Follow-up book to The Eudaemonic Pie, which was a readable story about a group of genius chaos theory physicists attempting to beat roulette. In The Predictors Thomas Bass picks up the story some ten years later as the original Eudaemonic founders Doyne Farmer and Norm Packard move on to real prey: they want to use chaos theory to beat the stock market.
They start what could be called one of the early quant funds, and the story intersects with a who's who of the late '80s/early '90s investing scene, including Paul Tudor Jones, D.E. Shaw, Tom Dittmer (whose firm, Refco, infamously transformed Hillary Clinton's cattle futures account from $1,000 into $200,000), Solomon Brothers, etc.
You learn about the founders' doubts, their fears, and their moments of astounding overconfidence, where they wade into domains where they haven't a clue what's going on. Any experienced investor would be mortified to watch these guys attempt to trade oil futures at the NYMEX, while not knowing that their orders are so big and so predictable that everyone front-runs them every single day. They couldn't figure it out until they sent one of their employees into the pit to learn how things actually worked.[*]
They started trading while not really understanding the way spreads work, and without grasping slippage costs at all. "Transaction costs are going to be steeper than we thought." "We should have focused on this earlier." Their computer models rarely seems to work as expected, sometimes the models produce catastrophic problems, and when they do work they only produce meaningful profits when heavily levered up or with tremendous amounts of outside capital.
It is fascinating, however, to think about doing not just one but two kooky startup enterprises that each literally dominate decades of your life. It just doesn't seem worth it: dividing up (more like fighting over) equity, failing over and over again to get funding, struggling with payroll, living on a shoestring for years without even knowing if you'll ever have anything to take the marketplace. Whatever these guys earned, it wasn't enough. And they were among the moderately successful!
When you think about the zillions of startups that fail, invisible and unremembered, when you envision all those business efforts we never hear about and that no one writes about, it's amazing (and a blessing) that there are people out there who have the stomach to do this.
The book could use another heavy editing pass: as an example, there's one passage where the author explains the "debate" between efficient markets believers and fundamental analysis believers twice within a 40 page span, including reusing the phrase "lucky monkey." The author also has an odd habit of describing everybody's clothing in minute detail. And there are entire chapters meandering tangentially on derivatives, the hedge fund industry, the Swiss banking industry, the history of the Chicago Merc, commodities trading, etc. These might interest some readers (ding!), but probably should have been cut.
[*] This seems a good time to remind everyone of another investor who had his face ripped off in the futures markets, so I link here to Jim Cordier's bathos-filled "rogue wave" video, made as his firm Optionsellers.com collapsed, losing all client money (and then some). Do not sell volatility unless you know what you are doing.
Notes:
* Right on the first page the author slightly misuses some Wall Street jargon, conflating the witching "hour" with options triple witching.
* A few different types of non-vanilla options that show up in currency markets, all new to me: "up and in" options, "down and out" options, also "touch" options.
* The founders structure the company a lot more intelligently than the prior book (where everybody in on the Eudaemonic Pie venture got an equal share regardless of how much or how little they did, not to mention that company never made any meaningful money). This company is a lot more pyramid-shaped in its comp and equity structure, thus much more appropriately done. Note they also manage to secure an anchor investor, James Pelkey, who's very intelligent and savvy, who helps them enormously. Note that Pelkey his own google-worthy story: shot in the spine by his ex-wife, never walked again, in constant pain, yet a key investor in Silicon Valley's early days.
* Long tangent on the history of Santa Fe here, the annual Zozobra festival, the Santa Fe Institute, Doyne's work using complex systems analysis to describe economics.
* Interesting how the author argues that general equilibrium theory and efficient market theory both assume what they should instead prove. I hadn't thought about this as a true killshot for both of these (fallacious) models of reality.
* Types of self-organizing systems: markets, biological systems, immune response, origins of life. Viewing the economy as an adaptive nonlinear network. Contrast with neoclassical economics and equilibrium-based economics (see the original notion from Leon Walras in 1874 of a "clockwork universe" like Newtonian mechanics).
