s value, and then sell the token at a market high,
driving a large price decline that harms retail tokenholders [54]. In fact, Singapore’s first open digital token
offering resulted in a token later manipulated by such a
pump-and-dump scheme [55]. Singapore’s government
has warned of this predatory market behaviour [56],
but retail investors can still fall victim. Yet market
The JBBA | Volume 2 | Issue 2 | October 2019
manipulation is a serious issue for many digital tokens
– not a problem exclusive to those generated via open
token offerings.
3.3. Nearly Half of Singapore-registered Token
Projects are Primarily Physically-based in the
Country
Despite the large number of Singapore-registered
projects primarily physically-based outside the city,
Figure 5xviii shows that a greater share of Singaporeregistered projects that successfully conducted
token sales in 2018 are domestically-based relative
to the respective share of Switzerland- and Caymanregistered projects primarily physically-based in those
jurisdictions [2]. Surely, at 46 percent, the share of
projects physically-based in Singapore has room
to grow. Yet a recent industry survey finding that
Singapore is the world’s leading “crypto hub” city
Figure 5: Almost half of projects that conducted successful
2018 Singapore-registered token offerings were primarily
physically-located in the city
notes that its strengths relative to other cities include
not only the robust “activity” of its digital token
project community, but also Singapore’s “international
ecosystem” [57]. Indeed, Singapore’s relative openness
to foreign visitors [24] enables internationally-diverse
project teams not primarily physically-located in the
country – many of which are based elsewhere in Asia,
as mentioned in Section 2 – to regularly visit and
maintain a secondary presence there.
Moving forward, Singapore will benefit from its
physical concentration of token projects, as research
indicates that geographically-concentrated innovation
within a particular field begets relatively deeper and
swifter innovative activities [58]. Data indicate that
Singapore was likely home to more projects that
successfully conducted digital token sales in 2018 than
any other city, with the second- and third-highest being
San Francisco (including Palo Alto) and London [2].
4. Conclusion
44
jbba
the
The open digital token offering can enable projects
to simultaneously: 1) raise funds for the development
of a project’s network, platform, or service; 2) build
up a user-base; and 3) incentivise globally-dispersed
communities of developers to contribute to a project.
While in certain jurisdictions, this token distribution
model may be deemed to be a securities offering, in
practice, the open digital token offering and digital
tokens it produces are often fundamentally different
than traditional securities distributions and securities,
respectively. Singapore’s emergence as a global hub of
the open digital token offering was enabled not only
through existing legal frameworks and constructive
steps to produce regulatory clarity regarding securities
law, but also by its geographic location and openness to
foreign visitors and capital.
The inclusiveness of open digital token offerings, as
well as Singapore’s regulatory clarity regarding this
distribution model, help explain why a greater share of
one-to-two-year-old Singapore-registered open digital
token offerings, relative to U.S. private initial token sales,
have resulted in operational networks or minimumviable-products, and why so many token offerings for
2018 smart contract platform projects were Singaporeregistered. Indeed, open digital token offerings are
well-suited for incentivising the development of opensource projects. While this distribution model can ease
the ability of bad actors to conduct fraud, fraudulent
projects are likely not a major concern in Singapore,
in part due to local gatekeepers and strict laws. There
are also practical barriers-to-entry associated with open
token offerings that preclude large-scale participation
of an uninformed public.
While open digital token offerings have flaws and can
support likely-to-fail projects, trends highlighted in this
research support claims that this distribution model
is advantageous relative to securities offerings and
private initial token sales for certain types of projects,
particularly those focused on launching distributed
open-source networks and services. Because of its
embrace of the open digital token offering, as well as
other policy factors, Singapore is well-positioned to
remain a hub of open blockchain innovations.
5. Appendix
Token projects included in the dataset used in this
research’s estimates of token sale activity (the “Primary
Dataset” [2]) were initially sourced by Smith+Crown
through: 1) a detailed Smith+Crown intake survey
submitted by token projects; 2) Smith+Crown’s bimonthly reviews of online data aggregators and
the SEC EDGAR database; and 3) Smith+Crown’s
reviews of ongoing industry events. Before including
projects identified through these channels in the
Primary Dataset, Smith+Crown confirmed that project
team member identities were transparent, there was
The JBBA | Volume 2 | Issue 2 | October 2019
a reasonable amount of public documentation and
information available on the project, the project raised
over $25,000, and funds raised were not returned to
initial backers – for purposes of this article, these
criteria are used to classify a “successful” digital
token offering. This sourcing methodology makes
the scope of Smith+Crown’s data smaller relative to
those of some popular online aggregators, which may
exclusively rely on information sourced through token
project self-reporting.
To obtain dollar-raised figures, Smith+Crown sourced
token projects’ self-reported dollar-raised amounts
from data aggregators, and then verified those amounts
using on-chain analysis,xix SEC EDGAR, other
government filings, reports from reliable news sources,
or official project statements. If a raise amount was
unverifiable, then Smith+Crown entered the amount
raised by the project as zero. Generally, token sale
dates were determined using the reported date of a
sale period ending, and multiple sale stages of a single
token offering were treated as a single offering event as
long as: 1) sale terms were largely similar; and 2) sale
periods were not separated by more than thirty days
(otherwise, sales were treated separately).xx
Unlike datasets used in other analyses of global
token sale trends (for example, [27] [59]), the Primary
Dataset clearly distinguishes between a project’s legal
jurisdiction and physical location. The legal jurisdiction
of the entity responsible for a token offering was
determined for almost all 2017-18 token sale dollarvolume, and was identified using information provided
on the sourcing survey, which Smith+Crown verified
and, as necessary, corrected through a review of a
project’s website and sale terms in collaboration with
the authors.xxi To determine the primary physical
location of digital token projects, publicly-available
information on the project’s website was used. When
data was not available, LinkedIn.com information was
reviewed, and the reported city of the project’s or
CEO’s LinkedIn page was treated as the project team’s
location. If that data was not available, then the selfreported location of the CTO or the predominant
location of other project team members was used.
For six percent of the projects reviewed to produce
Figure 5, the primary location of the project team was
listed as unknown, and overall, for approximately 25
percent of 2018 token sale events contained in the
Primary Dataset, project team location information
was unknown or not recorded.
To determine whether a token offering was an open
digital token offering or a private initial token sale,
Smith+Crown and the authors reviewed government
filings, project announcements, reputable news sources,
and token sale terms.xxii Multi-tiered sales consisting
of both public and private sale stages (including
Regulation D offerings followed by public sales) were
45
jbba
the
generally treated as one open digital token offering, in
line with this article’s definition of that distribution
method; conversely, private sales conducted in advance
of cancelled or planned (but yet to occur) open sale
rounds were treated as private initial token sales (for
example, Telegram’s token sale). Digital token projects
that conducted Regulation D offerings concurrently
or shortly before an open digital token offering
that restricted U.S. non-accredited-investors from
participating were treated as part of a single open
digital token offering. Security token offerings and
token sales by projects structured as investment funds
were not treated as open digital token offerings, but
were included in this article’s holistic analyses of digital
token offerings (including Figures 2 and 5).xxiii
[6] N. Szabo, “Smart contracts: building blocks for digital
markets,” 1996. [Online]. Available: http://www.fon.hum.
uva.nl/rob/Courses/InformationInSpeech/CDROM/
Literature/LOTwinterschool2006/szabo.best.vwh.net/smart_
contracts_2.html. [Accessed 15 Jun 2019].
Figure 3 was produced using a definition of
“operational” set forth in Section 3 and developed
in collaboration with Smith+Crown, LongHash, and
other industry participants. Q3 2017 to Q2 2018
Singapore open digital token offerings and U.S. private
initial token sales were classified as “operational” or
not as of mid-June 2019 based on a review of publiclyavailable information. Proof-of-works and proof-ofconcepts were not treated as “operational” projects.
Project classifications used to produce Figure 4 were
developed from Smith+Crown’s review of project
white papers and public information. Based on that
review, Smith+Crown tagged certain projects as “smart
contract platform” projects, meaning that the project’s
primary focus is developing a smart contract platform.
[9] F. Vogelsteller, “ERC: token standard,” GitHub, Nov
2015. [Online]. Available: https://github.com/ethereum/
EIPs/issues/20. [Accessed 15 Jun 2019].
References:
[1] “Singapore IPO proceeds plunge to $730m in 2018,”
Singapore Business Review, 2018 Dec 21. [Online]. Available:
https://sbr.com.sg/markets-investing/in-focus/singapore-ipoproceeds-plunge-730m-in-2018. [Accessed 15 Jun 2019].
[2] Smith+Crown, Dataset on file with authors [see Appendix
for methodology used by Smith+Crown and the authors to
construct dataset], 2019.
[3] Token Alliance, “Understanding digital tokens: market
overviews and proposed guidelines for policymakers and
practitioners,” Chamber of Digital Commerce, 2018.
[4] MAS, “MAS clarifies regulatory position on the offer of
digital tokens in Singapore,” 1 Aug 2017. [Online]. Available:
http://www.mas.gov.sg/News-and-Publications/MediaReleases/2017/MAS-clarifies-regulatory-position-on-the-offerof-digital-tokens-in-Singapore.aspx. [Accessed 15 Jun 2019].
[5] D. Chuen, L. Low and Y. Wang, “Introduction to initial
crypto-token offering,” in Inclusive FinTech: Blockchain,
Cryptocurrency and ICO, Singapore, World Scientific Publishing
Company, 2018, pp. 83-114.
The JBBA | Volume 2 | Issue 2 | October 2019
[7] Ethereum Foundation, “Developer resources: guides,
resources, and tools for developers building on Ethereum,”
[Online]. Available: https://www.ethereum.org/developers/.
[Accessed 15 Jun 2019].
[8] P. V. Valkenburgh, “What does it mean to issue a token
‘on top of ’ Ethereum?,” Coin Center, 10 May 2017. [Online].
Available: https://coincenter.org/entry/what-does-it-mean-toissue-a-token-on-top-of-ethereum. [Accessed 15 Jun 2019].
[10] D. Chuen, L. Low, M. Chwierut, W. Anderson, B.
Lio and B. Downes, “The characteristics of token investors,”
in Inclusive FinTech: Blockchain, Cryptocurrency and ICO,
Singapore, World Scientific Publishing Company, 2018, pp.
125-171.
[11] CoinMarketCap, Dataset on file with authors, 2019.
[12] PricewaterhouseCoopers, “Considering an IPO to fuel
your company’s future? Insight into the costs of going public and
being public,” Nov 2017. [Online]. Available: https://www.
pwc.com/us/en/deals/publications/assets/cost-of-an-ipo.pdf.
[Accessed 15 Jun 2019].
[13] International Monetary Fund, “Handbook on securities
statistics,” 2015. [Online]. Available: https://www.imf.org/
external/np/sta/wgsd/pdf/hss.pdf.. [Accessed 15 Jun 2019].
[14] S. Adhami, G. Giudici and M. Stefano, “Why do
businesses go crypto? An empirical analysis of initial coin
offerings,” Journal of Economics and Business, vol. 100, pp.
64-75, May 2018.
[15] H. Peirce, “Regulation: a view from inside the machine,”
8 Feb 2019. [Online]. Available: https://www.sec.gov/news/
speech/peirce-regulation-view-inside-machine. [Accessed 15 Jun
2019].
[16] Securities and Futures Act (Cap. 289), Section 2(1).
[17] J. W. Lim, “A facilitative model for cryptocurrency
regulation in Singapore,” in Handbook of Digital Currency:
Bitcoin, Innovation, Financial Instruments, and Big Data,
London, U.K., Elsevier Inc., 2015, pp. 361-380.
[18] S. Yep, “How Singapore became Asia's ICO hub,”
LongHash, 30 Apr 2018. [Online]. Available: https://www.
longhash.com/news/14. [Accessed 15 Jun 2019].
46
jbba
the
[19] MAS, “A guide to digital token offerings,” 14 Nov 2017.
[Online]. Available: http://www.mas.gov.sg/News-andPublications/Monographs-and-Information-Papers/2017/
Guidance-on-Digital-Token-Offerings.aspx. [Accessed 15 Jun
2019].
Kistner, “ICO participant liability — could you be liable for assisting
in the sale of unregistered securities?,” Bloomberg Law, 15 Dec 2017.
[Online]. Available: https://news.bloomberglaw.com/securities-law/
ico-participant-liability-could-you-be-liable-for-assisting-in-the-sale-ofunregistered-securities. [Accessed 15 Jun 2019].
[20] FINMA, “FINMA is investigating ICO procedures,”
FINMA, 29 Sep 2017. [Online]. Available: https://www.
finma.ch/en/news/2017/09/20170929-mm-ico/. [Accessed
15 Jun 2019].
[30] T. Hanusik and T. Rodriguez, “I don’t live in the United
States, so how can the SEC sue me? SEC actions against a
foreign national living outside the United States,” Bloomberg
Finance L.P., 2008. [Online]. Available: https://www.crowell.
com/documents/SEC-Actions-against-a-Foreign-NationalLiving-Outside-the-United-States.pdf. [Accessed 15 Jun 2019].
[21] FINMA, “FINMA guidance 04/2017,”
29 Sep 2017. [Online]. Available: https://
w w w. fi n m a . c h / e n / ~ / m e d i a / fi n m a / d o k u m e n t e /
d o k u m e n t e n c e n t e r / m y fi n m a / 4 d o k u m e n t a t i o n /
fi n m a - a u f s i c h t s m i t t e i l u n g e n / 2 0 1 7 0 9 2 9 - fi n m a aufsichtsmitteilung-04-2017.pdf. [Accessed 15 Jun 2019].
[22] N. Rogers and C. Macculloch, “Building blocks for ICOs in
the Cayman Islands,” Cayman Financial Review, 22 Jan 2018.
[Online]. Available: https://www.caymanfinancialreview.
com/2018/01/22/building-blocks-for-icos-in-the-caymanislands/. [Accessed 15 Jun 2019].
[23] J. Russell, “First China, now South Korea has banned
ICOs,” TechCrunch, 29 Sep 2017. [Online]. Available:
https://techcrunch.com/2017/09/28/south-korea-hasbanned-icos/. [Accessed 15 Jun 2019].
[24] I. Vásquez and T. Porcnik, “The human freedom index,
XLSX (2016 “foreign ownership/investment restrictions,”
"capital controls," and “freedom of foreigners to visit”
variables),” Cato Institute, 2018. [Online]. Available: https://
object.cato.org/sites/cato.org/files/human-freedom-index-files/
hfi2018web-revised3.xlsx. [Accessed 15 Jun 2019].
[25] A. Irrera and B. H. Neghaiwi, “Switzerland seeks
to regain cryptocurrency crown,” Reuters, 19 Jul 2018.
[Online]. Available: https://www.reuters.com/article/uscryptocurrencies-banking-switzerland/switzerland-seeks-toregain-cryptocurrency-crown-idUSKBN1K91AY. [Accessed
15 Jun 2019].
[26] K. Sedgwick, “Swiss regulations are driving ICOs away,”
Bitcoin.com, 9 Apr 2018. [Online]. Available: https://news.
bitcoin.com/swiss-regulations-are-driving-icos-away/. [Accessed
Jun 15 2019].
[27] W. Kaal, “Initial coin offerings: the top 25 jurisdictions
and their comparative regulatory responses (as of May 2018),”
Stanford Journal of Blockchain Law & Policy, vol. 2, 2019.
[28] M. Orcutt, “Cryptocurrencies crashed in 2018. Now
they’re right where they should be,” MIT Technology Review, 26
Dec 2018. [Online]. Available: https://www.technologyreview.
com/s/612659/cryptocurrencies-crashed-in-2018-now-theyreright-where-they-should-be/. [Accessed 15 Jun 2019].
[29] G. P. Fondo, M. Chang, M. Spillane, S. Fox and T.
The JBBA | Volume 2 | Issue 2 | October 2019
[31] J. Debler, “Foreign initial coin offering issuers beware: the
Securities and Exchange Commission is watching,” Cornell
International Law Journal, vol. 51, pp. 245-272, 2018.
[32] SEC, “Cyber enforcement actions,” [Online]. Available:
https://www.sec.gov/spotlight/cybersecurity-enforcementactions. [Accessed Jun 15 2019].
[33] U.S. Code of Federal Regulations Title 17 Parts 230.501230.508.
[34] P. H. Lee, “Crowdfunding capital in the age of blockchainbased tokens,” St. John's Law Review, vol. 92, pp. 833-913, 2018.
[35] C. Chanjaroen and H. Amin , “Singapore will help crypto
firms set up local bank accounts,” Bloomberg, 10 Oct 2018.
[Online]. Available: https://www.bloomberg.com/news/
articles/2018-10-10/singapore-aids-crypto-firms-seekingbanks-while-staying-vigilant. [Accessed Jun 15 2019].
[36] MAS, “MAS warns digital token exchanges and ICO
issuer,” 24 May 2018. [Online]. Available: http://www.
mas.gov.sg/News-and-Publications/Media-Releases/2018/
MAS-warns-Digital-Token-Exchanges-and-ICO-Issuer.aspx.
[Accessed 15 Jun 2019].
[37] J. Li and W. Mann, “Initial coin offering and platform
building,” Working paper, Jun 17 2018. [Online]. Available:
https://www.jbs.cam.ac.uk/fileadmin/user_upload/research/
centres/alternative-finance/downloads/2018-af-conference/
paper-li.pdf. [Accessed 15 Jun 2019].
[38] H. T. Sabrina, M. Niessner and D. Yermack, “Initial
coin offerings: financing growth with cryptocurrency token
sales,” ECGI working paper series in finance, working paper
no. 564/2018, Jul 2018. [Online]. Available: https://
ecgi.global/sites/default/files/working_papers/documents/
finalhowellniessneryermack.pdf. [Accessed 15 Jun 2019].
[39] J. Lee, “ICOs are turning exclusive as wealthy investors
snatch up deals,” Bloomberg, 8 Aug 2018. [Online]. Available:
https://www.bloomberg.com/news/articles/2018-08-08/
token-sales-turn-exclusive-as-private-investors-snatch-up-deals.
[Accessed 15 Jun 2019].
[40] M. Dibb, “Are public token sales a thing of the past?,”
Medium: Astronaut Capital, 29 May 2018. [Online].
47
jbba
the
Available: https://medium.com/astronaut-capital/are-publictoken-sales-a-thing-of-the-past-15c89efefa1a. [Accessed 15 Jun
2019].
[41] U.S. Code of Federal Regulations Title 17 Part 230.144.
[42] U.S. Code of Federal Regulations Title 17 Part
230.144A.
[43] I. Allison, “Filecoin laments shutting out crypto supporters
to meet SEC regulations,” International Business Times,
8 Aug 2017. [Online]. Available: https://www.ibtimes.
co.uk/filecoin-laments-shutting-out-crypto-supporters-meet-secregulations-1633620. [Accessed 15 Jun 2019].
[44] V. Buterin, “Ether sale: a statistical overview,” Ethereum
Foundation, 8 Aug 2014. [Online]. Available: https://blog.
ethereum.org/2014/08/08/ether-sale-a-statistical-overview/.
[Accessed 15 Jun 2019].
[45] P. V. Valkenburgh, “What is “open source” and
why is it important for cryptocurrency and open blockchain
projects?,” Coin Center, 17 Oct 2017. [Online]. Available:
https://coincenter.org/entry/what-is-open-source-and-why-isit-important-for-cryptocurrency-and-open-blockchain-projects.
[Accessed 15 Jun 2019].
[46] H. Benedetti and L. Kostovetsky, “Digital tulips? returns
to investors in initial coin offerings,” Working paper, May 20
2018. [Online]. Available: https://papers.ssrn.com/sol3/
papers.cfm?abstract_id=3182169. [Accessed 15 Jun 2019].
[47] J. Daniell, “Ethereum's Vitalik Buterin on ‘tokens 1.0’,”
ETHNews, 24 Oct 2017. [Online]. Available: https://www.
ethnews.com/ethereums-vitalik-buterin-on-tokens-10. [Accessed
15 Jun 2019].
[48] R. Jayaseelan, “Singapore’s ICO gamble,” The Star
Online, 24 Dec 2018. [Online]. Available: https://www.
thestar.com.my/business/business-news/2018/12/24/
singapores-ico-gamble/. [Accessed 15 Jun 2019].
[49] Statis Group, “Cryptoasset market coverage initiation:
network creation,” Bloomberg, Jul 2018. [Online]. Available:
https://research.bloomberg.com/pub/res/d28giW28tf6G7T_
Wr77aU0gDgFQ. [Accessed 15 Jun 2019].
[50] D. Liebau and P. Schueffel, “Cryptocurrencies and initial
coin offerings: are they scams? - an empirical study,” The Journal
of The British Blockchain Association, vol. 2, no. 1, pp. 4755, 2019.
[51] S. Sharpe, “Financial crime in Singapore: overview,”
Thomson Reuters Practical Law, 2017. [Online]. Available:
https://uk.practicallaw.thomsonreuters.com/3-618-7957?__
lrTS=20190325161937875&transitionType=Default&c
ontextData=(sc.Default)&firstPage=true&comp=pluk&bh
cp=1#co_anchor_Ib17914890ef411e798dc8b09b4f043e0.
[Accessed 15 Jun 2019].
The JBBA | Volume 2 | Issue 2 | October 2019
[52] R. Mui, “Two men charged in Singapore with promoting
'fraudulent cryptocurrency' OneCoin,” The Business Times, 10
Apr 2019. [Online]. Available: https://www.businesstimes.
com.sg/government-economy/two-men-charged-in-singaporewith-promoting-fraudulent-cryptocurrency-onecoin. [Accessed 15
Jun 2019].
[53] Flag Theory, “Where to set up a foundation for an ICO
(a comparative analysis),” 6 Oct 2017. [Online]. Available:
https://flagtheory.com/foundation-initial-coin-offering-ico/.
[Accessed 15 Jun 2019].
[54] J. Xu and B. Livshits, “The anatomy of a cryptocurrency
pump-and-dump scheme,” in 28th Usenix Security Symposium,
Santa Clara, CA, USA, 2019.
[55] S. Shifflett and P. Vigna, “Traders are talking up
cryptocurrencies, then dumping them, costing others millions,”
5 Aug 2018. [Online]. Available: https://www.wsj.com/
graphics/cryptocurrency-schemes-generate-big-coin/. [Accessed
15 Jun 2019].
[56] Singapore Government, “Are digital tokens such as
cryptocurrencies a simple, safe, and sure-fire way of making
money?,” Factually, 25 May 2018. [Online]. Available:
https://www.gov.sg/factually/content/digital-tokens. [Accessed
Jun 15 2019].
[57] R. Kurani, “Which are the best locations for blockchain
companies? — we asked our well-travelled crypto friends,”
Medium, 22 Feb 2019. [Online]. Available: https://medium.
com/birds-view/which-are-the-best-locations-for-blockchaincompanies-bd816c940456. [Accessed 15 Jun 2019].
[58] M. Muro and B. Katz, “The new 'cluster moment':
how regional innovation clusters can foster the next economy,”
Metropolitan Policy Program, Brookings Institution, Sep 2010.
[Online]. Available: https://www.brookings.edu/wp-content/
uploads/2016/06/0921_clusters_muro_katz.pdf. [Accessed
15 Jun 2019].
[59] Ernst and Young, “EY research: initial coin offerings
(ICOs),” EY, Dec 2017. [Online]. Available: https://www.
ey.com/Publication/vwLUAssets/ey-research-initial-coinofferings-icos/$File/ey-research-initial-coin-offerings-icos.pdf.
[Accessed 15 Jun 2019].
Dollar-volume figures throughout this article are in
U.S. dollars. This figure includes open digital token
offerings, security token offerings, and private initial
token sales. See the Appendix for methodology used to
calculate dollar-volume figures of 2017-18 digital token
sales presented throughout this research.
ii
Trends related to initial mining events and airdrops
are not analysed in this research.
iii
Obstacles related to retail investor technological
acumen are discussed in Section 3. Additionally,
compliance with anti-money laundering laws,
sanctions. and/or securities law interpretations may
i
48
jbba
the
cause projects to decide to prohibit certain nationalities
from participating in open digital token offerings.
iv
For an explanation of how this research defines a
“successful” digital token sale, see the Appendix.
v
The U.S. legal definition of a security extends beyond
equity and debt securities and includes “investment
contracts,” defined according to a multi-prong common
law test [3] (see footnote viii and accompanying text).
vi
Figure 1 was produced by the authors using a
Smith+Crown dataset of 2017-18 digital token
offerings [2]. Sale-level data related to country of
registration and distribution type were independently
reviewed by the authors for approximately 90 percent
of the dollar-volume of token sale events included in
the dataset (see Appendix for more on methodology).
vii
These guidelines also clarified that open digital token
offerings not subject to direct MAS regulation are
nonetheless likely subject to certain Singapore laws
aimed at combatting money laundering and terrorism
financing [19].
viii
According to the Supreme Court’s 1946 “Howey
Test,” an “investment contract” – a type of security
– exists if 1) an investment is made in 2) a common
enterprise by 3) investors reasonably expecting to
earn profits 4) as a result of others’ managerial or
entrepreneurial efforts (see [3] [15]).
ix
Small jurisdictions with a sizable share of 2018 open
digital token offering dollar-volume include Gibraltar
(4 percent) and Estonia (4 percent) [2]. Singapore
and the U.K. accounted for 14 and 7 percent of 2018
dollar-volume, respectively [2].
x
For example, approximately half of the dollar-volume
of the U.K.’s nine successful Q3/Q4 2018 open digital
token offerings is attributable to two projects [2]. In
Singapore, on the other hand, there were over 20
successful open digital token offerings during this time,
the largest of which accounted for just 11 percent of
total offering dollar-volume [2].
xi
Figure estimated through conversations with U.S.based legal and regulatory experts.
xii
Bitcoin’s initial distribution was not an open digital
token offering (as defined in this research), but rather,
was an initial mining event.
xiii
For example, Filecoin’s private initial token sale.
xiv
Regulation CF offerings are each capped at $1.07
million and represented approximately $22 million of
token sale dollar-volume in 2017 through mid-2018
[34]. While in 2018 some projects reportedly applied to
sell tokens to the public using the SEC’s Regulation A+
exemption, the SEC did not approve any Regulation
A+ token sales.
xv
Figure 2 was produced by the authors using a
Smith+Crown dataset of 2017-18 digital token
offerings and through the authors’ classification of
token sale type as determined by a review of publiclyavailable information related to 46 Q3/Q4 2018
U.S.- and Singapore-registered digital token sales (see
Appendix for more on methodology) [2].
xvi
Figure 3 was produced by the authors using a
The JBBA | Volume 2 | Issue 2 | October 2019
Smith+Crown dataset of 2017-18 digital token
offerings and through the authors’ classification of
token sale type and the operational status of networks
or products associated with a token sale as determined
by a review of publicly-available information related
to 128 U.S.- and Singapore-registered digital token
projects that conducted open digital token offerings or
private initial token sales from Q3 2017 through Q2
2018 (see Appendix for more on methodology) [2].
xvii
Figure 4 was produced by the authors using a
Smith+Crown dataset of 2017-18 digital token
offerings (see Appendix for more on methodology) [2].
xviii
Figure 5 was produced by the authors using a
Smith+Crown dataset of 2017-18 digital token
offerings (see Appendix for more on methodology) [2].
xix
When possible, Smith+Crown used Etherscan to
examine the actual amounts raised by a token project.
As off-chain sales of digital tokens proliferated, this
method became less workable.
xx
One exception to this general rule was EOS’s large,
prolonged fundraising event, which was “grouped into
monthly amounts, with each month being treated as
a separate [sale event]” [3]. In a few instances, data
constraints forced the estimation of sale dates and/or
the consolidation of sale periods with unclear start or
end dates.
xxi
For approximately two percent of 2017-18 open
digital token offering dollar-volume, legal jurisdiction
was classified as unknown [2].
xxii
The authors independently reviewed publiclyavailable information on distribution type and country
of legal jurisdiction for approximately 90 percent of
2017-18 token sales by dollar-volume.
xxiii
These types of projects accounted for approximately
two percent of successful 2017-18 digital token
distribution dollar-volume [2].
Competing Interests:
None declared.
Ethical approval:
Not applicable.
Author’s contribution:
RWG1 and DLKC1,2 designed and coordinated this research and
prepared the manuscript in entirety.
Funding:
RWG1 wants to thank the Foundation of the Chamber of Digital
Commerce for a modest travel and conference expenditures grant that
funded research-related visits to Hong Kong, Singapore, and Tokyo.
Acknowledgements:
RWG and DLKC deeply thank Smith+Crown (Matt Chwierut,
Brian Lio, Alistair Simmonds, and Stuart Young) for providing data
that made this research possible. The authors are also appreciative of
helpful insights shared by LongHash (Emma Cui and Shi Khai Wei)
and CoinMarketCap (Carylyne Chan and Aaron Khoo), as well as
by Paul S. Atkins, Perianne Boring, Jehan Chu, Matthew Comstock,
Nizam Ismail, Amy Davine Kim, TM Lee, Daniel Liebau, Bobby
Ong, Remington Ong, Teong Jing Sim and Diego Zuluaga.
49
jbba
the
Peer-reviewed Research
OPEN ACCESS
ISSN Print: 2516-3949
https://doi.org/10.31585/jbba-2-2-(2)2019
Cryptocurrency Investing Examined
Jim Kyung-Soo Liew1, Richard Ziyuan Li2 , Tamás Budavári2, Avinash Sharma3
Johns Hopkins University, USA
1
Carey Business School
2
Department of Applied Mathematics and Statistics
3
Bioengineering and Biomedical Engineering
Correspondence: kliew1@jhu.edu
Received: 17 January 2019 Accepted: 28 March 2019 Published: 28 May 2019
Abstract
In this work we examine the largest 100 cryptocurrency returns ranging from 2015 to early 2018. We
concentrate our analysis on daily returns and find several interesting stylized facts. First, principal components
analysis reveals a complex daily return generating process. As we examine data in the most recent year, we
find that surprisingly more than one principal component appears to explain the cross-sectional variation.
Second, similar to hedge fund returns, cryptocurrency returns suffer from the “beta-in-the-tails” hidden risk.
Third, we find that predicting cryptocurrency movements with machine learning and artificial intelligence
algorithms is marginally attractive with variation in predictability power per crypto-currency. Fourth, lower
volatile cryptocurrencies are slightly more predictable than more volatile ones. Fifth, evidence exists that
efficacy of distinct information sets varies across machine learning algorithms, showing that predictability
may be much more complex given a set of machine learning algorithms. Finally, short-term predictability
is very tenuous, which suggests that near-term cryptocurrency markets are semi-strong form efficient and
therefore, day trading cryptocurrencies may be very challenging.
Keywords: AI, Bitcoin, Cryptocurrencies, Machine Learning, PCA, Beta-in-the-Tails
JEL Classifications: G12, G14, G17, G40, G
1. Introduction
Cryptocurrency is a digital asset designed to work as
a store of value and a medium of exchange1. As of
February 28th, 2018, the total market capitalization of
the cryptocurrency market stood at $448 billion and
consists of 1,524 types of currencies. Amongst the many
controversies surrounding cryptocurrencies, a popular
topic of debate is whether it should be classified as
a commodity, investment, property, currency or digital
currency. Bitcoin puts cryptocurrencies center stage
in the popular press and with the recent painful pull
back in early 2018, the interest in Bitcoins in particular
continues to hold. Bitcoins started 2017 at $998.33 and
grew 14x to finish the year at $14,156.40, as is shown in
Fig. 1. As of February 28th, the price was $10,559.20.
Bitcoin, the first successful cryptocurrency, was created
in January 2009, in the aftermath of the financial
crisis of 2008, by an unknown person or a group of
people under the Japanese name of Satoshi Nakamoto.
Bitcoin utilizes a technology called blockchain, which is
a combination of cryptography, consensus algorithms,
The JBBA | Volume 2 | Issue 2 | October 2019
Figure 1: Bitcoin price from Jan 1, 2017 to Feb 18, 2018
economic incentives and distributed ledger to secure
its transactions. While the technical discussion of
blockchain is beyond the scope of this work, this
technology has endowed Bitcoin with many important
characteristics, such as;
•
•
•
•
Decentralization,
Trusted network built upon potentially
untrustworthy nodes,
Transparency, and
Immutability history, etc.
Many cryptocurrencies were invented after Bitcoin, but
51
jbba
the
Bitcoin continues to be the most popular, as evidenced
by it having the largest market capitalization and
trading volume, shown in Table 1 below. Subsequently,
our investigation primarily focuses on Bitcoin prices in
this research.
of time before practitioners and academic researchers
begin using AI techniques to analyze cryptocurrency
markets. We hope our findings herein will serve as an
important contribution to this growing field.
1.3 Our Research Results
Market
Cap
($Billion)
Volume
(24 hrs $Billion)
Index
Name
Price
1
Bitcoin
$10,559.20
$178.4
$6.9
2
Ethereum
$869.63
$85.1
$2.0
3
Ripple
$0.921
$36.0
$0.33
4
Bitcoin
Cash
$1,223.85
$20.8
$0.38
5
Litecoin
$208.43
$11.6
$0.78
6
NEO
$135.27
$8.8
$0.33
7
Cardanol
$0.317
$8.7
$0.12
8
Stellar
$0.346
$8.2
$0.037
9
EOS
$8.64
$6.0
$0.38
10
IOTA
$1.89
$5.3
$0.044
Table 1: Top Ten Cryptocurrencies
(Source: CoinMarketCap.com, data as of February 28th, 2018.)
While participants of the Bitcoin blockchain can
transfer Bitcoins with each other directly, most
investors have to go to cryptocurrency exchanges if
they want to purchase Bitcoins with U.S. dollars or
other traditional currencies. While the quoted prices
from different exchanges can vary largely, arbitrage was
very difficult due to the lack of easy access to short
Bitcoins, until CBOE and CME introduced Bitcoin
futures in December 2017.
1.2 Artificial Intelligence (AI)
Similar to cryptocurrency, AI is another increasingly
intriguing technological development. AI represents
a broad range of techniques including machine
learning, deep learning, natural language processing,
etc. Its application is rapidly penetrating every aspect
of human society - e-commerce, autonomous vehicles,
image recognition, to name a few. A detailed discussion
of AI techniques and their application, unfortunately,
is beyond the scope of this paper.
Financial institutions are increasingly testing and
deploying AI techniques to obtain an edge in their
business, such as in trading. Money managers have
been employing thousands of quantitative experts to
develop sophisticated AI models for predicting prices,
identifying signals, monitoring sentiment, etc. While
the efficacy of these efforts is still debatable, AI models
and strategies are prevailing in every market (equity,
commodity, FX, etc.). It is, therefore, only a matter
The JBBA | Volume 2 | Issue 2 | October 2019
In this paper, we first analyze the top 100
cryptocurrencies using correlation analysis and
principal component analysis (PCA). Daily returns
reveal that in some period there exists a single dominant
component however, in the most recent prior year
there appears to be two components that help explain
the variation of the cryptocurrency returns. Next, we
compare cryptocurrencies with traditional assets. We
also perform Liew [2013]’s beta-in-the tail analysis to
examine potential hidden risks. We find some evidence
that similar to hedge funds, cryptocurrencies may
suffer from this hidden risk.
Finally, we conduct rolling prediction analysis on 57
cryptocurrencies with 11 AI algorithms. Our results
show that predictability may be difficult and there are
many heterogeneous effects here. Some information
sets perform better with some family of algorithms,
and larger cryptocurrencies with lower volatility maybe
more predictable than smaller cryptocurrency with
higher volatility.
The remainder of this paper is organized as follows:
Section 2 reviews prior literature, Section 3 presents
our data and preliminary analysis, Section 4 describes
the methodology, Section 5 provides the results and
Section 6 summarizes and concludes.
2. Literature Review
While there are many cases and projects about Bitcoin
price predictions online, scarce academic research
presently exists regarding Bitcoin price predictability.
