We need a globalised co-ordinated financial marketplace to avoid future crashes, according to RICHARD L HUDSON
IN APRIL of this year, a gaggle of economists and bankers flocked to the tall glass headquarters tower of the European Central Bank in Frankfurt for a conference on economic statistics. Dull? Not necessarily. With the credit crunch by then eight months old and soon to become a global crisis, the timing was right for some serious brainstorming about what the central bankers know about the economy, and how they know it.
It's a pity nothing much happened. New ideas and new faces need not apply. Most of the 250 delegates worked for national banks or statistics agencies across Europe. It was the ultimate clique. Pesky outsiders - whether academics, journalists or private bankers - were scarce.
All but a few of the (mostly) men who read aloud from their carefully prepared speeches - heaven forbid they might accidentally say something wrong - focused on how to gather more numbers about the economy more cheaply, not about what to do with the numbers they already had. And the credit crunch? Some passing comments, at best.
Tunnel vision - that's what was on display in Frankfurt that day, and that's what has been the norm among our financial regulators, bankers and economists. We're paying the price right now, with a 30 per cent drop in world stock prices in the past quarter, scores of teetering banks and a gathering global recession.
Of course, lots of people got us here: greedy bankers, lax regulators, gullible investors. But there's another cause less obvious and more intractable: our over-confidence in what we know about globalised markets and economies.
We have built amazing computer models of the economy, lightning-fast trading programmes, breathtakingly clever financial plans - yet still, at the most fundamental level, we don't really understand how markets work, how prices move and how risks evolve. We have been running before we could walk.
What's needed now is a global, co-ordinated effort to correct that ignorance. Since 1963, Benoit Mandelbrot, a Yale mathematician and scientist, has been campaigning for enlightenment. His argument, recapitulated in a book he and I wrote (and just updated), is simple: that financial markets are a lot more complicated, and far riskier, than most economists assume.
Now 84 and still working in the Boston area, he advocates a branch of mathematics that he invented, fractal geometry, as an especially promising avenue of market study.
But Mandelbrot's point goes further. Markets must be studied scientifically, as any other phenomenon of nature or man. There is no room for comfortable assumptions or unmerited complacency - and the last few months have illustrated that point amply.
The sub-prime mortgages that undermined our great banks were written on the false assumption that what had been seen before would, more or less, persist into the future. Housing prices would keep rising, default rates would stay within a forecast range, hedging strategies that worked hitherto would keep on working. That kind of thinking has led to every financial bubble in history - from the tulips of 17th century Holland to the dotcoms and sub-primes of our age.
The 2007/8 credit crisis has been amplified by the computer-driven power of our financial models. With them, Wall Street banks placed multi-billion-dollar bets on the calculated probabilities of default of the mortgages they were indirectly financing. Fannie Mae, the US home-finance banker, undertook a $2 trillion (€1.5 trillion) programme of insuring mortgage securities. In 2004, the Securities and Exchange Commission (SEC), the US stock-market regulator, decided to place its faith on market models in writing its regulations.
Based on computer calculations of default probabilities, it decided security firms didn't need to keep as much cash on hand as in the past.
It relaxed the so-called net capital rule, a regulatory cushion against default dating back to the 1930s. The result was that Wall Street borrowed more than ever before.
At Bear Stearns, for instance, the leverage ratio of how much it borrowed to how much it actually owned skyrocketed to 33:1. The US government continued its optimism well into the credit crunch. "We have a good deal of comfort about the capital cushions at these firms at the moment," said the SEC chairman just six months before the 2008 crash.
But it wasn't mortgage financing alone that was the problem. Complex, model-driven debt-trading strategies spread throughout the world. A striking example is Constant Proportion Debt Obligations (CPDOs) first developed in 2006 by Dutch bank ABN-AMRO.
The CPDOs were billed as a safe way to make money out of the booming market for corporate debt, and offered returns of 2 percentage points above standard, international bank-lending rates.
The investment strategy was toxically simple: when the managers were on a winning streak, they were to close out their positions as soon as they had amassed enough money to pay the promised rate. When they were on a losing streak, they were to raise the bet - on the assumption that, all else being equal, the extra "leverage" would let them recover lost ground when their luck changed.
Incredibly, the early CPDOs earned ultra-safe, AAA credit labels from the debt-rating agencies. But any Las Vegas gambler knows that raising your bet when you're losing isn't a smart idea - it's called chasing losses.
For it to work, on Wall Street or in Vegas, the game has to proceed with a stereotypical, even-handed randomness and a bottomless wallet - no nasty surprises that wipe you out before your luck turns. Alas, nasty surprises are not scarce in financial markets. According to the text-book market formulae, none of the stomach-churning lurches in stock prices of the past few months should ever have happened.
The first drop of 7.7 per cent in the Dow Jones Industrial Average, on September 29th, 2008, was improbable beyond belief. The odds of that happening, say the standard models, were about one in a billion. Yet it did happen.
So what's wrong? In part, it's the assumptions. The textbook financial models, as taught in business schools around the world, have some kind of financial paradise in mind: mostly moderate price changes, each unrelated to the last, in a market populated by rational investors with all the information they could possibly need to make a trading decision. They assume that prices go up, and prices go down, in a "random walk" pattern.
