On Thursday, March 28, 1985, almost exactly five years after Ronald Reagan visited the New York Stock Exchange as a candidate, he returned, the first sitting president to visit the Big Board.
At his side was his new chief of staff, Donald Regan, who greeted John Phelan as an old friend. As they threaded their way across the trading floor, the president shook hands and waved. Regan and Phelan edged him determinedly toward the stone balcony where Reagan would deliver his prepared remarks and ring the opening bell.
“This is a great view from up here,” the president joked, once he reached the balcony. “It’s kind of like being at a Saturday night tag-team wrestling match at the Garden.”
He detailed how his administration had helped restore health to the stock market during his first term and promised to do more in his second term. “That’s our economic program for the next four years,” he concluded. “We’re going to turn the bull loose!” The opening bell rang right on time, amid shouts of “Ronnie! Ronnie! Ronnie!”
Phelan escorted Reagan and his chief of staff to the gilded Edwardian boardroom on the sixth floor, where the president fielded polite questions from the guests gathered there and told a few familiar stories. Then he was on his way.
While the visit had been a credit to the team Phelan had put together, he still needed a right-hand man, someone to take on administrative duties so he could cultivate the political relationships that had paid off so handsomely that morning. In short, he needed someone who could be for him what he had been for Mil Batten.
For weeks, Phelan had been quietly courting Robert J. Birnbaum, the able president and chief operating officer of the American Stock Exchange, to take the number-two job at the NYSE. Birnbaum had joined the Amex in 1967 and had held his job there for the past eight years. He was no-nonsense in his business dealings; his unsentimental style reminded some people of a surgeon short on bedside manner.
When Phelan discussed the NYSE job with Birnbaum over lunch in his private dining room, he mentioned that he was talking with four candidates. “They’re all great,” Phelan said, “but we’re also looking at you.” He beamed expectantly.
“Well, of course you are,” Birnbaum said. “Unless you were all retarded, of course you’d be looking at whoever was in my job.”
Perhaps his blunt style spoke to Phelan’s inner marine. In any case, a short time later, Bob Birnbaum signed on as Phelan’s top lieutenant. His move sparked rumors of a merger between the NYSE and the American Stock Exchange, promptly squelched by the Amex’s chairman, Arthur Levitt. The merger whispers reflected the business pressures the NYSE was facing as giant institutional investors pushed for faster, cheaper trading services and a closer alignment with the latest derivative products to support their new financial strategies.
The American Stock Exchange and a few other regional stock markets had developed healthy options markets to compete with the Chicago Board Options Exchange, but the NYSE had not been successful in such ventures. Its offspring, the New York Futures Exchange, had struggled from the moment its ribbon had been cut in 1980.
And it was harder to see a future built around the simple business of trading blue-chip stocks, face-to-face, on the NYSE floor in Lower Manhattan. Some giant investors had begun to trade around the clock, around the world; more would surely follow. And Wall Street firms already had the computer connections necessary to fill their big clients’ largest stock orders “upstairs,” at their own trading desks, without routing them to the floor. The NYSE’s defense against this “upstairs trading” had been rules that limited such off-floor trades and efforts to maintain the deepest and most transparent market for its listed stocks, with prices reported for all to see and specialists who were willing to use their own capital to maintain an orderly market.
If the day came when big institutional customers cared more about secrecy and speed than about transparency and price, or when the NYSE’s specialists could not maintain order in the midst of a herdlike stampede—what then?
John Phelan was automating whatever and wherever he could, but those investments would be wasted if the Big Board did not keep its blue-chip listings.
The takeover mania had created an existential threat to those NYSE listings. In July 1984, three prominent Big Board companies, General Motors, Dow Jones and Co., and Coastal Corporation, had created “super-voting” classes of stock to ward off corporate raiders, in violation of the NYSE’s bylaws. That action should have disqualified these companies from being listed on the exchange.
If the NYSE stood its ground, raider-wary companies could move their listings to a safer market. Both the American Stock Exchange and the Nasdaq market permitted their listed companies to have multiple classes of stock.
So, Phelan did not kick the three blue chips off the Big Board. Instead, he appointed an advisory committee to study various options. In January, the committee recommended that the NYSE allow listed companies to have more than one class of stock, so long as the common shareholders approved.
The forces opposed to the NYSE changing its listing standards were formidable. Institutional shareholders, corporate raiders, regulators, and powerful members of Congress argued that “one share, one vote” was the essence of shareholder democracy and should never be abandoned. Congressional hearings were planned, and Phelan would be expected to testify.
In his heart, he believed that the existing rule against “super-voting” shares was the right one. The question was whether preserving “shareholder democracy” for giant institutional investors would end up destroying the New York Stock Exchange.
