Customer Reviews

The Quants: How a New Breed of Math Whizzes Conquered Wall Street and Nearly Destroyed It

5 star:		(36)
4 star:		(28)
3 star:		(19)
2 star:		(11)
1 star:		(27)

 

 

Scott Patterson's _The Quants_ was thoroughly terrible. Patterson manages to make a dizzying array (to borrow a term he overuses) of errors, packaged in a mass of hyperbolae and confused statements.

It had a few good qualities, which I'll start with. It was pretty entertaining, especially the first half, and it was a quick and easy read. It also had some interesting bits that don't appear in other books (that I'm aware of): the "second forty hours" at Renaissance and the description of AQR deciding to go back into the markets on the Friday just after the quant liquidation in August 2007. Finally, I applaud the message that risk management policies based on the normal distribution can be deeply pernicious. But the problems with this book were monumental.

The first problem with Patterson's book is that it's wrong at its core. Quant traders weren't guilty of causing the credit crisis. Some of them were victimized by it (when Lehman went bust, it took with it a bunch of money belonging to some very good, honest, and hardworking quant traders that were Lehman's prime brokerage clients). It's foolish to claim that market neutral trading, CTAs, and high frequency traders were somehow responsible for investment banks' over-leveraged, toxic balance sheets. The responsibility for this falls squarely on the shoulders of banks' managers, and perhaps also on the shoulders of free-market disciples who believe, despite all the evidence throughout history to the contrary, that regulation of human behavior is bad. The crime in this is that it dramatically changes the focus from the real source of the problem that nearly buckled our economic system--namely unchecked greed, incompetent or impotent risk managers, screwed up incentive structures, and misguided regulation--to a group of traders that people are naturally inclined to hate anyway. If Patterson's disingenuous take on the credit crisis is widely adopted, it will make for a very convenient scapegoat enabling greedy, ego-hungry Goldman Sachs execs once again to make the very same kinds of bets that (at least nearly) brought them down to begin with. Did these execs use statistics to justify their position? Sure. But to make it sound like quants are somehow responsible for the stupidity or greed of their bosses who didn't (want to?) understand the weaknesses of a model is moronic.

Another fundamental problem with this book was the arbitrariness of Patterson's use of the label "quant." Whenever it was convenient (when it sounded evil), he labeled or insinuated the activity as being quant. But math is used pretty much everywhere in finance, and it always has been. Patterson:
- Treats the computation of a price-to-book ratio (P/B) as "value investing" but taking the difference in two interest rates (X minus Y) as a "quant carry trade". Why is subtraction "quant" and division "value"? Patterson also ignores the fact that the bulk of carry trading is done by discretionary traders, such as those in the global macro space.
- Confuses financial engineers, derivatives experts employed by the sell-side investment banks to create products like Principal Guaranteed Notes, Collateralized Debt Obligations, and compute VaR with buy-side quant trading outfits that are simply speculating their own, or their clients', capital in the markets alongside everyone else.
- Calls the belief that investors are rational a "quant theory," which is stupid. It's a basic tenet of economics and not a premise of quant trading.
- Treats the efficient market hypothesis as central to quants. By definition, quant traders believe the market is at least somewhat inefficient.
- Refers to capital structure arbitrage and distressed debt trading, respectively, as a though they are quant strategies. They're not. Cap structure arb is at the intersection of legal and accounting expertise. Deciding to buy a bunch of toxic assets from a company to which you already have lots of exposure (E*Trade) is not a quant trade either.
- Equates the move by banks to take huge risk off their balance sheets through tricky accounting practices with quants.
- Somehow treats Jerome Kerveil's very plain vanilla long equity futures trade as a "complex derivatives trade," which (for the author) puts it under the heading of quant. This was a fully discretionary trade that moved markets down by 8-9% as it was unwound.
Saving the worst for last...Patterson writes: "The quants were killing Bear Stearns." This is so foolish that it should make anyone with half a brain question his integrity. Because two funds with quant trading activities withdrew their funds' capital from a brokerage house rumored to be on the brink of failing, they are somehow quants killing a bank? Are the quants who trusted Lehman (and had their money evaporate as a result) called martyrs for the cause of our financial system because they kept their capital there too long? Is it a quant model that is responsible for the manager of a fund deciding it was a matter of common sense and fiduciary responsibility to move his cash to a safer haven? What kind of nonsense is Patterson trying to peddle here? This kind of arbitrary labeling is helpful for his rhetoric, but it's also garbage. In reality, quants are no better or worse citizens of humanity than George Soros (who was responsible for breaking the Bank of England in 1992 and maybe for bringing Asian economies to the brink of collapse in 1997) or Warren Buffett.

