Remembering Black Monday, When Computers Traded Too Many Stocks and Wall Street Crashed
Twenty years ago, automated trading systems contributed to investor panic and a historic drop in the stock market. Historians still draw lessons about market volatility and regulatory oversight from the Black Monday crash.
CIO — DuWayne Peterson always believed in making time for business and IT strategy. In fact, on the morning of Monday, Oct. 19, 1987, Peterson, then vice president of systems, operations and telecommunications at Merrill Lynch, went to his office in the World Financial Center in lower Manhattan while a good portion of his senior staff attended a strategic offsite in Princeton, N.J. The previous Friday, to Peterson and the rest of his Wall Street colleagues, the market looked grim. The the Dow Jones Industrial Average (DJIA) fell 108 points to close at 2246.74. On Thursday, U.S. Treasury Secretary James Baker expressed concerns to the media over the massive sell-offs. Over the weekend, investors began to sweat. Pundits, economists and politicians went on television, hedging their bets about the market.
“Friday we saw trouble,” Peterson says. “But we had no idea what we’d walk into that Monday morning.”
Black Monday, as it became known, would represent the largest decline in U.S. stock market history. The DJIA dropped an astonishing 508 points (22 percent) to close at 1739. Investors sold off stock in a frenzy. Peterson says Merrill Lynch’s systems, which usually saw 100 million to 200 million shares a day traded on the New York Stock Exchange, buckled on Black Monday as volume soared to around 604 million. Very quickly, Peterson knew he’d need help.
“In the morning, my senior person called me from New Jersey and asked me if I wanted them to come back,” Peterson remembers. “But I said, ‘No, strategic planning is important. We’ll handle this.’ Well, by mid-morning, the volume just started roaring, and they had to hustle back.”
And Peterson’s situation was hardly unique. Financial services firms and their technical people worked frantically throughout that Oct. 19 to keep up with the massive volume. As Black Monday celebrates its 20th anniversary today, economists, business leaders and politicians believe that bizarre day of chaos taught the country just how profound an effect information technology—and the way humans behave using that technology—plays on the market. Technology, just like humans, can make mistakes. Systems built to withstand loads of information can be overwhelmed, and computers don’t stop and take a deep breath in times of crisis to assess why it’s happening. Twenty years later, Black Monday serves as a reminder that balancing technology-enabled efficiencies with the realities of human behavior is difficult if not impossible without key safeguards.
Questions Before the Crash
Back in the summer of 1987, a few months before Black Monday, something had been bothering Rep. Edward Markey (D-Massachusetts). A robust stock market had been slowing down, and Markey says he worried about the role of program trading—where computers automatically buy and sell stocks based on algorithms set by stock trading companies. He was concerned about the possibility that stocks, options and related futures markets would become volatile. If the market experienced rapid sell-offs, for instance, driving the price of stocks down to a certain level, the computers would then be programmed to sell very quickly without consideration to the human panic or hysteria of a tough day on Wall Street.
As chairman of a House subcommittee on telecommunications and finance, Markey called a hearing to look at program trading in July 1987. “While many of the industry witnesses argued that there was nothing to worry about, I was not convinced they were right,” he remembers.
After the market dropped 91 points on Oct. 6, Markey wrote to officials at the Securities and Exchange Commission (SEC), asking them to research what happened. Twelve days later, he watched it unfold on his television. “Watching the crash take place on CNN was a gut-wrenching experience,” he says. “My worst fears were realized.”
Although economists say that program trading wasn’t the only factor at work on Black Monday—dried up liquidity, a lack of visibility into market conditions and irrational panic on the part of investors also played key roles—most acknowledge it was a factor. “We haven’t agreed whether or not program trading was the primary culprit, but there’s been consensus it played a role,” says Paolo Pasquariello, assistant professor of finance at the University of Michigan’s Stephen M. Ross School of Business.
Program Trading Takes Off in the 1980s
Program trading had started as early as the 1970s, but it wasn’t utilized by a large percentage of brokers until the mid-1980s. In 1987, it’s widely believed program trading accounted for approximately 10 percent of the trades taking place at the New York Stock Exchange. While they’d been carefully designed and were cutting edge for their time, the algorithms behind the computers weren’t nearly as complicated as they are today, economists say. In 2007, trading systems enjoy complex algorithms that take a multitude of economic models into consideration. In 1987, brokers would set a certain price and once the stock reached that price, it would automatically be trained to buy or sell. “If a stock fell to $50, for instance, it might be programmed to sell,” explains Pasquariello. “But the computer didn’t stop and ask the question, ‘Should I sell?’”
If the computers in 1987 hurt the market from being too primitive and unable to predict human behavior, the ones used today can in some cases do just as much damage from thinking their algorithms are good enough to make that premonition. “There is a fair amount of hubris there,” says Pasquariello. “Some hedge funds who think their models are unbeatable have actually shown huge margins of error,” he adds.
In addition, Pasquariello says, there’s the worry today that if we program all our systems in similar ways, we’ll all either win or lose at the same time. “We all go to the same grad schools, read the same textbooks and read about the same economic theories, then our computer-based strategies might be the same,” he says. In other words, in a healthy market, everyone shouldn’t win and lose at the same time. There needs to be winners and losers.
A Stock Options System “Dead in the Water” on Black Monday
On Black Monday, according to Peterson, Merrill Lynch’s mainframes were stretched thin. One system, in particular, had been designed to produce options every time any of the stocks the company was tracking moved to a certain price level. The system “was dead in the water,” he remembers. “It wasn’t going to produce any more options because its memory was filled.”
