Contrarian Investment Strategies

by David Dreman

  Then there was index arbitrage (also called program trading), which allowed S&P index funds and scores of other major institutions to buy or sell either the S&P stocks or the S&P futures if the prices of either one got out of line with the other. If, for example, the value of all the stocks in the S&P 500 was at 250 and S&P futures were trading at 252, an alert institutional investor could, using computerized trading, quickly short sell the S&P futures and buy the stocks, locking in a return of 0.8 percent before commissions. When the futures and the S&P index went back to their normal value, the transaction would be reversed. Such a small return seems hardly worth the effort, but if the sums were large, say $100 million, if the purchases were on minuscule margin, and if the transactions were repeated many times in a month or two, the profits could be very large.

  PORTFOLIO INSURANCE: THE FINAL COMPONENT OF THE CRASH

  Portfolio insurance was a product designed to protect the capital of institutional investors in a down market while giving them all the upside in a rising one. In brief, it was a way to have your cake and eat it too. First of all, it was not insurance at all, although it was slicky marketed as such. It purported to protect an institution’s portfolio, if markets went down, by short selling S&P 500 futures. The more the market dropped, the more futures would be sold, progressively lowering the institution’s stock exposure to the market.*26 If the portfolio dropped 3 percent, the computers of Leland O’Brien Rubinstein Associates and other portfolio insurance firms, which licensed their model, would sell futures akin to 10 percent of their portfolio. If the S&P went down another 3 percent, another 10 percent of the futures would be sold. If the S&P subsequently went up 3 percent, the 10 percent futures position would be repurchased and little would be lost, and so on.

  If the price of the S&P 500 went down, the futures would be sold, not the stocks of a pension fund, a hedge fund, or any other institution that owned them. Futures were considereded exceptionally liquid, and, markets being totally efficient according to EMH, could be sold in the blink of an eye. According to its adherents, it was a perfect system, said to maximize market gains and protect against losses, supposedly at only a small cost to the owner.

  Why? Because the S&P futures had infinite liquidity. This is a central belief of efficient marketers and the core concept of portfolio insurance. If markets went down, swarms of sophisticated buyers would come out of the woodwork and immediately buy the futures at lower prices. If the market went still lower, even larger hordes of buyers would be there to scoop up the depressed values. After all, a cardinal EMH belief was that since markets were entirely rational, increasing numbers of buyers would appear as prices declined. End of discussion.

  “Liquidity begets liquidity,” the academics chanted. But lo and behold, they turned out to be all wet.

  Did portfolio insurance actually protect a portfolio? Well, no. It would lessen the institution’s losses if executed well, but it could not prevent them, nor was it designed to do so. Moreover, if a market was quite volatile it might end up costing big, even if prices did not move much, because of the small losses and commissions generated by frequently getting into and out of the market.

  The keen-eyed Barron’s editor Alan Abelson cut through the hype immediately. “It is an exotically labeled version of the small speculator’s stop-loss order,” he wrote. Under its complex mathematical formulas, it was simply another market-timing device. The previous chapter briefly discussed the uninspiring record of market-timing products, but as with so many others, this one, with the backing of academics steeped in EMH lore, became wildly popular. The consistency of academic thinking on market timing might seem a touch questionable here.

  The mathematical calculations for portfolio insurance were based on the Black-Scholes Option Pricing Model, which for sheer complexity possibly rivaled the mathematical formulas designed to launch a space shuttle. Designed by Fischer Black and Nobel laureate Myron Scholes, this method soon became the standard way of pricing stock options. Hayne Leland and Mark Rubinstein, both in the Department of Finance at the University of California at Berkeley, adapted the model to portfolio insurance. Leland was known as the father of portfolio insurance, partly because of an article published in 1980 in the Journal of Finance, the most prestigious academic and pro-EMH journal.5 Behind the awesome mathematics was a disarmingly simple but fallacious idea: you could all but guarantee yourself against downside by eliminating risk when the market dropped, clients were told, while reaping the benefits of a bull market by keeping a larger percentage of the portfolio in stocks.