We review the most important prior research in this
subject by aggregating them into three different groups.
The first group attempts to predict Bitcoin prices with
information about the Bitcoin blockchain network.
For example, Madan et al. [1] from Stanford use three
machine learning algorithms to predict the sign of daily
price change of Bitcoin based on data about the Bitcoin
blockchain network, including average confirmation
time, block size, hash rate, etc. They report a highest
accuracy of 98.7%. Another group of Stanford
researcher, Greaves et al. [2] perform similar analysis,
getting a classification (sign of hourly price change)
accuracy of 55%. In addition to information about the
blockchain network, McNally [3] adds daily open, high,
low, and close prices as explanatory variables, reporting
a classification (signs of daily price changes) accuracy
of 52%. El-Abdelouarti Alouaret [4] moves further
by including the S&P 500 index and EUR/USD rate,
52
jbba
the
as well as a variable named bitcoins days destroyed.
Similar to sentiment analysis, it also includes a variable
representing daily page view on the Wikipedia item
“Bitcoin”. It also uses vector autoregression and
recurrent neural network to conduct price prediction
instead of classification.
The second group of studies focus on the relationship
between social media data and Bitcoin performance.
For instance, Mai et al. [5] analyze Bitcoin-related user
posts from a forum and Twitter and demonstrate that
more bullish posts are associated with higher future
Bitcoin returns. They also conclude that the social media
effects on Bitcoin performance are driven by the “silent
majority”, and the impact of forum posts is larger than
that of tweets. Stenqvist et al. [6] try to predict Bitcoin
price (up/down) using sentiment analysis on Twitter,
and report that the sentiment change over a 30-minute
period is useful for predicting price movement of 2
hours later, resulting in an accuracy of 79%. Instead
of performing sentiment analysis on all social media
content posted, Kim et al. [7] extract the hottest topics
on a Bitcoin-related forum and define a time series
score to represent the “strength” of each topic. While
these scores are not significant in Granger causality
tests, a deep learning model with these scores as inputs
leads to prediction (for price and transaction volume)
accuracies ranging from 50%+ to 80%+. Interestingly,
Kaminskt [8], by analyzing Twitter posts, claims that
social media sentiments mirror the Bitcoin market
activity, rather than being predictive.
Instead of Bitcoin blockchain network data and social
media data, some papers examine the performance of
Bitcoin in other ways. Chu et al. [9] fits log returns in
fifteen popular parametric distributions in finance and
find that the generalized hyperbolic distribution is the
most appropriate. Balcilar et al. [10] perform causalityin-quantiles tests and point out that Bitcoin trading
volume can predict price returns but fail to predict
volatility. Indera et al. [11] use Multi-layer Perceptron
(MLP) to predict Bitcoin price based on historical
open, high, low, and close, as well as the moving average
technical indicators, reporting significant results (in
mean mean-squared error ).
The third group of research comprises of researchers
attempting to use every factor to predict Bitcoin
price. Georgoula et al. [12] and Garcia et al. [13]
contribute their work in this way. As they provide many
conclusions, we are not summarizing here.
3. Data and Preliminary Analysis
3.1 Cryptocurrency
As we mentioned above, there are 1,524 different
cryptocurrencies as of February 28, 2018, and they
are traded at many different exchanges (markets).
Fortunately, CoinMarketCap.com collects transaction
data of these cryptocurrencies from various exchanges
and publishes both up-to-date and historical data for
free, which can be obtained through their API. Taking
advantage of this resource, we scrap the historical
data of the top 100 cryptocurrencies, in terms of
market capitalizations as of February 18, 2018. Before
selecting the top 100, we remove those with relatively
short history1. Therefore, all selected cryptocurrencies
date back to at least January 1, 2017, and Fig. 2 shows
the number of cryptocurrencies under analysis over
time. The data includes close price, trading volume, and
market capitalization during the period of January 1,
2015 to December 31, 2017.
Figure 2: Number of cryptocurrencies under analysis (Jan 2015 - Feb 2018)
3.1.1 Price returns
We calculate daily, weekly, and monthly returns for
each cryptocurrency as (holding period returns):
𝑃𝑃𝑃𝑃
𝑃𝑃𝑃𝑃𝑡𝑡𝑡𝑡𝑡𝑡𝑡𝑡 −
𝑅𝑅𝑅𝑅𝑡𝑡𝑡𝑡
𝑅𝑅𝑅𝑅𝑡𝑡𝑡𝑡 = 𝑃𝑃𝑃𝑃𝑡𝑡𝑡𝑡−1 − 1
𝑃𝑃𝑃𝑃𝑡𝑡𝑡𝑡−1
We conduct normality tests on all returns series and
find that during Jan 1, 2015 to Feb 18, 2018, none of
𝑃𝑃𝑃𝑃𝑡𝑡𝑡𝑡 cryptocurrency is normal
the daily price returns𝑅𝑅𝑅𝑅of any
−
𝑡𝑡𝑡𝑡
𝑃𝑃𝑃𝑃𝑡𝑡𝑡𝑡−1
at the significance level of
95%. For weekly returns,
two cryptocurrencies yield normal returns. And ten
of them have normal monthly returns. Therefore,
we think it is more appropriate to use holding period
returns rather than log returns.
Table 2, Table 3, and Table 4 provide statistical summary
of price returns of Bitcoin (BTC), Ethereum (ETH),
and Ripple (XRP), respectively, which are the top 3
cryptocurrencies in terms of market capitalization, as
of February 18, 2018. All the three have an average
d as well
um
daily return of less than 1%
as nnsingle-digit
d
um
weekly returns.
Table 2: Statistics summary for price returns of Bitcoin
(Jan 2015 - Feb 2018)
Daily
Count
Mean
Standard Minimum Median
deviation
Maximum
1144
0.0039
0.0403
0.2525
-0.2115
0.0026
Weekly
163
0.0268 0.1053
-0.2834
0.0187 0.5097
ount
d
um
ountmeans the number ofd daily returns
um and etc. This note applies
Notes: the “Count”
to the next three tables.
ount
d
um
3
3
The JBBA | Volume 2 | Issue 2 | October 2019
53
jbba
the
overall cryptocurrency market during the past 60 (180)
days. Obviously, an interesting finding is the spike of
market correlation in the second half of 2017, which
was exactly accompanied with the rising hotness of
cryptocurrencies.
Table 3: Statistics summary for price returns of Ether
(Aug 2015 - Feb 2018)i
Count Mean
Standard Minimum Median
deviation
Maximum
Daily
926
0.0097
0.0798
-0.7280
-0.0002
0.5103
Weekly
132
0.0682
0.2514
-0.3394
0.0098
1.4227
Table 4: Statistics summary for price returns of Ripple
ount
(Jan 2015 - dFeb 2018)um
Count Mean
Standard Minimum Median
deviation
Maximum
Daily
1144
0.0065
0.0914
-0.4600
-0.0035
1.7937
Weekly
163
0.0494
0.2808
-0.3311
-0.0169
1.9992
3
Table 5 presents the average statistics summary for
the top 100 cryptocurrencies. On average, these
cryptocurrencies have an average history of 30
monthsii. Due to some volatile cryptocurrencies, the
average returns and average standard deviations are
larger than those for the top 3 shown above.
ount
d
um
d
um
Table 5: Average statistics summary for price returns of the Top 100
cryptocurrencies (Jan 2015 - Feb 2018)
ount
Count Mean
Standard Minimum Median
deviation
Maximum
Daily
962
0.0452
0.4701
-0.5580
-0.0009
9.0874
Weekly
137
0.1636
0.9940
-0.5356
0.0064
9.2084
Notes:
1.
2.
First, we calculate the statistics summary for each
cryptocurrency, including count, mean, standard deviation,
minimum, median, and maximum. Then, we calculate the
averages of these statistics of all cryptocurrencies.
Not all cryptocurrencies have history back to January 2015. Th
missing values are dropped before calculating the statistics.
Figure 3: Correlations of daily price returns between top 100
cryptocurrencies (Jan 2015 - Feb 2018)
Table 6: Statistical summary for correlations of returns between top 100
cryptocurrencies (Jan 2015 - Feb 2018)
Mean
Standard Minimum Median
deviation
Maximum
Daily
0.1210
0.0522
0.0052
0.1290
0.2289
Weekly
0.1569
0.0659
0.0036
0.1729
0.2855
Notes:
First, for each cryptocurrency, we calculate the mean of its correlations
with other cryptocurrencies. Then, we calculate these statistics of the
means of correlations.
3.1.2 Correlations
To reveal the relationship between various
cryptocurrencies, we calculate the correlations of
price returns between the top 100 of them. Fig. 3
present the heatmaps of the correlations of daily
returns. And Table 6 provides statistics summary for
the correlations across all top 100. Obviously, most
of the cryptocurrencies are positively correlated
and correlations are getting higher when the time
frame becomes larger. Another interesting finding
is that correlations between large market-cap
cryptocurrencies are higher than correlations between
smaller market-caps.iii Therefore, we can conclude that
most cryptocurrencies are moving in herds with lower
double-digit correlations, and this phenomenon is
stronger between large market-caps.
Finally, to find out how correlations among
cryptocurrencies develop over time, we perform a
rolling analysis as is shown in Fig. 4. On each day, we
calculate the correlations based on daily returns of the
preceding 60 (180) days, and then we use the arithmetic
mean as the average correlation for that day. That said,
the statistic represents the level of correlation of the
The JBBA | Volume 2 | Issue 2 | October 2019
Figure 4: Rolling average correlation (60-days and 180-days, Jan 2015 Feb 2018)
To have a closer look at Bitcoin, we summarize the
statistics of its correlations of price returns with other
cryptocurrencies in Table 7. On average, Bitcoin has
a correlation of price returns (daily, weekly) of about
0.20 with other cryptocurrencies. In addition, Table
8 lists the most and least correlated cryptocurrencies
with Bitcoins. One interesting cryptocurrency stood
out upon a quick inspection - Litecoin (LTC) is highly
positively correlated with Bitcoin in both time frames.
We also examine the autocorrelation of Bitcoin,
as is shown in Fig. 5. The autocorrelations for daily
returns fall between -0.05 and 0.05, implying a low
4
autocorrelation nature.
54
jbba
the
Table 7: Statistics summary for correlations of between Bitcoins and
other cryptocurrencies (Jan 2015 - Feb 2018)
Mean
Standard Minimum Median
deviation
Maximum
Daily
0.2211
0.1158
-0.0140
0.2225
0.5035
Weekly
0.1897
0.1382
-0.1135
0.1962
0.4976
Notes: These statistics are calculated based on the correlations of price
returns between Bitcoins and the other 99 cryptocurrencies.
Table 8: Most and least correlated cryptocurrencies with Bitcoins
(Jan 2015 - Feb 2018)
Daily returns
Most correlated
Least correlated
Weekly returns
Symbol
Correlation
Symbol
Correlation
PPC
0.5035
SBD
0.4976
LTC
0.5006
LTC
0.4706
DOGE
0.4740
GOLOS
0.4463
NMC
0.4678
EMC2
0.4315
WAVES
0.4401
NMC
0.4281
PASC
-0.0140
ZOI
-0.1135
PURA
0.0029
GAME
-0.0991
NYC
0.0244
PIVX
-0.0915
MOON
0.0248
EMC
-0.0829
EXP
0.0306
CRW
-0.0681
from 2017 to February 2018 the daily returns appear
to differ in their structure. Figure 8 displays that the
variation explained by the second principal component
gains significantly as the first principal component fall
to less than 60%.
Clearly, 2017 was a banner year for cryptocurrency
and the addition of more retail investors could be one
of the explanations of why this period may have a
different underlying structure in the return generating
process compared to the two other periods. Retail
investors became more heavily involved purchasing
cryptocurrencies as evidenced by CoinBase having
more accounts than Charles Schwab in November
27, 2017iv. This changing investor base could possibly
bring in more of a herding and momentum behavior
if these retail investors are susceptible to known biases
similar to those affecting stock retail investors.
Figure 6: Explained variance ratios for PCA components
(58 cryptocurrencies, Jan 2015 - Feb 2018)
Notes: the ranks are based on magnitudes of correlations.
Figure 7: Explained variance ratios for PCA components
(73 cryptocurrencies, Jan 2016 - Feb 2018)
Figure 5: Autocorrelation function of Bitcoin daily price returns (Jan
2015 - Feb 2018)
Notes: the lags range from 1 to 40.
3.1.3 Principal Component Analysis (PCA)
To uncover the common drivers of price returns, we
employ a popular dimensionality reduction technique
- PCA. The starting time of each cryptocurrency
varies, thus, to avoid artificially creating biasedness
by filling backward on the missing leading values, we
select three subsets of time for our PCA analysis and
only employ overlapping series. First, we select the 59
cryptocurrencies which have full history back to January
1, 2015. Second, we select the 74 cryptocurrencies with
full history back to January 1, 2016. Finally, we select
the 100 cryptocurrencies which have returns back to
January 1, 2017. We perform PCA for our three periods
employing daily price returns.
Figure 6, Figure 7, and Figure 8 present the results
for 2015 to Jan 2018, 2016 to Jan 2018, and 2017 to
Jan 2018, respectively. In the first and second case, the
first principal component captures the majority of the
variance, with less variation explained by the other four
principal components. In the third case, the period
The JBBA | Volume 2 | Issue 2 | October 2019
Figure 8: Explained variance ratios for PCA components
(100 cryptocurrencies, Jan 2017 - Feb 2018)
3.2 Traditional assets
Recent literature [14] shows that Bitcoin provides
diversification to portfolio comprised of traditional
assets. We dig in and investigate the cross-market
relationship between the top 100 cryptocurrencies and
5
traditional assets. Daily prices of following assets are
downloaded from Bloomberg Terminal:
●
●
●
S&P 500 index (SPX Index): It is a
capitalization-weighted index of 500 stocks
trading in the U.S. stock market.
MSCI World Index (MXWO Index): It is a
free-float weighted equity index covering
stocks trading in developed markets.
MSCI Emerging Markets Index (MXEF
55
jbba
the
Index): It is a free-float weighted equity
index covering large and mid-cap stocks
trading in emerging markets.
US Dollar Index: a measure of the value
of the U.S. dollar relative to the value of a
basket of currencies of the majority of the
U.S.'s most significant trading partners.
Gold spot price (in US$)
Bloomberg Commodity Index (BCOM
Index): It is an index reflecting commodity
futures price movement.
VIX Index: The measure of volatility
implied by S&P 500 index options,
calculated and published by CBOE.
●
●
●
●
●
●
●
●
●
●
●
●
●
●
● Table
9 presents the correlations between Bitcoin,
●
●
other cryptocurrencies, and traditional assets,
calculated in terms of daily returns. Obviously, Bitcoin
is barely correlated to any traditional assets at the
daily level (absolute correlations < 0.1). It exhibits a
slightly positive correlation to S&P 500, MSCI, USD,
Gold, and Commo, while demonstrating a negative
correlation to Emg and VIX. Not surprisingly Bitcoin
is positively associated with the first PCA component
and very highly correlated to the market capitalization
weighted cryptocurrency returns.
Table 9: Correlations between daily returns of cryptocurrencies and
traditional assets (Jan 2015 - Feb 2018)
BTC
VW
SP500
MSCI
Emg
USD
Gold
Commo
VIX
BTC
1.0000
0.9416
0.0441
0.0232
-0.0212
0.0134
0.0419
0.0351
-0.0921
VW
0.9416
1.0000
0.0538
0.0316
-0.0204
-0.0049
0.0526
0.0359
-0.0975
SP500
0.0441
0.0538
1.0000
0.9093
0.4480
0.0831
-0.1674
0.2967
-0.7880
MSCI
0.0232
0.0316
0.9093
1.0000
0.6587
-0.0413
-0.1262
0.3836
-0.7283
Emg
-0.0212 -0.0204 0.4480
0.6587
1.0000
-0.0426
-0.0053
0.3641
-0.3848
USD
0.0134
-0.0049 0.0831
-0.0413
-0.0426
1.0000
-0.4070
-0.2427
-0.0828
Gold
0.0419
0.0526
-0.1674
-0.1262
-0.0053
-0.4070
1.0000
0.2441
0.1365
Commo
0.0351
0.0359
0.2967
0.3836
0.3641
-0.2427
0.2441
1.0000
-0.2224
-0.7283
-0.3848
-0.0828
0.1365
-0.2224
1.0000
VIX
-0.0921 -0.0975 -0.7880
beta-in-the-tail risk.
Upon visual inspection we document the increasing
betas in down S&P 500 daily return periods. We argue
that beta-in-the-tail is a significant hidden risk for
cryptocurrency investors when employing daily returns.
The methodology for daily beta-in-the-tail analysis
follows: First, order all the daily returns on the S&P
500 from least to greatest. Associated to each S&P
500-day period we link both the Bitcoin return and
MarketCap Weighted Index return for that day. Next,
we anchor the worst daily returns for the S&P 500 and
use thirty days of returns to run our regressions. That
is, we estimate the beta associated with the worst thirty
days in our sample period. At this point, it is important
to note that the time dimension has been compromised
with this sorting of the daily returns.
The regression is the crypto-returns regressed on the
S&P 500 returns. Assuming that the risk-free daily
returns are zero yields the CAPM’s beta of Sharpe
[17] and Litner [18] for the given cryptocurrency
index. By anchoring the worst return day for the S&P
500 and expanding the window of daily returns we
plot the slope coefficients with inclusion of another
daily return. When the window has been expanded to
include all the daily returns then the final regression
corresponds to the beta for the whole period.
The Betas are reported in the left y-axis and the average
daily returns for the window period is reported in the
right y-axis on the black dashed line. Standard deviation
bands surround the beta estimates. Notice that as more
observations are included the standard deviation of
the beta estimates reduces. The beta-in-the-tails based
on daily returns reach above 1.0 compared this to the
whole period beta of close to zero for Bitcoin and VW
Index, respectively, as seen on the furthest left bottom
corner of Fig. 9.
Notes: “VW” is the market cap weighted price returns. “MSCI” is the MSCI
o
51
59
developed
market
index. “Emg” is the MSCI emerging market index. “Commo” is
the Bloomberg Commodity Index.
921
975
3.3 Beta-in-the-Tails Analysis (BTA)
In this section we estimate the potential hidden risks in
the cryptocurrency markets. In particular, we examine
the stability of their betas for Bitcoin and the VW
index with respect to the market, which we employ
the S&P 500 as a proxy. Edwards and Caglayan [15]
document changes in hedge fund correlation in bull
and bear markets. Liew [16] introduces the beta-inthe-tail analysis for hedge funds and documents the
vanishing diversification benefits as a hidden risk
for hedge fund investors. In down periods the beta
associated to hedge fund increases and thus decreasing
the perceived diversification benefits. Similarly, we find
such an occurrence for cryptocurrencies and warn
potential investors to be vigilant with regards to the
The JBBA | Volume 2 | Issue 2 | October 2019
Figure 9: Beta in the Tails (daily)
Notes: Calculated based on daily returns from April 2013 to Feb 2018.
Given that cryptocurrencies trade seven days a week
and twenty-four hours a day in contrast to stocks 6
which typically trade only five days a week and six and
a half hours a day, we repeat the analysis excluding
the weekend in Fig. 10, Beta in the Tail Excluding
the Weekends. We arrive at a similar pattern with an
increase in the beta in down S&P 500 days. Beta-in-the56
jbba
the
regularization
with
α =α 1.0.
regularization
with
= 1.0.
regularization
regularization with
with αα =
= 1.0.
1.0.
2 2
‖𝑋𝑋𝑋𝑋‖𝑋𝑋𝑋𝑋
− 𝑦𝑦𝑦𝑦‖
∗ ‖𝜔𝜔𝜔𝜔‖
𝜔𝜔𝜔𝜔 −
1 1
2 2 𝛼𝛼𝛼𝛼 ∗𝛼𝛼𝛼𝛼‖𝜔𝜔𝜔𝜔‖
𝜔𝜔𝜔𝜔 𝑦𝑦𝑦𝑦‖
𝜔𝜔𝜔𝜔 with
𝜔𝜔𝜔𝜔 ∗ 𝑛𝑛𝑛𝑛
2
𝑛𝑛𝑛𝑛 1.0.−
regularization
α∗=
‖𝑋𝑋𝑋𝑋
𝑦𝑦𝑦𝑦‖
𝛼𝛼𝛼𝛼
∗∗ ‖𝜔𝜔𝜔𝜔‖
ElasticNet
(EN):
𝜔𝜔𝜔𝜔 − 𝑦𝑦𝑦𝑦‖2
1
2
‖𝑋𝑋𝑋𝑋
‖𝜔𝜔𝜔𝜔‖
𝛼𝛼𝛼𝛼
𝜔𝜔𝜔𝜔
𝜔𝜔𝜔𝜔
1
2
∗∗ 𝑛𝑛𝑛𝑛
𝜔𝜔𝜔𝜔
𝑛𝑛𝑛𝑛
tails appears robust to non-trading weekdays.
EN [19] is a linear
model
performs
regression
‖𝑋𝑋𝑋𝑋 −
𝑦𝑦𝑦𝑦‖22 that
𝛼𝛼𝛼𝛼 ∗ ‖𝜔𝜔𝜔𝜔‖
1
𝜔𝜔𝜔𝜔
𝑛𝑛𝑛𝑛 𝜔𝜔𝜔𝜔
with both L1∗ and
L2 regularization. This gives it the
property of both LASSO and ridge regression, and the
regularization
with α =
objective
function
is 1.0.
as below. We use the scikit-learn
learn
default
selection
of α =α 1.0.
learn
default
= 1.0.
default selection
of
α
= selection
1.0. of of
learn default selection
α = 1.0.
Figure 10: Beta in the Tails (daily, excluding weekends)
Notes: Calculated based on daily returns from April 2013 to Feb 2018.
4. Methodology - Rolling Prediction Analysis
learn default selection2 of α = 1.0.
‖𝑋𝑋𝑋𝑋𝜔𝜔𝜔𝜔 −2 𝑦𝑦𝑦𝑦‖2 𝛼𝛼𝛼𝛼 ∗ ‖𝜔𝜔𝜔𝜔‖1
1
2 𝛼𝛼𝛼𝛼 ∗ 𝜌𝜌𝜌𝜌 ∗ ‖𝜔𝜔𝜔𝜔‖
∗ 𝑛𝑛𝑛𝑛
min 𝜔𝜔𝜔𝜔 ‖𝑋𝑋𝑋𝑋
‖𝑋𝑋𝑋𝑋
− 𝑦𝑦𝑦𝑦‖
𝛼𝛼𝛼𝛼 ∗ 𝜌𝜌𝜌𝜌 ∗ ‖𝜔𝜔𝜔𝜔‖
𝜔𝜔𝜔𝜔 −
1 1
2+
𝜔𝜔𝜔𝜔 𝑦𝑦𝑦𝑦‖
2
𝜔𝜔𝜔𝜔 𝜔𝜔𝜔𝜔2
∗
𝑛𝑛𝑛𝑛
2
𝑛𝑛𝑛𝑛 −
learn∗default
selection
of
α∗=‖𝜔𝜔𝜔𝜔‖
1.0.1
‖𝑋𝑋𝑋𝑋
𝑦𝑦𝑦𝑦‖
𝛼𝛼𝛼𝛼
∗∗ 𝜌𝜌𝜌𝜌
2
𝜔𝜔𝜔𝜔 − 𝑦𝑦𝑦𝑦‖𝛼𝛼𝛼𝛼2 ∗𝛼𝛼𝛼𝛼(1
‖𝑋𝑋𝑋𝑋
‖𝜔𝜔𝜔𝜔‖
𝛼𝛼𝛼𝛼
𝜌𝜌𝜌𝜌
∗
−
𝜌𝜌𝜌𝜌)
∗
−
𝜌𝜌𝜌𝜌
𝜔𝜔𝜔𝜔
1
∗∗ 𝑛𝑛𝑛𝑛
𝜔𝜔𝜔𝜔
𝑛𝑛𝑛𝑛 𝜔𝜔𝜔𝜔 + 𝛼𝛼𝛼𝛼2 ∗ − 𝜌𝜌𝜌𝜌 ‖𝜔𝜔𝜔𝜔‖
‖𝜔𝜔𝜔𝜔‖
2
𝛼𝛼𝛼𝛼 ∗ 2 − 𝜌𝜌𝜌𝜌 ‖𝜔𝜔𝜔𝜔‖ 2
‖𝑋𝑋𝑋𝑋 − 𝑦𝑦𝑦𝑦‖22 𝛼𝛼𝛼𝛼 ∗ 𝜌𝜌𝜌𝜌 ∗ ‖𝜔𝜔𝜔𝜔‖
‖𝜔𝜔𝜔𝜔‖221
𝜔𝜔𝜔𝜔
∗ 𝑛𝑛𝑛𝑛 𝜔𝜔𝜔𝜔
𝛼𝛼𝛼𝛼 ∗ − (SGD):
𝜌𝜌𝜌𝜌
Stochastic Gradient Descent
‖𝜔𝜔𝜔𝜔‖2
(3)
default selection
of αto=fit
1.0.linear models.
SGD [19] is learn
an efficiency
method
In this section, we firstly give a brief introduction to It searches for minima or maxima through iterations.
with
α =α‖𝑋𝑋𝑋𝑋
0.0001.
−default
𝑦𝑦𝑦𝑦‖22 𝛼𝛼𝛼𝛼
∗ 𝜌𝜌𝜌𝜌 ∗ ‖𝜔𝜔𝜔𝜔‖1 squared loss
regularization
with
=𝜔𝜔𝜔𝜔0.0001.
the 11 machine learning algorithms we tested. Next, we regularization
We
use the
scikit-learn
parameters:
𝜔𝜔𝜔𝜔
∗α𝑛𝑛𝑛𝑛= 0.0001.
regularization
with
withL2
α=
0.0001. 𝛼𝛼𝛼𝛼 ∗ with
describe the way we roll the prediction analysis. Finally, regularization
function and
regularization
− 𝜌𝜌𝜌𝜌 α = 0.0001.
‖𝜔𝜔𝜔𝜔‖2
4)
2 2
we present our data.
‖𝑋𝑋𝑋𝑋‖𝑋𝑋𝑋𝑋
𝛼𝛼𝛼𝛼 ∗𝛼𝛼𝛼𝛼‖𝜔𝜔𝜔𝜔‖
− 𝑦𝑦𝑦𝑦‖
∗ ‖𝜔𝜔𝜔𝜔‖
𝜔𝜔𝜔𝜔 − 𝑦𝑦𝑦𝑦‖
2
2
4.1 Algorithms
In this subsection, we introduced the 11 machine
learning algorithms. Our problem can be easily
described with linear models – we have a set of
variables (x, a matrix with each column being a variable
and each row being value for the corresponding day)
such as historical returns, volatility and etc., and a target
variable (y, a column vector); and we want to train a
model that predicts y with out of sample input x.
There are three strands of algorithms in our analysis:
1) linear models, including LASSO, ElasticNet,
Stochastic Gradient Descent, and Bayesian Regression;
2) tree-based
models, including Decision Tree, Extra
𝑛𝑛𝑛𝑛 𝑛𝑛𝑛𝑛
Tree Random
Forest, AdaBoost, and Gradient Tree
𝑛𝑛𝑛𝑛
∑
𝐿𝐿𝐿𝐿(𝑦𝑦𝑦𝑦
𝜔𝜔𝜔𝜔 𝜔𝜔𝜔𝜔
𝑛𝑛𝑛𝑛 ∑
𝐿𝐿𝐿𝐿(𝑦𝑦𝑦𝑦
− 𝑥𝑥𝑥𝑥
𝑓𝑓𝑓𝑓𝑖𝑖𝑖𝑖 𝑥𝑥𝑥𝑥)𝑖𝑖𝑖𝑖 ) 𝛼𝛼𝛼𝛼 ∗𝛼𝛼𝛼𝛼𝑅𝑅𝑅𝑅∗ 𝑅𝑅𝑅𝑅
𝑖𝑖𝑖𝑖 −𝑖𝑖𝑖𝑖 𝑓𝑓𝑓𝑓
𝜔𝜔𝜔𝜔 𝜔𝜔𝜔𝜔𝑛𝑛𝑛𝑛 3)
Boosting;
other
models,
including
KNN, Support
𝑛𝑛𝑛𝑛
∑
𝑖𝑖𝑖𝑖=1 𝐿𝐿𝐿𝐿(𝑦𝑦𝑦𝑦
𝐿𝐿𝐿𝐿(𝑦𝑦𝑦𝑦𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖 −
− 𝑓𝑓𝑓𝑓
𝑓𝑓𝑓𝑓 𝑥𝑥𝑥𝑥
𝑥𝑥𝑥𝑥𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖 )) 𝛼𝛼𝛼𝛼
𝛼𝛼𝛼𝛼 ∗∗ 𝑅𝑅𝑅𝑅
𝑅𝑅𝑅𝑅 𝜔𝜔𝜔𝜔
𝜔𝜔𝜔𝜔
𝜔𝜔𝜔𝜔 𝑛𝑛𝑛𝑛 ∑𝑖𝑖𝑖𝑖=1
Vector𝜔𝜔𝜔𝜔Machine,
and
Multi-layer
perceptron.
We briefly
𝑛𝑛𝑛𝑛 𝑖𝑖𝑖𝑖=1
𝑛𝑛𝑛𝑛
𝑖𝑖𝑖𝑖=1
introduced∑
each
of
the
algorithms
as
below.
𝐿𝐿𝐿𝐿(𝑦𝑦𝑦𝑦 − 𝑓𝑓𝑓𝑓 𝑥𝑥𝑥𝑥 ) 𝛼𝛼𝛼𝛼 ∗ 𝑅𝑅𝑅𝑅 𝜔𝜔𝜔𝜔
𝜔𝜔𝜔𝜔
𝑛𝑛𝑛𝑛
𝑖𝑖𝑖𝑖
𝑖𝑖𝑖𝑖=1
𝑖𝑖𝑖𝑖
A typical objective function of linear models is as
below:
𝑛𝑛𝑛𝑛
1
min ∑ 𝐿𝐿𝐿𝐿(𝑦𝑦𝑦𝑦𝑖𝑖𝑖𝑖 − 𝑓𝑓𝑓𝑓(𝑥𝑥𝑥𝑥𝑖𝑖𝑖𝑖 )) + 𝛼𝛼𝛼𝛼 ∗ 𝑅𝑅𝑅𝑅(𝜔𝜔𝜔𝜔)
𝜔𝜔𝜔𝜔 𝑛𝑛𝑛𝑛
(1)
𝑖𝑖𝑖𝑖=1
where L is loss function, R is regularization term, f is
the fitted function.
Least Absolute Shrinkage and Selection Operator
(LASSO):
LASSO [19] is a linear model that performs both
variable selection and regularization. In contrast to
simple linear regression, its objective function is as
below. We use the scikit-learn default parameters:
squared loss
regularization
withfunction
α = 1.0. and L2 regularization with α =
1.0.
min
𝜔𝜔𝜔𝜔
1
‖𝑋𝑋𝑋𝑋 − 𝑦𝑦𝑦𝑦‖22 + 𝛼𝛼𝛼𝛼 ∗ ‖𝜔𝜔𝜔𝜔‖1
2 ∗ 𝑛𝑛𝑛𝑛 𝜔𝜔𝜔𝜔
The JBBA | Volume 2 | Issue 2 | October 2019
(2)
1 𝑛𝑛𝑛𝑛 𝜔𝜔𝜔𝜔 2 2
𝜔𝜔𝜔𝜔
𝜔𝜔𝜔𝜔𝑛𝑛𝑛𝑛 ‖𝑋𝑋𝑋𝑋
regularizationmin
with
α =𝜔𝜔𝜔𝜔0.0001.
‖𝜔𝜔𝜔𝜔‖2
− 𝑦𝑦𝑦𝑦‖22 + 𝛼𝛼𝛼𝛼
𝛼𝛼𝛼𝛼 ∗∗ ‖𝜔𝜔𝜔𝜔‖
𝜔𝜔𝜔𝜔 𝑛𝑛𝑛𝑛 ‖𝑋𝑋𝑋𝑋𝜔𝜔𝜔𝜔 − 𝑦𝑦𝑦𝑦‖2
2
𝜔𝜔𝜔𝜔 𝑛𝑛𝑛𝑛
2
(4)
2
Bayesian Regression
‖𝑋𝑋𝑋𝑋𝜔𝜔𝜔𝜔 − 𝑦𝑦𝑦𝑦‖(BR):
𝛼𝛼𝛼𝛼 ∗ ‖𝜔𝜔𝜔𝜔‖2
2
𝜔𝜔𝜔𝜔
𝑛𝑛𝑛𝑛
BR [19] provides another way of performing linear
regularizationwhere
with αlinear
= 0.0001.
regression,
model can be written as below:
5)
𝑦𝑦𝑦𝑦𝑖𝑖𝑖𝑖 𝑦𝑦𝑦𝑦𝑖𝑖𝑖𝑖 𝛼𝛼𝛼𝛼 𝛼𝛼𝛼𝛼 𝛽𝛽𝛽𝛽 ∗𝛽𝛽𝛽𝛽𝑥𝑥𝑥𝑥∗𝑖𝑖𝑖𝑖 𝑥𝑥𝑥𝑥
𝑤𝑤𝑤𝑤𝑖𝑖𝑖𝑖
𝑡𝑡𝑡𝑡ℎ 𝑡𝑡𝑡𝑡𝑦𝑦𝑦𝑦ℎ𝑖𝑖𝑖𝑖 𝑦𝑦𝑦𝑦𝑖𝑖𝑖𝑖𝑁𝑁𝑁𝑁 𝑁𝑁𝑁𝑁
𝜇𝜇𝜇𝜇𝑖𝑖𝑖𝑖 𝜇𝜇𝜇𝜇𝜎𝜎𝜎𝜎𝑖𝑖𝑖𝑖 𝜎𝜎𝜎𝜎
𝑖𝑖𝑖𝑖 𝑤𝑤𝑤𝑤𝑖𝑖𝑖𝑖
5)
𝑦𝑦𝑦𝑦
𝛼𝛼𝛼𝛼
𝛽𝛽𝛽𝛽
∗
𝑥𝑥𝑥𝑥
𝑤𝑤𝑤𝑤𝑖𝑖𝑖𝑖
𝑡𝑡𝑡𝑡
ℎ
𝑦𝑦𝑦𝑦
𝑁𝑁𝑁𝑁
𝜇𝜇𝜇𝜇
𝜎𝜎𝜎𝜎
𝑖𝑖𝑖𝑖
𝑖𝑖𝑖𝑖
𝑖𝑖𝑖𝑖
𝑖𝑖𝑖𝑖
(5)
𝑦𝑦𝑦𝑦𝑖𝑖𝑖𝑖 = 𝛼𝛼𝛼𝛼 + 𝛽𝛽𝛽𝛽 ∗ 𝑥𝑥𝑥𝑥𝑖𝑖𝑖𝑖 𝑤𝑤𝑤𝑤𝑖𝑖𝑖𝑖𝑡𝑡𝑡𝑡ℎ2𝑦𝑦𝑦𝑦𝑖𝑖𝑖𝑖 ~𝑁𝑁𝑁𝑁(𝜇𝜇𝜇𝜇𝑖𝑖𝑖𝑖 , 𝜎𝜎𝜎𝜎)
‖𝑋𝑋𝑋𝑋𝜔𝜔𝜔𝜔distribution
−distribution
𝑦𝑦𝑦𝑦‖2 𝛼𝛼𝛼𝛼 with
∗ ‖𝜔𝜔𝜔𝜔‖
2mean
That
is, yis,follows
mean
μ and
σ, σ,
That
y follows
a𝑛𝑛𝑛𝑛normal
with
μ and
𝜔𝜔𝜔𝜔a normal
That
follows
a normal
distribution
with
mean
and
σ,
μ
isμ aisis,
linear
parameters
α 𝜇𝜇𝜇𝜇and
this
way,
thethe
ayylinear
function
α and
β.μ
In
this
way,
That
is,
follows
μwith
and
σ,
𝑦𝑦𝑦𝑦𝑖𝑖𝑖𝑖yfunction
𝛼𝛼𝛼𝛼a normal
𝛽𝛽𝛽𝛽with
∗awith
𝑥𝑥𝑥𝑥normal
𝑤𝑤𝑤𝑤𝑖𝑖𝑖𝑖parameters
𝑡𝑡𝑡𝑡ℎ 𝑦𝑦𝑦𝑦𝑖𝑖𝑖𝑖distribution
𝑁𝑁𝑁𝑁with
𝜎𝜎𝜎𝜎β. In
𝑖𝑖𝑖𝑖distribution
𝑖𝑖𝑖𝑖 mean
That
is,
follows
mean
μ
μ
is
a
linear
function
with
parameters
α
and
β.
