ISDA International Swaps and Derivatives Association, Inc.

February 7, 2011

Mr. David Stawick

Secretary
Commodity Futures Trading Commission
Three Lafayette Centre
1155 21st Street, N.W.
Washington, DC 20581

Re: (1) RIN 3038–AD08 - Real-Time Public Reporting of Swap Transaction Data;
(2) RIN 3038–AD19 - Swap Data Recordkeeping and Reporting Requirements; and
(3) RIN 3038–AC96 - Reporting, Recordkeeping, and Daily Trading Records
Requirements for Swap Dealers and Major Swap Participants

Dear Mr. Stawick:

The International Swaps and Derivatives Association, Inc. 1 (“ISDA”) and the Securities Industry
and Financial Markets Association 2 (“SIFMA”) (hereinafter referred to as the “Associations”) are
writing in response to three Notices of Proposed Rulemaking: Real-Time Public Reporting of Swap
Transaction Data (the “Real-Time Reporting NPR”); Swap Data Recordkeeping and Reporting
Requirements (the “Swap Reporting NPR”); and Reporting, Recordkeeping, and Daily Trading
Records Requirements for Swap Dealers and Major Swap Participants (the “SD/MSP
Recordkeeping NPR”, and together with the Real-Time Reporting NPR and the Swap Reporting
NPR, the “NPRs”) issued by the Commodity Futures Trading Commission (the “Commission”) to
implement provisions of Title VII of the Dodd-Frank Wall Street Reform and Consumer Protection
Act (“Dodd-Frank Act”).

The Associations respectfully submit the following comments regarding the NPRs. The comments
are organized as follows:

• The first section identifies issues and presents our suggestions for future action relating to
block trade exemption rules, which we regard as a critically important element of the
reforms contained in Title VII of the Dodd-Frank Act.

1
ISDA, which represents participants in the privately negotiated derivatives industry, is among the world’s largest global
financial trade associations as measured by number of member firms. ISDA was chartered in 1985 and today has over
800 member institutions from 54 countries on six continents. Our members include most of the world’s major institutions
that deal in privately negotiated derivatives, as well as many of the businesses, governmental entities and other end users
that rely on over-the-counter derivatives to manage efficiently the risks inherent in their core economic activities. For
more information, please visit: www.isda.org.
2
SIFMA brings together the shared interests of hundreds of securities firms, banks, and asset managers. SIFMA’s
mission is to support a strong financial industry, investor opportunity, capital formation, job creation and economic
growth, while building trust and confidence in the financial markets. SIFMA, with offices in New York and Washington,
D.C., is the U.S. regional member of the Global Financial Markets Association. For more information, please visit:
www.sifma.org
 2

• The second section sets out some general considerations that apply to all areas of the
NPRs.

• The third section addresses specific points relating to the reporting of trade information
under the Real-Time Reporting NPR.

• The fourth section deals with considerations relating to the reporting of collateral and
valuation information under the Swap Reporting NPR.

• The fifth section responds to the Commission’s questions relating to responsibility for
reporting, including consideration of issues relating to extraterritorial application of the
NPRs.

• The sixth section responds to the Commission’s questions relating to swaps on

commodities.

• The seventh section addresses considerations relating to recordkeeping and daily trading
records requirements.

There are two Annexes to this letter. The first contains a table mapping the comments in the
different sections of this letter to the specific questions contained in the NPRs. The second contains
a study entitled “Block trade reporting for over-the-counter derivatives markets” (the “Block
Trading Study”), which has been prepared by ISDA and SIFMA, with support from Oliver
Wyman, to begin addressing considerations relevant to block trades, as explained in further detail
in Section I below.

I. Block Trades - Appropriate Block Size Threshold and Public Dissemination Delay

The Associations consider the development of appropriate block trading exemptions from certain
of the requirements of real time public dissemination of swap information to be of critical
importance to the successful implementation of Title VII of the Dodd-Frank Act for the swap
market. This is also explicitly recognized in the Dodd-Frank Act, which requires the Commission
to specify the criteria for determining what constitutes a large notional swap transaction (block
trade) for particular markets and contracts and to take into account whether the public disclosure
will materially reduce market liquidity3.

The importance of appropriate block trade exemptions can be demonstrated through the following
examples. If a corporate end-user plans to raise a significant amount of capital by issuing a large
bond to investors, it is exposed to the risk that interest rates may rise by the time it is ready to issue
the bond. It can hedge that risk by entering into an interest rate swap with a market maker that is
willing to provide liquidity. The market maker would then typically hedge the risk it has just taken
on by entering into one or more interest rate swap or other hedging transactions with other market
participants, indeed the price of the interest rate swap will likely be related to the price at which the

3
Section 2(a)(13) (E) of the Commodity Exchange Act, as amended by the Dodd-Frank Act.
 3

market maker believes it can hedge the risk. If however the interest rate swap with the corporate
end-user is reported to the market, then other potential counterparties will know that a market
maker has executed a large swap and will be looking to hedge that risk in the market, and will
change their prices accordingly, causing a risk of loss to the market maker. A rational market
maker might react to this increased risk by either refusing to enter into the large transaction with
the corporate end-user (thereby reducing liquidity), or by increasing the price of the interest rate
swap offered to the corporate end-user to provide a buffer against the increased risk. The end-user
may react by choosing to break the trade into smaller pieces, thus exposing itself to the liquidation
risk that previously the market maker was tasked with managing. Any of these results is clearly
detrimental to the end-user’s interests, and will have a negative impact on that end-user’s ability to
raise capital, damaging investment in the U.S. economy.

Alternatively, if the corporate end-user, instead of issuing a bond, plans to raise capital using a
loan, the lender may hedge its credit risk to that borrower by buying single name credit default
swap protection on the borrower from a market maker that is willing to offer liquidity. In this case
the lender’s willingness to lend or the price of the loan it is willing to offer the borrower will in part
be determined by the price of that credit default swap offered by the market maker. The market
maker will, in turn, typically hedge the risk it has just taken on by entering into one or more credit
default swaps or other hedging transactions with other market participants. If however the credit
default swap entered into by the lender and the market maker is reported to the market, then other
potential counterparties will know that a market maker has executed a large credit default swap and
will be looking to hedge that risk in the market, and will raise their prices accordingly, causing a
risk of loss to the market maker. A rational market maker might react to this increased risk by
either refusing to enter into the large transaction with the lender to the end-user (thereby reducing
liquidity), or by increasing the price of the credit default swap protection offered to the lender. The
lender may react by choosing to break the trade into smaller pieces, taking on liquidation risk. Any
of these outcomes may result in a more expensive loan for the end-user. As in the example above,
this will reduce the end-user’s ability to raise capital.

There will also be instances when dealers assume significant risk when another dealer exits a
market and wishes to pass its entire derivative portfolio to another dealer – as has happened in the
commodity markets – or when a dealer fails. In these cases special accommodations need to be
made. Each individual transaction may not constitute a block but the portfolio as a whole generates
so much risk that any public reporting would deter a dealer from assuming such high levels of risk.

From the examples above, it can be seen that the risk of adopting block trading rules that are not
proportionate to the available liquidity of an OTC derivatives market is that end-users’ ability to
hedge their risk will be compromised or such hedging will become more expensive through a
reduction in the opportunities to hedge that risk or through an increased cost of that hedging
activity. The final rules should be constructed so that block trades can be both executed and hedged
without negatively impacting liquidity or end-user funding and issuance costs. The Associations do
not believe that either the distribution test or the multiple (social size) test of block threshold size
discussed in the Real-Time Reporting NPR is likely to be a sufficiently well-calibrated test to avoid
this risk. Furthermore, given that the distribution of transaction sizes in the swap market is likely to
be discontinuous and fat tailed, it is natural to expect that a significant percentage of swap
transactions would qualify as block transactions, making the suggestion of placing an aggregate
cap upon their occurrence unadvisable. It is, in our view, an error to extrapolate from the infrequent
use of block trades in futures markets that similar infrequency in OTC markets will not
 4

compromise liquidity. The current ability to trade in the OTC market without a block limits regime
is the reason why block trades in futures are infrequent. We appreciate the need to harmonize the
block levels across these markets but we are concerned that once swaps no longer provide this
outlet, a recalibration of block levels will be required, if liquidity is not to be materially impacted.

To develop appropriate and well-calibrated block trading exemption rules, the Associations believe
that significant detailed research on swap markets must be performed before the appropriate block
size threshold and reporting delay for particular swap transactions can be determined. The Block
Trading Study, attached as Annex 2 to this letter, was prepared by ISDA and SIFMA to begin the
research process, and is submitted for consideration by the Commission. The Block Trading Study
was undertaken to help inform decisions about appropriate block trade reporting rules for OTC
markets. It explores the goals of transparency, the importance of block trade reporting exemptions
and the experience of other markets with transparency regimes and then uses trade-level data to
identify unique characteristics of the OTC interest rate and credit derivatives markets. It also
includes specific analysis of the proposals contained in the Real-Time Reporting NPR. While the
Block Trading Study concludes that transparency can be increased in the OTC derivatives markets
while preserving liquidity, it also finds that the Real-Time Reporting NPR would have a significant
adverse effect on trading in less liquid instruments, because the proposed rules would impose block
minimum size requirements without appropriately differentiating between instruments with very
different levels of liquidity.

ISDA and SIFMA believe that, while the Block Trading Study is a significant contribution to the
analysis undertaken to date on this subject, substantial additional research into appropriate block
trade exemptions is still required. We therefore strongly support the Commission’s intention to
collect and analyze additional data on the swap market in the coming months and suggest that
research should be directed towards determining the size of a transaction that would likely “move
the market” (i.e. change the prices that market participants would demand or accept for a particular
swap transaction). The Associations recommend that relevant considerations should include the
average daily trading volume for the relevant product and the size of two-way markets typically
made by market makers, and that further investigation is required to ascertain whether these are in
fact determinative factors. The analysis should be performed separately for different asset classes
(in particular, applying the concepts discussed in the Block Trading Study to asset classes beyond
interest rates and credit derivatives) and likely for different products within each asset class, as the
appropriate test for one product may not be appropriate for another product; in fact, it may be
appropriate to use different tests to determine the appropriate block size threshold and/or reporting
delay for different products. 4 For large notional swaps that are not centrally cleared it can be
assumed that there is some non-standard element that involves the assumption by the swap dealer
(“SD”) of some non-standard risk category, for example the risk in tranches of credit derivatives
indices (which trade infrequently) or correlation risk across the yield curve. It may be appropriate
that uncleared swaps receive substantially lower minimum block sizes and longer reporting times

4
The Commission may also find instructive the Committee of European Securities Regulators (“CESR”) proposal which
supports deferred publication of equity transactions. We recommend the Commission focus its attention on the CESR
framework, which establishes reporting intervals based on a matrix that looks both to the characteristics of the individual
transaction and the liquidity characteristics of the market for the relevant underlying security. The CESR proposal
permits reporting to occur at the end of day and where there are potential reductions in liquidity close to the end of a
trading day, CESR recommend extending the end of day deadline to early the following trading day for trades executed
late in the day. This approach is designed to ensure that the vast majority of deferred trades are reported no later than the
end of the trading day on which they are executed while still providing protection for trades occurring late in the day.
 5

than cleared swaps. Accordingly we recommend that the study give particular attention to the topic
of large notional swaps.

The Associations recommend that independent academic research be undertaken to supplement the
Block Trade Study and to determine the appropriate methodology for determining block size
thresholds, public dissemination delays and the information publicly disseminated for block trades.
ISDA has previously helped to co-ordinate similar research that examined the status of
transparency in interest rate and credit derivative markets. This research was first committed and
then presented to an international group of supervisors, including the Commission 5. ISDA would
be pleased to work with the Commission to help co-ordinate a similar study in relation to block size
thresholds and reporting delays, and recommends this course of action to the Commission.

The type of study envisioned above would require sufficient time to arrange and complete. We
estimate that work could be completed by the end of the first quarter of 2011 (or within three
months of the commencement of the study). This timing may be later than the Commission’s
anticipated publication of specific block trade thresholds. However it should be stressed that this
need not delay promulgation of the rules in the Real-Time Reporting NPR, merely the calibration
of the block size thresholds and the appropriate reporting delay for block trades, which could be
determined and published at a later date, independently of the other elements of the Real-Time
Reporting NPR.

The definition of “swap instrument” should be defined with sufficient resolution such that block
sizes established in relation to each swap instrument are appropriate. Although we recognize the
need to balance simplicity with precision, we believe that too simplistic an approach will be
damaging to liquidity. Even within the interest rate swaps category there are some types of swaps
that merit different treatment. This need is pronounced in commodities markets

Determining block sizes for options based simply on notional size of trades fails to have regard for
the risk profile of an option, which varies as much by strike as notional size.

In whatever methodology is eventually selected to determine block size thresholds, it is important

that specific block size thresholds be updated frequently, at a minimum of once every three months,
to reflect the latest market data, because liquidity in OTC markets can change quickly 6.

We agree with the Real-Time Reporting NPR that if a transaction is a block trade, then the size of
that transaction (other than the fact that it is a block trade) should not be disclosed at any time,
similar to the Financial Industry Regulatory Authority’s Transaction Reporting and Compliance
Engine system (“TRACE”) and as further discussed in the Block Trade Study.

Referring to the distinction drawn in Section II below between “execution” level data and
“allocation” level data, the final rules should be clear that the determination of whether a
transaction is a block trade occurs at the execution level (in any event as a practical matter, for the
reasons noted below in Section II(a)(iii), only the execution level data may be available in real time
to determine whether the transaction is a block trade). Where a transaction is executed
electronically, this may already be implied because the electronic platform will not receive any

5
For details of the commitment, please see the letter dated March 1, 2010, available on the website of the Federal
Reserve Bank of New York: http://www.newyorkfed.org/newsevents/news/markets/2010/100301_letter.pdf
6
For example as “on-the-run” products become “off-the-run”.
 6

allocation information and will therefore record the transaction at the execution level. This
clarification is therefore particularly applicable where the transaction is not executed electronically.

As with any proposed rule, the Commission could adjust the block size threshold tests over time to
reflect market impact, and time should also be allotted to account for the fact that block trade size
thresholds are new and a trade reporting system will have to be designed and implemented to
address the proposed rules. We therefore recommend that block trade requirements be phased in
and kept under periodic review. Please see further comments on phase-in generally in Section II(e)
below.

II. General Considerations

In this section we set out some general considerations that apply across the broad spectrum of
points relating to the NPRs.

(a) Consistency between CFTC and SEC rules and overseas regulators

Many market participants will likely be subject to parallel reporting requirements imposed by the
Commission, the Securities Exchange Commission (the “SEC”) and overseas regulators. To
remove inefficiencies, simplify compliance obligations and enhance regulatory agency capabilities,
it is vital that the Commission, the SEC, and overseas regulators adopt consistent reporting
requirements, including a common implementation effective date, particularly where transactions
in certain asset classes (such as credit derivatives) reported to the relevant swap data repository
(“SDR”) may be subject in some cases to the Commission’s rules and in other cases to SEC rules.
Inconsistencies between the NPRs and the SEC’s Proposed Regulation SBSR—Reporting and
Dissemination of Security-Based Swap Information 7 (the “SEC Proposed Regulation”) should be
minimized to enhance compliance.

We have identified the following specific points that we think necessitate consistent ruling between
the Commission and the SEC:

(i) The set of information to be publicly reported in real-time is quite different between
the two sets of proposed regulations. The Real-Time Reporting NPR is more specific
in terms of the set of information that is required, and also asks for a broader set of data
elements.

(ii) In the model for reporting swap continuation data across asset classes, it is also critical
to have consistency in the regulatory approaches. We would suggest that the approach
for reporting swap continuation data (i.e. life cycle approach or snapshot approach) not
be prescribed by regulation as proposed, but instead SDRs should be allowed to
develop in the most efficient way to meet the objectives for the relevant asset class.
Additionally, SDRs should be allowed to develop and improve on processes that will
provide operational benefits to the industry, as well as meet regulatory requirements.
We also note some issues specific to the life cycle event data reporting requirements in
Section II(f) below.

7
75 Fed. Reg. 75208 (December 2, 2010)
 7

(iii) Within each separate product type, the Commission and the SEC should harmonize
rules to define when the timeline for reporting a transaction will commence for that
product. In particular, the time at which a transaction becomes legally binding may not
be the same for all products. Where the reporting timeline is based on market activity
such as “affirmation”, “execution” and “confirmation”, the use of those terms should
reflect long-standing market conventions that differ according to the type of underlying
reference asset. Harmonization of use of such terms in the Commission's and SEC's
rules for a particular product type will foster operational efficiency, lessen the
incidence of errors, and place fewer burdens on reporting parties. Further observations
on the use of these terms and their application to total return swaps (“TRSs”), in
particular, are set out below:

(A) In the Real-Time Reporting NPR, “affirmation” is proposed to be defined as

counterparties’ verifying that they agree on primary economic terms but not
necessarily all of the terms, as distinguished from confirmation and, in many
cases, execution if execution does not occur when the parties affirm agreement
to primary economic terms. “Execution” is the agreement between the parties
that legally binds them. 8

(B) For example, and as noted in the Associations’ submission to the SEC in
response to the SEC Proposed Regulation, for certain equity TRSs, “affirmation”
addresses initial steps undertaken in advance of execution or confirmation; a
swap order is initiated at the “affirmation” stage but is neither executed nor
confirmed at this time. Affirmation can occur at the time or shortly after a trade
is preliminarily discussed between two counterparties but occurs before material
terms such as price and quantity are determined and the swap is executed or
confirmed. Following affirmation, intra-day hedge transactions are executed on a
regulated exchange and reported in real-time, in connection with, but separate
from, the TRS which has yet to be executed or confirmed. Any hedge
transactions entered into in advance of the TRS transaction are executed and
confirmed independently of the TRS. In order for reporting to be meaningful, the
material terms of the TRS must be available to be reported. If price, a material
term of the TRS, is not arrived at until after the hedge is consummated, then the
parties cannot confirm the swap until such time. The legally enforceable TRS is
made by way of swap transaction confirmation, which is agreed upon after the
time that preliminary swap terms were affirmed and after independent hedge
transactions are executed. For TRSs involving material terms such as pricing,
which occurs derivatively based on the price available in the market end of day,
the full terms of the TRS are not formed until end of day and therefore the TRS
is not executed until end of day and confirmed thereafter. In these circumstances,
after the TRS is confirmed by written trade confirmation, it may be reported in
real-time.