  Major pension and other institutional funds were keen to line up to participate in this sure thing. At one New York investment conference in the summer of 1986, the number of institutional investors who wanted to attend a workshop on portfolio insurance became so large that the workshop was expanded into the entire meeting. “It was madness,” said John O’Brien of Leland O’Brien Rubinstein.6 But it was madness he liked, as his firm was reeling in tens of billions of dollars of institutional assets that it managed. According to the Brady Commission report on the 1987 crash, $60 billion to $90 billion of funds were invested in portfolio insurance programs just prior to the crash.

  PREVIOUS WARNINGS IGNORED

  Something was wrong. Perhaps it seemed coincidental at first, but the vast growth of index arbitrage, portfolio insurance, and index and stock options didn’t do what the efficient-market advocates had assured regulators and the public they would: decrease market volatility. Quite the opposite: there were violent stock market downturns of almost 100 points, first on July 7 and 8, 1986, and then came a minicrash on September 11 and 12 of the same year, followed by another near-collapse in the averages of almost 100 points in late March 1987. Concern began to mount among a number of senior investment professionals, including John Phelan, the president of the New York Stock Exchange. I wrote a column in Forbes entitled “Doomsday Machine,” with the subtitle financial index features + program trading*27 + low margins = potential disaster. The article, published on March 23, 1987, described in detail how the interaction of index arbitrage and portfolio insurance would cause a crash. Unfortunately, it happened pretty much as described six months later.

  Because of the Forbes article, I was invited to sit on a panel with some eminent academics, the majority of whom strongly backed portfolio insurance and index arbitrage trading techniques. Also present were three commissioners of the SEC, all believers in efficient markets. The conference was held at the University of Rochester, New York, a bastion of EMH, in June 1987. Almost to a man, the prestigious academics on the panel scoffed when I spoke and stated why I felt a market crash could happen. Some of the great men snickered; others merely yawned politely. Who was a mere Wall Street money manager to debate with professors of their stature?

  Worse was in store for the two panelists holding views similar to mine. Robert Shiller, whom we met earlier, was taken to task and practically called a heretic. “With your training, how can you subscribe to such ridiculous views?” a member of the audience demanded. Jay Patel, then a professor of finance at MIT, received similar treatment. Despite three sharp downturns caused by the violent interactions of index arbitrage and portfolio insurance, those vehicles were not only given a clean bill of health at the meeting but were considered essential to keeping markets efficient. Then the unthinkable happened.

  THE CRASH THAT WAS THEORETICALLY IMPOSSIBLE

  What surprised me about the 1987 crash was the ferocity of the decline, in spite of what I knew about the dangers of the interaction of portfolio insurance and index arbitrage, as well as the tens of billions of dollars in play. The darkest scenarios I could envision didn’t even come close.

  Although many explanations were bandied about, there were no major events responsible for this crash, nor was the market wildly overvalued. Sure, there was some concern in the marketplace—there always is—but the anxiety level was below amber. What was disturbing was that a long-overdue correction was distorted into a disaster by
the new S&P derivatives, which in times of rapidly falling prices magnified the drop sharply. That’s exactly what happened on Black Monday, October 19, 1987. The result? A minor correction in the Dow turned into a 508-point panic for the day. Wow.

  The Dow Jones plummeted 22.6 percent on Black Monday alone, a plunge almost twice as large as the percentage drop on Black Tuesday, October 29, 1929. It started on Wednesday morning, October 14, 1987, with the Dow just over 2,500. By noon the next Tuesday, October 20, the Dow was slightly above 1,600, a decline of more than one-third. It was the worst crash in U.S. history to that time. (The 508-point drop on October 19 was only 60 percent of the total decline.)

