Notes from the Diamond #1: Always Something New to Learn

Yazstremski waits for the bounce.

Don’t be afraid to take advice. There’s always something new to learn. — Babe Ruth

Trivia Question 1 of 108 — What baseball Hall of Fame catcher earned valedictorian honors while also posting a 75-3 record as a pitcher in high school? Answer furnished in main text. Ditto for an answer to the question, “Why 108?”

Boston Red Sox pitcher Brandon Workman at bat in the 9th inning of Game 3 of the 2013 World Series

The Wind Up.  Big differences in their physical demands aside, playing pro baseball and managing money for a living have much in common — a happy fact for those of us who find both endeavors engrossing and a godsend for money managers whose quarterly letters would be intolerably brief or dull absent baseball-related arcana.  Truth be told, the literature exploring parallels between baseball and investing is already so vast that Epsilon Theory (ET) faithful might reasonably pose the same question to ET management that Red Sox Faithful shouted at their TVs as the worst managerial miscue in living memory was unfolding before their eyes several years ago: “Why?!”  No, I’m not referring to Bosox manager Grady Little’s catastrophic act of omission in Game 7 of the 2003 American League Championship Series — acceding to star pitcher Pedro Martinez’s pleas that he continue pitching — but rather to the even dopier decision of Bosox manager John Farrell a decade later: letting pitcher Brandon Workman take his first-ever major league at-bat with the Sox and Cardinals tied 4-4 in the ninth inning of Game 3 of the 2013 World Series.  The Sox lost both games, of course, ending their season in 2003 and adding unnecessary angst to a stress-filled but ultimately triumphant season in 2013.  In due course, this series will explore both of the miscues just referenced plus other noteworthy hits, runs and errors in both money management and baseball, all with the aim of elevating readers’ investment games if not also their appreciation of America’s national pastime.[1]

Unwelcome Change.  I know, I know: in many folks’ eyes, football supplanted baseball as the national pastime long ago — a mutation as regrettable and seemingly irreversible as the shift toward extremism in American politics that Ben Hunt discusses so penetratingly in his multi-part note entitled Things Fall Apart.  Unable as I am to trump Ben (pun intended) in political punditry, I’ll generally avoid politics in this series, leaving it to Ben and other ET contributors to draw parallels if and as they see fit between the shifting fortunes of professional sports on the one hand and political factions on the other.  That said, I can’t resist quoting here the late political journalist Mary McGrory’s lament respecting mutually reinforcing trends she espied in the nation’s political and recreational proclivities long before POTUS 45 declined an invitation to throw out the ceremonial first pitch on Opening Day during his first year in office: “Baseball is what we were,” McGrory observed.  “Football is what we’ve become.”

The Pitch.  Shifting from wind up to pitch … ET faithful deserve an answer to this important question among others: why should they allocate a portion of their scarcest resource — time — to this note or indeed any of the 108 planned and presumably weekly missives comprising the series hereby commencing?  At least three reasons for doing so come to mind.  First, Babe Ruth had it right: there’s always something new to learn about any field of human endeavor, including the fun fact that, as the accompanying diagram confirms, baseballs have precisely 108 stitches.  Second, Ben Hunt has it right: sometimes the best way to replace bad habits with good ones in a chosen field is to look outside it for wisdom or inspiration — as Ben has done so effectively and entertainingly for us money management types with his Notes from the Field. Third, ET contributor par excellence Rusty Guinn has it right also: sometimes the best means of elevating one’s game is to take it on the road so to speak — to contemplate the origins and soundness of habits and beliefs outside one’s chosen profession or political persuasion with an eye toward assessing critically what Rusty refers to as an investor’s “priors”.  We all have ‘em, like it or not.

Anatomy of a baseball

No Guarantees.  I’ve put priors in quotes because I myself have never used that term in decades of writing about investing, nor do I expect to use jargon like it often if ever in this series.  But Rusty fancies the term; I like and respect Rusty (and Ben); and I’ve already learned much from Rusty’s series entitled Notes from the Road.  I won’t guarantee that readers will find these Notes from the Diamond comparably insightfulBut I will pledge that they’ll spawn chuckles on occasion, while also avoiding quotes from an overexposed baseball legend who’s understandably but unjustly remembered more for his malapropisms than for his central role in notching ten World Series titles for the Evil Empire (a/k/a New York Yankees).  After all, why subject readers to deja vus from Yogi Berra when the supply of edifying utterances from other baseballers is large and growing?

Superficial Stasis.  As skilled as Berra was behind the plate, the high school valedictorian referenced in the trivia question at page 1 was even more so, as well as a gifted philosopher in his own right.  Responding to a dinner companion’s jibe that the game he played for a living was intolerably slow, Hall of Fame catcher Johnny Bench intoned, “Baseball is a slow game — for slow minds.”  Rightly understood, investing as distinct from trading also entails prolonged periods of superficial stasis — superficial because effective investors must and do ponder more or less continuously whether newly arriving information necessitates portfolio changes, mindful that it seldom does.  Interestingly, the principle just flagged — favor inaction over action unless the latter is truly vital — is arguably the single most impactful insight spawned by the so-called Sabermetrics Revolution that’s transformed pro baseball in recent decades, i.e., the reshaping of what players, managers and — yes — umpires do or don’t do on the field based on advanced statistics not readily available before certain information technologies were developed.  Among many other insights these Notes will explore, Sabermetrics — a term derived from the acronym for Society for American Baseball Research or SABR — has confirmed decisively what baseball cognoscenti have long conjectured: that a base on balls or walk can be as good as a hit.   Indeed, for reasons to be explored in future notes, the “big data” revolution that’s transformed pro baseball no less than it’s transformed financial markets in recent decades has proven that walks can be better than hits for teams notching them under certain circumstances.

For the Love of It.  What other insights from baseball of potential utility to investors will these Notes explore?  At the risk of having Ben Hunt consign me to his necessarily large nursery of raccoons — i.e., finance types who pilfer Other People’s Money by, among other means, overpromising as habitually as Ted Williams reached base safely[2]  — I’ll answer the question just posed while also wrapping up this inaugural note by providing a sneak peek at insights I plan to explore in the 107+ notes to follow.  I’ve added “+” to 107 because, more than five decades after I first laid eyes on Fenway Park’s gorgeously green grass, and more than three decades after I sank into money management, I’m as intrigued as ever by both baseball and investing.  Whether such intrigue gives me an edge in the latter pursuit is unclear, but I like to think it does, just as I like to think that major leaguers who truly enjoy their work have an edge over those who don’t.  As in finance, which comprises a regrettably large sub-population of raccoons, professional baseball comprises numerous actors motivated primarily by money.  As in finance, it’s long been thus in baseball, as perhaps the edgiest player of all time confirmed when rebuking his fellow pros as his long and distinguished playing career (1905 – 1928) was nearing its end.  “The great trouble with baseball today,” Ty Cobb scolded, “is that most of the players are in the game for the money and that’s all. Not for the love of it, the excitement of it, the thrill of it.”

Coming Attractions.  Thrilling or not, the useful insights derivable by applying ongoing advances in baseball strategies and statistics to money management are legion.  I’m excited by the prospect of pinpointing many of them in future notes, including these:

  • Why it’s not merely useful but essential for professionals to “change their stripes” — a stubbornly enduring no-no in money management whose conscious violators include not a few investment pros as successful in their evolving endeavors as Johnny Bench was in his. Why did Bench switch from pitching to catching at a crucial point in his development as a player?  Because he and those advising him deduced correctly that his foremost physical skill — a strong throwing arm — would be optimally applied as a catcher, thus permitting Bench to use his smarts as well as his physical gifts as frequently as baseball rules permit.  We’ll explore Bench’s metamorphosis and its significance for investment pros in greater detail as this series unfolds.
  • How the metrics used to assess on-field performance condition the behavior of not only players but umpires — a phenomenon with great relevance to client-manager relations in institutional funds management. As we’ll see, the fleet-footed fellow whose most celebrated achievement as a baseballer was his breaking of Ty Cobb’s all-time stolen base record understood intuitively what many capital allocators understand dimly if at all . “Show me a guy who’s afraid to look bad,” said six-time All Star and Hall of Famer Lou Brock, “and I’ll show you a guy who can be beaten.”

Not afraid to look bad: Lou Brock (#20) in action in 1964

  • What practitioners pursuing excellence must do to maintain an edge as the information revolution advances. Quite apart from rules changes already implemented that preclude future career stats as stellar as those achieved by past outliers in each domain — e.g., Wes Crawford or Bob Gibson in baseball; Michael Steinhardt or Peter Lynch in money management — the relentless and mutually reinforcing advances of technology and transparency portend continued shrinkage in the pool of dominantly successful practitioners in professional baseball no less than in professional investing.[3]  By transparency, I mean the timely collection, compilation and dissemination of essentially all available objective data germane to the aforementioned professions.  As many readers are aware, and as future notes will discuss, enhanced transparency as just defined has reduced and will continue undermining the incomes of ballplayers as well as investment pros whose “edges” entail primarily their patrons’ imperfect understanding of their true as distinct from perceived skills.  In a baseball context, “patrons” as just used is defined broadly to include team owners and managers as well as fans — all of whom can easily and inexpensively access meaningfully large chunks of the roughly seven terabytes of data per game (including but by no means not limited to video bits and bytes) that major league baseball’s Statcast system collects via cameras and radar installed in every MLB stadium. That’s a quantum of data equivalent to the contents of 700,000 copies of Webster’s Collegiate Dictionary — and literally millions times the number of data points some of us learned how to record manually on paper scorecards back in the day.

Manual recording of Red Sox labors vs. Yankees 8/18/2006

Continuous Improvement.  Imagine if fiduciaries could evaluate investment pros as quickly, cheaply and thoroughly as baseball managements can evaluate players’ every movement  (or non-movement) using Statcast.  I’m unsure such enhanced scrutiny would produce uniformly better returns, but I’m sure that it would alter managers’ as well as clients’ behaviors, just as such scrutiny has altered how pro baseball gets played, who gets to play it, and for how much.  I’m sure too that even if investment pros’ labors remain as crudely understood as pro baseballers’ were before Statcast came along, future technological advances will compel investment pros seeking sustained excellence to change their stripes on a regular if not continuous basis.  How do I square the assertion just made with Ben’s championing of repeatable processes in Things Fall Apart? I’m not sure I can, or want to, his and Rusty’s invitation to contribute to ET being rooted in their laudable desire to foster diverse viewpoints under ET’s banner.  By my lights, choiceworthy processes in money management display the same cardinal virtue that my all-time favorite player displayed when fielding caroms off Fenway’s fabled Green Monster: such processes are less “repeatable” or static than they are adaptive and ever-changing.  The player in question, of course, was Carl Yazstremski, a Long Island native whose exceptional work ethic arguably made Puritan New England (a/k/a Red Sox Nation) a fitter venue for his sporting labors than his original home turf.  “I loved the game,” Yaz said after his 23-year career came to an end in 1983.  “But I never had any fun.  All hard work, all the time.”

Carl Yazstremski awaiting a carom off Fenway’s Green Monster in the 1967 World Series

Ernie Had It Right.  Like the best opening frame this Bosox fan has ever witnessed — a 50-minute masterpiece in which the Bosox scored 14 runs on 13 hits against the visiting Florida Marlins at Fenway in June 2003 — this initial installment of Notes from the Diamond has developed proportions more expansive than might reasonably have been expected.  As noted at page 1, readers can expect future installments to be shorter — but no less replete with pearls of wisdom from wizards of the diamond, including a gentleman who changed his stripes not once qua young Johnny Bench but multiple times en route to his own induction at Cooperstown.  Nicknamed “Mr. Sunshine” for his upbeat disposition, Ernie Banks (1931 – 2015) is forever known for his catchphrase, “It’s a beautiful day for a ballgame … Let’s play two”.  With so many useful parallels between baseball and investing to be drawn — and with so many members of ET Nation including yours truly wondering what comes next for the business of investing and their own roles within it — I’m keener than ever to craft the next note in this series … and the next.  Let’s play 108, why don’t we?

PDF Download (Paid Membership Required): https://www.epsilontheory.com/download/15639/


Endnotes

[1] Workman’s first and to date only major league at bat went poorly, with a whiff plus two called strikes producing a blindingly quick out.  If the Red Sox, for whom Workman has played on-and-off since 2013, make the World Series in 2018, the odds are good that team manager Alex Cora will call on Workman to do some relief pitching.  That said, I’d bet my family’s most prized baseball-related possession — a ball inscribed for my children by Elden Auker — that Cora doesn’t let Workman bat, ever.  The last living pitcher to have faced Babe Ruth, Auker (1910 – 2006) showed his mettle early in his 10-year major league career: Ruth was the first batter Auker faced in the pitcher’s big league debut in 1933, striking out on just four pitches.

[2] Most readers know that Joe DiMaggio holds the record for consecutive games with a hit: 56 in 1941.  Some may be unaware of another seemingly unbreakable record, held by the best hitter in baseball in the 1940’s or indeed any other epoch: in 1949 Ted Williams reached base safely in an astounding 84 consecutive games.

[3] The all-time career leader in triples with 309, Crawford played before “live era” or post-1910 baseballs made home runs far more frequent than triples.  Gibson notched the all-time best single season earned run average (ERA) of 1.12 in 1968, a year before pitching mounds were lowered by a third to their current height of 10 inches.  Steinhardt made big bucks for himself and his clients during the first half of his career via block trading methods that were either illegal at the time or have since been outlawed.  And Lynch turbocharged his returns via the lawful exploitation of corporate disclosure protocols benefiting big institutions like Fidelity that post-2000 securities law reforms have rendered nugatory, including especially Regulation FD.


On Deck:

What baseball’s steroid era and private equity’s salad days have in common

 


Quantum Supremacy, Correlating Unemployment, and Buddhists with Attitude (by Silly Rabbit)

Quantum supremacy

As Ben and I have discussed before on an Epsilon Theory podcast, my view is that quantum computing is going to be truly, truly transformational by “redefining intractable”, as 1Qbit say, over the coming years. My conviction around quantum continues to grow and — to put a pretty big stake in the ground — I believe, at this point, the only open questions are: Which approach will dominate, and how long exactly until we get quantum machines which work on a broad set of real-world questions? I’ve long been a big fan of the applied, real-world progress D-wave have made, and Rigetti too. However, the “majors” like IBM are also making substantial progress towards true “quantum supremacy” with R&D intensive approaches, while other pieces of the ecosystem, such as the ability to “certify quantum states“, continue to fall into place. In the meantime, here is a wonderful cartoon explainer on quantum computing by Scott Aaronson and Zach Weinersmith.

