Popular Delusions Issue December 2019 PDF
Popular Delusions Issue December 2019 PDF
Popular Delusions Issue December 2019 PDF
POPULAR
DELUSIONS
"The world is at all times the dupe of some bubble or other."
- Col William Rafter
WORDS ACTIONS
Cargo Cult Finance and the smart Postcards from the depths of
beta bubble ................................................. 01 despair: jack-up edition .......................... 05
Richard Feynman warned academia of the dangers of Here’s a summary of today’s oil market: distressed; dumb
research malpractice in 1974, effectively predicting the money selling; smart money buying. Interested? You
“replicability crisis” currently sweeping across university should be. We check out the current situation in the oil
research departments the world over. Those departments field services sector and find some extraordinary value.
complain that the pressure to meet funding requirements
and to publish ‘something’, and ideally ‘something novel’
is to blame. But those same pressures apply to the BOOK REVIEW
financial industry too. Why should it be immune?
Actually, it isn’t. The academic evidence for factors is
The Man Who Solved the
dubious at best. Yet the ‘smart beta’ industry has sold over Market by Greg Zuckerman .................. 10
$1tr worth of it to investors. Is the breakdown of the
“value factor” merely the beginning of a slow burning “There’s Renaissance Technologies and there’s everyone
scandal? else” opined the Economist newspaper once, when
referring to the phenomenon that is Jim Simons and his
team of scientists. Unlike the charlatans peddling
Percentage of asset owners smart-beta, the guys at Rentec do real science, and
adopting smart beta strategies operate on a different plane from the rest of the industry.
Indeed, so unique are they that you might have
60%
wondered, as have I, if there was anything mortals like
us could learn from their story. Well, wonder no more. I
50%
just read Zuckerman’s fascinating book, and it turns out
there is.
40%
30%
From all of us at Calderwood we’d like to say thank
20%
you for your support and wish you all a very Merry
Christmas and a Happy New Year. See you in 2020!
10%
0%
2014 2015 2016 2017 2018 2019
In 1923, Robert Millikan won the Nobel Prize justifying new arguments by basing them on
for Physics for his oil drop experiment, arguments that they already like to believe in
measuring the charge of an electron, (to the extent that people then believe their
conducted with PhD student Harvey Fletcher own motivated arguments, they get trapped in
between 1909 and 1913. It remains a a kind of internal echo chamber.)
controversial episode in the history of science
Psychologists believe the effect of motivation in
though, partly because it appears that Millikan
reasoning is strongest for beliefs which
forced Fletcher to relinquish his claim on
intertwine with how a person sees him/herself.
co-authorship as a condition for receiving his
For example, suppose I go out and buy a
PhD. Fletcher, it seems, should have been a
second hand car. I might start out being very
co-prize winner. But it is controversial for other
rational about it: reading lots of online
reasons too.
reviews, renting a similar model for a weekend
We now know that the measure he calculated just to be sure I’m going to like what I buy,
was slightly too low (by about 0.6%). But if evaluating its service history etc. But then
you plot other scientists’ estimates of the suppose that after I’ve made the decision and
charge of an electron in the years following bought the thing I find that it’s uncomfortable,
Millikan’s 1913 paper, we find that the first one sluggish to overtake and more difficult to
is a little bit bigger than Millikan’s, the next a handle than I remember. I can’t get as much
little bigger still, the next a little bit bigger than luggage in the back as I thought either.
that, and so on. They only converge on the
If I was super rational I’d admit my mistake. I’d
higher (correct) value several decades later.
take the car back to the dealer and start the
Why didn’t they discover that the number was process of buying a new car again. Except
higher right away? During his commencement doing that will involve a time cost, because I’ll
lecture at Caltech in 1974, Richard Feynman have to start all that research from square one.
offered the following answer: Worse, I’ll take a financial hit because the
dealer will only buy the car back at a lower
“It's a thing that scientists are ashamed of - this price than that which I paid for the car. But
history - because it's apparent that people did worst of all, I spent ages telling everyone how
things like this: When they got a number that clever I was doing all this research, and how
was too high above Millikan's, they thought excited I was about my impending purchase. If
something must be wrong and they would look I return the car I’ll look like a bit of a dummy,
for and find a reason why something might be and I don’t like to see myself as a dummy.