  What created this disaster? Let’s look more closely at the role of portfolio insurance and index arbitrage during those five terrifying days. A number of studies on market volatility, which examined index arbitrage and portfolio insurance, had previously been commissioned by the commodity exchanges after the sharp 100-point drops previously discussed. Those “impartial studies” were performed for the exchanges by consultants salivating at the giant commissions the programs were bringing them. The studies, like others we will examine later, were flawed, as they covered short and relatively quiet market periods.*28 (Remember the law of small numbers.) The researchers concluded that the interaction between the two would not affect market volatility. Famous last words.

  Who started the selling? Computers, followed by humans in shock. What looks good in the lab can act like nitroglycerin when shaken. Unfortunately, the nitro was shaken. Bam! Crash!

  The S&P was down 3 percent on Wednesday, October 14, and the portfolio insurance formula went into action. Leland O’Brien Rubinstein and Wells Fargo began to sell futures by October 14 or 15 as prices fell more than 3 percent. Still the S&P futures stayed at a small premium to the S&P, in line with where they were expected to be.*29 On Friday, October 16, the S&P 500 Index took a sharp drop of more than 5 percent and the futures went to a minuscule discount to the S&P 500 stocks, of less than two-tenths of 1 percent.

  One would think that the EMH gurus running the portfolio insurance money management firms would have been elated with how well their portfolios were performing in a rapid market downturn. The truth was that they were terrified. The liquidity of their catchphrase “Liquidity begets liquidity” was starting to dry up. Buyers of futures were becoming increasingly scarce, and they lowered their bids as the decline in S&P futures increased. EMH theory said that this was impossible, but it was happening. The result was that the portfolio insurers could sell only a small fraction of the S&P futures their formulas called for. By Friday, October 16, with an estimated $60 billion to $90 billion in portfolio insurance and sales far behind the PI formulas required, it would turn into a serious problem if the market went lower the next week. But what if the unimaginable happened, EMH liquidity theory was wrong, and liquidity dried up? The problem would become a nightmare.

  The professors at the portfolio insurance firms may or may not have known about another equally devastating problem. Like the British who used the Enigma machines in World War II to decipher the most important German military codes, thus learning exactly what the enemy would do, the major brokerage houses and S&P futures traders also knew almost precisely how much the portfolio insurance managers still were behind in their futures selling. It wasn’t nearly as complex as Enigma to decode. Remember, if the market dropped 3 percent, the insurers were supposed to sell 10 percent of the futures to reduce the equity exposure to the predetermined lower level in order to remain delta neutral (the level of stocks and bonds the formula called for).

  The academics turned portfolio insurers made another small error. They didn’t count on how cutthroat the brokers and the futures traders who acted for them, as well as other players, could be. After all, that wasn’t factored into EMH theory. But everyone in the game knew the portfolio insurance managers were well behind the predetermined selling they were required to do. If the other brokerage houses and traders sold futures en masse or shorted stocks quickly early on Monday, they would get the jump on the portfolio insurers and make jillions buying the futures back at much lower prices. The stage was set.

  On Black Monday, October 19, 1987, the impossible happened: liquidity dried up completely. The market opened and began to drop rapidly, with heavy S&P 500 futures selling leading the way, dropping 3.5 percent on the first trade. Every knowledgeable pro was shorting S&P futures. As the futures dropped, computerized program traders (index arbitragers) rushed in, buying the discounted futures at increasingly lower prices and selling stocks, dropping their prices sharply. This forced the portfolio insurers to sell futures even more frantically at increasingly large discounts to the value of the S&P 500 as they fell further and further behind the amount they had to short to protect their clients’ portfolios. The death spiral continued. Major investment houses and commodity traders, knowing the desperation of the portfolio insurance managers, continued to short futures ahead of the portfolio insurers. The portfolio insurance managers were forced to sell at any price to realign their portfolios. So utter was the collapse of portfolio insurance that as the market continued to free-fall, the theory blew apart completely. S&P futures traded at such a large discount to the S&P 500 that portfolio insurance managers were forced to sell their clients’ stock, which they had never conceived of selling under any circumstances. This pushed the S&P stocks down even further.