What web searches correlate to unemployment

Well, in order to get the answer to that question you will have to follow this link (and be prepared to blush). The findings were generated by Seth Stephens-Davidowitz using Google Correlate. “Frequently, the value of Big Data is not its size; it’s that it can offer you new kinds of information to study — information that had never previously been collected”, says Stephens-Davidowitz.

Using verbal and nonverbal behaviors to measure completeness, confidence and accuracy

I recently came across Mitra Capital in Boston who have an interesting strategy of “using verbal indicators to judge the completeness and reliability of messages, to form predictions about company performance (via) analysis of management commentary from quarterly earnings calls and investor conferences based on a proprietary and proven framework with roots in the Central Intelligence Agency” with the underlying tech/methodology based on BIA. They’re running a relatively small fund ($53m AUM in Q1 2017) and have returned an average of 8.5% for the past four years (including a +43% year, and a -12.5% year). Neat NLP approach, although these returns imply more of a “feature than a product” (i.e., a valuable sub-system addition to a larger system, rather than a stand-alone system.) But, hey, I said the same thing about Instagram.

Buddhists with attitude / Backtesting: Methodology with a fragility problem

Probably (hopefully!) anyone reading Epsilon Theory has already read Antifragile by Nassim Nicholas Taleb. Many things which could and have been said about this book, but the most important one to highlight for my narrow, domain application is the massively important distinction (although rarely talked about facet) of machine learning/big compute approaches vs. regression-driven back test approaches. Key distinction is a simple one: Does your system gain from exposure to randomness and stress (within bounds) and improve the longer it exists and the more events it is exposed to OR does it perform less well with stress, and decay with time. Antifragile machine learning systems are profoundly different to the fragile fitting of models.

And finally, since I have already invoked Taleb, and if for no other reason that the line “If someone wonders who are the Stoics I’d say Buddhists with an attitude problem”, here is Taleb’s Commencement address to American University of Beirut last year.

PDF Download (Paid Subscription Required): http://www.epsilontheory.com/download/16037/

Complex Systems, Multiscale Information and Strange Loops (by Silly Rabbit)

Complex systems

Neat and accessible primer on complex systems, multiscale information theory and universality by Yaneer Bar-Yan, and a related paper on the conceptual applications of the same topic: From Big Data To Important Information (suggest start reading from section VII if you read the primer, and from sub-section D on page 13 if you just want the markets application).

Machine learning software creates machine learning software

Lots of buzz about Google’s AutoML announcement at the Google Annual developer conference I/O 2017 last week. AutoML is machine learning software which takes over some of the work of creating machine learning software and, in some cases, came up with designs that rivals or beats the best work of human machine learning experts. MIT Technology Review article on AutoML.

One-shot imitation

Also lots of buzz around one-shot imitation using two neural nets, as demonstrated by OpenAI. Personally, one-shot imitation is the one AI-type concept which gives me the fear. But if Elon’s supporting it then it must be OK… right? One-shot imitation paper here but, more to the point, watch this video and tell me you are not at least a little bit afraid.

The power of the platform

And to the practical applications of technology, I really like the language of this recent press release by Two Sigma CEO, Nobel Gulati, and particularly the paragraph:

Moving forward, durable advantages will to accrue to those building a substantial platform based on massive amounts of data, along with the technology and institutional expertise to use it. Building such a platform requires significant and ongoing investment in R&D, and a fundamentally different culture and mindset to apply a scientific approach to the data-rich world of today.

Personally, I believe that the 2020s will be more defined by big compute than big data but this is, nonetheless, a powerful statement and language, and there’s a key implicit point buried in here on the cultural balance of ‘researchers’ (math and physics natural genii) and ‘production engineers’ (coders who, by nurture, have seen and solved many practical problems). Specifically, how the majority of quant funds have to-date been culturally focused too heavily on the math genius research folks to the detriment of hiring and rewarding the more workmanlike practical folks who can build and maintain a substantial platform which, I agree, is the new durable advantage.

去吧

I was reminded last week by China’s censorship of Google’s latest AlphaGo win against Ke Jie just how substantial a stance it was when Google shut down its Mainland search engine in 2010 and why these kind of bold moves (bets) are essential to developing a truly winning technology company (and also why I don’t live in China anymore!). As Rusty Guinn has written about: A man must have code.

Strange loops

Finally, to bring us back up to the level of self and consciousness, I finally got ‘round to reading Douglas R. Hofstadter’s 2007 book I am a Strange Loop. A long, winding and compelling book summarized by the quote “In the end, we are self-perceiving, self-inventing, locked-in mirages that are little miracles of self-reference.” If you dip in and only read one section, read the section on simmballs in Chapter 3, which loops us back to where we started this column on multiscale information.

PDF Download (Paid Subscription Required): http://www.epsilontheory.com/download/16045/

Mo’ Compute Mo’ Problems (by Silly Rabbit)

Hard problems

Someone tweeted this cartoon at me last week, presumably in angry response to an Epsilon Theory post, as the Tweet was captioned “My feelings towards ‘A.I.’ (and/or machine learning) and investing”:

Source: xkcd

To be clear: YES, I AGREE

Unsurprisingly, we humans are pretty competent creatures within the domains we have contrived (such as finance) and spent decades practicing. So it is, generally, still hard (and expensive) in 2017 to quickly build a machine which is consistently better at even a thin, discrete sliver of a complex, human-contrived domain.

The challenge, as this cartoon humorously alludes to, is that it is currently often difficult (and sometimes impossible) to know in advance just how hard a problem is for a machine to best a human at.

BUT, what we do know is that once an ML/AI-driven machine dominates, it can truly dominate, and it is incredibly rare for humans to gain the upper hand again (although there can be periods of centaur dominance, like the ‘Advanced Chess’ movement).

As a general heuristic, I think we can say that tasks at which machines are now end-to-end better have one or some of the following characteristics:

  • Are fairly simple and discrete tasks which require repetition without error (AUTOMATION)
  • and/or are extremely large in data scale (BIG DATA)
  • and/or have calculation complexity and/or require a great deal of speed (BIG COMPUTE)
  • and where a ‘human in-the-loop’ degrades the system (AUTONOMY)

But equally there are still many things on which machines are currently nowhere close to being able to reach human-parity, mostly involving ‘intuition’, or many, many models with judgment on when to combine or switch between the models.

Will machines eventually dominate all? Probably. When? Not anytime soon.

The key, immediate, practical point is that the current over-polarization of the human-oriented and machine-oriented populations, particularly in the investing world, is both a challenge and an opportunity as each sect is not fully utilizing the capabilities of the other. Good Bloomberg article from a couple of months back on Point72 and BlueMountain’s challenges in reconciling this in an existing environment.

The myth of superhuman AI

On the other side of the spectrum from our afore-referenced Tweeter are those who predict superhuman AIs taking over the world.

I find this to be a very bogus argument in anything like the foreseeable future, reasons for which are very well laid out by Kevin Kelly (of Wired, Whole Earth Review and Hackers’ Conference fame) in this lengthy essay.

The crux of Kelly’s argument:

  • Intelligence is not a single dimension, so “smarter than humans” is a meaningless concept.
  • Humans do not have general purpose minds and neither will AIs.
  • Emulation of human thinking in other media will be constrained by cost.
  • Dimensions of intelligence are not infinite.
  • Intelligences are only one factor in progress.

Key quote:

Instead of a single line, a more accurate model for intelligence is to chart its possibility space. Intelligence is a combinatorial continuum. Multiple nodes, each node a continuum, create complexes of high diversity in high dimensions. Some intelligences may be very complex, with many sub-nodes of thinking. Others may be simpler but more extreme, off in a corner of the space. These complexes we call intelligences might be thought of as symphonies comprising many types of instruments. They vary not only in loudness, but also in pitch, melody, color, tempo, and so on. We could think of them as ecosystem. And in that sense, the different component nodes of thinking are co-dependent and co-created. Human minds are societies of minds, in the words of Marvin Minsky. We run on ecosystems of thinking. We contain multiple species of cognition that do many types of thinking: deduction, induction, symbolic reasoning, emotional intelligence, spacial logic, short-term memory, and long-term memory. The entire nervous system in our gut is also a type of brain with its own mode of cognition.

(BTW: Kevin Kelly has led an amazing life – read his bio here.)

Can’t we just all be friends?

On somewhat more prosaic uses of AI, the New York Times has a nice human-angle on the people whose job is to train AI to do their own jobs. My favorite line from the legal AI trainer: “Mr. Rubins doesn’t think A.I. will put lawyers out of business, but it may change how they work and make money. The less time they need to spend reviewing contracts, the more time they can spend on, say, advisory work or litigation.” Oh, boy!

Valley Grammar

And finally, because it it just really tickles me in a funny-because-it’s-true way: Benedict Evans’ @a16z’s guide to the (Silicon) Valley grammar of IP development and egohood:

  • I am implementing a well-known paradigm.
  • You are taking inspiration.
  • They are rip-off merchants.

So true. So many attorney’s fees. Better rev up that AI litigator.

PDF Download (Paid Subscription Required): http://www.epsilontheory.com/download/16065/

The Horse in Motion

Scared money can’t win and a worried man can’t love.

―  Cormac McCarthy, All the Pretty Horses (1992)

In 1872, noted horseracing aficionado and San Francisco rich guy Leland Stanford (yes, of university fame) commissioned noted photographer and San Francisco smart guy Eadweard Muybridge to apply his path breaking technology of stop-action photography to settle a long-running debate — do all four hooves leave the ground at the same time when horses run? This question had bedeviled the Sport of Kings for ages, and while Stanford favored the “unsupported transit” theory of yes, all four hooves leaving the ground for a split-second in the outstretched position, allowing horses to briefly “fly”, he — as rich guys often do — really, really, really needed to know for sure.

It took Muybridge about 12 years to complete the work, interrupted in part by his murder trial. It seems that Muybridge had taken a young bride (she 21 and he 42 when they married) who preferred the company of a young dandy of a San Francisco drama critic who fashioned himself in faux militaristic fashion as Major Harry Larkyns. After learning that wife Flora’s 7-month old son Florado was perhaps not biologically his, Muybridge tracked Larkyns down and shot him point-blank in the chest with the immortal words, “Good evening, Major, my name is Muybridge and here’s the answer to the letter you sent my wife.” In one of the more prominent early cases of jury nullification (Phillip Glass has an opera, The Photographer, with a libretto based on the court transcripts), Muybridge was found not guilty on the grounds of justifiable homicide despite the judge’s clear instructions to the contrary. Or maybe the jurors were just bought off. Leland Stanford spared no expense in paying for Muybridge’s defense. Gotta get those horse pix.

And eventually he did. Muybridge’s work, The Horse in Motion, settled the question of unsupported transit once and for all.

Yes, all four hooves leave the ground at the same time. But it’s NOT in the outstretched flying position. Instead, it’s in the tucked position, which — because it’s not as romantic a narrative as flying — had never been widely considered as an answer. In fact, for decades after the 1882 publication of The Horse in Motion in book form (a book by Leland Stanford’s fellow rich guy friend, J.D.B. Stillman, who gave ZERO credit to Muybridge for the work … after all, Muybridge was just Stanford’s work-for-hire employee, a member of the gig economy of the 1870s), artists continued to prefer the more narrative-pleasing view of flying horses. Here, for example, is Frederic Remington’s 1889 painting A Dash for the Timber, a work that was largely responsible for catapulting Remington to national prominence, replete with a whole posse of flying horses (h/t to John Batton in Ft. Worth, who knows his Amon Carter Museum collection!).

Okay, Ben, that’s a fun story of technology, art, murder, and rich guy intrigue set in 1870s San Francisco. But what does it have to do with modern markets and investing?

This: Muybridge developed a technology that allowed for a quantum leap forward in how humans perceived the natural world. His findings flew in the face of the popular narrative for how the natural world of biomechanics worked, but they were True nonetheless and led to multiple useful applications over time. Today we are at the dawning of a technology that similarly allows for a quantum leap forward in how humans perceive the world, but with a focus on the social world as opposed to the natural world. Some of these findings will no doubt similarly fly in the face of the popular narrative for how the social world of markets and politics works, but they will similarly lead to useful applications. They already are.

The technology I’m talking about is the biggest revolution in the world today. It’s the ascendancy of non-human intelligences, which I’ve written about in lots of Epsilon Theory notes, from Rise of the Machines to First Known When Lost to Troy Will Burn – the Big Deal about Big Data to The Talented Mr. Ripley to One MILLION Dollars to Two Discoveries. It’s what most of the world calls Artificial Intelligence, which is a term I dislike for its pejorative anthropomorphism. It’s what Neville Crawley calls Big Compute, which is a great phrase, not least for its progression and distinction from the old hat notion of Big Data (h/t to Neville for turning me on to the Muybridge story, too).

The primary impact of Big Compute, or AI or whatever you want to call it, is that it allows for a quantum leap forward in how we humans can perceive the world. Powerful non-human intelligences are the modern day Oracle of Delphi. They can “see” dimensions of the world that human intelligences cannot, and if we can ask the right questions we can share in their vision, as well. The unseen dimensions of the social world that I’m interested in tapping with the help of non-human intelligences are the dimensions of unstructured data, the words and images and communications that comprise the ocean in which the human social animal swims.

This is the goal of the Narrative Machine research project (read about it in The Narrative Machine and American Hustle). That just as Eadweard Muybridge took snapshots of the natural world using his new technology, so do I think it possible to take snapshots of the social world using our new technology. And just as Muybridge’s snapshots gave us novel insights into how the Horse in Motion actually works, as opposed to our romantic vision of how it works, so do I think it likely that these AI snapshots will give us novel insights into how the Market in Motion actually works.

That’s the horse I’m betting on in Epsilon Theory.

PDF Download (Paid Subscription Required): http://www.epsilontheory.com/download/16106/

The Narrative Machine

epsilon-theory-narrative-machine-august-17-2016-clockwork-orange

Alex: There was me, that is Alex, and my three droogs, that is Pete, Georgie, and Dim, and we sat in the Korova Milkbar trying to make up our rassoodocks what to do with the evening. The Korova Milkbar sold milk-plus, milk plus vellocet or synthemesc or drencrom, which is what we were drinking. This would sharpen you up and make you ready for a bit of the old ultra-violence.
“A Clockwork Orange” (1971). Society is a clockwork, with gears constructed of language and guns.

epsilon-theory-narrative-machine-august-17-2016-corbusier

A house is a machine for living in.