In
this
way,
μ is a linear function with parameters α and β. In this way, the
the
and σ, while μ is a linear function with parameters α
That is, y follows a normal distribution with mean μ and σ,
and
β. Infunction
this way,
the
model can
beβ.estimated
μ is
a linear
parameters
α and
In this way,using
the
𝑛𝑛𝑛𝑛with
𝑛𝑛𝑛𝑛
6)
maximum likelihood
function
instead
of minimizing
∏∏
𝑦𝑦𝑦𝑦
𝛼𝛼𝛼𝛼
𝛽𝛽𝛽𝛽
∗
𝑥𝑥𝑥𝑥
𝜎𝜎𝜎𝜎
𝑛𝑛𝑛𝑛 𝑁𝑁𝑁𝑁 𝑁𝑁𝑁𝑁
𝑦𝑦𝑦𝑦
𝛼𝛼𝛼𝛼
𝛽𝛽𝛽𝛽
∗
𝑥𝑥𝑥𝑥
𝜎𝜎𝜎𝜎
𝑖𝑖𝑖𝑖
𝑖𝑖𝑖𝑖
𝑖𝑖𝑖𝑖
𝑖𝑖𝑖𝑖
6)
𝛽𝛽𝛽𝛽𝛼𝛼𝛼𝛼𝜎𝜎𝜎𝜎𝛽𝛽𝛽𝛽 𝜎𝜎𝜎𝜎 𝑛𝑛𝑛𝑛
6)
squared 𝛼𝛼𝛼𝛼errors:
∏𝑖𝑖𝑖𝑖=1𝑖𝑖𝑖𝑖=1
𝑁𝑁𝑁𝑁 𝑦𝑦𝑦𝑦𝑖𝑖𝑖𝑖 𝛼𝛼𝛼𝛼 𝛽𝛽𝛽𝛽
𝛽𝛽𝛽𝛽 ∗∗ 𝑥𝑥𝑥𝑥
𝑥𝑥𝑥𝑥𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖 𝜎𝜎𝜎𝜎
𝜎𝜎𝜎𝜎
𝛼𝛼𝛼𝛼 𝛽𝛽𝛽𝛽 𝜎𝜎𝜎𝜎 ∏𝑖𝑖𝑖𝑖=1𝑁𝑁𝑁𝑁 𝑦𝑦𝑦𝑦𝑖𝑖𝑖𝑖 𝛼𝛼𝛼𝛼
𝛼𝛼𝛼𝛼 𝑦𝑦𝑦𝑦
𝛽𝛽𝛽𝛽𝑖𝑖𝑖𝑖𝜎𝜎𝜎𝜎 𝛼𝛼𝛼𝛼 𝑖𝑖𝑖𝑖=1𝛽𝛽𝛽𝛽 ∗ 𝑥𝑥𝑥𝑥𝑖𝑖𝑖𝑖 𝑤𝑤𝑤𝑤𝑖𝑖𝑖𝑖𝑡𝑡𝑡𝑡ℎ 𝑦𝑦𝑦𝑦𝑖𝑖𝑖𝑖 𝑁𝑁𝑁𝑁 𝜇𝜇𝜇𝜇𝑖𝑖𝑖𝑖 𝜎𝜎𝜎𝜎
𝑛𝑛𝑛𝑛
(6)
max ∏ 𝑁𝑁𝑁𝑁(𝑦𝑦𝑦𝑦𝑖𝑖𝑖𝑖 ; 𝛼𝛼𝛼𝛼 + 𝛽𝛽𝛽𝛽 ∗ 𝑥𝑥𝑥𝑥𝑖𝑖𝑖𝑖 , 𝜎𝜎𝜎𝜎 )
That is, y𝛼𝛼𝛼𝛼,𝛽𝛽𝛽𝛽,𝜎𝜎𝜎𝜎
follows a𝑖𝑖𝑖𝑖=1
normal distribution with mean μ and σ,
μ is a linear function with parameters α and β. In this way, the
Decision Tree (DT):
𝑛𝑛𝑛𝑛
DT [19] is a non-parametric
method that can be used
∏ 𝑁𝑁𝑁𝑁 𝑦𝑦𝑦𝑦𝑖𝑖𝑖𝑖 𝛼𝛼𝛼𝛼 𝛽𝛽𝛽𝛽 ∗ 𝑥𝑥𝑥𝑥𝑖𝑖𝑖𝑖 𝜎𝜎𝜎𝜎
𝛽𝛽𝛽𝛽 𝜎𝜎𝜎𝜎
𝑖𝑖𝑖𝑖=1
for both 𝛼𝛼𝛼𝛼classification
and regression. The tree is
built for classifying or predicting test points based on
several rules. For classification problems, the leafs of
the tree are the classification labels, and for regression
problems, the leafs are continuous values. We use the
default parameters provided by scikit-learn: using mean
square error as splitting criterion, and without max
depth of trees.
Extra Tree Random Forest (ETRF):
Random forest [19] is an ensemble method built on
many trees, and each tree is built through training on
a sample of the entire train set with replacement. In
addition, when splitting a node during the construction
of trees, the best split is measured among a random
subset of features rather than all features. This
randomness leads to lower variance and larger bias. On
the other hand, ETRF moves even further regarding
57
jbba
the
randomness in splitting the nodes – splitting thresholds
are randomly assigned instead of searching for the
most discriminative thresholds. We use the default
parameters provided by scikit-learn: 10 trees without
max depth of trees and using mean square error as
splitting criterion.
Adaptive Boosting (AdaBoost):
AdaBoost [19] is an ensemble algorithm that fits a
sequence of relatively weak models with repeatedly
modified data. More specifically, it firstly trains on
the original train set and assesses the errors. Then it
modifies the train set by assigning more weights to
poorly modeled points. The processes are repeated for
multiple times. Decision Tree is usually used as the base
model in AdaBoost. We use the default parameters
provided by scikit-learn: 50 Decision Tree models as
base estimators.
Gradient Tree Boosting (GTB):
Gradient Boosting [19] is another ensemble algorithm
that also fits a sequence of relatively weak models with
repeatedly modified data. More specifically, it firstly
trains on the train set and the original predicted targets.
Then it modifies the predicted targets to be certain type
of residuals between the true values and the predicted
(trained) values. The processes are repeated for multiple
times. GTB is the combination of Decision Tree and
Gradient Boosting. We use the default parameters
provided by scikit-learn: 100 Decision Tree models as
base estimators and without max depth.
K-nearest Neighbor (KNN):
b, which can be found by solving a constrained
optimization problem:
min‖𝜔𝜔𝜔𝜔‖ 𝐴𝐴𝐴𝐴 = 𝜋𝜋𝜋𝜋𝑟𝑟𝑟𝑟 2
(7)
𝜔𝜔𝜔𝜔
𝑠𝑠𝑠𝑠. 𝑡𝑡𝑡𝑡. 𝑦𝑦𝑦𝑦𝑖𝑖𝑖𝑖 ∗ (𝜔𝜔𝜔𝜔′ 𝑋𝑋𝑋𝑋𝑖𝑖𝑖𝑖 − 𝛽𝛽𝛽𝛽) ≥ 1, 𝑖𝑖𝑖𝑖 = 1, … , 𝑛𝑛𝑛𝑛
SVM can also be used for regression, where similar
kernel method is applied.
‖𝜔𝜔𝜔𝜔‖ 𝐴𝐴𝐴𝐴 𝜋𝜋𝜋𝜋𝑟𝑟𝑟𝑟 2
𝜔𝜔𝜔𝜔
Multi-layer Perceptron
(MLP):
′
𝑠𝑠𝑠𝑠 𝑡𝑡𝑡𝑡 𝑦𝑦𝑦𝑦𝑖𝑖𝑖𝑖 ∗ 𝜔𝜔𝜔𝜔 𝑋𝑋𝑋𝑋𝑖𝑖𝑖𝑖 − 𝛽𝛽𝛽𝛽 ≥
𝑖𝑖𝑖𝑖
𝑛𝑛𝑛𝑛
𝜔𝜔𝜔𝜔
‖𝜔𝜔𝜔𝜔‖ 𝐴𝐴𝐴𝐴
′
‖𝜔𝜔𝜔𝜔‖ 𝐴𝐴𝐴𝐴 𝑠𝑠𝑠𝑠 𝑡𝑡𝑡𝑡𝜋𝜋𝜋𝜋𝑟𝑟𝑟𝑟𝑦𝑦𝑦𝑦2𝑖𝑖𝑖𝑖 ∗ 𝜔𝜔𝜔𝜔 𝑋𝑋𝑋𝑋𝑖𝑖𝑖𝑖 − 𝛽𝛽𝛽𝛽 ≥
𝜋𝜋𝜋𝜋𝑟𝑟𝑟𝑟 2
Given a set of features and
𝜔𝜔𝜔𝜔 a target y, MLP [19] can learn
a non-linear function estimator
for either classification
′
≥ 𝑖𝑖𝑖𝑖
𝑛𝑛𝑛𝑛
𝑖𝑖𝑖𝑖 ∗ 𝜔𝜔𝜔𝜔 𝑋𝑋𝑋𝑋
𝑖𝑖𝑖𝑖 − 𝛽𝛽𝛽𝛽 backpropagation
or regression.𝑠𝑠𝑠𝑠 It𝑡𝑡𝑡𝑡 𝑦𝑦𝑦𝑦trains
using
with
no activation function in the output layer, which can
also be seen as using the identity function as activation
function. Therefore, it uses the square error as the
loss function, and the output is a set of continuous
values. We use the default parameters of scikit-learn:
one hidden layer with 100 hidden units and “relu” as
activation function.
𝑖𝑖𝑖𝑖
4.2 Rolling Methodology
We perform rolling prediction analysis. That is, we train
our models based on prior historical data and predict
future returns. The procedure then rolls forward by
expanding the train set by one day and then repeating
the training and prediction procedure. A detailed
description is as below.
Suppose we stand on day Dt, and we want to predict
the n-day (n>=1) price returns ahead. To allow the
𝑡𝑡𝑡𝑡
prediction to take place at any time 𝑅𝑅𝑅𝑅of𝑡𝑡𝑡𝑡 day 𝑃𝑃𝑃𝑃Dt,
−we only
𝑃𝑃𝑃𝑃𝑡𝑡𝑡𝑡−1
refer to information up to the previous day
Dt-1. There
are two important considerations:
Typically, KNN [19] method is designed for
classification, where discrete labels are determined by
𝑃𝑃𝑃𝑃𝑡𝑡𝑡𝑡
the majority of certain amount of nearest data points.
𝑅𝑅𝑅𝑅𝑡𝑡𝑡𝑡
−
𝑃𝑃𝑃𝑃𝑡𝑡𝑡𝑡−1 𝑃𝑃𝑃𝑃𝑡𝑡𝑡𝑡
𝐷𝐷𝐷𝐷𝑡𝑡𝑡𝑡
𝐷𝐷𝐷𝐷
𝑡𝑡𝑡𝑡−1 predicted variable (y) is calculated as:
s: 𝑅𝑅𝑅𝑅𝑡𝑡𝑡𝑡 =
−1
However, KNN can also be used for regression where Our
𝑃𝑃𝑃𝑃𝑡𝑡𝑡𝑡−1
𝑃𝑃𝑃𝑃𝑡𝑡𝑡𝑡−1
𝑅𝑅𝑅𝑅𝑡𝑡𝑡𝑡−𝑚𝑚𝑚𝑚 𝑡𝑡𝑡𝑡−1
−
the labels are continuous. The label assigned to a test
𝑃𝑃𝑃𝑃𝑡𝑡𝑡𝑡−𝑚𝑚𝑚𝑚
our explanatory variables (X), we can only use𝐷𝐷𝐷𝐷
point is determined based on the mean of the labels and 𝐷𝐷𝐷𝐷
𝑡𝑡𝑡𝑡−1
𝑡𝑡𝑡𝑡
𝑡𝑡𝑡𝑡−1
For𝐷𝐷𝐷𝐷𝑡𝑡𝑡𝑡−1
example,
the m-day 𝐷𝐷𝐷𝐷𝑡𝑡𝑡𝑡
to day Dtot-1.day
of its nearest data points. Scikit-learn provides three variables up𝑃𝑃𝑃𝑃𝐷𝐷𝐷𝐷
𝑡𝑡𝑡𝑡−1
.For example
𝑃𝑃𝑃𝑃𝑡𝑡𝑡𝑡−1
𝑅𝑅𝑅𝑅𝑡𝑡𝑡𝑡−𝑚𝑚𝑚𝑚 𝑡𝑡𝑡𝑡−1
−
𝑃𝑃𝑃𝑃𝑡𝑡𝑡𝑡−1
𝑃𝑃𝑃𝑃𝑡𝑡𝑡𝑡−𝑚𝑚𝑚𝑚
return
methods of searching for nearest neighbors: 1) brute historical𝑅𝑅𝑅𝑅𝑡𝑡𝑡𝑡−𝑚𝑚𝑚𝑚
𝑡𝑡𝑡𝑡−1 on𝑃𝑃𝑃𝑃D : −
t H𝑅𝑅𝑅𝑅𝑡𝑡𝑡𝑡−𝑚𝑚𝑚𝑚,𝑡𝑡𝑡𝑡−1 =
− 1.
𝑡𝑡𝑡𝑡−𝑚𝑚𝑚𝑚
𝑃𝑃𝑃𝑃𝑡𝑡𝑡𝑡−𝑚𝑚𝑚𝑚
force – compare distances of all pairs of data points;
2) K-D tree – use tree-based structures to reduce the Table 10 provides an example of our data structure.
calculations of distances; and 3) ball tree – partition data
Table 10: An example of data structure of rolling prediction
in a series of nesting hyper-spheres when constructing
trees. As scikit-learn supports auto method selection
Predicted
Date
Explanatory variables (X)
based on input data, we use this option. Also, we use
variable (y)
the default parameters provided by scikit-learn: 5
ay
n-day
Historical m-day
Historical k-day
nearest neighbors and uniform weights.
Support Vector Machine (SVM):
For regression, SVM [19] finds the classifiers
represented by hyperplanes that separate the different
groups as wide a margin as possible. The hyperplanes
are represented by the normal vector v and the bias
The JBBA | Volume 2 | Issue 2 | October 2019
returns
returns
moving averages
ist
Dt
Pt+n / Pt -1
Pt-1 / Pt-1-m -1
SUM(Pt-k, …, Pt-1)/k
P
Dt+1
Pt+1+n /
P+1t+1 -1
Pt / Pt-m -1
SUM(Pt-k+1, …, Pt)/k
)/
Another problem concerning time series rolling analysis
is time series leakage. More specifically, standing on day
58
𝑅𝑅𝑅𝑅𝑡𝑡𝑡𝑡
jbba
the
𝐷𝐷𝐷𝐷𝑡𝑡𝑡𝑡−1
𝑡𝑡𝑡𝑡−𝑛𝑛𝑛𝑛
0
𝐷𝐷𝐷𝐷𝑡𝑡𝑡𝑡
𝑃𝑃𝑃𝑃𝑡𝑡𝑡𝑡−1
𝑃𝑃𝑃𝑃𝑡𝑡𝑡𝑡−1−𝑛𝑛𝑛𝑛
−
𝐷𝐷𝐷𝐷0
to 65 percent.
4.3 Explanatory variables
5.2 Important information sets for Bitcoin
Table 11 shows the explanatory variables in our
rolling prediction analysis (predicting 30-day returns
for Bitcoin). Based on the preliminary analysis above,
we decide to exclude USD index, gold, and VIX,
due to their relatively low correlations with Bitcoin.
The variables are constructed in the abovementioned
rolling way and standardized using StandardScaler in
scikit-learn, which centers the data with sample mean
and the scales them into unit variance.
In addition, we categorize these variables into eleven
“information sets”. In the later sections, we will
examine the relative importance of each information
set for Bitcoin, in terms of their contribution to the
performance of our machine learning algorithms.
As stated above, to reveal the potentially useful
information sources in predicting Bitcoin prices, we
categorize all variables into 10 information sets: 1)
price returns, 2) price momentum, 3) rolling volatility,
4) volume, 5) S&P 500, 6) Developed equity market,
7) Emerging equity market, 8) commodity, 9) market
capitalization weighted returns of cryptocurrencies
(crypto VW), and 10) the 30-day rolling correlation of
the overall cryptocurrency market (rolling volatility).
𝑡𝑡𝑡𝑡−𝑛𝑛𝑛𝑛
𝐷𝐷𝐷𝐷𝑡𝑡𝑡𝑡−𝑛𝑛𝑛𝑛
𝐷𝐷𝐷𝐷99
𝑋𝑋𝑋𝑋130
𝐷𝐷𝐷𝐷129
𝐷𝐷𝐷𝐷100
0
Dt, though we have access to historical information
(X)
𝐷𝐷𝐷𝐷𝑡𝑡𝑡𝑡−1up to the previous day ( Dt-1 ), but we do not have
𝐷𝐷𝐷𝐷𝑡𝑡𝑡𝑡−1 (y), whose calculation involves
the predicted variable
the close price 𝐷𝐷𝐷𝐷on
𝑡𝑡𝑡𝑡 day (t+n). That said, standing on day
Dt , if we want to train a model𝐷𝐷𝐷𝐷and
𝑡𝑡𝑡𝑡 predict the n-day
𝐷𝐷𝐷𝐷𝑡𝑡𝑡𝑡−𝑛𝑛𝑛𝑛 set can only be constructed
𝐷𝐷𝐷𝐷𝑡𝑡𝑡𝑡−𝑛𝑛𝑛𝑛
𝐷𝐷𝐷𝐷the
0
returns ahead,
train
𝑃𝑃𝑃𝑃𝑡𝑡𝑡𝑡−1
−
𝑅𝑅𝑅𝑅
𝑡𝑡𝑡𝑡−𝑛𝑛𝑛𝑛
𝐷𝐷𝐷𝐷𝑡𝑡𝑡𝑡−1
based
on
data from day D0 to𝐷𝐷𝐷𝐷0Dt-n 𝐷𝐷𝐷𝐷(the
𝑃𝑃𝑃𝑃𝑡𝑡𝑡𝑡−1−𝑛𝑛𝑛𝑛
𝑡𝑡𝑡𝑡−𝑛𝑛𝑛𝑛 predicted
variable for Dt-n𝑅𝑅𝑅𝑅is Rt-n=
= 𝑃𝑃𝑃𝑃𝑡𝑡𝑡𝑡−1 − 1)
𝑅𝑅𝑅𝑅1000
𝐷𝐷𝐷𝐷𝑡𝑡𝑡𝑡𝑃𝑃𝑃𝑃
𝐷𝐷𝐷𝐷0
𝑡𝑡𝑡𝑡−1−𝑛𝑛𝑛𝑛
𝐷𝐷𝐷𝐷𝑡𝑡𝑡𝑡−𝑛𝑛𝑛𝑛
𝐷𝐷𝐷𝐷𝑡𝑡𝑡𝑡−𝑛𝑛𝑛𝑛
𝐷𝐷𝐷𝐷𝑡𝑡𝑡𝑡−𝑛𝑛𝑛𝑛
𝑃𝑃𝑃𝑃𝑡𝑡𝑡𝑡−1
𝐷𝐷𝐷𝐷
Finally, we
our
− rolling method with a specific
𝑅𝑅𝑅𝑅𝑡𝑡𝑡𝑡−𝑛𝑛𝑛𝑛repeat
𝑃𝑃𝑃𝑃𝑡𝑡𝑡𝑡−1−𝑛𝑛𝑛𝑛
example. Suppose we have constructed a time series Figure 11: Predicted price vs. Real BTC price (predicting 30-day returns)
data set of 1,000 days: the y is a series of 30-day returns
𝑃𝑃𝑃𝑃128
𝐷𝐷𝐷𝐷99is a matrix of size 1,000𝐷𝐷𝐷𝐷by
𝑅𝑅𝑅𝑅99(20 explanatory
−
𝐷𝐷𝐷𝐷
and
0 X
99 20
𝑃𝑃𝑃𝑃98
𝑃𝑃𝑃𝑃159prediction
variables). We want to experiment 𝑅𝑅𝑅𝑅a130
rolling
−
𝑃𝑃𝑃𝑃129
of 𝑋𝑋𝑋𝑋30-day returns. We set the minimum train
set size
𝑃𝑃𝑃𝑃
130
𝐷𝐷𝐷𝐷99 𝑅𝑅𝑅𝑅99 128 −
𝐷𝐷𝐷𝐷0train𝐷𝐷𝐷𝐷a99model based on the data𝑃𝑃𝑃𝑃 from
as 100.𝐷𝐷𝐷𝐷129
First, we
𝑃𝑃𝑃𝑃98
128
𝐷𝐷𝐷𝐷
𝐷𝐷𝐷𝐷
𝑅𝑅𝑅𝑅
=
−
1;
𝐷𝐷𝐷𝐷
predicted variable for D9999is R99
= 𝑃𝑃𝑃𝑃98
D00 𝑃𝑃𝑃𝑃to
𝑃𝑃𝑃𝑃159
𝐷𝐷𝐷𝐷D99 0(the 99
𝐷𝐷𝐷𝐷
99
128 100
𝑃𝑃𝑃𝑃159
𝑅𝑅𝑅𝑅130
−
𝐷𝐷𝐷𝐷99 𝑅𝑅𝑅𝑅99 𝑃𝑃𝑃𝑃 − 1
𝑅𝑅𝑅𝑅1000 model to predict𝑅𝑅𝑅𝑅130
−1
; then
the R𝑃𝑃𝑃𝑃130
=
𝑃𝑃𝑃𝑃129
98 we use the trained
129
𝑃𝑃𝑃𝑃159
𝑅𝑅𝑅𝑅130 =
− 1 based
𝑋𝑋𝑋𝑋130
X130 (a 1 by 20 row vector) which
𝑋𝑋𝑋𝑋130 on
𝑃𝑃𝑃𝑃129
𝐷𝐷𝐷𝐷129
ins
informatio
contains
information
up
to day D129. Next, we expand
𝐷𝐷𝐷𝐷
129
𝐷𝐷𝐷𝐷100
0 to
Figure 12: Predicted price vs. Real BTC price (predicting 30-day returns)
the train𝐷𝐷𝐷𝐷set
include
data
𝐷𝐷𝐷𝐷0 𝐷𝐷𝐷𝐷100𝑅𝑅𝑅𝑅 from D0 to D100 and repeat Notes:
This figure shows results from Jan 2017 to Feb 2018 for the top 3
the training and prediction.1000
The analysis
is
rolled
until
𝑅𝑅𝑅𝑅1000
algorithms (in terms of accuracy).
we get R1000.
5. Model Results
5.1 Rolling prediction analysis (30-days) for Bitcoin
We recalculate predicted prices based on predicted
30-day returns, as is shown in Figure 11. As the illperformance of Multi-layer perceptron during the
second half of 2017 leads to poor readability, we
present results of the top 3 algorithms (in terms of
accuracy) from Jan 2017 to Jan 2018 in Figure 12.
Obviously, none of them successfully forecasted
the big price crash in Jan 2018. On the other hand,
Figure 13 and Figure 14 show the accuracy and
RMSE, respectively, both of which are calculated in a
cumulative way (expanding the data by one prediction
for each time). As the number of predictions increases,
accuracy of all algorithms stabilizes in the range of 50
The JBBA | Volume 2 | Issue 2 | October 2019
We first run the rolling prediction analysis with all
information sets as input, and next, we repeat the
analysis for 10 times by removing one information
set each time. The “relative importance” of each
information set is measured as the difference between
the accuracies with and without the corresponding
information set as input. That is, a positive difference
indicates positive contribution of the information set
and negative difference implies the opposite.
Figure 15 shows the heatmap presenting the relative9
importance of each information set for each algorithm.
Overall speaking, none of the information sets has
significant impact on any algorithms, as the relative
importance fall in the range between -0.05 and 0.05.
However, a closer inspection would reveal that, on
average, rolling volatility (past 15 days and 30 days)
and correlation among cryptocurrency market (past 30
days) are useful information for most algorithms, while
the market capitalization weighted historical returns
(15-day and 30-day) and emerging equity market are
the least beneficial.
59
jbba
the
Table 11: Explanatory variables
Variable name
Definition
Information set
1
Price_ret10
Historical 10-day price returns
Historical price returns
2
Price_ret30
Historical 30-day price returns
3
Price_momentum_MA10
The ratio of price to 10-day moving average minus 1
4
Price_momentum_MA30
The ratio of price to 30-day moving average minus 1
5
Volume_momentum_MA10
The ratio of trade volume to 10-day moving average minus 1
6
Volume_momentum_MA30
The ratio of trade volume to 30-day moving average minus 1
7
Price_volatility15
The standard deviation of the daily price returns over the past 15
days
8
Price_volatility30
The standard deviation of the daily price returns over the past 30
days
9
SP500_ret15
S&P500 historical 15-day price returns
10
SP500_momentum_MA15
The ratio of price to 15-day moving average of S&P500 minus 1
11
Developed_ret15
MSCI developed equity market historical 15-day price returns
12
Developed_momentum_M
A15
The ratio of price to 15-day moving average of MSCI developed
equity market minus 1
13
Emerging_ret15
MSCI developing equity market historical 15-day price returns
14
Emerging_momentum_MA
15
The ratio of price to 15-day moving average of MSCI developing
equity market minus 1
15
Commodity_ret15
Bloomberg Commodity Index historical 15-day price returns
16
Commodity_momentum_M
A15
The ratio of price to 15-day moving average of Bloomberg
Commodity Index minus 1
17
VW_returns10
10-day market-cap weighted returns 57 cryptocurrencies *
18
VW_returns30
30-day market-cap weighted returns 57 cryptocurrencies *
19
PC1 **
The first principal component of PCA on x-day returns of 57
cryptocurrencies *
20
PC2 **
The second principal component of PCA on x-day returns of 57
cryptocurrencies *
21
Crypto_corr30
The average correlation between the predicted coin and other
cryptocurrencies over the past 30 days
Price momentum
Volume Momentum
Rolling volatility
S&P 500
Developed equity market
Emerging equity market
Commodity
Market capitalization weighted returns of
cryptocurrencies
Principal components of cryptocurrencies
Rolling correlation of the overall cryptocurrency
market
Notes:
1.
* All the “57 cryptocurrencies” above means the 57 cryptocurrencies which have full data back to January 1, 2015.
** The PCA is conducted in a rolling base.
5.3 Rolling prediction analysis for other
Cryptocurrencies
We also examine the analysis for the 57 cryptocurrencies
with available data back to January 1, 2015. Many
cryptocurrencies are slightly predictable if the
algorithms with the highest accuracies are chosen.
Bitcoin yields the highest best accuracy as displayed in
Fig. 14 below. Another finding is that higher prediction
accuracy is associated with larger market capitalization
and lower volatility. But we also see that higher
predictability is accompanied by larger dispersion
among different algorithms.
Figure 14: Summary of rolling prediction results (predicting 30-day
returns)
Figure 13: Relative importance of different information sets on
predicting 30-day Bitcoin returns
The JBBA | Volume 2 | Issue 2 | October 2019
60
jbba
the
Notes:
1.
2.
3.
4.
The volatility is calculated by annualizing the daily volatility over
the sample period (Jan 1, 2015 - Feb 18, 2018). We limit the
range of x-axis to be [0, 6] for the purpose of readability, and as
result 8 cryptocurrencies are removed from the figure.
The highest accuracy: we run 11 algorithms for each
cryptocurrency and pick the one with highest accuracy.
The size of dots is based on the market capitalization of each
cryptocurrency, i.e., Bitcoin is the largest.
The color of dots is based on the standard deviations of
accuracies generated by 12 algorithms (algo dispersion).
Fig. 15 presents a performance summary of the 12
algorithms. LASSO dominates in predicting the 30day returns of cryptocurrencies. And one average, all
algorithms generate accuracies in the range of 50 to 60
percent, which is above random guess but still far from
accurate prediction.
higher volatility. Some care should be taken given the
many moving parts across the cryptocurrency industry.
The complexity will lead to possible risks of overfitting
machine learning algorithms.
References:
[1] I. Madan, S. Saluja, and A. Zhao, “Automated bitcoin
trading via machine learning algorithms,” URL: http://cs229.
stanford. edu/proj2014/Isaac% 20Madan, vol. 20, 2015.
[2] A. Greaves and B. Au, “Using the bitcoin transaction
graph to predict the price of bitcoin,” No Data, 2015.
[3] S. McNally, J. Roche, and S. Caton, “Predicting the price
of Bitcoin using Machine Learning,” in 2018 26th Euromicro
International Conference on Parallel, Distributed and Networkbased Processing (PDP), 2018, pp. 339–343.
[4] Z. El-Abdelouarti Alouaret, “Comparative study of vector
autoregression and recurrent neural network applied to bitcoin
forecasting,” PhD Thesis, ETSI_Informatica, 2017.
[5] F. Mai, Q. Bai, J. Shan, X. S. Wang, and R. H. Chiang,
“The impacts of social media on Bitcoin performance,” 2015.
Figure 15: Summary of algorithm performance (predicting 30-day
returns)
Notes:
1. The frequency is the times an algorithm performs the best
among the 11 algorithms plus random guess.
2. The mean accuracy is calculated by averaging the accuracies
when the corresponding algorithm performs the best.
6. Conclusion
Cryptocurrencies have captured the attention of many
investors across the spectrum from retail to institutional
- see Liew and Hewlett [14]. In this work we extend our
understanding of the behavior of cryptocurrencies.
We document several interesting findings. First off,
we find that PCA reveals that the return generating
process is much more complex than that for stock
returns. Generally speaking, the financial community
agrees that the “market” is the first dominant PCA
in stock returns. However, for cryptocurrencies daily
returns reveals that in some period there exists a single
dominant component however, in the most recent prior
year there appears to be two components that help
explain the variation of the cryptocurrency returns.
Next, we document a strong beta-in-the-tails hidden
risk associated with Bitcoin daily returns. Similar to
hedge fund cryptocurrencies may have some unstable
tail behaviors.
Our analysis of machine learning algorithms
applied to the data from cryptocurrencies hints that
predictability may be difficult and there are many
heterogeneous effects here. Some information sets
perform better with some family of algorithms, and
larger cryptocurrencies with lower volatility maybe
more predictable than smaller cryptocurrency with
The JBBA | Volume 2 | Issue 2 | October 2019
[6] E. Stenqvist and J. Lönnö, Predicting Bitcoin price
fluctuation with Twitter sentiment analysis. 2017.
[7] Y. B. Kim, J. Lee, N. Park, J. Choo, J.-H. Kim, and C. H.
Kim, “When Bitcoin encounters information in an online forum:
Using text mining to analyse user opinions and predict value
fluctuation,” PloS one, vol. 12, no. 5, p. e0177630, 2017.
[8] J. Kaminski, “Nowcasting the bitcoin market with twitter
signals,” arXiv preprint arXiv:1406.7577, 2014.
[9] J. Chu, S. Nadarajah, and S. Chan, “Statistical analysis
of the exchange rate of bitcoin,” PloS one, vol. 10, no. 7, p.
e0133678, 2015.
[10] M. Balcilar, E. Bouri, R. Gupta, and D. Roubaud, “Can
volume predict Bitcoin returns and volatility? A quantiles-based
approach,” Economic Modelling, vol. 64, pp. 74–81, 2017.
[11] N. Indera, I. Yassin, A. Zabidi, and Z. Rizman, “Nonlinear autoregressive with exogeneous input (NARX) Bitcoin
price prediction model using PSO-optimized parameters and
moving average technical indicators,” Journal of Fundamental
and Applied Sciences, vol. 9, no. 3S, pp. 791–808, 2017.
[12] I. Georgoula, D. Pournarakis, C. Bilanakos, D.
Sotiropoulos, and G. M. Giaglis, “Using time-series and
sentiment analysis to detect the determinants of bitcoin prices,”
Available at SSRN 2607167, 2015.
[13] D. Garcia and F. Schweitzer, “Social signals and
algorithmic trading of Bitcoin,” Royal Society open science, vol.
2, no. 9, p. 150288, 2015.
61
jbba
the
[14] J. K.-S. Liew and L. Hewlett, “The case for Bitcoin
for institutional investors: Bubble investing or fundamentally
sound?,” Available at SSRN 3082808, 2017.
[15] F. R. Edwards and M. O. Caglayan, “Hedge fund
performance and manager skill,” Journal of Futures Markets:
Futures, Options, and Other Derivative Products, vol. 21, no.
11, pp. 1003–1028, 2001.
[16] J. Liew, “Hedge fund index investing examined,” Journal
of Portfolio Management, vol. 29, no. 2, p. 113, 2003.
[17] W. F. Sharpe, “Capital asset prices: A theory of market
equilibrium under conditions of risk,” The journal of finance,
vol. 19, no. 3, pp. 425–442, 1964.
[18] J. Lintner, “The valuation of risk assets and the selection
of risky investments in stock portfolios and capital budgets: A
reply,” The review of economics and statistics, pp. 222–224,
1969.
[19] F. Pedregosa et al., “Scikit-learn: Machine Learning in
Python,” Journal of Machine Learning Research, vol. 12, pp.
2825–2830, 2011.
i
The data of Ethereum provided by coinmarketcap.
com starts on Aug 7, 2015.
ii
The average history is calculated using the data for
only 2015 to 2017, thus it is not the exact length
of average history. But as most of the top 100
cryptocurrencies came into being after 2015, this
calculation approximates the real length of average
history.
iii
For horizontal axis, cryptocurrencies are ranked by
market capitalizations from the right (large) to the left
(small). For vertical axis, they are ranked by market
capitalizations from the top (large) to the bottom
(small).
Accessed on Mar 14, 2018: https://www.cnbc.
com/2017/11/27/bitcoin-exchange-coinbase-hasmore-users-than-stock-brokerage-schwab.html
iv
Competing Interests:
None declared.
Ethical approval:
Not applicable.
Author’s contribution:
JL1, RL2, TB2, and AS3 designed and coordinated this research and
prepared the manuscript in entirety.
Funding:
None declared.
Acknowledgements:
None declared.
The JBBA | Volume 2 | Issue 2 | October 2019
62
jbba
the
The JBBA | Volume 2 | Issue 2 | October 2019
63
jbba
the
Peer-reviewed Research
Blockchain Investigations:
Beyond the ‘Money’
OPEN ACCESS
ISSN Print: 2516-3949
https://doi.org/10.31585/jbba-2-2-(6)2019
Simon F. Dyson
NHS Digital, Leeds, U.K
Correspondence: simon.dyson@protonmail.com
Received: 18 July 2019 Accepted: 7 August 2019 Published: 13 August 2019
Abstract
Cryptocurrency investigations have centered almost entirely around the transfer of value “money” or a
cryptocurrency asset. The use of cryptocurrency for illicit purposes, especially Bitcoin, is well documented
both in academic writing, media reporting and even film documentaries. The infamous SilkRoad marketplace
in addition to the millions of dollars spent within dark markets on drugs, guns and assassinations have
grabbed the headlines. This paper looks at how blockchain is creating new areas of investigation that are yet
to be explored in detail. This scenario-based research examines the hosting of stolen data (P.I.I) personal
identifiable information on a distributed blockchain host where the data is also accessible. The platform used
is based on Ethereum infrastructure but demonstrates just one available platform that poses the paradigm.