Meanwhile, the exchange was being pressed daily to keep up with the growth in trading volume. Records were being set and broken too often to bother mentioning them. The automated DOT system, originally set up to expedite small retail orders, already was handling larger orders, many of them arising from the complex index arbitrage strategies of institutional investors. That arbitrage trading was rapidly tying the NYSE closer to the futures and options markets in Chicago.
Those new links were demanding attention on Wall Street. Earlier in the year, on the third Friday in March, the Dow had eked out a small gain during the day and then, suddenly, had fallen by a full percentage point in the last hour, on heavy trading. According to the traders on the floor, the dramatic end-of-day drop had occurred because several brokerage houses needed to unwind positions that exploited price differences between stocks and the options on those stocks that expired on the third Friday of the month. To do this, they had instituted automated selling programs to unload stocks in the last hour of trading. Similar last-minute nosedives had been occurring occasionally on these “third Fridays” going back at least a year.
As the third Friday in April approached, the New York Times sounded an early alert. “Superstitious or not, many stock investors have come to regard the last hour of trading on the third Friday of each month as the ‘witching hour,’ a time when the market seems to plunge for no apparent reason,” the paper’s commodity markets columnist noted. That was the expiration day for the Amex’s Major Market Index option, known as the MMI, an option contract that closely tracked the Dow Jones index.
There was also an MMI futures contract traded on the Chicago Board of Trade. Days when stock index futures expired on the same day as stock index options were known as “double witching hours.” However, four times a year (on the third Friday of March, June, September, and December), stock options for the company stocks included in various indexes also expired along with the index options and index futures, creating a “triple witching hour.” That was what was in store for June.
“Whatever happens on April 19,” one trader said, “we think June’s witching hour will be most interesting.”
* * *
THE STOCK MARKET’S witching hour in April was surprisingly benign; trading closed with barely a ripple, leaving the Dow essentially flat. Trading was relatively calm in May as well, though there were a few moments of sharp selling just before the closing bell.
But all eyes were on the third Friday in June, a “triple witching hour.” In the final moments of trading, stock prices leapt higher, sending the Dow close to a new record and pushing trading volume to more than 125 million shares, a 50 percent jump over the previous day.
For most investors, an abrupt rally is always more pleasant than a sudden nosedive. Here, what was worrisome to professionals was the frenzy of the late trading. “Typically, expiration Friday is a psychotic day in the market,” one analyst noted. But this day’s outburst, which strained the NYSE’s computer capacity, was crazy enough to be worrisome.
Clearly, the market’s metabolism was changing. It was hard to see how these regular spasms up and down could be explained by “rational” changes in investor sentiment occurring in an “efficient” market.
* * *
IN THE SPRING of 1985, Leo Melamed saw an opportunity to expand the Chicago Merc’s influence in Washington. President Reagan’s chief trade negotiator, William Brock, had been named secretary of labor in mid-March, leaving an opening for a new U.S. trade representative. Melamed heard the news on the car radio as he headed toward his vacation home in Arizona. “The idea struck me like lightning: Clayton Yeutter,” he recalled. Merc president Clayton Yeutter had held the same post in the Ford administration, and his return to Washington would add luster to the Merc’s reputation.
In a masterful exercise of political power, Melamed worked the telephones for an entire weekend, calling three dozen chief executives to urge them to lobby for Yeutter. The effort paid off, and Yeutter was soon sworn in as the new U.S. trade representative.
Yeutter’s move to Washington opened the door for William J. Brodsky to assume the presidency of the Chicago Mercantile Exchange. Bill Brodsky had joined the Merc from the American Stock Exchange, leaving roots that reached deep into the bedrock of the Manhattan markets. His father had worked on Wall Street for almost 60 years, and Brodsky had easily snagged summer jobs on the NYSE floor while he was in high school. At Syracuse University, he was president of his fraternity, and after finishing law school there, he went to work for the brokerage firm where his father was a partner.
With the bear market of 1973 bearing down on the firm, Brodsky’s wife, Joan, encouraged him to pursue some research that might grow into a law review article to burnish his credentials. He chose to explore the Chicago Board Options Exchange, which was just opening its doors, and gained a solid legal grounding in the world of options. Meanwhile, he landed a job in the legal department of the American Stock Exchange, where he was soon overseeing the Amex’s ambitious options trading operation. A few years later, he was appointed to the board of the Options Clearing Corporation, which acted as the central clearinghouse for settling all options trades, and he traveled regularly to Chicago for board meetings.
The trips had become a bit of a romance for the lifelong New Yorker. “I was smitten by the ingenuity and creativity and willingness to take risks that I found in Chicago,” he later recalled. “There was an openness, a freshness, a lack of haughtiness in Chicago. The question here was ‘are you smart?’ not ‘what’s your pedigree?’”