wrong. When they got a number close to
Millikan's value they didn't look so hard. And so So I don’t take the car back. I keep it, and I
they eliminated the numbers that were too far double down. I tell myself how much I love it:
off, and did other things like that.”1 its lack of acceleration is actually great,
because it disciplines me into driving more
Feynman’s description of the mindset of the safely; the cushioning isn’t so bad either once
scientists following up on Millikan’s you get used to it; and there’s plenty of space
experiment would today be a textbook for everything you might need on a trip, so
example of what psychologists call motivated long as you pack carefully; etc etc
reasoning in which people believe things .
because they want to believe them (ie they are Obviously, I’m just paraphrasing from the
motivated to believe them). Motivated textbook examples, because I’ve never actually
reasoners get trapped in a kind of loop, done anything like this before … but the
1
See http://calteches.library.caltech.edu/51/2/CargoCult.pdf for the entire pdf, or the final chapter of his book “Surely
you’re joking, Mr. Feynman!” For an adapted version.
01
19TH DECEMBER 2019 | WWW.CALDERWOODCAPITAL.COM
motivated reasoning literature highlights that cannot do that, because the experiment has
there are many factors which influence a already been done and you would be wasting
decision or judgement. Critically, ‘the truth’ is time. This was in about 1947 or so, and it seems
only one of those factors, and often to have been the general policy then to not try to
subordinate to emotional ones. repeat psychological experiments, but only to
change the conditions and see what happened.
We’ve all seen investors go public on a banner
investment, only for it to blow up in their faces. Nowadays, there's a certain danger of the same
Have you ever wondered if they’d have stuck thing happening, even in the famous field of
with their losing position for so long had they physics. I was shocked to hear of an experiment
not gone so public about it? The public show of being done at the big accelerator at the National
cleverness having the effect of adding to their Accelerator Laboratory, where a person used
cost of reversing the position when the deuterium. In order to compare his heavy
investment seems to be going wrong, thus hydrogen results to what might happen with
increasing the motivation for them to reason light hydrogen, he had to use data from someone
that they are in fact right, notwithstanding the else's experiment on light hydrogen, which was
onslaught of contrary evidence? done on different apparatus. When asked why,
he said it was because he couldn’t get time on the
My car example was made up (honestly). But
program (because there’s so little time and it’s
I’ve stuck with hires, stayed in jobs, and held
such expensive apparatus) to do the experiment
onto views and investments for too long
with light hydrogen on this apparatus because
because I didn’t want to admit to myself that I
there wouldn’t be any new result. And so the men
had been dumb. In Feynman’s example, the
in charge of programs at NAL are so anxious for
physicists seem to have been motivated by the
new results, in order to get more money to keep
fear of looking like idiots for straying too far
the thing going for public relations purposes,
from the conclusion of one of the day’s great
they are destroying - possibly - the value of the
scientists. Other ways researchers might be
experiments themselves, which is the whole
motivated to find untrue ‘truths’ might be the
purpose of the thing. It is often hard for the
pressure an academic department faces to
experimenters there to complete their work as
publish new and exciting findings, as in
their scientific integrity demands.”
Feynman’s example of the situation he
encountered at Cornell’s psychology As Charlie Munger has said, “Show me the
department:2 incentive and I’ll show you the outcome.”
“When I was at Cornell, I often talked to the
people in the psychology department. One of the
Cargo cult finance
students told me she wanted to do an experiment Anthropologists have developed the concept of
that went something like this - it had been found “cargo cults” to describe the phenomena of
by others that under certain circumstances, X, pre-industrial cultures briefly encountering
rats did something, A. She was curious as to more technologically advanced ones, and then
whether, if she changed the circumstances to Y, trying to replicate the wealth of their departed
they would still do A. So her proposal was to do visitors by replicating their material trappings.
the experiment under circumstances Y and see if For example, in some of the South Sea islands
they still did A. Americans used as landing bases during WW2,
I explained to her that it was necessary first to the natives built mock runways,
repeat in her laboratory the experiment of the communication towers, a wooden man for a
other person - to do it under condition X to see if controller, complete with wooden headphones.
she could also get result A, and then change to Y They thought that if they faithfully imitated the
and see if A changed. Then she would know the form of the Americans’ technology they would
real difference was the thing she thought she had coax more cargo drops from the sky.
