October 5, 2008

Wall Street Quants and the inherent failures of risk management

Obviously, there has been a gigantic failure by Wall Street rocket scientists at "risk management." This isn't my area of expertise, but I think I can point out a basic mistake. Going back decades to Burton Malkiel's book "A Random Walk Down Wall Street," sophisticated financial thinking (e.g., the "efficient markets theory" of the 1970s) has been dominated by the concept of randomness: events are distributed on a bell curve-shaped probability distribution.

For example, to take a simplistic example, people tend to default on mortgages when they have bad luck: dad gets cancer and dies and then mom gets depression and loses her job. The bank forecloses. If you hold 1,000 mortgages, that kind of bad luck happens to, say, 15 each year. Of course, your 1000 people might have worse luck than normal. Say you study millions of mortgage and determine the standard deviation is 5 per 1000. So, for 99.75% of the bundles of 1000 mortgages, the number of defaults in a year will range from 0 to 30.

Here's the problem: human life really isn't all that random. That's because human beings respond to incentives. If you treat human beings as if they are just mindless probabilistic events, whose risks you can diversify away by dealing with large numbers of them at a time, they will outsmart you. They will put down inflated incomes on their mortgage applications. They will claim to be owner-occupiers when they are just speculators who will rent out the property to Section 8 tenants when they get into a cash flow bind. They will bribe appraisers to report a higher than actual value.

Another common pattern in life is that things build to a climax of the greatest risk and reward, where predictability is at a minimum. The events that we are most interested in are those that are hardest to predict. We know exactly when the sun will set on December 21, 2008. That is a hugely important fact, but it's not a very interesting one to us because it has already been taken into account. We're more interested in things like who will win the World Series or be elected President or whether the stock market will go up or down ... because those are so hard to predict.

Let's look at a sports example of risk vs. reward. Say you are an Olympic boxer, one of 32 contenders in your weight class. You have a particular power punch that you are fond of which requires you to drop your defenses for a fraction of a second as you wind up to deliver it. In the first round, against a boxer from Sikkim, you throw it seven times with no bad results for you. In the second round, against the Ghanian fighter, you use it five times with no ill effects. In the third round against the Slovenian fighter you throw it six times and suffer one glancing blow. In the semifinal round against the Korean boxer, you throw it seven times and suffer two glancing blows.

Okay, so, in the first four matches, you've thrown it 25 times and suffered three glancing blows. Only a 12% problem rate, and those problems aren't that bad: just glancing blows. You run a 1000 Monte Carlo simulations, and using that punch pays off in 973 of them. You like those odds!

Now you are in the final against the Cuban, who is the World Champion and defending Olympic gold medalist. You immediately rear back to throw your power punch ... and wake up in the infirmary with your silver medal on the bedside table.

What happened?

Non-randomness. The whole Olympics were set up to pit the two best boxers in the final round. The Cuban, who might be the professional champion of the world if he were allowed out of Castro's paradise, is just plain better than anybody you fought before. In hindsight, you can see a trend in the data but you simply couldn't predict from it how hard you'd get hit.

A lot of things in real life work out roughly along the same lines as in organized tournaments, building to a climax. First, Hitler conquers Czhecoslovakia, then Poland, then Denmark and Norway. So, feeling lucky, he invades France. Then in 1941, with all that positive data on the high rewards and low risks involved in starting wars available to him, he invades the Soviet Union and declares war on the United States. Notice a pattern?

In retrospect, things tend to evolve toward maximum unpredictability. The pre-1914 alliance system of Europe tended, because the weaker side at any point had the incentive to offer more to neutrals to join them, tended to evolve

This doesn't mean that this has to happen. There are lots of periods without that kind of disastrous evolution. But things like World Wars or financial crashes are what catch our eye in hindsight, and with good reason.

This suggests that there is no way to avoid disasters permanently. That's no doubt true. But we can make them rarer and less catastrophic just by being less stupid. Consider two economies, both of which either grow 5% per year or shrink 5% per year. The first economy is more bubble-prone, so it grows for seven years then shrinks for three years. The second economy grows for ten years, then shrinks for two years. Over the course of sixty years (six cycles for the first economy, five for the second), the second economy will end up over twice as big.

The Albanian economy collapsed in 1997-98 due to the entire population, who had only been introduced to capitalism less than a decade before, becoming entranced by simple pyramid scams. That was so stupid that it probably won't happen again in Albania for a long time.

Americans don't fall for simple pyramid schemes. We need more complicated scams.

My published articles are archived at iSteve.com -- Steve Sailer

24 comments:

  1. I think the Wall Street boys understood your point, up to a point, but got into a position where they could not defend against the threat adequately.

    First, pooling and reselling bundles of bad and good mortgages protected againts exceptional losses on one or a few mortgages, as you note, so they did that first.

