Charlie [Munger] and I are of one mind in how we feel about derivatives and the trading activities that go with them: We view them as time bombs, both for the parties that deal in them and the economic system.
—WARREN BUFFETT
© Phil McCarten/Reuters/Corbis
It was December 10, 1997, and Myron Scholes was among the dozen laureates gathered on the stage in the Stockholm Concert Hall to receive the annual Nobel Prize in the memory of Swedish industrialist Alfred Nobel. Scholes was sharing the 1997 award for economics—officially, the Sveriges Riksbank Prize in Economic Sciences—with his colleague Robert Merton. Twenty-five years earlier they had created what is known throughout the financial world as the Black-Scholes option pricing model and what the Nobel committee described as “a new method to determine the value of derivatives.”
In his Nobel Prize lecture, Scholes cited Alfred Nobel’s stipulation that the prize was to be awarded for an “important discovery or invention.” Scholes pointed out that he and Merton did not invent derivatives; options contracts, he reminded his audience, were traded on the Amsterdam Stock Exchange as early as the seventeenth century and actively traded on the Chicago Board of Trade since the 1930s. But despite his acknowledgment, the Black-Scholes model was indeed a discovery, because it provided a breakthrough in the understanding of the risk and return characteristics of options. Furthermore, the model was presented with such clarity and simplicity that it quickly led to dramatic growth in the uses and markets for stock options and in a dizzying variety of newly created derivatives beyond the simple option contract.
For most Nobel Prize winners, the generous award from the Nobel committee represents a once-in-a-career windfall. But for Scholes, and for a somewhat lesser extent Merton, the Nobel Prize was a pittance compared to the fortunes they had already amassed as partners of Long-Term Capital Management, the hundred-billion-dollar hedge fund they helped cofound in 1993—and whose strategies included real-world applications of their model.
Within a year of the receipt of their prizes, however, the LTCM hedge fund collapsed and their fortunes evaporated. The failure of their once high-flying hedge fund serves as a reminder that brilliance does not always ensure financial success. It also may reinforce Warren Buffet’s famous contention that “derivatives are a form of mass financial destruction.”
The Making of a Finance Scholar
Scholes’s fascination with the stock market started when he was growing up in Timmins, Ontario, a gold-mining area in the remote reaches of Canada. As he recalls, however, his parents’ occasional forays into the “penny stocks” issued by regional mining companies were largely unsuccessful. Likewise, his own investments in the stock market while in high school in Hamilton, Ontario, and in college at McMaster University did not pan out, despite long hours devoted to studying the market. The world of finance was not an apparent fit for him, and he planned to work for his uncle’s publishing business after he finished an MBA at the University of Chicago. Loquacious, confident, and dashing, with a drive for business success, Scholes seemed to be the logical family member to eventually take over that business.
But in 1963, while still enrolled in the MBA program, Scholes found a summer job as a junior programmer in the university’s computer facility. That position changed his career path. At the time, he knew next to nothing about programming and very little about academic finance. But his understanding of the financial markets and computer science quickly improved. During the 1960s, the University of Chicago had become a hotbed of finance research, and Scholes found himself working closely with professors who would, like him, eventually become Nobel Prize winners. He learned from them, but they also learned from him as he began to make suggestions to his eminent professors on the format of their papers. He described the four-and-a-half months at the computer lab as a time of “falling in love with computers and the researchers that I met.”1 One of those researchers, Merton Miller (a Nobel laureate in 1990), steered him into Chicago’s PhD program, at which point Scholes abandoned any notion of returning to Canada or his uncle’s business.
After completing his doctoral work in 1968 (with a dissertation focused on the dynamics of securities trading), he joined the faculty of the Massachusetts Institute of Technology. At MIT he met Fischer Black, then a consultant with Arthur D. Little, and began work on what would eventually become known as the Black-Scholes option pricing model. In 1973, Black and Scholes published their results in the Journal of Political Economy, in an article titled “The Pricing of Options and Corporate Liabilities.” That article, written in the dry and mathematics-laced language of an academic journal, provided the intellectual framework for the pricing of derivatives—and in a short time added a new dimension to the business of Wall Street. (Fischer Black was recognized by the Nobel committee for his contribution to the Black-Scholes model. But he died in 1995, and since the Nobel Prize is not awarded posthumously, he is not listed as an official laureate.)