  This doomsday interaction between portfolio insurance and program trading continued through the day and was strongly reinforced by larger and larger amounts of selling by institutional and individual investors reacting to the panic and stampeding for the exits. Prices dropped more than ever before in modern times.

  The panic had spread like a rampaging forest fire out of the futures pits of Chicago into the stock market itself, where it sharply magnified the impact. The crapshooting spirit of the commodity pits had been injected into the floors of the country’s major stock exchanges, with devastating results.

  THE IMPOSSIBLE HAPPENS

  The S&P 500 was down 20.5 percent at the close on October 19, and the S&P 500 futures dropped 28.6 percent; this left them at a 23.3-point discount (10.4 percent) to the S&P 500. According to EMH theory, futures cannot ever fall below the cash price for the same expiration date (in this case the end of December 1987). The reason is simple: when an investor buys a stock or the S&P 500 on margin, he or she has to pay the broker an interest charge to carry it. The interest (about 5 percent at the time) is priced into the S&P 500 futures contract, so it must always be at a premium to the cash price of the S&P itself. This should apply even in a panic if people are rational. If the S&P 500 futures contract goes to a discount, the trader makes an automatic profit by selling the S&P 500 stocks and immediately buying the discounted futures. When the futures contracts expire less than two months later, he pockets his gains.*30

  But as we saw the S&P 500 futures didn’t stay at a premium; they dropped 10.4 percent below the cash price. This gave knowledgeable investors such as index funds and other investment firms with cash an incredible return. They simply had to buy the future and sell or short sell the S&P 500 stocks instantaneously.

  Let’s go through the mechanics to see how this was done. Since they had to put only 5 percent down on the futures because of the ridiculously low margin requirement, they made nearly 228 percent on the money they actually put down in margin, on an absolutely safe, completely hedged investment. When the contract expired in late December 1987, they would sell the S&P stock portfolio at the same time the futures contract expired; this is rudimentary index arbitrage. At expiration of the futures’ contract, the price of the S&P futures, bought at a large discount to the cash price, would rise to the price of the S&P 500 stock index, capturing the full original 10.4 percent discount on the S&P 500 futures.

  Take the example of an institution or brokerage firm that bought $10 million of futures for its own account. It would have put up only $500,000 in margin and would come up with $1,1
37,900 in profit on the contract’s expiration, or the 228 percent shown before.*31 This action makes a mockery of the omnisciently rational investor of EMH theory.

  As noted, this is the type of riskless, high-return transaction that EMH theorists state can never happen in an efficient market. Occurring in one of the most widely traded, carefully followed markets in the world, large, riskless profits available to knowledgeable investors are a very serious challenge to EMH. No explanation by the EMH theorists, to my knowledge, has ever been forthcoming. This event also puts an enormous dent in Professors Roll’s and Fama’s theory of rational crashes, which we’ll look at shortly.

  As I see it, many of the major pillars supporting the efficient-market hypothesis were destroyed in the 1987 crash. Five should be noted.

  1. Liquidity dried up. A crucial assumption of EMH, that there is always sufficient liquidity in markets, was disproved, as the above events indicate. The lack of understanding by the academics and the portfolio insurance managers of liquidity was a primary cause of the 1987 crash.

  2. The argument that rational investors keep stock prices in line with their value, another core assumption of EMH, was seriously challenged. When was the market efficient? When professional and other sophisticated investors took the S&P 500 down from 315 to 216 (31 percent) in five trading days or when it recovered all of its losses nineteen months later, with only minor changes in the underlying fundamentals?

  3. Another example of glaring market inefficiency was the 28.6 percent drop in prices of S&P 500 futures to the actual 20.5 percent drop in the price of the S&P itself on October 19, allowing savvy investors to make gigantic profits on their capital in two months. The cornerstone of EMH theory states that such actions by rational investors are impossible because of their automaton-like rationality.