8
As the Commission points out in the Real-Time Reporting NPR, “execution can occur immediately following or
simultaneous with (the pre-execution) affirmation; the proposed definition of execution does not attempt to define what
constitutes a legally enforceable contract, only that execution occurs if and when the parties have formed a legally
enforceable contract, which is a matter to be decided by applicable law.” (Real-Time Reporting NPR (75 Fed.Reg.76140
at page 76144)).
 8

(C) TRS transactions in other asset classes often also involve different stages. For
the majority of swap transactions in Commodity index TRS, price is typically
determined after affirmation. For these transactions, it is more appropriate to
report when price or quantity have been determined, which occur later in the day
or at the end of the day, rather than report at the affirmation stage. Similarly, for
bespoke credit or interest rate TRS there may be instances where affirmation
occurs separately from execution or confirmation, such that the material terms of
the swap are not available until a time after affirmation occurs. Under these
circumstances, reporting should occur after the full description of the trade
becomes available.

(b) Trade allocations

It is common practice in the OTC derivatives markets for an asset manager to enter into a
transaction with a counterparty for a particular notional size for an agreed price (the “execution”
level), and for the asset manager to then allocate parts of that notional amount to multiple
underlying funds (the “allocation” level). Each fund is a separate legal entity, and so the agreement
at the execution level will ultimately result in several separate transactions at the allocation level.

As a result, we recommend the following:

(i) For the purpose of public real time trade reporting, the objective of which is
transparency, participants should report the trade as executed by the desk. The
reporting counterparty will not need to receive the allocation information from the
client for the purpose of meeting the real time reporting obligations. Furthermore, this
report will effectively reflect the pricing and size of the trade. This is also consistent
with reporting under TRACE.

(ii) For the purpose of trade reporting to the SDR, by contrast, the allocation of the trade to
the respective counterparties will be essential to understanding the final dispersion of
risk derived from the initial trade. For transactions where the counterparty allocates to
multiple funds (or other entities), therefore, the requirement to report should be
triggered from the time when the reporting party receives the allocation from the
customer - which is not typically within the reporting counterparty’s control.

(c) Unique identifiers

The Associations believe it is critical to introduce one set of uniformly-applied unique identifiers
within the derivatives industry for legal entities/counterparties, products and transactions, each as
discussed further below. We encourage the Commission, together with the SEC, the Department of
the Treasury and other regulators (including overseas regulators), to explore current best in class
models and mechanisms and adopt best practices for the derivatives industry (e.g. DTCC gold
standard). Industry utilities (meaning not-for-profit, industry-governed solutions) should be
considered for assigning these unique IDs. Furthermore, we encourage the Commission to attempt
to leverage existing market constructs used in the cash securities markets. Confidentiality and the
protection of information is also critical.
 9

The Commission should consider adopting a convention for assigning unique IDs and
incorporating a pilot or early adopter program for certain products and participants that will allow
for end-to-end testing and a proof of concept. For example, as previously suggested to the SEC, a
pilot program could consist exclusively of single name CDS traded by security-based swap dealers.
The identifiers need to be universally adopted and the industry is committed to use the standard
identifiers as and when they become available but allowing for an appropriate implementation
period. There are a number of trade events, such as allocation, clearing, novation and compression,
which will need to be described with respect to unique identifiers. The requirements for unique
swap identifiers (“USIs”) must complement these events rather than dictate how they function. For
example, the requirement for USIs should not prevent a DCO holding positions, as opposed to
individual trades, for standardized instruments. A newly formed ISDA cross-product data working
group, with representatives from sell side and buy side institutions, will look at proposed solutions
and the practical implications of unique identifiers for the derivatives industry. Such an exercise
should also cover the application of unique identifiers to pre-enactment swaps.

For legal entity identifiers (“LEIs”), we broadly support the principles set forth by the Office of
Financial Research (“OFR”) 9 and believe that the LEI should serve as the unique counterparty ID.
The Commission proposes a universal, international standard based on a voluntary consensus
standards body and states that it will prescribe its own method to create unique counterparty
identifiers (“UCIs”) to be used in reporting if no internationally accepted identification system
acceptable to the Commission is available prior to the implementation date of the final regulations.
We strongly recommend that a single ID be implemented and that one entity administer the unique
ID system to avoid the development of inconsistent standards. The solution needs to be
international; the entity operating the LEI issuance should be not for profit and operate on the
principle of cost recovery. The industry should decide on the appropriate model for cost recovery.
Additional input is needed to decide the right key minimum elements and their definition, which
should also be determined by the industry. In a letter to the OFR on their statement, the
Associations and several other trade associations commented that they “agree with the Linchpin
Discussion Paper that the number of data elements be kept at to the minimum necessary to assure
the uniqueness of each legal entity”. 10 We request that the Commission clarify that its UCI will be
the same as the LEI to avoid any ambiguity and further the goals of harmonizing identifiers.
Finally, we strongly urge the Commission to coordinate with all of the major domestic and global
financial services regulators to ensure this standard identifier system is enacted and enforced on a
consistent, global basis.

ISDA is willing to assume responsibility for developing the product identifiers for OTC derivatives
products that reflect the FpML standard. ISDA would work closely with the Commission in this
standardization effort. For this process we will follow the same general principles laid out for LEI.
In the first instance, this work will focus on product identifiers for cleared products. ISDA/FpML is
currently working on a pilot project with certain derivative clearing houses to provide a normalized
electronic data representation through an FpML document for each OTC product listed and/or
cleared. This work will include the assignment of unique product identifiers. We believe that
before the requirement for the mandatory reporting of trading activity is implemented, the industry

9
OFR discussion paper: “Creating a linchpin for Financial Data: The Need for a Legal Entity Identifier”, December 10,
2010.
10
For further industry commentary on LEIs, please see the letter from the Associations and several other trade
associations addressed to the Office of Financial Research dated January 31, 2011 Re: Statement on Legal Entity
Identification for Financial Contracts.
 10

standardization effort will need to have been completed and the industry given sufficient time to
adopt these changes.

The Associations fully support the effort to also define a formal product topology, which would be
essential to perform aggregations and promote transparency. Considering the evolving nature of
OTC derivatives, we recommend that the Commission should not be prescriptive in this respect.
The definition of the actual product topology would be more effectively developed through
professional organizations like ISDA or SIFMA, which could bring together participants and
regulators to (i) define an effective topology (as an example, the proposed “Contract Type”
topology might not capture cases such as interest rate cap), and (ii) adapt it to reflect the evolution
of OTC products. The FpML standard includes product schemes, which should be used as the
starting point for such product topology. In considering the efficient development of product
identifiers, the Associations believe that additional dialogue between the Commission, other
interested regulators, and the industry is needed to understand the purpose and intended use of the
product identifiers so that they may be appropriately tailored. For example, while we agree that a
granular product identifier could provide an informational benefit for “asset based” derivatives that
are linked to underlying cash or physical products (CDS or commodities, for example), the benefit
of this data becomes less clear when applied to interest rate swaps that lack this linkage. Clarity on
the aims of the Commission in collecting this data will best ensure development of the appropriate
model.

With respect to USIs, we similarly suggest that the Commission state clear objectives rather than
be prescriptive as to what the exact implementation will be, because it would be extremely difficult
to define upfront an exact implementation that would cover all use cases. As an example, the
proposed approach of having the USI be assigned by the reporting party may result in unexpected
behavior in the case where that party assigns the contract (partially or completely) to another party.
Also, while the proposed approach of having the reporting participants assign the USI is valid in
the case where the role of the SDR is limited to collecting and reporting the trades, there may be a
need for the SDR also to assign its own USI if its role extends to operational functions (as
illustrated by the DTCC implementation for credit derivatives). Furthermore, as the OTC
derivatives markets are currently characterized by a diversity of business processes across asset
classes (and even within asset classes), standardization will have a dramatic impact upon the
participants’ systems and workflows. In the OTC derivatives space, an initial “transaction”
typically defines a unique instrument that only exists between the two parties to the contract. Other
“transactions” may change the counterparties to the instrument (i.e. assignment, partial assignment,
novation, etc.) or modify the transaction (e.g. an amendment) but the instrument remains unique.
At the end of each day, a party to the instrument will have a position in that instrument, sometimes
referred to as a “trade” or “contract”. For certain instruments, they may become standardized
enough that they cease to be unique (for example index CDS trades), in which case all reporting
could be based on a party’s net position in that standardized instrument, in a similar manner to
exchange-traded derivatives. If the Commission can state clearly their objectives, then the industry
will work quickly to come up with ways of implementing as it has previously under the
commitment letter process.

(d) Error reporting

The Associations support the objective of prompt correction of errors by the reporting
counterparty. We however want to point out that most market participants rely upon systems that
 11

do not record the specific reason for an amendment. As a result, we recommend that while such
errors should promptly be adjusted by market participants, the specific root cause of such
amendments (for example a booking error or a trade amendment between parties) could be omitted
in line with current practice in listed markets. In addition, we urge the Commission to clarify that
reporting parties are not responsible for data which is inaccurately transcribed or corrupted after it
has been submitted to a SDR, and also have no duty to correct data errors of which they are
unaware.

While the industry has done much to improve the speed at which trades are confirmed in recent
years, it has done so over time and without sacrificing accuracy. The time frames proposed by the
Commission are significantly more aggressive than what the industry has committed to in the past
and it would be unfortunate if this were to lead to an increase in errors. We recommend the
Commission aim for an appropriate balance between speed and accuracy in proposing time frames
for regulatory reporting.

(e) Phase-in implementation

It is difficult to comment on the appropriate phase-in periods for the rules contained in the NPRs
until the precise details of all reporting obligations are available in final form. However, in general
terms, the phase-in period should be sufficient to afford the industry the time needed to build the
technology infrastructure required to comply with regulations. We believe that virtually all existing
systems would have to be significantly overhauled to satisfy real-time reporting obligations of the
Real-Time Reporting NPR. The phase-in period should take account of the work needed for market
participants to establish connectivity to the SDR for the relevant asset classes once the final
standards for data provision are known, including the determination of unique identifiers, as well as
the time needed for the SDRs themselves to be properly established. This phase-in could take the
form of staggered or delayed effective dates for regulations, as contemplated by the SD/MSP
Recordkeeping NPR. We expect that it will be technologically challenging to establish an SDR in
each asset class 11, however given sufficient time, we do believe this will be achieved. 12 (However,
it is possible that sufficient differences may exist within the Commodities asset class to warrant
separate SDRs at a sub-class level, e.g. Metals, Energy). We also expect that each SDR will operate
across the G20. If a set of SDRs per jurisdiction results, this will require duplicative
implementation and costs that would be carried through to future enhancements. Additionally, this
would carry the risk that a population of trades is not reported to any SDR impacting the
completeness and accuracy of information available to the Commission and other regulators. While
there will be challenges in providing a single global access model to global regulators, we urge the
Commission to address this in consultation with overseas regulators. We request clarity on time

11
ISDA has previously notified the Commission that the designation of a single registered SDR per class of swap would
provide the Commission and market participants with valuable efficiencies and expressed views regarding the adoption
of Financial Products Markup Language (“FpML”) as the protocol for reporting swap transactions to a SDR or the
Commission. We re-iterate those views in the context of the NPRs. Please see the letter from ISDA and the Futures
Industry Association to the Commission dated November 12, 2010 Re: Interim Final Rule for Reporting Pre-Enactment
Swap Transactions (75 Fed. Reg. 63080).
12
By way of examples: There is currently no infrastructure in place to support alternative approaches for data reporting
for commodity swaps. The ISDA Commodities Steering Committee is working on building out an SDR, which will begin
by reporting on financial oil. More detailed information is available by coordinating with the Commodities Steering
Committee. For equity swaps, industry participants are in the early stages of being able to report to a SDR. While there is
significant additional work required to further this effort, to the extent that some undertakings have been made these
alternatives should be explored further before the Commission implements entirely new reporting rules that require new
and costly operational infrastructure to support.
 12

lines, which we suggest should be estimated based on SDR registration and capability testing.
Requiring compliance via non-electronic methods is not recommended, as this would increase
systemic risk within the industry. Similarly, for the Commission to have to receive raw data from
market participants would likely not be effective; clarification of how this would work in practice
is required.

The industry has worked successfully with regulators in recent years to develop an industry
infrastructure that has proved effective in reducing systemic risk and promoting regulatory goals,
notably the process of commitment letters delivered to the Federal Reserve Bank of New York and
other regulators. The Associations would welcome the opportunity to work further with the
Commission and other regulators in a similar framework to structure the necessary development in
the most effective manner and monitor progress towards established goals. For such an approach to
be successful, the Associations would suggest that implementing rules reflect the outcome of such
work and can be executed within such a framework.

As an example, the Credit Derivatives Trade Information Warehouse was implemented using a
phase-in approach; new trades for dealers were first sent to the warehouse 12 months after work
commenced and phased implementations over the following two years addressed on-boarding of
clients and back-loading of trade populations. Over time the population of credit derivatives
included in the warehouse has increased and timeliness of confirmation has improved through the
industry commitment process outlined above.

One aspect of phase-in that is not contemplated in the NPRs is a gradual phase-in of the targeted
timeframe for reporting “real-time” information. By analogy with TRACE, the time required for
reporting when the system was first introduced was 75 minutes, and over a period of several years
this was reduced to 15 minutes as evidence was compiled that such reductions could be safely
achieved technologically and without adverse market impact. The reporting requirements for swaps
are significantly more complex than for TRACE, therefore the phase-in should reflect this degree
of complexity.

In addition, any concerns related to confidentiality of data should be addressed prior to the Real-
Time Reporting NPR being implemented. The fields to be publicly disseminated should be clearly
defined in the final rules.

(f) Life cycle event data

The Swap Reporting NPR would require the reporting of all life cycle event data on the same day
in which any life cycle event occurs. Life cycle events would include any corporate action affecting
a security or securities on which the swap is based (e.g., a merger, dividend, stock split, or
bankruptcy). We would note two main concerns with this definition and the related reporting
requirement.

• First, we refer to the SEC Proposed Regulation, which expressly excludes any “event that
does not result in any change to the contractual terms” of the security-based swap from its
definition of “life cycle event”. This exclusion is appropriate, particularly since credit and
equity swaps typically contain contractual provisions to address adjustments needed to
reflect events, such as a merger, dividend, stock split or bankruptcy. Any adjustments to
the credit or equity swaps are made pursuant to the contractual terms of such credit or
 13

equity swaps; therefore, such adjustments should not be considered a reportable event.
Consequently, we would suggest that the Commission adopt the approach taken by the
SEC. 13

• Second, as a practical matter, the reporting party will not necessarily know that such a
corporate action has taken place or all the relevant details, on the same day that such a
corporate action takes place. In fact, the relevant details of the corporate action may not be
available to the market until the issuer of the security makes a public filing (or even a
series of public filings) in connection with such corporation action or the information
becomes otherwise available through press releases or reporting by the media. The timing
of such public filings or when information becomes available is outside the control of the
reporting party; therefore, the reporting party may not be able to comply with the
requirement to report all life cycle event data on the same day in which such life cycle
event occurs.

With respect to any event that is not already addressed by the contractual terms of the swap, an
alternative approach would be to require the reporting of any change to data previously reported
with respect to a swap at the time an adjustment to such data is made due to a life cycle event. We
would also suggest that the life cycle event itself does not need to be reported since that
information would be in the public domain. This approach would still achieve the objective of
keeping the swap data up-to-date, but at the same time allow the reporting party to be able to
comply with the reporting requirement in a timely manner.

III. Reporting of Trade Information

ISDA and SIFMA support the objective of real time reporting for swap transactions contained in
the Dodd-Frank Act and the Real-Time Reporting NPR. Initial trades reported should carry a
primary reference number, and all amendments of that trade would then produce iterations of the
original reference number. Initially trades would be submitted with primary economic data. Upon
receipt of additional information pertaining to the original trade (e.g. trade specific allocations,
partial or full termination), a subsequent version of trade will be submitted reflecting associated
amendments.

(a) Information to report

We make the following specific recommendations regarding the set of information that has been
identified to be reported:

(i) The Commission specifies the set of required data elements required for both real-time
public reporting and swap data recordkeeping and reporting. Our understanding of the
intent of the NPRs is that the former set of data elements will be a subset of those

13
On a similar note, we would also note that the SEC has excluded events such as a scheduled expiration of the security-
based swap and a previously described and anticipated interest rate adjustment. These events would not need to be
reported under the SEC Proposed Regulation. These same events are included in the definition of “contract-intrinsic
event” under the Swap Reporting NPR and are required to be reported. We would suggest that the Commission adopt the
approach taken by the SEC on these events as well.
 14

required for swap data recordkeeping and reporting. As a result, we recommend that
the Commission specify the data elements required for swap data recordkeeping and
reporting in addition to the information to be reported in real time. This would provide
more clarity in the regulation and avoid the risk of inconsistencies when specifying
those data elements. In this context, we again note the crucial need for coordination
and consistency between the data requirements adopted by the Commission, the SEC
and overseas regulators. In addition, we respectfully suggest that the “Minimum
Primary Economic Terms Data” specified as part of the Real-Time Reporting NPR are
inconsistent across asset classes in some ways that are not justified by economic
differences, nor by differences in the information required for regulatory supervision.
The proposed rules require a set of data elements for credit and equity swaps (the
qualification of the counterparty, the execution and clearing venues, the settlement
terms, the data elements necessary to determine the market value of the transaction)
which could be applicable to other asset classes. Conversely, the contract type and the
timestamp for submission to the SDR are required only for FX, interest rates and other
commodity swaps. This also does not appear to be justified by differences in the asset
classes.

(ii) We suggest that the following requirement, part of the “Minimum Primary Economic
Terms Data” specified in the Real-Time Reporting NPR, be clarified: “If the
transaction involved an existing swap, an indication that the transaction did not involve
an opportunity to negotiate a material term of the contract, other than the
counterparty”. If this is intended only to refer to an assignment or novation of an
existing transaction, then this should be made explicit.

(iii) The requirement to associate the execution time, to the second, with each of the
reported trades, would prove extremely challenging and invasive in the case of voice
trades, for which the entry time in the participants’ systems is typically provided, but
not the execution time.

(iv) We make two further recommendations in relation to customized swaps:

(A) We believe that real time reporting and public dissemination of information
relating to customized swaps, meaning a swap having any amount of
customization away from benchmark/standard swaps, will add little to no price
discovery value as their terms will not be comparable with benchmark/standard
swaps. Furthermore, we believe that such reporting would introduce the risk of
providing price information that could potentially be misunderstood by some
market participants. As a result, we recommend that such trades be excluded
from the public dissemination of real time information (but not from regulatory
reporting requirements under the Swap Reporting NPR).

