Learning From the London Market

Eleven months ago, without much warning, stock values plunged in record-breaking declines on record-breaking volume. Prophets of gloom immediately began to warn us that we must end our profligate and speculative ways. The government budget deficit and computerized stock trading were the most common causes cited as responsible for the crash. Unless something was done about one or both problems, we could expect to continue down the slippery slope.

Looking back, we see that these warnings were wide of the mark. True, there were some minor changes in the government budget late last year. But once we discard the usual government budget hype, the 1987 budget changes would be hard to separate from the large errors typically made by government and private forecasters. True, also, there were some changes in regulation of stock trading and some restrictions on computerized trading, but these were small changes that cannot account for the failure of the gloomy predictions to be realized.

The economy has been robust so far this year, and the stock market, while not buoyant, has given an average rate of return. Of course, the stock market is difficult to predict, so large prediction errors are not entirely surprising.

The October, 1987, crash is a rare example of markets moving together all over the world. The decline of the New York Stock Exchange was one of the smallest declines experienced in any country. Yet there has been more discussion and investigation of the causes of the decline here than in any other country that I know. Six or seven reports by agencies of the U.S. government or by the individual markets have appeared. Some run to hundreds of pages. In contrast, most countries either have not studied the decline in their markets or have not made their reports public.

The many U.S. reports and studies of the October crash share two striking, but little noted, features. First, none of the reports that recommend restrictions on computerized trading, margin requirements or more regulation claim that the crash would have been prevented if their proposed restrictions had been in place. Second, the reports and the subsequent discussion have ignored what happened to foreign markets during the crash.

Richard Roll, a distinguished professor of finance at UCLA, compared the decline in 23 stock markets around the world. Although markets differ by the presence or absence of computerized trading, futures markets, margin requirements and other trading rules, Roll found little evidence that these rules affected the size of the October, 1987, decline.

The most important difference among markets was their past volatility. More volatile markets, such as Mexico or Hong Kong, typically rise more in expansions and fall more in declines. That’s what happened last October: The more volatile markets declined most. Market organization, computerized trading and the like add nothing to the explanation of the size of the declines once allowance is made for past volatility.

Comparisons between U.S. and foreign markets also reveal possible flaws in market organization unrelated to the use of computers and other new techniques. Although studies by the Brady Commission and others praise the performance of the U.S. markets during the week of Oct. 19, they note several problems. Some people who would have liked to trade on Oct. 19 and 20 could not reach their brokers. Specialists on the New York Stock Exchange and other U.S. markets were unable to establish prices for some securities for long periods during the trading day. Some market makers reportedly withdrew from trading over-the-counter stocks, reducing the public’s ability to buy and sell.

Trading volumes became so heavy that the reporting of transaction prices to buyers and sellers was delayed for long periods--for days in some cases. Some of the official reports note that the absence of information on transactions and prices may have inhibited trading and increased uncertainty. This could occur if potential buyers did not know what they had purchased or sold, the prices to which securities had fallen, or the prices at which purchases could be made.

Trading halts were common and long lasting. There were 195 trading delays and halts on the NYSE on Oct. 19 and 280 on Oct. 20. For stocks in the S&P; 500, the average duration of trading halts was 51 and 78 minutes on the two days. The delays added to confusion and uncertainty, and contributed to the large spread between stock and futures prices.

The London stock market fell by a percentage similar to the decline in New York, but the London decline was more orderly. Market makers in London had fewer difficulties making markets function. A study by the International Stock Exchange in London found that most securities traded continuously, without the halts that were so common in New York. Spreads between bid and ask prices did widen, and market makers limited quotes to smaller blocks of stock.

For shares of the highest quality, bid-ask spreads rose to 3.5% from about 1% before Oct. 19, and the average quote size dropped to about one-third of normal. But investors who wanted to sell could do so. Trading halts were rare.

A general conclusion of studies of the London market is that the volatility of the Dow Jones industrial average for the United States was greater during the crash than similar measures for London (the Financial Times index) or Tokyo (the Nikkei Dow). This contrasts with data for July through September, 1987, which shows little difference in the price volatility in the three markets.

The London experience suggests that regulators should be less concerned about portfolio insurance, computerized trading and other modern trading techniques. They should worry more about why London was able to execute transactions more reliably than New York during the crash. One reason, suggested by the report on the London market, is that competitive market makers replaced specialists a year before the market crash. London’s market makers have an incentive to make continuous markets even in periods of panic selling: If they stop making a market in a stock, they cannot trade it again for 90 days.

The New York Stock Exchange, the Securities and Exchange Commission and Congress should be asking whether the United States should copy the London example. Competitive market makers could replace the single, exclusive agent, called a specialist, who keeps records of offers to buy and sell and executes all transactions. The change might be appropriate for all traded stocks, as in London, or only for the largest, most actively traded stocks.

In the past 10 years the volume of trading in stocks of large companies has increased to levels at which alternative methods of trading may now be more efficient. Further, the specialist system is geared to trading individual stocks, whereas the rising demand is for a system that permits pension funds and other institutional investors to buy and sell entire portfolios. That’s one reason the index options, traded in the Chicago exchange, have become popular.

There were, of course, many differences between London and New York in October, 1987. London, unlike New York and Chicago, had very little portfolio insurance in force. The use of index options is limited, in part by differences in tax treatment. Although trading was at record volume during the crash, futures markets in London are small relative to the size of the London market or U.S. futures markets.

London, unlike Washington, hasn’t spent the months since last October arguing about new regulation. Instead, regulators there have been saying that further development of the futures market and the modern techniques used in the United States would improve the market’s ability to handle transactions. Japan has moved in an orderly way to develop futures and new portfolio techniques.

Instead of looking for new regulations and restrictions on trading arrangements, Congress, the SEC and the New York Stock Exchange might take note of London’s performance and encourage more competition among market makers in our stock markets.