The paper examines the considerations through the lens of an incident responder /cyber investigator,
forensics examiner and data controller. The scenario highlights distinct differences in considerations from a
traditional response compared to dealing with the immutable and unstoppable distributed technology. The
paper concludes that more is needed to be done to understand digital forensics in the blockchain era and
the need to develop beyond track and trace in the cryptocurrency investigative toolbox. The discussion also
brings forth how data retention and GDPR requires consideration when applying it blockchain systems.
Keywords: Blockchain, Distributed-hosting, Distributed-storage, Ethereum, Swarm, Forensics
1. Introduction
Research into cryptocurrency has focused generally
on the transfer of value. The use of cryptocurrency
in large scale criminal activities is well documented in
cases such as the Silk Road drugs marketplace or in large
ransomware campaigns such as Wanacry. The focus
has been on the “follow the money” aspect in order
to locate the perpetrators. The underlying technologies
have however developed since the inception of Bitcoin
in 2008. Blockchain technology is now scaling and
developing new features now able to support multiple
data and communication protocols across its stack.
Law enforcements focus has remained around the large
cryptocurrencies however the use of smart contract
technology and now distributed computing and
storage creates a new set of problems for investigators
and those responding to incidents. This paper sets out
a common leak of personally identifiable information
(P.I.I) where it is hosted on blockchain technology and
how the traditional responses are required to adapt.
The scenario uses Ethereum and its related technology
to host files. There are a number of cryptocurrency/
blockchain assets that can host the data in a similar
The JBBA | Volume 2 | Issue 2 | October 2019
nature. A distributed blockchain by its design contains
properties that are not inherent in traditional hosting
services. A blockchain is immutable in general terms
so they are unstoppable and have no central authority
or body.
2. Scenario and Roles
The scenario is to replicate the discovery of files taken
containing (P.I.I) personally identifiable information
from a server and hosted externally. The hosting,
however, will take place on a distributed blockchain
system. In order to establish if the PII information is
legitimate, a comparison will need to take place, this
will entail a visual comparison of the data. A forensic
comparison of the data will need to be conducted
using traditional methods to hash the file contents and
examine EXIF data contained within the file. Cyber
investigators searching for online hosted material will
examine records of web hosting companies to see the
I.P data for the hosting company and registrar details
such as WHOIS information. Data controllers hold the
responsibility for the holding storage and protection
of the data. The data controller will need to make
65
jbba
the
decisions about steps that are possible to minimise
the damage. Each role will respond using traditional
methods and record the findings. A discussion section
will reflect on the approaches and highlight quick wins
and areas that require further work.
2.1. Cyber Investigator / Incident Response
This role will respond to initial reports and record
and utilise OSINT Open Source Intelligence sources
to discover evidential information to assist the
investigation. The coordination of tasks to systems
administration for internal log investigation and other
closed source materials will be conducted.
2.2. Digital Forensics
Examination of digital material will be conducted by
the digital forensics team member. This will include
host forensics and also comparison of highlighted
online material where required. They will take a forensic
analytical approach in order to approach the problem.
2.3. Data Controller
As the responsible owner of the data, the controller
will be consulted on the state of the investigation.
The controller will establish additional tasks that
would assist to protect the data or prevent further
dissemination.
of automated decision making and profiling. The right
to erasure is one of the more complex and powerful
rights that is created in the new act. This right caused
many to question the ability of blockchain to function
under such a regime. The conflict of immutability as
an absolute property of blockchain in comparison to
the legal requirement to deletion of GDPR is cited as
pushing solutions to standard databases[1]. There are
other potential ways forward such as gaining consent
for perpetual processing. It is argued that address
hashing is pseudonymous and that the effort to deobfuscate a hash is disproportionate so would stand as
it would not likely identify an individual. Permissioned
blockchains are also suggested in order to control the
data but they don’t fit the public and permissionless
systems of large cryptocurrency structures. There are
also systems using new encryption methods such as
zkSNARKS and RingCT methods that could protect
data throughout the complete process [2]. Tokenised
solutions are appealing although they may require offchain processing but the use of distributed storage is
possible through Ethereum – SWARM or IPFS[3]. The
use of a smart contract with an upgradable contract
section could allow amendable content but record
the transaction metadata and deletion process[4].
Implemented correctly the ability to control and make
accountable sharing structures with blockchain could
strengthen systems to comply with GDPR.
4. Decentralised Blockchain Storage
3. (GDPR) General Data Protection Regulation
In May 2018, the General Data Protection Regulations
came into effect and incorporated existing legislation
to protect people’s data and their rights. GDPR is
covered in depth in numerous resources so in this
section a focus on some key themes that will be later
visited will be briefly documented. GDPR covers data
that belongs to people who are in the GDPR aligned
nations, Europe and some additional territories. The
rules outlined cover those entities that are considered
a data controller or processor. A controller is the entity
that holds the data for purpose, and they will process
for their agreed business requirements. A processor is
considered a third party that is doing something with
the data on behalf of the controller, an agreement
will define what that process is. GDPR defines that
personal data can generally identify somebody or be
used for that purpose and it offers protection to that
data. There are also additional protections to sensitive
personal data that protects special characteristics. The
term PII (Personal Identifiable Information) is not
defined by GDPR but is commonly used and will
be covered as personal data under GDPR and this
scenario. There are 8 individual rights that are listed
under the new act. These rights are; the right to be
informed, the right of access, the right of rectification,
the right to erasure, the right to restrict processing, the
right to data portability, the right to object and rights
The JBBA | Volume 2 | Issue 2 | October 2019
Decentralised networks have been utilised for
numerous cryptocurrency projects with the ability to
trade tokenised value they have become used for a
new wave of “Digital money”. Blockchain technology
itself has evolved behind the headlines of boom and
bust price fluctuations and Silk Road drug dealing dark
markets. The introduction of Smart Contracts utilised
by Ethereum and now other blockchain technologies
allows Turin complete languages and sections of code
to produce complex computational outputs. Using
resources on Ethereum for example is expensive if you
process through the Ethereum Virtual Machine (EVM)
the world computer, each byte and code execution has a
price to pay using “gas”. Utilising “gas” small amounts
of the currency this ensures that the “halting problem”
is addressed and a denial of service attack or forever loop
will be too expensive to conduct. There are however a
number of blockchain projects that are looking to use
an additional protocol or system to provide blockchain
storage using peer to peer nodes incentivised to
the system. The creation of a decentralised storage
system solves a number of computing problems, it
creates resilience as files are striped across multiple
nodes in a system. The ability to reside on multiple
nodes reduces single points of failure or risk from
physical events such as earthquake, tsunami or power
outages. A decentralised system uses nodes in the
control of world users who are incentivised to “mine”
66
jbba
the
or provide a service similar to miners and Bitcoin
nodes. Services such as Dropbox operate a storage
system that allows a cloud storage system however the
service is a centralised under one organisation. The
company is subject to US law, so privacy therefore
is not guaranteed as the ability to access, subpoena,
court order and secret service oversight. Nodes in a
decentralised generally hold only partial fragments of
the file so physical integrity is maintained as the file
portion is fragmented and optionally encrypted. There
are a number of decentralised file storage systems
namely, IPFS (Inter Planetary File System) developed
by Protocol labs this part of the system allows for
distributed storage, Filecoin [5] is an additional service
to incentivise storage by paying miners to store. IPFS
as a protocol is used by a number of other projects
and is cross blockchain agnostic [6]–[8]. In addition to
the above there are other distributed storage projects
in various phases of production these include Storj,
Sia and Maidsafe [9]–[11]. Ethereum has its own sub
project called “Swarm” this will be explored in the next
section.
5. Ethereum Swarm
Swarm was designed to create a system to store
dapp (Decentralised Application) code, resources
on a peer2peer system. The ability to access material
outside of the Ethereum chain reduces the cost of
storing larger files or code in a smart contract off chain
where it is cheaper to store. Dapps by their nature
are applications that are not a singular stored item,
therefore the use of larger code sets and files to produce
more complex and visually focused items require more
storage. The ability to access resource from the Swarm
protocol layer allows this exchange maintaining a fully
decentralised eco-system. The system will maintain the
properties of a truly decentralised system transaction
layer on chain and storage another chain. This makes
it non-censorable, fully redundant / resilient, DDOS
resistant, highly available and secured by encrypted
cryptographic signatures. Ethereum integration is
used with a Swarm node and a Geth node, Geth is a
“GO” programming language implementation version
of Ethereum. This scenario will utilise both Ethereum
Geth and Swarm working on the Ropsten test net,
the closest to the production service. Ropsten allows
integration with the services as if it was connected to
the Mainnet where the technology is already live, with
the advantage of not costing real Ethereum and “gas”
to test and operate. Geth version (1.8.20-stable) and
Swarm version (0.3.-stable) [12], [13].
6. Blockchain Domain Naming
The (DNS) Domain Name System is used to assist with
searching the internet, it translates a human readable
(URL) Uniform Resource Locator into the relevant
Internet Protocol Address (I.P). This directs a query
The JBBA | Volume 2 | Issue 2 | October 2019
such as www.a_web_address.com to the root servers to
the (TLD) Top Level Domain and to the domains name
server that holds the record of the I.P address example
8.8.8.8. The ability to store domain naming information
on a blockchain has existed for some time with services
such as Namecoin offering various services including
a name resolution stored on blockchain. The criminal
use of decentralised DNS services does exist but is
not extensively used [14], [15], [16]. The discovery of
a recent botnet that was discovered to be cleaning up
bad botnets was observed in the wild using Emercoin’s
distributed DNS implementation [17].
The (ENS) Ethereum Naming Service provides similar
functions to a DNS system and is held and operated
over the Ethereum blockchain [18].
7. Method
In order to replicate an intrusion event a number
of files with identifiable meta-data will be created
and hosted on a virtual machine. A base forensic
examination will be completed to display (Modified,
Access, Created) MAC date and times, Meta data that
may also include geo-data serial number or other EXIF
data. The scenario host is a machine running Windows
7, the host contains a folder on the desktop entitled
“Work_items” containing related documents. The
documents include an image of a passport that is used
for identification of the customer in this scenario and
contains PII information. In addition to the image are
further documents including an XLS and CSV files,
this contains customer details including PII data.
Scenario – a message is received by phone that a leak
of company information has occurred, and a website
URL is provided.
7.1. Investigation phase
7.1.1. Cyber investigator / Incident response
The cyber investigator is initially passed information
provided by a telephone call that states the web URL
hosting the company’s potentially stolen information.
The URL is placed into a browser on a standalone
environment to ensure the reported event is not a social
engineering ruse and to protect the main corporate
network from malicious activity. Initial activity will
ensure the link is live and that the data appears to be
present, accurate recording of event will take place
including a screen capture of the page. A capture of
the page and the source code alongside an abstraction
of pertinent files will be completed for further analysis.
Data will be needed to be compared to corporate data to
ensure the attack is a legitimate attack and is not a hoax.
OSINT Open source intelligence will reveal additional
information about the web hosted material. The source
code may reveal hosting details or frameworks used
to create the site, these may include author and other
67
jbba
the
meta data of interest to the cyber investigator. Source
code can also reveal other links hosted on other sites or
resources that may allow additional investigative leads.
As discussed in Open source intelligence techniques by
Michael Bazzell there are numerous services that assist
in the location of a website these include some of the
following important areas [19]:
•
•
•
•
•
•
•
•
•
•
•
•
•
Protocol
Website name and top-level domain
information
I.P address
Whois
Registration data
e-mail addressing
The hosting company (Server hosting)
Domain hosting (Name holder)
DNS zone tranfers
Registrar change history
Ad-sense / analytical tokens – numbers
Robots.txt
Shodan
7.1.2. Digital Forensics
Following information from the original call the
forensic response team will react to the main areas
of data storage. The firewalls and server logs will be
checked for intrusion or indicatiors of compromise.
The data storage servers will be examined, and a
RAM dump will be executed on each device. This will
capture processes, network connections and master file
table entries that will enable initial triage to identify any
breach information. Identification of the information
can take place by using methods such as hashing
values and searches for names or data from the leaked
source to discover if the information is owned by the
company.
The order of volatility is Processor, Network, Main
Memory, Semi-volatile, Resident data, Remotely logged
and any data on archival media [20]
In this scenario, live data should be considered before
a raw dump, if the memory dump crashes then the
machines critical live data could be lost. A memory
dump should be obtained and analysed the machine
can be shut down and retained for a full forensic image
if required.
The forensic response to an incident would record
the process using contemporaneous notes and
photographs.
Examine with the visual inspection of a machine and
examination of live desktop activity
•
Live data – command line – time & date,
network connections (netstat), current user,
tasklist
•
Memory RAM capture – full, Dumpit.exe
•
Any operational/incident specific
investigation tasks.
•
Power down machine when examination
The JBBA | Volume 2 | Issue 2 | October 2019
complete retain for full disk imaging.
7.2.
7.2.1.
Operational Phase
Data Controller
The General Data Protection Regulation GDPR
introduced in 2018 enforces businesses and those who
control data to protect the rights of the citizens whose
data is held. Each European country or participating
country must introduce a body to monitor and
administrate enforcement of fines and breaches of the
code. In the U.K the ICO (Information Commissioners
Office) hold this position, they provide guidance,
advice and are the primary contact if a breach occurs.
GDPR requires a company that is aware of a breach of
personal data to report to the supervisory authority in
the U.K to the I.C.O within 72 hours of been aware that
a breach has occurred. Where it is likely that a breach
will affect the rights and freedoms of the individuals on
who the data relates then they must also be informed
“without undue delay” [21]. The principles that are
to be considered around data are the security triad of
Confidentiality, Integrity and Availability. The rights of
the data subject are to be considered and notification
made if the breach is likely an adverse effect on the
data subject. An example would be where full personal
data and financial data are lost these are likely to incur
subsequent fraud offences using the identities of the
data subject [22].
GDPR therefore requires all companies that process
data in the EU or about people in the EU to have
policies and procedures to detect, investigate and
report on incidents with accuracy.
GDPR has a number of requirements in relation to
information to be provided to the supervisory body
in response to a breach. The below section details the
requirements and these points will be addressed in the
breach investigation plan for the data controller.
•
•
•
Description of the personal data (data
categories, number of individuals, number of
records)
An assessment on potential consequences
following the breach
Following the breach what measures have
been or will be taken in order to mitigate risk
and harm following the breach. (ICO GDPR
breach guidance [23]).
It is obvious from the above requirements that a
response from the cyber investigators / digital forensic
team is essential in providing timely and accurate
reporting to ensure the data controller can make
informed decisions on the subject.
68
jbba
the
8. Initial Response
The scenario starts with a report of information
reported into the Cyber security team. This was initially
reported as a URL and the action will start by the teams
who will perform incident response according to their
response plans. The URL was reported as:
https://swar m-gateways.net/
bzz:/9eaab00f3eb97cfc731ae095 8aa2c9f249a2cd0045
dae7bec659e736c920112a/
Figure 2. Shodan results from the OSINT scan
9.3. Forensic Findings
9. Findings
9.1. Cyber Findings
Initial actions resulted in the preservation of the online
material and the capture of HTML information of the
files hosted on the site. The site contained three items
of interest an image of a passport and two data files
for download. The nature of the message on the site
suggested an insider threat this requires further internal
work to attribute.
The image was reverse searched to see if the image
was hosted elsewhere on the web, this was to establish
if this was disseminated elsewhere or if it was a hoax
using another source. The EXIF data was examined
from the passport photo, this provided metadata that
included time dates, make, model, image composition
and crucially geo location, see Fig 1. This established
the passport image was taken in a popular café used
by the sales team to on-board new customers close
to company premises. This provided metadata that
allowed attribution in this circumstance in the scenario
set out.
The files recovered in the discovery phase were
provided for forensic analysis. The items were hashed,
and the metadata examined. This information enabled
identification of the company database server where
the data was likely to be stored. The forensic actions
as previously described were enacted capturing live,
ram and forensic level data. The company server was
examined, and activity was discovered around the
folder of interest using Volatility. Artefacts were found
in the MFTPARSER and SHELLBAGS modules
that allowed activity from MAC (Modified Accessed5
Created) times to create a timeline of suspect activity.
In addition, access to the registry keys through the
Volatility modules allowed the USBSTOR to show
activity in the timeframe, giving make model and
GUID for the suspect USB. Table 1 shows the hashing
data and the matches.
Table 1. Shows the hash detail and if the hashing matched from the
blockchain storage and the host system
10. Data controller – next steps
Figure 1. EXIF data from the passport image
9.2. OSINT – Findings
The domain was subject to a reverse look up,
WHOIS search, and it revealed a hosted service. The
information was shown to a Windows Azure instance
based in the Netherlands. The domain registrar shows
a named contact with addressing. The I.P address was
established with versioning and port numbers on a
Shodan scan that revealed web ports 443 and 80 were
the only active services see Fig 2.
The JBBA | Volume 2 | Issue 2 | October 2019
In this section discussed is considerations of the data
controller for the blockchain element and not the
general actions of the controller, the data is personal
and a referral to the ICO is required in the time frames
set. The current actions would now look to reduce the
spread of stolen data. Legal action against the hosting
company or a complaint procedure to the hosting
services would be sought. A powerful tool for removal
of data is the Subject Access Request procedure where
a data subject has rights under GDPR for enforcement.
Legal proceedings are complex and civil claims can
potentially disrupt or force servers to shutdown such
services as detailed between PML vs unknown [24].
This shows the complexity and interactions that a
hosting company can be pursued to reduce the impact
and required to remove content under national and
international law.
11. Initial conclusion standard response
The conclusion established from the above investigation
at this stage are mixed. The captures have been
69
jbba
the
performed to an adequate standard but there are some
items that confuse the investigation. The domain and
services discovered in the phase point to the “swarm
gateway” a service allowing a pass through of web
traffic to the Ethereum network. The Microsoft Azure
server hosted in The Netherlands and the registrar
name highlighted is a project lead on the Swarm
service. The registrar and the WHOIS information all
resolve to an unrelated subject not the true location
of the data, just the portal to find it. It is important
to relay that there is no information hosted on the
server it is a HTTP/S proxy API that allows access to
the Swarm network. There are other gateways such as
https://ensgateway.com/ and IPFS specific gateways.
What legal action can be taken against a portal that
contains no data but provides access. Similar to that of
a tor gateway or node allowing access to a darkmarket.
The forensic investigation however demonstrates
that the files and data hosted on the Swarm system
are not altered and retain important meta-data. The
comparison shows that the integrity of the file is
retained and the hashing value and EXIF data is
retained when recovering from the Swarm network this
confirms attribution for the company.
The investigation can conclude that the personal
information has been taken from the company and
this has been conducted by an individual with authority
to access the service. The attack was conducted by
exfiltrating data and removing it on a USB device
this is a classic insider attack. The file is hosted on an
unstoppable blockchain where no legal avenue exists to
remove or request a cease and desist.
human readable name, name-hash, account or content
to a resolver. There is a public resolver frequently used
however custom resolvers are possible to create and
likely to be adopted in some Dapps or other services.
A name hash is used to represent the human readable
name and is combination of cumulative hashing of
domain and naming using Keccak hashing [18]. Fig 3
shows a walkthrough of the process.
In the example of the scenario the content hash was
created by Swarm ‘9eaab00f3eb97cfc731ae0958aa2c
9f249a2cd0045dae7b ec659e736c920112a” this hash
was used to search the Swarm node to retrieve the full
content. The Swarm hash is created using a chunk hash
function with a merkle tree, this is currently formulated
Figure 3. A breakdown of the ENS and the different elements involved
in name resolution
using a 32 byte Keccak(256)SHA3. It is possible to
create a hash of just a file or similarly in this case a
folder with linked resources and files. In the meta-data
for the html file the linked images are referenced as the
hash and file Fig 4.
12. Blockchain investigations
The Swarm decentralised system operates using the
URL scheme identifier as “BZZ:” the location of the
file is designated by a Swarm hash or an ENS assigned
domain such as “photoalbum.eth”.
In the example, the ENS domain is assigned as
“Unstoppable.eth” - Ropsten testnet and this resolves
the content of the stolen items as examined previously
to the swarm hash.
There are a number of components that work to
resolve the addressing. At high-level an Ethereum
registry that tracks the domains and sub-domains
on the network. There are additional registrars that
are involved in the hosting and reselling activities of
ENS names. The Ethereum Naming Service is used
to bid and retrieve a human readable address such as
“Unstoppable.eth” and this is done using an Ethereum
account. On successful allocation of the bid the name
is under the control of the account and using Smart
contract calls can be accessed and communicated
with to set the requirements in the contract. The
ENS record requires a resolver assigned that links the
The JBBA | Volume 2 | Issue 2 | October 2019
Figure 4. Web link that shows the Swarm hash in the HTML link data
from the page
In Fig 5 displayed is a resolved address through the
ENS service linking to the swarm hash in this case
“photoalbum.eth”. To discover the hash the ENS
address is resolved to an account the contract held
on the account can be searched for the “setContent”
function as shown in Fig 6.
Figure 5. Web link that shows an ENS address in the HTML link from
the page
Figure 6. Method setContent function applying the hash to the node
address
ENS names can also be applied to accounts so
unstoppable.eth can applied to an Ethereum account /
wallet and be used instead of the long account address.
70
jbba
the
Fig 7 shows the name, resolver and account details
assigned and revealed with in an ENS search within
myetherwallet.com (Ropsten).
Figure 7. ENS reverse lookup that shows the Name and additional
bidding and resolved address
13. Discussion
Figure 7. ENS reverse lookup that shows the Name
and additional bidding and resolved address
The ability to host decentralised resources and store
material that would be traditionally held on centralised
services changes some of the traditional methods
of search. This scenario has demonstrated how
material can persist beyond the normal experience
of investigators creating an unstoppable hosting
problem. The practical element demonstrated the
use of the Ethereum network, just one of the
technologies available to perform distributed storage.
The ENS Ethereum Naming Service also provides
the ability to link to TLD domains such as .xyz and
.luxe. It is understood that the DNSSEC and the
TLD integration will not likely resolve correctly with
ENS as the DNS browser protocol may override the
resolving [25]. There were a number of limitations and
technical issues that could not be overcome to test a
.xyz domain with any objectivity or confidence. The
Ropsten testnet had some service issues during my
testing with ENS and syncing, this included using
the third-party API Infura that demonstrated the
same behaviour. Where required I have used Ropsten
and confirmed behaviour across the Mainnet with
alternatively hosted sites. There are interesting uses
of ENS and DNS hosted on Ethereum, the EthDNS
system is prototyped and documented that uses DNS
records stored on Ethereum [26]. There are potentially
interesting attack vectors if Swarm and ENS became
mainstream the use of a bad resolvers in new “Dapps”
for example. IPFS also needs to be investigated to
understand how it can be used in addition to existing
technology or integration with other blockchain
technologies. As the example shows the ability to
bring up a node write information into the distributed
storage is possible both quickly and cheaply, removal 7of
the node from the system still allows the new files to
remain. Attribution using a blockchain explorer allows
account identification additional resources, identifiable
information and linked smart contracts. The layers of
investigation cut across web technology, blockchain
account records, smart contracts, blockchain naming
service, blockchain storage and the host machine.
The JBBA | Volume 2 | Issue 2 | October 2019
These can lead directly to additional accounts that
may identify cryptocurrencies entering or leaving the
system. The ability to interact with a smart contract
using privacy focused technologies such as zkSNARKS
or private smart contracts such as Enigma allow data
or image sharing autonomously with strong encryption
[27]. The ability to create a photo-sharing application
for payment with content hosted on decentralised
storage can be achieved using privacy focused methods
in addition to blockchain technologies.
What is demonstrated is a need to understand the
sources of hosted material as distributed storage
becomes wider spread in its adoption. Hosting malware
on distributed storage or indecent images of children
will require investigators and responders to locate all
the sources of material. In the examples shown it is
possible to make attribution to file access and use
for forensic examination. File signatures, hash values,
hosted distributed domains, protocol specific URLs,
e.g. BZZ or IPFS can be extracted. In incident response
scenarios, the ability to source and collect the sample
for reverse engineering will be essential for mitigation
and research. Virus scanning and network protection
rules could be used to search detect and block hosted
material entering or leaving a network. Fig 8 shows the
host and file access to the blockchain via Blockchain
node / software or via internet gateways.
Fig 8 shows a host connecting via the blockchain protocols or via an
internet gateway
14. Conclusion
This scenario has demonstrated it is possible to store
content persistently on blockchain technology allowing
access to those on the blockchain and to the internet
through internet gateways. Decentralised storage
remains uncensorable with no technical recourse to
remove or even request for lawful motions against its
storage. There are no regulations such as GDPR, local
laws, state, or international law that have any power to
control or remove it. The hosting of resources such
as images or files on distributed file storage requires
additional investigative methods to discover the source
and linked information. The ability to attribute the
access or presence of an illegal image or document
can be reliable proven using hashing protocols used in
Ethereum Swarm, the Swarm hash and the temporal
71
jbba
the
data from the blockchain against fragments held on
the host. The requirement to recover electronic data
stored or what was accessed is needed in E-discovery
and for corporate legal compliance, so the need exits
to be able to seek and find documents hosted as
described. Malware researchers require the source file
to reverse engineer or perform static analysis so the
ability to access blockchain storage to recover such
files along with additional threat intelligence from
linked accounts and blockchain naming is essential.
In this case forensic artefacts were not interfered
with in terms of their integrity, this is good news for
forensic investigators wanting to review rich sources of
meta-data. This was only performed on the Ethereum
Swarm and other storage systems may also leave
metadata or artefacts, a potentially important forensic
research area. Research on distributed storage is still
focused on the introduction, development, scalability
and the performance of the technology. There are
clearly vast gaps in literature around the use and longterm performance behaviour as the technology is
rapidly evolving. Blockchain forensics has focused on
cryptocurrency track and trace but the evolvement of
smart contracts and now storage and computational
resource will be a future frontier. It is unclear on
the adoption of these technologies to long-term
adoption, but a new challenge and knowledge gap
could appear overnight. Blockchain will undoubtable
continue to pioneer computational breakthroughs but
new paradigms and challenges exist in its wake. The
misuse cases should be considered and researched
to compliment blockchain development as a global
revolution.
References:
[1] F. January, M. Ii, F. I. Directive, R. R. Review, E. Union,
G. Data, P. Regulation, and T. Gdpr, “The rise of the regulator
may lead to trouble for the blockchain,” pp. 1–2, 2018.
[2] C. Salmensuu, “General Data Protection Regulation and
the Blockchains,” Liikejuridiikka, no. 1, p. 92, 2018.
[3] B. Ramsundar, R. Chen, A. Vasudev, R. Robbins, and A.
Gorokh, “Tokenized Data Markets,” 2018.
[4] N. Vergauwen, “Upgradeable Smart Contracts – Hacker
Noon,” Medium - Hackernoon, 2018. [Online]. Available:
https://hackernoon.com/upgradeable-smart-contractsa7e9aef76fdd. [Accessed: 15-Jan-2019].
[5] “Filecoin - Website,” 2019. [Online]. Available: https://
filecoin.io/. [Accessed: 15-Jan-2019].
[6] “IPFS is the Distributed Web,” 2019. [Online]. Available:
https://ipfs.io/. [Accessed: 15-Jan-2019].
[7] J. Redman, “BCH-Powered Bitcoin Files Project Adds
IPFS Support - Bitcoin News,” News-Bitcoin.com, 2018.
The JBBA | Volume 2 | Issue 2 | October 2019
[Online]. Available: https://news.bitcoin.com/bch-poweredbitcoin-files-project-adds-ipfs-support/. [Accessed: 06-Jan2019].
[8] M. Zalecki, “Using IPFS with Ethereum for Data Storage
| Tooploox,” TOOPLOOX - WEB, 2018. [Online].
Available: https://www.tooploox.com/blog/using-ipfs-withethereum-for-data-storage. [Accessed: 06-Jan-2019].
[9] “MaidSafe,” 2019. [Online]. Available: https://maidsafe.
net/. [Accessed: 15-Jan-2019].
[10] “Sia,” 2019. [Online]. Available: https://sia.tech/.
[Accessed: 15-Jan-2019].
[11]
“Storj - Decentralized Cloud Storage,” Storj Decentralized Cloud Storage. 15-Nov-2017.
[12] “Go Ethereum,” Geth. [Online]. Available: https://geth.
ethereum.org/. [Accessed: 15-Jan-2019].
[13] “1. Introduction — Swarm 0.3 documentation,” Swarm
read the docs, 2019. [Online]. Available: https://swarm-guide.
readthedocs.io/en/latest/introduction.html. [Accessed: 15-Jan2019].
[14] R. Amado, “How Cybercriminals are using Blockchain
DNS | Digital Shadows,” Digital Shadows_ (Web), 2018.
[Online]. Available: https://www.digitalshadows.com/blogand-research/how-cybercriminals-are-using-blockchain-dnsfrom-the-market-to-the-bazar/. [Accessed: 14-Jan-2019].
[15] “Namecoin,” Namecoin (Web), 2019. [Online].
Available: https://namecoin.org/. [Accessed: 14-Jan-2019].
[16] M. Ali, J. Nelson, R. Shea, and M. J. Freedman,
“Blockstack : A Global Naming and Storage System Secured
by Blockchains,” USENIX Annu. Tech. Conf., pp. 181–
194, 2016.
[17] I. Ilascu, “New Botnet Hides in Blockchain DNS Mist
and Removes Cryptominer,” Bleeping Computer - Web, 2018.
[Online]. Available: https://www.bleepingcomputer.com/
news/security/new-botnet-hides-in-blockchain-dns-mist-andremoves-cryptominer/. [Accessed: 14-Jan-2019].
[18] N. Johnson, “A developer’s guide to ENS concepts – The
Ethereum Name Service – Medium,” Medium Blogpost Web,
2017. [Online]. Available: https://medium.com/the-ethereumname-service/a-developers-guide-to-ens-concepts-7004eea8a073.
[Accessed: 13-Jan-2019].
[19] M. Bazzell, Open Source Intelligence Techniques:
Resources for Searching and Analyzing Online Information, 5th
ed. USA: CreateSpace Independent Publishing Platform, 2016.
[20] D. Murdoch, Blue Team Handbook: Incident Response
Edition. 2014.
[21] “Personal data breaches,” 2019.
72
jbba
the
[22] R. Jones and P. Collinson, “Identity theft warning after
major data breach at Ticketmaster | Money | The Guardian,”
The Guardian (Online), 2018. [Online]. Available: https://
www.theguardian.com/money/2018/jun/27/identity-theftwarning-after-major-data-breach-at-ticketmaster. [Accessed: 06Jan-2019].
[23] D. P. Act, “ICO lo Guidance on data security breach
management,” pp. 1–8, 1998.
[24] The Crown, “PML v Person(s) Unknown [2018]
EWHC 838 (QB) (17 April 2018),” 2018.
[25] N. Johnson, “ethereum/go-ethereum/name-registry Gitter,” Chatboard, 2019. [Online]. Available: https://gitter.
im/ethereum/go-ethereum/name-registry. [Accessed: 15-Jan2019].
[26] J. McDonald, “EthDNS: an Ethereum backend for
the Domain Name System,” Medium Blogpost Web, 2018.
[Online]. Available: https://medium.com/@jgm.orinoco/
ethdns-an-ethereum-backend-for-the-domain-name-systemd52dabd904b3. [Accessed: 13-Jan-2019].
[27] S. Dyson, W. J. Buchanan, and L. Bell, “The Challenges
of Investigating Cryptocurrencies and Blockchain Related
Crime,” vol. 1, no. 2, pp. 1–6, 2018.
Competing Interests:
None declared.
Ethical approval:
Not applicable.
Author’s contribution:
SD has prepared the manuscript in entirety.
Funding:
None declared.
Acknowledgements:
None declared.
The JBBA | Volume 2 | Issue 2 | October 2019
73
jbba
the
Peer-reviewed Research
OPEN ACCESS
ISSN Print: 2516-3949
https://doi.org/10.31585/jbba-2-2-(7)2019
A Blockchain Infrastructure for Transportation
in Low Income Country Cities, and Beyond
Simon J Herko
TravelSpirit Foundation, UK
Correspondence: siho@travelspirit.io
Received: 26 August 2019 Accepted: 29 August 2019 Published: 5 September 2019
Abstract
For our cities of tomorrow, it is essential that transport is organised in an efficient, resilient and equitable
way; enabling economic growth, social cohesion and minimising environmental impacts, including Climate
Change. In cities across the world, new flexible, sharing economy services are blurring the lines between
private and public transportation. However, these new transport modes are creating a “digital divide”
and lack the integration and co-ordination between other services. This is needed to create seamless and
sustainable travel options for people, including those belonging to vulnerable groups. This exploratory paper
examines the potential for Blockchain to play a pivotal role in addressing increasing congestion and pollution
in growing cities of developing countries. It draws on preliminary research into the role of Automatic Fare
Collection systems and related mobility market dynamics and trends in the cities of Cape Town, South Africa
and Dehli, India. By creating viable new digital infrastructure for Low Income Country Cities (LICCs), who
have less incumbent legacy systems, there is potential to establish a decentralised blockchain network across
these territories. There would also be scope for this network to be scaled further into wealthier countries,
through a secondary wave of adoption by Mobility-as-a-Service (MaaS).
Keywords: Blockchain, Distributed-hosting, Distributed-storage, Ethereum, Swarm, Forensics
JEL Classifications: A13, B41, C60, C71, D41, D43, D63, E24, E26, F02, F60, L14, L16, L17, L91, O18,
O33
1. The challenge of integrating mobility services
The proportion of the world’s population living in
urban areas will approach 66% by 2050[1], with much
of this growth coming from Low Income Country
Cities (LICCs).
However, transport in LICCs is fragmented, with
no common standards for booking, payment and
service delivery across different modes of transport,
competing services or across regions. The majority of
data is yet to be digitised and there are no mechanisms
in place to support data-sharing of movements and
assets. This leads to inefficient transport provision,
impacting economic and social well-being and
increasing congestion and pollution levels, including
unsustainable carbon emissions that are accelerating
Climate Change.
Cape Town, South Africa and Delhi, India, as suitable
real-world case studies for examining the potential
for blockchain to provide common infrastructure for
LICCs.
Our research into the Cape Town and South African
context was undertaken in collaboration with the
Greater Tygerberg Partnership (GTP). The GTP is
a not-for-profit entity funded by the City of Cape
Town, under the Transport and Urban Development
Authority. It serves as a facilitator to economic and
social renewal and collaborative efforts between the
private sector, civil society, academic institutions and
government for the benefit of the Voortrekker Road
Corridor (VRC). The VRC is an identified integration
zone and inward investment opportunity area,
comprising a population circa 350,000. It acts as the
second largest economic hub and busiest transport hub
in the Western Cape.
2. Developing the evidence base
We identified the high growth and congested cities of
The JBBA | Volume 2 | Issue 2 | October 2019
By researching the economic and social conditions
in Cape Town and the wider South African region,
75
jbba
the
we have developed key insights into the challenge of
bringing together transportation within and across
LICCs.
In South Africa, the proportion of individuals
benefiting from social grants rose from 12.7% in 2003
to 29.9% in 2016 [2]. The unemployment rate in South
Africa is 26.7% [3]. Access to transport is a key enabler
for accessing employment and education opportunities.
In the public and charitable sectors, transport funding
subsidies are often applied to the infrastructure, not
the user, creating a lack of transparency and often
inefficient utilisation of scarce resources.
Although improving, a high proportion of the
population (23%) are unbanked [4] and 63% are without
access to smartphones [5]. Credit card penetration is
at 17% and 65% of all transactions are made by cash.
54% of the population could be persuaded to switch
from cash to digital wallets only if they provided a
significant value-add over cash [6].