A Brief Primer on Options
The derivatives revolution had its roots in the world of common stock options—“puts” and “calls.” Options give investors the ability to profit from changes in the price of a stock without having to actually buy the stock.
Suppose an investor holds a view that within a few months AT&T’s common stock will increase in price. To act on that opinion, she could purchase 100 shares of the company’s stock outright at the market price of, say, $34.75 per share, for a total of $3,475. If after three months she was correct and the market price of the shares moved up to $36.50 per share, she could sell the 100 shares for $3,650 and realize a $175 short-term profit on the transaction. That’s a 5 percent return on her investment for the three-month holding period; on an annualized basis, the return would be 20 percent.
That is an attractive return on investment, but the options markets provide a way for her to greatly magnify that return with a much smaller investment. Rather than investing $3,475 to act on her hunch, she could buy a three-month call option (hence, a “call”) that would allow her to purchase the stock at a “strike price” of, say, $35 per share—an amount that is just slightly above the current $34.75 market price. That call option would likely cost around $0.50 per share ($50 for the 100 shares) and would give her the right (but not the obligation) to purchase the 100 shares of AT&T stock at $35 per share at any time during the following three months. After three months the option would expire. If, as expected, the price rises to $36.50, she could “exercise” her call option to buy 100 shares of AT&T at the $35 per share strike price—and simultaneously sell the shares at $36.50 in the stock market for a $150 profit. After subtracting the $50 cost of the option, she would realize a $100 net profit, which amounts to a 200 percent return on her $50 option investment during the three-month period; annualized, the return would be 800 percent.
The downside? To make a profit, the stock price must be above the exercise price by at least $0.50 per share in light of the option’s $0.50 cost; that is, the stock must move up to at least $35.50 during the three months. If she exercises the option when the stock is selling between $35.00 and $35.50, she will make a small gain, but not enough to defray the total $0.50 per share cost of the option. And if the stock never climbs above $35 per share, the option will expire worthless, a 100 percent loss. In reality, the majority of call options expire without value, resulting in pure gain for the seller of the contract (called a “writer” in the lingo of option trading) and a total loss for the investor. (This is not to say, however, that the option purchaser is necessarily at a disadvantage; as the examples above illustrate, positive returns may not be frequent, but they can be very large when an option performs in accordance with the holder’s expectations.)
A “put” is the flip side of a call, giving the investor the right to sell shares of common stock at the strike price on or before a set date. The mechanics and costs of a put transaction are similar, but in this case the investor is speculating that the price of the stock will fall. If the investor is correct about a lower stock price in the future, the investor will produce a profit through a simultaneous purchase of the stock in the market at the lower price and sale to the writer at the higher contracted strike price.
Model Building
Of course, success in options trading is contingent on the ability to predict the future price of the stock, and so would seem to rely on fuzzy and immeasurable variables like the investor’s risk profile or the economic “utility” of an option—or the instincts of street-smart traders. The Black-Scholes model did away with such fuzziness. The first version was fairly simple; all one needed to include were variables that can be directly observed and precisely measured. To find the option price in the emerging model, one just plugged in the volatility of the stock’s price, the current stock price, the current interest rate, and the two governing terms of the option contract—the strike price and the option’s expiration date.
The most important of those five variables was the stock’s price volatility—and it was also the trickiest to get right. Among finance academicians, a stock’s risk is measured by the variability of its price over time and, specifically, by the standard deviation of its returns. It only makes sense that the price of an option contract for a volatile stock will exceed that for one that is less volatile. For instance: historically, AT&T stock does not fluctuate as widely as the overall stock market. However, a stock like Yahoo fluctuates more widely than the market. For several years, the stocks of Yahoo and AT&T sold for similar prices (between $30 and $40 per share), but the options of Yahoo were always much more expensive. That was because the Yahoo stock was more volatile and therefore more likely than the AT&T stock to exceed the call option strike price during the contract’s time frame, thereby earning an attractive trading profit for the holders of Yahoo’s call options. And as Yahoo was the more volatile stock, its shares were also more likely to fall below the option’s strike price and produce a profit for holders of put options. Consequently, a Yahoo option, whether a put or a call, was a better speculation for profit than the AT&T option and therefore always fetched a higher price.