My second problem with this book is that it is poorly written. It is full of confused statements and errors. Patterson:
- calls diversification "quant magic" (p. 180)... what the hell?
- mistakenly refers to buying credit default swaps when in fact the transaction described is a sale, carrying this mental midgetry throughout the rest of the example and drawing wrong conclusions from it (pages 189-191).
- claims that "virtually the entire quant community...embraced the derivatives explosion wholeheartedly," (p. 192) which is pretty much the opposite of correct. The derivatives explosion also resulted in the widespread selling of volatility by banks, which itself was no small pain in the neck for quants (and other trading-oriented alpha-seekers).
- claims that the August 2007 quant liquidation was "making a hash of (mom-and-pop investors') 401(k)s and mutual funds." (p. 230) The quants that liquidated in August 2007 were market neutral. This means they held roughly equal quantities of long and short positions, and that they liquidated roughly equal quantities of long and short positions. The S&P was basically flat through this crisis, meaning that no one's 401(k) was being hashed.

The style of the writing reminded me of a cross between the National Enquirer and a Batman comic. Every one of the following phrases appears in this book, many more than once, and some countless dozens of times: "nerd king," "math whiz," "math wizard," "value king," "whiz-bang," "crack team," "it was nuts," and "whiz kid." Patterson also continuously used overwrought, mixed and confused metaphors, such as: "churning wheels of the Money Grid," and later, "tentacles of the Money Grid." On p. 197, he claims that the carry trade was a "frictionless digital push-button cash machine based on math and computers--a veritable quant fantasyland of riches." This horrible abuse of the English language is also hyperbolic nonsense. On p. 270, he likens investors in 2008 to "frightened children in a haunted house," a trivializing and wholly inappropriate description. On p. 273, a nonsense sentence appears: "...its hedge funds held about $140 billion in gross assets on $15 billion in capital, or the stuff it actually owned." He climaxed on p. 307, with this gem: "Lo's view of the market was more like a drum-pounding heavy metal concert of dueling forces that compete for power in a Darwinian death dance." That, I think, sums it up. I'd say I was disappointed that the press has adopted Patterson's deeply flawed views wholesale, but in reality, I guess I didn't expect any better.

The subtitle begins, misleadingly, with the word, "how." There is no "how" in the book. It is a beach read, an adventure story full of noir language, but short on words such as nonlinear, leptokurtic, nonparametric, adaptive. I think he uses the term "standard deviation" once.

If Scott Patterson edited The Joy of Cooking, his recipe for chocolate cake would read as follows:

"Irma Rombauer shuffled nervously. A little flour, a few eggs, and - wham, bam - through the magic of French chef-ery, a cake would magically appear. It had always worked in the past. Surely it would again this time, wouldn't it? Rombauer stared pensively at the oven."

Entertaining, but hardly enlightening, and not useful.

I heard an NPR interview pertaining to this book and immediately bought a copy. Many of the personalities woven into Patterson's tale are very intriguing, to say the least.

This book neatly retraces the influences of several quantitative traders ("quants", got it?). The author provides a history spanning the early work of Ed Thorp ("The Godfather") up to the current generation of quants who currently run the high-frequency trading strategies on Wall Street.

All in all, the book is a good read... but it also could have been assembled in a more informative way. The author set out a rather difficult task for himself: on the one hand, he has to tell the anecdotes in a way that will reach a wide audience; on the other hand, he has to provide a thorough enough treatment of topics that could easily be found in an advanced textbook. Patterson's approach goes right down the middle, so that sometimes it is condescendingly basic, while at other times unintelligibly riddled with market lingo. Hence most readers will not be able to read it at a consistent pace.