(B) The marketplace experience in the development and usage of FpML as an

electronic algorithmic representation of OTC derivatives leads us to recommend
a pragmatic approach for the trade representation of such customized swaps
which would be limited to a set of generic fields and be supplemented
(potentially at a later point upon request) by the actual confirmation (through a
format such as PDF). This would be consistent with the current approach for
 15

“Copper” records in credit derivatives and would facilitate the support of all
trades in an asset class within a single SDR. Further, this approach would
facilitate the monitoring of customized swaps and help direct efforts to expand
the population of fully supported trades.

(v) As a general matter, the Associations urge the Commission to limit real time reporting
requirements to new trading activity (including stepping into an existing transaction by
assignment or novation). For example, transactions resulting from portfolio
compression exercises do not reflect trading activity and therefore contain no market
information. As a result, we recommend that these types of events be excluded from
the real time reporting requirement for price discovery purposes, but be included as
part of the ongoing trade update reporting to the SDR at the end of the day (as they will
impact trades that would have already been reported). We would further suggest that
an inventory of activity that should be excluded from real time public reporting is
established by asset class with input from industry groups.

(vi) We believe that, in the case of some asset classes, there is not a universal definition of
the notional amount of a trade. This is particularly the case where the notional is not
confirmable information. We therefore recommend that, as part of the NPRs, the
Commission provide guidelines for reporting the notional amount, such as those
already developed by the Federal Reserve Bank of New York 14.

(vii) The final rules should be clear that the information required to be publicly
disseminated cannot identify the participants to a swap or provide information specific
to the participants. Such information would include the title and date of any master
agreement, and premiums associated with margin, collateral and independent amounts.
The data element “Additional Price Notation” should not be included in real-time
public reporting as this provides information on one party’s view on the
creditworthiness of its counterparty which could have a negative impact on that
counterparty if disclosed. Additionally, bilaterally executed trades may contain a
premium over market value that, while not associated with margin collateral or
independent amounts, would need to be excluded from Real Time Public Reporting in
order to prevent the price of the trade being misinterpreted by market observers.

(viii) The Commission requests comment on whether date information for swaps should be
rounded to the nearest tenor/month. Many swaps meet specific requirements for end-
users. To limit or manipulate data elements that are part of the Primary Economic
Terms in order to allow trades with differing terms to be aggregated will reduce post
trade transparency. We recommend that this proposal not be implemented.

(ix) The Commission requests comment on whether any data fields in the Master Reference
Generic Data Fields List should be included in one or more of the Tables of required
Minimum Primary Economic Terms. We have concerns that prescribing data elements
to report may adversely affect the use and development of open industry standards,
such as FpML, in the transmission of trade information to SDRs. We would

14
Guidelines are included under “Line Item Instructions for Derivatives and Off-Balance-Sheet Items Schedule HC-L” in
the Board of Governors of the Federal Reserve System’s “Instructions for Preparation of Consolidated Financial
Statements for Bank Holding Companies Reporting Form FR Y–9C”.
 16

additionally point out that prescribing data elements to report could result in product
types that come to market in the future not being adequately described by the data
elements prescribed. As a result, and in keeping with our suggestion of harmonization
of rules between the Commission and the SEC, we strongly urge the Commission to
follow the same approach as the SEC by stating reporting requirements in broad-based
generic terms. In the event that the Commission determines that a precise definition of
data elements to report is required we would suggest that this information is put
together by asset class with input from industry groups. At this time we would provide
the following examples of items to address should the Commission wish to precisely
define data elements to report:

(A) The “Start date” data element specified as part of the “Data Fields and Suggested
Form and Order for Real-time Public Reporting of Swap Transaction and Pricing
Data” is ambiguous, and we recommend that the Commission instead require
participants to report the “Effective Date”, consistently with the SEC Proposed
Regulation.

(B) In relation to the Commission’s request for comment on whether additions

should be made to the proposed list of Minimum Primary Economic Terms Data
based on the Master Reference Generic Data Fields List, we make two specific
suggestions for fields that should not be added:

• The Order Entry Timestamp, because the OTC market is characterized by a

multiplicity of client channels, a number of which do not translate into
formal orders. Request for quotes would be an example of such a channel.

• The Parent Counterparty, because this would be more effectively determined

centrally by either the SDR or regulators, as a result of the UCI initiative.

(C) The proposals for Primary Economic Terms including the data elements
necessary for a person to determine the market value of the transaction would
not be achievable for some complex trades. In order for a third party to value
these transactions access would be required to proprietary market data and
pricing models that would not be in the public domain. We request that the
Commission consider how any requirement to provide valuation information
interact with requirements under proposed “Business Conduct Standards for
Swap Dealers and Major Swap Participants With Counterparties”15 and proposed
“Swap Trading Relationship Documentation Requirements for Swap Dealers and
Major Swap Participants” currently under consideration by the Commission.
Please also see the general comments regarding valuation data in Section
IV(b)(i) below.

(D) Reference is made to reporting of Primary Economic Terms, Affirmation and

Confirmation at different times, which would lead to duplicative reporting. It
would be preferable to have one set of required data elements, including

15
75 Fed. Reg. 80638 (Dec. 22, 2010)
 17

confirmation status submitted after the trade is confirmed. Any updates may be
reported if there are changes to the data set.

(E) The requirement in the Swap Reporting NPR that “[a]ny other primary economic
term(s) of the swap matched by the counterparties in verifying the swap” be
reported is unclear. We request that the Commission either clarify what these
data elements are intended to capture, or adopt an approach similar to the SEC
by requiring more generically the data elements necessary to determine the
market value of the transaction. We recommend the latter.

As noted in Section II, above, under “Phase-in Implementation”, compliance with the reporting
requirements under consideration will require development of substantial technology infrastructure
across a diverse range of asset classes. We therefore encourage the Commission to consider
existing confirmation models and their requirements regarding economic fields that should be
matched to confirm a transaction. Confirmation data can be relayed by derivatives clearing
organizations (“DCOs”), swap execution facilities (“SEFs”) and middleware providers (including
unregulated platforms). To promote successful implementation of the reporting regime, we strongly
believe the Commission should leverage and build upon investments made within the industry over
recent years. Specifically, the Commission should seek to pursue solutions based upon the benefits
seen in existing trade repositories such as the Credit Derivatives Trade Information Warehouse,
specifically that:

• leverage bi-laterally matched legally binding (“gold”) records,

• handle most if not all lifecycle events,

• provide all participants with access to key operations controls and efficiencies such as
central settlement, credit event, re-organization and rename processing, and

• provide regulatory access to key market and industry data.

(b) Total return swap transactions

There should be a general exemption from public dissemination of data with respect to TRSs and
trades otherwise designed to offer risks and returns proportional to a position in the security,
securities or loan(s) on which the TRS is based. TRS pricing information is of no value to the
market because it is driven by many considerations including the funding levels of the
counterparties to the TRS and therefore may not provide information about the underlying asset for
the TRS.

(c) Inter-affiliate transactions

Information relating to transactions undertaken within an organization to manage risk within the
organization should not be publicly disseminated. For example, if a counterparty chooses to enter
into a swap with a particular entity within an organization, such as a U.S. subsidiary, although the
non-U.S. parent of the organization group is in a better economic position to incur the counterparty
exposure from a risk management standpoint, the inter-affiliate transaction entered into between
 18

the inter-company entities (not with the counterparty) does not contain any additional price
information beyond that contained in the transaction with the customer. As a result, we recommend
that such inter-affiliate transactions be excluded from the scope of public real time reporting for
price discovery purposes.

(d) Multi-asset swaps

We recommend that multi-asset swaps be reported as one trade only, to one specific SDR, and not
be decomposed among their underlying asset class constituents for reporting purposes. The SDR
designation could be determined by the reporting counterparty as the most significant asset class
component (in practice, it will most often be the asset class of the desk that trades the swap). A
specific indicator could be associated with the trade for such purpose.

IV. Reporting of Collateral Information

In Section IV(b) below, we offer specific comments on aspects of the Swap Reporting NPR
relating to the reporting of collateral and valuation information. Before addressing these specific
points, we would stress a few general points of clarification regarding collateralization in the OTC
derivative market, distinguishing between uncleared and cleared transactions, as set out in Section
IV(a) below.

(a) General comments regarding collateral in uncleared and cleared transactions.

In relation to uncleared transactions, the following points are critical to defining correctly the set of
data fields in order to achieve the Commission's objectives for reporting and transparency. The
Commission may find it helpful to refer to two documents that were published in 2010: the Market
Review of OTC Derivative Bilateral Collateralization Practices published by ISDA (March 1,
2010) 16, which provides an overview of the bilateral collateralization process and explains the use
of collateral as a credit risk mitigant and the Independent Amounts white paper published by ISDA,
SIFMA and the Managed Funds Association (March 1, 2010) 17, which describes the usage and
purpose of Independent Amount (“IA”) together with some of the risks and challenges associated
with IA segregation.

Bilateral collateralization in the uncleared OTC derivatives market has several key distinguishing
features that are materially different from margin arrangements relating to futures, options and
securities transactions. For example:

• Collateral flows in both directions between the counterparties, according to the exposure
that each has to the other at different times

16
The full Market Review can be found on ISDA’s website: http://isda.org/c_and_a/pdf/Collateral-Market-Review.pdf
17
The full IA White Paper can be found on ISDA’s website: http://isda.org/c_and_a/pdf/Independent-Amount-
WhitePaper-Final.pdf
 19

• The total collateral requirement comprises two elements, exposure collateral (“EC”) 18,
which is present in all standard agreements and IA, which is optional according to bilateral
negotiation. These two elements are netted to produce the total collateral requirement.

• EC is required to cover the net estimated mark-to-market value of the portfolio of

transactions between two parties at the time of a collateral call.19 Importantly, the
calculation of required collateral is performed at a netted portfolio level and collateral
cannot be attributed at the transaction level - it is simply not possible to identify the
specific exposure collateral associated with any particular transaction.

• IA is an optional additional amount of collateral that two counterparties may negotiate. In

ISDA’s responses to other Notices of Proposed Rulemaking issued by the Commission,
Independent Amount has been defined as “money, securities, or property posted by a party
to secure its obligations pursuant to the terms of a swap agreement and either (i) specified
as an “Independent Amount” in the relevant agreement of the parties or (ii) the amount of
which is calculated based upon terms agreed between the parties (in addition to and
separately from any Exposure Collateral requirement)” 20. Its purpose is to protect the IA
holder against adverse movement in the net mark-to-market value of the portfolio that
occurs before additional EC can be obtained to cover that exposure. The calculation of IA
generally takes into account the estimated period it would take to unwind trades and/or
portfolios along with the volatility of the positions in a portfolio. IA is roughly analogous
to Initial Margin in the existing futures and options markets. Within a single collateral
agreement, IA may apply to all, some, or none of the transactions in the portfolio. If IA
does apply to a particular collateral agreement, it may be specified at transaction level, at
portfolio level, at some intermediate level (a combination of product type, currency and
maturity, for instance), and possibly a hybrid of all three. Therefore it may or may not be
possible to identify the IA associated with a particular transaction, but as a general matter
this association cannot be reliably made.

• Generally, swap counterparties do not employ transaction-level collateral arrangements;

instead collateral arrangements are managed and processed at the portfolio level. In rare
cases of a single transaction with a specific collateral arrangement, this can be considered
to be a special case of portfolio collateralization, but for a portfolio of one trade.

18
Exposure Collateral and Variation Margin are not defined terms in the bilateral space, however, to avoid confusion
with the term “Variation Margin” in the cleared space which is used very differently, the Associations strongly believe
that we need a different term for the uncleared derivatives market. ISDA has offered the following definition in several
other response letters: “Exposure Collateral” means money, securities, or property posted by a party to secure its
obligations pursuant to the terms of a swap agreement, the amount of which is based on an estimate of the net mark-to-
market exposure of all transactions under the master swap agreement.
19
Specifically, the estimate (typically at mid-market) is of the amounts that would be payable between the parties if the
transaction(s) were terminated, and is typically referred to as the “Exposure” of the party that would be entitled to receive
a payment in the event of an early termination.
20
In light of these proposed changes to the definitions of Initial Margin and Variation Margin, we would propose a
corresponding replacement of the definition of “Margin” with the following: “Uncleared Swap Collateral” means both
Exposure Collateral and Independent Amount.
 20

The implications of the points above for uncleared transactions should be considered in developing
solutions to achieve the Commission’s objectives for reporting and transparency, as discussed
further below in our comments on specific provisions in the Swap Reporting NPR.

In relation to cleared transactions, the situation is substantially simpler. We suggest that the most
pragmatic solution to creating transparency of valuation and collateral for cleared derivatives
would be for DCOs to report the transactions/positions (as appropriate) and collateral and
valuations on a portfolio level. We also suggest that in particular, DCOs’ values should be used for
all cleared transactions. Because of its clarity, we would recommend this approach be adopted by
the Commission.

(b) Specific comments in response to the Swap Reporting NPR

The remaining comments in relation to collateral information relate to specific provisions in the
Swap Reporting NPR, which are quoted, together with relevant footnotes from the Swap Reporting
NPR, in italics below:

(i) Valuation Data Reporting for all Swaps in All Swap Asset Classes

Valuation data is defined in the proposed regulations to mean all of the data elements
necessary for a person to determine the current market value of a swap, including, without
limitation, daily margin, daily mark-to-market, and other measures of valuation to be
determined by the Commission prior to promulgation of its final swap data reporting
regulations. Swap valuation data is essential to a variety of the regulatory functions of
many financial regulators, and is crucial to fulfillment of fundamental purposes of Dodd-
Frank, including systemic risk reduction and increased transparency of the derivatives
marketplace to regulators. The Commission and other regulators would use valuation
information regarding swaps reported to SDRs for prudential oversight, to monitor
potential systemic risk, and to monitor compliance with regulatory requirements for SDs
and MSPs. The importance of reporting swap valuation data to SDRs is recognized
internationally.-- Swap Reporting NPR (75 Fed. Reg.76574 at 76584)

The Associations fully support the need for the supervisory and regulatory community to
have access to valuation data. The largest SDs in the U.S. market are all prudentially
regulated. In order to start collecting this information, the Commission should work closely
with their prudential regulators, the Federal Reserve Banks and the Office of the
Comptroller of the Currency.

Longer term, the Associations recommend the creation of a “Counterparty Exposure

Repository” as described below. We propose that this requirement would be in place of
requiring valuation data on a transaction or swap level, which as discussed earlier is not
possible due to the portfolio nature of collateral. Additionally, the computation of the
mark-to-market value for a derivative position is quite complex for all derivatives. To
perform this calculation one must be in possession of all the economic terms of a
transaction (which may be over 100 data fields per trade) and all of the lifecycle structure
parameters of a transaction, to permit the cash flow structure of the transaction to be
constructed and the appropriate discounting and estimation of future value to take place. To
 21

perform these latter calculations, it is necessary to maintain a full set of current market data
parameters, forward rates and the history of such market data. It is a computationally
intensive and technically difficult task for each firm to compute the valuation each day for
each transaction - all firms have invested significantly in the technology and staff to
undertake this daily valuation, and even so it is not straightforward. Rather than the
Commission duplicating these measures, we suggest that a data feed of general transaction
data, plus the submitting firm's computed valuation, should be sufficient for market
surveillance use, and will avoid the Commission incurring large expense in replicating
existing computations. We note also that the prudential regulators of the SDs have the
power (and frequently exercise the power) to review firms' internal valuation models,
which would provide assurance that the valuation results being provided to the
Commissions are sound, and would permit more in-depth analysis of valuation methods
and parameters if necessary.

Referring specifically to the text quoted above, as a technical matter, we note that “daily
margin” is not a defined term in the bilateral OTC market. We remind the Commission that
margin under Credit Support Annexes (“CSAs”) is typically not collected until the day
following notice requesting such margin and that the routines that run the valuation of the
portfolio are customarily run overnight. Accordingly, valuation data for uncleared swaps
provided for end of day reporting to SDRs or other locations will not be “same day” but
will refer to portfolio valuation on the close of the preceding day. Cleared swaps valuation
may be reportable at the end of the trading day, but this will largely depend on the
capabilities of the DCO. In addition, we caution that the implementation of any valuation
methodology requires significant operational and infrastructure development, and we are
therefore concerned that the text quoted above indicates that the Commission’s final swap
data reporting regulations may require further measures of valuation without further
consultation.

(ii) Capturing Counterparty Risk Concentrations

“TRs should collect data to enable monitoring of gross and net counterparty exposures,
wherever possible, not only on notional volumes for each contract but also market values,
exposures before collateral, and exposure value net of collateral with a full counterparty
breakdown. This would allow for the calculation of measures that capture counterparty
risk concentrations both for individual risk categories as well as for the overall market .--
CFTC: NPR Reporting and Recordkeeping”52 -- Swap Reporting NPR (75 Fed. Reg.76574
at 76584)
52
FSB, Implementing OTC Derivatives Market Reforms: Report of the OTC Derivatives
Working Group, October 20, 2010, at 48.

As noted above, the Associations support transparency to the regulators and supervisors,
but only at the portfolio level in the aforementioned “Counterparty Exposure Repository”.
We do not believe that this information should be provided on a transactional level. ISDA
had proposed much the same idea to the OTC Derivatives Supervisor's Group in July 2010
and again to the Federal Reserve Board in August 2010. We set out below the data
elements that would be required to effect such monitoring. At the outset, we caution that an
immense amount of data exists in firms’ internal systems, much of which is not helpful in
providing the kind of risk concentration monitoring specified by the Commission. Having
 22

had extensive experience with the large and expensive technology platforms that firms
must use to manage all aspects of the collateralized portfolios, we would urge the
Commission to follow the doctrine of “less is more”, meaning that by careful selection of a
smaller number of key data elements the burden imposed on the Commission and market
participants will be smaller and less costly to manage, and yet the essential data to permit
proper market oversight can be obtained. We believe that an optimal balance can be struck.
In this regard, firms could submit for each portfolio the following information to the
repository:

(A) Current net portfolio mark-to-market value in US$ of the collateralized

portfolio 21, from the perspective of the reporting party;

(B) Current net portfolio mark-to-market value in US$ of the total portfolio
covered by the ISDA agreement 22, from the perspective of the reporting
party;

(C) Currently applicable unsecured Threshold, both for the Reporting Party
and for the Reporting Party’s Counterparty;

(D) Currently applicable Independent Amount 23, both for the Reporting Party
and for the Reporting Party’s Counterparty;

(E) Currently applicable Minimum Transfer Amount, both for the Reporting
Party and for the Reporting Party’s Counterparty; and

(F) Currently held collateral balance, excluding collateral in transit, collateral

to be itemized (e.g. USD cash, US treasuries, letters of credit, etc. by ISO
CCY code and/or CUSIP/ISIN) and, for each of the collateral held by
Reporting Party and held by Reporting Party’s Counterparty, broken down
into (I) USD equivalent of cash collateral held before haircuts are applied,
(II) USD equivalent of cash collateral held after haircuts are applied 24, (III)
USD equivalent of securities collateral held before haircuts are applied,
(IV) USD equivalent of securities collateral held after haircuts are applied,
(V) the USD equivalent total before haircuts are applied which would be
the sum of items (I) and (III) above, and (VI) the USD equivalent total
after haircuts are applied which would be the sum of (II) and (IV) above.

