A Demon of Our Own Design: Markets, Hedge Funds, and the Perils of Financial Innovation [ハードカバー]

In recounting his time as risk manager at a number of prominent houses (Morgan Stanley, Salomon Brothers, Citigroup etc.), Bookstaber completes the i-banking trifecta. First there was the Michael Lewis classic, Liar's Poker, detailing the juvenile bravado and macho antics of the trading floor. Then Jonathan Knee gave an intimate portrait of the i-banker deal making culture with The Accidental Investment Banker.

And now, in A Demon of Our Own Design, we get a glimpse at the risk management side of things... a sort of master plumber's walking tour through the bowels of the system, with technical descriptions of exactly what happens when pipes burst and boilers explode. (Some will find Bookstabers' level of detail intolerably dull; others will find it quite fascinating. I was in the fascinated camp.)

Nature of the beast

In describing the finer points of risk arbitrage, Bookstaber explains why it's normal -- expected even -- for trading desks to take a good whack every so often. The nature of the beast is to make relatively steady profits, month in and month out, and then give back a chunk of those profits when something goes haywire. (That's how you move huge sums on an arb desk; grind out small bets that are almost guaranteed to work, juice up the returns with leverage, and try not to be in the vicinity when the rare position goes kablooey.)

In light of this general modus operandi, perhaps it isn't surprising that the "quant" funds recently took a major hit (as of September 2007). They had been minting money for an extraordinarily long period, had the leverage to show for it, and now, after the recent "oops," seem to be generally back in business.

In fact it appears natural for much of Wall Street to work in this "make a little, lose a lot" fashion... the key idea being that all the little updrafts make up for the once-in-a-blue-moon downdrafts. (Such calculus works better for the fee collectors than the fee payers, but that's a different kettle of fish.)

Bookstaber's detail-rich description of the various trades that investment houses put on, many of them lasting years, is also enlightening. The details seem to confirm that, by and large, Wall Street is a gigantic, slow moving, conventional-returns type machine. (And what else could it be, really, with such an ocean of capital to allocate and so many jobs to fill? There is only so much creativity and contrarianism to go round.)

A dangerous combination

Risk manager war stories aside, Bookstaber's goal is to hammer home a key philosophical point regarding risk. He wants readers to understand that financial markets are inherently unstable, and this reality places limits on how far we (or anyone) should go in pursuit of outsized returns.

To make his point, Bookstaber uses various analogies to describe how the market is a highly complex, tightly coupled system... and to explain why the combination of high complexity and tight coupling is particularly dangerous.

The counterexample Bookstaber gives of a highly complex, loosely coupled system is the US Postal Service. The USPS has countless potential points of failure and myriad moving parts, but there are no catastrophic linkages involved. A lost package does not set off a disastrous daisy chain of events in which millions of packages are lost.

In contrast, the classic example of a highly complex, tightly coupled system is a nuclear reactor. The reactor is tightly coupled because any point of failure can lead to a knock-on chain reaction; one small thing going wrong can set the entire mechanism on a path to disaster. Being a highly complex, tightly coupled system, the market is less like the postal service and more like the nuclear reactor, in that the combination of aggressive leverage, complex methodologies and heavily interlocking parts leads to significant potential for catastrophe.

Exquisitely adapted

Another serious problem is Wall Street's deeply ingrained tendency to push the envelope. (Richard Lowenstein put it exceptionally well in his book Origins of the Crash: "Finance has its own Peter Principle, by which a successful model will be adapted to progressively riskier causes until it fails.")

In this habit of fighting for every inch of profit, Wall Street is like a self-evolving animal overquick to embrace the particulars of its immediate environment. The more precisely an animal is attuned to a particular "fitness landscape," the better that animal can thrive... in the short term at least, as long as everything stays just so. To be exquisitely adapted (as opposed to robustly adapted) is to be vulnerable to the slightest change.