I suspect that the main frustration for most readers will be that extremely important bits are incompletely explained at the outset. For example, the distribution curve on page 30 has no axis labels... either you know what the author is talking about or you don't. Various terminology is not explained well at all; for example, the author's description of warrants will likely send you straight to wikipedia for clarification:

""Warrants are basically long term contracts, much like a call option, that investors can convert into common stock."

Basically? Call option? Common stock?! How about a more through lexicon at the back of the book for everyone who isn't a daytrader?! The small glossary that is provided won't be helpful to most readers. You'll definitely want to pair this book with "Trading for Dummies" if you aren't already up to the Cramer level of lingo. As an armchair daytrader, I still found myself re-reading certain key passages many times before they clicked.

Another minor annoyance is that effusive terms like 'genius' and 'brilliant' and 'whiz' etc. are used so freely that they completely lose their impact. Okay, the quants did well in statistics in some well-known Colleges; we got it! But real geniuses (like Mandelbrot) open the door to entirely new theory; quants program those ideas into their computers. There is a big difference, and the book's overall hypothesis is a lot less surprising if you understand that.

Nevertheless, the stories and anecdotes are very enjoyable, and the book does thread together quite well as a whole. Patterson does ultimately form an alarming hypothesis that will have you trailing your stops by chapter 14! I recommend it highly, with a few provisos.

James Cameron would be the perfect person to make the movie "Quants." They'd be great in 3D, with really big brains and really fast computers.

With their really big brains and really fast computers they would beat the market and lord it over lesser life forms. And it would all work until the market took revenge in a giant act of economic hara kiri that brought down the Quants and, very nearly, civilization as we know it.

But, since Cameron isn't available, you'll have to make do with Scott Patterson's book, Quants: How a New Breed of Math Whizzes Conquered Wall Street and Nearly Destroyed it. I chose it to review, as opposed to other excellent books on the economic crisis, because this is the book with the most lessons to teach.

The lessons I was looking for are human lessons. If you want to understand the investment strategies or the mathematics of quantitative investing, this is not the book for you. If you want to dig into the causes of the recent recession, you should buy a different book. But if you want a well-paced story of human folly, set in the world of investing, this book is worth reading.

Quants begins and ends with Ed Thorp. He was a math professor who was fascinated with using mathematics to figure out game strategies. That led him to gambling games, especially blackjack.

Thorp's first book was Beat the Dealer (1962). In it he showed how you could overcome a casino's house advantage at blackjack by using a card counting strategy.

The book was a success, but it wasn't long till Thorp was looking for a bigger casino. He applied his math skills to developing a strategy for investing in the stock market. In 1967, he wrote Beat the Market: A Scientific Stock Market System with Sheen Kassouf.

The strategies worked and soon the math professor was also a fund manager. Between the book and his own success Ed Thorp became the role model for what a bright young mathematician could do in the world of finance.

Graduate students of mathematics suddenly had something to strive for besides published papers and a tenured faculty position. All you needed to do, they thought, was use your super-power mathematical brain and find a giant computer to crunch your numbers.

In 1974, the Wall Street Journal published a front page article headlined "Trading by Formula." It outlined how some investors were using sophisticated mathematics and powerful computers to beat the market.

The academics provided help. The Black-Scholes Pricing Model gained a Nobel Prize for its developers. It also gave traders a recipe to use for investing in options that didn't depend on whether you thought the stock was going up or going down.

This is about the point in history where Patterson begins devoting his narrative to the people who ran the quant funds. You'll get good profiles of Ken Griffin, Boaz Weinstein, Peter Muller and Cliff Asness. There are briefer portraits of others.

For a while, things looked great. But, as we now know, that didn't last forever. When things began to unravel, the models simply didn't work that way all the theory and careful planning said they would.

The models were rational. The problem is that investors aren't rational and when they started acting irrationally, engaging in a "flight to liquidity" for example, the models broke down.

The models were based the idea that the market was efficient and that really smart investors with really fast computers could exploit temporary inefficiencies. That exposed them to real losses when the market didn't bounce back quickly enough for the model. As Keynes said, "In a crisis, markets can remain irrational longer than you can remain solvent."