So, when a recruiter called in the summer of 1982 about a spot as executive vice president at the Chicago Merc, Brodsky was tempted. He flew west to meet with Melamed and the exchange’s executive committee. He saw the Merc’s new home under construction and was astounded that the exchange was prosperous enough to finance the building largely from the recent sale of new memberships. He soon signed on and, over the next year, organized the management staff for the successful move into the new facility in the fall of 1983.
Despite his embrace of Chicago, Brodsky never burned his bridges to New York. He easily maintained cordial relationships with people that some of his Merc colleagues considered adversaries—exchange officials in New York and regulators in Washington and executives of the upstart Chicago Board Options Exchange. One of his close friends was Bob Birnbaum, the new president of the NYSE.
* * *
IN NEW YORK, witching hours continued to create uncertainty and unease. After the turbulence in June, the July expiration date was more orderly. A trading desk executive had an explanation for this. He said that the giant institutional investors “seemed interested in reducing the volatility, and therefore [they] spread the buying over several hours,” instead of slamming into the closing bell.
There is no way of knowing whether money managers working for Gordon Binns at the General Motors pension fund were among the traders showing restraint. Binns had authorized trading in the S&P 500 futures contracts, after one of his senior staffers invited Leo Melamed and Bill Brodsky to visit Binns at the GM building.
Melamed had been delighted to make the trip, and Brodsky soon became an admirer of the warm, intellectually curious pension fund manager. They invited Binns to attend various conferences to talk about his use of futures contracts. They knew that the cachet of the GM pension fund, the largest in the corporate world, could only burnish the Merc’s reputation.
Derivatives were not the only experiment Gordon Binns had undertaken. For nearly a year, he had watched the activities of the Council of Institutional Investors and was confirmed in his opinion that corporate pension funds would always be an awkward fit there. So, in the fall of 1985, he and several other corporate pension fund executives formed their own organization, the Committee on Investment of Employee Benefit Assets, or CIEBA.
The new organization was a way for Binns and his corporate counterparts to have a seat at the table when Congress investigated market developments important to these institutional investors. Program trading was one of those developments; Binns believed it was crucial to preserve program trading—and CIEBA was a way for another army of giant institutional investors to make that case to Congress.
* * *
IN NOVEMBER 1985, Jerry Corrigan learned firsthand that human panic wasn’t the only way a crisis in one market could spread to another. In an increasingly automated market running at an accelerating pace, a computer malfunction would work just as effectively, with less warning.
At 8 a.m. on Thursday, November 21, a vice president at the Bank of New York called a New York Fed official to report a delay in the bank’s computerized tally of the previous day’s trading in Treasury securities. The Fed official was not alarmed; minor problems like this occurred frequently, and the Bank of New York was one of the busiest processing hubs in the government bond market.
The New York Fed and the Bank of New York were both part of a massive electronic network that could fairly be called the nation’s financial circulatory system. Insiders called it “the payment system.” It was the banking industry’s modern equivalent of the “back offices” on which Wall Street once relied to send stock certificates to the right buyer and cash to the right seller. So-called “clearing banks” such as the Bank of New York served as the back office for legions of other financial firms and institutions engaged in Treasury trading and other complicated high-speed activity.
Unlike the old engraved stock certificates, however, Treasury securities existed only as distinct bookkeeping entries in a vast electronic ledger. Each had a unique identifying number so that it could be tracked through the system. Clearing banks such as the Bank of New York made sure that a Treasury bond that had been sold was removed from its seller’s electronic ledger and entered into the electronic books of its buyer, and that the money the buyer paid was electronically subtracted from the buyer’s bank account and added to the seller’s.
Like the back offices of the late 1960s when the “paper crunch” hit, this was not glamorous work, but it was work that could bring the markets to a halt if it didn’t get done. Normally it was carried out at blinding speed and in astonishing volume, as computers received information from other computers and automatically sent instructions on to still other computers, with the Fedwire (the New York Fed’s electronic money-moving service) as the primary artery in the system.
Jerry Corrigan sometimes likened the financial payment system to the nation’s interstate highway system. Instead of cars and trucks, there were electric impulses, which shuttled virtual securities and intangible cash between buyers and sellers. On a typical day, about $200 billion in transactions traveled smoothly from bank to bank, around the country and out into the wider world.
November 21, 1985, would not be a typical day.
At about 10:15 a.m., the Bank of New York alerted the New York Fed that the bank’s customer data files, belatedly updated with Wednesday’s transactions, had somehow been corrupted. For hours, programmers at the bank and its outside software consultants tried to find the flaw.