Instead of sending all of the necessary pieces of information required to calculate

counterparty exposure, we are proposing that the macro risk picture between a firm and a
counterparty can be represented as a single record (containing a few critical data points).
This would typically be computed as of the previous day’s close.

21
This would only include the mark to market of the portfolio covered by the CSA.
22
Frequently, certain trades are excluded from collateralization under a CSA, but would be included in the exposure in
the event of default.
23
Independent Amount can be defined at the level of the portfolio of transactions between two parties, or can be defined
uniquely for each individual transaction but for purposes of the proposed repository, the aggregate Independent Amount
is important.
24
Although not common in practice, a haircut may also be applied to cash in currencies other than the “Base Currency”
to protect against adverse movements in exchange rates.
 23

This suggested list of information above would provide visibility of net mark-to-market
exposures, thresholds and the offsetting collateral, which together may be a valuable
addition to the body of data available to regulators. The Associations concur that, just as
this information is valuable for each firm to effectively risk-manage its portfolio of
counterparty risk, this information may have utility for regulators to review this risk
management by firms, and potentially to model the systemic effects of unsecured risks (if
any) to which firms may be exposed.

The data could help regulators to better understand the interconnectedness of systemically
important firms by virtue of their exposures to one another and any offsetting collateral.
Greater visibility of this data would potentially allow faster and more fact-based
understanding of emerging credit stress events, and allow an assessment of whether the
demise of certain market participants might or might not have material adverse impact on
other market participants - in other words, providing a qualitative assessment of “too big to
fail”. This could be very useful input to promote informed public sector decision making
around times of market stress, and thus enhance market stability.

The Associations believe that this data should be reported by all SDs and Major Swap
Participants (“MSPs”) for their entire portfolios. This will allow the Commission to see if
there are any material valuation differences between systemically important counterparties.
Further, the Associations believe that only non-cleared portfolios should be shared this
way. DCOs are best positioned to provide any required reporting on the exposures between
the DCO and clearing members.

We further note that the calculations used by prudentially regulated institutions are already
subject to significant oversight as they form the basis of the inputs to the Basel II
calculations. Therefore, the Commission should not be concerned with being able to
recreate the calculations but should be comfortable relying on the calculations provided.

(iii) Minimum Valuation Data

Independent amount; Independent amount currency; Independent amount payer;

Independent amount receiver; Initial margin; Variation margin; Mark-to-market; Non-
cash collateral; Non-cash collateral valuation.

-- Swap Reporting NPR (75 Fed. Reg.76574 at 76607)

The Commission lists sample fields which should be required for “valuation data”.
However, in light of our comments made above about the portfolio versus transaction level
at which certain data elements exist, we respectfully suggest that this list be revised. The
two types of data elements cannot be inter-mixed, and as noted earlier we strongly
encourage that the Commission obtain valuation data primarily at the portfolio level, or at
least at the summary transaction level, rather than obtaining the lowest level data elements
and attempting to replicate the trade valuation and portfolio netting calculations performed
by firms. We propose that the data elements listed (A) through (F) under our response for
Capturing Counterparty Risk Concentrations would meet the objectives of the Commission
in the most efficient and streamlined manner.
 24

(iv) Collateral Warehouse System

Should a separate collateral warehouse system be established as part of an SDR to enable

systemic risk and prudential regulators to monitor collateral management and gross
exposure on a portfolio level for swap participants? How should this be done?
-- Swap Reporting NPR (75 Fed. Reg.76574 at 76586)

The term “Collateral Warehouse” has been used in many discussions with supervisors and
regulators over the past few months. If by “Collateral Warehouse”, the Commission is
asking for a central entity to maintain all collateral and/or provide record keeping for that
collateral, we do not believe that would provide meaningful results for the purposes
described, however, as stated above, we do believe that a “Counterparty Exposure
Repository” should be created to provide for surveillance of portfolio level risk.

We do believe that it would be useful for regulators to have oversight of the credit risk that
exists at a portfolio level between pairs of counterparties, particularly those that may be
considered to be systemically important. These same methods could be used for all
counterparties, but the cost effectiveness of the data collation and analysis effort is
significantly reduced when including pairs of counterparties where the risk involved is
relatively small and thus non-systemic.

A “Counterparty Exposure Repository” could be created to contain the net mark-to-market

exposure for each counterparty portfolio and the corresponding collateral. The data must be
maintained at a portfolio level in the repository, and firms currently maintain this
information for internal risk management purposes, therefore, on a relative basis, this
should be straightforward to implement. We do not believe that having access to current
exposure information on a trade by trade basis would support supervisors in monitoring
and managing systemic risk since the risk is governed by portfolio based Master
Agreements that provide for netting and credit support. To have a true picture of the risks a
party is running, one needs to look at the exposure as if one of the counterparties had
defaulted under the terms reflected in appropriate legal agreements and supported by legal
opinions.

We would propose that the Counterparty Exposure Repository would receive and house the
list of data elements listed (A) through (F) under our response for Capturing Counterparty
Risk Concentrations. The Counterparty Exposure Repository should operate on a non-
profit basis and be subject to similar requirements as swap data repositories. We also
suggest that registered swap data repositories should make accessible, at a reasonable fee,
to the Counterparty Exposure Repository the primary economic terms and valuation data of
any swap and security-based swap transaction reported to the swap data repository. This
would provide the information that we believe the systemic risk and prudential regulators
need to perform their respective oversight functions.

(v) Master Agreement Library

Should a separate master agreement library system be established as part of an SDR? How
should this be done?-- Swap Reporting NPR (75 Fed. Reg.76574 at 76586)
 25

Master agreements are a crucial risk mitigation technique because they provide the
contractual basis for netting between a pair of counterparties. However, they are negotiated
and then rarely amended, sometimes being used for decades without substantial change.
Therefore, due to the low velocity of change in master agreements there is no particular
value in dynamic monitoring by regulators of changes in the overall pool of master
agreement used by market participants. Firms have generally invested heavily in
technology to track master agreements and other documentation, including images of those
contracts. To the extent that the Commission needs to examine particular master
agreements, the parties to any contract would be able to furnish them readily. It should also
be noted that the master agreement document does not contain the entirety of the
agreement between counterparties - typically there may be a CSA, side letters, amendment
agreements and operative industry protocols that need to be read with the master agreement
in order to understand the entire contractual basis of the relationship. For example, the
CSA contains terms which establish the collateral relationship between two parties.
Variables including the Independent Amount, Threshold, Minimum Transfer Amount, and
Eligible Collateral are all examples of data that is maintained on proprietary collateral
systems that is required for calculating margin calls and can be provided to regulators upon
request. There are also a wide range of product-specific definitions and of course the
individual trade confirmations that are also relevant. Therefore, a centralized effort to
capture documentation would need to be much wider than the master agreement, would be
duplicative of existing industry investments, would not provide regulators with particularly
meaningful data given the slow rate of change in these documents, and would not provide
any information above and beyond that which could already be readily obtained from
regulated firms.

It should also be noted that each prudentially regulated participant in the swap market
already provides access to its legal agreements and any stress testing that is performed on
its portfolio based on triggers the agreements may contain. The Associations strongly
recommend that the Commission work closely with the prudential regulators to gain access
to this information as needed.

Because of the complexity and the availability of the information through alternative
means, we recommend that the Commission does not establish a master agreement library
at this time. After all SDRs required to facilitate swap data reporting are established, then
the Commission and the OTC Derivatives Supervisors Group could engage the industry to
decide if such an investment is warranted.

V. Reporting Responsibilities

(a) Reporting responsibilities

We believe that the Real-Time Reporting NPR captures the relevant parties in the derivatives
markets that should be the reporting party for a swap. However, the Associations consider a
requirement from the Commission that one or more entities other than a swap counterparty, such as
a registered SEF, a national securities exchange, a DCO, or a broker, report swaps to be
unnecessary in light of the likely prevalence of competition to provide reporting services and given
the ability of market participants to contract with the appropriate vendors to achieve the most
 26

efficient allocation of reporting responsibilities. We therefore recommend that, similarly as in the

approach adopted in the SEC Proposed Regulation, the counterparties to the swap should primarily
be responsible for reporting but that third-party agents may report on their behalf, thereby enabling
the market to allocate reporting responsibilities in the most efficient manner. The Commission
should consider whether in certain markets a third-party service provider may assume the legal
responsibility for reporting the transaction to the SDR. If so, it may be helpful to incorporate this
construct into the rulemaking.

We are strongly supportive of allowing third party facilitation of swap data reporting for the
reasons noted in the Swap Reporting NPR. While it is difficult to anticipate the market structure
that may develop in this area pending the promulgation of the final rule requirements, SEFs,
exchanges, DCOs, brokers, and stand-alone data reporting vendors are all potential providers of
this service, either across asset classes or for particular products or transaction states (e.g., with
respect to cleared trades). Consideration should also be given as to whether a particular entity such
as a SEF, DCO or SDR will hold the authoritative record of a trade and whether that information
should be leveraged for reporting purposes. 25 In regards to technology, the industry should look to
use standardized data interchange formats.

We agree with the Commission’s proposal that swap markets satisfy their public dissemination
requirement by either sending to a registered SDR that accepts and disseminates swap transaction
and pricing data or by publicly disseminating through a third-party service provider, provided that
the SDR is an independent, third party and that the information is available within a reasonable
time frame. We do not think that there should be other means of reporting because this will fracture
the overall market picture and reduce the utility of reported information.

We do not see an obvious benefit from requiring both counterparties to a swap to report swap data.
In fact, this may be inefficient and lead to duplicative efforts and operation costs. Bilateral consent
to confirmation data is sufficient to ensure the accuracy of such data. Regarding whether selection
of the reporting counterparty should be the same for cleared swaps as for non-cleared swaps, we
believe that the answer is dependent on the evolution of the cleared swap workflow and the
introduction of SEFs into the market. The venue where the matching takes place should be the
reporting party for this purpose. If the counterparties have the same hierarchical status, then there
are already market conventions which will suggest where the responsibility for reporting should lie.
For example, in the dealer to dealer CDS market, the seller of protection is responsible for
confirming the trade. The Commission should adopt these market standards where possible.

We agree that the distinction between the two categories of counterparty (SD or MSP, versus non-
SD/MSP) is appropriate and fully consistent with the statute. Given the hub and spoke nature of the
derivatives markets, a SD will likely have numerous counterparties while a MSP will have
relatively few. Consequently, the quantity of data available from SDs will be greater and will assist
regulators in developing a full picture of the market.

In the case of an end-user claiming an exemption from the clearing requirement for a swap, we
request that the Commission clarify the nature of regulation and enforcement to ensure the
accuracy of the claim.

25
It should be noted that the authoritative record may transfer between entities at certain points during the life of a trade,
for example the authoritative record of a trade executed on a SEF and then cleared would initially reside at the SEF and
then move to the DCO.
 27

(b) Extraterritoriality

The Associations strongly urge the Commission to base its rulemakings on certain core principles
related to extraterritorial scope and international comity. We believe these core principles should
be as follows:

• Section 752 of the Dodd-Frank Act requires the Commission “consult and coordinate with
foreign regulatory authorities on the establishment of consistent international standards
with respect to the regulation...of swaps...[and] swap entities...”. Accordingly, the
Commission should consult with foreign regulators before establishing the extra-territorial
scope of the rules promulgated under Title VII and, when appropriate, should defer to
substantially similar foreign regulation that serve similar policy interests to those of Title
VII. Many of the provisions of Title VII and the European Market Infrastructure
Regulation (“EMIR”), for instance, are conceptually similar but different in specific
implementation. Because market participants will have significant issues complying with
both sets of regulations if applied to the same transactions, we urge the Commission to
seek international harmonization in derivatives regulation through memoranda of
understanding or other international cooperative measures. We are concerned that without
such international outreach there could be regulatory chaos as different regulators compete
to regulate overlapping parts of the global derivatives business.

• The Commission should seek to avoid the regulatory uncertainty and ambiguity (and
potential room for regulatory arbitrage) and additional expense that will ensue if market
participants are required to comply with inconsistent or redundant regulations. This is
particularly true where, as in the case of trade reporting, complex, novel, and expensive
information technology and operational systems must be developed over extended time
periods. In addition, resolving potential regulatory uncertainty and ambiguity between
foreign and U.S. regulation will facilitate the continued provision of capital, liquidity and
risk management solutions to U.S. corporations and institutional investors by foreign SDs,
thereby reducing the concentration of risk and enhancing the strength of the U.S. capital
markets.

• Jurisdictional boundaries are essential to implementation of Title VII. The Commission

should define the universe of transactions that they seek to regulate in a clear and
unambiguous manner so that the industry can implement the significant systems and
operational changes necessary to give effect to the regulations by the relevant effective
dates. The jurisdictional boundaries should also be tailored to promote and effectuate the
public policy objectives underlying the specific rule under consideration. To this end, the
Commission should craft differing jurisdictional boundaries that reflect the policy
objectives of the rule in question as opposed to crafting a “one-size-fits-all” framework.
This approach will also help the Commission more precisely harmonize Title VII with
parallel international regulation. Ultimately, this will allow the Commission to manage
 28

their scarce resources without sacrificing the important public policy considerations behind
Title VII.

Applying these core principles to the proposed reporting and recordkeeping requirements, we urge
the Commission to work to reduce duplicative reporting, recordkeeping and other requirements in
overlapping regulations. Avoiding overlap is important with respect to reporting, particularly if the
overlapping data cannot be easily reconciled. For example, EMIR will also require reporting of
OTC derivative transactions likely resulting in some swaps being reported more than once unless
the Commission works with its foreign counterparts. Absent international coordination to reduce
redundant reporting or, where unavoidable, establish standard data so that redundant records can be
easily reconciled, overlapping and inconsistent reporting regimes may serve to obfuscate rather
than clarify the true nature and size of the global swap markets for international regulators. Instead
of implementing swap reporting rules unilaterally, we request the Commission work with global
regulators to devise systems that efficiently operate together to which such global regulators have
access to data relevant to the performance of their responsibilities.

We do not believe that the Commission should require reporting of transactions between two non-
U.S. counterparties, nor is it clear that the Commission has the authority to do so. With respect to a
transaction between two non-U.S. persons that is cleared through a DCO having its principal place
of business in the U.S., the real time public reporting requirement should not apply to either of the
two non-U.S. persons, although the DCO can provide information for regulatory purposes. In
addition, we believe that the Commission should reach international agreements with other
regulators before requiring that all transactions with any U.S. person (even if entered into outside
the U.S. or cleared with a foreign DCO) be reported under Title VII for the reasons discussed
above.

The Swap Reporting NPR requires that all transactions between a non-U.S. person and a U.S.
person must be reported by the U.S. person, even if the non-U.S. person is a foreign SD. Given that
end-users are unlikely to have the internal systems and processes necessary to support this
reporting, we are concerned that the practical result would be an inadvertent exclusion of foreign
SDs from the U.S. market, which could decrease liquidity, further concentrate the U.S. swap
market and thereby increase systemic risk. Accordingly, we urge the Commission to reconsider this
provision and follow the general precepts that SDs, even foreign SDs, are responsible for reporting
transactions with non-SDs.

As noted in section II (General Considerations), we strongly recommend the establishment of

SDRs that operate globally and encourage the Commission to take the same stance. An access
model for regulatory authorities across jurisdictions will need to be established around global
SDRs. The Commission would then have access to trades reported under overseas regulation
without requiring reporting from non-U.S. entities and end-users as described above. With a global
SDR, imposing Dodd-Frank reporting requirements in the scenarios described above of a trade
between two non-U.S. counterparties being cleared through a U.S. DCO and a trade between a U.S.
end-user and a non-U.S. SD reported under the reporting requirements of overseas regulatory
authorities would be redundant. We respectfully request that this subject is included in consultation
with foreign regulatory authorities.

Lastly, we would ask the Commission to consider carefully and provide for consistency with,
foreign privacy laws, some of which carry criminal penalties for wrongful disclosure of
 29

information. Alternatively, and at the very least, the requirements should be made subject to any
such mandatory restrictions on disclosure binding on the relevant parties. Failure to do so would
create potentially insurmountable challenges, both for foreign SDs who wish to participate in the
U.S. swaps market, with concomitant decreases in liquidity and concentration of risk in the U.S.
capital markets, and also for U.S. SDs who have entered into a transaction with a non-U.S.
counterparty who is protected under such privacy laws.

VI. Commodities

In this section, we set out responses to the Commission’s questions relating specifically to swaps
on commodities; however first we consider two more general points in relation to the commodities
asset class. First, financial and physically settled commodity transactions are fungible from a risk
management perspective and we therefore believe that they should be treated in a similar manner.
In this regard, please see the concerns that we identify with respect to the reporting of physically
settled transactions below. Second, many participants that use commodities derivatives extensively,
and thus, will be classified as SDs or MSP, do not have systems for quick reporting. This is driven
by them being physical players and bona fide hedgers but still with large net derivative portfolios.
Please see also the general comments regarding the need for phase-in requirements in Section II(e)
above.

Should the asset class for other commodity be divided further (e.g., agricultural commodity, energy
commodity, etc.)? If so, how should it be divided?

As discussed below, subdividing the asset class for “other commodity” would be advisable for
certain aspects of the data proposals. While it may not be necessary for certain of the reporting
requirements, it may be necessary for some other aspects of rulemaking. For example, “block
trade” sizes should be tailored for the underlier. The commodities asset class is heterogeneous and
establishing a “block trade” size across all constituents would be problematic.

Would this rounding convention be appropriate for all swaps? For example, would this apply to
swaps with an underlying asset that is a physical commodity with a specific delivery point? If not,
why and what additional rounding convention may be needed?

Commodities swap notional amounts should be reported by units of measure (mmbtus for gas, MW
for power, etc.) rather than in dollar amounts as proposed by the CFTC. An example of this in
practice is that exchanges set block trades and establish contracts by reference to units of measure
and not dollar amounts.