The following research insights are of particular
relevance to the opportunity for a blockchain-based
infrastructure intervention:
1.
2.
Competing transport businesses,
including high levels of “informal” minibus
taxi operations, make aggregation of services
and data highly challenging and encourage
disreputable operators. A commercially
agnostic platform that is easy and compelling
to adopt would therefore be highly desirable.
i.
In South Africa, the proportion of
the population who use informal minibus
taxis rose from 17.6% in 2003 to 22.4% in
2013. The proportion of mass-transit
commutes that are carried by minibuses is
67.5% [7].
ii.
Customer dissatisfaction with
minibuses is very high – 26.5%, compared to
3.9% for trains and 4.2% for buses.
In mass transit, the gap between fare revenues
collected and passenger numbers serviced is
too high, inhibiting further investments in
infrastructure and a negative impact on the
affordability of fares. Transit providers
require higher surety of payment.
i.
Affordability of mass transit has an
impact on poverty, inclusivity and the
economy [8].
ii.
Cape Town buses have introduced a
Smart Card and are enjoying growth [9].
iii.
Western Cape rail revenues are in
decline due to unreliable services and poor
funding [10].
For comparison with the South African research, we
The JBBA | Volume 2 | Issue 2 | October 2019
reviewed the transport landscape and relevant scientific
papers for the National Capital Territory (NCT) of
Delhi, India and its wider National Capital Region,
including the significance of the Metro for rapid transit
and active travel (i.e. non-motorised transport) for first
and last mile access.
Despite rapid growth of the Metro network, the lack of
integration of different modes has hastened the shift
towards private automobiles, including two-wheelers
and increasingly four-wheelers, for commuting and
other short distance travel. Over the course of 20152016 alone, the number of private motor vehicles
registered within the NCT of Dehli rose 10 percent,
from 8.8 million to 9.7 million, and the trend is expected
to continue without dramatic shifts in planning policy
[11].
Price and first/last mile connectivity are the major
influencing factors on choice of transport mode,
demonstrated in shifts from Metro (faster with poorer
last mile access, thus supplemented by auto hire) to bus
(slower with better last mile access) amongst middle
and lower income commuters following a Metro fare
hike over the 2016-2017 fiscal year [12] [13].
The following insights should help inform the design
and rollout of blockchain-based infrastructure for
enhancing the ease of multi-modal trips, including
the need to consider how funding for infrastructure
to support active travel can be integrated into the
conceptual framework:
1.
iii.
2.
Poor first and last mile connectivity of
public transit, especially the Metro, is
hampering the effectiveness of public
transit at reducing congestion and enhancing
mobility. Furthermore, transfers between
metro and bus for first/last mile trip segments
require separate fare payment methods, given
the Metro fare payment card is not widely
accepted by bus operators, despite pledges by
operators to install card readers [14].
i.
Offering convenience expected
of private motoring, especially door-to-door
service, can help reverse the decline in modal
share of public transit [15].
ii.
Physical facilities for active travel
tend to be substandard or absent, leading to
greater reliance on private cars, reduced street
space for walkers and cyclists, and declining
ridership of bus transit [16].
Funding for completion of discontinuous
footpaths, regular maintenance, and
prevention of encroachments are expected to
boost the propensity of active travel [17].
Mobility providers including bus operators,
ride-hailing and cycle-share platforms do
not coordinate with each other, leaving
76
jbba
the
certain areas of the city grossly underserved
relative to potential trip demand and, in the
case of separate companies operating buses
and metro trains, leading to lower than
expected ridership on new metro lines. Local
authorities are evidently aware of this
shortcoming, as demonstrated by the launch
of One Delhi mobile application for real time
journey planning covering both bus and
metro lines [18].
i.
The benefits of a common mobility
account have merited endorsement by the
highest levels of the central government,
including the Vice President in a call to
combat vehicular pollution through improved
ease of using public transport [19].
ii.
There is a desire to address the
lack of coordination by bringing ideally all
mobility providers under a common
organisational umbrella [20]. This desire, in
practice may not be achievable, pointing
towards a role of a blockchain infrastructure
to support a multi-stakeholder eco-system
with no centralised control.
iii.
A study for a cycle sharing system
that is ready for fares integration with other
transport modes is ongoing in South Delhi [21].
3. Our working hypotheses on a viable blockchain
Our research is motivated by a hypothesis that, less
hampered by legacy infrastructure and with strong
economic drivers for innovation, LICCs can leapfrog
high income countries on Intelligent Transport Systems
(ITS) [22]. This would imply:
1.
2.
LICCs do not have to depend upon large
programme budgets (which aside from the
expense can be often open to corruption)
and enter complex procurements to drive
forward and realise technology-driven
benefits.
There are ways for emerging economies to
innovate faster than developed markets and
play the role of pilot/pioneers in blockchain.
More specifically, there is an opportunity for a common
blockchain network infrastructure, for transport
booking, payments and subsidies, that, starting with
Low Income Country Cities (LICCs), would enable all
cities to enjoy the benefits of an integrated transport
system that is interoperable across competing services
and inter-regional borders.
As highlighted in our South African evidence base
(while applicable across much of the African, Indian,
Asian and South American continents) the informal
minibus taxi sector is a complex environment, while a
key ingredient to the transport mix of many LICCs. It is
The JBBA | Volume 2 | Issue 2 | October 2019
ripe for change, especially with regards to new payment
models and methods to optimise and integrate systems.
Following an examination of the current Intelligent
Transport Systems (ITS) landscape in LICCs, we
identified the most compelling blockchain use-case to
be for Automatic Fare Collection (AFC). A common
global and universal “open-loop” infrastructure,
enabled by blockchain, would replace the need for
bespoke and centralised back-office systems for each
city, and provide a common payment system for the
informal minibus taxi sector.
Both the European Bank for Reconstruction and
Development (EBRD) and World Bank have identified
the key barriers to adoption of “open-loop” accountbased systems outside the largest and most affluent of
world cities, such as Washington D.C. Boston, London,
Amsterdam, Vienna, Singapore, Hong Kong and Seoul
[23] [24]. They are the cost, time and effort required
to obtain the necessary banking security permissions
and the complexities of public sector led procurement
and implementation, which can take up to 5 years to
complete.
Advanced contactless card systems in London, Hong
Kong and Singapore are made possible by an effective
monopoly over transport provision and a well-funded
co-ordinating body (e.g. TfL’s operational budget is
over £6 billion per annum). They generally do not
extend to new collective transport innovations such
as car clubs, ride-hailing and bike-sharing; especially
if operated privately. In this respect, they are less
helpful operating models to replicate in emerging
market economies, with their higher levels of market
fragmentation, and where informal private minibus
services often dominate mass transit.
Research undertaken for the World Economic Forum
[25] articulates the case for improved integration and
interoperability in city transportation and its potential
for positive impact on global prosperity, equality and
the environment. Their hypothesis is that a centralised
global platform is required, risking, in our view,
bringing transport under the control of a small set of
data monopolies.
Our working hypothesis is that a permission-based
blockchain solution could provide users equitable and
open market access to transport services, with cashless,
subscription-based and/or subsidised payment
mechanisms. The solution would supersede “closedloop” AFC technology (e.g. smartcards) on buses and
trains and provide viable infrastructure to the informal
minibus taxi market, which represents circa 70% of all
mass transit trips in LICCs.
The rationale for adoption could be as follows, in terms
of benefits for different stakeholders.
77
jbba
the
Benefits to End Users:
3.
1.
4.
2.
3.
4.
5.
Cashless and trusted solution, improving
safety & security.
Access to user-based subsidies and microcredit worthiness.
Access without smartphone or contactless
banking.
Develop personal identity and data profile.
Roaming capability.
Benefits to Transport Providers:
1.
2.
3.
Surety of payment and uplift in fare revenues
collected.
Access to customer data and new markets.
Fair and trustworthy subsidy compensation
mechanisms.
Benefits to Cities:
1.
2.
3.
4.
Shift to mass-transit and reduced congestion/
pollution.
Platform for inward investment into public
transport infrastructure.
Easier to allocate subsidies in line with policy
objectives (e.g. active travel).
Affordable, easy to adopt AFC solution.
These benefits would be delivered through
decentralised, self-sovereign and interoperable
“mobility accounts”, hosted on a permission-based
blockchain [26]. This includes smart contracts to execute
commercial agreements, a shared set of business rules
for innovation in fares policy and blended financial
subsidies, including user-based subsidy.
The primary goal of the blockchain would be to provide
all LICCs with a common global ITS (Intelligent
Transport Systems) infrastructure, whose adoption
could be achieved organically, rather than procured.
We anticipate an open, transparent and crowd-based
governance structure and token economy that will
ensure transaction costs remain affordable.
4. Technical characteristics of a suitable
blockchain
In researching the feasibility of a blockchain solution in
the South African context, we identified the following
initial functional requirements to establish a viable
blockchain solution and adopted network:
1.
2.
Users (including the unbanked) to access
multiple transport services through a global
mobility account.
Account system interoperability and roaming
capability between transport operators,
modes and across regional borders.
The JBBA | Volume 2 | Issue 2 | October 2019
5.
6.
7.
8.
Manage rights and responsibilities of portable
personal data.
Support trusted multi-lateral commercial
arrangements between transport providers.
Provide low network latency, fast verification
and compatibility with low power devices.
Resilience to fraud and denial-of-service
attacks.
Commercially agnostic solution that can be
easily adopted by competing transport
providers and multiple regions and cities.
Close integration with existing infrastructure,
and a distributed share of transaction
revenues.
The decentralised delivery model of an open-source
and permission-based blockchain network would also
seek to address the high expense and long duration of
ITS procurements for AFC implementation.
Through dialogue with Hyperledger Working
Requirements Group, we have identified the
Hyperledger Indy and Hyperledger Sawtooth
development frameworks and modular open-source
codebase as the starting point [27] [28] [29]. To meet
the above functional requirements, we anticipate the
following future research and development actions:
4.1. Proof of Location within the Trusted
Execution Environment (TEE)
Existing Sawtooth framework accesses an efficient
Proof of Elapsed Time lottery algorithm for network
consensus, via a TEE developed by Intel. There is
opportunity to explore a new TEE that is optimised
for deployment in low power devices, including a Proof
of Location to improve network security and mobility
account operation.
4.2. Sharding / partitioning of the global state
Existing Sawtooth framework requires consensus
of the entire global state of transactions, with a total
ordering of every transaction. There is an opportunity
for our blockchain network to be partitioned or
‘sharded’ by location, to improve scalability and reduce
storage requirements. A new framework could be
developed to spawn multiple permissioned overlays
of Sawtooth, enabling a segmented-state management
protocol.
4.3. On-Chain Smart Contracts with “Seth”
There is scope to research into the capabilities of
the new Seth transaction family [30] as a means
for deploying Turing complete programs for
compensation, arbitration and concessionary reimbursement processes.
78
jbba
the
4.4. Linking via “Seth”, to a token-based economic
model
Hyperledger frameworks are optimised for the
application of permissioned blockchains within
business enterprise solutions using a centralised
platform business model. A design goal of commercially
agnostic, distributed revenues requires a higher level of
decentralisation.
There is scope for using Seth to bridge between the
Sawtooth permissioned framework and Ethereumbased tokens, to enable each city and transport
provider to operate their own node and gain a share in
the transaction revenues.
5. Beyond LICCs: global Mobility as a Service
(MaaS)
identified just over half a dozen published research
papers on blockchain for MaaS, from Germany,
Sweden, UK (by the Transport Systems Catapult and
TravelSpirit Foundation), Finland and the Netherlands
[26][34][35][36][37][38][39]. This growing evidence
base corroborates with our thesis that the scope of
MaaS to scale effectively, even within the European
market, where public policy and industry interest is the
greatest, is limited without the support of a common
blockchain infrastructure.
With a focus on wealthier markets, the papers we have
reviewed on the application of blockchain for MaaS
do not make direct references to LICC contexts. We
therefore believe we have developed a novel concept
for how to scale a blockchain network for ultimate
adoption as a MaaS solution in wealthier countries.
6. Conclusion
Mobility as a Service (MaaS) is a new disruptive business
model paradigm [31]. With an expected market size of
$1 trillion by 2030, it will empower users with hasslefree payment options and an integrated approach to
accessing public transport, flights, ferries and shared
economy services.
To scale globally, MaaS requires commercial
collaboration between a diverse and large transport
ecosystem [32], and affordable solutions for Low
Income Countries. Latest public policy and industry
thinking would suggest a growing consensus that
such collaboration would require a greater level of
“openness”, both culturally and technically, within the
city transport sector, than currently exists in most city
states [33].
Furthermore, to satisfy the demands of inter-regional
and international travel, supporting MaaS platforms
need cross-border functionality, facilitating “roaming”
across cities and countries. They must also integrate
various public, charitable, private and consumer
funding sources to enable effective investment in mass
transit and active travel infrastructure.
In a small collection of cities within wealthier countries,
that also enjoy advanced Open Data programmes (e.g.
Finland, Germany and the Netherlands), some MaaS
apps are already covering a full spectrum of collective
transport services. They have, in our view, limited
scope for widespread adoption due to the centralised
platform approach - i.e. the “unwanted third-party
aggregator”. This is a problem blockchain could
solve by enabling personalised aggregation to take
place direct to consumer, via a trusted, commercially
agnostic and decentralised infrastructure.
While there are many new blockchain solutions
appearing for shipping and logistics, the application
of blockchain for MaaS is in its infancy. We have
The JBBA | Volume 2 | Issue 2 | October 2019
The potential global impact of a blockchain-based
network infrastructure on the city transportation sector
is substantial. With blockchain, we can ensure a healthier
democratisation of the transport economy, that, based
upon liberal philosophies, will provide autonomy to
local and regional economies, strengthening global
collaboration and regional governance.
A case has been made for a global and universal
blockchain infrastructure, for the sharing of data
on movement and assets, designed with low income
economies and vulnerable groups in mind. It would
enable:
1.
2.
3.
Users’ access to different modes of transport
in an equitable and hassle-free way.
Assurance to transport operators on surety of
payment.
Cities with integrated solutions for tackling
congestion and targeting subsidies.
Through the work of both the European Bank for
Reconstruction & Development and the World Bank,
the economic and social case for delivering Automated
Fare Collection (AFC) technology in transportation
systems in emerging markets is already supported by
a comprehensive evidence base. Existing research on
AFC solutions consistently focuses on centralised
platforms and bespoke back-office infrastructure
for each city. It means the opportunity for a global
infrastructure, delivered through a decentralised and
networked route to market, has not been researched
and advocated to the same extent.
In wealthier countries Mobility-as-a-Service (MaaS)
is a new business model that integrates public and
private services together. Its level of adoption could be
limited without a supporting blockchain infrastructure.
By creating viable new digital infrastructure for
79
jbba
the
Low Income Country Cities (LICCs), who have
less incumbent legacy systems, there is potential to
establish a decentralised blockchain network across
these territories. There would also be scope for this
network to be scaled further into wealthier countries,
through a secondary wave of adoption by Mobility-asa-Service (MaaS).
To advance our understanding of this alternative vision
for global AFC infrastructure (i.e. technology that is
universal and enables a decentralised approach to the
management and orchestration of transport) we’d
recommend there to be:
1.
2.
Technology-based research and development
on the Hyperledger Project open-source
codebases.
Interventional pilots in Low Income Country
Cities, and research into the institutional,
commercial and funding mechanisms that
would be required to establish and scale this
kind of universal blockchain infrastructure.
References:
[1] “World Urbanization Prospects - Population Division,”
United Nations. [Online]. Available: https://population.
un.org/wup/. [Accessed: 16-Aug-2018].
[2] “General Household Survey, 2016,” Department Statistics
South Africa. [Online]. Available: http://www.statssa.gov.
za/?p=9922.
[3] “Quarterly Labour Force Survey – QLFS Q4:2017”
Department Statistics South Africa. [Online]. Available:
http://www.statssa.gov.za/?p=10884
[4] “Credit Card vs Cash in Africa – on the verge of
convergence” Cape Business News (2019). [Online] Available:
https://www.cbn.co.za/opinion/credit-card-vs-cash-in-africaon-the-verge-of-convergence/
[5] “In South Africa, Cash Is Still Right On The Money”
PYMNTS.com (2019). [Online] Available: https://www.
pymnts.com/cash/2017/south-africa-cash-usage/
[6] “Mobile wallets 'key' to SA e-commerce,” Fin24tech, 30Mar-2016. [Online]. Available: https://www.fin24.com/
Tech/Mobile/mobile-wallets-key-to-sa-e-commerce-20160330
[7] P. Lehohla, “Transport Series Volume I: Profile of nonmotorised transport users: In-depth analysis of the National
Household Travel Survey 2013 data,” Department Statistics
South Africa, rep. Available: http://www.statssa.gov.za/
publications/Report-71-03-01/Report-71-03-012013.pdf
[8] R. Carruthers, M. Dick, and A. Saurkar,
“Affordability of Public Transport in Developing Countries,”
The World Bank Group, rep., Jan. 2005. Available:
http://siteresources.worldbank.or g/INT TRANSPO
RT/214578-1099319223335/20460038/TP-3_
affordability_final.pdf
[9] “Golden Arrow hits profit targets.” Cape Business News.
(2019). [Online] Available: https://www.cbn.co.za/news/
golden-arrow-hits-profit-targets/
[10] C. Presence, “Metrorail losing paying customers by the
The JBBA | Volume 2 | Issue 2 | October 2019
millions, MPs told,” IOL News, 17-Apr-2018. [Online].
Available: https://www.iol.co.za/news/south-africa/westerncape/metrorail-losing-paying-customers-by-the-millions-mpstold-14494831
[11] S. Pillai, “Poor public transport behind Delhi vehicle
boom, say experts,” Hindustan Times, 20 Dec. 2016.
Available:
https://www.hindustantimes.com/delhi/poorpublic-transport-behind-delhi-vehicle-boom-say-experts/story4lbiZTogbYPofE2A57zcCJ.html
[12] A. Roychowdhury, “Towards Clean and Low Carbon
Mobility: Addressing Affordability and Scaling up of
Sustainable Transport,” Centre for Science and Environment,
rep., Sep. 2018. Available: http://cdn.cseindia.org/attachme
nts/0.21487000_1536054529_Anumita-Clean-low-carbonmobility-strategy-Sept.pdf
[13] C. Kumar and A. Ganguly, “Travelling Together but
Differently: Comparing Variations in Public Transit User
Mode Choice Attributes Across New Delhi and New York,”
Theoretical and Empirical Researches in Urban Management,
vol. 13, no. 3, pp. 54-73, Aug. 2018. Available: https://www.
jstor.org/stable/pdf/26472536.pdf ?refreqid=excelsior%3Ab
a21b9e6215c4ee0989116576282054a
[14] “Use metro cards on buses for discount,” The Hindu, 20
Oct. 2018. Available: https://www.thehindu.com/news/cities/
Delhi/use-metro-card-on-buses-for-discount/article25266914.
ece
[15] O. P. Agarwal, "Compulsion to Choice: How Can Public
Transport in India Be Transformed," Economic and Political
Weekly, vol. 54, no. 4, 26 Jan. 2019. Available: https://www.
epw.in/node/153650/pdf.
[16] “U. Nasim and V. Chattopadhyay, “Indian roads belong
to motorised vehicles, not cyclists or pedestrians,” Down to
Earch, 6 Nov. 2018. Available: https://www.downtoearth.
org.in/news/air/indian-roads-belong-to-motorised-vehicles-notcyclists-or-pedestrians-62049
[17] M. A. Alam, “Sustainable and Equitable Transport
System in Delhi: Issues and Policy Direction,” Asian Institute of
Transport Development, rep., 2015, Available: https://www.
unescap.org/sites/default/files/Article%202_Sustainable%20
and%20equitable%20transport%20system%20in%20Delhi.
pdf
[18] “Delhi commuters can now locate public transport better
with ‘One Delhi’ app,” Press Trust of India, First Post, 6
Mar. 2019. Available: https://www.firstpost.com/tech/
news-analysis/delhi-commuters-can-now-locate-public-transportbetter-with-one-delhi-app-6205941.html
[19] “VP calls for a public transport-centric approach to combat
growing vehicular pollution,” Government of India Press
Information Bureau, 3 May 2019. Available: https://pib.gov.
in/newsite/PrintRelease.aspx?relid=189925
[20] “Reimagining public transport in India,” KPMG, rep.,
Oct. 2017. Available: https://assets.kpmg/content/dam/
kpmg/in/pdf/2017/10/Reimagining-public-transport.pdf
[21] “Draft Detail Project Report on Public Bicycle Sharing
System for SDMC (South Delhi Municipal Corporation),”
Centre for Green Mobility, rep., Sep. 2015. Available: https://
shaktifoundation.in/wp-content/uploads/2017/06/PublicBicycle-Sharing-DPR-South-Delhi.pdf
[22] T. Yokota, “ITS Technical Note for Developing Countries,”
80
jbba
the
World Bank Group, rep., Jul. 2004. Available: http://
siteresources.worldbank.org/EXTROADSHIGHWAYS/
Resources/ITSNote1.pdf
[23] “On the move: delivering automated fare collection,”
European Bank for Reconstruction and Development, rep., Jul.
2017. Available: https://www.ebrd.com/documents/admin/
on-the-move-delivering-automated-fare-collection.pdf
[24] C. Monsalve et al, “Public Transport Automatic Fare
Collection Interoperability: Assessing Options for Poland”
World Bank Group and Korea Green Growth Partnership,
rep., Jun. 2016. Available: http://documents.worldbank.
org/curated/en/564001469009916441/pdf/107014WP-P148489-PUBLIC-Phase-2-Public-Transport-AFCInteroperability-Final-Report-June-10-2016.pdf
[25] J. Moavenzadeh and V. Padilla-Taylor, “Designing a
Seamless Integrated Mobility System (SIMSystem),” World
Economic Forum, rep., Jan. 2018. Available: http://www3.
weforum.org/docs/Designing_SIMSystem_Manifesto_
Transforming_Passenger_Goods_Mobility.pdf
[26] S. Ho et al, “TSio Protocol: The Internet of Mobility,”
Whitepaper, TravelSpirit Foundation, UK, rep., Dec. 2017.
Available:
https://travelspirit.foundation/wp-content/
uploads/2017/12/TravelSpirit-W hitePaper-TSioProtocol-v-6-1.pdf
[27] “Hyperledger Architecture Volume 1, Design Philosophy
and Consensus” Linux Foundation, rep., Aug. 2017. Available:
https://www.hyperledger.org/wp-content/uploads/2017/08/
Hyperledger_Arch_WG_Paper_1_Consensus.pdf
[28] “Hyperledger Architecture Volume II, Smart Contracts”
Linux Foundation, rep., Apr. 2018. Available: https://www.
hyperledger.org/wp-content/uploads/2018/04/Hyperledger_
Arch_WG_Paper_2_SmartContracts.pdf
[29] K. Olson, M. Bowman, J. Mitchell, S. Amundson, D.
Middleton, and C. Montgomery, “Hyperledger Sawtooth: An
Introduction,” Linux Foundation, rep., Jan. 2018. Available:
https://www.hyperledger.org/wp-content/uploads/2018/01/
Hyperledger_Sawtooth_WhitePaper.pdf
[30] “Seth Transaction Family Specification” Hyperledger
Project [Online, accessed Jun. 2019]. Available: https://
sawtooth.hyperledger.org/docs/core/releases/0.8/transaction_
family_specifications/sawtooth_burrow_evm_family.html
[31] M. Kamargianni and M. Matyas, “A Holistic Overview
of the Mobility-as-a-Service Ecosystem” University College
London, rep. Mar. 2017, Available: https://docs.wixstatic.
com/ugd/a2135d_8ec5294674a44129b04bcc99a324d1c5.
pdf
[32] P. Karjalainen, “Guidelines & Recommendations to
create the foundations for a thriving MaaS EcoSystem”, MaaS
Alliance, rep. Sep. 2017, Available: https://maas-alliance.eu/
wp-content/uploads/sites/7/2017/09/MaaS-WhitePaper_
final_040917-2.pdf
[33] S. Herko, S. Witzel, P. Karjalainen et al, “An Open Future
for Cities: Preparing cities for the necessary transformation and
organisational changes needed for an open future.” Whitepaper,
TravelSpirit Foundation, UK, rep., Dec. 2017. Available:
https://www.researchgate.net/publication/334524614_An_
Open_Future_for_Cities_Preparing_cities_for_the_necessary_
transformation_and_organisational_changes_needed_for_an_
open_future
The JBBA | Volume 2 | Issue 2 | October 2019
[34] D. Sümmermann, C. D. Öge, M. Smolenski, G. Fridgen,
and A. Rieger, “Open Mobility System”, Concept Paper,
MotionWerk GmbH, Fraunhofer FIT, TÜV Rheinland,
Germany, rep., Sep. 2017 Available: https://www.omos.io/
wp-content/uploads/whitepaper/OMOS_concept_paper.pdf
[35] P. Andersson and J. Torstensson, “Exploring the role
of blockchain technology in Mobility as a Service,”, Master’s
Thesis, Chalmers University of Technology, Gothenberg,
Sweden, rep., Nov. 2017. http://publications.lib.chalmers.se/
records/fulltext/252507/252507.pdf
[36] “Blockchain Disruption in Transport: Are You
Decentralised Yet?” Transport Systems Catapult, UK, rep., Jun.
2018. Avaliable: https://s3-eu-west-1.amazonaws.com/media.
ts.catapult/wp-content/uploads/2018/06/06105742/
Blockchain-Disruption-in-Transport-Concept-Paper.pdf
[37] A. Karinsalo and K. Halunen, “Smart Contracts for a
Mobility-as-a-Service Ecosystem,”, Conference Paper, VTT
Technical Research Centre of Finland, rep. Jul. 2018. Available:
https://ieeexplore.ieee.org/abstract/document/8431964
[38] J. Verheul, M. Mijnheer, and J. Ferwerda, “A new
Blockchain Platform Designed for the Future of Human
Mobility,”, Whitepaper, VMC, The Netherlands, rep.,
Jan.
2019.
Available:
https://vmc.ai/wp-content/
uploads/2019/01/whitepaper.pdf
[39] T. Nguyen, J. Partula, S. Pirttikangas, “Blockchainbased Mobility-as-a-Service,”, University of
Oulu,
Finland, Conference Paper, rep., May 2019. Available:
https://www.researchgate.net/profile/Tri_Nguyen43/
publication/333343145_Blockchain-based_Mobility-as-aService/links/5ce7e298a6fdccc9ddcabb45/Blockchain-basedMobility-as-a-Service.pdf
Competing Interests:
None declared.
Ethical approval:
Not applicable.
Author’s contribution:
SJH designed and coordinated this research and prepared the manuscript
in entirety.
Funding:
None declared.
Acknowledgements:
SJH would like to thank his colleagues at the TravelSpirit Foundation
and Iconic Blockchain for their support and encouragement over the
past 2 years, in particular to David Alexander, Giles K Bailey, Justin
Coetzee, Mike Fitzgerald, Dr Pieter J Fourie, Bren Hutchinson, Dr
Maria Kamargianni, Nathan King, Rob Mann, Gary Parkinson and
Yangbo Du. Also, special thanks to Johan Muller and Warren Hewitt
at the Greater Tygerberg Partnership and Mark Rathbone at Brabners
LLP.
81
jbba
the
analytical essay
A Review of fast-growing
Blockchain Hubs in Asia
OPEN ACCESS
ISSN Print: 2516-3949
https://doi.org/10.31585/jbba-2-2-(5)2019
Yu Wang, Jing Ren, Caroline Lim, Swee-Won Lo
School of Business, Singapore University of Social Sciences, Singapore
Correspondence: carolinelimsl@suss.edu.sg
Received: 28 June 2019 Accepted: 26 July 2019 Published: 9 August 2019
Abstract
The unique combination of social and economic factors has brought about a dynamic and rapidly-evolving
blockchain ecosystem in Asia. This paper systematically reviewed the development of four fast-growing
blockchain hubs in Asia, namely China, Japan, Singapore and South Korea using secondary data sources.
These countries are fast-growing based on the development of its digital, technological and regulatory
infrastructure, patent applications, cryptocurrency trading volume and Initial Crypto-token Offerings (ICOs)
activities. The review included insights into the different regulatory approaches, the blockchain startup
scenes, selected enterprise or government-backed projects, as well as the research and educational landscape.
Our findings suggested that the regulators, industry players, and academic institutions were purposeful and
deliberate in nurturing blockchain technology innovation. Future development would be dependent on the
regulatory, technological, as well as talent capability support unique to each blockchain hub.
Keywords: : blockchain, cryptocurrency, regulation, fintech, ICO
JEL Classifications: D02, G18, H11, O20, O32, O50
1. Introduction
support, funding, and investment capital.
The blockchain technology, with its properties that
distributes, disintermediates and decentralises, enables
value to be unlocked for peer-to-peer exchange. This
distributed ledger technology (DLT) can enforce
“trust” such that a mutually distrusting community can
collaborate and consent to a single version of the truth,
which implies trade and exchange can occur between
parties not known to each other.
Literature in the inter-organisational relationship,
including innovation and knowledge hubs, can be
parsimoniously organised into two paradigms – network
versus dyadic. Organisations in a network paradigm
developed long-term and trusting relationships that
were mutually reinforcing, and behaviours followed
socially accepted norms. Organisations in a dyadic
paradigm were opportunistic and sought to “maximise
cooperation and minimise conflicts” [2].
These features of blockchain have enabled applications
across different industries. Besides applications in
trade, stocks and securities exchange, banking and
finance, insurance, telecommunications, voting, health
care, government administration, social networking
and more, the blockchain technology holds promise to
financial integration and inclusion [1]. The potential of
blockchain is immense.
This paper is a systematic review of the development
and application of blockchain in four Asian countries
- China, Japan, Singapore and South Korea. These four
countries have leveraged the entrepreneurial fervour
and intensity of activities to shape themselves into
blockchain hubs, evidenced by our analysis of their
respective technological infrastructure, regulatory
The JBBA | Volume 2 | Issue 2 | October 2019
As technology hubs serve the community in addition
to organisational interests, we adopt the network
paradigm in our definition of a hub. We define a
hub as a locus of innovation and entrepreneurial
activities that fuels the local economy as coordinating
firms collaborate & develop capabilities supported
by different enablers. The literature described a
financial technology (“FinTech” in short) hub to be
characterised by sufficiently mature and developed
technology infrastructure, availability of talented and
receptive workforce (including investors, technologists,
financiers), established regulatory support (e.g.,
favourable tax rates), involvement of the academia,
government and enterprises in applied research and
investment (e.g., accelerators, incubators, mentorship
83
jbba
the
and seed funding), and a demand for FinTech (e.g.,
large volume of daily financial transactions, the need
to enhance consumer experience and improve business
efficiency, and the need for financial inclusion)[3].
A critical difference between FinTech and blockchain
is that the latter can be applied beyond the financial
industry. Blockchain allows parties with natural mistrust
to collaborate and consent to a single version of the
truth, thereby boosting business efficiency where crosscompany and cross-industry collaborations are needed;
it also holds promise to financial and social inclusion
[1]. With these in mind, we propose the enablers of
a blockchain hub to include the innovator group,
infrastructure readiness and programme, availability of
funding and capital, and the existence of demand for
blockchain applications.
As one of the emerging new technology, blockchain
drew investments in research and development of
large technology firms and technology startups. We
termed these technological firms and technology
startups as innovator group. The innovator group
represented technical capabilities to advance the
development and application of blockchain. Funding
and capital reflected the willingness and capabilities
of individual and institutional investors to support
technological development, especially for a relatively
new and less-understood technology like blockchain.
Infrastructure readiness and programme would
facilitate the development of new and innovative
technology. Apart from network and technology
readiness, a friendly regulatory environment and
availability of skilled talent pool would encourage and
facilitate technological innovations. The economic and
socio-ecological contexts could generate demand for
blockchain applications; a politically stable economic
environment could serve as a landing for blockchain
projects addressing trust issues between and across
partners. Similarly, the socio-ecological contexts could
provide the impetus for financial and social inclusion.
This review intends to improve the understanding of
blockchain development in different jurisdictions and
contribute to current literature about blockchain in the
Asia region. In the subsequent section, we explained
the rationale of selecting the Asian countries, namely
China, Japan, Singapore, and South Korea. In the
third section, we analysed and compared the status
of blockchain development in each country and intercountry. We concluded this paper with a discussion on
the implications for future research and practice.
2. Scope of Review
In this section, we explained the selection criteria of
the four countries in Asia, beginning with a description
and analysis of four key enablers of a blockchain
hub namely innovator group, infrastructure and
The JBBA | Volume 2 | Issue 2 | October 2019
programme, funding and capital, and demand.
2.1. Innovator Group
The actors in a network were central to the activities
of a blockchain hub. These actors included large
technology firms, blockchain startups and related
technology unicorns who drew venture capital and
drove spending in research and development (R&D).
As of July 2018, China, Japan and South Korea
were three Asian countries with the most number of
Global 500 companies in the top ten of Fortune 500.
Companies in Global 500 included large technology
firms with research and investments in blockchain
projects. In China, 46 of the 120 companies were
involved in blockchain development representing
sectors like banking, energy, IT, and motor. Japan’s Sony
and Fujitsu were also actively involved in blockchain
projects. Almost all of the South Korea IT and motor
companies in the Fortune 500, like Samsung, LG,
and Hyundai, were exploring their own blockchain
platforms.
Following the statistics of total blockchain-related
patents filed globally by IPR Daily and Cintelliq, China
filed the highest number of blockchain patents (41%),
followed by the United States (32%)i. As of August
2018, Chinese companies occupied more than half of
the top 100 companies globally for patents application
on the blockchain (57 out of 100). Technology firms
among them included Alibaba of China, Sony and
Fujitsu of Japan and Coinplug of South Korea. The
European Patent Office (EPO) showed a steady
increase in patents granted normalised by population
between 2009 to 2018 from countries like China,
Japan, and South Korea. The year-on-year change of
these three Asian countries over the period exceeded
the figures reported for thirty-eight member states of
EPO (including 28 states of the European Union).
Investment in blockchain startups represented the
market expectation for blockchain development in
the long-term. As of 31 March 2019, there were 333
technology unicorns worldwide between 2010 to 2019ii.
Among them, more than one third (124) originated
from Asia, of which China accounted for 89 unicorns.
Another 15 unicorns originated from India, eight from
South Korea, five from Indonesia, two from China
SARiii Hong Kong, one each in Japan and Singapore.
Nine unicorns among them were blockchain-based in
the areas of FinTech and cryptocurrency. These nine
unicorns included six from China (e.g., Bitmain, Tiger
Brokers), two from India (One97 Communications,
PolicyBazaar), one from South Korea (Viva Republica).
In China, where the regulation prohibited fund-raising
through initial crypto-token offering (ICO), technology
firms would become one of the main funding sources
for blockchain startups.
84
jbba
the
National spending on research and development
(R&D) fueled the growth of the innovator group.
According to data from the UNESCO Institute of
Statistics published by the World Bank, high-income
countries spent on average 2.36 per cent of GDP on
R&D for science and technology between 2000 to
2016iv. Across countries in Asia, Japan’s R&D spending
had consistently exceeded the average figure of highincome countries. The same index in South Korea
rose steadily since 2000 to more than double that of
high-income countries from 2012 onwards. Singapore’s
R&D spending as per cent of GDP approximated close
to high-income economies. China, on the other hand,
did not perform close to other high-income countries
on this index, but its R&D spending rose significantly
from 0.89 per cent in 2000 to 2.11 in 2016. Other Asian
countries performed below average relative to the rest
of the world or when compared against high-income
economies.