The rest of the market was up and running, the start of a busy Thursday. Securities were being electronically delivered to the Bank of New York for processing; the sellers expected cash, and the buyers expected to be credited with ownership and interest on the securities. With its computer system disabled, the bank was the equivalent of a drawbridge that had frozen in the open position. Government securities piled up on its books and had to be paid for with the bank’s cash. The bank couldn’t deliver the securities to, or collect cash from, the buyers on the other side of the broken bridge.
But there was some breathing room. The New York Fed, via the Fedwire, could take cash from the Bank of New York’s account at the Fed and transfer it to the sellers of the securities, but with no cash flowing into the bank’s account, it was soon overdrawn. And this was no garden-variety overdraft. By 11:30 a.m., the Bank of New York’s account at the New York Fed was overdrawn by $12 billion.
Back at the bank’s data processing center, as one team worked to find and fix the database flaw, another worked on a temporary patch to allow the bank to start clearing transactions before the Fedwire’s daily closing time of 2:30 p.m. Around 2 p.m., the first patch was tested. It failed. Work was begun on a second one, and a third one; none was successful. A senior bank executive called the New York Fed to ask that the Fedwire be kept open past its normal closing time.
By 4 p.m., both Corrigan and J. Carter Bacot, the chairman and chief executive of the Bank of New York, had been alerted to the problem.
Corrigan agreed to keep the Fedwire open as late as possible on Thursday night, to give the bank additional time to fix its system and catch up on the mountainous logjam of unprocessed business. In a fortuitous decision, he also told his legal team to immediately prepare the paperwork for an emergency overnight loan to the bank, just in case.
As the afternoon slipped into the evening, the Bank of New York’s overdraft at the Fed continued to grow. By 8 p.m., the bank was overdrawn by almost $32 billion. At that point, Corrigan sent an envoy to hand-deliver the special loan documents to the bank’s headquarters; the bank’s chief financial officer signed them, just in case. The documents called for the bank to pledge, as collateral for the loan, every penny of its domestic assets, a stack of collateral that totaled about $36 billion.
Finally, a patch held up and the bank was able to process enough business by 10 p.m. to reduce its overdraft by about $8 billion. But these transactions were moving slowly through the computer system, and people were getting tired.
Soon after midnight, Corrigan called Carter Bacot in his war room at the bank—they had talked countless times since 10 p.m.—and told him that 1 a.m. was positively the last Fedwire deadline for the bank. Almost as soon as Bacot hung up, he was told that the latest computer patch had failed and the computer had crashed again.
At 2:15 a.m. on Friday, November 22, the New York Fed officially made an historic $23.6 billion loan to the Bank of New York, the largest overnight loan it had ever made, a loan many times larger than the credit the Fed extended to Continental Illinois in its hours of crisis.
And there was still the rest of the Treasury market to worry about.
When the business day opened on Friday, the Bank of New York system was still not fully operational, and it once again began to overdraw its account. By 11:30 in the morning, with another $2 billion added to the bank’s overdraft, Corrigan made another fateful—and almost disastrous—decision.
He ordered the Fed to stop accepting securities being delivered by other banks to the Bank of New York’s account at the Fed, as the bank could not pay for them. He explained that he was trying “to see whether it was practical to prevent further increases in the overdraft without causing excessive disruption in the market.”
By lunchtime on Friday, Corrigan had his answer—and it was “absolutely not.” Within an hour of his decision the vast Treasury market had started to freeze up. Wary traders were suddenly unwilling to complete trades with one another. Some were trying to back out of deals they had already done with firms that relied on the Bank of New York for back-office processing. And the banks that could no longer complete transactions through the Bank of New York were suddenly incurring their own Fedwire overdraft problems, because they could not get paid.
Anyone in Jerry Corrigan’s shoes might well have recurring nightmares about what would have happened on that nerve-racking Friday if the Bank of New York had not been able to repair its computer system as quickly as it did. Fortunately, by 1:30 p.m., the bank’s processing system was more or less functional and Corrigan was able to lift the temporary and almost ruinous freeze on the bank’s account at the Fed.
Weeks later, he tried to share the awesome warning delivered by this episode with the congressional committees that quizzed him about this unprecedented “bank bailout.” As things now stood, Corrigan cautioned his critics, there was no other option. “It is unrealistic to think of ‘disconnecting’ a major participant except in circumstances which, in the end, might require closing the market as a whole.” And closing any large market “as a whole” could create problems that would dwarf those the system had just survived.
Obviously, there was no such thing as a “minor computer problem” if fixing it meant more than a very brief shutdown of any part of the financial market’s circulatory system. If a shutdown lasted longer, or had to be done in the middle of some other crisis—well, as Jerry Corrigan would say, “Katie, bar the door.”