Also, the size of commodity trades is typically smaller that interest rate trades. Therefore, subject
to the general points regarding block trade size threshold made in Section I above, the sizing of the
relative threshold/reporting amounts should be lower (i.e., large notional trades for commodities
would be smaller than the $250,000,000 threshold set for other asset classes) and block trade sizes
should be established by reference to market size of the relevant commodity. Once again, amounts
would be converted to volumes.

By way of example, NYMEX allows block trades in one of its most liquid futures contracts (Henry
Hub natural gas) for trades of 100 futures contracts. Based on prices towards the end of December,
the value of this trade would be $4,260,000 (100 contracts x 10,000Mmbtus x $4.26).
 30

How should the Commission determine an appropriate time delay for large notional swaps?

Certain commodities markets are different to markets in equity, credit, currency and interest rate
derivatives. Commodity markets are generally smaller, less liquid and therefore less anonymous
than those other markets. Any data disclosures to the market are therefore likely to have a more
pronounced effect on that market. It is likely that such data disclosures will be easier to reverse
engineer, meaning that it is possible, if not likely, that market participants will not be able to act
anonymously, since many commodities are so regionally specific as to be associated solely with
producers in that locality. This will make them less willing to assume or warehouse market risk and
less willing to quote market based prices for transactions, thereby negatively impacting liquidity,
price discovery and, most importantly, access to these markets.

Other accommodations may be required for bespoke transactions and material non-public
information (“MNPI”). The transparency benefit of bespoke transaction public disclosure to the
broader market is questionable specifically because these transactions have no comparables.
However the cost to participants engaging in these transactions if anonymity is not completely
assured is potentially high. The Real-time Reporting NPR does not include detail about how to
accommodate MNPI associated with financings including hedges for VPPs (volumetric production
payments) and CDS associated with a financing, among others. At a minimum, with respect to
swaps with listed entities, i.e. entities which are subject to securities-based disclosure standards, the
Commission should ensure that its disclosure standards are consistent with MNPI disclosure
standards, such that a party’s obligations to report real time swap transaction information to the
market are consistent with any obligations that party may have to announce the deal to which the
transaction relates to the market pursuant to MNPI reporting requirements.

In the same way as there is a bona fide hedge exemption for Position Limits, there should be
special treatment for hedging real physical risk (whether this is done physically or financially) in
the large notional swap space.

We believe very extended reporting timelines are often appropriate in these cases and in some
cases there should be no requirement to report publicly at all.

With regard to the time delay for large notional swaps in the other commodity asset class, the
Commission recognizes a longer time delay may be necessary due to the hedging strategies that
are associated with such swaps. What time delay would be appropriate for swaps in the other
commodity asset class and why?

For the reasons outlined above, any rules should draw certain distinctions and afford adequate time
for hedge activity to occur throughout the day, as dictated by the particular parameters of a
transaction.

VII. Recordkeeping and daily trading records requirements

The Associations are supportive of the goal stated in § 23.202 requiring SDs and MSPs to ensure
that they preserve all information necessary to conduct a comprehensive and accurate trade
reconstruction for each swap. We recommend however that the implementation approach for
determining whether it is more appropriate to maintain each transaction record as a separate
 31

electronic file should be left to the respective reporting counterparties. SDs and MSPs have
invested considerable time and effort in development of systems to store trade data in an efficient
manner. This routinely involves storing data across a number of systems. For example, a risk
management system would not normally store records of oral and written communication that leads
to the execution of a swap but would hold the terms of the swap. Aggregating transaction data from
all systems into a single electronic file will require enormous investment across market participants
and will require a substantial implementation period.

The requirement to maintain records of all communications “conveyed by telephone, voicemail,

facsimile, instant messaging, chat rooms, electronic mail, mobile device, or other digital or
electronic media” extends the remit of record keeping obligations to include voice communications
whether conveyed by telephone, voicemail or mobile device. We recommend that further analysis
and consultation is performed on the costs and benefits of holding these records for the life of a
swap plus five years, as we believe this new requirement could impose a heavy cost burden to
implement and maintain, for only a small incremental benefit. We would be more supportive of a
voice recording obligation aligned to the rules of the UK Financial Services Authority, which are to
retain recordings for a minimum period of six months.

The Associations request further clarification on the requirement in the Swap Reporting NPR that
SDs and MSPs have records “readily accessible via real time electronic access by the registrant
throughout the life of the swap, and for two years following the final termination of the swap” 26.
Current record keeping practice is for records to be readily accessible, which normally means
within a reasonable period of time, such as up to two working days. We are concerned that the
phrase “real time” implies that records must be instantly accessible, which is impractical to achieve
given the volume of day to day transactions. We suggest the existing requirement for records to be
“readily accessible” is sufficient. In addition, the Associations do not see any value in retaining
records for a period of ten years from the termination of a swap. This would impose a significant
additional burden in terms of costs and maintenance.

Regarding the requirement for SDs and MSPs to retain information of cash or forward transactions
that are related to swaps, we respectfully point out that the hedging and risk mitigation activities
referred to in the SD/MSP Recordkeeping NPR are typically not executed with respect to specific
trades; rather they are executed against the overall positions of business units such as trading desks.
As such, although records will be retained, it would not be possible to link cash and forward
transactions to a specific swap. The reference to “hedge” also requires clarity to know the extent to
which it comports with existing CEA definitional standards.

Reliance on retention protocols for swaps booked out of entities that are already regulated by a
prudential regulator should be clarified and relied upon. The SD/MSP Recordkeeping NPR states
that “[u]nder sections 4s(f)(1)(B)(i) and (ii), the Commission is authorized to prescribe the books
and records requirements of “all activities related to the business of swap dealers or major swap
participants,” regardless of whether or not the entity has a prudential regulator”. This standard
warrants clarification to avoid duplication. We request that the Commission clarify the extent to
which counterparties may rely upon SDRs to retain records beyond the time periods that
counterparties currently retain such records.

26
§ 45.2(d)(2) of the Swap Reporting NPR.
 32

Many of the retention requirements would be difficult for the relevant parties to meet, as current
industry participants do not typically capture all this data. This data is not required to be captured
for purposes of the securities or bond markets and so significant additional infrastructure
development would therefore be required before this data could be captured and stored. The
following requirements would be particularly problematic given the current operational capabilities
of market participants:

• Maintain records of all pre-execution oral and written communications provided or

received concerning quotes, solicitations, bids, offers, instructions, trading, and prices, that
lead to the execution of a swap. Where these records are maintained today they are
captured ahead of the execution of a swap and as such they are not linked to a trade. While
it may be possible to search by counterparty, potentially with some investment required, it
would not be possible to search by transaction. The infrastructure to back-populate these
records to link to a transaction should they result in the execution of a swap is not in place
today and the procedural and technical feasibility has not been contemplated nor evaluated.
We strongly recommend that the Commission limit this proposal to describing data
required as part of a trading record without dictating how this should be stored and, in
particular, that the Commission exclude oral communication from the requirement to
enable electronic searchability by transaction.

• Maintain record of the date and time, to the nearest minute, using Coordinated Universal
Time (UTC), by timestamp or other timing device, for each quotation provided to, or
received from, the counterparty prior to execution. Moreover, the value derived by moving
the industry to UTC appears minimal when compared to the costs involved.

• Keep a record of each swap portfolio reconciliation, including the number of portfolio
reconciliation discrepancies and the number of swap valuation disputes (including the
time-to-resolution of each valuation dispute and the age of outstanding valuation disputes,
categorized by transaction and counterparty).

An active and ongoing dialogue between the Commission and the industry is vital to understand the
necessary capabilities of the systems that industry participants will need to design, build, and put
into operation to meet requirements of these rules with respect to management of the data that is
captured, in particular the degree to which retained data will need to be identifiable and searchable.
 33

* * *
We appreciate the ability to provide our comments on the Proposed Rules and look forward to
working with the Commission as you continue the rulemaking process. Please feel free to contact
us or our staff at your convenience with any questions.

Sincerely,

Robert Pickel
Executive Vice Chairman

Kenneth E. Bentsen, Jr.

Executive Vice President
Public Policy and Advocacy
 34

ANNEX 1

The table below maps the comments in the different sections of this letter to related
questions contained in the Real-Time Reporting NPR, Swap Reporting NPR and SD/MSP
Recordkeeping NPR. In the case of each NPR, we refer to the page and column of the
Federal Register version of the NPR in which related questions are located.

Section of Letter Related Questions

I. Block Trades - Appropriate Block Size Real-Time Reporting NPR: Pg. 76152, col. 2; Pg.
Threshold and Public Dissemination Delay 76153, col. 3; Pg. 76154, col. 1; Pg. 76158, col. 3;
Pg. 76165, col. 1; Pg. 76165, col. 2; Pg. 76167, col.
(Please also refer to the Block Trading Study) 1; Pg. 76167, col. 2;

II. General Considerations

(a) Consistency between CFTC and SEC rules Real-Time Reporting NPR: Pg. 76142, col. 1; Pg.
and overseas regulators 76158, col. 3; Swap Reporting NPR: Pg. 76586, col.
3;
(b) Trade allocations Swap Reporting NPR: Pg. 76586, col. 2;

(c) Unique identifiers Swap Reporting NPR: Pg. 76589, col. 1; Pg. 76591,
col. 2; Pg. 76592, col. 1; Pg. 76592, col. 3;

(d) Error reporting Swap Reporting NPR: Pg. 76594, col. 3;

(e) Phase-in implementation Real-Time Reporting NPR: Pg. 76143, col. 1; Pg.
76165, col. 2; Swap Reporting NPR: Pg. 76585, col.
3; Pg. 76586, col. 2; Pg. 76586, col. 3; Pg. 76593,
col. 3; Pg. 76597, col. 3;
(f) Life cycle event data

III. Reporting of Trade Information

(a) Information to report Real-Time Reporting NPR: Pg. 76142, col. 1; Pg.
76146, col. 3; Pg. 76150, col. 2; Pg. 76151, col. 2;
Pg. 76158, col. 3; Pg. 76159, col. 1; Swap Reporting
NPR: Pg. 76581, col. 1; Pg. 76586, col. 2; Pg.
76594, col. 2;
(b) Total return swap transactions

(c) Inter-affiliate transactions

(d) Multi-asset swaps Swap Reporting NPR: Pg. 76586, col. 3;

IV. Reporting of Collateral Information

(a) General comments regarding collateral in

uncleared and cleared transactions.

(b) Specific comments in response to the Swap Swap Reporting NPR: Pg. 76585, col. 3; Pg. 76586,
Reporting NPR col. 3;
 35

V. Reporting Responsibilities

(a) Reporting responsibilities Real-Time Reporting NPR: Pg. 76146, col. 3; Pg.
76147, col. 3; Swap Reporting NPR: Pg. 76586, col.
2; Pg. 76593, col. 1; Pg. 76593, col. 2;

(b) Extraterritoriality Real-Time Reporting NPR: Pg. 76146, col. 3;

VI. Commodities Real-Time Reporting NPR: Pg. 76144, col. 2; Pg.

76152, col. 1; Pg. 76167, col. 1;

VII. Recordkeeping and daily trading records Real-Time Reporting NPR: Pg. 76149, col. 2;
requirements SD/MSP Recordkeeping NPR: Pg. 76668, col. 2; Pg.
76669, col. 3; Swap Reporting NPR: Pg. 76580, col.
1;
 ANNEX ANNEX 2

January 18, 2011

Block trade reporting

for over-the-counter
derivatives markets
Contents

Executive summary............................................................................................................. 1

1. Transparency and block trading .................................................................................... 3

1.1. Goals of transparency ...........................................................................................3
1.2. The cost of transparency – Illiquidity ...................................................................3
1.3. Block trade exemptions ........................................................................................5
1.4. Considerations for implementation.......................................................................6
2. Transparency in securities and futures markets ............................................................ 8
2.1. Trade reporting in the equity markets: the experience of the LSE .......................8
2.2. Trade reporting in the US futures markets............................................................9
2.3. Trade reporting in the corporate bond markets: the experience of TRACE .......11
3. The OTC derivatives markets ..................................................................................... 13
3.1. The rates markets ................................................................................................15
3.1.1. Interest rate swaps ...................................................................................15
3.1.2. Other OTC rates derivatives products .....................................................18
3.2. The credit derivatives markets ............................................................................20
4. Analysis of proposed rules .......................................................................................... 23
4.1. CFTC proposal ....................................................................................................23
4.2. SEC Proposal ......................................................................................................25
4.3. European proposals .............................................................................................27
5. Conclusion .................................................................................................................. 28

Appendix 1 ........................................................................................................................ 30
Appendix 2 ........................................................................................................................ 31
Appendix 3 ........................................................................................................................ 32
Executive summary

The Dodd-Frank Wall Street Reform and Consumer Protection Act (Dodd-Frank)
requires the Commodity Futures Trading Commission (CFTC) and the Securities and
Exchange Commission (SEC) to establish rules that provide for the real-time public
reporting of swaps 1 transactions, as well as exemptions to the real-time reporting rules for
large notional swap transactions and block trades (referred to collectively as “block
trades” throughout this paper).

A major challenge facing the CFTC and SEC is balancing the benefits of increased post-
trade transparency in over-the-counter (OTC) derivatives markets with potentially
adverse effects on market liquidity and pricing for end users. Both agencies have
proposed reporting rules that include exemptions for some large trades, though the CFTC
and SEC proposals differ substantially in how such block trades are treated.

The International Swaps and Derivatives Association (ISDA) and the Securities Industry
and Financial Markets Association (SIFMA) have jointly prepared this paper, with
support from Oliver Wyman, to help inform decisions about appropriate block trade
reporting rules for OTC markets. After reviewing the goals of transparency as well as the
importance of block trade reporting exemptions, the paper reviews and assesses trade
reporting regimes used in the securities and futures markets. Using trade-level data from
the interest rate and credit swap markets, it then illustrates distinctive market
characteristics that should inform an appropriate trade reporting approach for the OTC
derivatives markets. Finally, it assesses the CFTC and SEC proposals, identifying a
number of potential shortcomings and providing recommendations on how they could
be refined.

While not the primary focus of our research, one of the central conclusions of this paper
is that transparency can be increased in the OTC derivatives markets while preserving
liquidity. Other key findings include

 Special rules for block trades have been effectively used in equity, bond, and futures
markets to ensure that dealers are able to execute block trades on behalf of clients
without taking on unmanageable levels of risk, thus maximizing liquidity.
Introducing similar rules in the OTC derivatives markets will have an equally
beneficial effect

 Mechanisms used to balance the benefits and costs of transparency for large trades
include minimum block trade size thresholds, reporting delays, and limited disclosure
of block trading terms

1
“Swaps” is used throughout this paper to refer to OTC derivatives subject to regulation under Dodd-Frank by both the
CFTC and the SEC (which has authority to regulate “security-based swaps” in the legislation), unless otherwise noted.

1
 Trade reporting rules typically are developed and refined over time. TRACE, for
example, was phased in over three years for the US corporate, municipal, and agency
bond markets. Reporting rules for the London Stock Exchange experienced several
adjustments since 1986 to cope with changing market conditions. Trade reporting
rules for OTC derivatives should likewise be phased in, allowing regulators time to
test and refine preliminary standards

 Liquidity in OTC derivative markets is fragmented and varies considerably depending

on the specific product and terms of the contract (reference entity for CDS, maturity
for all products, etc.) traded, making a “one size fits all” approach to trade reporting
exemptions problematic

 The existing CFTC and SEC proposals for block trade reporting would likely increase
(rather than decrease) costs for end users, including institutional investors and
corporations, seeking to manage risk or raise capital

 The CFTC proposal establishes thresholds and reporting delays for block trades that
would have a significant adverse effect on trading in less liquid instruments. The
proposed rules would impose block minimum size requirements without appropriately
differentiating between instruments with very different levels of liquidity

 The SEC proposal, requiring full disclosure of notional trade size (albeit on a delayed
basis) for block trades, would likely impair liquidity for larger transactions in the
credit default swap (“CDS”) market, potentially leaving end users with significant
credit risk exposures

 TRACE-type volume dissemination caps should be employed for all OTC derivatives
products to ensure end users have sufficient sources of liquidity

Block trade rules should be set so that liquidity is not impaired, in order to preserve the
ability of investors and companies to hedge their risks in a cost-effective way. Rules
should be tailored to products – reporting rules for less liquid products should reflect
differences from more liquid products, for example. New rules for trade reporting should
be introduced using a phased approach. Reporting rules should be re-evaluated on a
regular basis to ensure they reflect the changing characteristics of the market.

2
1. Transparency and block trading

1.1. Goals of transparency

The Dodd-Frank Wall Street Reform and Consumer Protection Act (Dodd-Frank) calls on
the Commodity Futures Trading Commission (CFTC) and the Securities and Exchange
Commission (SEC) to adopt final rules for the public reporting of transaction and pricing
data for all “swap transactions” by July 15, 2011. Similar reforms are also being drafted
by regulators in Europe.

A major policy objective of Dodd-Frank is to bring greater transparency to the OTC

derivatives markets in the United States, while not adversely impacting liquidity in these
markets; in this regard, Dodd-Frank mandates that regulators take into account the impact
of liquidity when issuing rules regarding transparency. 2 The SEC and CFTC state in their
recent notices of proposed rulemaking 3 that the objectives of increased transparency are

 To provide regulators with access to comprehensive and timely market data,

facilitating the task of ensuring the safety and soundness of the financial system

 To promote lower transaction costs, greater competition, broader participation, and

improved liquidity through the public dissemination of trade data

These objectives are meant to be achieved, in part, through real-time, public reporting of
all OTC derivatives transactions (real-time is defined to be as soon as practicable).

1.2. The cost of transparency – Illiquidity

There is broad agreement that transparency can enhance market liquidity. However,
some forms of trade transparency can impair liquidity. Immediate reporting of large
trades will make hedging the risk in those trades more difficult as other market
participants anticipate the hedging trades that will be needed. These extra hedging costs
will be passed on to end users such as pension funds and companies. The result will be
higher costs for end users that rely on the OTC derivatives markets to manage risk.

2
See Dodd-Frank Sec. 727, which states that rules issued regarding the public availability of transaction and pricing
data for swaps shall contain provisions “that take into account whether the public disclosure will materially reduce
market liquidity.”
3
See Real-Time Public Reporting of Swap Transaction Data; Proposed Rule, Commodity Futures Trading
Commission, December 7, 2010 (http://www.cftc.gov/ucm/groups/public/@lrfederalregister/documents/file/2010-
29994a.pdf) (“CFTC proposal”) and Regulation SBSR – Reporting and Dissemination of Security-Based Swap
Information, Securities and Exchange Commission, November 19, 2010 (available at
http://www.gpo.gov/fdsys/pkg/FR-2010-12-02/pdf/2010-29710.pdf) (“SEC proposal”) for the detailed notices of
proposed rulemaking.