under control. Feynman used the term Cargo Cult Science to
She was very delighted with this new idea, and refer to a form of pseudoscience, specifically
went to her professor. And his reply was, no, you about those mimicking the form of science
(ostensibly careful experiments, equations,
2
Gregory Zuckerman, The Man Who Solved The Market, Penguin, Random House, UK, 2019
02
19TH DECEMBER 2019 | WWW.CALDERWOODCAPITAL.COM
technical language), but not its substance cherry picking, p-hacking and hypothesising
(data openness, calculation sharing, code after results are known (HARKing) has been
auditing, and basically bending over cited too.4
backwards to show where you might be wrong
So there seems to be something fundamental
in conception, experimentation, simulation
to research, beyond the domain of that
etc). To anyone reading his lecture today, his
research, which pushes it to ‘over-claim’
description of Cargo Cult Scientists reads like a
knowledge. Feynman illustrated it in the
prophetic warning of the “replication crisis”
physical sciences over forty years ago. Fiona
currently sweeping through academia.
Fidler is drawing attention to it in the social
For those not familiar with this crisis, it’s what sciences today. Why should financial research
it sounds like: an embarrassingly large number be any different? Where might we expect to
of academic findings in esteemed journals from find a replication crisis in finance?
various disciplines don’t bear up to scrutiny. In
Well, we’d have to look for a corner of the
some cases this means that there have been no
market with a large incentive to push spurious
attempts to reproduce results and in others,
conclusions while at the same time making the
attempts to replicate findings have failed.
work appear rigorous and dare we say it,
Chart 1, constructed using one presented by
‘scientific’. Perhaps a corner of the industry
Fiona Fidler and colleagues3 suggests that as
which stood to meaningfully gain from selling
few as 40% of findings in the psychology
its own repackaged version of such academic
literature can be independently replicated
looking research without actually consuming
(suggesting the other 60% might be made up).
those products itself. Perhaps that industry
Cargo Cult Science seems to be alive and well.
might use the cover of calling itself “smart”, to
conjure up an image of being technologically
Chart 1
advanced, like a “smart phone” or a “smart
Percentage of results which were
successfully replicated watch”. Maybe that industry would call itself
80%
something like “smart beta”. But maybe we
70% should think of it more as Cargo Cult Finance.
60%
Smart beta vendors all have two things in
50%
common. The first is that they flaunt their ties
40% to academic finance. Dimensional Advisors,
30% whose website touts its success in “putting
20% financial science to work for investors” and
10% which is advised by none other than Ken
French, one of factor investing’s founding
0%
Economics Psychology Science & Nature fathers is far from atypical. Blackrock’s factor
(n=18) (n=151) (n=21)
team is run by a former Prof at Columbia
Source: Fiona Fiddler et al Business School. The second is that the smart
beta “providers” are typically asset gatherers.
The causes of this crisis have been attributed to In other words, firms that have made their
the pressure to publish and the bias towards money selling smart beta products to others,
publishing something novel (only 0.1% of not by investing in smart beta themselves.
psychological papers attempted to replicate the
No wonder. When McLean & Pontiff (2015)5
work of others) much as Feynman described in
looked into the claimed returns of 85 published
his address. This may in part be because of the
“anomalies” documented in a range of
way that research is currently funded, but a
financial, economics and accounting journals
culture which permits and even encourages
(they started out with 97 but found that they
Questionable Research Techniques (QRTs) like
3
https://www.youtube.com/watch?time_continue=993&v=26rznwTa3cE&feature=emb_logo
4
At least there seems now to be an acknowledgement of the problem, as well as various attempts to root the problem
out. The University of Melbourne’s RepliCATS project for example, is leading with an attempt to estimate the replicability
of 3,000 published claims in social science, including economics, psychology, education and sociology (the CATS in
RepliCATS stands for Collaborative Assessment for Trustworthy Science)
5
See “Does academic research destroy stock return predictability?” By David McLean and Jeffrey Pontiff, Journal of
Finance, October 2015
03
19TH DECEMBER 2019 | WWW.CALDERWOODCAPITAL.COM
05
19TH DECEMBER 2019 | WWW.CALDERWOODCAPITAL.COM
been more US shale-oil producer bankruptcies which, like the rest of the energy complex, has
by August of this year than there had been been in one of its worst slumps in history.
throughout 2018, even with modestly higher
So politicians are selling, industry tycoons are
crude prices in 2019. And in the stock market,
buying. Which group would you rather invest
energy looks like it will close as the worst
along side?
performing sector this year, its current 4%
S&P500 weighting comparing to 13% ten years Offshore oil and the jack-up
ago.
