DON'T FRET ABOUT PROGRAM TRADING It has been blamed for making stocks more volatile and for at least part of the big drop in early September. Sure, computer-generated buy and sell orders can lock in mind-boggling profits. But they're not increasing the risk for individual investors.
By Daniel Seligman REPORTER ASSOCIATE Susan Lindauer

(FORTUNE Magazine) – WHEN YOU OPENED the morning newspaper to the stock market pages on Friday, September 19, the main news story was not the usual account of the previous day's market action. The main story in most papers was about what might happen to stock prices later that day. Friday was a day featuring the infamous Triple Witching Hour, and on such days in the past, ''program trading'' has often made the market more volatile. This time, however, nothing special happened; the witching hour proved to be a non-event. Wall Street has been feeling a bit spooked by program trading. That term refers to buy and sell orders, typically huge, based on profit opportunities discernible only to investors who have certain computer programs working for them. Some people believe program trading is making the stock market much more volatile, and not only on Triple Witching Days. Some believe it is affecting the level of stock prices; the Wall Street Journal blamed program trading for some part of the 87-point tumble taken by the Dow Jones industrials on September 11. Some believe it is giving large institutional investors a significant new advantage over individuals in the market. A huge majority of investors, it seems safe to say, do not have the faintest idea what is going on here but worry that program trading is making the stock market a riskier place to be. A pretty good case can be made that they should stop worrying. Program trading reflects the exploding popularity of all those mind-bending ''derivative instruments'' that have poured out of Wall Street and Chicago in the past decade or so. These include options on particular stocks, which give you the right to buy or sell a certain amount of the stock (the usual unit is 100 shares) at a specified price during a given time period. There are also options on stock market indexes like the New York Stock Exchange Composite, in which you get the right to buy or sell a multiple of the index at a specified price during a given period. Then there are futures contracts on stock market indexes, in which you commit to buy or sell a multiple of the index by a specified future date at a specified price. When these varied instruments expire, the Street often sees frenzied trading. On four trading days each year, all three expire at the same time; these are the days of triple witchery. Playing off these instruments, the Street's professionals have several different ways to make money on program trades. You can use them to hedge portfolios against large losses, to engage in aggressive speculation, or to lock up arbitrage profits. Although they account for only around 10% to 20% of all program trading, the programs that have attracted the most attention -- because they seem to be most deeply implicated in market swings on expiration dates -- are those designed for arbitrage. You get arbitrage profits by looking for disparities between the prices of index futures and those of the underlying stocks. You can have a guaranteed profit any time the futures prices deviate from the stock prices by more than the ''cost of carry'' (in effect, the cost of short-term money less dividends received during your holding period). To latch on to the profit, you might sell the futures and buy the stocks. When the contract expires, you will assuredly receive more for the futures than you laid out for the stocks. In principle you could arbitrage the opposite way, by buying the futures and selling the stock. If you already owned the stock you needed, that would work just fine. But if you didn't own enough stock, and had to go short in order to sell, you would run into a large practical difficulty. You are not allowed to sell short except on an uptick, and if everyone was trying to sell on that particular day, you might never get a chance to implement your arbitrage program. So most arbitrage starts out by selling futures and buying stock. Figuring the cost of carry is simple enough; it can be done with an ordinary hand-held calculator. You need a computer, however, to combine the cost-of-carry calculation with real-time pricing of all your stocks, which, of course, you need in order to stay abreast of the spread. The program used in Kidder Peabody's Financial Futures Department, for example, is always showing the bid, offer, and last sale price for every stock in the S&P 500. Let's say you have sold the S&P index futures and bought the underlying stock, figuring you have a locked-in arbitrage profit. In fact, you don't have to stick with arbitrage. The Kidder program is always prepared to tell you the potential return from ''unwinding'' the transactions before expiration -- that is, selling the stocks and covering the futures contract -- as opposed to holding on and settling for arbitrage profits. To be sure, the potential profits are not always realized because of what is delicately called ''execution slippage.'' In other words, prices change in the few minutes or seconds between your decision to unwind and the moment at which your order is executed. Still, there are moments when spreads get so large that almost any large money manager would decide to unwind immediately rather than wait for expiration. A spread of 2 1/2 or three points, including the cost of carry, would look pretty irresistible to many fund managers. A number of firms prefer not to deal with the entire S&P roster, and immense theoretical labors have been performed in efforts to find baskets of, say, 50 stocks that reliably simulate the 500. Kidder Peabody itself has moved away from these simulations. Steve Wunsch, head of Financial Futures there, says the firm's experience persuades him that the higher transaction costs associated with the full basket are worth paying -- that is, they're less than the costs arising from imperfect modeling of the 500. It can take a stake of around $25 million to arbitrage the entire 500. You can't just buy a few shares of each stock because the S&P is weighted to reflect each company's capitalization, and so you are driven to buy huge chunks of some issues, like IBM. Wunsch believes that in general it is not practical for wealthy individuals to undertake program trading unless they are very wealthy. ''Maybe a billionaire,'' he crisply puts it. There is no doubt that program trading has put a new premium on fast footwork in the financial markets. In the past, academic research on the markets typically established frameworks in which you looked at rates of return over years, even decades. Now we have studies of program trading looking at the returns over minutes. For example, a study by Hans R. Stoll of Vanderbilt University and Robert E. Whaley of the University of Chicago has examined returns over successive one-minute and 15-minute intervals on expiration days and the following market days. However, repeated studies have shown that the stock market in the age of program trading is not in general more volatile than it used to be and is probably less volatile. A Salomon Brothers study published this summer shows that the standard deviation of daily price changes in the last couple of years is about average, looking at the Seventies and Eighties as a whole. (Looking at the 16-year period, you see that peaks in volatility are closely associated with market bottoms.) This finding conforms to several studies done by the Federal Reserve and the Securities and Exchange Commission. The Stoll-Whaley study offers only a modest amount of ammunition to anyone wishing to regulate or somehow abolish program trading. The study establishes that the markets are more volatile on days when futures contracts expire and also that closing prices on those days are somewhat affected by the programs. How can anybody know that prices are affected? Because the price trend on futures expiration dates is usually reversed on the next trading date. The price effects of program trading are not especially large; on average they are less than the effects of large block trades. (The magnitude of the reversals averages around 0.4% of the market indexes.) Still, it is marginally possible to trade profitably on the price effect, and the reversals occur often enough so that you can tell yourself it is 95% certain the pattern is not just a chance event. THE SEC, LIKE Stoll and Whaley and most scholars who have looked at program trading, does not believe program trading is much of a problem. For the September 19 Triple Witching, Chairman John Shad proposed, and the New York Stock Exchange instantly accepted, a scheme in which specialists were supposed to disclose any major imbalances on their books for orders to buy and sell in the last half hour of trading; the idea was to prevent the market from being blind-sided by either buy or sell orders being placed in the last few minutes, as program trades were unwound. When you talk to Shad about this initiative, you rather get the sense that he is a very reluctant regulator. He believes that arbitrage and, for that matter, speculation are good for the market, not bad. They increase market efficiency and tend to smooth out price movements. He sounds frustrated by the fact that the undesirable effects of program trading -- the volatility and price swings on expiration dates -- get all the publicity, while the desirable effects by their nature go unnoticed. Asked whether program trading is a real problem or just a matter of perceptions, he answered sadly: ''We live in a world of perceptions.'' He is certainly right about that. Also about program trading being a great big non-problem.