And the models were based on historical records. That created two problems.

First, the models only had a limited amount of historical data. As Niall Ferguson said of the Long Term Capital Management's model, "They had plenty of mathematics, but not enough history."

And the models could only look backward. They simply couldn't predict things that hadn't happened before.

But, in the end, the real problem wasn't with the models at all. It was with the people who made the models and put their faith in the models and in themselves.

That brings us back to Ed Thorp. He got out of the market in 2002 because he saw too many people and funds showing up and acting like things would never go bad.

A quick reading of history will tell you that things always go bad at some point. And the quants did something no gambler or investor should ever do. They bet more than they could afford to lose. They bet that way because they were sure they would never lose.

In the end this is a book where the narrative is about investing and hedge funds and the troubles we've made for ourselves. But the story is as old as the Greeks, and probably older.

It's the story of what the Greeks called "hubris," extreme arrogance and overestimating a person's capabilities. For the Greeks, and for the quants, hubris was followed by Nemesis.

That's the important lesson. It's also one we have to learn again and again.

Scott Patterson's book, The Quants, is worth reading for lessons about how human beings have been proving their fallibility from Ancient Times down to the present.

That's the important lesson and it's worth the price of the book. But there are problems with the book, too. First, Mr. Patterson seems never to have met a cliché he didn't like.

Fact checking is a problem. Harry Reid, for example, is identified as a Senator from Arizona. He's from Nevada.

Things like that, combined, with some fast and loose uses of financial terms of art, make me wonder about Patterson's understanding of the investment vehicles and strategies he describes. He does not seem to have the comfortable understanding that other authors, like Andrew Ross Sorkin have brought to similar topics.

Bottom Line: The Quants is a good book for the human story of the quants themselves. It's a poor choice if you want to understand their methods or instruments or the mathematical and investment theories they used.

This book is among several that have been published in the last two years that attempts, without success, to lay the blame for the current financial crisis on the practice of financial modeling. These books are full of hyperbolae and exhibit an incredible naiveté on the role of mathematics in finance and in financial trading. Further, the authors lack the mathematical insight that is required to explain some of the ideas involved in financial modeling, and they frequently display an excess of veneration for the mathematicians that are hired to do modeling. It is not that the mathematical techniques used in financial modeling are difficult, some of them are and some are not. The issue is rather that the authors of these books cannot discuss them in a way that is understandable to a reader who is not familiar with them. And too often one can stand in awe of something that one cannot understand. The mathematicians themselves of course do not usually help in promoting understanding of their results.

A lot of this book therefore is just plain silly, and the writing incredibly sophomoric when discussing some of the mathematics used in financial modeling. As an example, the author speaks of the "Great Hedge Fund Bubble", characterizing it as a "true bubble" without defining precisely and quantitatively what a "true" bubble really is. He also refers to the use of quantum physics to "wring billions in profits from the market." No explanation is given as to how quantum theory is used in financial modeling, and its use would be surprising to those readers familiar with it. There has been discussion on "quantum finance" in recent years, but this work was not referenced in this book. Artificial intelligence is also said to have been used, but the author does not bother to delineate for the curious reader what ideas from this area were deployed. The author only gives a cursory treatment of automated trading, disappointing readers who, like this reviewer, are strong advocates of it.

The historical anecdotes in the book are somewhat entertaining, as are the brief biographies of some of the managers and employees of a few of the major financial institutions on Wall Street. These serve to make the book a little bit more palatable but they do not assist at all in giving the reader insight into how financial modeling "nearly destroyed" Wall Street. The personalities and idiosyncrasies of the people he discusses in the book may titillate some readers, but the space devoted to discussing them is a complete distraction from the author's main thesis. No doubt many of them were colorful people, and the type of people one would like to work with, but it is how they used algorithms and mathematical finance that is the most important, not how well they played poker.