opportunity
Worse, the world seems determined to wean
itself off its products. According to Reuters, So let’s recap: an industry in one of the worst
over one thousand funds with nearly $9tr in downturns in its history; blood in the streets;
assets have now committed to divest oil and dumb money selling; smart money buying.
gas holdings, including Norway’s sovereign This is very obviously an alpha-rich
wealth fund, New York’s public pension fund environment. But the ‘energy sector’ that we’ve
and the Rockefeller heirs. In October, 300 been talking about until now is very broad, and
British MPs called for the parliamentary not a particularly useful term. So let’s drill
pension fund to rid itself of all fossil fuel assets. down (.. sorry) into a more specific area. In
particular, I want to focus on the off-shore
Politicians vs billionaires services sector, and a good way to look at that
is through the jack-up market.
But if politicians and large institutional fund
managers are saying loudly and clearly that Jack-ups are rigs which operate in shallow
hydrocarbons don’t have a future, some waters (typically less than 120m, but some can
self-made billionaires and entrepreneurs go as deep as 190m). They are usually pulled
disagree. In May of this year, when Warren into position on the site by tugs, before
Buffett funded $10bn towards Occidental’s bid planting their legs on the sea-bed, jacking up
for Anadarko he told CNBC that it was “a bet their hulls to above the sea surface and from
on the fact that the Permian Basin is what it is here, sinking their drills.
cracked up to be.” The important thing to understand about
Sam Zell, who sold his commercial property jack-ups is where they fit into the hierarchy of
assets to Blackstone in 2007, just before the risk when it comes to oil projects. The riskiest
market collapsed, is buying oil assets in are the biggest and most capital intensive
California, Colorado and Texas in a JV with (typically) deepwater projects, which can
Colony Capital founder Tom Barrack. As he require tens of billions of upfront capital
told Bloomberg TV in an interview a few weeks without any payback for four or five years. In
ago, his investment vehicle, Equity Group contrast, more shallow water projects,
Investments has a history of investing in especially those which drill in, or close to
“dislocated situations” in which there is a existing fields are some of the least risky: the
shortage of capital: infrastructure is already in place, the waters
are known and the cashflow can be expected
“I compared it recently to the real estate industry within a year or so of project commencement.
in the early 1990s, where you had empty Jack-ups, in other words, tend to be associated
buildings all over the place, and nobody had cash with the less risky offshore projects.
to play. That’s very much what we’re seeing
But how healthy is the market for offshore oil
today.”
projects? One of the stranger effects of the
Closer to home, legendary Norwegian tycoon surge in shale production is the assumption
John Fredriksen - Norway’s (actually, now that it is low cost output. Commentators see
Cyprus’) richest man with a net worth the low market prices caused by the increase in
estimated at $10bn earned mainly by shale output and assume it means that the
aggressively and opportunistically trading shale output has a lower cost of production.
boom-bust cycles in the shipping and drilling But the two things are not the same. True,
industries – recently formed a dedicated North American shale gas is cheap to produce,
investment company to buy off-shore rigs and seems to be sustainable at around
06
19TH DECEMBER 2019 | WWW.CALDERWOODCAPITAL.COM
$3/mBtu. But onshore shale oil isn’t. Once you We’ve already seen that shallow water projects
account for the amount of fluids you have to are much less risky. So in an industry which
truck back and forth, and the short life span of remains emotionally fragile, low on confidence
each well, you end up with North American and generally cautious, one would expect the
shale oil breakevens of around $60/barrel. shallow water market to revive sooner than the
Which isn’t too dissimilar to the breakevens deep water market.
you see in offshore projects.
Indeed, this seems to be the case. According to
Borr Drilling (the Norwegian driller founded
Chart 1 by Tor Olav Trøim, the former right hand of
Offshore oil is easily economic John Fredriksen) the market for modern
at current prices jack-up rigs is already strong. The utilisation of
30 100%
the world’s premium jack-up capacity has
25
already reached 90% with day rates running at
80%
around $100kpd (compared to $50k during
Resources
the late 2017 nadir and nearly $250k during
# economic barrels (bn)
20
% of total
60% the 2007 boom). To some degree too,
% of Total
20% Chart 3
5
Jack-up supply is now contracting
0 0%
5 0 5 0 5 0 5 0 5
<30 30-3 35-4 40-4 45-5 50-5 55-6 60-6 65-7 70-7
550
price range of oil per barrel ($)
thought.