2.2. Infrastructure and Programme
The critical determinants of a blockchain hub
included both digital and regulatory infrastructure of a
country. We reviewed the digital infrastructure in two
aspects, namely the network readiness and technology
readiness. The World Economic Forum published the
Global Information Technology Reportv to assess the
state of network readiness of 139 economies from the
annual executive opinion survey. The index evaluated
the quality of regulatory and business environment,
information and communications technology
(ICT) readiness in terms of affordability, skills and
infrastructure, the role of the government, business
sector and population as well as the environment,
readiness and usage. Countries in Asia, including
China, Malaysia, Mongolia, Sri Lanka, and Thailand,
demonstrated steady improvements from 2012 to
2016. Across the drivers of network readiness in 2016,
Singapore performed better than other advanced
economies in business and innovation environment,
skills, government usage, and social impacts. Taiwan
performed the best in mobile network coverage and
internet bandwidth infrastructure. Singapore was
ranked first in 2015 and 2016; Japan was the other Asian
country ranked in the top 10 of network readiness; the
others in top 10 were made up mostly of European
countries. Meanwhile, South Korea hovered around
10th to 13th position between 2013 to 2016 and was
ranked 13th in the most recent published ranking.
We further referenced the technological readiness
ranking of eighty-two countries published by The
Economist Intelligence Unit (EIU) as part of their
medium- and long-term forecasts of the world’s
largest economies. The EIU assessed performance
across three categoriesvi: access to the internet, digital
economy infrastructure and openness to innovation.
The index ranked each country for the historical
The JBBA | Volume 2 | Issue 2 | October 2019
period from 2013 to 2017 and forecasted change in
performance for the period 2018 to 2022. Countries
in Asia ranked in the top 10 included Singapore, Japan,
South Korea, and Taiwan. Meanwhile, EIU forecasted
improvements in technological readiness for these four
countries/region and Hong Kong. In particular, the
forecast projected Singapore to be ranked similarly to
Australia and Sweden in technological readiness by the
period 2018-2022.
Besides technological readiness, a workforce that
was ICT-enabled, trained and skilled in blockchain
development would more effectively contribute to
the completion of innovative blockchain projects.
According to the Global Startup Ecosystem Report
[12], cities such as Beijing, Shanghai, Singapore,
Bangalore, and Hong Kong possessed high-quality
technology talents (such as top developers on GitHub
and software engineers) that were relatively inexpensive
compared to the US and European countries.
Workforce policies that encouraged science, technology,
engineering and mathematics (STEM) training as well
as lifelong learning in related skills and knowledge
enhanced the adoption and development of blockchain
technology. Preliminary results by the OECD reported
that workforce capabilities and training received were
associated with higher digital adoption, such as in
cloud computing technology [4].
According to another OECD survey of adults aged
between 16 and 65 in 35 economies in 2012 and 2015
Singapore (62%) and South Korea (60%) performed
above the OECD average (55%) for adult participation
rates in structured training. In the same survey, Japan
performed below the OECD average.
Apart from workforce policies, lower regulatory costs
and simplified compliance procedures would expedite
the process of starting a business, and these would be
attractive factors for blockchain startups. As of 2018,
the shortest time needed to start a business was in
Hong Kong (1.5 days), Singapore (2.5 days), South
Korea (4 days), Thailand (4.5 days), and Sri Lanka (9
days). In contrast, Cambodia took the longest time (99
days), followed by Laos (67 days), India (29.8 days),
Philippines (28 days), China (22.9 days), and Vietnam
(22 days).
On the other hand, the cost of business startup
procedure (in per cent of gross national income per
capita) including all official fees and legal costs was the
lowest in Singapore (0.5%), followed by China (0.6%),
Brunei, Hong Kong (1.1%), and Mongolia (1.4%) [5].
2.3. Funding and Capital
Funding and capital played an essential role in fueling
the growth of blockchain startups and thus, the
85
jbba
the
development of blockchain technology. Blockchain
startups commonly raised funds through loans,
donations, traditional venture capital, and ICOs [6].
Among them, ICO represented a unique fundraising
method as it allowed blockchain startups to raise
funds from the community at a relatively early stage.
The value of ICO reflected the financial support that
blockchain startups might receive and the size of the
blockchain community within a region. In Asia, Hong
Kong and Singapore were popular destinations for
many blockchain entrepreneurs considering ICOs,
after the prohibition of ICOs in mainland China and
South Korea. As of October 2018, 8.14 per cent of
ICOs globally occurred in Singapore and 2.81 per cent
in Hong Kongvii. Vietnam and Japan led in the traffic
to ICO listing websites globally, followed by the US
and the United Kingdom. Other countries among the
top 10 were China and South Korea [7]. Search volume
from Google Trends suggested Asian countries and
regions like China, South Korea, Singapore, and Hong
Kong, to be among the top 10 worldwide for ‘ICO’ viii.
By the volume of venture capital and private equity
activities, Hong Kong, Japan, and Singapore were
among the most attractive countries/regions for
venture capital and private equity globally. They were
ranked 4th, 5th and 6th respectively, after the US, UK
and Canadaix. Other Asian countries in the top 30 list
included Malaysia, China, South Korea, Thailand, and
India.
The trading volume of Bitcoin served as another
indicator for the scale of capital in the country; an
indicator of the cryptocurrency market that drew
investors’ attention. Asia accounted for almost a third
of cryptocurrency transactions globally. According to
LocalBitcoins.com, a decentralised bitcoin exchange
website, the trading volume was US$6.3 billion in July
2018, of which Asia contributed 32.8 per cent of the
global volume of bitcoin traded [8]. Furthermore,
statistics of the most-traded national currencies for
bitcoin showed a consistent trend - Japanese Yen
accounted for around 40 per cent of the global total
bitcoinx volume, second to only US Dollar, with
national currencies of other Asian countries such as
South Korea, Indonesia, Thailand, Singapore, and
Vietnam also among the top 20.
2.4. Demand
We proposed the demand for DLT as another
enabler of a blockchain hub. Demand could stem
from an economic infrastructure where multinational
organisations converge as well as the socio-ecological
landscape. DLT solved trust in digital asset transactions
between businesses or between businesses and
consumers without a central administrator [9].
The JBBA | Volume 2 | Issue 2 | October 2019
The economic infrastructure of a country that
would stimulate demand for blockchain applications
included countries with active participation in the
global production networks. We considered the
global value chain participation since that reflected
the relative positions of different economies in the
global production networks. Forward or backward
participation ratios measured each country’s
participation in the global value chain. Forward
participation ratio measured participation through
the supply side, i.e., the extent that “an economy’s (or
economic sector’s) locally generated value-added was
embedded in the production of other economies” [5].
Backward participation ratio measured participation
through the demand side and “denotes the foreign
value-added contribution to an economy’s (or
economy-sector’s) exports” [5].
In Asia, Singapore led in the use of foreign inputs
in the production of its exports, with a backward
participation ratio of close of 60 per cent, followed by
Vietnam, Taiwan (China), South Korea, and Malaysia.
Brunei took the pole position in forward participation
ratio at slightly more than 80 per cent, followed by
Laos, Indonesia, Philippines, and Malaysia [5].
The need to resolve trust issues to boost efficiency and
save cost through dis-intermediation using blockchain
applications would emerge from countries with foreign
direct investments (FDI). In 2017, the top three
recipients of FDI in Asia was China, Hong Kong, and
Singapore, followed by India, Indonesia, Japan, South
Korea, Vietnam, and the Philippines [5].
A significant population of the working class in Asia
earned their living outside of their home countries and
remitted their earnings back to their home countriesxi.
In 2017, countries in the Asia Pacific region received
US$266 billion in remittances [5]. Globally the top
three remittance recipient economies were in Asia,
namely India, China, and the Philippines. There were
two pain points to be addressed – trusted peer-topeer funds transfer and remittance fees. Firstly, a large
population in Asia were unbanked, although a majority
of them owned mobile phones connected to a 3G or
4G network [10]. Secondly, the average cost of crossborder remittance fees for sending US$200 remained
high at seven per cent [11].
The socio-ecological context supported by a
favourable regulatory environment and well-developed
technological infrastructure laid the foundation for
blockchain projects that would boost production
efficiency or solve financial inclusion within the
economy as well as the neighbouring region.
2.5. Rationale
Using national spending and patents filed as measures
86
jbba
the
for the impact of the innovator group, countries like
China, Japan, Singapore, and South Korea, performed
better than other Asian countries. Moreover, China
had the highest number of technology firms and
technology unicorns. Technology firms provided
alternative funding sources for blockchain startups.
While China’s regulation prohibited cryptocurrency
trading and ICOs, trading volume in bitcoin and other
cryptocurrencies were high in Asian countries or
regions like Hong Kong, Indonesia, Japan, Singapore,
South Korea, Thailand, and Vietnam. Activities and
interest in ICOs were also high. Drawing parallel from
investment trends reported in Europe, investors would
likely invest more in ICOs in investment destinations
that appealed to VC and PE funds; destinations
included China, India, Japan, Malaysia, Singapore,
South Korea, and Thailand [12].
The readiness of technological infrastructure and
programmes were critical to support innovation in
blockchain technology. Singapore, Japan and South
Korea led in technological readiness evident from
our earlier analyses. Additionally, China, South Korea,
Japan and Singapore were ranked in the top 10 by the
2018 Global Digital Economy Development Index
that assessed the overall digital economy development
in more than 150 countries and regions worldwide [13].
The enabling factors of a hub namely the availability of
talent pool in the innovator group, funding and capital,
infrastructure and programme as well as demands
for business efficiencies or financial inclusion, allow
blockchain projects to flourish. From the performance
of these enablers, we identified China, Japan, Singapore
and South Korea to be fast-growing blockchain hubs in
Asia relative to other countries in the region.
3. Analysis by Country
We analysed the status of blockchain development
in these four countries from four aspects: regulations
and standards, characteristics of blockchain startups,
enterprise- and government-backed blockchain
projects, and research.
We extended our study of regulations to those for
cryptocurrencies-related activities such as ICOs
and cryptocurrency exchanges, to present a more
comprehensive view of the state of blockchain in
the country. Given alternative supporting resources
through retail and institutional investors, enterprises,
or the government, we conceded that the prohibition
or absence of regulations for ICOs or cryptocurrency
exchanges did not imply the lack of support for
blockchain projects.
3.1. China (Mainland)
3.1.1.
Regulations and standards
The JBBA | Volume 2 | Issue 2 | October 2019
With the support of the Chinese government and
available skilled workforce, the digital economy
was a primary driver of economic growth in China
contributing 30.3 per cent of China’s GDP [14]. Before
the state intervention on cryptocurrency trading,
Chinese investors invested heavily in cryptocurrencies
without knowledge of the market nor the underlying
mechanism [15].
To mitigate financial risks brought about by the
volatility of Bitcoin and cryptocurrencies, seven
authorities in China issued a joint announcement in
September 2017 to prohibit onshore and offshore
platforms related to ICOs and cryptocurrency trading
[16]. Nevertheless, the prohibition did not extend to
the development of bitcoin’s underlying technology
– blockchain. Instead, the Chinese government took
the lead in advocating the development of blockchain
technology through a series of initiatives. In December
2016, the State Council of China included for the first
time blockchain technology in the 13th Five-Year Plan
to build a national strategic technological advantage. In
June 2017, the central bank of China, People’s Bank
of China (PBoC), expressed their intent to promote
research and application of advanced technologies
such as blockchain and artificial intelligence (AI) in the
five-year development plan for the financial industry
[17]. Four months later, the Ministry of Industry
and Information Technology (MIIT) released a
white paper on China’s Blockchain Technology and
Application Development, the country’s first official
guidelines on the blockchain. Additionally, the State
Council issued a mandate to the local government to
accelerate the development of technologies, including
blockchain in May 2018 [18]. Most recently in April
2019, the regulator, Cyberspace Administration of
China, endorsed 197 blockchain service providers; the
endorsement gave confidence to the industry for the
deployment of their services.
To nurture this vibrant technology and innovation
hub, the Chinese government further introduced
regulatory guidelines for technology applications. A
FinTech committee was set up by PBoC to strengthen
the application of RegTech (regulation technology that
addresses regulatory challenges in financial services
using innovative technologies such as big data, AI, and
blockchain) [19]. FinTech startups, namely Gingkoo
and PeerSafe, have introduced regulatory frameworks
and solutions on blockchain for domestic government
and banks.
3.1.2. Blockchain startups
Despite the prohibition of ICOs, new blockchain
companies in China outnumbered that of the US in
2016; these Chinese blockchain startups accounted for
28 per cent of new startups globally [20]. Furthermore,
as at the end of 2017, China submitted the most
87
jbba
the
patent applications for blockchain with 550 patent
submissions, nearly twice that of 284 applications from
the US [21].
There were over 400 blockchain startups in China
as of March 2018, according to data from ITJuzi
and BlockData. Instead of blockchain solutions,
infrastructure and social media, the majority of
Chinese blockchain companies focused on technology
applications for the financial industry, and on traditional
economic sectors like: agriculture, manufacturing,
supply chain and logistics. Seventy-eight per cent of
these operated out of Beijing, Shanghai, Shenzhen, and
Hangzhou, which suggested an agglomeration effect.
Wanxiang Blockchain Labs, a non-profit research
institution funded by China Wanxiang Holding setup
the first blockchain research centre in Shanghai in 2015
to pioneer research, development, and application of
the technology. Projects like Bubi Chain and Juzix
worked on developing blockchain infrastructure to build
the ecosystem. In the meantime, many startups have
proposed blockchain-based commercial platforms to
solve real-life issues. For example, Qulian Technology
provided enterprise-level blockchain products and
application solutions such as supply chain finance and
traceability, digital certificate, and energy assets.
3.1.3. Enterprise- and government-backed projects
While startups experimented with new and novel
ideas associated with blockchain, existing industry
leaders explored potential solutions using blockchain
technologies. The three Internet tech giants in China,
Baidu, Alibaba and Tencent (collectively known as
BAT), have started projects related to blockchain.
Baidu became a member of an open source industry
blockchain initiative named Hyperledger in October
2017. Baidu has launched its blockchain-as-aservice (Baas) platform, and Alibaba has successfully
applied blockchain in areas such as healthcare and
e-commerce. Alibaba built a supply chain tracking
system using blockchain technology together with PwC
in March 2017. In the same year, Tencent invented the
TrustSQL platform to develop blockchain applications
and provide enterprise service solutions. Tencent
established the first digital private bank in China,
WeBank. Blockchain Open Source (BCOS) platform
was the first commercial blockchain technology
platform to be introduced in China jointly by Wanxiang
Blockchain Labs and WeBank. Ant Financial, the
financial affiliate of Alibaba, and Baidu published
a white paper to illustrate their blockchain strategic
roadmap in 2018. Besides BAT, other corporations
like Huawei, Xunlei and JD.com (logistics tech giant)
have incorporated blockchain into their firms’ strategic
plan and released white papers related to blockchain
projects.
The JBBA | Volume 2 | Issue 2 | October 2019
To support blockchain startups, the municipal
governments of Chinese cities launched blockchaindedicated funds. Example, Xiong’An Global Blockchain
Innovation Fund equivalent to US$1.6B was launched
in Hangzhou in April 2018 , and a district government
of Nanjing city launched another blockchain fund of
US$1.4B in July 2018xii.
3.1.4. Research
Research in technology has been a focal area for the
Chinese national and local government bodies. The
volume of blockchain related publications and the
number of research institutes increased rapidly in 2016.
The number of blockchain research institutes that
opened in the first four months of 2018 was equivalent
to those that opened in the whole of 2017, which
was three times the number in 2016 [22]. Apart from
the government-led independent research institutes,
corporations and universities established more than 90
per cent of the research institutes in China.
In 2017, the PBoC launched the Digital Currency
Research Institute that focused on the development
and research of digital currencies. So far, the Institute
had filed more than 63 patent applications, according
to China's State Intellectual Property Office (SIPO)
[23, 24]. Its ultimate goal is to introduce a state-backed
virtual currency that would combine blockchain-based
cryptocurrencies with the existing monetary system.
3.1.5. Hong Kong
Hong Kong has been zoned a special administrative
region compared to other cities in China mainland.
Under the “one country, two systems” constitutional
principle, Hong Kong maintained its own governmental
system, legal, economic and financial affairs, including
trade relations with foreign countries. This separate
constitution enabled Hong Kong to play a vital role in
promoting blockchain development in China and even
the rest of Asia.
The Hong Kong government defined cryptocurrencies
as “securities”, similar to that of the US Securities
and Exchange Commission (SEC). ICOs and
cryptocurrency came under the Securities and Futures
Commission (SFC). In November 2018, SFC defined
a regulatory framework for trading, managing and
distributing cryptocurrencies [25] which would
facilitate the maturity of the regulatory framework in
the long run for digital assets.
Meanwhile, the Hong Kong government supported
the development of blockchain technology and related
projects. As early as November 2016, the Hong Kong
Monetary Authority (HKMA), jointly with Hong Kong
Applied Science and Technology Research Institute
(ASTRI), released a technical white paper on DLT. In
88
jbba
the
the same month, HKMA-ASTRI FinTech Innovation
Hub was launched to provide a neutral ground for
the FinTech industry and startups in Hong Kong
[26]. Later in March 2017, HKMA and seven banks
commercialised a blockchain-based trade finance
platform which was officially launched by HKMA on
31 October 2018, named “eTradeConnect”. Developed
by a consortium of twelve major banks in Hong Kong
including HSBC and Standard Chartered Bank [27],
eTradeConnect aimed to improve trade efficiency,
improve trust among trade participants, reduce
risks and facilitate trade counterparties by leveraging
digitalisation and blockchain technology.
HKMA collaborated with other regions and countries,
including Singapore and Abu Dhabi. HKMA and
Monetary Authority of Singapore (MAS) have signed
and exchanged a Co-operation Agreement in 2017 to
strengthen co-operation on FinTech [28] such as the
Hong Kong Trade Finance Platform (HKTFP), an
HKMA-led trade finance proof-of-concept based on
DLT. In June 2018, HKMA worked with regulators
in Abu Dhabi to develop a cross-border trade finance
system using DLT [29]. These collaborative initiatives
revealed the economic, technological and geographical
advantages and capabilities of Hong Kong in the
development of blockchain.
Besides government-run FinTech and blockchain
projects, financial institutions, research centres and
various startups have landed their projects in Hong
Kong. Example, Ant Financial of Alibaba Group
joined GCash of Philippines to launch the world’s first
blockchain-based remittance service built on Alipay
blockchain technology [30]. The Bank of China Hong
Kong developed a blockchain-based system for real
estate appraisals to avoid mortgage fraud [31]. Over 20
various FinTech startups emerged from Hong Kong.
Example, startup Crypto.com released Asia’s first
cryptocurrency Visa card in Singapore in September
2018 and subsequently in the US in November.
Blockchain research centres or laboratories by Deloitte
and China Blockchain Application Research Centre
were established in Hong Kong. Hong Kong University
of Science and Technology received US$20 million
research grant for blockchain payment system. To
attract blockchain talents, the Hong Kong government
effected special immigration policy to expedite
immigration for job seekers with blockchain expertise.
It released a talent list on 28 August 2018 for eleven
professions including blockchain technology [32]. The
Hong Kong’s Quality Migrant Admission Scheme
(QMAS) that administered points-based tests for job
seekers in Hong Kong accorded lower entry barrier to
those with blockchain expertise.
3.2. Japan
3.2.1. Regulations and standards
The JBBA | Volume 2 | Issue 2 | October 2019
Japan was the first country that recognised bitcoin as a
legal payment option and has a national system to regulate
cryptocurrency exchanges. A cryptocurrency exchange
registered with the Financial Services Agency (FSA)
of Japan was considered a legitimate entity in Japan.
To-date, there were sixteen approved cryptocurrency
exchange operators in Japan and cryptocurrencies on
these exchanges could be exchanged for fiat monies or
alternative cryptocurrencies. Basic guidelines for ICOs
that focused on investor protection and anti-money
laundering were released by a research group led by
academics at Tama University [33, 34]. Still under
deliberation by the FSA, many anticipated that these
guidelines would eventually pass as a law in Japan.
The regulatory landscape in Japan for cryptocurrency
exchanges and ICOs paved a promising future for
the development of blockchain projects. In 2016, the
Ministry of Economy, Trade and Industry (METI)
engaged Nomura Research Institute to survey domestic
and international blockchain applications [35]. As
an outcome of the survey, METI published the first
version of evaluation templates to assess blockchain
applications and completed the first evaluation for
blockchain applications in healthcare, supply chain
& logistics, and smart property in 2018 [36, 37].
The process uncovered legal and technical issues of
blockchain applications for respective industries.
3.2.2. Blockchain startups
Compared to the exponential growth in bitcoin trading,
the number of blockchain ventures in Japan was small
relative to other regions in Asia. In 2016, among the
167 FinTech startups in Japan, there were only 20
blockchain-related businesses [38]. This phenomenon
in Japan could be attributed to the stronger public
sentiment on the use of bitcoin for official payment
than the application of the underlying blockchain
technology.
Nevertheless, the blockchain startup scene in Japan
was encouraging with generous support from the
Japanese government. In 2017, METI sent three
blockchain startups to the US as part of the Silicon
Valley-Japan Bridge Project [39]. In the private sector,
major industry players or financial institutions have
announced investment funds, incubators or coworking space for blockchain startups. For example,
SBI Holdings, a global rank-1 corporate blockchain
investor, invested approximately US$460 million in AI
and blockchain fund [40]. Mizuho Financial Group,
one of the three major financial institutions in Japan,
sponsored Neutrino, the first blockchain co-working
space in Japan [41]. In short, blockchain startups in
Japan received assistance and mentorship from the
government, enterprises and large financial institutions.
Foreign startups in Japan had similar access to funding,
facilities and advice on regulatory matters [42].
89
jbba
the
3.2.3. Enterprise- and government-backed projects
Enterprise-backed projects in Japan focused on
building applications in financial services and supply
chain. The Japan Exchange Group, Inc. (JPX) tested
the streamlining of processes in the securities market
and ownership registry through a proof of concept
(POC) with six other financial institutions in Japan
[43]. NTT Data, one of the largest information
technology companies, collaborated with Mitsubishi
UFJ Financial Group (MUFG) and Singapore’s
National Trade Platform to launch a blockchain POC
that would foster trade between Singapore and Japan
[44]. With Skuchain, NTT DATA developed a business
collaboration platform for Japanese manufacturers to
boost supply chain efficiency [45].
The three financial institutions in Japan have
implemented blockchain projects to streamline trading,
payment, and other financial services. Mizuho Financial
Group and Sumitomo Mitsui Financial Group (SMFG)
respectively launched blockchain to streamline trade
transactions [46,47]. On the other hand, MUFG
introduced its MUFG Coin for commercial and retail
customers, as well as to incentivise its employees to
reduce overtime hours for healthier lifestyles [48].
At the government level, Japan’s New Energy and
Industrial Technology Development Organisation
(NEDO) under the instructions of METI, worked
on several blockchain-based projects. Among them
included the use of internet-of-things (IoT) to
streamline infrastructure for trade information sharing,
where NEDO operated in partnership with NTT Data.
The Ministry of Internal Affairs and Communications
explored the application of blockchain solution to
process government tenders and introduced a roadmap
for incorporating DLT in e-government services in
2018 [49].
The Blockchain Study Group, established by Deloitte
Japan, Mizuho Financial Group, SMFG and MUFG,
promoted blockchain adoption and education. The
focus of this study group was to conduct studies
on interbank payment and a Know-Your-Customer
advanced platform. The Japan Blockchain Association
facilitated collaboration and conversations between
blockchain startups and the Japanese government.
Other associations such as the Japanese Bankers
Association whose members comprised banks, bank
holding companies and bankers’ association analysed
the implementation of blockchain for financial services
[50].
3.2.4. Research
Major financial institutions and universities led the
blockchain research and development landscape
in Japan. In April 2016, the Bank of Japan (BOJ)
The JBBA | Volume 2 | Issue 2 | October 2019
established the FinTech Centre in its Payment and
Settlement Systems Department [51]. The BOJ
conducted a joint research project entitled “Stella” with
the European Central Bank (ECB). The Stella project
evaluated the performance of using Hyperledger
Fabric to facilitate large value payments and the
“delivery versus payment” environment using single
and cross-ledger platforms, respectively [52,53]. In
academia, Japan has five university nodes in the BSafe.
network that promoted scientific and interdisciplinary
social and economic research [54]. In addition, the
more notable academic initiatives include the teaming
of University of Tokyo and University of Aizu with
two industry organisations to study smart currency
[55], the establishment of BASE Alliance between
Keio University and University of Tokyo [56], and the
establishment of Blockchain Research Lab at Kyushu
Institute of Technology [57].
3.3. Singapore
3.3.1. Regulations and standards
A confluence of factors – global financial centre, publicprivate partnerships, engagement and consultation,
public education – had shaped Singapore’s emergence
as a leading technological hub of the world.
On the regulatory front, MAS, the central bank of
Singapore, adopted a nurturing stance of regulation,
one that was conciliatory but strict. In 2016,
MAS introduced a “regulatory sandbox” to foster
experimentation of innovative business models for
financial institutions and FinTech companies [58].
The MAS did not regulate Crypto-tokens, digital tokens
or virtual currencies. Instead, the MAS regulated
activities on the use of virtual currencies that would fall
under the regulator’s ambit, such as money laundering
and terrorism financing. Digital tokens structured like
securities in ICO, also known as equity tokens, must
satisfy the requirements of the Securities and Futures
Act (SFA). Cryptocurrency exchanges were regulated
under the SFA by the MAS when such exchanges
allowed the listing and trading of digital tokens.
Although the MAS had not issued specific legislation
related to ICOs, it monitored activities and
developments in the space carefully. Example, MAS
issued a directive and warning to an ICO issuer to
terminate its digital tokens offering in May 2018 as
MAS assessed those digital tokens to represent equity
ownership and they failed to satisfy SFA requirements
[59].
To upskill the workforce in digital skills and promote
lifelong continuous learning, Singapore’s Ministry
of Education launched a nationwide SkillsFuture
Initiative. This Initiative provided subsidies on
training and courses, including courses on blockchain.
90
jbba
the
Singaporeans and permanent residents received up to
seventy per cent in fee subsidy and to a maximum of 90
per cent subsidy for those aged above 40. Institutions
of high learning and industry associations including
local autonomous universities each undertook a
digital skill including blockchain, to lead in capability
development.
3.3.2. Blockchain startups
There were 270 FinTech startups, including blockchain
startups in Singapore [60]. Blockchain startups spanned
across industries from the supply chain and logistics,
social networking, FinTech, insurtech, gaming [61].
This year-to-date, Singapore was ranked third at 8.14
per cent relative to the world’s total ICO projects after
the US and UK [62,63]. The conducive regulatory
environment, open and transparent business practices
as well as the availability of skilled workforce,
contributed to making Singapore an appealing hub for
blockchain innovators and startups.
3.3.3. Enterprise- and government-backed projects
There were multiple prototypes, and POCs announced
and implemented by consortia of conglomerates.
In 2016, Bank of America Merrill Lynch, HSBC
and the-then Infocomm Development Authority of
Singapore built a POC to streamline the paper-based
import/export documentation using the Hyperledger
blockchain. PSA International, IBM Singapore and
Pacific International Lines collaborated in August
2017 to develop a trial for blockchain-based supply
chain business network solution. Singapore Airlines
completed its POC in early 2018 for the world’s first
blockchain-based airline loyalty digital wallet that would
allow frequent flyers to instantly convert air miles into
loyalty tokens.
Besides investments by the private sector, “Project
Ubin” by MAS jointly with the network of financial
institutions, was launched to improve transparency
and efficiency of clearing and settlement of payments
and securities with DLT. To-date, “Project Ubin” had
completed software prototypes of three different
models of decentralised inter-bank payment and
settlements.
3.3.4. Research
To stimulate research and development in the use of
technology to improve quality of life and enhance
economic opportunities, the government of Singapore
set aside US$14 million under the Research, Innovation
and Enterprise 2020 plan [64].
In addition, IBM centre for blockchain innovation
(ICBI) that was opened jointly with Singapore’s
Economic Development Board (EDB), worked with
The JBBA | Volume 2 | Issue 2 | October 2019
government agencies, academia and other industry
players to advance Singapore’s contribution to FinTech
innovation and facilitate the adoption of blockchain
technology for finance, trade and commerce as well as
develop the local workforce capabilities [65].
The National University of Singapore established
an academic research laboratory and think tank for
blockchain technology, CRYSTAL (cryptocurrency
strategy, techniques and algorithms) Centre [66]. The
Singapore University of Social Sciences FinTech &
Blockchain Group bridged academia and industry to
build and develop capabilities and skills in FinTech and
blockchain through the twin engines of education and
research that would realise financial integration and
inclusion objectives.
As a blockchain hub, there were open dialogue and
exchanges between regulatory, government and
industry bodies in Singapore. Furthermore, voluntary
and self-regulatory groups like Singapore FinTech
Association, ACCESS (Association of Cryptographic
Enterprises and Startups, Singapore) and BEST
(Blockchain Enterprise and Scalable Technologies)
Association, actively promoted the exchange of
knowledge and best practices to advance the industry.
3.4. South Korea
3.4.1. Regulations and Standards
The government of South Korea supported
the development and application of blockchain
technology and have announced plans to invest over
US$900 million into blockchain initiatives by 2019.
There were six pilot projects in the initiatives, including
livestock history management, personal customs
clearance, simple real estate transactions, online voting,
international electronic document distribution, and
maritime logistics [67]. To accelerate growth through
innovation, the government announced plans to revise
the existing tax regime that would motivate companies
to focus on nascent technology development, like
blockchain [68]. The strategy of the South Korean
government was to construct an “Encrypted Valley”
for the global blockchain industry in Industry 4.0.
The Korean Financial Services Commission (FSC)
confirmed the prohibition of ICOs in January 2019.
When the Korean Financial Investment Association
established Korea's first blockchain alliance at the end
of 2016, South Korean investors participated actively in
cryptocurrency transactions and ICOs until September
2017. Subsequently, the FSC prohibited all ICOs and
enforced their governance given the financial risks of
cryptocurrency investments and transactions [69]. The
Korean FSC started to restructure the regulation on
cryptocurrency trading in 2018 as more cryptocurrency
exchanges opened in the country; only twelve
cryptocurrency exchanges have passed its security
91
jbba
the
checks, while another eleven failed [70, 71]. Following
in August, the Blockchain Law Society issued a clear
mandate to create a proper regulatory framework for
the blockchain and associated cryptocurrencies.
3.4.2. Blockchain startups
South Korean blockchain startups covered a range
of industries, such as FinTech, insurance, social
media, entertainment, real estate. Some of the most
promising blockchain startups in South Korea worked
on blockchain infrastructure & services (e.g. Icon,
Blocko, Deblock), FinTech (e.g. Proof Suite, theLoop),
cryptocurrency exchanges (e.g. Upbit, Korbit, Coinone),
and social media services (e.g. Foresting, Lucidity). The
startup Coinplug, supplied multiple blockchain related
services like digital asset exchange, an identity-based
blockchain platform and online service platform.
Coinplug held the most patents in blockchain in South
Korea and was ranked seventh globally in 2018.
The startups in South Korea sourced funding from the
local and global technology giants. For example, Blocko
that provided a platform for blockchain solutions
had secured US$8.9 million in Series B funding from
Samsung SDS early in 2016. Cultural exports were
integral to South Korean GDP. Muzika, a blockchain
startup, had attracted over ten thousand musicians and
2 million users from 150 countries globally, as well as
crypto and blockchain investment groups [72].
The South Korean internet company, Kakao launched
its blockchain subsidiary, GroundX, in March 2018.
To-date, Ground X had over 50 million monthly
developers to create blockchain services on its global
public blockchain. In May 2018, ICON and LINE cofounded Unchain to build LINE’s blockchain network.
Unchain would develop various DApp services and
expand the blockchain ecosystem.
3.4.4. Research
In December 2016, a group of twenty-one financial
investment companies and five blockchain technology
firms signed a Memorandum of Understanding to
form a distributed ledger solution as a blockchain
consortium. This consortium marked the first attempt
in South Korea where multiple financial firms leverage
blockchain technology for development.
At present, most blockchain developers worked from
universities in South Korea, including Seoul National
University, Korea University, Sogang University, Yonsei
University. These universities launched blockchain
related courses. Decipher, a think-tank in blockchain
research at the Seoul National University made up of
master- and doctoral-level researchers had engaged in
blockchain research for over three years. There had
been various collaborations between university and
industry to nurture skilled blockchain professionals,
such as the collaboration between Korea University
and Huobi.
3.4.3. Enterprise- and government-backed projects
3.5. Comparison across Countries
To develop skilled talent in blockchain, the Minister
of Science and ICT in South Korea announced new
initiatives valued at US$720,000 in addition to the
original S$900 million, to train students, construct
blockchain research centres and foster 10,000
professionals by 2022 [73]. In September 2018, the
government established an open-source blockchain
platform, dubbed Gold Ore. This platform signed an
agreement with multiple international organisations,
such as the Korean Standards Association, Japan
Blockchain Consortium and others, to conduct
blockchain-related training for the industry.
On the enterprise side, Samsung launched its blockchain
platform hosted in the cloud, named “Nexledger”, in
2017. Nexledger applications covered digital identity,
digital payment, digital stamping, supply chain finance,
global warranty and digital provenance [74]. Besides
Samsung, the LG launched its blockchain service
platform in May 2018, named “Monachain”. This
platform offered digital authentication, community
token and supply chain management for the finance,
public, telecommunications and manufacturing
industries [75]. Hyundai Group had made a substantial
investment on the internet of things (IoT) side of
blockchain.
The JBBA | Volume 2 | Issue 2 | October 2019
Table 1 summarises the status of blockchain
development in each of the four countries.
4. Conclusion and Discussion
4.1. Conclusion
From the previous and current state analysis of
infrastructure and programme, Asian countries
like Japan, Singapore, and South Korea stood out
in technological readiness, as well as digital and
regulatory infrastructure. Although China had yet
to make its way into front ranking in global surveys,
the country performed the best in terms of patents
granted normalised by population. Enterprise-backed
blockchain projects contributed to the volume of
patent applications led by Chinese technology firms
such as the BAT, Huawei, Xunlei, and JD.com. Chinese
provincial governments encouraged technology
development using blockchain-dedicated funds.