* See also the irony of Newton himself losing all his money in the South Sea stock bubble: "I can calculate the motion of heavenly bodies, but not the madness of crowds." None of these equilibrium economic theories had any way to explain market crashes, euphoria, speculative behavior, technological change, etc.
* "Why are we wasting our time on discussion of whether or not markets are efficient? Let's just assume profits are possible and figure out how to make them." Doyne, losing patience at the 1991 Santa Fe Institute meeting.
* "Don't work on a shoestring; don't try to do everything yourself, like building computers from scratch; and avoid the temptation to hire troublesome geniuses." These are key lessons the founders learned from starting The Eudaemonic Pie.
* See the striking images of gases combining in the 1974 paper "On Density Effects and Large Structures in Turbulent Mixing Layers" by Garry Brown and Anatol Roshko: turbulent flow between gases mixed with each other at high speed, forming fractal whirlpools mirrored in different sizes, indicating a structuring and evolving system. Patterns are an intrinsic part of turbulent fluid flow, likewise their patterns are intrinsic to complex financial markets.
* "We know from control theory that lags between phenomena and the controls placed on them will make things oscillate. A similar lag exists between receiving news and trading on that news. Time lags and delays in market perception create oscillations."
* Interesting (and kind of farcical) how Doyne totally rejects Efficient Markets theory but then buys into Modern Portfolio Theory: the idea that risk and return can be optimized with a proper portfolio of riskless and risky assets.
* On "burning" data: you can run a trading algorithm over and over against a (past) dataset but all it's going to do is perfectly fit a curve and just "predict the past"... thus this is a type of data mining. You need to try the algorithm on different datasets, different periods, etc., and you only have so many passes through a given dataset before datamining tendencies will make your model useless for the future.
* These guys really intersect with everybody, a genuine who's who of the 80s and 90s investor scene: Paul Tudor Jones, Tom Dittmer (who ran Refco), D.E. Shaw, the guys at Solomon, O'Connor & Associates (the famous options trading firm), etc.
* Doyne keeps a mental notebook of his "anthropology of Wall Street." Learning different types of people and personalities, finding (to his great surprise) a relatively high level of integrity and rigor, much higher than he expected.
* Also notable some of the ethical (and metaphysical) thoughts of the company members, like one quote: "what is the secret to being rich without being an asshole?" How much money does it take to be rich; is it ethical to siphon money off a system in a zero-sum game, for us to be winners when others have to be losers; on playing games with other people's money etc; or is it ethically okay because we are a stabilizing factor in the market running a program like this; or does this issue even matter?
* They finally find "the perfect girlfriend" in a business partnership with O'Connor and Associates, the options trading firm which signs an attractive deal with them. Note that David Weinberger, one of the MDs at O'Connor and Associates, reads The Eudaemonic Pie try to understand Doyne and Packard. Now that is legit due dilligence.
* Two intervening chapters: one explaining the roulette project, the other explaining Packard's and Doyne's careers leading up to their current effort: starting The Chaos Cabal, later Doyne working at Los Alamos and Norman working at The Institute of Advanced Studies in Princeton.
* Doyne describes a second law of self-organization ("matter tends to organize itself") paralleling or contravening the Second Law of Thermodynamics (matter tends towards entropy), this is a an effort to address self-organization self-ordering systems.
* The guys get their whole idea for this investing company from the son-in-law of Aristide Lindenmayer (the famous Hungarian botanist). The son-in-law had just sold a robot arm company for a large sum of money and suggested to them that their work on machine learning and chaos might be useful for predicting the stock market; he notes that the stock market is "a chaotic system good for modeling." Funny how a random comment from someone sets you in a totally new direction.
* Norm Packard has a change of perspective on statistics too: at first he passed on Doyne's business idea saying "I wasn't keen on learning all the classical statistics you need to know" then later has change of heart and discovers he likes it, considers statistics a deep philosophical issue. Fascinating from an epistemic perspective how you can "know" that you don't want to know something... how can you know this when you don't know enough about it to know?