Source: IHS Petrodata
Chart 2
The offshore project pipeline So it feels like everything is beginning to fall
is healthy into place - decent offshore pipeline, capacity
120
utilisation rising, supply contracting, smart
money moving in … You might think that
90 shares in the companies exposed to that
market would be in a sweet spot. But you’d be
wrong.
# projects
60
30
The jack-up melt-up?
To understand why, we have to understand
0
that even if the jack-up market is plugged into
the lower-risk part of oil producers’ capex
2009
2010
2011
2012
2013
2014
2015
2016
2017
2018
2019
2020
2021
07
19TH DECEMBER 2019 | WWW.CALDERWOODCAPITAL.COM
$bn
oil price, so the drillers are like a second 30
is tight; and do badly when it isn’t (like now). Source: Bloomberg Debt Equity
consolidation. 10
$bn
0
3 3 14 5 6 6 17 8 9 9
r-1 c-1 p- n-1 r-1 c-1 p- n-1 r-1 c-1
Ma De Se Ju Ma De Se Ju Ma De
Source: Bloomberg
1
“U.S. Drilling Slowdown Triggers Oil Bankruptcy” The Wall Street Journal, July 1st, 2019
08
19TH DECEMBER 2019 | WWW.CALDERWOODCAPITAL.COM
09
19TH DECEMBER 2019 | WWW.CALDERWOODCAPITAL.COM
After poking around the grubby world of Cargo We’ll get to some of that in a moment. But let’s
Cult Finance it’s nice to spend some time in the get a few priors out of the way before we do.
company of real financial scientists, by which I The first is that while The Man Who Solved the
mean of course, Jim Simons and the team he Market is a compelling read (I think “page
assembled at Renaissance Technologies turner” is the appropriate phrase) don’t expect
(Rentec). I doubt anyone reading this doesn’t to come away thinking you’ve found the “secret
know who Jim Simons and Rentec are so I’ll sauce”. Rentec are incredibly (and
skip the detailed introduction. Nevertheless, understandably) secretive and Simons himself
since it’s important we’re all on the same page, only reluctantly agreed to talk to Zuckerman
and that we fully grasp the enormity of the when he realised that the book was going to be
returns achieved by his team, the following written with or without him. I couldn’t help
chart lays it out quite clearly. wondering how much more there really was to
the Rentec story.
Chart 1 The second, and related point to make, is that
Comparing the greats: annualised notwithstanding its title I didn’t find the book
annual lifetime investment returns so much about Jim Simons ‘the man’ as about
40% Rentec ‘the firm’. Simons’ pre-Rentec existence,
for example, is covered in just two chapters.
35%
Yet during this stage of his life he proved
30% himself a prodigious mathematical talent,
25% winning the prestigious Oswald Veblen Prize in
Geometry aged just 38, and developing with
20%
Shiing-Shen Chern what is now known as the
15% Simons-Chern form, which would prove
10% central to the development of string theory,
5%
among other things. He was a code breaker at
the Institute for Defence Analysis (IDA) where
0%
s tt
he assembled and ran a team of
on ) ros n
he ) nc
h
ffe )
Sim-2019 e
g -
So 000)
2
Co 2003
t e r Ly 990)
1 n Bu 2018 mathematicians and scientists to crack Soviet
Jim1988 or 69 92
-
Pe 1977
-
rre 65-
( Ge (19 (19 ( Wa (19 codes for four years, before running foul of his
boss for speaking out against the war in
Source: Zuckerman 2019 Vietnam. And he set up an all-star team of
mathematicians at Stony Brook, turning a
hitherto run-of-the-mill state college into one
As can be seen, Rentec’s performance beat that of the most prestigious mathematics
of Warren Buffett and George Soros hands departments in America.