3
For example, when a corporation plans to raise a significant amount of capital by issuing
a fixed-rate bond, it is exposed to the risk that interest rates rise by the time it is ready to
issue the bond. The firm can hedge that risk by entering into an interest rate swap with a
market maker. The cost of the interest rate swap to the corporation will be directly
related to the price at which the market maker believes it can hedge the risk. If, however,
the terms of interest rate swap with the corporate end-user are reported in real time to the
market, then other potential counterparties will know that a market maker has executed a
large swap and needs to hedge the risk. As a result, these counterparties are likely to
adjust pricing (bid-offer spreads) in anticipation of the trade, increasing the risk of loss to
the market maker. 4 A rational market maker might react to this increased risk by (1)
refusing to enter into the large transaction with the corporate end-user (thereby reducing
liquidity), or (2) by increasing the price of the interest rate swap offered to the corporate
end-user (thereby increasing the firm's financing costs) to provide a buffer against the
increased risk. Either result is clearly detrimental to the end-user’s interests, and will
have a negative impact on that end-user’s ability to raise capital, damaging investment in
our economy. 5

Post-trade transparency in one transaction effectively leads to pre-trade signaling for

subsequent hedging related transactions. The knock-on negative effects – including
decreased liquidity, reduced ability to trade, and increased costs to hedge risks – will be
passed on to swaps end-users and those whose interests they represent. A reduced ability
to hedge risk or an increased cost to hedging risk will ultimately affect the economic
activity of companies and the savings and pensions of individuals.

The impact of transparency rules in major markets has been the subject of a number of
academic studies. 6 Several studies have found evidence of an adverse impact of
transparency in a range of markets. Madhavan, Porter and Weaver (2005), writing about
the Canadian stock markets, report “that the increase in transparency reduces liquidity.
In particular, execution costs and volatility increase after the limit order book is publicly
displayed.”

4
The size and direction of a transaction can be inferred before size is publicly disseminated based on the liquidity
premium in the reported price.
5
Similarly, a lender may wish to hedge a portion or all of a large new lending commitment to a corporation using credit
derivatives. If this new large hedging transaction is reported to the public before market makers can hedge their risk,
the cost and availability of the hedge will be negatively affected. This will then impact the lender’s ability to extend
credit or result in an increase in the cost of credit provided. Either event would, in turn, affect the corporation’s ability
to finance and expand its operations, and ultimately have a negative effect on the economy and job creation.
6
Bessembinder, H., Maxwell, W., Venkataraman, K., 2006. Market transparency, liquidity externalities, and
institutional trading costs in corporate bonds. Journal of Financial Economics 82, 251-288.
Edwards, A., Harris, L., Piwowar, M., 2007. Corporate bond market transaction costs and transparency. The Journal of
Finance 62, 1421–1451.
Madhavan, A., Porter, D., Weaver, D., 2005. Should securities markets be transparent?. Journal of Financial Markets 8,
265-287.

4
The same impact has been observed in other geographies. When the London Stock
Exchange (LSE) abolished fixed commissions in 1986, it initially required immediate
publication of prices. After experiencing a reduction in liquidity, the exchange allowed
the prices of trades exceeding £100,000 to be published after a 24-hour delay. In 1991,
the LSE changed its rules once again to introduce a 90-minute delay for trades that
exceeded a “social threshold” 7 of three times a normal market size trade. The LSE has
since changed the rules numerous times to achieve a better balance between transparency
and liquidity.

Futures exchanges have also recognized the impact of real-time reporting on liquidity of
listed futures and options. Some exchanges allow members to execute large transactions
bilaterally provided the terms are reported to the exchanges after a short delay. Chicago
Mercantile Exchange (CME) and Chicago Board of Trade (CBOT) rules require reporting
within five minutes for interest rate products during regular trading hours and 15 minutes
at other times.

Futures are relatively simple, fungible instruments that trade in markets with thousands of
participants, including large numbers of individual investors. Contracts are of small size
and liquidity can run to hundreds of thousands of trades per day. Block trades are very
rare (less than one per day) for many products, as block minimum sizes are very high
relative to the average ticket size and the trading that can be executed during the short
delay periods. End users either execute transactions piecemeal, taking basis and market
risk, or rely on OTC markets to conduct large trades.

1.3. Block trade exemptions

To preserve a high level of liquidity, market regulators frequently allow reporting

exemptions for block trades. In defining block trade exemption rules, market governing
bodies have three general mechanisms at their disposal: (1) minimum block trade size
thresholds, (2) trade reporting delays, and (3) limited disclosure.

 Minimum trade size thresholds – By definition, block trade exemptions require

clear definitions of the criteria that qualify transactions as block trades subject to
special reporting requirements. This threshold or “minimum block size” is commonly
a function of the average trade size or the cumulative distribution of trades for a
specific instrument. Market regulators frequently target a percentage of transactions
that will qualify as block trades, but also take into consideration a wide range of
market factors (e.g. average daily trade volume).

 Reporting delays – Reporting delays of appropriate length allow market participants

to hedge the market risk of block trades during the delay period. The delay
mechanism is most effective when instruments or contracts are very liquid and either

7
Social thresholds are based on trade sizes that are representative of a particular product or asset class, which is usually
an average trade size for that product or asset class.

5
 fungible or highly standardized, 8 and minimum block sizes are set at reasonable
levels. If these requirements are met, participants are able to hedge entirely the
market risk of block trades during the reporting delay.

 Limited disclosure – Many products do not have sufficient liquidity to ensure that
risks from a block trade can be sufficiently hedged during a relatively short reporting
delay period. In many cases, markets permit participants in block trades to report
limited information regarding block trades. The most common form is a volume
dissemination cap – the market is informed that a transaction above the cap has
occurred, but not the exact size of the transaction. Markets may also grant volume
dissemination caps for more liquid products in cases where the block trade is a
multiple of the block minimum. The limited disclosure mechanism ensures that price
discovery remains intact for block trades while protecting post-block trade hedging
needs from being anticipated by other market participants.

1.4. Considerations for implementation

When establishing rules for block trade exemptions, market governing bodies should
consider a number of factors

 Block trade thresholds should be set so that disclosure of such trades does not
adversely impact liquidity. The purpose of block trade exemptions is to maximize
liquidity by allowing traders to efficiently cover the risks associated with the
execution of large trades.

 Rules should be tailored to products and assume one size does not fit all. The
OTC derivatives market contains a wide variety of products. Some products are
reasonably liquid and standardized, and block trading rules can be defined with some
degree of confidence as to their effect on liquidity. Other products may have much
less liquidity and a large percentage of this small volume may consist of block trades.

 Reporting rules for less liquid products should reflect differences from more
liquid products. Block minimum size for these illiquid products need to be smaller,
delays longer, and information less complete to ensure end users get the best
possible pricing.

 In some markets, the aggregate size of block trades represents a significant share
of overall turnover. For example, 45% of trading turnover on the LSE is subject to a
delay in trade reporting (but only 5% of the number of trades). This seems to be a

8
Standardized products are those for which market quotes are easily available. They include stocks, bonds and futures
contracts. In the OTC markets, credit default swaps and some credit indices have become highly standardized. Interest
rate swaps with spot start and 3- or 6-month LIBOR as the floating rate index also exhibit reasonably high levels of
standardization.

6
 natural consequence for many OTC derivatives products given their large average size
and low level of trading frequency.

 All market participants should be able to (cost effectively) hedge their risk.
Block trading rules should be designed to allow market makers to cover their risks,
and thereby provide efficient, low-cost liquidity to other market participants. In
liquid, standard instruments trading volumes need to be examined relative to
minimum block sizes and reporting delays. For illiquid and customized (non-
standard) products, market makers are not able to offset risk in short periods of time
and the disclosure of limited information may be the only viable alternative.

 For highly customized products, price transparency may be uninformative and

misleading. An OTC derivative contract can be customized to such a degree that its
transparency does not meaningfully inform the rest of the market. In fact, reporting
prices for such products can be misleading for market participants trading similar, but
different products.

 New rules for trade reporting should be introduced cautiously, as the impact on
market liquidity for OTC derivatives is unpredictable. Raising thresholds over
time does not risk damage to market liquidity in the same way that immediate
introduction of high thresholds would. Experience bears this out. The LSE initially
implemented real-time reporting, but soon had to introduce 24-hour reporting delays
for some trades given the initial impact on liquidity. Conversely, TRACE gradually
phased in shorter block trade reporting delays (moving from 75 to 15 minutes).

 Block trading formulas should be re-calibrated regularly and methodologies

reviewed periodically to ensure they both remain appropriate for
changing markets.

 Great care should be taken to ensure that the specificity of trade data reporting
does not compromise the anonymity of participants.

7
2. Transparency in securities and futures markets

Real-time post-trade reporting requirements have been introduced in a number of markets

in the US and Europe. Almost all efforts to implement real-time reporting have
recognized the need for block trading exemptions to preserve market liquidity.
Regulators and other market governing bodies have recognized that dealers will only
make markets when given the ability to hedge risk economically. Each of the
mechanisms described in Section 1 (minimum block trade size thresholds, reporting
delays, and limited disclosure of transaction data) are commonly used, often in
combination with one another, to balance transparency and liquidity.

Below we briefly review the evolution of trade reporting for UK equities on the LSE, the
trade reporting regime for US exchange-traded futures and the impact of the introduction
of the TRACE trade reporting system for US corporate, municipal and agency bonds.
Collectively and individually, these case studies demonstrate that inadequate block
trading exemptions impair liquidity and affect market structure. Indeed, the challenge is
to devise a post-trade transparency framework where the overall benefit of increased
transparency is maximized by preserving market liquidity.

2.1. Trade reporting in the equity markets: the experience of the LSE

The LSE trade reporting experience highlights the need for accommodating block trades
through exemptions to real-time reporting rules even in highly liquid markets. Rules
governing the trading of equity shares in the London markets were the subject of
sweeping changes on October 27, 1986, an event widely referred to as the “Big Bang.”
The changes included abolishing fixed commissions, eliminating most of the restrictions
on the ownership of brokers and introducing electronic trading.

As part of these changes, the LSE introduced a trade reporting regime designed to
promote total transparency. It required all trades in major stocks to be reported within
five minutes. It became apparent that near immediate and full transparency hurt liquidity
as market makers faced increased risks with their equity positions known virtually
instantaneously. 9 Real-time reporting rules were modified in early 1989, when the LSE
permitted trades in excess of £100,000 to be reported on a delay of up to 24 hours
after execution.

As illustrated in detail in Appendix 1, block trading rules continued to evolve, becoming

more flexible and detailed over time. Some of the first social thresholds (block size
thresholds defined as a multiple of normal trade sizes) were incorporated in the early
1990s. Current rules provide for reporting delays that vary from 60 minutes up to three
trading days for very large trades. Throughout this period, the LSE has set its size

9
Ganley, J., Holland, A., Saporta, V., Vila, A., 1998. Transparency and the design of securities markets. Financial
Stability Review 4, 8-17.

8
thresholds and reporting delay periods in a manner that enables dealers to offset risk
during the reporting delay period.

The current post-trade reporting delay regime has produced very interesting results. In
terms of the number of trades, almost 95% of trades are reported without any delay; in
terms of value, approximately 55% of trade value is reported without any delay, and a full
30% is reported at the end of the current trading day or later. 10 These data show that the
market still supports significant levels of block trading, albeit with a multi-tiered
reporting delay framework, a fact that might be difficult to ascertain from the assessment
of the LSE reporting delays contained in the CFTC’s December 7, 2010 proposal. 11

Table 1: Current LSE equity deferred publication framework10

End of End of End of End of
Delay band No delay 60 mins 180 mins day day 2 day 3 day 4
Value of trades 55.4% 7.7% 6.9% 17.0% 3.1% 6.5% 3.3%
Number of 94.8% 2.7% 0.9% 0.5% 0.3% 0.7% 0.1%
trades

The evolution of the LSE rules demonstrates that the right mix of real-time reporting and
block trading exemptions is a difficult balance to strike. A real-time reporting regime,
even in highly liquid securities, requires ongoing analysis and frequent review.

2.2. Trade reporting in the US futures markets

The unique characteristics of the US futures markets highlights the potential

consequences of block trade thresholds set well above normal trade sizes and should
guide the implementation of any trade reporting regime for OTC derivatives (where block
trades are more common and critical to market liquidity).

Futures markets are generally highly liquid and well-suited to central order books that
accommodate small trades and broad market participation. Futures trade in standardized,
small contracts (in contrast to the OTC markets, in which each contract is customized and
can be very large). Futures markets require reporting as soon as trades are executed.
Block trades are permitted with brief reporting delays that generally range from 5 to
15 minutes.

10
www.londonstockexchange.com TradElect parameters.
11
“The London Stock Exchange (“LSE”) allows the publication of the trade to be delayed, if requested, for a specified
period of time which is dependent on the volume of the trade compared to the average daily turnover, as published by
LSE, for that particular security. LSE rules require member firms to submit trade reports to LSE as ‘close to
instantaneously as technically possible and that the authorized limit of three minutes should only be used in exceptional
circumstances.’” (CFTC proposal, p. 76166)

9
The delay allowed for reporting futures block trades can be examined in light of the level
of trading for each product. Table 2 below provides block trading and other market
details for selected CME Group products. The table shows, for select futures contracts,
the potential number of block trades (e.g. 200 contracts for gold futures) that could be
completely offset over the course of a typical five-minute delay period. We calculate the
average number of contracts that are traded during the delay period (e.g. 2,196 for gold
futures) based on the year-to-date average daily volume, and then calculate how many
minimum block trades this would accommodate.

Table 2: Block trading details for selected CME Group futures products 12
Contracts
traded in
Minimum 5-minute Number of Average
block size delay block Average number of
(number of period trades trade size block
Futures contracts) 2010 YTD based on offset in (number of trades per
Contract (A) ADV (B) ADV (C) delay (C:A) contracts) day
Gold 200 171,277 2,196 11 2 <1
Silver 200 42,120 540 3 2 <1
Copper 100 40,842 524 5 2 <1
Natural Gas 100 246,663 3,162 32 2 10
Light "Sweet" 100 679,282 8,709 44 3 >50
Crude Oil
Ethanol 10 2,477 32 3 3 3
30-day Fed 2,000 52,009 667 0 50 <1
Funds
30-Year 3,000 326,481 4,186 1 10 <1
Treasury
Bonds
5-year 5,000 509,712 6,535 1 15 <1
Treasury
Notes

As shown in the table, most block trades in energy products and metals can be offset
during the delay. However, block trades in interest rate products cannot typically be
offset during the reporting delay despite significant activity in these contracts. The table
also shows that block trades are relatively rare in all the contracts in the table and are
virtually non-existent in the contracts where the delay provides the least opportunity to
offset risk.

A natural outgrowth of the high block trading thresholds is small average trades and a
scarcity of transactions of even modest size. Contracts for Natural Gas and US Treasury

12
Trading data for November 21, 2010, CME Group.

10
Notes futures illustrate this point, shown in Figure 4 and Figure 5 of Appendix 2. We
examined trading activity for both of these contracts on the CME on November 21, 2010.
98% of transactions in Natural Gas futures included less than ten contracts; likewise, 98%
of transactions in 5-year US Treasury Notes futures had an underlying principal of less
than $5 MM (with a single trade exceeding the $500 MM block minimum).

As a result of this market and reporting structure, participants that wish to buy relatively
large contracts (e.g. $200-300 MM of 5-year US Treasury Notes futures) need to split the
order into many smaller orders, thereby assuming aggregation risk as other market
participants infer from the initial trades that there are more trades to come. The aggregate
trade can easily become expensive, as it takes longer to execute and markets move
adversely. Practically, the futures market block trading rules have resulted in larger users
moving to other markets – primarily to US government securities markets themselves and
the OTC derivatives markets.

For a market such as OTC derivatives where the trade sizes are less concentrated in small
transactions (in fact, the SEC proposal acknowledges that for products with very low
trading frequencies most trades can actually be considered block trades, as each trade
makes up a significant portion of daily volume 13), it will be challenging for real-time
transparency to support active trading in the sizes that market participants require for
active risk management unless minimum block sizes are set appropriately.

2.3. Trade reporting in the corporate bond markets: the experience of TRACE

In 2002, The Trade Reporting and Compliance Engine (TRACE) mandated the public
dissemination of corporate, municipal, and agency bond trading data.

Similar to the OTC derivatives market, these bonds are traded over-the-counter on a
secondary basis. Market makers collectively hold inventory in thousands of different
bonds in order to meet the expected demand of the market and to support client activities.
The TRACE bond reporting system was introduced in phases, starting in 2002. It initially
applied only to 500 large investment grade securities and 50 high yield issues, and
instituted a 75-minute delay for block trades. TRACE was subsequently applied to about
4,650 debt securities in 2003, and the block reporting requirement reduced to 45 minutes.
This phased introduction allowed the market impact of the changes to be assessed.

The current TRACE reporting timeframe was introduced in 2005. Under these rules,
dealers are required to report trades within 15 minutes of their execution. Reporting
consists of the particular bond, time and date, price, yield, whether the bond was bought
or sold, and the size. Size is disclosed if a trade is less than $5 MM for investment grade

13
“For example, a single trade that is equivalent in size to a full- or half-day’s average volume may be considered out-
sized. On the other hand, if a particular SBS trades only once or twice per day then every trade would be equivalent to a
full or half-day’s average size.” (SEC Proposal, p. 75231)

11
bonds, and if less than $1 MM for non-investment grade bonds; otherwise, size is
reported as being above those thresholds.

There is a significant body of research on the effects of TRACE on market practices

including research that addresses TRACE’s impact on liquidity. Bessembinder and
Maxwell (2008) 14 present a number of interesting findings. The authors find that trading
costs decreased for smaller trades following the introduction of TRACE. This occurred
because less-active market participants that typically trade in smaller sizes now had a
better informed view of market prices, which improved their bargaining position. This
conclusion was arrived at independently by several studies. 15

With an average trade size of $2.7 MM for institutional corporate bond trades in the OTC
market and 85% of trades greater than $1 MM, 16 it is clear that a block level of $5 MM
for investment grade bonds and $1 MM for non-investment grade bonds is indeed
relatively low. This exemption provides for real-time transparency for the majority of
trades, but at the same time limits the disclosure of trade size for the significant portion of
trades that qualify as block trades. The framework provides transparency, and also
accommodates trading in large sizes.

TRACE’s introduction has achieved one of its primary objectives – to better inform
smaller investors about recent bond trading prices and has done so while allowing block
trades to continue.

