Is anyone else tempted to take all their money out of their respective bank accounts and IRAs and just hide everything under the mattress, so we'll have ready cash after the apocalypse? That's how I'm feeling at the moment, after watching one of the nation's most venerable financial institutions, Bear Stearns, crash and burn in the space of just a few days. A few weeks ago, their shares were valued at around $170. By last week, that had dropped to $70. And this morning, I awoke to read that, thanks to a bailout by the feds, J.P. Morgan will be buying Bear Stearns for the rock-bottom price of $2 a share. The entire Bear Stearns headquarters building in Manhattan is worth more than that! Maybe they'll throw in a set of Ginzu knives with the purchase. ("Now how much would you pay?")
How art the mighty fallen. It wasn't that long ago that Bear Stearns was listed as one of the most admired financial institutions. Turns it out it was all an elaborate house of cards, much like the Enron meltdown of a few years ago, except this time it's worse, because stocks are tumbling globally, not just in the US. That's what happens when a company (or a nation) over-extends itself and takes on far too much debt. I'm sure James Cayne, the former CEO, feels just terrible -- maybe even badly enough to pause for reflection while rolling in the great wads of cash he raked in during his tenure: $232 million in compensation from 1993 to 2006. Personally, I think he should be forced to give some of it back to help bail out the company he helped drive into the ground. Call it a return to the traditional conservative values of accepting responsibility for financial imprudence.
No doubt Bush and Cronies will dismiss this as one of those statistical outliers, a "rare event" in the financial markets, instead of something that could have been avoided had management been a bit more prudent. I find myself wondering what Eugene Stanley of Boston University would have to say about the mess. He's one of the pioneers of econophysics, and was on hand at the APS March Meeting in New Orleans to talk about his latest research: namely, that these so-called "outlier" rare events actually occur in regular patterns, and thus should be incorporated into economic theories, which to date have dismissed them as "anomalies."
Stanley can make this statement with some degree of confidence because he's just completed analysis (with the help of numerous grad students and post docs) of an enormous amount of financial data -- 200 million transactions on the New York Stock Exchange spanning a two-year period. That's far more than has ever been included in such analyses before (10<8> data points, compared to 10<4> data points). "Classic economic theories not only fail for a few outliers, but there occur similar outliers of every possible size," he said. "So ignoring them is not a responsible option." He's applying the tools of physics to figure out if there are underlying unifying principles (equivalent to physical "laws") that dominate the NYSE and other financial bodies and institutions -- i.e., whether there is "an identical set of laws hat hold across diverse markets, and over diverse time periods."
Econophysics is a relatively new field, emerging in the mid-1990s thanks to the work of several physicists who decided to apply the tools of statistical mechanics to the complex problems posed by financial markets in particular. It was the right time for this to happen: not only did huge amounts of data suddenly become available in the 1980s, but there were an increasing number of PhD physicists fleeing the stagnant job markets in their fields for Wall Street, finding work as "quants" -- basically, sophisticated financial analysts. At the same time, according to Wikipedia, "It became apparent that traditional methods of analysis were insufficient. Standard economic methods dealt with homogenous agents and equilibrium, while many of the more interesting phenomena in financial markets fundamentally depended on heterogenous agents and far-from-equilibrium situations."
Lots of different physics models have been applied to financial systems, including percolation models, diffusion theory (the famed Black-Scholes equation garnered a Nobel Prize in Economics), models with self-organizing criticality of complexity, models developed for earthquake prediction, even chaotic models originally developed to study cardiac arrest. That was the topic of another paper at the March Meeting, in fact. Nothing that fractal analyses of cardiac rhythms suggest that healthy people have complex cardiac behavior -- compared to the rhythms of unhealthy people, which are more random or periodic in behavior -- researchers at Brigham Young University are looking into whether similar complexity might be an indication of a healthy company. The title of their paper: "Fractal Hearts are Healthy Hearts -- Are Fractal Companies Healthy Companies?" (If so, I'd bet Bear Stearns would have failed any test devised along those lines.)
Stanley has used a spin glass model to describe stock market fluctuations. I've heard of biophysicists adopting a similar approach to, say, mutations of the flu virus. Apparently spin glass models are pretty generic in their applicability. They can be used whenever you have a complex system made up of lots of units: eg, stock market traders who all have different opinions, interact with each other, and make decisions based on the relative strengths of those interactions. The stronger the interaction -- or the more trustworthy a trader deems a colleague -- the more influence that interaction has. But the strength of those interactions can change with time, for example, if a trader loses confidence in a colleague. (Jen-Luc Piquant gives a rousing vote of "no confidence" to James Cayne, just for the record.)