down. Rentec outperformed Steve Cohen and
Peter Lynch too even though those investors’ So don’t read this expecting Zuckerman to
record covers a much longer time horizon. have really gotten inside Simons’ head in the
These eye-popping returns - 39.1% net over a way that, for example, Alice Schroeder did
thirty year period - don’t tell the whole story. with Warren Buffett in The Snowball. To be
On a gross (pre-fee) basis, Medallion fund clear, I’m only saying this as an observation,
returns averaged a scarcely believable 66.1%. not a criticism. Schroeder got several years of
How did he do this? How is it even possible? access to Buffett himself, his inner circle, his
And what, if anything, can mere mortals like us family, and was even granted time in his office
learn from it? watching him work. Zuckerman got a few
10
19TH DECEMBER 2019 | WWW.CALDERWOODCAPITAL.COM
grudgingly agreed hours. But he isn’t trying to Rentec ends in a cringeworthy confrontation at
be Robert Caro, and it works incredibly well a charity poker match, what he’s hoping for is
indeed as a more narrowly focussed financial his chance to rebuild bridges with his old pals,
story. to shake hands and move on. What he gets is a
scowling Rebekah Mercer (Bob’s daughter)
So with that out of the way, I’m not going to go
“You’re pond scum … get out of here!”
through the whole Rentec story because I don’t
want to spoil it for you. But I will say that I But having such a cast of characters isn’t what
have followed the firm for some time now and makes the firm what it is. Rentec is
that Rob Crenian, my partner in the Capital fundamentally not like other financial firms.
Management business we’re aiming to launch Indeed, it sees itself as a science company,
next year is a Rentec alumni. So before sitting hiring pretty much exclusively hard scientists:
down to read the book, I felt I already knew a mathematicians, physicists, astrophysicists,
lot of the story. computer scientists. But not just any physicists,
astrophysicists and computer scientists, the
But it turns out I didn’t. From the firm’s
very top people in their respective fields,
beginnings as a traditional macro/CTA fund
including Nobel Prize winners. The MBAs,
with Simons at the head, punting around on
financial economists and investment theorists
the gold price with Lenny Baum and James Ax
who populate the industry at large (ie people
(as Barry Ritholz exclaimed in his fascinating
like us, the triers) are typically given a wide
podcast interview with Zuckerman1, “He was a
berth. Is there anything people like us can
macro tourist!”) to the state-of-the-art de facto
learn? I think there is.
signal processing unit it seems to be today,
there were twists and turns, successes and The first is that it’s not enough to be good at
failures, dramas and heartaches along the way what you do. Being smart, focused and
which I’d expect most people wouldn’t be hard-working is a necessary but insufficient
aware of (eg accidentally cornering the potato condition. You also have to be different. In the
market). 1980s Sandor Straus gathered, cleaned and
stored huge (generally but not exclusively)
There are some incredibly colourful characters
financial data sets from libraries around the
too. For example, Bob Mercer, the computer
world to store on Rentec servers. This was
scientist who is as brilliant as he is socially
before Bloomberg existed, or Reuters sold its
awkward, and perhaps now more famous for
data to the industry. They didn’t even pay
his role in bringing Donald Trump to the US
much for it because the value of data wasn’t yet
Presidency, features throughout. It’s difficult
appreciated. As part of this effort, Rentec quite
not to be mesmerised, as Zuckerman clearly is,
possibly created and maintained the world’s
by the guy’s apparently bizarre contrast of
first tick-database, to which they then applied
hyper-rationality (when it comes to data
their data science and big-data skills long
processing), but oddly motivated reasoning
before Google and Facebook.
(when it comes to niche politics and conspiracy
theories). Peter Brown features heavily too. As Rentec continued to find its feet in the
First working with Mercer at IBM and almost 1990s, the prevailing investment wisdom was
subliminally in-sync with him, despite being that since trading was costly, the number of
more outgoing and ‘normal’. Together, they trades should be minimised. Excess returns
make some of the firm’s key breakthroughs in would only accrue to long-term investors who
equity markets. Or David Magerman, the kept portfolio turnover low. But Rentec
lost-soul computer scientist who never quite thought differently. Henry Laufer pointed out
finds a home at Rentec, despite his that from a data-science perspective, long term
accomplished work with Mercer and Brown. signals were too few to be properly and reliably
He becomes so outraged by Mercer’s judged. The shorter-term the strategy, the
involvement in Trump’s politics that he larger would be the sample size of signals
commits the cardinal sin of going public on the which was statistically robust enough to be
firm with a letter to the WSJ. When his time at convincingly tested.
1
https://www.bloomberg.com/news/audio/2019-10-30/gregory-zuckerman-on-the-quant-revolution-podcast
11
19TH DECEMBER 2019 | WWW.CALDERWOODCAPITAL.COM
The second is that being different isn’t an easy comparison to many financial economists who
sell. After ten years in the hedge fund business will only accept statistical correlations which
Simons was running only $43m and was make intuitive sense on the very reasonable
struggling to raise external capital, so different grounds that signals which don’t make
was his approach. Luminaries like Donald intuitive sense are more likely to be illusory.