Thus when the fitness landscape DOES change -- as it inevitably will -- the heavily specialized competitors tend to get crushed (if not go extinct). If a strategy-gone-sour broadsides a large enough group of market participants, the entire financial ecosystem can be thrown into turmoil. When the turmoil from this upheaval spills into the broader economy, wreaking havoc in its wake, the "demon" spoken of in the book's title is unleashed. (As this reviewer interprets it anyway.)

Wisdom of the cockroach

So the problem, in sum, is Wall Street's tendency to `overadapt' to every appealing landscape it encounters, building up complexity and leverage to dangerous levels in doing so.

Bookstaber's suggestion is to heed the wisdom of the cockroach.

The cockroach has survived a longer time span, and a wider variety of harsh environments, than humans could ever match. It is one of the creatures man cannot wipe out no matter how hard he tries. And yet, the cockroach's key risk management strategy is embarrassingly simple... simpler, even, than putting in a stop loss. The deeper point is that simple equals robust; by refusing to get fancy, and sticking with the tried-and-true, the cockroach ensures its reign as champion survivor.

Bookstaber uses the cockroach (and other examples from nature) to argue that we, too, should consider cutting back on our excessively specialized ways. The cost of a rough-edged strategy is forgoing excess profits in accomodative environments... but the benefit is increased likelihood of survival in a much wider range of environments, including the truly harsh ones. (As Jim Grant likes to joke, if so many of these credit-driven vehicles can barely handle prosperity, how are they supposed to fare when adversity hits?)

Harrumphs all round

Bookstaber's finger-wagging solution (be less fancy; take less risk) has the ring of common sense to it, especially in the way it frustrates all those market participants determined to have their cake and eat it too.

For those who seek to wring every last nickel out of the market (as LTCM used to brag of doing), Bookstaber argues persuasively that flying too close to the sun will always be perilous. The commitment to leveraging every edge on a broad scale inevitably leads to disaster-prone configurations, no matter how smart the players.

For those who think the answer is greater regulation of markets, i.e. more rules, Bookstaber shows how extra layers of bureaucracy can actually bring about the exact opposite of the intended affect. Perversely, layers of red tape can (and often do) make a situation more risky, by increasing confusion and complacency simultaneously.

Nor is greater information disclosure the answer. If the market's traditional liquidity providers (traders, market makers, speculators etc.) are forced to disclose their positions to the world in real time, they will react in the manner of poker players forced to play their hands face-up. To the extent that disclosure resolves uncertainty, it also drives market participants from the game. And because "liquidity is a coward" as the old saying goes, always running away when you need it most, strict disclosure rules would likely make bad market conditions worse at the least opportune times.

Some left smiling

Two groups in particular may be left smiling at the end of this book -- value investors and trend followers. In both the theory and practice of their normal operations, value investors and trend followers intuitively embraced Bookstaber's message a long long time ago, favoring longevity and robusticity over the temptations of adjusting to the moment.

It is perhaps not surprising, then, that value investors and trend followers are arguably the most profitable market participants by far on an absolute-dollar basis, hauling in hundreds of billions in profit over the course of many decades. They are champion survivors too... with a touch more class than the cockroach.

This is an entertaining account of 25 years of financial disasters by a smart insider who is also a keen observer and witty storyteller. Some authors take simple ideas and cloud them with hairsplitting definitions and complex equations, real mathematicians like Bookstaber effortlessly work in deep concepts like sufficient statistics and state variables without any equations or formal mathematics.

The best part of this book is the context. If you read books on individual disasters the situations come across as complex and the people as tragic geniuses. If you read the one paragraph versions favored by the business press, the underlying trades seem impossibly simple, and the protagonists seem to be morons. This book shows the euphoria of winning trades when money flows in like magic, and the confusion and shock that result from unanticipated losses. There are many reactions to these losses: cut them quickly, ride out the storm, even double up the bets. Each of these sometimes works and sometimes makes the situation worse. Only after reading all the permutations and outcomes can you understand the stark choices posed as the disasters unfold. The players are neither idiots nor geniuses, they are smart but ordinary people, facing understandable human dilemmas.