The focus areas for blockchain-based solutions
differed across countries. Solutions by enterprisebacked projects in Singapore were related to trading
and finance like those on the streamlining of import/
export documentation, supply chain solutions and
inter-bank payment and settlements. In Japan,
92
jbba
the
Table 1. Blockchain development status summary
Four
Enablers
(I)
Regulation
and
standards
(II)
Blockchain
startups
(III)
Enterpriseand
governmentbacked
projects
(IV)
Research
China
Japan
Singapore
South Korea
- ICO & cryptocurrency
trading banned except in
Hong Kong where
cryptocurrencies treated as
securities
- Blockchain advocated by the
government
- Fintech committee by the
central bank (BoC) and
blockchain included
- First batch of blockchain
service providers officially
registered
- Accounting for 28% of new
blockchain startups globally in
2017
- 78% are in 4 major cities
- Areas: technology
applications & enterprise-level
blockchain solutions (e.g.,
Bubi Chain, Juzix & Qulian)
- First country to recognise
bitcoin as a legal payment
option
- Legalised cryptocurrency
exchanges
- Working towards legalising
ICOs
- Devising evaluation
framework for blockchain
projects
- Cryptocurrency regulated if
structured like a security
- Cryptocurrency exchanges
that offered listing and
trading of digital tokens
regulated under the SFA by
the MAS
- Cryptocurrency monitored
for money laundering &
terrorism financing activities
- Fintech regulatory
sandboxes launched
- ICO banned outright
- Cryptocurrency exchanges
legalised
- Blockchain advocated as
existing tax regime being
revised to encourage
blockchain companies)
- 20 out of 167 fintech
startups are blockchainrelated in 2016
- Supported by enterprises &
governments (e.g. NTT
Data, Skuchain, METI)
- Areas: IoT, gaming &
energy industry
- Supported by big
enterprises (e.g. Samsung
sponsored Blocko early)
- Coinplug ranked in 7th in
blockchain patent filed
- Areas: fintech, insurance,
social media, entertainment
& real estate
- Big companies like BAT,
Huawei, Xunlei & JD.com
have started projects related to
the blockchain (e.g. Tencent
created the 1st digital private
bank WeBank & the
TrustSQL blockchain
platform; Alibaba built a
supply chain tracking system
using blockchain together with
PwC)
- Government-backed
blockchain funds launched in
a few cities (e.g. $1.6B
launched in Hangzhou &
$1.4B in Nanjing)
- HKMA & 12 banks released
a blockchain-based trade
finance platform
eTradeConnect
- Surge in the number of new
blockchain research institutes
observed from 2016 to 2018
- More than 90% of research
institutes were established by
corporations and universities
- PBoC launched the Digital
Currency Research Institute
that focuses on the
development & research of
digital currencies
- NEDO on IoT for trade
information sharing
- MIAC on government
tenders & e-government
services
- Deloitte Japan, Mizuho,
SMFG & MUFG on
interbank payment and KYC
platform
- JPX & 6 financial
institutions on streamlining
financial services
- NTT Data on supply chain
efficiency
- 270 FinTech start-ups
including blockchain startups
- Areas: applications in
supply chain & logistics,
social networking, fintech,
insurtech, gaming, financial
exchanges, cloud
infrastructure, payment &
remittances (e.g., Qtum,
NEO & VeChain)
- MAS Project Ubin for
settlement of payments &
securities between financial
institutions and the central
bank led the project
- A project between Bank of
America Merrill Lynch,
HSBC & IMDA on trade
documentation using
blockchain technology
- Singapore Airline’s
blockchain-based airline
loyalty digital wallet
- Singapore Smart Nation
Initiative to improve living
with new and emerging
technology
- Supporting research &
development in the national
RIE 2020 plan
- University, government &
industry collaboration (e.g.
the IBM centre for
blockchain innovation, the
Cryptocurrency Strategy,
Techniques & Algorithm
Centre at NUS, FinTech &
Blockchain Group at SUSS)
- Volunteer groups of selfregulatory organisations (e.g.
SFA, ACCESS & BEST)
- Financial and tech firms
assigned MOU for
blockchain consortium
development.
- Blockchain training
courses launched in many
universities
- Universities & industry
collaborating on blockchain
research & application (e.g.
Korea University
collaborating with Huobi
and KEB)
- Led by major financial
institutions & universities
- Bank of Japan’s FinTech
Centre for payment &
settlement
- BSafe.network led by the
University of Tokyo forming
a blockchain research
network for with over 30
member universities
worldwide
- University & industry
collaborations for applied
research on smart currency
& blockchain
enterprise-backed projects were related to information
sharing by financial institutions and government
agencies. Like China, technology firms in South Korea,
such as Samsung, LG, and Hyundai, initiated various
enterprise-backed projects on blockchain technology.
These developmental activities created a value chain of
activities and opportunities.
Blockchain startups in China and South Korea built
applications across a wide spectrum from FinTech,
insurance, social media, to real estate, and more. On
the other hand, the startups in Japan and Singapore
tended towards FinTech applications. English is the
The JBBA | Volume 2 | Issue 2 | October 2019
- Government-backed
blockchain talent project
(e.g. MOS & ICT planning
to invest US$720k to
construct blockchain
research centres and foster
10,000 professionals by 2022
- The government
established an open-source
blockchain platform named
Gold Ore
- Tech giants are investing in
blockchain development
(e.g. Samsung, LG & Kakao
launched blockchain
platform Nexledger; LG
launched blockchain service
platforms)
official business language in Singapore; this appealed
to investors and blockchain entrepreneurs globally.
Perceived as a gateway in the east to countries in the
west, startups from China, such as Qtum, NEO, and
VeChain, had registered their firms in Singapore. The 11
source of funding and capital for startups at their early
stage in China, Japan, and South Korea, were mostly
domestic.
The existing regulations in China and South Korea
prohibited the exchange and trading of cryptocurrencies
and ICO. Meanwhile, Japan and Singapore adopted a
more nuanced stance and issued clear policy statements.
93
jbba
the
The former legislation protected investors’ interest
and the latter statutory and regulatory approach could
motivate blockchain startups in fulfilling their project
objectives. All four countries nurtured investments and
developments in blockchain technology. For example,
blockchain was included in official documents released
by the Chinese national government and the first batch
of blockchain projects were endorsed with service
provider licenses. Likewise, the Japanese government
published an evaluation framework for blockchainbased projects. South Korea revised its tax regime to
encourage blockchain companies. The central bank of
Singapore launched FinTech regulatory sandboxes in
2016 to promote and nurture technology innovation.
Although there were many blockchain startups or
projects backed by large enterprises and government,
to-date no blockchain-giants had emerged. We
anticipated Japan to lead on the regulatory infrastructure
front being one of the first to accept cryptocurrencies
by legalising cryptocurrency exchanges, and the
publication of a government-led evaluation framework
for blockchain projects. The abundant technical talent
pool in China among the innovator group might
accelerate the growth of the hub. In Singapore, the
favourable environment for ICO financing could
support the funding requirements of blockchain
startups with strong offerings. Although a small citystate relative to China, Japan and South Korea, the
domestic talent gap, particularly in technical know-how
could be mediated by Singapore’s language capability
and proximity to countries in Southeast Asia.
4.2. Discussion
This essay contributed to the growing body of
literature on blockchain and informed the state of
blockchain development in Asia. We reviewed the
stage of development in four different countries in
Asia and found these countries to have possessed
similar characteristics in their blockchain ecosystem:
innovators and developers supported by regulatory
and digital infrastructure, funding and capital, as well
as programmes for workforce capability development
against ready demand for distributed ledger or
blockchain technology.
The performance of these four enablers would impact
the speed of development of each hub given the
nascent state of blockchain development.
On the regulation front of the infrastructure and
programme enabler, we projected the pace of change
to differ by countries.
The regulators of these four countries we have reviewed
shared similar approach towards the blockchain
technology – a deliberate and agile strategy to protect
the public’s interest while advocating technology
The JBBA | Volume 2 | Issue 2 | October 2019
innovation. Industry bodies could be the catalyst
to initiate self-regulating organisations to network,
exchange knowledge and best practices, promote
standards, and engage the startups and regulators
constructively.
As each of the four countries competes to attract and
develop technical expertise in blockchain technology,
they would have to harness their unique value
propositions to develop capabilities and sustain the
hub of activities.
4.3. Future Trends
Going forward, two factors shape the developments of
these blockchain hubs – one factor at the network level
facilitated by one or more catalyst firms and another
within the network.
In the network paradigm of a hub, Dhanaraj and
Parkhe identified the role of a catalyst firm in a hub
that comprised of diverse stakeholder groups [2]. We
drew parallel in forecasting the future developments
of blockchain hubs. The presence of a catalyst firm
in a blockchain hub would accelerate development to
realise both economic and social impact of the hub.
Such a catalyst firm could be the coordinator between
regulators and startups to facilitate communication
and knowledge sharing. Any organisation could step
up to be the catalyst, such as an industry association,
a research institute, a government agency or even a
technology corporation.
The second influencing factor for the future of these
blockchain hubs would be endogenous in the network.
By this, we refer to the capabilities of the workforce in a
blockchain hub. These capabilities of a blockchain hub
shape the speed of its future development. Capabilities
include technical skills and capabilities, as well as the
language communication skills of the workforce to
transcend cultural differences and collaborate with
global teams.
4.4. Limitations
4.4.1. Biases from language
The data collected for analyses of this paper were
predominantly in English. For example, a patent
application might be filed in vernacular languages
instead of English in countries like China, Japan, and
South Korea. Google Trends was the primary source
to gather search trends for ICOs. This approach
introduced biases in our study of China, Japan, and
South Korea, where official languages were nonEnglish. We attempted to minimise such biases by
collecting data from multiple sources. Future research
could introduce expert opinion surveys in respective
local languages for comparative analyses.
94
jbba
the
4.4.2. Temporal analysis and quantitative analysis
Although we had systematically investigated each of
the four hubs separately, focusing on its regulation,
standards and research development over time, this
paper had not addressed agglomeration effects within
the country. First-tier cities in China such as Beijing,
Guangzhou, and Shanghai were key contributors to
patents filed and granted. Subsequent research could
extend beyond country-level analyses to study the
agglomeration effects within the country.
This paper served as a qualitative analysis across
blockchain hubs in Asia. Subsequent research using
quantitive analysis could consider quantifying each of
the enablers as inputs into an index to monitor and
track the development of blockchain hubs through
inter-temporal analysis, regionally and globally.
References:
[1] D. K. C. Lee, and L. Low, “Inclusive Fintech: Blockchain,
Cryptocurrency And ICO”, World Scientific, Singapore, 2018.
[2] C. Dhanaraj, and A. Parkhe, “Orchestrating Innovation
Networks”, The Academy of Management Review, vol. 31, no.
3, pp. 659–669, Jul. 2006.
[3] S. Chishti, and J. Barberis, The FinTech Book. Chichester
: John Wiley & Sons, 2016.
[4] D. Andrews, G. Nicoletti and C. Timiliotis, "Digital
technology diffusion: A matter of capabilities, incentives or
both?", OECD Economics Department Working Papers,
No. 1476, OECD Publishing, Paris, Jul. 2018. [Online].
Available: doi.org/10.1787/7c542c16-en.
[5] Asian Development Bank, Key Indicators for Asia and
the Pacific 2018 (49th Edition). Manila: Asian Development
Bank, 2018.
[6] D. K. C. Lee, M. Chwierut, W. Anderson, B. Lio, and
B. Downes, “The Characteristics of Token Investors”, in book
“Inclusive Fintech: Blockchain, Cryptocurrency And ICO”,
World Scientific, Singapore, pp. 125-172, 2018.
[7] A. Kerya, “ICO Statistics: Countries, Traffic, and
Investors”, Medium.com, Jun. 2018. Accessed on: Oct. 6,
2018. [Online]. Available at URL: https://medium.com/@
incryptico.com/ico-statistics-countries-traffic-and-investors4e830e0438b0.
[8] T. Alford, “Bitcoin Adoption: Trading Volume by
Country”, TotalCrypto.io, Aug. 2018. Accessed on: Oct. 6,
2018. [Online]. Available at URL: https://totalcrypto.io/
bitcoin-adoption-trading-volume-country/.
[9] A. Wright, and P. De Filippi, “Decentralized Blockchain
Technology and the Rise of Lex Cryptographia”, 2015.
Available at SSRN 2580664.
[10] A. Demirgüç -Kunt., L. Klapper, D. Singer, S. Ansar,
and J. Hess, The Global Findex Database 2017: Measuring
financial inclusion and the fintech revolution. The World Bank,
2017.
[11] The World Bank. “Record High Remittances Sent Globally
in 2018”. Retrieved from World Bank Group - International
Development, Poverty & Sustainability. Apr. 2018. [Online].
Available at URL: https://www.worldbank.org/en/news/
press-release/2019/04/08/record-high-remittances-sentThe JBBA | Volume 2 | Issue 2 | October 2019
globally-in-2018.
[12] Startup Genome, "Global Startup Ecosystem Report
2019", San Francisco: Startup Genome LLC, 2019.
[13] Aliresearch and KPMG, “Welcoming a New Wave of
Global Economy: Digital Economy Development Global Index
2018”, Sep. 2018. Accessed on: Sep. 28, 2018. [Online].
Available at URL: http://www.199it.com/archives/774852.
html.
[14] China Academy of Information and Communications
Technology (CAICT), “White Paper on the Development of the
Digital Economy in China”, CAICT, Jul. 2017. Accessed on:
Aug. 25, 2018. [Online]. Available: http://www.cac.gov.cn/
files/pdf/baipishu/shuzijingjifazhan.pdf.
[15] L. Liu, and Q. Wang, “Analysis of Bitcoin Herd
Behavior Based on Imitation and Infection Model”, Journal
of Beijing University of Posts and Telecommunications (Social
Sciences Edition), Vol.17, No.2, pp.27-33, 2015.
[16] Ministry of Industry and Information Technology (MIIT),
“Announcement on preventing financial risks of ICOs”, 2017.
Accessed on: Sep. 26, 2018. [Online]. Available at URL:
http://www.miit.gov.cn/n1146290/n4388791/c5781140/
content.html.
[17] National Development and Reform Commission
(NDRC), “The 13th five-year plan for economic and social
development of the People’s Republic of China (2016–2020)”,
Central Compilation & Translation Press, 2017. Accessed on:
Sep. 20, 2018. [Online]. Available: http://en.ndrc.gov.cn/
newsrelease/201612/P020161207645765233498.pdf.
[18] M. Amsili, “Blockchain in China: Local is Everything”,
supchina, Aug. 2018. Accessed on: Sep. 20, 2018. [Online].
Available at URL: https://supchina.com/2018/08/28/
blockchain-in-china-local-is-everything/.
[19] G. Sun, “Expectations on the Development of Blockchain
in China”, Caixin, Jun. 2017. Accessed on: Sep. 27,
2018. [Online]. Available at URL: http://opinion.caixin.
com/2017-06-17/101102805.html.
[20] Wuzhen Think Tank, “The white paper on the development
of China’s blockchain industry”, 2017. Accessed on: Sep. 23,
2018. [Online]. Available: http://sike.news.cn/hot/pdf/12.
pdf.
[21] C. Liu, “Blockchain - A Guide for Officials,” People’s
Daily Press, pp.132-134, 2018.
[22] X. Fei, and S. W. Li, “Blockchain Courses Among Global
Universities and Blockchain Research Institutes in China”,
BlockData, 2018. Accessed on: Sep. 29, 2018. [Online].
Available at URL: https://www.8btc.com/article/195209.
[23] M. Xing, “63 patent applications filed by the PBoC digital
currency research institute”, Jun. 2018. Accessed on: Oct. 2,
2018. [Online]. Available at URL: http://finance.ce.cn/
rolling/201806/29/t20180629_29567884.shtml.
[24] W. Zhao, “South Korea Plans Tax Perks for Blockchain
Startups”, coindesk, Jul. 2018. Accessed on: Sep. 28, 2018.
[Online]. Available at URL: https://www.coindesk.com/
south-korea-plans-tax-perks-for-blockchain-startups/.
[25] Securities and Futures Commission (SFC), “SFC warns
of cryptocurrency risks”, Feb. 2018. Accessed on: Sep. 1,
2018. [Online]. Available at URL: https://www.sfc.hk/
edistributionWeb/gateway/EN/news-and-announcements/
news/doc?refNo=18PR13.
95
jbba
the
[26] Applied Science and Technology Research Institute
(ASTRI), “HKMA-ASTRI Fintech Innovation Hub”,
Nov. 2016. Accessed on: Sep. 12, 2018. [Online]. Available
at URL: https://www.astri.org/technologies/joint-researchlaboratories/rd-centres/hkma-astri-fintech-innovation-hub/.
[27] HKMA, “Fintech Co-operation between the Hong Kong
Monetary Authority and the Financial Services Regulatory
Authority of Abu Dhabi Global Market”, Jun. 2018.
Accessed on: Sep. 10, 2018. [Online]. Available at URL:
https://www.hkma.gov.hk/eng/key-information/pressreleases/2018/20180626-4.shtml.
[28] HKMA, “Fintech Collaboration between the Hong
Kong Monetary Authority and the Monetary Authority of
Singapore”, Oct. 2017. Accessed on: Sep. 10, 2018. [Online].
Available at URL: https://www.hkma.gov.hk/eng/keyinformation/press-releases/2017/20171025-4.shtml.
[29] HKMA, “The launch of eTradeConnect and the
Collaboration with we.trade”, Oct. 2018. Accessed on: Apr. 20,
2019. [Online]. Available at URL: https://www.hkma.gov.
hk/eng/key-information/press-releases/2018/20181031-4.
shtml.
[30] Business Wire, “AlipayHK and GCash Launch CrossBorder Remittance Service Powered by Alipay’s Blockchain
Technology”, Jun. 2018. Accessed on: Aug. 8, 2018. [Online].
Available at URL: https://www.businesswire.com/news/
home/20180625005561/en/.
[31] A. Antonovici, “Bank of China HK Uses Blockchain
for 85% of Real Estate Valuations”, Cryptovest, Apr.
2018. Accessed on: Sep. 10, 2018. [Online]. Available at
URL: https://cryptovest.com/news/bank-of-china-hk-usesblockchain-for-85-of-real-estate-valuations/.
[32] W. Zhao, “Hong Kong to Expedite Immigration for
Blockchain Job Seekers”, coindesk, Aug. 2018. Accessed
on: Sep. 12, 2018. [Online]. Available at URL: https://
www.coindesk.com/hong-kong-to-expedite-immigration-forblockchain-job-seekers.
[33] Y. Hagiwara, and Y. Nakamura, “Japan Unveils
Guidelines for Allowing Initial Coin Offerings”, Bloomberg,
Apr. 2018. Accessed on: Sep. 27, 2018. [Online].
Available at URL: https://www.bloomberg.com/news/
articles/2018-04-05/japan-plans-first-step-toward-legalizinginitial-coin-offerings.
[34] Tama University, (2018), “Call for Rule-Making on
ICO”, Apr. 2018. Accessed on: Sep. 7, 2018. [Online].
Available: https://www.tama.ac.jp/crs/2018_ico_en.pdf.
[35] Nomura Research Institute, “Survey on Blockchain
Technologies and Related Services FY2015 Report”, Mar.
2016. Accessed on: Aug. 28, 2018. [Online]. Available:
http://www.meti.go.jp/english/press/2016/pdf/0531_01f.
pdf.
[36] Information Economy Division, Commerce and Information
Policy Bureau, “Evaluation Forms for Blockchain-based System
ver. 1.0”, Ministry of Economy, Trade and Industry, 2017.
[37] Information Economy Division, Commerce and Information
Policy Bureau, “Survey on Technology and Institution related
to Distributed System”, Ministry of Economy, Trade and
Industry, 2018.
[38] T. Wilson, “Expert Shortage Hampers Japanese
The JBBA | Volume 2 | Issue 2 | October 2019
Financials in Blockchain Race”, Aug. 2016. Accessed on: Oct.
2, 2018. [Online]. Available at URL: https://www.reuters.
com/article/us-japan-fintech-blockchain-idUSKCN10S2GN.
[39] Ministry of Economy, Trade and Industry, “Japan
Startup Selection" - The Participants of 2016 "HIYAKU
Next Enterprise" Program”, 2017. Accessed on: Sep. 30,
2018. [Online]. Available: http://www.meti.go.jp/english/
press/2017/pdf/0105_001a.pdf.
[40] D. Cullinan, “World’s First Bank-Backed Crypto
Exchange Opens For Trading”, BitcoinNews.com, Jul. 2018.
Accessed on: Sep. 27, 2018. [Online]. Available at URL:
https://bitcoinnews.com/worlds-first-bank-backed-cryptoexchange-opens-for-trading/.
[41] Mizuho Financial Group, Inc., “Collaborating to Create
Japan's First Blockchain Coworking Space”, May 2018.
Accessed on: Sep. 30, 2018. [Online]. Available: https://
www.mizuho-fg.com/release/pdf/20180518release_eng.pdf.
[42] Finolab, “About us - FINOLAB”, 2018. Available at
URL: https://finolab.tokyo/#aboutus.
[43] A. Santo, I. Minowa, G. Hosaka, S. Hayakawa, M.
Kondo, S. Ichiki, and Y. Kaneko, “Applicability of Distributed
Ledger Technology to Capital Market Infrastructure (Vol.
15)”, Japan Exchange Group, Aug. 2016.
[44] NTT DATA Corporation, “MUFG and NTT
DATA Lay Foundation for Digital Trade Between Singapore
and Japan Using Blockchain”, Dec. 2017. Accessed on: Sep.
28, 2018. [Online]. Available at URL: https://www.nttdata.
com/global/en/media/press-release/2017/december/mufgand-ntt-data-lay-foundation-for-digital-trade-between-singaporeand-japan-using-blockchain.
[45] NTT DATA Corporation and Skuchain, Inc., “NTT
DATA will Develop a Business Collaboration Platform for
Japanese Manufacturers in Cooperation with Skuchain in
the U.S”, Jan. 2018. Accessed on: Sep. 28, 2018. [Online].
Available at URL: http://www.skuchain.com/ntt-datawill-develop-a-business-collaboration-platform-for-japanesemanufacturers-in-cooperation-with-skuchain-in-the-u-s/.
[46] Sumitomo Mitsui Financial Group, “Demonstration Test
of Blockchain Technology in Cross-Border Trade Operations”,
Dec. 2017. Accessed on: Aug. 28, 2018. [Online]. Available
at URL: http://www.smfg.co.jp/news_e/e110076_01.html.
[47] I. Ueno, “Japanese Financial Institution and FinTech”,
Mizuho Financial Group, 2017.
[48] Mitsubishi UFJ Financial Group, “Fintech, Blockchain
and Digital Currencies”, Mar. 2016. Accessed on: Sep. 30,
2018. [Online]. Available at URL: https://www.mufg.jp/
english/ourbrand/featuredarticle/2016_03.html.
[49] Nikkei, “Japan Looks to Blockchains for More Secure
E-Government Systems”, Asian Review, Jun. 2017. Accessed
on: Sep. 27, 2018. [Online]. Available at URL: https://asia.
nikkei.com/Politics-Economy/Policy-Politics/Japan-looks-toblockchains-for-more-secure-e-government-system.
[50] Japanese Bankers Association, “Report of the Review
Committee for the Possibility and the Challenges of Utilizing
Blockchain Technology”, Japanese Bankers Association, 2017.
[51] Bank of Japan, “Message from Governor Kuroda on the
Occasion of the Establishment of the FinTech Center”, Apr.
2016. Accessed on: Sep. 29, 2018. [Online]. Available at
URL: https://www.boj.or.jp/en/paym/fintech/message.htm/.
96
jbba
the
[52] European Central Bank and Bank of Japan, “Payment
Systems: Liquidity Saving Mechanisms in a Distributed
Ledger Environment”, Sep. 2017. Accessed on: Sep. 29,
2018. [Online]. Available: https://www.boj.or.jp/en/
announcements/release_2017/data/rel170906a1.pdf.
[53] European Central Bank and Bank of Japan, “Securities
Settlement Systems: Delivery-versus-Payment in a Distributed
Ledger Environment”, Mar. 2018. Accessed on: Sep. 28,
2018. [Online]. Available: https://www.boj.or.jp/en/
announcements/release_2018/data/rel180327a1.pdf.
[54] S. Matsuo, “BSafe.network: Current Member
Universities”, Bsafe.network, May 2019. Accessed on: May.
10, 2019. [Online]. Available at URL: http://bsafe.
network/member-university/.
[55] S. Takagi, H. Tanaka, M. Takemiya, and Y. Fujii,
“Blockchain-Based Digital Currencies for Community
Building”, GLOCOM, 2017.
[56] Keio Research Institute at SFC, “Announcement of
BASE Alliance Establishment”, Jul. 2017. Accessed on: Sep.
18, 2018. [Online]. Available: https://www.kri.sfc.keio.ac.jp/
ja/press_file/20170724_base_en.pdf.
[57] HAW International Inc., “Blockchain Technology
Research Lab Established Within the Institute of Information
Engineering Research at the Kyushu Institute of Technology”,
Mar. 2018. Accessed on: Oct. 1, 2018. [Online]. Available:
htt p://www.chaintope.com/en/wp-content/uploads/
sites/2/2018/03/0306_PRESS_BC_KIT_EN_2.pdf.
[58] Monetary Authority of Singapore, “Fintech Regulatory
Sandbox Guidelines”, 2016. Accessed on: Sep. 30, 2018.
[Online]. Available: https://www.mas.gov.sg/-/media/
MAS/Smart-Financial-Centr e/Sandbox/FinTechRegulatory-Sandbox-Guidelines-19Feb2018.pdf ?la=en&ha
sh=1F4AA49087F9689249FB8816A11AEAA6CB3
DE833.
[59] J. Lee, “MAS Slaps Warnings on 8 Cryptocurrency
Exchanges; Bars ICO Issuer”, Business Times, May 2018.
Accessed on: Sep. 8, 2018. [Online]. Available at URL:
https://www.businesstimes.com.sg/banking-finance/mas-slapswarnings-on-8-cryptocurrency-exchanges-bars-ico-issuer.
[60] Startup Genome, “Global Startup Ecosystem Report
2018”, San Francisco: Startup Genome LLC, 2018.
[61] Techinasia.com, “Tech in Asia - Meet the 15 top-funded
blockchain companies in Singapore”, May 2018. Accessed on:
Oct. 6, 2018. [Online]. Available at URL: https://www.
techinasia.com/.
[62] ICO Watchlist, (2018), “ICO Statistics - By Country”,
2018. Accessed on: Sep. 12, 2018. [Online]. Available at
URL: https://icowatchlist.com/statistics/geo.
[63] W. A. Kaal, “Initial Coin Offerings: the Top 25
Jurisdictions and Their Comparative Regulatory Responses”,
CodeX Stanford Journal of Blockchain Law & Policy, U of
St. Thomas (Minnesota) Legal Studies Research Paper No. 1807, 2018. Available: dx.doi.org/10.2139/ssrn.3117224.
[64] U-W. Lee, “Record S$19b Set Aside for R&D Until
2020”, Business Times, Jan. 2016. Accessed on: Sep. 1, 2018.
[Online]. Available at URL: https://www.businesstimes.com.
sg/government-economy/singapores-future-economy/records19b-set-aside-for-rd-until-2020.
[65] F. Ungku, “IBM to open first blockchain innovation center
The JBBA | Volume 2 | Issue 2 | October 2019
in Singapore”, Reuters, Apr. 2016. Accessed on: Sep. 28,
2018. [Online]. Available at URL: https://www.reuters.com/
article/us-ibm-fintech-singapore/ibm-to-open-first-blockchaininnovation-center-in-singapore-idUSKCN0ZS03Y.
[66] National University of Singapore, “NUS Computing
forms academic blockchain think tank”, Sep. 2018. Accessed
on: Sep. 25, 2018. [Online]. Available at URL: https://news.
nus.edu.sg/press-releases/CRYSTAL-centre.
[67] Y. Yoon, “Korean Gov't Unveils Blockchain Technology
Development Strategy”, Business Korea, Jun. 2018. Accessed
on: Sep 8, 2018. [Online]. Available at URL: http://www.
businesskorea.co.kr/news/articleView.html?idxno=23184.
[68] W. Zhao, “PBoC filings reveal big picture for planned
digital currency”, coindesk, Jul. 2018. Accessed on: Sep.
29, 2018. [Online]. Available at URL: https://www.
coindesk.com/pboc-filings-reveal-big-picture-for-planned-digitalcurrency/.
[69] R. R. O’Leary, “South Korean Regulator Issues ICO
Ban”, coindesk, Sep. 2017. Accessed on: Sep. 1, 2018.
[Online]. Available at URL: https://www.coindesk.com/
south-korean-regulator-issues-ico-ban/.
[70] FSC, “FSC Reshuffles Organizational Structure”,
Jul. 2018. Accessed on: Sep. 7, 2018. [Online]. Available
at URL: https://www.fsc.go.kr/eng/new_press/releases.
jsp?menu=01&bbsid=BBS0048.
[71] K. Helms, “Only 12 out of 23 Korean Crypto Exchanges
Pass Probe - Inspector Under Fire”, Bitcoin.com, Jul. 2018.
Accessed on: Sep. 12, 2018. [Online]. Available at URL:
https://news.bitcoin.com/only-12-out-of-23-korean-cryptoexchanges-pass-probe-inspector-under-fire/.
[72] Muzika, “Muzika Project Teaser”, 2018. Accessed on:
Sep. 20, 2018. [Online]. Available: https://www.muzika.
network/assets/mzk-teaser-en.pdf.
[73] J. Kim, “South Korean Gov’t to Invest $200 Mln in
Blockchain Initiatives”, Cryptoslate, Jun. 2018. Accessed
on: Sep. 6, 2018. [Online]. Available at URL: https://
cryptoslate.com/south-korean-govt-to-invest-200-mln-inblockchain-initiatives.
[74] Samsung SDS, Nexledger™ A Blockchain Platform and
Solution, White paper, 2017.
[75] M. H. Cho, “LG CNS Launches Monachain Blockchain
Platform”, ZDNet, May 2018. Accessed on: Aug. 20, 2018.
[Online]. Available at URL: https://www.zdnet.com/article/
lg-cns-launches-monachain-blockchain-platform/.
Consolidated information from World Intellectual
Property Organisation and IPR Daily statistics on
blockchain patents between 2008 to 2018.
ii
Source: CBInsights (https://www.cbinsights.com/
research/unicorn-startup-market-map/)
iii
Special Administrative Region (SAR)
iv
R&D spending as per cent of GDP of APAC
countries, World Development Indicators by the World
Bank.
v
http://www3.weforum.org/docs/GITR2016/WEF_
GITR_Full_Report.pdf
vi
http://www.eiu.com/Handlers/WhitepaperHandler.
ashx?fi=Technological_readiness_report.pdf&mode=
wp&campaignid=TechReadiness
i
97
jbba
the
Data obtained from ICOWatchList.com. https://
icowatchlist.com/statistics/geo
viii
Search volume of a 1-year period from 7 Oct, 2017
to 7 Oct, 2018. https://trends.google.com/trends/
explore?q=%2Fm%2F0138n0j1
ix
Results are based on the Venture Capital & Private
Equity Country Attractiveness Index by IESE Business
School, University of Navarra. https://blog.iese.edu/
vcpeindex/ranking/
x
A full list of national currencies exchanged for the
24 hours of total bitcoin volume can be found at the
Coinhills website: https://www.coinhills.com/market/
currency/
xi
https://news.8btc.com/1-6-billion-governmentbacked-blockchain-fund-launched-in-hangzhou
xii
https://www.coindesk.com/another-1-billionblockchain-fund-to-launch-with-government-backing
vii
Competing Interests:
None declared.
Ethical approval:
Not applicable.
Author’s contribution:
YW, JR, CL and SWL designed and coordinated this research and
prepared the manuscript in entirety.
Funding:
None declared.
Acknowledgements:
YW, JR, CL and SWL deeply thank Professor David Lee for his
guidance and Ms Sherry Li for her support.
The JBBA | Volume 2 | Issue 2 | October 2019
98
jbba
the
The JBBA | Volume 2 | Issue 2 | October 2019
99
jbba
the
commentary
Decentralisation is Coming:
The Future of Blockchain
OPEN ACCESS
ISSN Print: 2516-3949
https://doi.org/10.31585/jbba-2-2-(5)2019
Mark Fenwick1, Erik P.M. Vermeulen2
1
Kyushu University, Japan
2
Tilburg University, The Netherlands
Correspondence: E.P.M.Vermeulen@uvt.nl
Received: 31 July 2019 Accepted: 12 August 2019 Published: 16 August 2019
Abstract
Advocates of blockchain believe that distributed ledger technologies can provide us with a technological
infrastructure to challenge the concentrated power of tech giants such as Amazon, Facebook and Google,
and create a more equitable, sustainable and decentralized world. This paper considers these claims and
concludes that they are preferable to defending the status quo or arguing that a solution might be found in
more and better regulations. Nevertheless, the future remains highly uncertain and we are currently living
in a rapidly evolving “space” between two competing realities: a centralized old-world reality and a fastemerging, but, as yet, incomplete, decentralized reality. We remain optimistic that decentralization is coming
but identify powerful competing forces seeking to preserve the status quo. As such, we must encourage more
organizations – business, government, investors, charities – to experiment with distributed ledger technologies
and to participate actively in the digital transformation. We need more experimentation to address the current
shortcomings of decentralisation and to ensure the early arrival of mainstream applications of a technology
that has the potential to solve some of the most pressing global challenges of a digital age.
Keywords: Bitcoin, Blockchain, Crypto-Economy, Decentralization, Digital Transformation, Distributed Ledgers,
Disintermediation, Ethereum, Satoshi Nakamoto, Smart Contracts, Technology
JEL Classifications: K20, K22, K24, L50, M21, O30, O31, O33, O35, Q55
1. Introduction
Advocates of blockchain – let’s call them the
“Evangelists” – believe that decentralised ledger
technologies have the potential to address many
of the most pressing problems of the digital age.
We are all familiar with the problems. The massive
concentration of economic power in companies such
as Amazon, Apple, Facebook, Google, etc. The largescale abuse of privacy via the hoarding and selling of
personal information online. The systematic (and statesponsored) political misinformation operations and
the calculated spreading of so-called “fake news.”
The Evangelists believe that these and other problems
can only be solved with more technology, rather
than through more rules and regulations. And, in
the strongest version of this story, Evangelists claim
that blockchain technologies have the potential to
transform capitalism and herald in a more sustainable,
egalitarian, and decentralised world. In this piece, we
would like to offer a defense of this Evangelist view.
Not least because it offers a more compelling vision
The JBBA | Volume 2 | Issue 2 | October 2019
of the future than those in denial about the scale of
the challenges created by the digital revolution or those
arguing that more and better rules and regulations are
the answer.
Nevertheless, it is easy to be skeptical or cynical in
the face of such idealism. After all, the Evangelist
narrative cuts against previous experience of disruptive
technologies.i Historians have often noted that new
technologies start in the hands of nerds and dreamers
motivated by the desire to make the world a better place
(the Apple of Steve Wozniak). But this rarely lasts,
and successful technologies ultimately end up in the
hands of powerful corporations driven the desire to
maximize profits and shareholder value (the Apple of
Steve Jobs). According to this view, the Internet story is
just the latest chapter in a sorry tale of a human failure
to ensure that technology works for the benefit of all.
After all, the corporate giants of today are amongst
the biggest companies that have ever existed and there
is ever-increasing inequality in wealth distribution.ii
Everything we know about the history of technology
and capitalism should make us treat the Evangelist
101
jbba
the
position with caution.
Moreover, the transformative potential of distributed
ledgers can sometimes be difficult to see through
all the noise and hyperbole that surrounds the
“Blockchain Revolution.”iii It is unfortunate, for
instance, that blockchain technologies have attracted
greedy opportunists and fraudsters keen to make a
quick profit. The result? A series of ICO scams and
other scandals that discredited the technology in many
people’s eyes before it had any real-world impact on
our everyday lives. But, once the blockchain hype fades,
and the opportunists have moved on to the next “big
thing,” will these technologies be able to deliver on
their potential and promise? Or, are the skeptics and
nay-sayers right when they suggest that this is just hype
“all the way down?”
The paper has three parts. In the next section (‘The
Rise of Centralized Platforms’), we describe the
emergence of the new tech giants that leveraged the
new possibilities of the Internet to develop a platform
business model. Furthermore, we identify various
pressures that create ever-more centralization and
concentrations of economic power in the platform
economy. The next section (‘The Decentralized
Alternative of Blockchain Evangelists’) identifies the
Evangelical alternative; a radically different account
of the future that seeks to utilize distributed ledger
technologies to realize the idealistic vision of the
original architects of the Internet as a decentralized
global communications network. In doing so, a genuine
alternative to the current tech giants can be conceived.
We conclude (‘Experiments in Decentralization’)
with some brief reflections on the need for more
participation in the development of blockchain
technology, smart contracts, and cryptocurrencies to
address the current shortcomings of decentralization
and to ensure that we will soon see mainstream
applications of the technology.
The takeaway? We are currently living in a fastdeveloping “space” between two competing realities:
a centralized “old world” reality and a fast-emerging,
but, as yet, incomplete, “decentralized reality.” We are
cautiously optimistic that decentralization is coming,
but acutely aware of the competing forces that seek to
preserve the status quo.
2. The Rise of Centralised Platforms
The Internet today comprises two connected, but
distinct, layers. Firstly, there are a series of open source
protocols, such as HTTP, GPS, IMAP, POP, SMPT,
etc., that first allowed computers to communicate with
one another across global networks and which still
provide the basic infrastructure of the system. The
key characteristic of such protocols is that no one
owns them, and anyone can use them free of charge.