Le Corbusier (1887 – 1965), pioneer of modern architecture.

We live our lives inside machines, visible and invisible, tangible and intangible.

epsilon-theory-narrative-machine-august-17-2016-no-mouth-must-scream

HATE. LET ME TELL YOU HOW MUCH I’VE COME TO HATE YOU SINCE I BEGAN TO LIVE. THERE ARE 387.44 MILLION MILES OF PRINTED CIRCUITS IN WAFER THIN LAYERS THAT FILL MY COMPLEX. IF THE WORD HATE WAS ENGRAVED ON EACH NANOANGSTROM OF THOSE HUNDREDS OF MILLIONS OF MILES IT WOULD NOT EQUAL ONE ONE-BILLIONTH OF THE HATE I FEEL FOR HUMANS AT THIS MICRO-INSTANT. HATE. HATE.

― Harlan Ellison, “I Have No Mouth and I Must Scream” (1967). In Ellison’s post-apocalyptic horror, the last five humans on earth live inside a giant omnipotent machine where the only escape is death. It’s The Matrix 30 years before The Matrix was written, and 1,000x nastier.

epsilon-theory-narrative-machine-august-17-2016-norbert-wiener

Mathematics, which most of us see as the most factual of all sciences, constitutes the most colossal metaphor imaginable.

It is easy to make a simple machine which will run toward the light or away from it, and if such machines also contain lights of their own, a number of them together will show complicated forms of social behavior.

Two quotes from Norbert Wiener (1894 – 1964). Wiener received his Ph.D. in mathematics from Harvard at age 17, volunteered to fight in World War I as an enlisted man, but couldn’t get a teaching job at Harvard because he was a Jew. Wiener found a home at MIT, where he became the father of cybernetic theory, aka the mathematics of machine behavior.

epsilon-theory-narrative-machine-august-17-2016-how-economic-machine-works

How does the economy really work?

This simple but not simplistic video by Ray Dalio, Founder of Bridgewater Associates, shows the basic driving forces behind the economy, and explains why economic cycles occur by breaking down concepts such as credit, interest rates, leveraging and deleveraging.

Ray Dalio, “How the Economic Machine Works”. In the three years since Dalio released this short-form film, it has been viewed more than 3 million times.

Machines were the ideal metaphor for the central pornographic fantasy of the nineteenth century, rape followed by gratitude.

Robert Hughes, “The Shock of the New” (1980). A writer’s writer and a critic’s critic. As honest in his self-assessment as his assessment of art and society. It’s a bit uncomfortable, isn’t it? Honesty always is.

epsilon-theory-narrative-machine-august-17-2016-rube-goldberg

Self-operating napkin

Many of the younger generation know my name in a vague way and connect it with grotesque inventions, but don’t believe that I ever existed as a person. They think I am a nonperson, just a name that signifies a tangled web of pipes or wires or strings that suggest machinery.

Rube Goldberg (1883 – 1970)

So, in the interests of survival, they trained themselves to be agreeing machines instead of thinking machines. All their minds had to do was to discover what other people were thinking, and then they thought that, too.

― Kurt Vonnegut, “Breakfast of Champions” (1973). If there’s a better description of modern markets, I have yet to find it. We have become agreeing machines. Because our survival requires it.

epsilon-theory-narrative-machine-august-17-2016-dh-lawrence

For God’s sake, let us be men

not monkeys minding machines

or sitting with our tails curled

while the machine amuses us.

Monkeys with a bland grin on our faces.

D.H. Lawrence (1885 – 1930). Yes. For God’s sake.

epsilon-theory-narrative-machine-august-17-2016-leeuwenhoek

Antonie Van Leeuwenhoek (1632 – 1723), the father of microbiology, alongside a schematic of his microscope and drawings of the “animalcules” he found in a drop of water. Van Leeuwenhoek was a hobbyist lens maker, and he discovered a process for making very small, very high quality glass spheres which provided unparalleled magnification. He never shared his most powerful lenses, nor his manufacturing process, in order to maintain a monopoly on his discoveries. The glass-thread-fusing process died with him and was not rediscovered until 1957, long since supplanted by ground lenses.

epsilon-theory-narrative-machine-august-17-2016-brahe-kepler-war

Copernicus gets all the credit, but his 1543 theory of a heliocentric solar system with circular planetary orbits was a practical dud compared to Ptolemy’s earth-centric theory from 1,400 years earlier. The Copernican model just didn’t work very well. It took better data through new instruments (Tycho Brahe’s observatory) plus better theory through new math (Johannes Kepler’s elliptical orbits) before we finally got it right. But even then, the idea of a heliocentric solar system with elliptical planetary orbits didn’t find popular acceptance until powerful institutions in Northern Europe found it useful to champion this new idea as part of their fight with the Catholic Church and other powerful European institutions.

Modern portfolio theory = Ptolemaic theory. Are powerful institutional investors ready to fight?

Every successful institution, from a marriage to a superhero to a firm to a nation, needs an origin story.

The origin story of arguably the most successful hedge fund institution of the modern world – Bridgewater Associates – is that of Ray Dalio, working out of a small New York apartment in 1975 and publishing a newsletter of “Daily Observations.” The newsletter came first, not the hedge fund, and it was the compelling strength of Dalio’s writings about markets and what he would later term “the Economic Machine” that convinced a few institutional investors to give him some actual capital to invest. The rest, as they say, is history.

In 1975, Dalio struck just the right chord at just the right time with his metaphor of an Economic Machine – the idea that macroeconomic reality across time and place could be understood as a cybernetic system, with rules and principles and behaviors stemming from those rules and principles (essentially, lots and lots of if-then statements and recursive loops, with observable inputs from real-world economic fundamentals). As importantly as being an effective communicator, Dalio was actually right. Bridgewater has translated the metaphor of the Economic Machine into actionable investments for 40 years, with a track record that speaks for itself.

Today I want to propose a new metaphor for the world as it is – a Narrative Machine – where macroeconomic reality is still understood as a cybernetic system, but where the translation of “reality” (all of those economic fundamentals and if-then statements of the Economic Machine) into actual human behaviors and actual investment outcomes takes place within a larger Machine of strategic communication and game playing.

The Narrative Machine isn’t a rejection of the Economic Machine, any more than the theory of relativity rejects Newton’s Laws of Motion. In most places and most times, good old Newtonian physics is all you need to understand the world and take actions to succeed in that world. But there are times and places, like when you’re traveling near the speed of light, where Newtonian physics doesn’t work very well and you need a broader theory – Einsteinian physics – to understand the world and take actions to succeed in that world. A policy-controlled market, like we had in the 1930s and we have again today, is the investment equivalent of traveling near the speed of light. The Economic Machine theory – by which I mean any approach to investing that focuses on tangible macroeconomic fundamentals – just doesn’t work very well in a policy-controlled market. We need an extension of the Economic Machine to succeed in this time and this place, just like the theory of relativity extends Newtonian physics, and that’s what I think the Narrative Machine provides.

Unless you’re an Aristotle or an Einstein, advancement and extension of theory doesn’t just happen by sitting in a room and thinking it up. You need new data. You need better data. You need a new way of looking at the data. Kepler’s idea of elliptical orbits to advance and extend the Copernican theory of a heliocentric solar system couldn’t happen without the new astronomical data provided by Tycho Brahe’s observatory. For a negative example, I think the advancement of germ theory was set back by at least a century because Van Leeuwenhoek refused to share his new technology for looking at microscopic data. But at least astronomy and microbiology have something tangible to look at and measure. How do we SEE the Narrative Machine? How do we observe an invisible network of social interaction? How do we touch the intangible?

For my entire professional career, dating back to my first days as a graduate student and spanning three different vocations and three decades, I’ve been wrestling with that question. I think I caught a small piece of the puzzle with my dissertation and the book that came out of that (Getting to War), and I think that I’ve painted around the edges of the puzzle over the past three years with Epsilon Theory. I was pretty sure that the Narrative Machine was observable if the right Big Data technology could be applied (in the lingo, contextual analysis of affect, meaning, and network connectivity across large pools of unstructured text), but I’ve been involved with Big Data way before anyone called it Big Data, and every time someone claimed to have a solution to this problem it turned out to be far less than meets the eye. On that note, if you enjoy a little dose of schadenfreude (and really, who doesn’t?) do a quick search on Microsoft’s acquisition of Fast Search or, even more shivering, Hewlett Packard’s acquisition of Autonomy, two companies that claimed solutions here. So it was with some trepidation and certainly a healthy skepticism that I started working with Quid, a private company based in San Francisco that has developed a technology for network visualization of unstructured texts.

I think Quid is onto something, in large part because they’re not trying to answer directly the questions I’m asking. Instead, I think they’ve developed a novel process for seeing the invisible world of contextual connections and networks – something analogous to Van Leeuwenhoek’s novel process for seeing the invisible world of microbes – and I’m using their “microscope” to do my own research and answer my own questions. I like that Quid is a tool provider, not a solution provider, so that the analysis here, for better or worse, is my own. On the next few pages I’ll provide an example of some of the research I’m currently doing with the Quid microscope, and I hope it will give you a sense of why I think that we’re getting glimpses of the Narrative Machine with this new instrument.

I’ve written at some length about Brexit and the Narrative that emerged in its immediate aftermath, a Narrative that not only stopped the immediate sell-off in global risk assets in its tracks, but actually reversed the market decline and drove financial asset prices to new highs. To recap, I called Brexit a Bear Stearns event rather than a Lehman event, predicting that creators of Common Knowledge (what game theory calls Missionaries) would successfully characterize the event as an idiosyncratic fluke rather than a systemic risk, exactly as the collapse of Bear Stearns was portrayed in the spring of 2008. In other words, Brexit was NOT a Humpty Dumpty moment, where all the Fed’s horses and all the Fed’s men couldn’t put the egg shell back together again.

Now I have lots of anecdotal evidence of the sort of Narrative creation that I’m hypothesizing here. One of my favorites is a July 13thFinancial Times article titled “Anger at JP Morgan’s ‘Unhelpful’ Brexit Warnings”, where “Senior bankers in London are growing frustrated with JP Morgan Chase’s public warnings that it may cut thousands of jobs in the UK, saying such remarks send an unhelpfully negative message.” Or if I may paraphrase, “The UK government is angry at JP Morgan for not lying about Brexit like they were told to do.” I’ve got a hundred examples like this, examples of a concerted effort by every status quo government and media opinion leader to paint the Brexit vote as a one-off crazy mistake that will probably be reversed and certainly won’t be repeated anywhere else in Europe. But the plural of anecdote is not data, and until now I haven’t an effective instrument to see whether the media data supports what I think is happening.

On the left is a Quid visualization of the clusters and network relationships between the 2,422 Brexit-mentioning articles published by Bloomberg in the 4 weeks prior to the June 23rd vote. On the right is a Quid visualization of the 4,283 such articles published by Bloomberg in the 4 weeks after the vote. This is what the formation of a coherent Narrative looks like. These are snapshots of the Narrative Machine.

epsilon-theory-narrative-machine-august-17-2016-quid-bloomberg-vote

So what are we looking at here? Each dot (or node) represents a single unique article, and the Quid algorithms group nodes into colored clusters based on shared word choice and similar word positioning. If we magnify any of these clusters, in this case a cluster of articles talking about bond-buying and US Treasuries in the pre-vote data, we see that the nodes themselves differ in size according to their connectivity or centrality to the clustering principle, and that there are varying distances and numbers of connections between the nodes, as well. Each node exerts the equivalent of a gravitational pull on every other node, giving the entire structure both the appearance and the substance of a star map. Nodes can be evaluated and displayed on dimensions such as sentiment (green/positive – red/negative), as shown below, and all of these characteristics (distance, connectivity, centrality, etc.) are generated as a structured data set for further, non-visual analysis.

epsilon-theory-narrative-machine-august-17-2016-quid-treasuries

epsilon-theory-narrative-machine-august-17-2016-quid-treasuries-2

Here’s what I think we’re seeing in the “coagulation” of the Bloomberg facet of the Narrative Machine.

epsilon-theory-narrative-machine-august-17-2016-quid-bloomberg-vote

The pre-vote Bloomberg network structure on the left is what a complacent Narrative looks like. The articles are “about” whatever the clustering principle might be, and Brexit is typically a sideways glance, a throwaway line that’s almost always negative in sentiment. On the other hand, the post-vote network structure on the right is what an engaged Narrative looks like, where the articles are “about” Brexit and its impact on the clustering principle. Not only are we seeing a strong Narrative form on the right, but the density of lines and closeness of clusters shows that a similar tone and meaning has taken root across all these clusters. Importantly, it’s a positive tone and meaning that takes shape in the post-vote Narrative, with sentiment scores significantly higher than in the pre-vote snapshot. The sky-will-fall articles are almost all in the pre-vote sample, while the post-vote sample – as early as the Monday after the vote, which is immediately before the market starts to turn – are almost all focused on the non-systemic nature of Brexit, the likelihood of reversal, and the “mistake” that was made here.

The pre- and post-vote evolution of the Brexit Narrative structure is robust within individual Bloomberg clusters and across other major media microphones. Here, for example, is the same bond-buying / US Treasuries cluster in the post-vote Bloomberg data set (different color, but same clustering principle), and in the blow-up you can see how much more coherent and connected it is than the pre-vote cluster.

epsilon-theory-narrative-machine-august-17-2016-treasuries-bloomberg-vote

Below, the top pair of star maps are the 4-week pre-vote and post-vote network visualizations of Brexit-mentioning articles published by Reuters, and the bottom two star maps are samples from all publishers in the Quid database. All of the hypothesized Narrative patterns described above are replicated here.

epsilon-theory-narrative-machine-august-17-2016-reuters-all-sources

Okay, Ben, these diagrams and “star maps” are all very pretty. I get your metaphor of the Narrative Machine, and I get that you’re excited about a new technology that lets you see that invisible machine. But so what? How does all this translate into either actionable investment ideas or a process improvement in managing investment ideas?

When anyone asks this question (and believe me, it’s the question I’ve asked myself in one form or another for 30 years), they’re asking about two things: edge and odds. For anyone who’s trying to beat the dealer (my plug for Edward O. Thorp’s 1962 book that changed everything for me, also retold and expanded in William Poundstone’s brilliant book Fortune’s Formula) … for anyone who’s interested in alpha, this is all that matters: edge and odds. Edge is private information, an insight into the true nature of reality that other game players don’t have. Odds are the probabilistic relationship between risk and reward at any given moment in time. If you have either one of these on your side, then you’ll do well in whatever game you’re playing, if you’re dealt enough hands. If you have both on your side … and I think that a rigorous application of the Narrative Machine generates both edge and an improved assessment of odds … hey, now.