* "All the bad news comes after the contract is signed" one of Jim McGill's maxims (business manager/CFO of Doyne and Packard's business)
* Note the increasing abstraction when you go from the primary market (say GM offering stock in an IPO) to a secondary market (trading/betting on GM stock on an exchange) and then to a tertiary market like an options market (trading/betting on other's trading/betting on GM stock).
* Marcel Ospel, leading Swiss Bank (later called UBS) into the options world and acquiring O'Connor and Associates; later it turns out he was forced to resign with the 2008 recapitalization of UBS.
* The helium balloon in a plane question: what happens when the plane accelerates? The helium balloon moves forward due to gas density differential.
* Rafael deNoyo: here's a guy to read about as he bounces from lawsuit to lawsuit while trying to make a fortune.
* The firm does amazingly well in paper trading, but then as soon as they go live with actual money the system crashes, unable to handle a rollover day and a half-day session around Thanksgiving.
* Company morale plummets and everybody gets stressed out after the successful deal with O'Connor and Associates and while they're starting to go live: Doyne gets insomnia, Norman's face breaks out in a rash, they end up having to get rid of two of their employees, they create an authoritative management-employee hierarchy, "It's the end of our Eudaemonic ideals"
* Fascinating/horrifying that these guys don't even really understand the way spreads work in futures markets, and they don't understand the cost of slippage.
* "...every system they develop, no matter how well it performs on historical data, falls apart when it moves into the present." Paper trading and backtests are not the same as making actual money in the real world!!
* Market "regime shifts" and "nonstationary behavior": basically these are euphemisms for tectonic shifts in the tenor of the market, as well phrases for when markets do things when you don't expect or want them to.
* Sometimes "there is no such thing as an average" (market prices, levels of Nile River, etc,, are never equal to their "average" level); see clustering behavior of outlier events; long stretches of low volatility; fat tails (and Hurst exponents to measure them). The way The Prediction Company deals with this is to create "ensembles of models" with overlapping predictions that are used as a voting system. Basically there is nothing stable or stationary about markets!! Even at this point this late in the game these guys doesn't really have a model that works yet. They can't seem to make money.
* Other famous traders to look up Michael Marcus, Joe Ritchie.
* The curse of dimensionality; using a state space diagram where the dimensions in state space correspond to the variables that describe the system. State space reconstructions help determine if patterns exist in a stream of data; ironically, the reader can't help but get the impression that the most predictive piece of data in this whole book is the company's dog, who either barks or doesn't bark to indicate a profitable or an unprofitable day. (!!!)
* Also the team keeps intervening in the models and cutting back the bets: "...they sometimes think their judgment is better than that of their models, and, oddly enough, the more money the models make, the more inclined people are to override them. This sometimes happens to gamblers who become increasingly timid with each winning bet... Later, when they study the numbers, they realize that the portfolio would have made a lot more money if it had been allowed to run hands free." (Lots of Ed Thorpe-level insights here, lots to chew over in this paragraph)
* On management by walking around, explained by Jim Pelkey: "'Management by walking around' doesn't mean you're a friendly guy who pokes your nose in everyone's business. It works best when you perform a peculiar inversion. You will role play working for your employees. You sit down and ask them, 'What can I do to make your life more productive?' They may want you to run interference with Swiss Bank or buy them new tools, but employees get juiced on this kind of attention from management."
To Read:
Kevin Kelly: Out of Control (online copy here)
Ken Ustin: The Big Player
Doyne Farmer and Alletta Belin: "Artificial Life: The Coming Evolution" (1989 paper on human-made systems capable of reproducing themselves)
Robert E. Fink and Robert B. Feduniak: Futures Trading: Concepts and Strategies
***Mark Ritchie: God in the Pits
Richard Ney: The Wall Street Jungle
Frank Partnoy: Fiasco: The Inside Story of a Wall Street Trader
George Soros: The Alchemy of Finance
David Packard (Norman Packard's cousin): The HP Way: How Bill Hewlett and I Built Our Company