    But default rates and (especially) default costs would depend on the realism of housing prices, relative to price and real income trends. So when investment banks spotted big potential problems here, they started doing default hedges and swaps to protect themselves on this front.

    Initially, they did what is normal in executing hedges, check out counter-party risk, that is the risk that the hedge partner will not be able to make good on its commitment if it has to. That means looking at the maximum losses the hedge partner can sustain from all its exposed positions, not just your own.

    The investment banks eventually found themselves dependent on too many partners that could not make good on their hedge commitments, especially when oil prices joined rising default costs and slowing economies.

    ReplyDelete
  2. Everything you write is well founded and interesting. I would like to know what the global shakeout of the meltdown is going to be? 1929 again, in slow motion this time? Or a realignment of global power?

    ReplyDelete
  3. Why are you blaming the quants? The real idiots are the customers. You would think that the customers would know better than to buy structured paper. Structured paper is a bad deal from the start due to its 2-3% markup. Why buy pay the markup when other instruments don't have have such high transaction costs?

    Bottom line: investment bank customers are idiots.

    ReplyDelete
  4. I like the boxing analogy, because at its root what drives this evolution toward unpredictability is competition, or at least ambition.

    However, the disasters are fully predictable if one can only see the trend building. When there is no containment of this competitive impulse it grows until it can't sustain itself anymore. Maybe we need a new Malthusian explanation for finance and politics.

    ReplyDelete
  5. Sikkim is not a country!

    ReplyDelete
  6. An interesting post. This might also be why "moneyball" type baseball teams built around probabistic sabermetric thinking do so well during the regular season, but often fail in the playoffs and Series. Or not, but it's certainly something to think about.

    ReplyDelete
  7. Your saying that things aren't random is equivalent to saying that the underlying model is wrong. So the use of the concept of increasing "randomness" is off the mark. The "right" models would predict the stupid behavior.

    You may want to examine the concept of "Melioration" (this actually came from Richard Herrnstein of the bell curve fame from his work on animal learning) as it applies to choice across time and it critically examines the role of incentives. The original paper appears in J Econ Perspectives by Herrnstein and Prelec but hasn't picked up that much interest. A recent abstract from another paper on this topic reads:

    We examined how people allocate choices between two alternatives when the payoff from each alternative varied as a function of the allocation of recent choices. On any one trial alternative A had a higher immediate payoff than alternative B, but across all trials B had a higher overall payoff than A. Rational choice theory requires that participants allocate all their responses to the alternative with the greatest overall payoff irrespective of which has the higher immediate payoff. Melioration, in contrast, proposes that participants are motivated to choose the alternative with the higher immediate payoff, irrespective of the consequences for future returns. We report four experiments in which we varied the nature of the payoffs. Participants exhibited self-control consistent with rational choice theory when payoffs varied in magnitude, but exhibited impulsiveness consistent with melioration when the payoffs varied in probability. Finally, we show that impulsivity when payoffs varied in probability can be overcome following un-reinforced practice.

    ReplyDelete
  8. Have you started to read Nassim Nicholas Taleb?

    ReplyDelete
  9. Burton Malkiel's book "A Random Walk Down Wall Street," sophisticated financial thinking (e.g., the "efficient markets theory" of the 1970s) has been dominated by the concept of randomness: events are distributed on a bell curve-shaped probability distribution.

    I recommend The Black Swan for a look at an author, Nassim Nicholas Taleb. who takes the notion that we cannot predict damn near anything to its ridiculous conclusion.

    The event which sparked the ideas that Talib outlines in the book was the Lebanese Civil War. Talib says the people of his native Lebanon were shocked (SHOCKED!) that Lebanon could descend into chaos. If the Lebanese themselves couldn't have seen it coming, then who, he wonders, could have predicted that Lebanon might be such a powder keg? The randomness of the event is mind-boggling!

    Apparently, we are supposed to believe that because the Lebanese didn't see trouble brewing, that no one could have predicted Lebanon might be more volatile than, say, Finland.

    ReplyDelete
  10. Good analysis. The additional point is that the default probabilities are not independent, but linked.

    ReplyDelete
  11. Nassim Taleb had a somewhat similar discussion in The Black Swan. He relates how The Mirage casino conceptualized "risk" by focusing on the house's odds in its games of chance. As things developed, its big problems were:

    1. Siegfried and Roy's tiger mauled Roy Horn, sinking the casino's premiere show;

    2. A disgruntled contractor attempted to dynamite the building;

    3. An erratic employee neglected to file forms with the IRS documenting gamblers' winnings; and

    4. The owner dipped into the casino's reserve to ransom his kidnapped daughter.

    Wall Street had an excellent handle on the risks of mortgage-backed securities, all right. The Quants even paid S&P and Moody's and Fitch to quantitate and categorize them (AAA, AAa, etc.). The problems weren't in the calculations. They were in the assumptions behind the calculations.