But in constructing their pricing model, Black and Scholes encountered a major practical problem with volatility: by the time one accumulated enough data to measure volatility correctly, it was too late to have any immediate value. In fact, the early form of the Black-Scholes model, based on the initial empirical tests, would have produced significant losses for option traders. As Scholes would explain it, “These losses were incurred because using simple estimates of the volatility ignored information on future volatility that the market was using to price the options.”2 So at that point, Scholes and Black turned to another MIT finance professor, Robert Merton, who literally brought rocket science to bear on the problem. Using a branch of mathematics called Itō calculus—the same mathematics that makes it possible to continuously adjust the trajectory of a rocket in flight—Black and Scholes modified the final version of the model to adjust to changes in the price and risk of the underlying stock in real time.3 With that crucial refinement to the Black-Scholes model, investors for the first time had a way to eliminate risk, the four-letter word of finance. By knowing the correct price to be paid for an option, they presumably knew how much it would cost to eliminate a stock’s risk.
At the time the Black-Scholes model was published in 1973, the Chicago Board Options Exchange had, coincidentally, just been organized as the first centralized exchange for option trading. Wayne Luthringshausen, the longtime chairman of the Options Clearing Corporation, recalls that before the CBOE was founded, the average number of option contracts traded was a mere 9,000 per month.4 Now, with the new Black-Scholes model, retail investors and institutional portfolio managers had a way of precisely protecting a capital gain through the purchase of a put; or, conversely, they had a way of realizing a specified capital gain without putting much cash at risk. In 2008, the CBOE reached a level of one billion contracts per year, and observers of this phenomenal growth point to the Black-Scholes model as the necessary catalyst.
In his Nobel Prize acceptance speech, Scholes rightfully gave credit to the model for spurring the growth of the options markets:
Although it is hard to prove, I do think that the success of the CBOE and other exchanges, in part, can be attributable to option-pricing models. As traders became familiar with these modes, bid-offer spreads narrowed. As traders became more familiar with risk-management techniques they could take on larger position sizes to support the market. With a deeper and more efficient market, investors began to use options to facilitate their own investment strategies.5
Former CBOE vice chairman, Gerry Lahey, is less equivocal about the impact of the Black-Scholes model on options trading: “There’s no question that without the Black-Scholes model we would not be here today. It is clearly the most significant formula developed during my lifetime and maybe before.”6 The CBOE, which began operating in the smoking room of the Chicago Board of Trade, now conducts a worldwide options business from its multistory building in downtown Chicago.
Today exchange-traded options are available for most publicly traded common stocks, and the shares of large capitalization companies—Exxon, IBM, General Electric, Apple, Amazon, and the like—have become the basis for a multitude of options that trade based on specified strike prices and expiration dates. There is not just a single put or call option on those companies’ stocks; there are usually more than one hundred different options contracts available on each. Wall Street traders routinely use Black-Scholes calculators, and an explanation of the Black-Scholes option pricing model has become a part of every textbook on investing.
With the remarkable growth in the market for options and other derivative instruments, the investing and noninvesting public has become much more aware of their presence. Yet misunderstandings remain, including how derivatives are created and how their economic impact is measured. One common misconception, for instance, is that companies create their own puts and calls for their stock. And while it’s true that AT&T’s management may create employee stock options for key managers as a form of incentive compensation, these kinds of employee stock options are not listed on an exchange and are a very different breed of option from a traded derivative. The mushrooming of option derivatives traded on the CBOE is not a concoction of the companies whose stock serves as the “underlier” for the option but purely of the writers, usually trading firms. (As noted earlier, most options expire unexercised and without value, like losing lottery tickets and horse race bets. But that doesn’t usually result in pure profit for the writer, who is typically providing liquidity to the market by taking hedged positions on both sides of the trades.)
Another misunderstanding about derivatives has to do with “notional value,” an amount that is often used to describe the size of derivative markets. While the AT&T option contract described earlier has a market value of around $50, the amount of common stock under the “control” of the contract is much larger, about $3,500. The latter is the notional value and explains why reference to notional value can produce preposterous-sounding numbers. In the case of the Yahoo and AT&T examples, for instance, the combined notional value of the hundreds of options contracts on their stocks is likely to be in excess of the companies’ market capitalizations.
