05/15/2010 11:20 am ET | Updated May 25, 2011

The Stock Market Spasm

Investors and the public watched in horror as the stock market had a $1 trillion spasm on May 6. Investigators now think a 70-year-old conservative mutual fund in Kansas made the trades that set it off. Whether this finding holds up or not, people want to know how such a small match can ignite such a large fire? Why is the financial system so vulnerable to these extreme shocks?

Economists, analysts, regulators, and traders will continue to debate that question. But I think it can be traced to a convergence of three factors: arbitrage, leverage, and opacity. Separately, each factor can magnify profit -- and loss -- thus creating more volatility in a market. Together, each factor reinforces the others to create conditions for extreme volatility.

It is very easy to give the formula for making money in any trading situation: buy low, sell high. Easier said than done, of course. In ordinary stock trading, the buy and sell are separated in time. In international commerce, the buy and sell occur in different countries (buy copper in Peru where it is abundant and little used; sell it in Japan where it is scarce and much needed).

In arbitrage, the buy and sell occur close in time (seconds, or even microseconds), but on different exchanges. Often, the exact security bought is different from, but related to, the one sold. One may be the stock of a registered company, the other a future contract or derivative, based on that stock. Or a commodity related to the business of the company in question. Or an index fund. Or whatever.

The opportunities to creatively recognize arbitrage opportunities abound, because the markets in question value the securities differently, which means that the price differential can persist for a long time. However, once a canny trader recognizes the arbitrage opportunity and begins to trade on it, the trades move both markets. When only one early trader is involved, the volume traded is small compared to all of the trades in the affected securities, so the markets don't move much.

But the nature of competition is that success breeds imitation. So more people start to trade based on the same arbitrage opportunity. The volume of arbitrage trades increases and moves the securities on both markets, usually in a way that brings the values on each exchange closer together. In trader's parlance, it reduces the spread. And that starts a vicious cycle.

To see how it works, consider an example. Suppose that on one exchange, the security sells for $10 while the corresponding security on the other exchange sells for $11. By committing $10, the trader can make $1 on each trade. But now the prices move closer together: say $10 on one exchange and $10.10 on the other. Now it takes $100 to make $1. Now let's put the computers to work grinding out these trades at superhuman pace: now it's $10 on one exchange and $10.01 on the other. Now it takes $1000 to make $1. Now add a bunch of competing trading houses using automated systems and the spread gets even thinner. Successful arbitrage makes the next trades harder and harder and harder: the risk goes up while the return goes down.

When it takes more and more capital to make a profitable trade, the risks go up. Suppose that instead of $10 to $10.01, the trade goes at $10 to $9.99. Instead of making $1 on $1000, the trader loses $1 on $1000. Not too bad, at this point. But it does illustrate that arbitrage, like any good tool, can cut both ways.

As traders start to require more and more capital to make profitable trades, the temptation is to use other people's money for part of the deal: leverage. Suppose that the trader puts up $100 of his firm's money and borrows $900 to make the $1000 trade that yields a $1 profit. Now the deal looks much better: the firm made 1% on each trade instead of 0.1%. But what if the trade goes the other way? Suppose there is a real glitch: $10 on one exchange and $9 on the other. The firm would see its entire position wiped out. Leverage cuts both ways: very, very sharply.

Finally, consider opacity: the lack of specific knowledge of what is going on in market by the general public, or at least the broad reach of the financial community. Anyone who can operate with reasonable success in an opaque market can command premium fees for their services. Opacity makes for numerous and inviting arbitrage opportunities to those with the skill, connections, and luck to uncover them.

Regulation discouraged opacity during the years the Glass-Steagall Act was in effect (1933 to 1999), more as a side-effect than a main thrust of that legislation. In the years since financial deregulation, more and more of the dollar volume of trading moved into organizations (such as hedge funds), instruments (such as derivatives), and markets (other than public exchanges) that do not have to report regularly to the SEC or the public. The finance industry prospered because of so many more opportunities for premium pricing and arbitrage.

But there is a downside. Opacity conceals risk, not just from the uninformed, but from the expert. The reason is that even the most attentive expert simply cannot know everything that is going on in the market. Something unexpected is bound to occur -- for example, the "country bumpkins" in Kansas selling index funds.

If the unexpected occurs in an opaque market, normally little should happen. Unexpected events should go in every direction, and broadly speaking cancel out over an hour, or day, or year of trading.

But what if the unexpected gets magnified by others indulging in leverage, arbitrage, or worse still, both at the same time? The minor movement gets magnified into a seismic shift. A Kansas firm's decision to sell an index fund to hedge other trades somehow grows into a 1,000 point drop on the Dow, making $1 trillion evaporate, if only for an hour. The exact sequence of how that happens can be very complicated: the experts are still trying to unravel the story of May 6.

But with opacity, arbitrage, and leverage now much, much more common, the number of opportunities for simultaneous occurrence of self-reinforcing circumstances is orders of magnitude greater than we experienced in prior decades. Between 1933 and 1999, there was only one sickening plunge. The 1987 drop was traced to program (computer-driven) arbitrage trading between the stock market and the futures market, which had grown in volume to the point where it could no longer be treated as a small incremental part of the story, but as a major force driving prices.

Since 1999, we suffered several episodes of extreme volatility. If anything, they seem to be getting more frequent, and more furiously paced. It is no surprise that the public lost confidence in the financial industry as what seem like 100-year storms occur every few months.

But if my thesis is correct, this will keep happening - until something is done to bring more transparency and less leverage to the financial system. Besides reducing general risk, that should reduce the number of chances for self-reinforcing catastrophic chain events, we will all sleep better at night.