    "Gee, it's so complex and so mathematical and statistical! And these Wharton grads and math PhDs speak with such assurance! I can't understand all this stuff, so what a relief that they do. Let's add another AAA mortgage-backed security to the portfolio."

    ReplyDelete
  12. The Albanian economy collapsed in 1997-98 due to the entire population, who had only been introduced to capitalism less than a decade before, becoming entranced by simple pyramid scams...Americans don't fall for simple pyramid schemes. We need more complicated scams.

    No, we fall for only slightly more complex pyramid schemes, like the one that says we can keep importing uneducated immigrants ad infinitum, turn them culturally educationally economically into European Americans, and we'll all get rich and import more immigrants to start at the bottom.

    It's basically the same attitude, only a little less obvious. And living (temporarily) in Utah, I can assure you that even the simple pyramid schemes still manage to sometimes succeed. The current model seems to be that we're all going to get rich selling $40 bottles of health drinks to each other, using the juice of fruits that no one can even patent. "Of course that berry's juice is perfect for your health - just look how successful the Nepalese are!"

    You can substitute the Nepalese for Pacific Islanders or Amazonian tribesmen - all groups known for their intellectual prowess and long lifespans.

    ReplyDelete
  13. Oh, and I would add to above that Utah buys heavily into both types of pyramid schemes. That's not, I think, a coincidence. Our politcians are as supportive of mass immigration (in state tuition, driver's licenses, etc.) as our people are supportive of pyramid schemes.

    ReplyDelete
  14. Steve --

    I think you are not recognizing the underlying political struggle that created the crisis. The new elites, stagnant since around 1986 or so, when the generation of the 1960's first started to run things (the awful Boomers) have engaged in a pure class struggle.

    The Boomers want finance, law, media, and entertainment to dominate the economy, and resort to bubble speculation to create this. Deliberately shorting manufacturing, mining, oil/gas, etc. It's the struggle between the cool/hip new "knowledge" workers vs. the grimy industrialists. New School vs. Old School with the New School winning.

    So of course we resort to various tricks, including importing scads of illegal aliens, and so on. PC and multiculti up the wazoo. Naturally, this also fits the desires of a growing class of young single women for status/power in urban settings, a global phenomena captured by City Journal's Kay Hymnowitz and others. Europe is in the same boat with the same New School stagnant elites running things for the same results.

    After all, who would a hot young thing want to hook up with? Some nerdy chemical or mechanical engineer, or a hip, happening media executive who knows celebrities?

    Don't underestimate THIS angle in how certain industries/occupations/scams get favored over others. Steady, boring growth might get you a better economy, but you'll have a bunch of nerdy, celibate engineers wondering why they can't get dates. Over the hip, happening, celebrity driven bubble economies elsewhere where girls are everywhere.

    ReplyDelete
  15. The "Wall Street rocket scientists" most likely believe, like all 'serious' economists, that the business cycle is a natural phenomenon, not man made. Freidman came up with the famous idea that booms and busts are as natural as winter and summer.

    Real Theories Of Genius: Here's to you Mr. Chicago School Economist.

    The fundamentals are strong, yes they are, yeah!

    ReplyDelete
  16. Wall Street had an excellent handle on the risks of mortgage-backed securities, all right. The Quants even paid S&P and Moody's and Fitch to quantitate and categorize them (AAA, AAa, etc.). The problems weren't in the calculations. They were in the assumptions behind the calculations.

    The role of the bond rating firms is underreported. They were suppose to provide an independent judgment of an investment's risk of default and instead basically acted as the mortgage industry's PR firms.

    What's worse is, they judged state and local bonds by a tougher scale. These lowballed ratings meant local taxpayers were stuck choosing between lower infrastructure spending or higher interest payments.

    CA Treasurer Bill Lockyer pointed out that the historical default rate for BBB rated municipal bonds (.32 percent) is HALF the default rate of AAA rated corporate bonds (.60 percent).
    http://www.sacbee.com/110/story/838473.html

    Since many investment funds can only hold AAA bonds, a lot of money that would have gone to financing bridges, schools and highways was steered by S&P and its peers into feeding the mortgage bubble instead. Was this greed, stupidity or Grover Norquist's "starving the beast" in action?

    ReplyDelete
  17. "In retrospect, things tend to evolve toward maximum unpredictability. The pre-1914 alliance system of Europe tended, because the weaker side at any point had the incentive to offer more to neutrals to join them, tended to evolve

    This doesn't mean that this has to happen. There are lots of periods without that kind of disastrous evolution. But things like World Wars or financial crashes are what catch our eye in hindsight, and with good reason."