14
Bessembinder, H., Maxwell, W., 2008. Transparency and the corporate bond market. Journal of Economic
Perspectives 22, 217-234.
15
Bessembinder, Maxwell, and Venkataraman (2006); Edwards, Harris, and Piwowar (2007); and Goldstein, M.,
Hotchkiss, E., Sirri, E., 2007. Transparency and liquidity: A controlled experiment on corporate bonds. Review of
Financial Studies 20, 235-273.
16
Bessembinder, H., Kahle, K., Maxwell, W., and Xu, D., 2009, Measuring abnormal bond performance. Review of
Financial Studies 22, 4219-4258.

12
3. The OTC derivatives markets

The over-the-counter (OTC) derivatives market emerged in the early 1980s in response to
inefficiencies in the global debt markets. Some borrowers were able to raise debt in the
floating rate markets at comparatively lower rates than the fixed rate markets, and vice
versa. Early interest rate swaps allowed borrowers to "swap" fixed versus floating rate
payments on a common notional amount, resulting in lower financing costs for
both parties.

Swaps proved to be extremely flexible risk management tools, allowing end users to
manage a wide range of interest rate and currency risk 17 as well as lower financing costs.
However, matching counterparties with perfectly offsetting requirements was often
impossible and hampered the growth of the market. Interest rate swaps only became
commonplace when financial intermediaries began taking the other side of contracts,
warehousing and hedging risk on a portfolio basis without actually matching offsetting
client positions. By the early 1990s, these contracts became the instrument of choice for
end users to manage interest rate and currency risk. Soon thereafter, a comparable
derivatives market for the management of corporate, sovereign, and other credit risk
emerged (though it pales in comparison to the size of the interest rate swaps market).

From its inception, the OTC derivatives market has been an institutional market with
almost no retail participation. Indeed, it is illegal for most individual investors to trade
OTC derivative contracts. The first users of the market were large borrowers –
corporations, banks, securities firms, sovereigns and supranational agencies, such as the
World Bank and the European Investment Bank – who used swaps to adjust the risk
profile of their liabilities. Institutional investors, mutual funds, hedge funds and
insurance companies subsequently emerged as key users (and, in some cases, providers)
of derivatives, employing them to implement a variety of investment strategies.

The OTC derivatives markets evolved to maximize the flexibility of instruments for end
users. Market participants made use of the flexibility of OTC contracts to disaggregate
and manage a range of complex risks in a very precise manner. This has produced a
number of unique attributes that distinguish OTC derivatives markets considerably from
securities and standardized futures and options

 Limited market activity – Despite the hundreds of trillions of dollars in notional

outstanding OTC rates derivatives contracts, there is actually limited trading activity in
the market. Roughly 5,500 contracts are executed each day across interest rates swaps,
caps, floors, swaptions and other debt-related products in over 20 currencies. 18 Even if
products are categorized into multi-year maturity buckets, the most liquid contracts
17
Interest rate swaps can be customized to nearly any underlying reference interest rate, currency, and starting and
ending dates; thus, users are able to offset unwanted risks very precisely by engaging in the OTC derivatives markets.
18
TriOptima trade-level interest rate swap repository data over a 45-trading day period from August 1 to September 31,
2010.

13
 with maturities between five and ten years only trade 500 times per day (or less than
one per minute globally assuming a 12-hour trading day). The global universe of
outstanding OTC interest rate products, approximately five million transactions,
consists of the same number of trades as conducted in exchange traded interest rate
products on the CBOT and CME over the course of just 15-20 days. 19, 20

 Large individual transactions – The OTC derivatives marketplace primarily serves

large institutions with the need for large transactions. Individual trades by large
institutions may well represent activity for hundreds or thousands of distinct accounts
managed on behalf of small institutions and retail investors. The average size of a ten-
year USD interest rate swap was $75 MM during 2010, 21 whereas comparable
transactions in futures and securities markets are substantially smaller ($2 MM for ten-
year US Treasury Notes futures 22 and $3 MM for US corporate bonds, 23 respectively).
Other OTC products also tend to have substantially larger average transaction sizes
than their futures and cash counterparts. In many markets, OTC derivatives markets
have been the preferred (or only viable) venue for block trades.

 Limited participation – The OTC derivatives market is an institutional marketplace

with a relatively small number of active participants. JP Morgan estimates that there
are only 500 active participants in USD interest rate swaps and less than 250 in the
credit derivatives markets. 24 Active participants tend to be large institutions, banks,
securities firms, insurance companies, asset management firms (which represent a
number of smaller investors) and major corporations – this is due largely to balance
sheet requirements for trading in these markets. By contrast, the number of active
participants in the most liquid futures contracts (e.g. WTI Crude, S&P Index contracts)
is in the tens of thousands and includes a significant number of retail investors.

 Customization – There is no theoretical limit to the number of unique contracts that

can be executed in the OTC derivatives marketplace. In vanilla interest rate swaps
alone, there are more than 100,000 discrete instruments, 25 differentiated by underlying
currency, maturity and floating rate indices; in the credit default swaps market, there
are hundreds of thousands of discrete single-name contracts, differentiated by coupons

19
As measured by the TriOptima Trade Repository Report as of December 17, 2010, available at
http://www.trioptima.com/repository/historical-reports.html.
20
CME Group Exchange ADV Report, October 2010; CME Group daily trading activity for January 10, 2011.
21
TriOptima trade-level interest rate swap repository data over a 45-trading day period from August 1 to September 31,
2010.
22
Trading data for November 21, 2010, CME Group.
23
Bessembinder, H., Kahle, K., Maxwell, W., and Xu, D., 2009, Measuring abnormal bond performance. Review of
Financial Studies 22, 4219-4258.
24
Active market participants are defined as those trading at least five times per year in that product; the number of
actual users is much greater.
25
J.P. Morgan internal research and analysis.

14
 (at least two per entity) and maturities (40 quarterly maturities out to ten years) on
thousands of unique reference entities.

 Privately negotiated transactions – Because a significant share of trades are

customized and liquidity is provided by a relatively small number of participants, the
OTC derivatives market has not naturally evolved into an exchange-traded market with
thousands of participants like other instruments.

 Professional risk intermediation – Dealers offer OTC derivative contracts with terms
that are difficult to perfectly match on a consistent basis. Because of this and the long
duration of most contracts, dealers need to manage large portfolios of outstanding
contracts with significantly different risk profiles. This activity requires a substantial
investment in specialized staff, advanced technology and capital resources. Roughly
15 to 20 bank dealers are major market makers and competition for client business is
extremely strong among this group.

Many of the key differences between OTC and exchange traded derivatives markets are
briefly summarized in the table below.

Table 3: OTC derivatives and exchange traded derivatives market size and
participation 26
Ratio of market Average
Active Total participants to number of
Product participants Instruments instruments trades per day
Exchange traded markets
WTI futures >20,000 70 >300 >250,000
S&P e-Minis >150,000 5 >30,000 >200,000
OTC derivatives markets
Single-name CDS 200 75,000+ <0.003 4,000
Index CDS 200 100 2.0 2,000
Vanilla interest rate swaps 500 100,000+ <0.005 1,000

3.1. The rates markets

3.1.1. Interest rate swaps

The OTC rates derivatives market is one of the largest and most important financial
markets in the world today, yet only several thousand transactions are executed daily
across a wide range of currencies, reference rates, and maturities.

26
J.P. Morgan internal research and analysis.

15
Liquidity in rates derivatives is highly fragmented. The interest rate swaps market (the
most liquid segment of the market) is generally characterized by

 Low volumes in specific buckets (currency, maturity, etc.)

 Highly volatile daily trading volumes within specific contracts
 Relatively large transaction sizes and concentrated trading volumes

Approximately 4,000 27 interest rate swap transactions across all currencies and maturities
are executed per day by the 14 largest dealers. 28 Of those, approximately 1,500 trades are
in USD contracts with 500 trades per day in the 5-10 year maturity range. The number of
transactions executed in specific maturity buckets is much smaller: on average fewer than
100 seven-year USD interest rate swaps are completed on a typical trading day. 29 USD
and Euro interest rate swaps are the most commonly traded OTC interest rate derivatives.
Trading in other currencies is significantly lower.

Liquidity (as measured by trading volume) fluctuates considerably over time. Figure 1
shows the daily trading activity for the 14 largest derivatives dealers in USD interest rate
swaps with 5-10 year maturities, the most common maturity range, from August to
September 2010. Trading volume across this broad set of contracts ranged from 300 to
1,000 contracts per day, with significant spikes in activity driving up the average daily
volume. Volatility within specific maturity buckets is even greater.

27
Compared to the 1,000 trades per day listed in Table 3, the estimate of 4,000 trades per day for all interest rate swaps
includes non-vanilla interest rate swaps with odd maturities, non-spot starts, and non-major currencies.
28
ISDA estimates that the 14 largest dealers hold approximately 80% of OTC interest rate derivatives contracts
outstanding (Mid-Year 2010 Market Survey Results).
29
TriOptima trade-level interest rate swap repository data over a 45-trading day period from August 1 to September 31,
2010.

16
Figure 1: Daily trading activity in USD 3-month Libor interest rate swaps at
5-10 year maturity30

1500
# of trades per day

1000

500

0
8/2/10text
Blank August 2010 9/2/10text
Blank September 2010

The average transaction size for US$ interest rate swaps in the 5-10 year maturity bucket
is $75 MM with a significant number of transactions in excess of $200 MM. This is in
stark contrast with the futures markets where trade sizes are much smaller and 95% of the
trades in five-year Treasury Notes futures are less than $5 MM in size. The distribution of
transaction sizes for comparable contracts in the OTC and futures markets is provided in
Figures 2 and 3 below.

Figure 2: Trade size distribution in USD 3-month Libor interest rate swaps at 5-10
year maturity 30

7000

6000

95th
5000 percentile of
trading
# of trades

4000 activity ≈
$250 MM
5 times the
3000
average trade
size ≈ $375
2000 MM

1000

0
0 100 200 300 400 500+
Trade size ($MM notional)

30
TriOptima trade-level interest rate swap repository data over a 45-trading day period from August 1 to September 31,
2010.

17
Figure 3: Trade size distribution for Dec 10 5-year US Treasury Note futures
product for November 21, 2010 31

7500

95% of trades
are for less
than $5 MM
5000
Number of trades

notional
Only ~1 block
trade ($500
MM notional
2500 size) per day

0
0 10 20 30 40 50 60 70 80 90 100 500+
Notional trade size ($MM)

Figure 2 also shows thresholds derived from the CFTC proposed rules on minimum block
size trades – $250 MM (95th percentile) and $375 MM (five times the average trade size).
The CFTC proposal would require real-time reporting for over 98% of the market.

One of the stated goals of real-time reporting regulation is to tighten pricing spreads in
the OTC markets. In a recent blind test conducted by Atrevida Partners, 32 three large
investment firms each solicited executable price quotes from dealers on five separate IRS
transactions. For each transaction, three quotes were requested The dealer quotes were
compared to Bloomberg screen pricing as well as to one another. The best quotes
averaged 0.001% (one-tenth of a basis point) from the mid-market yield on Bloomberg.
The average spread between the best and worst quote (of the three total quotes) was
0.0038% (0.38 basis points) and as a percentage of the average quote this spread was
0.30%. The test indicates that pricing in the interest rate swap market is very competitive
despite the low volume of trades done each day by dealers. In addition, the close
relationship between Bloomberg and dealer quotes indicates that pricing is highly
transparent for customers.

3.1.2. Other OTC rates derivatives products

In addition to interest rate swaps, the OTC rates derivatives products consist of many
other product categories. The largest of these include forward rate agreements (“FRAs”),

31
Trading data for November 21, 2010, CME Group.
32
“Interest Rate Swap Liquidity Test” - a report sponsored by ISDA and conducted by Atrevida Partners in conjunction
with market participants in November 2010.

18
swaptions, caps and floors, and basis swaps. In all, these products represent
approximately 27% of outstanding notional and 20% of outstanding contracts. 33 (Both of
these figures may overstate the relative percentage of actual activity in these products as
interest rate swaps undergo regular “compression” cycles in which contracts are torn up.)

TriOptima lists 12 distinct categories of rates products. A snapshot of each product and
key market data is presented below.

Table 4: Overall “snapshot in time” trade summary by product type33

Notional Trade Count Average Trade Size
($TN) (’000s) ($MM)
Interest rate swaps 291 3,030 96
Overnight index swaps 57 96 531
(OIS)
Sub total 342 3,116 110
FRAs 51 145 351
Swaptions 28 193 143
Basis swaps 20 89 223
Caps/floors 12 78 151
Cross currency swaps 8 115 72
Exotic IRS 6 78 76
Other products 5 76 65
Sub total 129 774 167
Total 471 3,890 121

TriOptima data is for the 14 largest dealers, which skews the average trade size data
considerably as does the methodology for double counting cleared transactions (primarily
interest rate swaps and OIS interest rate swaps). But the data is clear with respect to the
non-interest rate swap products – trade size also varies considerably. These variations
along with differences in trade frequency and risk characteristics require that the products
should be examined independently with respect to block minimums, reporting delays and
disclosure requirements.

The TriOptima data indicates that the 14 largest dealers have approximately four million
outstanding contracts. These dealers represent an estimated 80% of the total notional,
implying that approximately five million OTC rate contracts are outstanding globally. By
contrast, the CME Group trades approximately 300,000 tickets per day in the US
government and Eurodollar futures contracts. The entire population of OTC interest rate
trades represents slightly more than the 15 days of activity in the interest rate futures
market of the CME Group. Approximately 5,500 OTC interest rate derivative
33
As measured by the TriOptima Trade Repository Report as of December 17, 2010, available at
http://www.trioptima.com/repository/historical-reports.html.

19
transactions are executed globally each day, equal to just 2% of the number of trades
conducted in the corresponding CME Group futures contracts. US$ trades are less than
1% of the daily volume in corresponding futures markets.

3.2. The credit derivatives markets

Like other OTC derivatives markets, the OTC credit derivatives markets are marked by
low volumes and large transaction sizes. The market is composed of approximately 4,000
single-name reference entities, on which protection is written (sold) or purchased, and
100 indices comprised of single-name reference entities. Volume and size characteristics
of the CDS market are summarized on the following page (graphs containing additional
CDS market data are contained in Appendix 3).

Overall average daily volume is approximately 6,500 contracts, of which 4,500 are
single-name reference entities and 2,000 are credit indices. Approximately 1,000 single
name reference entities are traded more frequently and consistently. They include
approximately 930 corporate and 65 sovereign entities. In all, average daily trading
volume for these 1,000 names amounts to approximately three trades per day for each
reference entity. Each reference entity will have at least 80 quotable contracts: 40
different maturities and two different coupons. In all, there are over 80,000 individual
contracts for these 1,000 names. The vast majority of individual contracts trade
very infrequently.

Table 5: Summary of CDS trading behavior 34,35

Daily Trading Activity Trade Size
Number
of Average % of RE % of RE
90th
reference daily with <5 with >20 80th
entities trades trades trades Mean percentile percentile
(RE) per RE per day per day ($MM) ($MM) ($MM)
Single-name
Corporates 935 3 79% <1% 8 7 10
Sovereigns 65 8 56% 11% 13 16 24
Total 1000 3 77% 1% 8 8 11
Indices
High Grade 80 15 79% 14% 15 100 150
High Yield 35 20 65% 16% 20 30 55
Total 115 17 75% 15% 16 80 120

34
DTCC Credit Default Swap (CDS) trade repository for all trades from March-June 2010
35
Trade size distribution determined by number of transactions (e.g. for a sample of 100 trades, the 80th percentile
represents the threshold, in $MM, that separates the smallest 80 trades and the 20 largest trades)

20
Of the corporate reference entities, nearly 80% trade less than five contracts per day, with
many names that average less than one trade per day. The table above shows that only
two corporate reference entities traded 20 or more times per day (across all contracts
outstanding on a given reference entity) over the three-month period. In a 12-hour
trading day, this represents one trade done globally every 36 minutes.

It should also be noted that the table is a snapshot of the entire market on an average day.
This means that a reference entity that trades 20 times on a given day may trade less than
20 times on a subsequent day. Average trade size for corporate reference entities is
$8 MM and more than 90% of trades are for less than $10 MM

Of the sovereign names, approximately 55% trade less than five times per day. The table
shows that seven sovereign reference entities trade 20 or more times per day. Average
size for a sovereign name is $13 MM and 90% of trades are for less than $25 MM.

To show an example of trading in the sovereign CDS market, Figure 3 shows daily
trading activity for the Kingdom of Spain, one of the most frequently traded single-name
reference entities. Daily trade volumes have varied over a three-month period from fewer
than 10 contracts to as many as 125. The average number of contracts traded is 35 per
day and the average turnover of the “on-the-run” five-year contract is 21 trades per day.
This trading volume is in stark contrast to that of equity and liquid futures contracts.

Figure 3: Most actively traded sovereign CDS daily trading activity 36

150

100

50

March April May June

Kingdom of Spain

It is useful to compare the TRACE process with what might be appropriate for the CDS
market. TRACE took three years to implement and ended up with volume dissemination
caps of $5 MM for investment grade bonds and $1 MM for high yield. The average size

36
DTCC OTC CDS trade repository; 3 month data set of CDS trades from March to June, 2010.

21
trade in single name corporate CDS ($7 MM) is higher than the average investment grade
corporate bond trade ($2.7 MM) and trading activity is much lower in CDS and dealers
often take weeks or more to close out large positions. 37 We believe that trade reporting
requirements for CDS products should be phased in and adjusted over time, as was the
case with TRACE, both with respect to mechanics as well as volume dissemination cap
sizes.

There are far fewer credit indices traded compared to single-name reference entities.
Analyzing the aggregate trading in each index, we find there are about 100 liquid indices.
The ten most active indices make up 75% of the total daily volumes; the four most active
indices make up 50% of the market's total trading volume. Each of the top four indices
trades more than 100 times per day, whereas 75% of the remaining indices trade less than
ten times per day. The average contract size is approximately $75 MM for investment
grade indices and $30 MM for high yield indices. 38 We believe a process similar to
TRACE can be developed as well for credit indices, differentiating investment grade from
high yield instruments, and setting the volume dissemination caps at relatively low initial
levels to ensure liquidity remains in the market.

The OTC credit derivatives markets illustrate well a common feature of swaps markets in
general – the market is fragmented across a wide range of instruments. This market
fragmentation means that individual instruments trade infrequently, even in asset classes
considered to be relatively liquid. For example, CDS contracts on most reference entities
trade less than five times per day, and there are dozens of contracts per reference entity.
This distinctive level of trading frequency should directly inform the development of an
effective block trade reporting approach.