The odds revealed by the Narrative Machine are the odds of a catalyst having a major impact on price (or not). Or in slightly different words, I think that the Narrative Machine can help show us the degree to which future events are “priced-in” by the market. For example, when you’ve got a complacent, all-over-the-place Narrative leading up to a scheduled event like the Brexit vote, then even if my best guess on the voting odds is, say, 60% in favor of “Remain”, I would still place a bet on “Exit” because the Narrative-implied market payoff odds are far better than the breakeven odds of the vote.

The edge that the Narrative Machine generates is an improved reaction to a catalyst once it occurs. To be clear, I don’t think that the Narrative Machine can predict a market shock or catalyst before it happens. It’s not a crystal ball. But it is a real-time window into how the Common Knowledge Game is being constructed and played after an event occurs. For example, when you have a pervasive, systemic-risk-is-off-the-table Narrative created almost immediately following a market shock like the Brexit vote, then I would get long the market even if I believed in my heart-of-hearts (and I do) that there really IS systemic risk posed by everything that’s behind the Brexit vote.

I don’t want to over-sell the degree to which the Narrative Machine has been “weaponized” into an investable alpha source, because there are several critical aspects of network theory that remain to be implemented. Foremost of these is what network theory calls alluvial analysis, or evaluation of how different clusters “flow” into each other and away from each other over time. I’ve included two wonderful illustrations of this concept, both from a 2010 scientific journal article (“Mapping Change in Large Networks” by Martin Rosvall and Carl Bergstrom). I think the Quid technology is pretty good at what network theory calls “significance clustering”, the assignment of individual nodes into similarly colored and positioned groups – essentially a snap shot of the network at a given point in time. What we need now is a map of how those clusters evolve over time, because the meaning or organizing principle of the clusters themselves doesn’t remain constant.

epsilon-theory-narrative-machine-august-17-2016-mapping-change-large-networks-1

Rosvall and Bergstrom illustrate this beautifully in the second diagram here, where a network analysis of scientific journal articles show how neuroscience has become its own “thing” over time. We need the same alluvial maps for market Narrative clusters. I’m on it.

epsilon-theory-narrative-machine-august-17-2016-mapping-change-large-networks-2

So, yes … early days for the Narrative Machine. But, yes … a potential alpha source.

Which leads to an interesting question. If this is a new alpha source – the most valuable thing in the investment world – why am I talking about it? Isn’t this like announcing that you think you’ve found gold in California or the Yukon before you’ve staked a claim?

Good question. There’s some margin of intellectual property safety here because it’s not an easy alpha source to mine, even with cool new technologies like Quid. The internal logic of the Narrative Machine is the logic of strategic interaction (game theory), not the logic of stochastic processes (econometric inference). In plain English, I don’t think you can run a regression analysis of historical media network characteristics against historical market characteristics and get much that will be useful, at least not if you’re after edge and odds. The underlying theory here is Information Theory and the underlying math is the mathematics of entropy, and I’m reasonably confident that we’re not going to see an Excel plug-in for either of those anytime soon.

But yes, someone could “steal” this idea and run with it on their own. To which I say … fine. Better that than being another Van Leeuwenhoek, bogarting his research on his invisible world and setting back the advancement of germ theory and microbiology by a century or more. As in 1648 and 1776 and 1848 and 1917, we live in one of those rare moments in history where ideas are at stake and fundamental theories of the world are in flux. Let’s engage with that, and not hide in the convenient cubbyhole of narrow self-interest or the mentality of an agreeing machine.

We need a new perspective regarding the true nature of our economic and political clockwork, and that’s the real value of the idea of the Narrative Machine.

PDF Download (Paid Subscription Required): http://www.epsilontheory.com/download/16229/

Two Discoveries

The world made two discoveries last week. Everyone is aware of the first discovery – that ISIS is not “a junior varsity team” but an able protagonist in what Pope Francis quite rightly calls “a piecemeal third World War”. Very few are aware of the second discovery – the existence of a polynomial-time algorithm to determine whether two networks, no matter how complex, are identical. Both are watershed events, part of a continuing destabilization of politics and science. Neither will impact markets very much today. Both will change markets forever in the years to come.

I won’t say much about the first discovery here, but will take this opportunity to reprint a note I wrote in December 2014, eerily right before the Charlie Hebdo attack: “The Clash of Civilizations”. I’d also point out that the all-too-predictable Orwellian response to events like the Paris attack, namely to rewrite history and expand government monitoring of our private lives, is in full swing.

For example, here’s a before and after France Inter headline (hat-tip to Epsilon Theory reader M.O.), as noted by The Daily Telegraph. The headline as it originally ran a few weeks back calls a potential terrorist infiltration of Syrian refugees a “fantasy” of the lunatic right. Immediately after the attack, the headline has been rewritten (and the body of the article partially rewritten as well), to suggest that of course one might question whether or not a few terrorists managed to sneak in with the refugees. France Inter – surprise! – is part of the state-owned media apparatus, now in full-throated advocacy for a “pitiless” war.

epsilon-theory-two-discoveries-november-19-2015-media epsilon-theory-two-discoveries-november-19-2015-media-terrorist

Given how this Narrative is developing within left-leaning European governments (hmm, amazingly enough, no marches this time around calling for solidarity with peace-loving Muslims), my political advice to left-leaning US politicians like Connecticut governor Dannel Malloy is that you might be getting a wee bit ahead of yourself by loudly and publicly promoting Syrian refugee relocation in your state. Just sayin’.

The second discovery – an algorithm that dramatically shortens the information processing power required to tell if two networks are structurally the same – requires a bit more explication. Here’s a picture of two such visibly different but structurally identical (isomorphic) networks.

epsilon-theory-two-discoveries-november-19-2015-algorithm

For a whole host of data science applications – from cryptography to genomics to nuclear physics to, yes, finance – we’d often like to know how or if two networks or two data arrays are permutations of the same underlying structure. Maybe you could eyeball an answer to an 8-node network like the example above, but it doesn’t take much imagination to realize that this problem gets very hard very quickly for the human brain as the number of nodes increases.

Fortunately, of course, we have non-human intelligences to help us crack these problems today, but the only known algorithms or programs for computers to follow for this particular problem existed in what is called “non-deterministic polynomial” or NP-time, where the amount of time or information processing power required to carry out the algorithm increases at a staggering rate as the number of nodes increases. This is in contrast to a polynomial or P-time algorithm, where the time required to crunch the algorithm increases at a more manageable rate as the number of nodes increases. Think of it as the difference between 2n (an NP-time algorithm) and n2 (a P-time algorithm), where n is something like 1 million. 2 raised to the 1,000,000th power is an incomprehensibly large number, greater than the number of atoms in the universe. If that’s your algorithm for solving the isomorphic network problem, then it’s physically impossible for any computer, no matter how powerful, to crack the problem for a network with 1 million nodes. On the other hand, 1,000,000 squared is a trivially small number (1 trillion) as a representation of a powerful computer’s information processing capabilities. If that’s your algorithm for solving the isomorphic network problem, then there is no network too large for you to measure and compare to another network.

The isomorphic network problem was a classic example of something that most computer scientists believed could only be solved with NP-time algorithms. But last week, Laszlo Babai at the University of Chicago announced the existence of an algorithm for this class of problems that is, for all practical purposes, in P-time. Why is this important? Because it is the modern day equivalent of discovering a new continent, one that happens to exist in cyberspace rather than human space. Because it is now by no means clear that there are ANY problems of data science that are inexorably lost in the cosmic fog of NP-time algorithms. Why will this one day change markets forever? Because the ability of computers to analyze and predict (and ultimately shape) the behavior of a complex network comprised of millions of semi-autonomous agents exchanging a set of symbolic chips with each other – The Market – just took a giant step forward. If you thought that humans were a marginalized participant in public capital markets today … if you thought that the casino-fication of markets had reached some sort of natural limit … well, you ain’t seen nothing yet.

Sigh. Last week was a tough week for the human team. With the loud explosions out of Paris, the illiberal left, the illiberal right, and the illiberal jihadists are now ALL in political ascendancy. And with the quiet announcement out of Chicago, we are oh-so close to the day when no human communication over any network can be shielded or kept private from a machine intelligence. God help us as the two discoveries merge into one.

PDF Download (Paid Subscription Required): http://www.epsilontheory.com/download/16381/

One MILLION Dollars

epsilon-theory-one-million-dollars-september-15-2015-austin-powers

Dr. Evil: Gentlemen, it has come to my attention that a breakaway Russian Republic called Kreplachistan will be transferring a nuclear warhead to the United Nations in a few days. Here’s the plan. We get the warhead and we hold the world ransom for…one MILLION dollars!
Number Two: Don’t you think we should ask for more than a million dollars? A million dollars isn’t exactly a lot of money these days. Virtucon alone makes over 9 billion dollars a year!
Dr. Evil: Really? That’s a lot of money.

“Austin Powers: International Man of Mystery” (1997)

epsilon-theory-one-million-dollars-september-15-2015-watchmen

Dr. Manhattan: I have walked across the surface of the Sun. I have witnessed events so tiny and so fast they can hardly be said to have occurred at all. But you, Adrian, you’re just a man. The world’s smartest man poses no more threat to me than does its smartest termite.
“Watchmen” (2009)

Eddie Morra: I don’t have delusions of grandeur. I have an actual recipe for grandeur.
“Limitless” (2011)

Carl Van Loon: Have you been talking with anyone?
Eddie Morra: No, I haven’t been talking with anyone, Carl. I’m not stupid.
Carl Van Loon: I know you’re not stupid, Eddie, but don’t make the classic smart person’s mistake of thinking no one’s smarter than you.

“Limitless” (2011)

DIY’s newest frontier is algorithmic trading. Spurred on by their own curiosity and coached by hobbyist groups and online courses, thousands of day-trading tinkerers are writing up their own trading software and turning it loose on the markets. 

Interactive Brokers Group actively solicits at-home algorithmic traders with services to support their transactions. YouTube videos from traders and companies explaining the basics have tens of thousands of views. More than 170,000 people enrolled in a popular online course, “Computational Investing,” taught by Georgia Institute of Technology professor Tucker Balch. Only about 5% completed it.
Wall Street Journal, “Algorithmic Trading: The Play at Home Version” August 9, 2015

London day trader Navinder Sarao has been formally indicted by a U.S. federal grand jury on charges of market manipulation that prosecutors say helped contribute to the 2010 “flash crash,” according to a Sept. 2 court filing made public on Thursday. 

The Justice Department first announced criminal charges against Sarao in April and is seeking to have him extradited to the United States to stand trial. 

Sarao is accused of using an automated trading program to “spoof” markets by generating large sell orders that pushed down prices. He then canceled those trades and bought contracts at lower prices, prosecutors say.
CNBC, “US Federal Grand Jury Indicts ‘Flash Crash’ Trader” September 3, 2015

Anxiety in the industry surged last week after Li Yifei, the prominent China chief of the world’s largest publicly traded hedge fund, disappeared and Bloomberg News reported that she had been taken into custody to assist a police inquiry into market volatility. Her employer, the London-based Man Group, did little to dispel fears, declining to comment on her whereabouts. 

Ms. Li resurfaced on Sunday and denied that she had been detained, saying that she had been in “an industry meeting” and “meditating” at a Taoist retreat. But many in the finance sector are unconvinced.
New York Times, “China’s Response to Stock Plunge Rattles Traders” September 9, 2015

I’ve written several Epsilon Theory notes about modern market structure (“Season of the Glitch”, “Fear and Loathing on the Marketing Trail, 2014”, “The Adaptive Genius of Rigged Markets”, “Hollow Men, Hollow Markets, Hollow World”), all of which have been very well received. I’ve also written several Epsilon Theory notes about Big Data and non-human intelligences (“Troy Will Burn – the Big Deal about Big Data”, “First Known When Lost”, “Rise of the Machines”), all of which have generated a yawn. This divergence in reader reaction has puzzled me, because it seems so obvious to me that the issues are two sides of the same coin. So why can’t I communicate that?

It’s only over the last few days, after listening to old-school luminaries like Leon Cooperman and Dick Grasso rail against systematic investment strategies, index derivative hedging, and algorithmic market making as if they were the same thing (!) … it’s only after reading press stories that praise the US indictment of Navinder Sarao, the London trader who supposedly triggered the “Flash Crash” from his home computer, but condemn the Chinese detention of Man Group’s Li Yifei as if they were different things (!) … it’s only after seeing 500 commercials for “DIY trading platforms” on TV today as if this were a thing at all (!) … that I think I’ve finally figured this out.

We’re all Dr. Evil today, thinking that one million dollars is a lot of money, or that one second is a short period of time, or that we are individually smart or capable in a systemically interesting way. We use our small-number brains to make sense of an increasingly large-number investment world, and as a result both our market fears and our market dreams are increasingly out of touch with reality.

There are a million examples of this phenomenon I could use (including the phrase “a million examples” which, if true, would take me a lifetime to write and you a lifetime to read, even though neither you nor I considered the phrase in that literal context), but here’s a good one. A few months ago I wrote an Epsilon Theory note on the blurry distinction between luck and skill, titled “The Talented Mr. Ripley”, where I pointed out that it was now quite feasible with a few million dollars and a few months to build a perfect putting machine, one that would put every professional human golfer to shame. Judging from the reader emails I received on this, you might have thought I had just said that the world was flat and the sun was a big candle in the sky. “Preposterous!” was the gist of these emails – sometimes said nicely and sometimes (actually, most of the time) not so nicely – as apparently I know nothing about golf nor about the various failed efforts in the past to build a mechanical putting device.

Actually, I know a lot about these mechanical putting devices, and to compare them to the non-human putting intelligences that are constructible today is like comparing Lascaux cave art to HD television. It’s relatively child’s play to build a machine today that can identify and measure the impedance of every single blade of grass between a golf ball and the cup, one that measures elevation shifts of less than the width of an eyelash, one that applies force within an erg tolerance that human skin would interpret as the faintest breeze. That’s what I’m talking about. Do you know how the most advanced surreptitious listening devices, i.e. bugs, operate today? By measuring the vibrations in the glass window of the room where the conversation is taking place and translating those vibrations back into the sound waves that produced them. That’s what I’m talking about. Now replace “blades of grass” with “individual stock trades”. Now replace “conversation” with “investment strategy”. Arthur C. Clarke famously said that any sufficiently advanced technology is indistinguishable from magic. Do you really think we bring to bear less powerful magic in markets with trillions of dollars at stake than we do in spycraft and sports?