This is white noise, the kind of randomness that you hear in the static of a radio, or see in the white fuzz of a bad TV picture. These assumptions are surprisingly old. They originated in the doctoral thesis of a French mathematician, Louis Bachelier, in 1900, and they have been elaborated since then by a string of Nobel-winning economists. Theirs is a risky but ultimately manageable world.
And it's a popular way of thinking. These assumptions are built into the standard asset-pricing model used by three-fourths of the corporate world's financial officers. They're in the spread-sheets that generate those neat pie-charts you get from your broker telling you how much of your portfolio should be in stocks, bonds or cash.
They're in the classic pricing models for many of the derivatives traded on Wall Street. Their ease of use (they're retailed now in standard financial software) is what fuelled the massive growth of executive stock options, market index funds and many types of derivatives.
Among professional economists, it's known that these assumptions aren't correct - Mandelbrot, in 1963, was the first to point that out. But all models have at least some simplifying assumptions and these particular ones aren't bad for most purposes.
Where they can go spectacularly wrong - in professional options trading, for instance - a great deal of mathematical firepower has gone into making corrections. But, as the sub-prime crash amply demonstrated, they still get it wrong in extreme market conditions.
Mandelbrot's point: wouldn't it be easier to get the math right, from the first instance?
Continued on page 38
Continued from page 37
Our current system, of bad assumptions partially compensated by clever math, is exactly what astronomers were doing before Copernicus and Keppler came along - starting from a false premise - that the sun goes around the earth - and then using some hyper-sophisticated logic to reconcile it to the contradictory evidence coming from the new telescopes.
The following ideas are built into Mandelbrot's "multifractal" models of financial markets - though these are research exercises, rather than practical trading tools. Making money isn't his goal. His interest is scientific. To him, a stock exchange is a black box, a system of data inputs and outputs, to be studied with mathematical tools.
Get the facts right. Markets are not rational. They are turbulent, chaotic, dangerous. Even if they cannot be avoided, we must learn how to mitigate them. Our focus should be on the concentrated bursts of action that common economic wisdom says shouldn't happen and calls "statistical outliers". Mandelbrot himself is an outlier.
He is most famous for his work in fractal geometry - a maths used today in fields as diverse as the interpretation of brainwaves and the compression of computer files. He has used it to construct his own market models. They capture the wild risk of price variations, and they have inspired a small group of mathematically inclined fund managers to experiment on their own.
Markets are very, very risky. A sound trading strategy or portfolio metric would build this cold, hard fact into its foundations. Some commentators have called for a "richter scale" of market turbulence to judge the severity of impending problems.
Trouble runs in streaks. The first tremors of the sub-prime crash have continued reverberating throughout this winter - and will, in one way or another, shape trading for years and years to come. There is a kind of "memory" in the price records, which financial models must capture.
Markets have a personality. Prices are not driven solely by real-world events, news and people. When investors, speculators, industrialists and bankers come together in a real marketplace, a new dynamic emerges - greater than, and different from, the sum of the parts.
Markets mislead. Patterns are the fool's gold of financial markets. The power of chance suffices to create spurious patterns and pseudo-cycles that, for all the world, appear predictable and bankable. Likewise, bubbles and crashes are inherent to markets.
Market time is relative. Mandelbrot talks of a "trading time" - quite distinct from the linear clock time in which we normally think. This trading time speeds up the clock in periods of high volatility, and slows it down in periods of stability. It is almost as if dealing rooms need, besides the standard row of wall-clocks showing the time in Tokyo, London and New York, a fourth clock showing "Greenwich Market Time".
In Frankfurt, the ECB's research department has 58 employees. Yet on this tiny group lies the intellectual underpinnings of a continent's monetary policy. To be sure, their efforts are amplified by the collaboration of a network of other central bank researchers and academic economists across the world. Yet most of that research is into such "typical" issues as understanding the business cycle and the mechanics of inflation.
Its financial-market research is focused on the practical task of understanding how the bank's own interest-rate decisions ripple through the industry - the "transmission mechanism" of monetary policy, in central-banker terminology. Basic research, into the dynamics of pricing and volatility in a global marketplace, gets short shrift.
It's time for change. Financial economics, as a discipline, is where chemistry was in the 16th century.
A bank in which the research department thinks it has discovered something new and useful won't share it with anyone else.
That's where the public sector needs to step in. Fundamental research in economics needs expansion, in an environment that encourages information-sharing.
The National Science Foundation in Washington, the European Research Council in Brussels, the ECB and the Fed need to fund more basic research.
We need a Project Apollo for economics - a passionate, well-funded effort to pioneer a new frontier, the globalised financial marketplace. We need to understand how different kinds of prices move, how risk is measured and how money is made and lost. Without that knowledge, we are doomed to crashes, again and again.
Richard L Hudsonwas formerly managing editor of the Wall Street Journal Europe, and is currently chief executive and editor of a new London-based media company, Science Business Publishing Ltd