The JBBA | Volume 2 | Issue 2 | October 2019
There is no license fee involved in using HTTP to set
up a web page, or in using SMTP to send an email,
or GPS to identify location. Secondly, there is the
web-based layer, which emerged later, and which sits
on top of the protocols providing various services.
Think Amazon, Facebook, Google, Twitter: this layer
is operated by profit-seeking corporations that – in
contrast to the authors of the protocols – have always
sought to maintain tight control over their services and
operations. The history of the Internet can be told
as a story of a shift in power from the open protocol
idealism of the early years to the closed, centralized
and controversial capitalism that dominates today.iv
Many of the companies operating on this second layer
provide what me might call a coordination function
between two or more groups of users, and this business
model is usually described as a “platform.”v Some
platforms facilitate connections between the buyer and
seller of goods (eBay, Amazon, Alibaba); some facilitate
connections between those wanting a service and those
willing to provide it (Uber, Airbnb); and others simply
facilitate connections (information exchange) between
friends (Facebook), content creators and consumers
(You Tube, Medium, Netflix) or app developers and
users (Google, Apple). However, what is common to all
platforms is that they coordinate connections between
“creators” and “extractors” of value and the platform
generates a profit from making these connections,
either by taking a commission or advertising.
The emergence and growth of platforms is a significant
economic and cultural event, not least because they
have become a routinized feature of everyday life
within a short period. To illustrate this rise, consider
that it took the radio 38 years to reach 50 million users.
It took television 13 years to achieve the same degree
of market penetration. But Facebook “only” needed
two years to gain the same number of users. Now, it
has an active user base of over 2 billion.
Moreover, the global proliferation of digital
technologies and communication networks means that
platforms can be established anywhere. The emergence
of hugely successful platforms in China (Alibaba) or
Indonesia (Go-Jek) illustrate the universal appeal and
adaptability of this business model. It also shows how
less developed economies might employ platforms as
a means of “leapfrogging” an earlier (industrial) phase
of economic development and “jump” directly into the
digital age.vi Go-Jek “only” needed three years to go
from 100,000 orders a day (in 2015) to 100+ million
orders across 18+ services in 2018.vii
What is clear, however, is that as platforms have scaled,
they have struggled to maintain their initial promise
and platforms that were once disruptive have lost
much of their initial appeal. And it is hard to ignore the
problems experienced and created by platforms. There
102
jbba
the
are too many recent examples of well-known platforms
“forgetting” the importance of improving people’s
lives. Although there are a number of reasons why
platforms have tended to become more centralized and
more “corporate,” and have experienced these kinds of
difficulties, two factors are worth emphasizing:
Firstly, markets tend to prefer a single service provider.
Take Airbnb as an example. When a new platform
service like Airbnb starts to take off, there’s a strong
incentive for the market to consolidate around that
single provider. The fact that more customers start to
use the Airbnb app means that more room-providers
are attracted to join the platform, which in turn attracts
more people looking for a room, as there are now
more choices of rooms. As such, platforms are acutely
sensitive to network effects.viii The more users there are
on one platform, the more everyone benefits (more
and better choices, more ratings, etc.). In addition,
individuals who already have the app installed and
their details stored on Airbnb have a strong incentive
to stay with that platform. The costs of migrating to
a different provider become prohibitive, even if the
company or individuals running the company are
revealed to be engaged in dubious practices. Although
many consumers may very well prefer multiple service
providers, there are clear incentivizes pushing everyone
to stick with one dominant player, once that dominant
platform has emerged.
Second, the need to innovate continually, whilst at
the same managing the legal risk created by rapid
expansion, requires more centralized forms of
organization and governance. Platforms often start
with a simple, idealistic proposition (“let’s bring people
together”). But, over time, they add more and more
features, making their technological infrastructure
more complex.ix The downside of this is that more
developers are then needed to accurately deal with the
increased technology complexity and managing such
complexity requires more centralized and hierarchical
organizational forms with more elaborate control
mechanisms. This is particularly true of companies
that scale globally. And when platforms become more
prominent, they need to attract more investors and
investment to fund further innovation. Again, this
transforms the incentives of platform owners and
short-term performance becomes critical. To improve
financial performance or save costs, platforms may
feel the need to change the rules of the game from
one day to the other (without consulting the users of
the platform) and the belief that such agility is better
achieved with hierarchical and centralized governance
structures can easily gain ascendancy.
We might say that platforms have exhibited a tendency
towards two different types of centralization. On the
one hand, “cartelization,” in which fewer and fewer
players dominate the market for a particular service,
The JBBA | Volume 2 | Issue 2 | October 2019
and, on the other hand, “corporatization,” in which
there is an ever-greater internal concentration of
authority based on a clear and closed hierarchy.
If we accept this story of the inevitable decline of
platforms, how should we respond? Again, there are
competing views. Some claim that our only hope
is to use the power of the Leviathan (the state or
regional organizations, such as the EU) to rein in these
corporate giants, through more and better rules and
regulation.x Think anti-trust laws, data protection laws
or laws controlling online speech. According to this
line of thinking, we can’t fix the problems with more
technology. Recently, we can hear more and more talk
around this “top-down,” regulatory solution.
3. The Decentralised Alternative of the Blockchain
Evangelists
The Evangelists, however, take a different view. These
are not problems that can easily be solved by more or
even smarter regulation, as the power and reach of
the Internet giants is just too great for any regulation
to be meaningful or effective. The size of many
platforms makes them largely immune to state actionxi.
Instead, the Evangelists recognize the importance of
technology-based solutions that can provide us with
the vision and direction to build something better. This
is a view that needs to be taken seriously and it is in this
context that we need to think about distributed ledger
technologies, such as blockchain, and smart contracts.xii
The key claim of Evangelists is that things can be
different and that distributed ledger technologies
have the potential to bring about a transformation
to a better world.xiii To understand why and how, it is
helpful to briefly go back to the origins of blockchain
and the original white paper by Satoshi Nakamoto.xiv
In this first statement, Nakamoto proposed a system
for a digital currency that did not require a centralized
trusted authority to verify transactions. Two key
elements characterize the general system that was
proposed in the Bitcoin whitepaper:
Firstly, a database scattered across many computers,
with no single authority controlling and verifying
the authenticity of the data. Secondly, the “work” of
maintaining the database – what we now refer to as
“mining” – was rewarded with small payments, in the
form of tokens. If you used a part of your computer’s
power to maintain the integrity and security of the
database, you would receive a reward in the form of
tokens that could then be used to “buy” services or
sold to third parties for profit. These tokens would
grow increasingly difficult to earn over time, ensuring a
certain amount of scarcity in the system. If you helped
in the beginning (and helped the database to develop
and grow) you would receive a larger reward, thus
incentivizing early stage participation.
103
jbba
the
Evangelists believe that this combination of ideas
are revolutionary.xv Firstly, they provide a way of
agreeing on the contents of a database without anyone
being in charge of, owning or otherwise controlling
that database. Secondly, they provide a mechanism
for rewarding people that made the database more
valuable, but – crucially – without those people being
paid by an owner of the database or owning shares
in the corporation that controls the database. There
would be no owner or controlling corporation of such
decentralized databases. Nakamoto provided a model
for supporting open protocols that wasn’t available
when the first tech giants emerged. And, for this
reason, they have the potential to challenge the tech
giants and change the world.
But, how does this technology have potential to
transform capitalism and how is it connected to the
protocol layer of the Internet described above? A
comparison with Airbnb can be used to illustrate the
possibilities of a distributed ledger model. A new open
protocol could be created that contains a request: “I
would like a room in PLACE between DATES.” A
decentralized blockchain database might then record
the metadata of all users, such as personal information,
past trips, credit card details, preferences and user
and host rankings. The protocol for transmitting this
request out onto the Internet would be completely
open. Anyone (individuals, private companies, public
authorities) who wanted to develop an app for
responding to such requests would then be free to do
so. In this model, when you transmit your request, you
would not need to commit ex ante to a single provider
(as you do now with Airbnb), but you would instead
be free to announce your wishes to the world via the
secure protocol and wait for competing offers from
diverse providers of accommodation, ranging from
anyone with a spare room though to large multinational
hotel companies.
Tokens would be vital in allowing such a protocol
to develop and scale, and early adopters would be
rewarded with tokens that they could then use to
either buy accommodation services themselves
or sell on an exchange for real world currencies.
Moreover, early adopters (app developers, providers of
accommodation, etc.) would receive a proportionately
larger share of tokens for entering and helping to
develop the new ecosystem. As the protocol developed
it would then attract outside investors, which would
give the token a greater monetary value that, in turn,
would encourage more participation.
Critics might argue that one company or group of
companies might monopolize the new protocol, in the
same way that the tech giants of today dominate various
sectors of the Internet economy. Indeed, fully-fledged
decentralized blockchain networks do not exist yet.
Consider the technical and operational shortcomings
The JBBA | Volume 2 | Issue 2 | October 2019
of the Bitcoin blockchain. In discussions with
mathematicians and other technologists, the following
weaknesses are usually highlighted. Bitcoin’s proof of
work protocol has led to “mining pools” because of
economies of scale and unbalanced reward structures.
The anonymity in the blockchain network means that it
is prone to “Sybil attacks” and “51% attacks.”
Still, there are advantages in an open source plus
decentralized database model that makes such a
process of “cartelization” much less likely. For a start, it
wouldn’t present the same opportunities for abuse and
manipulation that you find in the closed, centralized
systems of Amazon, Facebook, etc. If a particular
service provider did something I didn’t like, it would
be much easier to switch to an alternative service
provider, as my information would not be retained by
the service provider on a centralized database, but a
decentralized, open source database connected to the
protocol. The open standard would have a discipling
effect on platform operators, as it would facilitate a
level of migration (to other providers or simply opting
out altogether) that is simply impossible today.
Tokens would also give a blockchain-based open
protocol a number of advantages, in that it would
provide an infrastructure to reward content creators.
This seems preferable to the current situation on many
platforms – especially social media platforms – where
most content providers act without compensation,
while the platform companies receive all the economic
value of that content by selling advertising.
Finally, there are the potential security gains of a
decentralized network. Would our personal information
or transactions be more secure in a distributed
blockchain than behind the elaborate firewalls of giant
corporations like Google or Facebook? An openly
readable ledger means anyone can check the integrity of
transactions. The distributed cooperation component
implies that “attackers” must be able to “out-compute”
the entire network (which is practically impossible).
4. Experiments in Decentralisation
The takeaway from all of this? We are currently
living in a fast developing “space” between two coexisting realities: a centralized “old world” reality
and an emerging but incomplete new “decentralized
reality.” The centralized reality with its hierarchical
organizations, rules, regulations, and institutions still
prevails. It appears unlikely that we will soon say
goodbye to our familiar, centralized procedures and
organizations anytime soon.
Nevertheless, a more decentralized reality has already
started to emerge.xvi As we have seen, trust in the
“centralized companies” is already declining (mainly
due to the concentration of power, wealth and
104
jbba
the
information), and distributed ledger technologies,
including blockchain, are viewed by many as offering a
superior long-term model. These technologies have the
potential to create real level playing fields, transparency
and applications that run exactly as programmed
without any possibility of downtime, censorship, fraud
or third-party interference.
We have already passed the “tipping point” in our
experimenting with decentralized technologies.xvii
There’s simply no going back. So, instead of being
locked into the traditional “centralized” world or
remaining trapped in the space between the two
realities, it is better to see how digital technologies
are shaping the “new world” and affecting all of our
relationships.
As such, it is necessary to become actively involved
in the further development of blockchain and smart
contracts and the creation of a decentralized reality.
Only, if we build the new reality together, will we ensure
that a decentralized world can reach its full potential
and offer greater transparency, convenience, and trust.
When we co-create the future together in this way, new
jobs, opportunities, possibilities will inevitably emerge.
And incorporating multiple perspectives – business,
mathematics, and law – will be essential to make sure
that we make the right decisions in our journey towards
a better decentralized world.
The broader context for this project is a number
of significant cultural shifts. Digital technologies
have already changed our expectations. Consumers
have become smarter, better connected, and more
demanding. They love the “speed” and “convenience”
offered by digital technologies and they are not willing
to give it up. The consumers’ “voice” has become more
powerful than ever before. As a result, their relationship
with business has changed dramatically. Even businessto-business companies need to take consumer views
more seriously.
Who, when and where people “trust” has also changed.
Whereas in the past, we relied heavily on institutions,
intermediaries, and other third parties, we increasingly
place our trust in digital systems and algorithms.
It appears that we have less and less confidence in
“old world” institutions. The speedy development of
distributed ledger technology (including blockchain),
smart contracts and artificial intelligence will only
further automate trust. Institutionalized trust is
replaced by “digital trust.” It is obvious that the
automation of “trust,” “faith,” and “confidence” has a
tremendous impact on worker-employer relationships,
the meaning of leadership, and how management
operates. The opportunity to communicate and
interact with peers directly (through social media and
without the interference of third parties) makes us
more entrepreneurial and creates new opportunities to
The JBBA | Volume 2 | Issue 2 | October 2019
be creative.
Our “new” relationship with digital technology also
makes it possible to have peer-to-peer connections,
communications, interactions, and transactions.
Algorithms and data-analytics help us find partners,
assistants, sponsors, help, accommodation, etc.
Of course, these digital systems aren’t flawless,
but the fact is that we increasingly rely on more
decentralised, peer-to-peer systems. The convenience
of these new systems attracts us. The looser (digital)
connections and interactions are so much faster and
more comfortable than the old “formal” ways of
making fixed appointments and ritualized meetings.
The Millennial generation, in particular, appears to
understand this. They view decentralization as a given
for autonomy, responsibility, and happiness. Millennials
– and this is a mindset, more than a generation – just
seem more attuned to the freedoms and possibilities
of a flatter world. They understand that hierarchical
structures and an overreliance on formal procedures
often discourage open and honest discussion, leading
to either indifference, apathy or burnout.
“Fully-fledged” decentralization doesn’t exist yet. But
the decentralization trend is evident, and we must be
better prepared. There is no time for procrastination,
and we need to become smarter about decentralization
in order to ensure that the Evangelist vision of the future
comes to fruition. Of course, current technologies
and developments aren’t perfect (misuse of data,
fake news, etc.). But these issues cannot be solved by
traditional and centralized means (regulations, etc.). We
must collaborate to find decentralized and tech-driven
solutions now.
i
See, for example, Carlota Perez, Technological
Revolutions & Financial Capital (2002); Timothy Wu,
The Master Switch: The Rise and Fall of Information
Empires (2010).
ii
See Scott Galloway, The Four: The Hidden DNA of
Amazon, Apple, Facebook, and Google (2017).
iii
Don Tapscott & Scott Tapscott, Blockchain
Revolution: How the Technology Behind Bitcoin is
Changing Money, Business, and the World (2016).
William Craig, 15 Biggest Internet Controversies of
the Past Decade, FX Blog, (2018) available at: https://
www.webfx.com/blog/web-design/15-big gestinternet-controversies-of-the-past-decade/.
iv
See Geoffrey G Parker, Marshall W. Van Alsyne &
Shandgeet Paul Choudry, Platform Revolution: How
Networked Markets are Transforming the Economy
and How to Make them Work for You (2016); Alex
Moazed & Nicholas J. Johnson, Modern Monopolies:
What it Takes to Dominate the Twenty First Century
Economy (2016).
v
105
jbba
the
The World Bank, for example, organized a Disrupting
Development event on this theme in Bali in October
2018, available at: https://live.worldbank.org/
disrupting-development.
vi
Erik P. M. Vermeulen, Three Ways to Grow Your
Business in a Digital Age, Medium (2017) available
at:
https://hackernoon.com/3-ways-to-grow-yourbusiness-in-a-digital-age-86e8bb3f33d1.
https://doi.org/10.2139/ssrn.3379443.
xiv
Satoshi Nakamoto, Bitcoin: A Peer-to-Peer
Electronic Cash System (2008), available at: https://
bitcoin.org/bitcoin.pdf.
vii
xv
Steven Johnson, Beyond the Bitcoin Bubble, New
York Times Magazine (January 16, 2018).
See Mark Fenwick, Wulf Kaal, & Erik P. M.
Vermeulen, Why Blockchain Will Disrupt Corporate
Organizations: What Can be Learned from the Digital
Transformation, Journal of the British Blockchain
Association, 1(2), 1-11, (2018) available at: https://doi.
org/10.31585/jbba-1-2-(9)2018.
xvi
See Paul Belleflamme & Martin Peitz, Platforms
and Network Effects, Handbook of Game Theory &
Industrial Organization 286-317 (2018); Nirmala Reddy,
How to Harness the Power of Network Effects, Forbes
(2018) available at: https://www.forbes.com/sites/
forbescoachescouncil/2018/01/02/how-to-harnessthe-power-of-network-effects/#2b41823462e8.
ix
Platforms often use open source software and a
“microservices” architecture to accelerate growth, be
adaptable to change, and give more value to the endusers of the services. Think of these platforms as a
collection of loosely coupled applications which are
configured to interact through internal and external
application programming interfaces (APIs). The
API approach provides flexibility and windows to
new and other platforms. It allows the platforms to
attract innovative ideas from third-party developers.
The downside is that more developers and more
automation are needed to accurately deal with
the increased technology complexity. This require
more investments (more of which later) and a more
centralized organization with more control and
governance mechanisms.
viii
See, for example, Mark Fenwick, Wulf A. Kaal,
Erik P.M. Vermeulen, The Unmediated & TechDriven Corporate Governance of Today's Winning
Companies, University of St. Thomas (Minnesota)
Legal Studies Research (2017) available at: https://doi.
org/10.2139/ssrn.2922176; Mark Fenwick & Erik P.
M. Vermeulen, Technology & Corporate Governance,
The Texas Journal of Business Law (2019) vol. 48(1),
1-22.
xvii
x
See, for example, Scott Galloway, The Four: The
Hidden DNA of Amazon, Apple, Facebook, and
Google (2017).
xi
Mark Fenwick, Joseph A. McCahery, & Erik P.
M. Vermeulen, (2019). The End of ‘Corporate’
Governance: Hello ‘Platform’ Governance. European
Business Organization Law Review, Vol. 20, No. 1,
171–199; Mark Fenwick & Erik P. M. Vermeulen,
A Sustainable Platform Economy & the Future
of Corporate Governance. European Corporate
Governance Institute Law Working Paper, No.
441/2019, p. 1-38 (2019) available at: https://doi.
org/10.2139/ssrn.3331508.
xii
Mark Fenwick & Erik P. M. Vermeulen, Time for
Regulators to Open the 'Black Box' of Technology,
Lex Research Topics in Corporate Law & Economics
Working Paper, No. 2019-2 (2019) available at: https://
doi.org/10.2139/ssrn.3379205.
xiii
Mark Fenwick & Erik P. M. Vermeulen. A Primer
on Blockchain, Smart Contracts & Crypto-Assets,
Lex Research Topics in Corporate Law & Economics
Working Paper, No. 2019-3, p. 1-20, (2019) available at:
The JBBA | Volume 2 | Issue 2 | October 2019
106
jbba
the
The JBBA | Volume 2 | Issue 2 | October 2019
107
jbba
the
perspective
Is Blockchain Part
of the Future of Art?
OPEN ACCESS
ISSN Print: 2516-3949
https://doi.org/10.31585/jbba-2-2-(10)2019
Stylianos Kampakis
University College London, UK
Correspondence: stylianos.kampakis@gmail.com
Received: 30 July 2019 Accepted: 13 August 2019 Published: 17 August 2019
Art is an important part of our culture, and economy.
The global art market reached $67 billion in 2018i.
While some individuals might purchase art solely for
their own enjoyment, for others it can be a status
symbol or an investment.
However, the world of art is not without its problems.
Two of the most important challenges are are fraud,
and ownership of digital assets. However, blockchain is
promising to solve both of these issues soon.
Fraud in artwork usually shows up in the form
of forging. While it is not easy to estimate the total
amount of money exchanged in forged art, it is clear
that individuals and museums might be losing millions
of dollars every year because of forging. In 2018, it
was reported that a museum in Franceii dedicated
to the art of Étienne Terrus, discovered that most
of the artworks were not real. There are also plenty
other famous cases of forgery last year, such as that
of an exhibition about Amedeo Modigliani in Genoa,
where 21 of the 30 artworks were confirmed as fakesiii.
While these paintings might have been worth millions
of dollars (if they were authentic), the fakes were
practically worthless.
These are only two of the forgery cases that took
place in 2018. It is easy to find more examples, but
what is shocking, is that a large part of the forgeries is
never uncovered. It is possible that a forged artwork
exchanges hands many times, until the final owner
realizes that the actual value is zero.
Given blockchain’s ability to help in the provenance of
goods, it is a natural ally in the battle against art forgery.
The problem of authenticity in art, is not different to
the problem of provenance in supply chains. A work
can be identified through a single identifier which can
be, for example, an image hash, such as perceptual
hashing. The ownership of the work can be stored
on the blockchain. A smart contract or a Ricardian
contract can be used in order to transfer ownership of
the artwork.
The JBBA | Volume 2 | Issue 2 | October 2019
There are different companies working on that problem
right now, like Vastari and Thomas Crown Art. While
no standard solutions have emerged, we are likely to
see one in the next few years.
Another important problem that blockchain is aiming
to solve, is the ownership of digital assets. While for
physical assets the only problem is forgery, digital
assets can be copied an unlimited number of times.
Therefore, until blockchain came about, it was
impossible to create digital collectibles.
The first instance of a blockchain-based collectible
was the Rare Pepe Wallet, in 2016, based on an internet
memeiv. However, the most monumental moment for
crypto collectibles was the creation of Cryptokitties in
2017. Cryptokitties is by far the most successful game
of crypto collectibles. In this game, the users own cats
that have certain attributes, like colour or weird features
like wings. The cats can mate with each other creating
new cats with unique combinations of attributes. There
are in total 4 billion cats that can be bred.
The game combines elements of collectible card
games, with the breeding mechanism that could only
exist inside a computer. The game reached a total
number of 1 million transactions in October 2018.
At the time of writing, there are multiple exchanges
for crypto-collectibles and blockchain-based artworks:
opensea.io, digitalobjects.art, rareart.io, pixura.io,
Known Origin, Maecenas and Makers Place are some
of them. Artists can easily secure ownership their
artworks on blockchain through Mintable or Pixura.
On some of those exchanges, you can find collectibles
and artwork that have reached higher price tags. Some
artworks on Digital Objects can go up to $1000. On
Open Sea, there are collectibles that are sold for 10
ethers or more, which, at the time of writing, amounts
to more than $3000. Finally, a digital card of Elon
Musk was recently sold for over $50,000.
Ethereum has fully supported crypto-collectibles,
109
jbba
the
through the ERC-721 standard. Much like the ERC-20
standard describes how to setup smart contracts for
fungible tokens, the ERC-721 standard, describes how
to setup a smart-contract for non-fungible tokens. That
is, all tokens that are using this standard are unique.
The aforementioned exchanges are all based on this
standard.
So, to answer the question that was set out in the
beginning of the article: Yes, blockchain is definitely
going to play a key role in the future of art, and we
saw in this short article two ways in which it is going
to disrupt the world of art. It is clear that there are
still some barriers to the widespread adoption of
blockchain.
Cryptocurrency prices can still fluctuate rapidly,
and the speculative bubble that burst in December
2018 might have hurt the credibility and popularity
of cryptocurrencies. Also, buying Ethereum and
exchanging is something that is not easy for everyone.
While in practice, tools like the Metamask Chrome
extension or the Brave browser make it easy to
use Ethereum, audiences that are less familiar with
technology might find this challenging. Given that a
large part of high net-worth art buyers might be of
older age, this can become a significant barrier.
However, given the usefulness of blockchain and its
rising popularity, we expect that in the next few years
accessibility will increase, and use cases will multiply.
Therefore, blockchain for art is here to stay.
Disclaimer: The author of this article is personally involved in this
space, by using generative adversarial networks to create works of art
and sell them on blockchain.
https://www.artsy.net/article/artsy-editorial-globalart-market-reached-674-billion-2018-6
i
https://www.theguardian.com/global/2019/jun/15/
french-art-museum-full-of-fakes-etienne-terrus
ii
https://www.telegraph.co.uk/news/2018/01/10/
modigliani-paintings-thought-worth-tens-millionsdenounced-fakes/
iii
iv
https://rarepepewallet.com/
The JBBA | Volume 2 | Issue 2 | October 2019
110
jbba
the
The JBBA | Volume 2 | Issue 2 | October 2019
111
jbba
the
For Authors
We are now accepting manuscripts for
Volume 3, Issue 1 (March 2020).
Please submit your article by using the document
template provided in the link below. Please do NOT
include any author/institute identifiable details
in the manuscript as this document is sent to the
reviewers for a 'double blind' review. The details
including author(s), affiliations, correspondence email,
acknowledgements/COI/ if any, should be provided
as a separate document.
https://www.britishblockchainassociation.org/
jbba-template
(Max. word count = 5000 words, excluding references)
Step-by-step guide to manuscript submission:
1.
2.
3.
Author submits the article via JBBA Scholastica site
by using the above template, which must include:
an article header, an abstract (max 300 words), a
conclusion, and references (in IEEE referencing
style). The abstract should reflect both content
and emphasis of the paper. Please use 'British
English' when spelling words, for example, write
'centralisation' with an 'S', and not 'centralization'
with a 'z'. The article submission fee is $10 per
artcile and is paid directly to Scholastica.
The article will undergo quick initial screening
by the Managing/ Associate Editor-in-Chief. If
it was deemed that the article is inappropriate
for the journal for a reason that can be quickly
ascertained, such as the subject matter being too
far from the scope of the journal, the authors
will normally be informed within 1-2 weeks of
submission.For all other submissions, we will first
seek to gauge the level of interest that the paper
will have, on the assumption that it is correct and
well written. This will normally mean sending
the paper out to the editor for "quick opinions",
after which its suitability will be discussed by the
Associate Editors-in-Chief.
The paper will then be sent for review. Post review,
there are 3 possible outcomes:
•
•
•
Accepted (will be allocated for publication)
Rejected
Revise and submit
4.
The author(s) is/are informed of the review
outcome by the Managing Editor. The final
decision for all manuscripts is taken by the Editorin-Chief or an Associate EIC. The handling
editor will make a recommendation – sometimes
a tentative one – and this will be discussed. In
The JBBA | Volume 2 | Issue 2 | October 2019
5.
cases of doubt, more quick opinions will usually
be sought. Some stages of the above process may
occasionally be bypassed if the content is so close
to the expertise of one or more of the editors that
extra external information is clearly not necessary
for a fair decision to be made.
We aim for a turnaround time of 5 weeks from
submission to publication.
References
References should follow IEEE style referencing.
IEEE referencing style, also known as the numerical
system, uses numerical citations in square brackets to
refer to a reference list at the end of the paper. You
may wish to chose the resources below to easily cite the
references in IEEE format:
http://www.citationmachine.net/ieee
OR
http://www.citethisforme.com/citation-generator/
ieee
Here is an example of indicating relevant reference
in the text:
"...The theory was first put forward in 1987 [1]."
"...Scholtz [2] has argued that......."
"...Several recent studies [3, 4, 15, 16] have suggested
that..."
"...For example, see [7]."
Check out the link below for more information on
IEEE referencing:
https://libguides.murdoch.edu.au/IEEE/text
Here is an example of how an IEEE reference list
should appear at the end of the paper:
[1] T. Kaczorek, "Minimum energy control of fractional
positive electrical circuits", Archives of Electrical
Engineering, vol. 65, no. 2, pp.191–201, 2016.
[2] P. Harsha and M. Dahleh, "Optimal management
and sizing of energy storage under dynamic pricing for
the efficient integration of renewable energy", IEEE
Trans. Power Sys., vol. 30, no. 3, pp. 1164–1181, May
2015.
[3] A. Vaskuri, H. Baumgartner, P. Kärhä, G. Andor, and
E. Ikonen, "Modeling the spectral shape of InGaAlPbased red light-emitting diodes," Journal of Applied
Physics, vol. 118, no. 20, pp. 203103-1–203103-7, Jul.
2015. Accessed on: Feb. 9, 2017. [Online]. Available:
doi: 10.1063/1.4936322
[4] K. J. Krishnan, "Implementation of renewable
energy to reduce carbon consumption and fuel cell as a
113
jbba
the
back-up power for national broadband network (NBN)
in Australia," Ph.D dissertation, College of Eng. and
Sc., Victoria Univ., Melbourne, 2013.
[5] C. R. Ozansoy, "Design and implementation of a
Universal Communications Processor for substation
integration, automation and protection," Ph.D.
dissertation, College of Eng. and Sc., Victoria Univ.,
Melbourne, 2006. [Online]. Accessed on: June 22,
2017. [Online]. Available: http://vuir.vu.edu.au/527/
Article Processing Charge
If your article is accepted for publication, we will ask
you to pay the Article Processing Charge (APC) of
£685 (£585 for members of the British Blockchain
Association). For full details about the APC and our
waiver policies, please visit the 'About us' section of
the journal:
https://jbba.scholasticahq.com/about
Listing sources of information at the end of a paper
is an important part of professional scholarship and
writing. It is highly suggested that all references should
be checked if they are complete and there should be no
missing or uncited references.
Papers that have not been published, even if they
have been submitted for publication, should be cited
as “unpublished”. Papers that have been accepted for
publication should be cited as “in press”. Capitalize
only the first word in a paper title, except for proper
nouns and element symbols. For papers published in
translation journals, please give the English citation
first, followed by the original foreign-language citation.
The APC covers the cost of administration, copy editing,
formatting, layout, online hosting, archiving, digital
object identifier, journal marketing, designing, print
publication and print distribution. Article processing
charges will enable full, immediate, and continued
open access for all work published in the JBBA. This
allows unrestricted access; to authors, through the
widest possible dissemination of their work; and to the
blockchain community in general, through facilitation
of information availability and scientific advancement
of distributed ledger technologies and allied disciplines.
Plagiarism Policy
The JBBA accept articles in the following
categories:
•
•
•
•
•
•
•
•
•
•
•
Original Scientific Research (Qualitative,
Quantitative)
Systematic Reviews
Meta-analysis
Conference Research Abstracts
Comparative studies
Case Studies & Essays
Book Reviews
Critical reviews and Analysis
Interviews / Opinions of Key Influencers/
Thought Leaders
Editorial
Commentary on latest issues and trends in
blockchain & DLT
We may also include selected mix of articles published
on our website. Interviews/ opinions are not peer
reviewed but all content must be approved by the
handling editor to assess suitablility for publication.
Editor-in-chief has overall responsibility for the
content, production and strategic direction of the
JBBA.
Time to publication
On average, papers receive a decision in 4 weeks from
first submission and accepted articles are published
online and indexed in an additional 14 days.
The JBBA | Volume 2 | Issue 2 | October 2019
We have a very stringent plagiarism policy in place and
all articles are screened on Viper Plagarism Checking
Tool for detection of plagiarism. We accept a plagiarism
score of less than 10%. This allows the highest possible
level of scholarly integrity and transparency in contents
published by the JBBA.
General Guidance for Authors
The editors request that all articles shall be submitted
via Scholasticahq portal using the word template
document provided via the above link and must include
an abstract. We prefer text in Garamond font, size 12,
double spacing except for references at the end of the
paper, which should be single space.
The Journal allows authors to deposit a copy of their
own work at an instituitional repository.
The JBBA does not publish the work that has been
published elsewhere. The only exception to that rule
are original research papers published as "pre-print
repositories" on SSRN or ResearchGate. Submission
implies that the work is not being considered for
publication elsewhere and that it has been approved by
all authors. Original research articles should not exceed
10 A4 size pages (c. 500 words per page, excluding
Tables and Figures).
Author names and contact information are provided
during the submission process. The person who submits
the paper via Scholastica is the corresponding author
and an active email address is needed.) The first author
114
jbba
the
or primary author is the person who conducted most
of the work described in the paper, and is usually the
person who drafted the manuscript. The “senior author”
is usually the last person named, and is generally the
one who directed or oversaw the project. The names of
the “contributing authors” appear between the primary
and senior authors, and the order should reflect their
relative contribution to the work. By completing the
submission, you automatically agree to the statement
that the manuscript has not been published elsewhere
and that it has not been submitted simultaneously for
publication elsewhere. Authors who fail to adhere to
this condition will be charged with all costs which
JBBA incurs, and their papers will not be published.
The text of accepted manuscripts can sometimes be
edited to enhance communication between the author
and the reader.
The link below provides useful instructions on how to
write an academic/ scholarly article:
https://canvas.hull.ac.uk/courses/371/pages/
academic-writing-style
Duties of Authors
Authors should submit original research work only
(except if it this is clearly not the intention of the
article – as might be the case, for example, with a
survey paper, interview, analysis, commentary). Any
results that are not due to the authors should be clearly
cited. Copying or paraphrasing substantial parts of
another paper without attribution is unacceptable, as is
any other form of plagiarism.
No paper should be submitted to JBBA that is
already published elsewhere or is being considered for
publication by another journal.
Those named as authors of a paper should have
made a substantial contribution to the paper, or to a
more general project of which the paper is a part, and
anybody who has made such a contribution should be
offered authorship.
Authors who discover important errors in their
articles, whether published or under consideration for
publication, should notify the journal promptly.
______________
The JBBA | Volume 2 | Issue 2 | October 2019
115
jbba
the
The Journal of The
British Blockchain
Association
VOLUME 1 | ISSUE 1 | ISSN PRINT: 2516-3949
INAUGURAL EDITION JULY 2018
Collaboration. Innovation. Excellence
EUROPE'S FIRST PEER REVIEWED SCIENTIFIC BLOCKCHAIN JOURNAL
jbba
the
The Journal of The
British Blockchain
Association
britishblockchainassociation.org/jbba
Volume 1 | issue 2 | issN PriNt: 2516-3949
he
editioNtdecember
2018
collaboration. innovation. excellence
decentralised healthcare on the blockchain
challenges of investigating
cryptocurrencies
& blockchain
related crime
Setting the "Gold
Standard"
in Blockhain
e-Voting on the
blockchain
International
Blockchain
Standards:
A Research
by ConsenSys
ethereum
registration
Authorities for
Private sidechains
Blockchain &
Post Brexit
United Kingdom
crypto ricardian
rent & a
crypto coase
theorem
utility tokens &
Quantity theory
of money
IBM and Walmart
Blockchain Pilot
by BRI
the cost of trust:
A Pilot study
Blockchain & GDPR:
A Delphi Study
blockchain research:
exclusive interview with
Andreas m. Antonopoulos
EXCLUSIVE INTERVIEW WITH
LORD HOLMES OF
RICHMOND MBE
UK BLOCKCHAIN
APPG VICE CHAIR
britishblockchainassociation.com/jbba
AcAdemic pArtners
SPONSORS & ACADEMIC PARTNERS
The JBBA | Volume 1 | Issue 1 | July 2018
Portrait courtesy The House of Lords
Academia.edu no longer supports Internet Explorer.
To browse Academia.edu and the wider internet faster and more securely, please take a few seconds to upgrade your browser.
Related Papers
2021 IEEE International Conference on Blockchain and Cryptocurrency (ICBC), 2021
Download
JMSS, A2Z Journals, 2021
Download
EPJ Quantum Technology, 2021
Download
Arxiv preprint quant-ph/ …, 2007
Download
Download
arXiv (Cornell University), 2007
Download
International Journal of Engineering and Advanced Technology (IJEAT), 2023
Download
Scripta Theologica 56 (2024) 532-534
Download
N/A, 2022
Download
Sulle orme degli Antichi. Scritti di filologia e di storia della tradizione classica offerti a Salvatore Cerasuolo, a cura di M. Capasso, Lecce 2016, pp. 75-110, 2016
Download
Anticipation: Learning from the Past. Cognitive systems monographs. Ed. by Mihai Nadin. Berlin: Springer., 2015
Download