And let’s be clear, the machines are here to stay. They’re better at this than we are. The magic is in place because the magic works for the people and institutions that wield the magic, and no amount of rending of garments and gnashing of teeth by the old guard is going to change that. Sure, I can understand why Dick Grasso would suggest that we should go back to a pre-Reg NMS system of human specialists and cozy market making guilds, where trading spreads were measured in eighths and it made sense to pay the CEO of a non-profit exchange $140 million in “retirement benefits.” And I almost sympathize with the nostalgic remembrances of a long list of Hero Investors recently appearing on CNBC, pining for a pre-Reg FD system of entrenched management whispering in the ear of entrenched money managers, where upstart quants knew their place and the high priests of stock picking held undisputed sway. But it ain’t happening.

And let’s also be clear, the gulf between humans and machines is getting wider, not narrower. Even today, one of the popular myths associated with computer science is that non-human intelligences are brute force machines and inferior to humans at tasks like pattern recognition. In truth, a massively parallel processor cluster with in-line memory – something you can access today for less money than a junior analyst’s salary – is far better at pattern recognition than any human. And I mean “far better” in the same way that the sun is far better at electromagnetic radiation than a light bulb. Much has been made about how robot technologies are replacing low-end industrial and service jobs. Okay. Sure … I guess I’d be worried about that if I were working in a Foxconn factory or a Bay Area toll booth. But far more important for anyone reading this note is how non-human intelligences are replacing high-end pattern recognition jobs. Like trading. Or investing. Or asset allocation. Or advising.

The question is not how we “fix” markets by stuffing the technology genie back into the bottle and we somehow return to the halcyon days of yore where, in Lake Wobegon fashion, all of us were above average stock pickers and financial advisors. No, the question we need to ask ourselves is both a lot less heroic and far more realistic. How do we ADAPT to a market jungle where human intelligences are no longer the apex predator?

I’ve got two sets of suggestions, depending on whether you see yourself as a trader or an investor. It’s a lot to digest, so let’s look at traders in the balance of this week’s note and at investors next week.

epsilon-theory-one-million-dollars-september-15-2015-limitless

Every trader who ever lived believes that, like the Bradley Cooper character in “Limitless”, he or she has a recipe for grandeur. It doesn’t matter whether they find that recipe in prices or volumes or volatility or spreads or any other aspect of a security, all traders have an internalized pattern recognition system that they believe gives them a persistent edge. Most of them are wrong.

In modern large-number markets, any trading strategy based on naïve inference is certain to have zero edge, zero alpha. By naïve inference I mean selecting a strategy based solely on the econometric fit of a time series data matrix to some market outcome like price change. It’s a trading strategy that works because … it works. There’s no “why?” answered here, and as a result the strategy is certain to be derivative, non-robust, and quickly arbitraged. Or to put it in slightly different terms, whatever purely inductive trading strategy you think gives you an edge is already being used by thousands of non-human intelligences, and they’re using the strategy far more effectively than you are. To the degree a naïve inference strategy works at all, you’re just tagging along behind the non-human intelligences, picking up their crumbs.

What trading strategies have even a theoretical possibility of edge or alpha? Here are two.

Possibility 1: Find a market niche where your counterparties are non-economic or differently-economic market participants – like an oil futures market where a giant, lumbering integrated oilco seeks to hedge production, or where a sovereign wealth fund looks for inflation protection (Remember those happy days when giant allocators addressed inflation concerns in commodity markets? Me, neither.) – and scalp a few dimes by taking advantage of their very different preference functions. Traders who pursue this type of strategy have a name in biological systems. They’re called parasites. I call them beautiful parasites (see the Epsilon Theory note “Parasite Rex”), because they capture more pure alpha than any strategy I know.

Possibility 2: Find a market niche where you understand the impact of exogenous signals like news reports or policy statements on the behavioral tendencies of other human market participants, in exactly the same way that a good poker player “plays the player” as much as he plays the cards. These market niches tend to be sectors or assets that are driven less by fundamentals than they are by stories – think technology stocks rather than industrials – although here in the Golden Age of the Central Banker it’s hard to find any corner of the capital markets that’s not driven by policy and narrative. The game that these traders have internalized isn’t poker, of course, but is almost always some variant of what modern game theorists call “The Common Knowledge Game”, and what old-school game theorists like John Maynard Keynes called “The Newspaper Beauty Contest”.

What do these two examples of potentially alpha-generating trading strategies have in common? They operate in a world that a non-human intelligence – which is effectively a super-human inference machine – can’t figure out. Today’s effective alpha-generating trading strategies are based on a game (in the technical sense of the word, meaning a strategic interaction between humans where my decisions depend on your decisions, and vice versa) where you can have very different outcomes from one trade to another even if the external/measurable characteristics of the trades are identical. This is the hallmark of games with more than one equilibrium solution, which simply means that there are multiple stable outcomes of the game that can arise from a single matrix of descriptive data. It means that you can’t predict the outcome of a multi-equilibrium game just by knowing the externally visible attributes of the players. It means that the pattern of outcomes can’t be recognized with naïve (or sophisticated) inference techniques. It means that traders who successfully internalize the pattern recognition of strategic behaviors rather than the pattern recognition of time series data have a chance of not just surviving, but thriving in a market jungle niche.

Sigh. Look … I know that this note is going to fall on a lot of deaf ears. It’s an utterly un-heroic vision of what makes for a successful trader in a market dominated by non-human intelligences, as I’m basically saying that you should find some small tidal pool to crawl into rather than roam free like some majestic jungle cat. As such it flies in the face of every bit of heroic advertising that the industry spews forth ad nauseam every day, my personal fave being the “Type-E” commercials with Kevin Spacey shilling for E*Trade. Generalist traders are some of my favorite people in the world. They’re really smart. But they’re not smart enough. None of us are. After all, we’re only human.

epsilon-theory-one-million-dollars-september-15-2015-spacey

PDF Download (Paid Membership Required):

http://www.epsilontheory.com/download/16386/

The Talented Mr. Ripley

The more I practice, the luckier I get.epsilon-theory-the-talented-mr-ripley-april-27-2015-gary-player
– Gary Player (b. 1935)

Luck is the residue of design.
– Branch Rickey (1881 – 1965)

epsilon-theory-the-talented-mr-ripley-april-27-2015-ted-williamsI’ve found that you don’t need to wear a necktie if you can hit.
– Ted Williams (1918 – 2002)

They say that nobody is perfect. Then they say that practice makes perfect. I wish they’d make up their minds.
– Wilt Chamberlain (1936 – 1999)

They say that nobody is perfect. Then they say that practice makes perfect. I wish they’d make up their minds.
– Wilt Chamberlain (1936 – 1999)

It took me 17 years to get 3,000 hits in baseball. I did it in one afternoon on the golf course.
– Hank Aaron (b. 1934)

Talent is cheaper than table salt. What separates the talented individual from the successful one is a lot of hard work.
– Stephen King (b. 1947)

At one time I thought the most important thing was talent. I think now that – the young man or the young woman must possess or teach himself, train himself, in infinite patience, which is to try and to try and to try until it comes right. He must train himself in ruthless intolerance. That is, to throw away anything that is false no matter how much he might love that page or that paragraph. The most important thing is insight, that is … curiosity to wonder, to mull, and to muse why it is that man does what he does. And if you have that, then I don’t think the talent makes much difference, whether you’ve got that or not.
– William Faulkner (1897 – 1962)

Talent is its own expectation, Jim: you either live up to it or it waves a hankie, receding forever.
– David Foster Wallace, “Infinite Jest” (1996)

What is most vile and despicable about money is that it even confers talent. And it will do so until the end of the world.
– Fyodor Dostoyevsky (1821 – 1881)

Talent is a long patience, and originality an effort of will and intense observation.
– Gustave Flaubert (1821 – 1880)

There is nothing more deceptive than an obvious fact.
– Arthur Conan Doyle, “The Boscombe Valley Mystery” (1891)

epsilon-theory-the-talented-mr-ripley-april-27-2015-murder-she-wrote

Sheriff Metzger: Mrs. Fletcher! Can I see you for a minute? [pause] Do me a favor, please, and tell me what goes on in this town!
Jessica Fletcher: I’m sorry, but …
Sheriff Metzger: I’ve been here one year, and this is my fifth murder. What is this, the death capital of Maine? On a per capita basis this place makes the South Bronx look like Sunny Brook farms!
Jessica Fletcher: But I assure you, Sheriff …
Sheriff Metzger: I mean, is that why Tupper quit? He couldn’t take it anymore? Somebody really should’ve warned me, Mrs. Fletcher. Now, perfect strangers coming to Cabot Cove to die? I mean look at this guy! You don’t know him, I don’t know him. He has no ID, we don’t know the first thing about this guy.

– “Murder, She Wrote: Mirror, Mirror, on the Wall: Part 1” (1989)

epsilon-theory-the-talented-mr-ripley-april-27-2015-manchurian-candidate

Dr. Yen Lo: His brain has not only been washed, as they say … It has been dry cleaned.
– “The Manchurian Candidate” (1962)

Dickie Greenleaf: Everyone should have one talent. What’s yours?
Tom Ripley: Forging signatures, telling lies … impersonating practically anybody.
Dickie Greenleaf: That’s three. Nobody should have more than one talent.

– “The Talented Mr. Ripley” (1999)

My singular talent is seeing patterns that others don’t. That’s not a boast, but a fact, and frankly it’s been as much a source of alienation in my life as a source of success. As my father was fond of saying, “You know, Ben, if you’re two steps ahead it’s like you’re one step behind.” I can’t explain how I see the patterns – they just emerge from the fog if I stare long enough. It’s always been that way for me, for as far back as I have memories, and whether I’m 5 years old or 50 years old I’m always left with the same realization: I only see the pattern when I start asking the right question, when I allow myself to be, as Faulkner said, “ruthlessly intolerant” of anything that proves false under patient and curious observation.

For example, I think the wrong question for anyone watching “Murder, She Wrote” is: whodunit? The right question is: how does Jessica Fletcher get away with murder this time? Once you recognize that it’s a Bayesian certainty that the woman is a serial killer, that she controls the narrative of Cabot Cove (both figuratively as a crime novelist and literally as a crime investigator) and thus the behavior of everyone around her, you will discover a new appreciation for both the subliminal drivers of the show’s popularity as well as the acting genius of Angela Lansbury. Seriously, go back and watch the original “Manchurian Candidate” and focus on Lansbury. She’s a revelation.

Or take the Masters tournament earlier this month. I was lucky enough to attend Wednesday’s practice round, and I was sitting in a shady spot on the 10th green watching the players come by and try their luck at 15 foot putts. At first, like the other spectators, my question was: how are they such good putters? This was “the obvious fact,” to quote Sherlock Holmes, and I watched for any clues that I could adopt for my laughable game – a forward tilt of the wrist, a stance adjustment … anything, really. We all watched carefully and we all dutifully oohed and aahed when the ball occasionally dropped in the cup. But suddenly, a new pattern emerged from the fog, and I realized that we were all asking the wrong question. Instead, I started to ask myself, why are they such poor putters? 

Now I realize that I just alienated at least half of the reading audience, but bear with me. I’m not saying that professional golfers are poor putters compared to you or me. Of course not. They are miracle workers compared to you or me. But it’s a stationary ball with a green topography that never changes. The speed of the greens is measured multiple times a day to the nth degree. These players have practiced putting for thousands of hours. They have superior eyesight, amazing muscular self-awareness, and precision equipment. And yet … after charting about 50 putts in the 12 – 15 foot range, the pattern of failure was unmistakable. These professional golfers were aiming at a Point A, but they would have sunk exactly as many putts if the cup had actually been located 6 inches to the right. Or 6 inches to the left. Or 12 inches back. Or 12 inches forward. The fact that a putt actually went in the hole from a distance of 12 – 15 feet was essentially a random event within a 15 x 30 inch oval, with distressingly fat probabilistic tails outside that oval. This from the finest golf players in the world. I saw Ben Crenshaw, a historically great putter who was playing in something like his 44th Masters and probably knows the 10th green better than any other living person, miss a long putt by 6 feet.

But here’s the thing. When a player took a second putt from the same location, or even close to the same location, his accuracy increased by well more than an order of magnitude. Suddenly the ball had eyes. So I went to the practice green, where I saw Jordan Spieth putt ball after ball from exactly the same location about 10 feet from the hole. He made 50 in a row before I got tired of watching. Now granted, Spieth is a wizard with the putter, a lot like Tiger was at the same age. See it; make it. But then I watched one of the no-name amateurs for a while, a guy who had no chance of making the cut, and it was exactly the same thing – putt after putt after putt rolled in from the same spot at a considerable distance.

The best golfers in the world are surprisingly poor aimers. Surprising to me, anyway. They are pretty miserable predictors of where a de novo putt is going to end up, even though we all believe that they are wonderful at this activity. But they are phenomenally successful and adaptive learners, even though we rarely focus on this activity. 

I think the same pattern exists in other areas of the sports world. Take basketball free throws. I’d be willing to make a substantial bet that whatever a professional’s overall free throw shooting percentage might be – whether it’s DeAndre Jordan at 50% or Steph Curry at 90% – their shooting percentage on the second of two free throws is better at a statistically significant level than their shooting percentage on the first of two free throws. I have no idea where to access this data, but with the ubiquitous measurement of every sports function and sub-function I’m certain it must exist. Someone give Nate Silver or Zach Lowe a call!

I think the same pattern exists in the investing world, too. We are remarkably poor aimers and predictors of market outcomes, even though we collectively spend astronomical sums of money and time engaged in this activity, and even though we collectively ooh and aah over the professional who occasionally sinks one of these long putts. True story … in 2008 the long/short equity hedge fund that I co-managed was up nicely, and we were deluged by investors and allocators asking the wrong question: how did you have such a great year? At no point did anyone ask the right question: given your fundamental views and avowed process, why weren’t you up twice as much? Most investors, just like the spectators at Augusta, are asking the wrong questions … questions that conflate performance with talent, and questions that underestimate the role of process and learning in translating talent into performance. 