This context is precisely what is missing in the chapter on engineering disasters. If you look only at disasters, when by definition all the safety precautions failed, it's no surprise that you'll conclude safety precautions are worthless. If you only look at the most dramatic disasters, it's no surprise you'll conclude that the most ambitious, advanced, complex and tightly-coupled systems are the most prone to catastrophe. Bookstaber relies on writers I call the dismal engineers. I think optimists like Duke civil engineering professor Henry Petroski have more to say to financial risk managers. Petroski wrote "no one wants to learn from mistakes, but we can't learn enough from successes to advance the state of the art."

The chapters on biological, mathematical and quantum mechanical limits on rationality are interesting speculations. Bookstaber appears to know, or perhaps to care, more about these fields than about engineering. However even in these cases his thesis is not entirely convincing within the realm of discussion, nor is the analogy he wants to draw to finance compelling. There's another good book (or three) in these ideas, this book gives only a taste of the arguments.

Despite being a very smart guy with a quarter century of experience in cutting-edge trading, Bookstaber cannot overcome the disadvantages of being trained as an economist, especially an MIT economist. In the final analysis, he believes risk is bad. Trading is defended as socially useful when it provides liquidity, when traders exploit pricing discrepancies caused by short-term supply and demand forces. These traders stabilize the market. But at least half of trading is trend-following, which exacerbates pricing discrepancies and sucks up liquidity, destabilizing the market. To a University of Chicago finance guy, these are symmetrical market forces, both good. Bookstaber correctly points out how financially-supplied liquidity ushered in the Industrial Revolution by converting frozen wealth to dynamic capital. But he doesn't mention the momentum traders who forced prices quickly to their eventual equilibrium, sweeping aside those who try to stand in the way of progress. This is the creative destruction of capitalism. Economists are comfortable with the markets facilitating real economic decisions, spreading the inherent risk among willing investors. They are less comfortable with the market forcing real economic decisions, creating virtual risk and imposing it upon unwilling actors.

This anti-risk bias shows in the account of Tulipmania and the reflexive comparison of financial disaster to engineering disasters. No one died at Long-Term Capital Management. It's true that some financial crises, like the 1997 Asian Crisis, impose human costs, but financial crises are virtual, and therefore less harmful than the same degree of failure with physical economic experimentation, and far, far less harmful than the retarded progress if experimentation is slowed. Bookstaber celebrates the liquidity provided by a shrewd swimsuit retailer who anticipates fashion and supplies her customers' demand without excessive price increases. Nowhere does he mention the businesses that create fads, including swimsuit fashions, and thereby the need for liquidity providers. Economists like businesses that respond to exogenous consumer preferences, they have trouble with businesses that manipulate consumer preferences. But both are needed for successful dynamic economies.

None of this takes away from the fact that this is an entertaining and insightful book, essential reading for people interested in financial risk. I disagree with Bookstaber's conclusions calling for slowing market innovation and trading, but I think these are preconceptions from his early training rather than the result of his experiences since.

Extremely well-written and thought-provoking. The author has a wide range of interests and knowledge, from yield curve behavior to cockroach survival traits, and explains them all lucidly, simply and in an entertaining and practical manner. He has a deep understanding of the workings of the financial markets, and shares several unique perspectives in this book which I have not read elsewhere, and it is extremely valuable for that reason alone. He is one of those rare geniuses who can keep his autobiographical urge to an interesting, useful and entertaining minimum, only mentioning his personal experiences when they provide insights into larger themes (cf. Black Swans). The author does occasionally use the book as a platform for payback to former colleagues who have done him wrong, but this is done with a stiletto, not a blunderbuss, and is fun to watch. Always well-written, occasionally entertaining. Very highest recommendation.