Readers cannot expect of course that hedge funds will reveal what algorithms, trading practices, and mathematical constructions they are using, so the author's opinions and speculations in this regard are no better than anyone else's. Therefore it definitely remains an open question whether the mathematicians, traders, and managers of these hedge funds "controlled the ebb and flow of billions of dollars coursing through the global financial system every day" as the author states. It definitely remains an open question as to the efficacy of the mathematical techniques they used for generating profits or indeed if many where used at all. It definitely remains an open question as to whether financial modeling "conquered" or "nearly destroyed" Wall Street. From the information that the author has given in the book, one could just as easily assert that the profits where generated by a thousand edible frog farms in Brazil, or something equally as nefarious.

"The Quants" is an interesting read but I compare it to a movie about a historical event. You suspect a lot has been filled in and may not even be accurate [no documentation but then that might not prove anything anyway], but if interested in any details can probably check into further or compare to other stories in the future.
I had this feeling about the accuracy but when I read that James Simons went to Berkeley to study physics, I really wondered. His degree there was in Math as Dissertation Abstracts and Mathematics Genealogy would have easily shown.
The author is obviously in awe of mathematicians since almost every quant is referred to as a "math whiz" or similar description. After while you wonder if the author knows how to tell the difference between levels of knowledge and skill and/or has much of a mathematics background.
People who have been in the industry will wonder why so many names and firms that were equally successful, though smaller and proprietary and had very successful quants and strategies, were omitted---but then you can't cover everything.

Disclosure: I work as a quant, primarily in risk management. This is not the only book I have examined, written by an author who clearly does not understand mathematics well enough to render a qualified opinion, to pin the blame on quantitative professionals in finance. I have seen the reality of it first hand for myself, time after time, where the sales and marketing people want to get out there and SELL SELL SELL financial products involving derivatives, and when times are good, they have big celebrations about how great we are, and how much we sold. However, they often have marketed these products by going over the heads of quant/risk professionals who were not given the opportunity to perform due diligence, or who have issued red flags over questionable assumptions. When the market turns down, however, the company has to scramble to cover their liabilities and losses. Don't say we didn't tell you so.

Patterson writes with no delicacy or particular intelligence about managers who have made money with quantitative strategies. The book is not about the strategies themselves; with the exception of a few words about statistical arbitrage, Patterson is content to leave the money-making algorithms as magical black boxes. This is fine, except that the profiles are written with no more skill or subtlety than a Harlequin Romance, and the selection of which people to profile seems largely arbitrary.

Actually, it's worse than that, we know exactly how he chose which people to profile, and it was a bad method. Patterson found out about a regular Wall Street poker game, which he decided was a useful metaphor and used as an organizing scheme for the entire book. This means that he gives little or no space to successful quantitative managers who don't play poker (or who play, but did not play in the particular game he uses to launch the book), while he devotes chapters to managers who aren't even quantitative but who happened to be at the game. Ken Griffin is an interesting character, but he's no more a quant than dozens of other managers of multistrategy hedge funds that happen to have quant divisions (Millennium, HBK, SAC, etc.)

To demonstrate how poorly written the book is, I offer the following excerpt, from a discussion of securitization on page 179: "One reason why banks engage in securitization is to spread around risk like jelly on toast. Instead of lumping the jelly on one small piece of the toast, leaving all the reward (or risk that it falls off the toast) for one bit, it's evenly distributed, making for lots more tasty bites -- and, through the quant magic of diversification (spreading the jelly), less risk."

In addition to being gross, that's just plain wrong. It's wrong about securitization, and it's frankly also wrong about jelly. Who spread around jelly in order to reduce the risk that it falls off the toast? Does that even make sense? Banks engage in securitization because (a) it enables them to remove risk from their books, so they can do more deals, (b) it allows them to sell complex assets to buyers who would not otherwise consider them, thus (c) generating fees. The jelly analogy is utterly inept.

Patterson tries to coin a few phrases here, but none of them will stick. Rather than saying "alpha" or "excess returns," he says "the Truth." Rather than saying "the markets," he says "the Money Grid." Both seem lame every time you see them.