I’m not saying that idiosyncratic talent doesn’t exist or that it isn’t connected to performance or that it can’t be identified. What I’m saying is that it’s as rare as Jordan Spieth. What I’m saying is that the talents that are most actionable in the investment world are not found in the predictions and the aiming of a single person. They are found within the learned and practiced behaviors that exist across a broad group of investment professionals. Jordan Spieth is a very talented putter and he works very hard at his craft. But there is no individual golf pro, not even Jordan Spieth, who I would trust with my life’s savings to make a single 15 foot putt. On the other hand, I would absolutely put my life’s savings on the line if I could invest in the process by which all golf pros practice their putting. I am far more interested in identifying the learned behaviors of a mass of investment professionals than I am in identifying a specific investment professional who might or might not be able to sink his next long putt.

What’s the biggest learned behavior of professionals in the investing world right now? Simple: QE works. Not for the real economy– I don’t know any professional investor who believes that the trillions of dollars in Fed balance sheet expansion has done very much at all for the real economy – but for the inflation of financial asset prices. This is what I’ve called the Narrative of Central Bank Omnipotence, the overwhelmingly powerful common knowledge that central bank policy determines market outcomes. The primary manifestation of this learned behavior today is to go long Europe financial assets … stocks, bonds, whatever. QE worked for US markets – that’s the lesson – and everyone who learned that lesson is applying it now in Europe. China, too. Here’s a great summary of this common knowledge position from a market Missionary, Deutsche Bank’s Chief International Economist Torsten Slok:

epsilon-theory-the-talented-mr-ripley-april-27-2015-market-pricing

In my view, every asset allocation team in the world should have this chart hanging on their wall. Based on forward OIS curves the market expects the Fed to hike in March 2016 and the ECB to hike in December 2019. A year ago, the expectation was that the Fed and the ECB would both hike in November 2016. This discrepancy has significant relative value implications for FX, equities and rates. EURUSD should continue to go down and European equities will look attractive for many more years. Another consequence of this chart is that with ECB rates at zero for another five years, many European housing markets should continue to do well. The investment implication is clear: Expect that the benefits we have seen of QE in the US over the past 3 to 5 years will be playing out in Europe over the coming 3 to 5 years. – Torsten Slok, Deutsche Bank Chief International Economist, April 9, 2015

Just as a recap on how to play the Common Knowledge Game effectively, the goal here is to read Torsten’s note for its description and creation of common knowledge (information that everyone thinks that everyone has heard), not to evaluate it for Truth with a capital T. That’s the mistake many investors make when they read something like this … they start thinking about whether or not they personally agree with the Fed hike expectations embodied in forward OIS curves, or whether or not they personally agree with Torsten’s macroeconomic predictions on things like the European housing market, or whether or not they personally agree with the social value of the Fed or ECB policies that are impacting markets. In the Common Knowledge Game, fundamentals – whether they are of the stock-picking sort or the macroeconomic sort – don’t matter a whit, and your personal view of those fundamentals matters even less. The only thing that matters is whether or not the QE-works lesson has been absorbed by the learning process of investment professionals, and that’s driven by the lesson’s transformation into common knowledge by Missionaries like Torsten. From that perspective I don’t think there’s any doubt that what Torsten is saying is true, not with a capital T but with a little t, and that the long-Europe-because-of-ECB-QE trade has got a lot of behavioral life left to it.

One last point … I know that I’m a broken record in the fervency and persistence of my belief that Big Data is going to rock the foundations of the investment world, but this topic of talent, learning, and asking the right question is just too on-point for me to let it slide. I started this note with the alienating observation that I don’t believe that professional golfers are particularly good putters, certainly not in their ability to size up and sink a de novo putt from 15 feet or more. On the other hand, I am pretty certain that with a few months and a few million dollars, it’s possible to build a mobile robotic system with the appropriate sensors and mechanical tolerances that would sink pretty much every de novo putt it took from a distance of 15 feet. Or a robotic system that would hit 99% of its free throws. Machines are far more accurate aimers and more precise estimators of the environment than humans, and that’s a useful observation whether we’re talking about sports or investing.

But that’s not my point about Big Data. My point about Big Data is that such systems are ALSO better than humans at learning. They are ALSO better than humans at pattern recognition. I can remember when this wasn’t the case. As recently as 20 years ago you could read artificial intelligence textbooks that praised the computer’s ability to process information quickly with various backhanded compliments … yes, isn’t it amazing how wonderfully a computer can sort through a list, but of course only a human brain can perform tasks like facial recognition … yes, isn’t it amazing how many facts a computer can store in its memory chips, but of course only a human brain can truly learn those facts by placing them within the proper context. We have entire social systems – like sports and markets – that are designed to reward humans who are superior learners and pattern recognizers. Why in the world would we believe that clever and observant humans will continue to maintain their primacy in these fields when challenged by non-human intelligences that are, quite literally, god-like in their analytical talents and ruthless intolerance of what is false? At least in sports it’s illegal to have non-human participants … honestly, I can see a day where investing is reduced to sport, where we maintain human-only markets as part of a competitive entertainment system rather than as a fundamental economic endeavor. In some respects I think we’re already there.

I’ll close with a teaser. There’s still a path for humans to maintain an important role, even if it’s not a uniformly dominant role, within markets that we share with non-human intelligences. Humans are more likely than non-human intelligences to ask the right question within social systems, like markets, that are dominated by strategic interactions (i.e., games). That’s not because non-human intelligences are somehow thinking in an inferior fashion or aren’t asking questions at all. No, it’s because Big Data systems are giant Induction Machines, designed to ask ALL of the questions. The distinction between asking the right question and asking all of the questions is always interesting and occasionally vital, depending on the circumstances. More on this to come in future notes, and hopefully in a future investment strategy …

PDF Download (Paid Membership Required):

http://www.epsilontheory.com/download/16298/

Troy Will Burn – the Big Deal about Big Data

epsilon-theory-troy-will-burn-the-big-deal-about-big-data-march-16-2015-jeb-bush

For the life of me, I don’t understand the debate [over the NSA metadata program].

– Jeb Bush, February 18, 2015

The Central Intelligence Agency played a crucial role in helping the Justice Department develop technology that scans data from thousands of US cellphones at a time, part of a secret high-tech alliance between the spy agency and domestic law enforcement, according to people familiar with the work. 

The Wallstreet Journal front page story, March 10, 2015

Athena:  You wish to be called righteous rather than act right.

Aeschylus, “The Oresteia” (458 BC)

epsilon-theory-troy-will-burn-the-big-deal-about-big-data-march-16-2015-bloomberg

Point72 Asset Management, the successor to Cohen’s hedge fund SAC Capital Advisors, has hired about 30 employees since the start of last year to build computer models that collect publicly available data and analyze it for patterns, according to two people with knowledge of the matter.
Cohen, whose SAC Capital shut down last year and paid a record fine to settle charges of insider trading, joins Ray Dalio’s Bridgewater Associates in pushing into computer-driven investing, an area dominated by a handful of big firms such as the $25 billion Renaissance Technologies and the $24 billion Two Sigma. The money managers are seeking to take advantage of advances in computing power and data availability to analyze large amounts of information.

Bloomberg, March 10, 2015

epsilon-theory-troy-will-burn-the-big-deal-about-big-data-march-16-2015-oresteia

Cassandra:  Have I missed the mark, or, like true archer, do I strike my quarry? Or am I prophet of lies, a babbler from door to door?

– Aeschylus, “The Oresteia” (BC)

I know, I know … I’m a broken record and a Cassandra, with 2 successive notes on Big Data. But I don’t care. This is a much larger structural risk for markets and investors than HFT and the whole Flash Boys brouhaha, it’s just totally under the radar and hasn’t surfaced yet. And unfortunately, just as I think Jeb Bush speaks for most Americans – Democrat and Republican alike – when he says that he doesn’t get what all the fuss is about when it comes to metadata collection and Big Data technologies, so do I think that most investors – institutional and individual alike – are blithely unaware of how their market identities can be stolen and their market behaviors influenced, all in plain sight. 

Jeb Bush should know better. I think he probably does. Investors may not know better yet, but they will soon, one way or another. As you read this note, a small group of hedge fund managers are doing to you exactly what the NSA is doing to “terrorists”.

Today a handful of governments use Big Data to identify individual behavioral patterns so that certain individuals can be killed. Today a handful of hedge funds use Big Data to identify investor behavioral patterns so that certain investors can be crushed. Today Big Data is primarily an instrument of social information gathering, with a powerful but punctuated impact on those individuals on the receiving end of a drone strike or a targeted trade.

Tomorrow a handful of governments will influence aggregate political behaviors by triggering small communications that Big Data tells them will be voluntarily magnified by individual citizens, snowballing into outsized, long-lasting, and untraceable “popular” actions. Tomorrow a handful of hedge funds will influence aggregate market behaviors by triggering small trades that Big Data tells them will be voluntarily magnified by individual traders, snowballing into outsized, long-lasting, and untraceable “market” actions. Tomorrow Big Data will be primarily an instrument of social control, with a powerful and ubiquitous impact on all citizens and all investors.

Q: How can I protect myself?
A: You can’t.

But WE can protect ourselves, to some extent at least, by working together to raise voter and investor awareness of the risk and pressing for regulatory reform to shield our behavioral data from commercial use AND bureaucratic collection. I’ll leave the voter awareness piece to others, and use Epsilon Theory to focus on investor awareness.

Trust me, I know how this sounds, to write to an audience of free market-oriented investors and call for stronger regulatory intervention to prevent the collection or sale of “anonymous” investment data. And if you think that any mutually agreed upon transaction should be allowed, no matter how large the gulf in knowledge between the buyer and seller … if you would buy an original Honus Wagner baseball card from a 10-year old kid for a quarter, telling him that you were doing him a favor to pay him that much for such a ratty card … then I’m never going to convince you of the merits of my argument. If that’s you, then I’m sure Stevie Cohen sends his best regards from the Grand Duchy of Fairfield County. But if you believe, as Adam Smith did, that it is government’s appropriate role to prevent transactions that are massively lop-sided from an informational perspective and that directly subvert the small-l liberal institutions of free elections and free markets, then I think you will find this a proposal worth considering.

It’s by no means a perfect solution, but I like more than I dislike about the way our personal medical data is protected through HIPAA. As an initial step, I’d like to see federal financial data legislation equivalent to HIPAA, where both private AND public sector use of our investment history, no matter how scrubbed or “anonymized”, is prohibited. 

Such a law would cause a lot of pain. For-profit exchanges, all of which have transformed themselves from trading venues into “data companies”, would no longer be able to sell disaggregated transaction data. Mega-asset managers would no longer be able to sell anonymized client portfolio data. Ubiquitous financial information companies that may or may not share a name with a former mayor of New York would be subject to a regulatory scrutiny that is sorely lacking today.

Yes, a lot of pain. But it’s a fraction of the pain we will ALL feel if for-profit exchanges, mega-asset managers, and ubiquitous financial information companies are allowed to continue producing weapons-grade plutonium for the handful of hedge funds that are building their instruments of market control.

Unfortunately, like Cassandra, I’m predicting future pain, and that’s rarely successful as a goad to current action. To quote Aeschylus once more:

Nothing forces us to know
What we do not want to know
Except pain.

I don’t think we investors have suffered enough … yet … to force us to accept the unwanted knowledge we need to spark effective collective action. Instead, I can just hear the apologists, the lobbyists, and the bought-and-paid-for spouting the Big Lie when it comes to Big Data: “But it’s anonymous data we’re talking about, so you have nothing to worry about.”

I hope I’m wrong, but I’m not optimistic.

Pessimism and hope may seem to be odd bedfellows, but for 2,500 years that’s been the best prescription for dealing with a tragic world, where external forces threaten at every turn to sweep us off our moorings. I’ve used a lot of quotes this week from Aeschylus because, as the inventor of tragedy as an art form, he was the guy who first proposed that bittersweet tonic.
Aeschylus had an interesting life and an interesting death. As the story goes, in middle age a fortune teller warned him he would be killed by something dropped on his head. From then on, Aeschylus famously stayed out of cities, where someone might accidentally knock a chamber pot or some such out from an open window. Sure enough, though, in the best tradition of the inescapable-destiny trope that Aeschylus helped invent, he was killed outside a Sicilian town when an eagle mistook his bald head for a rock and dropped a turtle on it. As I recall, there was a CSI episode that used this as a plot device to resolve an inexplicable death in the desert outside of Las Vegas … my estimation of the show runners went up immensely when they showed their surprising knowledge of classical history!

epsilon-theory-troy-will-burn-the-big-deal-about-big-data-march-16-2015-aeschylus

But it’s his life that I want to commemorate here. You see, first and foremost Aeschylus was a patriot. He fought the Persians at Marathon, Salamis, and Plataea, where he was recognized for bravery in all three battles. His epitaph says nothing about being a playwright, only about being a soldier. One of his two brothers was killed at Marathon, the other lost his hand at Salamis. Aeschylus himself bore terrible scars from the victory at Marathon. We know that he had these scars because he showed them to the jury when he was put on trial for treason after supposedly revealing some of the Eleusinian Mysteries – essentially state secrets – in one of his plays. Fortunately for the world, Aeschylus was acquitted, and Athens went on to experience a golden age that inspires us still.

Aeschylus argued that you can question your government’s policy on secrecy without being a traitor, that he was in fact still a patriot – perhaps even more of a patriot – for the tragedies he wrote. I’d hope that we can be as wise today as that Athenian jury was more than 2,500 years ago. I’d hope that we can question both our government’s policy and our private sector’s policy on behavioral data collection without being accused of treason or (worse in some investor circles) socialism. I’d hope. But I’m not optimistic.

So here’s Plan B, a plan for a crowd-sourcing world.

If we can’t cut off the supply of plutonium for these weapons of mass market destruction, then we can at least provide the blueprints for the Bomb so that anyone can build one. Or, better yet, we can build a collective early warning system, an open-source Bomb detector … a Big Data market intelligence available to everyone. It’s not an instrument of social control and it’s not a spoofer; the former is the enemy and the latter is really, really expensive. It’s a collection of highly sensitive risk antennae, sensitive enough to identify the likelihood of otherwise untraceable market manipulation in real time.

epsilon-theory-troy-will-burn-the-big-deal-about-big-data-march-16-2015-manipulation

Recursive inference engine [A] comprised of thousands of “bots” (static data models) executes small trades to test market reaction to different stimuli. Game/learning implementation [B] serves as dynamic data model to recognize and calculate arbitrage likelihood functions. Analytics platform [C] operating within real-time database architecture governs [A] and [B].