    Word processing error? Paragraph ends in mid sentence.......

    ReplyDelete
  18. Two books that have shed light on the subject for me - "Fooled by Randomness" by Taleb and "The Misbehavior of Markets" by Mandelbrot. Both of these authors point out that the assumption of normality in the markets is unwarranted. Normality requires independent observations, which we don't have in the markets as you point out. Perhaps better models would be those from epidemiology, which tracks the spread and transmission of diseases.

    ReplyDelete
  19. As Buffet recently said, "Beware geeks bearing formulas."

    ReplyDelete
  20. Steve, it's all much simpler than that.

    The "rocket scientists" were just that -- they knew NO economics.

    Worse, there professors knew no economics -- they were 4th rate math jocks with degrees that said "economics", but that's about as far as the understanding of economics went.

    KEY INSIGHT:

    The artificial boom / inevitable bust cycle is a SYSTEMATIC spontaneous dis-coordination in the economy -- this systematic disequilibrium dramatically shifts the nature of all risk, especially in systematically effected sectors.

    Read some FA HAYEK or some ROGER GARRISON.

    If the math jocks had understood any of this, they would have known their "models" were non-economic fairy tales.

    The math models of the modern macroeconomists and finance economists is BRAIN DEAD from the point of view of intertemporal coordination of plans and relative prices across the time and production structure of the economy.

    IT'S PSEUDOSCIENCE -- the misapplication of statistics to phenomena that has been mischaracterized, and NOT understood.

    Really.

    ReplyDelete
  21. As a minor point of reference, Steve, the Olympic Boxing Tournament has been bracketed by a random draw since the 1980's. The Soviet and Cuban fighters in any weight class are as likely to square off in the first round as a Barbadian and a Dominican.

    ReplyDelete
  22. Also check Hyman Minsky, who was a follower of Keynes (who himslef wrote a lot about the difference between risk and uncertainty, and the perils of confusing them) and thus was flushed down the memory hole by the Chicagoites until recently. He came up with the memorable phrase "In markets, stability breeds instability."

    http://en.wikipedia.org/wiki/Hyman_Minsky

    ReplyDelete
  23. headache: I would like to know what the global shakeout of the meltdown is going to be? 1929 again, in slow motion this time? Or a realignment of global power?

    What we are beginning to experience is not 1929AD, but rather 450AD.

    We are in the throes of the greatest collapse in the population of the civilized world since the fall of Rome:

    IQ and the Wealth of Nations
    en.wikipedia.org

    List of countries and territories by fertility rate
    en.wikipedia.org

    I'd suggest reading Spengler, over at the Asia Times:

    Why nations die
    atimes.com

    ...Roman population data are somewhat conjectural, and Strabo's estimates have been disputed by some scholars. Explanations have been forwarded that range from the collapse of the slave-based agricultural system to mass infanticide and venereal disease.

    Nonetheless, it seems clear that the Romans did not so much conquer Greece as to occupy its shell; that the Germanic tribes did not so much conquer Rome so much as to move into what remained of it; and that the Arabs did not so much conquer the Byzantine hinterland as migrate into it. On this last point, a new book by Yehuda Nevo and Judith Koren argues convincingly that the Byzantines ceded frontier territories to Arab foederati in the mid-seventh century and that the famous battles of the Islamic conquest in fact never took place. [3] In one form or another the antecedents of Western civilization died of existential causes, rather than external ones...


    What we had in the subprime crisis was a huge bulge of Caucasian Baby Boomers, in their early- to mid-fifties, at the height of their earning power, with literally trillions of dollars which they needed to invest, but because they failed to breed at replacement levels, back when they were in their twenties, they had no young Caucasians to whom they could loan the money, and hence they had to loan the money to very low IQ minorities, which, as anyone could have predicted, a priori, proved to be a disaster:

    USA Caucasian Population By Age

    And, in case you're wondering, this subprime mess is just the tip of the iceberg: Circa 2020, when those Caucasian boomers go [literally overnight, as they pass through their 65th birthdays] from paying tens of thousands of dollars per year in taxes to consuming tens of thousands of dollars per year in Medicare & Social Security giveaways, the USA will implode and cease to exist.

    Of course, as the fertility numbers above show only too clearly, this very same phenomenon will strike every other nation in the civilized world, many of which [such as e.g. Hong Kong and Japan and Italy and Spain] are already [even now] just about effectively extinct.

    ReplyDelete
  24. "We are in the throes of the greatest collapse in the population of the civilized world since the fall of Rome:..."

    Lucius Vorenus, you know what you have been chosen to do!

    It's time you got that little tallywhacker of your moving son!

    ReplyDelete

Comments are moderated, at whim.