At various points he flails about, including a few paragraphs or even a whole chapter on an unrelated topic that, I suspect, happened to be on CNBC as he was writing. Why are private equity funds discussed in this book at all? A chapter on dark pools, suggesting they are somehow evil? If Patterson learned anything about quant funds while writing this book, he would know that a major focus of their efforts is detecting the trading activity of large institutions, so they can trade ahead of it. This increases slippage for these large institutions, and thus for their investors. Effectively, it transfers money from pensioners and mutual fund investors to the hedge fund managers. We should *applaud* any market structure that allows these institutions to hide their activity from Renaissance, and D.E. Shaw, and Citadel, etc.

I can't think of anybody who would benefit from reading this book.

The author is a staff reporter for The Wall Street Journal, and it's clear that he has done a good deal of research for this book, which describes the history and the (sometimes colorful) characters behind the growth of quantitative methods in hedge funds and throughout Wall Street. The book begins and ends with the Wall Street Poker Night Tournament at the St. Regis Hotel in Manhattan--specifically, the tournaments of 2006 and 2009, where many quants would seek to strut their IQs. (Have you ever noticed that Wall Street quants seem to love poker, while fundamentally-oriented investors like Warren Buffett seem to prefer bridge? One possible reason is that compared to bridge, poker is much more easily characterized by math--specifically, math that computers and sharp people can master.) A lot happened between 2006 and 2009, of course, but before the author spends much time discussing all of the market turmoil of the last few years (and the role quantitative investing had), he spends perhaps half the book discussing the relevant background of quant methods, which dates back to the 1960s, 1950s, or 100 years ago, depending on how you look at things. Digesting this history, the reader will likely conclude that big events (in numerous disciplines) seldom come out of nowhere--they develop over a long time, and once a catalyst occurs, they can present themselves with full force. And so it is with quantitative investing. So it is entirely appropriate that author Patterson devote the time necessary to present this important background.

Well before the trading turmoil of 2007 - 09, there was the one-day market crash of 1987, which was caused in part by a disastrous overuse of "portfolio insurance." Patterson describes how many of the academics who promoted portfolio insurance didn't realize how transactions in derivative markets, such as stock index futures, could essentially flow into the cash markets--where most investors see them. This general problem would grow greatly in the 2007 - 09 experience, so it is important to review the 1987 episode.

It doesn't take the reader much insight to generalize that one of the fatal flaws of the quants was that while math can be useful to help researchers understand some of the complexities of the economic world, economic reality is not a purely deterministic endeavor--and, especially when people act in unusual ways (typically during times of extreme stress), seemingly regular market patterns aren't quite so regular. Indeed, some quants, such as Benoit Mandelbrot (discussed in the book) saw the limits to some of the quants' simplifying assumptions. That's another key point: Simplifying assumptions are just that--assumptions. People, whether they be Joe Six-pack or the head of the Fed, reserve the right to change their behavior. The fact that they don't do so frequently is what makes math models seem to work, and the fact that sometimes people do change behavior (perhaps from that suggested by raw logic to that described more by behavioral finance) makes the use of models more art than what many of the quants ever believed. Quants may assume that they don't really need to know much about "business fundamentals," because these fundamentals are already reflected in market prices. And often they are. But not always, which is why a non-quant like Warren Buffett can be so successful.

So, go ahead and read the book if you want to know more about the rich and interesting history of quant theory, with plenty of spice in the form of descriptions about some of the more colorful players. The seduction of math can be that it is sufficient to see through the complexity of the modern world into the depths of economic reality. The disruptive truth is that, while math certainly has its place, without a good understanding of the limits of math, you can go wrong. Very wrong.

The only thing more boring than playing poker is reading Scott Patterson's description of Wall Street guys who like to "butch up" by playing poker. That said, I am enjoying the book despite how badly written it is.

(Btw, I'm surprised it never occurred to Patterson that according to the Kelly criterion to which he attributes the success of Princeton/Newport Partners, you only go "all-in" when you have p=1 of winning. BUT, in poker, you gotta go "all-in" when p is not equal to 1. Ergo? Does poker induce brain damage in hedge fund managers?)

What makes the book enjoyable is the subset of quants he picks and the background story of their development and success.

Hitting page 180 though and having to suffer his scatter-brained description of CDOs is even worse than his obsession with the poker playing habits of his protagonists.