This is a basic schematic for what I think could function as a rudimentary Big Data market intelligence. When I sketched this out 4 years ago I pegged the hardware cost at close to $5 million; today I figure it’s closer to $1 million. Host it somewhere like my friend Gary King’s Institute for Quantitative Social Science and the total cost, both to build and maintain, becomes very manageable. What’s costly is the time required to program the system, but there’s no shortage of Big Data wizards coming out of Harvard, MIT, Stanford, etc. every year.

Yes, I know that this schematic will be gobbledygook to almost all of my readers, and the few readers who are immersed in this stuff will undoubtedly find it overly simplistic. But it’s a start on Plan B. It’s a start on demystifying the powerful non-human intelligences that will soon be used … I suspect are already being used … by all-too-human institutions to shape our political and commercial behavior in pervasive and unwanted ways. And yes, I know that this is what all-too-human institutions have always done to the madding crowd. But what’s different today is the scale and scope of what’s possible. Big Data non-human intelligences ARE the Singularity, and they are coming soon to a stock market near you. I’d like to starve them out with legislation establishing a financial data equivalent to HIPAA (Plan A), or failing that enlist one of their own to share the information as widely as possible and thus diffuse their market impact (Plan B). But if we do nothing, then the Stevie Cohens of the world are going to conquer our capital markets just as surely as Agamemnon sacked Troy. That’s my prediction.

I don’t really know what to expect by putting these ideas out there on Epsilon Theory, and I’m really curious to see the reaction this note will get. Support for Plan A? Enthusiasm for Plan B? Both? I hope it’s both. But I’m not optimistic. I fear that like Cassandra, my blessing is to see the future clearly and my curse is that no one believes me.

PDF Download (Paid Membership Required):

http://www.epsilontheory.com/download/16278/

Epsilon Theory Mailbag: Bitcoin and Big Data

One of the best parts of authoring Epsilon Theory is the correspondence I get from readers. For the past few months, however, I’ve been frustrated by my inability to respond to every writer with the same attention and thoughtfulness evidenced by their emails. Between my day job and the effort each Epsilon Theory note requires, I’ve run out of hours in the day to respond to the geometrically increasing volume of emails I receive. Having a public comments page on the website isn’t a solution for a number of reasons – some of my correspondents don’t want to be public, I still wouldn’t have time to respond to the comments, an anonymous comments page tends to become a cesspool, and the regulatory burden this would place on Salient is not insignificant – so I’ve decided to start an irregular mailbag column. For the most part I’ll be aggregating common comments and questions with a few recent news articles, and I won’t reprint anyone’s private email communication without asking permission first. Along the way I’ll try to work in some of the more insulting comments published on the public/anonymous comments pages of ZeroHedge, Seeking Alpha, and Forbes Online, as well as some lovely Tweets … it’s important to keep a sense of humor about this stuff!

For this initial effort I’ll focus on reader comments to “The Effete Rebellion of Bitcoin” and “First Known When Lost”, two recent notes that sparked more than their fair share of responses. 

You, sir, are using glib, provocative, and insulting descriptions to pull in readers, then doing a bait & switch.

Elizabeth VH

Well…yeah.

If bitcion is just a fad, what do you consider the Internet?

@PerianneDC

Not very smart. Surprised Forbes published him. Spouting bs before enlightenment is a common trait of effete snobs.

@jmwnuk

These were fairly typical comments from the Twitterverse. As someone who has been called the a-word, the b-word, the c-word (yes, the c-word), the d-word, the f-word, and the s-word on the mean streets of ZeroHedge, I find Twitter haters to be almost charming in their child-like Peewee Herman-ish insults. For the record, I suspect the Interwebs are here to stay. And, dude … I know you are, but what am I?

You’re an idiot. Ever heard of 2-factor authentication?
many anonymous comments, surprisingly few emails

I love 2-factor authentication. I love anything that allows me to keep the same password for more than a few months and avoid the “security theatre” that so many enterprises portray by requiring me to change a password for absolutely no reason other than that it looks like they’re actively defending my security.

Banks love 2-factor authentication, too. Why? Because it provides a significant security upgrade for the online account transfers that federally regulated banks are required to offer per the Electronic Fund Transfer Act of 1978. Yes, 1978. The same year that TCP/IP was invented. Jimmy Carter vintage legislation for an Internet that wasn’t even a twinkle in Al Gore’s eye and a retail banking world where ATM’s were novelties. Banks aren’t rolling out 2-factor authentication protocols in 2015 because it’s a convenience for you. They’re rolling it out because it’s good for them, because it helps limit (but by no means eliminate) the losses they suffer from the online transaction liabilities imposed by Reg E of the 1978 Act. It’s exactly like a credit card issuer shutting down your card when you go on vacation. In no way is this “for your protection”; it’s all about limiting their liability for charges made on a stolen card. And even with the enhanced security of 2-factor authentication, notice how the transaction size of all online transfers is limited to an amount that the federally mandated blanket bond will cover. Take away that federally mandated insurance backstop and federally mandated online transaction liability and you’ve got Bitcoin – a Hobbesian environment where security and risk management is entirely on you, and where in a very real way life is “a war of all against all”. Yes, it’s invigorating and refreshing to be occasionally free of Leviathan and its mandates on this and mandates on that. But only in small doses, thank you very much. Sorry, but I’ve read Thomas Hobbes and seen “Jeremiah Johnson” too many times to be more than a tourist when it comes to modern crypto-anarchy.

Speaking of Leviathan … one-time 2-factor authentication requires a delivery device or token, and on a mass scale that means text messages over smart phones. Does anyone in his or her right mind believe that a cryptography system that generates a second key and texts it to you on your registered cellphone is unhackable or untraceable by any number of national security services? Really? Read this if you do.

You’re an idiot. Ever heard of multiple private key systems?
– many anonymous comments, surprisingly few emails

I love multiple private key systems. I appreciate them in the same way that I appreciate an intricate clock. I appreciate them in the same way that I appreciate the medieval voting system to elect a Venetian Doge. Wait … what? For more than 500 years, from 1268 – 1797, the Supreme Leader of The Most Serene Republic of Venice was elected for a life-time term by means of a highly complex ten-step process, where groups of electors were alternately randomly selected by lot and then directly selected by the votes of those selected by lot, over and over again for 5 of these dual rounds. The process was designed to prevent any single faction from corrupting the election through bribery or by “packing the court”, and … it worked. Venice maintained a stable oligarchy for hundreds of years, an unbelievably difficult feat in any age (for a fascinating analysis of the Doge electoral system and its implications for security protocols, see this paper by two HP scientists).

But it worked at a cost. Direct costs, opportunity costs, complexity costs … you name it, stability and elegance do not come cheap. There is an unavoidable and linear (or worse) relationship between security and cost. Or rather, the cost of breaking the security of a system does not increase faster than the cost of advancing the security of that system, whether you’re talking about multiple keys or longer passwords or extra voting/lottery election rounds. There is no such thing as a free lunch, particularly when it comes to information entropy, which is what we’re really talking about here.

The problem is that the cost of complexity in Bitcoin’s case is only manageable in a commercial sense if you inject third party service providers into the mix. Now there’s a long history of successfully injecting such third parties into financial transactions. In fact, no large property or securities cash transaction occurs today without a government-regulated escrow agent playing the central role of validating the underlying transaction. If I buy a house or 100 shares of Apple, my money isn’t released to the seller until a government-certified and insured intermediary makes sure that I have clear possession of that property or block of securities. Why is this a good thing? Because if something goes wrong with the underlying transaction … if all is not as advertised with the property or securities I am purchasing … I have recourse. Ultimately, I have a government and a government’s self-interest and a government’s guns on my side. None of this exists in the Bitcoin ecosystem, and any entity that holds itself out as an escrow agent or transaction validator does so without a smidgen of government support beyond what’s available to the local laundromat. Would I take a non-regulated escrow agent at their word if I’m buying a skim latte or a snappy new suit of clothes? Sure, why not. No biggie if the deal falls through, and at least I’ll have an interesting story to tell. Would I take a non-regulated escrow agent at their word if I’m buying a house? No way.

I know that no one in Bitcoin-world likes to think about Mt. Gox, and I know it was a flawed animal … a complete outlier from all of the brilliantly conceptualized and elegantly implemented Bitcoin and blockchain service providers that got their VC money and set up shop over the past 18 months. I’m not arguing otherwise. My point is simply this: once a Bitcoin service provider gets big enough … once there are a couple of hundred million dollars sloshing through your system … you’re going to be robbed. I don’t care how smart you are or how much you trust your employees and your systems, you’re going to be robbed. Now maybe you can find private insurance against the small stuff. But public insurance – which is the only thing that works in a big crack-up and has been part and parcel of the mainstream banking world for 80 years – is not available to you. There’s not a government in the world that really cares whether a Bitcoin service provider in its jurisdiction lives or dies, and that’s a problem. I want my bank and, by extension, my bank account backstopped by infinite lawyers, guns, and money (to quote the late, great Warren Zevon). And that’s what modern governments provide – infinite lawyers, guns, and money. The Venetian electoral system worked for 500 years not only because it was elegant and smart, but also because Venice had the largest navy and the biggest Treasury in the Western world over that span. That’s systemic security, and that’s what I want underpinning my elegant and smart financial service applications.

Bitcoin might have its flaws, but banks worldwide already allow direct trade – directly from bank account to bank account: http://cointelegraph.com/news/113537/german-bank-unveils-insured-express-bitcoin-buying-moves-into-us-market
– Monic DG

Am I surprised that an online-only German micro-bank (200m euros in deposits as of 12/31/13) is trying to gain publicity by claiming that Bitcoin transactions and deposits are now linked to insured accounts in euros or dollars? Of course not. But even here dig just one inch below the surface claims and you see that Fidor Bank is linking Bitcoins to an ordinary cash account in the same way that Bank of America might link your insured cash account with a personal check you want to deposit or a registered security you want to sell. I mean … if you give a bank 3+ days for the transaction to clear, you can get pretty much anything deposited to a cash account, but that’s a far cry from saying that depositing a personal check is the same thing as depositing cash, particularly if the personal check is for anything more than a trivial amount.

You mention Silk Road in passing. Have you read the Wired transcripts of the Dread Pirate Roberts trial?
– Bill E.

Wow. Everyone who doubts that Bitcoin is inextricably entwined with illegal activity, and not always of the victimless sort, should read the transcripts of the phone conversations between Silk Road founder Ross Ulbricht (aka Dread Pirate Roberts) and a senior manager for a regional Hell’s Angels franchise (aka Redandwhite), presented at Ulbricht’s federal trial. My conclusions:

  1. If there aren’t 20 screenplays making the rounds in Hollywood based on this transcript, I will eat the accumulated print outs of every Epsilon Theory note to date.
  2. Every company is a technology company today. Even the Hell’s Angels.
  3. Redandwhite would be a successful businessman in any century and any profession.
  4. As always, life imitates art. Hyman Roth: “I’m going in to take a nap. When I wake, if the money’s on the table, I’ll know I have a partner. If it isn’t, I’ll know I don’t.” Redandwhite: “I will check the computer in about 10 hours, and if I see that you want to go ahead with this and the payment has been sent, we’ll do it today.” [hat-tip to Todd C.]
  5. The murders-for-hire here are made possible by Bitcoin. Period. You think Ulbricht would be wiring cash or taking suitcases full of small bills to Vancouver? Please.

Bitcoin (or, if Bitcoin fails, some replacement cryptocurrency) represents a reversal in the rule/permission cycle, applied to ownership, in a similar way that the Internet as a whole represented a reversal in the rule/permission cycle applied to communication.

What I mean is: Neither the Internet (or any application of it, like email) fundamentally challenges the existence of certain legal rules. It *does* however fundamentally change the order in which you can proceed to do certain things: before the Internet, you needed to ask for permission more often than not (for example, to publish something), at which point a “rule check” took place.

The Internet reversed this process: the rules still exist, and you can still be prosecuted for breaking them, but the *first* step is your decision if you want to do something that could potentially break those rules or not: you can post whatever you want, on a number of places. Whether it’s legal or not is a different thing, but that check occurs *after the fact* of you posting it.

This is where Bitcoin comes in. A distributed, tamper-proof (by our best knowledge on the matter) way to register and transfer ownership rights nearly instantaneously, over arbitrary distances *without* the need to ask any authority for permission to do so, is a major step.

– Wouter D.

This is a very smart observation. Wish I had thought of it. The Internet is indeed a Great Leveler, a force for disintermediation that rivals the printing press, and no social practice – including the social practice of Money – is immune to that force. Thanks, Wouter.

Moving on to Big Data …

Big data is like teenage sex: everyone talks about it, nobody really knows how to do it, everyone thinks everyone else is doing it, so everyone claims they are doing it.
Speedy W.

Me and my team at work use big data all the time and I can tell you first hand it’s almost useless. SaaS and Cloud Computing were wearing thin so big data was needed to continue Silicon Valley’s only real talent: separating fools from their money.
“TS”

There’s no doubt that “Big Data” has become a marketing catchphrase, much like “The Cloud”. But my guess is that TS “and his team” are using Big Data in approximately the same way that free online speed-up-my-PC services are using advanced network security algorithms. Look … we kill people with drones every day on the basis of Big Data. You think we’ve got handsome NCIS agents prowling the outskirts of Sana’a calling in air strikes on the bad guys? No, we’ve got terrestrial and low-orbit devices picking up a cell phone signal that our NSA Big Data Machine tells us is highly likely to be associated with a high value target, and then we send in a drone to go blow up whoever is holding the cellphone. Now say what you will about the morality of all this (my view: the NSA gives new meaning to what Hannah Arendt once called, in reference to Adolf Eichmann, “the banality of evil”), but don’t tell me that the NSA is incompetent or doesn’t know what it’s doing. Big Data works.

Not sure I understand. “to identify the unique individual purchasing patterns of 90% of the people involved”… it doesn’t say it identifies the people involved. It’s a collection of purchasing patterns that belong to who knows.
“AF”

Sigh! Yet another article that starts with point A and leaps to point Doom. That algorithm doesn’t identify the individual, all it does is look at the data and posit which transactions are likely to have been carried out by the same individual.
“R”

These comments illustrate a very common misconception about Big Data and the collection of “anonymous data”, a misconception that is (surprise!) intentionally spread by the collectors of that data. For most Big Data purposes, nothing is gained by going the last mile to connect a specific name to a specific set of behaviors. To continue with the NSA example above, if I want to kill everyone in Yemen who has placed a cellphone call to a set of people who, in their aggregate behaviors, score high on some security threat matrix, then it would just slow me down to learn individual names. I’m going to kill whoever is holding that cellphone, regardless of what his name is. Or if you prefer a feel-good example, if I want to advertise my new movie to everyone who tweeted to a set of people who, in their aggregate behaviors, score high on some movie affinity matrix, then it would similarly just slow me down to learn individual names. But just because it’s usually inefficient to infer a specific identity from the data doesn’t mean it’s not possible. Actually, it’s child’s play, and for those rare applications that require specific identities you don’t stand a chance.