Sussman and Paul Tudor Jones passed on For example, between 1999 and 2009 the
making an allocation because it was unproven. number of people in the US who drowned by
To be fair, Rentec’s futures trading business falling into a pool each year correlates very
wasn’t spectacular, and they spent years trying well with the number of films Nicolas Cage
to crack equity markets (they knew this was stars in. Most people would say that this
where the capacity was, and where the correlation is obviously spurious, and building
potential for much larger returns would be) a model to predict pool deaths using Nicolas
before figuring it out in the late 1990s. But Cage appearances wouldn’t be a good way to
when Simons told Sussman that he’d finally predict annual pool deaths. Requiring some
cracked it, Sussman passed again. Success kind of intuitive relationship between data
didn’t come quickly. series isn’t a dumb way of avoiding falling into
the trap of depending on a spurious
The third is that alpha is a zero-sum game, and
correlation.
that alpha strategies stop working the moment
you go around telling everyone about them It’s just that Rentec have figured out much
(actually, it’s not so much a lesson as a smarter ways. Their techniques have not only
master-class illustration). We already saw in allowed them to get comfortable backing
the section on Cargo Cult Finance that as soon correlations others wouldn’t trade even if they
as academics published papers documenting could find them, but also ensured that their
“pricing anomalies” those “anomalies” most profitable and reliable signals are so
basically disappeared. In the same section, I precisely because they make no sense. If they
also mentioned that when Rentec were trying made sense, others would trade them. In other
to crack equity markets they spent several words, they’re literally operating on a level
months trawling through the academic which is completely inaccessible to most of us.
literature on financial anomalies for
Even the cleverest alpha strategies are scale
hypotheses to test, before reaching the
constrained however, and Rentec are precise
conclusion that nearly all such studies were
about just how much profit they can squeeze
worthless. Indeed, it’s interesting to contrast
out of a signal over a given amount of time. By
the approach of smart beta “providers” who
Simons’ own estimation, they don’t think
openly publish their findings in academic
they’re the best at trading, but they are “the
journals, trumpet their work through regular
best at estimating the cost of a trade”.2
appearances on the conference circuit and hire
business school professors as “senior Beyond these financial insights the Rentec
consultants” with that of Rentec, who are story provides food for thought in the wider
highly secretive, tie their staff into tight domain of money and society which is
non-compete and confidentiality terms and especially pertinent today. For example, is it
compete for talent with Ivy League level right that a financial fund removes world-class
academic science departments, not investment scientists from their noble pursuit of eternal
banks. When Rentec find anomalies they don’t scientific truths and places them instead in the
go around telling everyone about them. pursuit of transient financial ones?
But it goes deeper still. Consider that as Rentec What scientific knowledge has been lost to
evolved their systems they eventually gave up society, sacrificed for the creation of a few
trying to make intuitive sense of their signals. more billionaires who haven’t even made their
If the science was sound, and the signal money by benefitting customers like pension
deemed valid and robust they were happy to funds or endowments, but by benefitting
allocate capital to it. This is an interesting themselves (the Medallion fund today is
2
Gregory Zuckerman, The Man Who Solved the Market, Penguin, Random House, UK, 2019, pg 349
12
19TH DECEMBER 2019 | WWW.CALDERWOODCAPITAL.COM
13
19TH DECEMBER 2019 | WWW.CALDERWOODCAPITAL.COM
14
19TH DECEMBER 2019 | WWW.CALDERWOODCAPITAL.COM
No liability whatsoever is accepted for any loss arising whether directly or indirectly as a result of the recipient
acting on such information or opinion or estimates. Calderwood Capital Research employees may have or take
positions in the markets, securities or other investments mentioned in this document. The stated price of any
securities mentioned herein is not a representation that any transaction can be effected at this price. Investing
entails risks. The investments referred to are not suitable for all investors and should not be relied upon in
substitution for the exercise of independent judgment. This material is not directed at you if Calderwood
Capital Research is prohibited or restricted by any legislation or regulation in any jurisdiction from making it
available to you. Calderwood Capital Research and its analysts are remunerated for providing independent
investment research to financial institutions, corporations, and governments.
DISCLOSURE: This Report is not to be copied, forwarded or otherwise disseminated to non-subscribers in
electronic or physical form without prior consent.
15