37
Bessembinder, H., Kahle, K., Maxwell, W., and Xu, D., 2009, Measuring abnormal bond performance. Review of
Financial Studies 22, 4219-4258.
38
DTCC OTC CDS trade repository; 3 month data set of CDS trades from March to June, 2010.

22
4. Analysis of proposed rules

4.1. CFTC proposal

Dodd-Frank has designated the CFTC as the primary market regulator for certain OTC
swaps contracts. It includes certain swaps tied to interest rates, currencies, commodities,
baskets or broad-based indices of equities and indices of indebtedness of groups of
reference entities (credit indices). The legislation requires real-time reporting (as soon as
practically possible) for certain swaps, but assigns regulators the task of developing
reporting rules that reflect the effects of real-time reporting on market liquidity. The
CFTC published its proposed rules on real-time reporting in the Federal Register on
December 7, 2010. In this section of the paper, we examine the proposed rules with
respect to interest rate and credit index swaps.

The proposed rules require that all swaps be reported in real time unless a transaction
meets the minimum block trading size, in which case the transaction is subject to a
15-minute delay in reporting. All transactions, whether executed on a swap exchange or
bilaterally, are subject to real-time reporting and subject to the same minimum trading
sizes in order to qualify for the 15-minute delay.

Minimum block trading sizes are determined generally by Swap Data Repositories
(SDRs). SDRs aggregate swap products within asset classes into smaller groups called
Swap Instruments. The rule itself defines a Swap Instrument as “a grouping of swaps in
the same asset class with the same or similar characteristics.” In the explanation of the
proposed rules, the CFTC “believes that it is appropriate to group particular swap
contracts into various broad (emphasis added) categories of swap instruments.” It goes
on to state, “the Commission believes that within each asset class there should be certain
criteria that are used to determine a category of swap instrument. For example, swaps in
the interest rate swap asset class may be considered the same swap instrument if they are
denominated in the same major currency (or denominated in any non-major currency
considered in the aggregate) and if they have the same general tenor.” Additionally, “... a
single category of swap instrument may be ‘US dollar interest rate swaps in a short
maturity bucket, including swaps, swaptions, inflation-linked swaps, etc. and all
underlying reference rates.’” With respect to credit indices, they all are presumed to be
the same Swap Instrument. 39

Public dissemination of the notional amounts of transactions is subject to a rounding

convention. This convention provides, among other things, that notional principal of
contracts in excess of $250 MM be reported as $250 MM+. The explanation of the
proposed rules cites the rounding convention as providing a degree of anonymity. As
discussed below, this is an important element in preserving the availability of block
trading.

39
CFTC proposal, pp. 76153, 76172.

23
The minimum block trading sizes are then subject to a two-part test. The first part, called
the Distribution Test, is the notional amount that is greater than 95% of the transactions of
a Swap Instrument, where the rounding convention has first been applied. The second
part, called the Multiple Test, is the result of multiplying a block multiple by the social
size of the Swap Instrument. The block multiple is proposed to be five and the social size
is the largest of the Swap Instrument's mode, median or mean. The minimum block
trading size is then simply the higher of the results produced by the Distribution Test and
the Multiple Test.

Analysis of the CFTC proposed rules

As proposed, we see three significant areas where improvements might be made to the
current CFTC proposal

 Narrower definition of swap instruments with appropriately tailored rules –

We believe the definition of Swap Instrument contained in the proposed rules is
excessively broad. For example, it classifies a two-year plain vanilla interest rate
swap and a three-year Bermuda options contract as the same Swap Instrument. The
liquidity of each of these products is vastly different and disclosure of a $250 MM
trade in each product will have a different impact on market liquidity for each one.
For interest rate products, it would be more advisable to retain the critical tenor
division but also allow for additional Swap Instruments in the interest rate product
market. For example, fixed rate interest rate swaps against major floating reference
bases might be grouped into three Swap Instruments (short, medium and long term).
Similarly, swaptions, caps and floors with European or American exercise provisions
could be another group of three Swap Instruments. Another grouping might apply to
liquid basis swaps and all other products might comprise one or more additional
groupings.

 Broader application of rounding convention – A second issue relates to the

rounding convention as its use mitigates the very short delay of 15 minutes. Many
large transactions, whether they are OTC derivatives, equities or corporate bonds,
cannot be offset within a relatively short reporting delay. This has been the
motivation for equity exchanges to permit long, multi-day delays while other markets
such as the corporate bond market have used volume dissemination caps. TRACE
uses such caps of $5 MM and $1 MM for investment grade and non-investment grade
bonds, respectively, in conjunction with a reporting requirement of 15 minutes. As
written, the rounding convention would permit the most liquid interest rate derivatives
products to be executed in very large size (e.g. $1 BN or more) and dealers would be
able to offset risk, confident that the market only knows of a $250 MM+ trade. The
rounding convention will not, however, provide similar protection to other swaps
products that may be less liquid. We believe it would be most useful to adopt
rounding conventions for each of the expanded set of Swap Instruments
recommended above, and that such rounding conventions reflect the liquidity
characteristics of the specific Swap Instruments.

24
 Broader test of block trading to account for average daily volume – The two-part
test used to define “block trades” may fail to capture the full breadth of block trading
activity. The example provided in the CFTC proposed rules provides an illustration
of a swap instrument with all transactions between $50-60 MM in notional size. 40
However, the “social size” for the instrument is $55 MM, yielding a minimum block
size of $275 MM. This text neglects to specify that the average daily volume was
$1,375 MM, placing the block size threshold at approximately 20% of daily trading
volume for the instrument. As a general matter, we believe block minimums for
single trades should be established at levels well below 20% of average daily volume.
Both the Distribution Test and the Multiple Test should be bounded by a percentage
well below 20% of average daily volume. We also believe that aggregate block
trading activity should not have a pre-determined limit. As noted in Section 2.1, LSE
block trading activity, amounts to 45% of aggregate trading volume without damaging
the transparency of overall prices.

 Initial reporting delay of greater than 15 minutes – The CFTC’s proposed delay
period is inadequate to allow market participants to hedge risks from large trades or
trades in illiquid instruments. The changes described above might eliminate the need
for longer reporting delays but longer reporting delays for blocks should also clearly
be considered.

4.2. SEC Proposal

Dodd-Frank has designated the SEC as the primary market regulator for security-based
swaps. These include swaps tied to equities of single entities as well as single-name CDS
and narrow-based baskets or indices of securities. The SEC published proposed rules on
November 19, 2010. In this section, we will examine the proposed rules with respect to
single-name CDS.

The proposed rules require that all security-based swaps be reported in real time unless a
transaction meets minimum block trading size. The proposed rules specify general
guidelines for setting block trading thresholds but do not set specific levels. The
proposed general guidelines appear to be less certain than the proposed rules for real-time
reporting from the CFTC. However, the SEC states that it will assess the distribution of
single-name CDS trades and determine some size cut-off which will be the block trading
minimum. The example used by the SEC suggests that the minimum block trade size will
be $15 MM to $30 MM. The minimum will not vary by maturity of the instrument or by
the type or liquidity of the reference entity.

Block trades will still require real-time reporting of execution and pricing but the notional
size will be suppressed for a minimum of eight hours and a maximum of 26 hours, based
strictly on the time of day a transaction is executed.

40
CFTC proposal, p. 76162.

25
Analysis of the SEC proposed rules

The SEC is proposing a methodology that differs substantially from the TRACE reporting
system. TRACE requires 15-minute reporting of all trades but has a volume
dissemination cap of $5 MM for investment grade securities and $1 MM for
non-investment grade securities. Trades larger than the caps are merely noted as such.
There is no second wave of transaction reporting that includes actual notional size. By
contrast, the SEC proposes reporting complete notional size transaction data (albeit with
substantial reporting delays).

We believe that this reporting of actual block trading notional amounts will impede the
execution of very large trades. This is problematic because the CDS market is
characterized by a significant number of very large trades relative to the cash corporate
bond market. This is due in part to the fact that corporate bond trades involve securities
of modest size, while the CDS market references an entity's entire stock of debt with the
same seniority. We agree that the CDS block sizes should be larger than TRACE's
volume dissemination caps, but we believe the CDS market is better suited for large
trades and does not have the same protection under the current proposal as does the
market of smaller trades (corporate bonds).

As noted in Section 4.3 below, another approach towards single-name CDS reporting has
been proposed by the Committee of European Securities Regulators (CESR). CESR will
require immediate reporting of transactions under the “social threshold” (€5 MM or
lower). Transactions greater than €5 MM and less than €10 MM will require end of day
trade size and price information. Trades in excess of €10 MM will be disclosed at the end
of the trading day without actual size data. This multi-tiered reporting system is more
appropriate for very large trades than the system proposed by the SEC. The disclosure of
very large trade sizes in relatively illiquid markets may impact liquidity and prices for
extended periods.

As we have noted, one product (corporate bonds) will have a more favorable reporting
environment for block trading than another (single-name CDS) if the SEC's proposal
becomes final. Another jurisdiction (Europe) is considering a second reporting
environment that also provides more protection to block trading than the SEC. We
believe that reporting of actual size trades, albeit with a delay, will reduce the number of
block trades and most likely the aggregate volume of single-name CDS trading. We do
not think a goal of the process of establishing minimum block trade sizes is to reduce the
actual number of block trades. Instead, the goal should be to balance the need for
transparency with its effect on liquidity.

The single-name CDS market is much different than the markets for much more liquid
instruments. Dealers are apt to have single-name CDS positions on their books for days,
if not weeks or months. Market knowledge of the existence of these positions will impact
prices for considerable periods of time. Both the TRACE process and the

26
recommendations of CESR contain volume dissemination caps. We believe these should
also be part of the block trading rules for CDS products.

4.3. European proposals

The rulemaking process regarding trade transparency in Europe started shortly after the
Markets in Financial Instruments Directive (MiFID) introduction in 2007, and the
rulemaking process continues (e.g. MiFID II). The directive brought significant changes
to the European regulatory framework for secondary markets. Already, CESR assessed
the impact of these changes for corporate bonds, structured finance products, and credit
derivatives markets, but since other OTC derivatives markets were not studied originally,
CESR is now considering a post-trade transparency regime for the following financial
instruments: interest rate derivatives, equity derivatives, foreign exchange derivatives and
commodity derivatives.

The general framework used by CESR (for CDS products) has been one of tiered trade
size buckets by asset class, with varying levels of transparency for each. In the lowest
bucket, price and volume reporting is proposed to be in real time, or as close to real time
as possible. In the middle bucket, price and volume reporting is proposed to be at the end
of the trading day. In the highest bucket, price reporting without actual volume (but with
an indicator that the trade is indeed in this highest bucket) is proposed to be at the end of
the trading day.

CESR recommends that the calibration of block thresholds and time delays for the
proposed regime should ideally be based on the liquidity of the instrument in question.
However, due to the nature of these OTC markets, there is currently an absence of trading
data which can reliably be used to calibrate a transparency regime. CESR therefore
recommends that initial calibration be based on the average trading size of each of the
markets in question. Once the regime is implemented this information will quickly
become available for regulators to further study the market and refine the proposed
framework. At the core of CESR’s recommendations is the need to undertake a post-
implementation review for all asset classes, with plans to reach conclusions one year after
introducing the new transparency obligations.

27
5. Conclusion

The foregoing discussion clearly demonstrates that a very high degree of transparency can
be introduced to the OTC derivatives market while preserving its liquidity. Building an
effective trade reporting system for the OTC derivatives market, however, is a significant
challenge, partly because there is no established framework for real-time public reporting
in OTC derivatives today. Models that function well in securities or futures markets are
poorly suited to OTC derivatives, which are characterized by a diversity of instruments,
low trade frequency but large transaction sizes for many instruments, and a relatively
small number of large, sophisticated participants. Regulators will need to walk a fine line
to effectively balance market transparency with liquidity.

The proposed rules of the CFTC and SEC recognize this goal, but are more appropriate
for transactions in cash securities or futures than for transactions in OTC derivatives. If
established, they could pose a significant risk of impairing market liquidity or
dramatically increasing execution costs.

Drawing on the lessons from three trade reporting regimes and market data on interest
rate and credit derivatives, we propose several considerations that an effective trade
reporting regime for OTC derivatives should reflect

 Block trade thresholds should be set so that liquidity is not impaired, in order to
preserve the ability of investors and companies to hedge their risks in a
cost-effective way

 Rules should be tailored to products and markets. Rules for less liquid products
should be different from rules for more liquid products. One size does not fit all

 New rules for trade reporting should be phased in and refined over time. Rules
should be re-calibrated and methodologies re-assessed in light of experience and
market changes

 Block trades may constitute a significant amount of trading volume for

certain products

 For highly customized products, price transparency may be uninformative

and misleading

 Volume dissemination caps such as those found in TRACE are important means of
mitigating the effects on liquidity of real time reporting for all OTC derivatives
products

The proposed rules by the CFTC and SEC should be modified with these considerations
in mind. Most importantly, rules should calibrate block trade thresholds to reflect trade

28
volume and liquidity for specific instruments and limit disclosure for certain large
block trades.

29
Appendix 1

Table 6: LSE experience with post-trade transparency regimes 41

Time Period Rule Reason for change
42
Oct ’86 – Feb ’89 All trades in actively traded stocks in 5 LSE considers transparency as an
43
minutes important feature of the new
trading system
Feb ’89 – Jan ’90 Prices in trades >£100,000 in actively To help increase low volumes and
traded stocks in 24 hours. Other trades mitigate losses made by market
as before makers
Jan ’90 – Jan ’91 Trades >£100,000 in actively traded To increase transparency
stocks same as before. Other trades in
actively traded stocks in 3 minutes
44
Jan 91 – Dec 93 Trades >3x NMS in 90 minutes. Other OFT report (1990) stated that
trades in 3 minutes current regime was uncompetitive
Dec 93 – Jan 96 Trades >75x NMS within 5 days or until These trades were viewed as
90 per cent unwound, whichever is the particularly informative and
earliest. 3x NMS - 75x NMS in 60 immediate publication would harm
minutes. Other trades in 3 minutes liquidity
Jan 96 – Dec 99 Trades >6x NMS within 60 minutes. OFT Report (1994) reiterated the
Trades >75x NMS as before. Inter-dealer conclusions of the 1990 report
trades excluded from publication delay. based on the empirical evidence
Other trades in 3 minutes of Gemmill (1996). Also, a SIB
report (1995) recognised the
possibility of a trade-off between
transparency and liquidity
… … …
45
Present day 4 average daily trading (ADT) bands To distinguish between different
created for each currency, with greater levels of trading across products
delays (60 minutes up to 3 trading days
after trade) allowed for transactions of
increasing size within each band

41
Ganley, J., Holland, A., Saporta, V., Vila, A., 1998. Transparency and the design of securities markets. Financial
Stability Review 4, 8-17.
42
The most actively traded securities in the Stock Exchange Automated Quotations System (SEAQ). About 100
securities came into this category when it was in official use by the London Stock Exchange. These were shares of
companies with high turnover and high market capitalization.
43
Publication refers to date, time and the name of the stock, whether the trade was a buy or a sell, its price and volume.
Until 1991, publication delays referred to price only. Subsequently, publication delays referred to both price and
volume.
44
NMS (Normal Market Size) is given by (2.5%/250x(customer turnover in the past 12 months)/(closing mid-price on
last day of quarter)).
45
www.londonstockexchange.com TradElect parameters.

30
Appendix 2

Figure 4: Trade size distribution for Dec 10 natural gas futures product for
November 21, 2010 46

20000

98% of trades
15000 are for less
than 10
Number of trades

contracts

10000
Only ~10
block trades
(100
5000 contracts)
occur per day

0
0 25 50 75 100+
Contracts per trade

Figure 5: Trade size distribution for Dec 10 5-year US Treasury Note futures
product for November 21, 201046

7500

95% of trades
are for less
than $5 MM
5000
Number of trades

notional
Only ~1 block
trade ($500
MM notional
2500 size) per day

0
0 10 20 30 40 50 60 70 80 90 100 500+
Notional trade size ($MM)

46
Trading data for November 21, 2010, CME Group.

31
Appendix 3

Figure 6: Trade frequency distribution of the 930 most actively traded single-name
corporate reference entities (all coupons and maturities) 47
50%

40%
% of corporate entities

30%

20%

10%

0%
0-2 3-5 6-8 9-11 12-14 15-17 18-20 20+
# of trades per day
Corporate CDS

Figure 7: Trade frequency distribution of the 65 most actively traded single-name

sovereign reference entities (all coupons and maturities)47
40%
% of sovereign entities

30%

20%

10%

0%
0-2 3-5 6-8 9-11 12-14 15-17 18-20 20+
# of trades per day
Sovereign CDS

47
DTCC OTC CDS trade repository; 3 month data set of CDS trades from March to June, 2010.

32
Figure 8: Trade size distribution of 5Y USD based single-name corporate CDS
reference entities 48
30%

20%
% of trades

10%

0%
0-2 2-4 4-6 6-8 8-10 10-12 12-14 14-16 16-18 18-20 20-22 22-24 24-26 26-28 28-30 30+
Trade size ($ MM)
% of 5Y Corporate CDS

Figure 9: Trade size distribution of 5Y USD based single-name sovereign CDS

reference entities48
30%

20%
% of trades

10%

0%
0-2 2-4 4-6 6-8 8-10 10-12 12-14 14-16 16-18 18-20 20-22 22-24 24-26 26-28 28-30 30+
Trade size ($ MM)
% of 5Y Sovereign CDS

48
DTCC OTC CDS trade repository; 3 month data set of CDS trades from March to June, 2010.

33
Figure 10: Trade frequency distribution for index based CDS contracts 49
50
•22 entities trade 10+
Number of index reference entities in
times/day

•10 trade 50+ times/day

bucket

•5 trade 100+ times/day

25

0
0 1 2 3 4 5 6 7 8 9 10+
Average num ber of daily trades

Figure 11: Trade size distribution of investment grade USD based index CDS
reference entities49
20%
% of investment grade USD
index CDS trades

10%

0%
0-10 50-60 100-110 150-160 200-210 250+
Trade size bucket ($MM notional)

49
DTCC OTC CDS trade repository; 3 month data set of CDS trades from March to June, 2010.

34
Figure 12: Trade size distribution of high yield USD based index CDS
reference entities50

% of high-yield USD index CDS 40%

30%
trades

20%

10%

0%
0-10 50-60 100-110 150-160 200-210 250+
Trade size bucket ($MM notional)

50
DTCC OTC CDS trade repository; 3 month data set of CDS trades from March to June, 2010.

35