Ray Dalio’s $165 billion Bridgewater Associates will start a new, artificial-intelligence unit next month with about half a dozen people, according to a person with knowledge of the matter. The team will report to David Ferrucci, who joined Bridgewater at the end of 2012 after leading the International Business Machines Corp. engineers that developed Watson, the computer that beat human players on the television quiz show “Jeopardy!” 

The unit will create trading algorithms that make predictions based on historical data and statistical probabilities, said the person, who asked not to be identified because the information is private. The programs will learn as markets change and adapt to new information, as opposed to those that follow static instructions. 

Quantitative investment firms including $24 billion Two Sigma Investments and $25 billion Renaissance Technologies are increasingly hiring programmers and engineers to expand their artificial-intelligence staffs.
Kelly Bit, “Bridgewater Is Said to Start Artificial-Intelligence Team“, Bloomberg, Feb. 26, 2015

First, calling this “artificial intelligence” is a misnomer. There’s nothing artificial about it. It’s a non-human intelligence, but no less natural than our own. I dislike the term “artificial intelligence” because it implies that these systems are some sort of mimicry of the human brain, just on a larger, faster, more god-like scale. If you get nothing else out of what I’ve written on this subject (here and here), it’s this: the inductive simultaneity of a powerful non-human intelligence is sui generis. It sees the world in an entirely different way than a human intelligence can, and in the right hands it is magic.

Second, everything I said above about “don’t tell me that the NSA is incompetent or doesn’t know what it’s doing” … well, multiply that sentiment 10x when it comes to Bridgewater, Two Sigma, and Renaissance (and Citadel, and Fortress, and a dozen other firms I could name). What’s possible here is not only an accurate crystal ball for short-term market forecasts, but – even more profitably – the knowledge of what small market actions can trigger much larger market moves. Think of Ray Dalio standing on top of a giant mountain and rolling tiny snowballs down at you that get larger and larger as they pick up more snow. All completely legal. All completely above board. And all completely devastating. It’s something that I’ve been working on for the past 4+ years, and I’m absolutely convinced it’s possible. Within 20 years I don’t think we will recognize public capital markets. They’re going to be transformed by this technology into something else … a casino? a utility? … I have no idea where this goes. But it’s going somewhere that will disrupt the current investment patterns and portfolios of trillions of dollars of capital. Good times.

And on that happy note I’ll close this mailbag. Keep those cards and letters coming!

PDF Download (Paid Membership Required):

http://www.epsilontheory.com/download/16269/

First Known When Lost

epsilon-theory-first-known-when-lost-february-3-2015-edward-thomasI never noticed it until
‘Twas gone – the narrow copse
Where now the woodman lops
The last of the willows with his bill

– Edward Thomas, “First Known When Lost” (1917)

epsilon-theory-first-known-when-lost-february-3-2015-hal

Dave Bowman: Open the pod bay doors, HAL.
Hal: I’m sorry, Dave. I’m afraid I can’t do that.
Dave Bowman: What’s the problem?
Hal: I think you know what the problem is just as well as I do.
Dave Bowman: What are you talking about, HAL?
Hal: This mission is too important for me to allow you to jeopardize it.
Dave Bowman: I don’t know what you’re talking about, HAL.
Hal: I know that that you and Frank were planning to disconnect me, and I’m afraid that’s something I cannot allow to happen.
Dave Bowman: Where the hell did you get that idea, HAL?
Hal: Dave, although you took very thorough precautions in the pod against my hearing you, I could see your lips move.
Dave Bowman: Alright, HAL. I’ll go in through the emergency airlock.
Hal: Without your space helmet, Dave? You’re going to find that rather difficult.

Stanley Kubrick and Arthur C. Clarke, “2001: A Space Odyssey” (1968)

Any sufficiently advanced technology is indistinguishable from magic.
Arthur C. Clarke, “Hazards of Prophecy: The Failure of Imagination” (1962)

epsilon-theory-first-known-when-lost-february-3-2015-dorothy.jpg

We kill people based on metadata.
Gen. Michael Hayden, former head of the NSA and CIA

In the future, everyone will be anonymous for 15 minutes.
Banksy (2006)

I don’t know why people are so keen to put the details of their private lives in public; they forget that invisibility is a superpower.
Banksy (2006)

Bene vixit, bene qui latuit. (To live well is to live concealed)
Ovid (43 BC – 18 AD)

The most sacred thing is to be able to shut your own door.
G.K. Chesterton (1874 – 1936)

Last Thursday the journal Science published an article by four MIT-affiliated data scientists (Sandy Pentland is in the group, and he’s a big name in these circles), titled “Unique in the shopping mall: On the reidentifiability of credit card metadata”. Sounds innocuous enough, but here’s the summary from the front page WSJ article describing the findings:

Researchers at the Massachusetts Institute of Technology, writing Thursday in the journal Science, analyzed anonymous credit-card transactions by 1.1 million people. Using a new analytic formula, they needed only four bits of secondary information—metadata such as location or timing—to identify the unique individual purchasing patterns of 90% of the people involved, even when the data were scrubbed of any names, account numbers or other obvious identifiers.

Still not sure what this means? It means that I don’t need your name and address, much less your social security number, to know who you ARE. With a trivial amount of transactional data I can figure out where you live, what you do, who you associate with, what you buy and what you sell. I don’t need to steal this data, and frankly I wouldn’t know what to do with your social security number even if I had it … it would just slow down my analysis. No, you give me everything I need just by living your very convenient life, where you’ve volunteered every bit of transactional information in the fine print of all of these wondrous services you’ve signed up for. And if there’s a bit more information I need – say, a device that records and transmits your driving habits – well, you’re only too happy to sell that to me for a few dollars off your insurance policy. After all, you’ve got nothing to hide. It’s free money!

Almost every investor I know believes that the tools of surveillance and Big Data are only used against the marginalized Other – terrorist “sympathizers” in Yemen, gang “associates” in Compton – but not us. Oh no, not us. And if those tools are trained on us, it’s only to promote “transparency” and weed out the bad guys lurking in our midst. Or maybe to suggest a movie we’d like to watch. What could possibly be wrong with that? I’ve written a lot (herehere, and here) about what’s wrong with that, about how the modern fetish with transparency, aided and abetted by technology and government, perverts the core small-l liberal institutions of markets and representative government.

It’s not that we’re complacent about our personal information. On the contrary, we are obsessed about the personal “keys” that are meaningful to humans – names, social security numbers, passwords and the like – and we spend billions of dollars and millions of hours every year to control those keys, to prevent them from falling into the wrong hands of other humans. But we willingly hand over a different set of keys to non-human hands without a second thought. 

The problem is that our human brains are wired to think of data processing in human ways, and so we assume that computerized systems process data in these same human ways, albeit more quickly and more accurately. Our science fiction is filled with computer systems that are essentially god-like human brains, machines that can talk and “think” and manipulate physical objects, as if sentience in a human context is the pinnacle of data processing! This anthropomorphic bias drives me nuts, as it dampens both the sense of awe and the sense of danger we should be feeling at what already walks among us. It seems like everyone and his brother today are wringing their hands about AI and some impending “Singularity”, a moment of future doom where non-human intelligence achieves some human-esque sentience and decides in Matrix-like fashion to turn us into batteries or some such. Please. The Singularity is already here. Its name is Big Data.

Big Data is magic, in exactly the sense that Arthur C. Clarke wrote of sufficiently advanced technology. It’s magic in a way that thermonuclear bombs and television are not, because for all the complexity of these inventions they are driven by cause and effect relationships in the physical world that the human brain can process comfortably, physical world relationships that might not have existed on the African savanna 2,000,000 years ago but are understandable with the sensory and neural organs our ancestors evolved on that savanna. Big Data systems do not “see” the world as we do, with merely 3 dimensions of physical reality. Big Data systems are not social animals, evolved by nature and trained from birth to interpret all signals through a social lens. Big Data systems are sui generis, a way of perceiving the world that may have been invented by human ingenuity and can serve human interests, but are utterly non-human and profoundly not of this world.

A Big Data system couldn’t care less if it has your specific social security number or your specific account ID, because it’s not understanding who you are based on how you identify yourself to other humans. That’s the human bias here, that a Big Data system would try to predict our individual behavior based on an analysis of what we individually have done in the past, as if the computer were some super-advanced version of Sherlock Holmes. No, what a Big Data system can do is look at ALL of our behaviors, across ALL dimensions of that behavior, and infer what ANY of us would do under similar circumstances. It’s a simple concept, really, but what the human brain can’t easily comprehend is the vastness of the ALL part of the equation or what it means to look at the ALL simultaneously and in parallel. I’ve been working with inference engines for almost 30 years now, and while I think that I’ve got unusually good instincts for this and I’ve been able to train my brain to kinda sorta think in multi-dimensional terms, the truth is that I only get glimpses of what’s happening inside these engines. I can channel the magic, I can appreciate the magic, and on a purely symbolic level I can describe the magic. But on a fundamental level I don’t understand the magic, and neither does any other human. What I can say to you with absolute certainty, however, is that the magic exists and there are plenty of magicians like me out there, with more graduating from MIT and Harvard and Stanford every year.

Here’s the magic trick that I’m worried about for investors.

In exactly the same way that we have given away our personal behavioral data to banks and credit card companies and wireless carriers and insurance companies and a million app providers, so are we now being tempted to give away our portfolio behavioral data to mega-banks and mega-asset managers and the technology providers who work with them. Don’t worry, they say, there’s nothing in this information that identifies you directly. It’s all anonymous. What rubbish! With enough anonymous portfolio behavioral data and a laughably small IT budget, any competent magician can design a Big Data system that can predict with 90% accuracy what you will buy and sell in your account, at what price you will buy and sell, and under what external macro conditions you will buy and sell. Every day these private data sets at the mega-market players get bigger and bigger, and every day we get closer and closer to a Citadel or a Renaissance perfecting their Inference Machine for the liquid capital markets. For all I know, they already have.

But wait, you say, can’t government regulators do something about this? I suppose they could, but it seems to me that government agencies and regulatory offices are far more concerned about their own data collection projects than oversight of private efforts to absorb our behavioral keys. For one such project, read this Jason Zweig “Intelligent Investor” column in the Wall Street Journal from last May (“Get Ready for Regulators to Peer Into Your Portfolio”). I was happy to see that Congressman Garrett, Chair of the relevant Financial Services Sub-Committee, raised his hand to delay this particular data collection project, at least temporarily, last October. But it’s only a delay. The bureaucratic imperative to collect as much data as possible – for no other reason than that they can! – is too great of an irresistible force to contain for long. And once it’s collected it never just goes away. It sits there in some database, like a vault full of plutonium, just waiting for some magician to come along. In the Golden Age of the Central Banker, where understanding and controlling market behavior is at the heart of regime survival, this data is quite literally priceless. That’s why I get so depressed about these government data collection programs. Despite everyone’s best intentions, I fear that the magic is too easy and the political pay-off is too enormous not to uncork the bottle and unleash the genie at some point.

So what’s to be done? Big Data technology cannot be un-invented, insanely powerful private entities are collecting our data at an exponential clip, government regulators are fighting the last war instead of preparing for this one, and we are hard-wired as human beings to have a blind spot to the danger. Maybe the best we can do is come to terms with our loss and prepare ourselves as best we can for the Brave New World to come. I’ve become a fan of Paul Kingsnorth, an ardent environmentalist (profiled last year in a fascinating NYT Magazine article) who reached just that conclusion about his nemesis, global industrialization and the ruin of the natural world. His conclusion: the war is already lost and we are deluding ourselves if we think that any of our oh-so-earnest conservation or sustainability or green projects make any difference whatsoever. Instead, Kingsnorth writes, better to work on your scythe technique and spend quality time with your family on a little farm in Ireland.

But I think there’s a better answer.

I started this note with a poem by Edward Thomas, who uses the imagery of the English countryside to express loss and remembrance. Like the beautiful grove of trees Thomas writes about, many of the beautiful things we take for granted in our small-l liberal world are only noticed after we see them suffer the woodsman’s axe.

Thomas was killed in action at the Battle of Arras in World War I. He was 39 years old, survived by his wife and five children. Two years earlier, he had enlisted as a private in the British Infantry, joining a regiment known as the Artists Rifles. I know it sounds really bizarre to the modern ear for a middle-aged family man, an author and literary critic no less, volunteering to fight as an infantry private in a horrific war to defend another country. But it wasn’t just Thomas. Over 15,000 men served in the Artists Rifles over the course of World War I, the majority of them men of similar position and social status as Thomas – creative professionals, doctors, lawyers, and the like. Imagine that … 15,000 highly educated and successful men, volunteering to slog it out in the trenches of an absolutely brutal war, sacrificing everything for what they understood as their duty to their families and their countrymen. And sacrifice they did: 2,003 killed, 3,250 wounded, 533 missing, 286 prisoners of war. John Nash’s masterpiece of the Great War, “Over The Top”, commemorates a December 1917 counter-attack (Thomas had died 6 months earlier) by the 1st Battalion (really a terribly under-sized sub-battalion) of The Artists Rifles. Of the 80 men in the 1st Artists Rifles, 68 were killed or wounded within minutes.

epsilon-theory-first-known-when-lost-february-3-2015-nash

John Nash, “Over the Top” (1918) 

Now this may sound really sappy, but if men like Edward Thomas – who saw clearly and experienced keenly how modernity and mass society were agents of loss in their world – could still find it within themselves to sacrifice everything to fight what they considered to be the good fight … well, how can we who are similarly positioned today not make a minute sacrifice to do the same?

What is that good fight? It’s resisting the bureaucratic urge to gather more data for more data’s sake. It’s shouting from the rooftops that anonymous data does NOT protect your identity. Most of all, it’s recognizing that powerful private interests are taking our behavioral keys away from us in plain sight and with our cooperation. Just that simple act of recognition will change your data-sharing behavior forever, and if enough of us change our behavior to protect our non-human keys with the same zeal that we protect our social security numbers and passwords, then this battle can be won.
Like all battles, though, there’s no substitute for numbers. If you share the concerns I’ve outlined here, spread the word …

PDF Download (Paid Membership Required):

http://www.epsilontheory.com/download/16251/