Of course, treating human beings as if they were mere particles has its limitations; human behavior is inherently unpredictable. And no model is likely to ever enable analysts to predict a specific event in the stock market, any more than one can precisely pinpoint the time, location, and severity of an earthquake. Stanley was unequivocal about this, calling the stock market "a very complex system and probably insoluble," emphasizing, "There is absolutely no way anyone has been, or will be able to predict the future.
One of the prevailing economic theories is the random walk hypothesis for stock market prices, which basically says the prices can't be predicted due to the lack of correlation of past and present. Just because a stock rises one day, there's no guarantee it will rise again the next. Here's an interesting anecdote: Princeton economics professor Burton Malkiel -- author of A Random Walk Down Wall Street -- conducted an experiment with some of his students, giving them a hypothetical stock worth $50 at the outset.
Each day, he would flip a coin to determine the closing stock price for the day: heads, the stock closed half a point higher; tails, it would close half a point lower. So it was pretty much a 50/50 chance. Malkiel mapped out the cycles and trends in a chart and graph form, then took it to what's known in the financial industry as a "chartist" -- a person whose job it is to predict how the stock market will behave in the future based on past patterns. The chartist, not knowing Malkiel's data was based on a coin toss, immediately wanted to buy the stock, and was disappointed when the truth was revealed. Probably a bit depressed, too, since if the market and stocks are indeed as random as flipping a coin, he's pretty much out of a job. (You can create your own random walk with this spiffy online game involving virtual turtles.)
There are naysayers to the random walk hypothesis, most notably Martin Weber, who specialized in behavioral finance. He observed the stock market over 10 years, analyzing market prices for any signs of trends. He concluded that stocks with high price increases in the first five years tend to under-perform in the following five years. In a different study, he found that stocks with an upward revision for earnings tend to outperform other stocks in the next six months, giving investors a bit of an edge when deciding which stocks to pull out of the market and which ones to leave in (the ones with the upward revision -- at least for the next six months). Other naysayers include MIT's Andrew Lo and Craig MacKinley, who argue in A Non-Random Walk Down Wall Street (1999) that "even the casual observer can look at the many stock and index charts generated over the years and see the trends. if the market were random... there would never be the many long rises and declines so clearly evident in those charts." They believe the stock market is predictable.
Hmmm. I'm no economist, but I have some doubts about those "trends." First, human beings are kind of designed to see patterns, so it's easy to see something that isn't really there, and make a lucky guess based on the perceived trend, causing confirmation bias to kick in. I think this is a particular risk when you're working with a limited data set. That's why Stanley's work is noteworthy -- he's working with an order of magnitude more data. And while he definitely sees patterns, he appears to be looking less for short-term trends, and more for universal principles of economics. I guess we'll just have to let the proponents of the various models fight it out in the Economics Octagon until there's One Model Standing.
[Jen-Luc is rooting for the Relativistic Economic Model, which seeks to incorporate relativistic effects into economic analysis (h/t: Arjendu of Confused At a Higher Level). You know, deposit a small amount of money in a bank account, then rocket off into space at the speed of light, returning to find hardly any time has passed for you, but your meager deposit has accrued tons of interest, and you are rich, my friend -- stinking rich! Until you find, as Woody Allen's hapless time traveler did in Sleeper, that $1 billion doesn't go very far when a phone call from a public payphone costs a couple million. Inflation is such a buzzkill for the nouveau pseudo-riche.]
Apparent trends there may be, but to date, it's still impossible to predict the stock market with 100% precision (although Doyne Farmer and Norm Packard of Prediction Company are no doubt still trying, continually refining their methods). But we can continue to analyze probabilities to reduce risk, and better models translate into better risk management. That's something everyone at Bear Stearns should be able to get behind, along with the rest of Wall Street. Ironically, many economists have resisted the encroachment of physicists onto their academic territory; econophysics has had the greatest impact on models for price fluctuations in financial markets, with less of an impact on general principles of economics. Considering how those prevailing principles seem to be faring, maybe they should at least consider giving physicists like Eugene Stanley a try. Still, I doubt anyone could have predicted Bear Stearn's demise -- that's gotta be the outlier of outliers.