Excursions from the Academy
In 1973, the year he published his option pricing model, Scholes moved back to the University of Chicago, where he stayed until he took a position at Stanford University in 1981. Like many other academic all-stars on the faculties of business schools, he found lucrative consulting opportunities and board positions in the private sector. He worked for Wells Fargo Bank in the late 1960s and early 1970s when the bank began to structure one of the earliest index funds—preceding John Bogle’s retail-oriented Vanguard products by several years—and later was active in the business of Dimensional Fund Advisors, a firm that developed index-like products for institutional investors. Scholes also acted as a consultant for a time to Donaldson, Lufkin & Jenrette, an investment firm that was building its option technology in anticipation of trading on the CBOE. His job at DLJ, as he described it, was to “marry the old-time trader types, with their mental set, with young mathematical modeling types, with their model assumptions, to help add value for the firm.”7 (William Donaldson, one of the founders of DLJ, is the focus of chapter 13.)
Another extracurricular venture began in 1990 when Scholes became a special consultant for Salomon Brothers, a major Wall Street firm at the time. By then, derivatives had grown well beyond puts and calls on common stocks. With the growth of index funds, put and call options were beginning to be applied to exchange-traded funds (ETFs). Most ETFs have the same investment objectives as index mutual funds, but they can be traded like common stocks—and therefore options can be written on them. The early ETFs tracked broad market indexes such as Standard & Poor’s 500 Index, but later ETFs spread to hundreds that track indexes based on industry sector, country, market capitalization, and other criteria.
By the 1990s, derivatives had also spread beyond equity securities to bonds and other fixed-income instruments and to currencies. Aided by powerful computer technology and sophisticated quantitative techniques akin to the Black-Scholes model, booming markets soon arose for debt-based derivative products. Such products included interest rate swaps—the typical arrangement involves two borrowers (“counterparties”), with one swapping its fixed-rate interest payments for the other’s floating-rate interest payments. Another debt-based instrument is a credit default swap—a form of credit insurance in which the buyer of the swap (usually the holder of a loan) makes periodic payments to the seller of the swap and, in the event the loan defaults, receives a payoff of the total loan amount. The seller of the swap receives those periodic payments and, upon default, takes possession of the loan, receiving whatever value it might eventually provide.
The logical place for the development and implementation of those new markets was at Salomon Brothers, which at the time was the world’s leading firm in trading bonds and other fixed-income securities. No one was better suited than Scholes to assist that firm in extending its leadership in fixed-income to derivatives. Scholes was awarded a lucrative managing directorship and later became the cohead of the fixed-income derivatives sales and trading department, all the while continuing in his teaching and research positions at Stanford.
But the position at Salomon didn’t last long; in 1991 one of its traders in government bonds fatally sullied the firm’s long-held reputation with the Federal Reserve by submitting false bids to manipulate the price of bonds that were newly issued by the U.S. Treasury. The discovery of the violation led the Treasury Department to revamp its auction procedures—and eventually led to the sale of the weakened Salomon Brothers firm to Travelers Group and, ultimately, its demise. But while Scholes’s position with Salomon lasted only a few years, it started a chain of events that would have a profound effect not only on Scholes but on the financial markets as a whole.
The Rise of a Celebrity Hedge Fund
The full-scale housecleaning at Salomon that followed the bid-rigging revelations included the forced resignation of John Meriwether, the forty-four-year-old head of bond operations. By all accounts, Meriwether was innocent of any involvement in the Treasury manipulation, and the traders who had worked for him remained fiercely loyal, especially those in Salomon’s highly profitable arbitrage group. In 1994, Meriwether formed a new hedge fund called Long-Term Capital Management and recruited many of his former top traders and arbitrageurs at Salomon to become his partners. Most of those partners held PhDs (overwhelmingly from MIT), and with the new developments in derivatives and financial engineering, Meriwether believed their collective trading prowess would propel LTCM far beyond the successes they had enjoyed at Salomon. Scholes was one of the people he recruited to help him achieve his goals, along with Merton, one of the codevelopers of the Black-Scholes model, and David Mullins Jr., who resigned his position as vice chairman of the board of governors of the Federal Reserve to become an LTCM partner.
The addition of Scholes seemed to make special sense. His brief consulting assignment with Salomon had given him an insider’s view of how the traders operated, and he had an unequaled understanding of the underlying mechanics of derivative strategies that would become an important part of LTCM’s operations. Moreover, he was also an expert in taxes (he was the lead author of a textbook, Taxes and Business Strategy, written while he taught at Stanford) and designed tax shelters to reduce the expected heavy tax burdens that LTCM’s investors would face. The fund, no surprise, was domiciled in the Cayman Islands.
As it happened, however, Scholes was at least as important in promoting the fund as in overseeing its management and handling its taxation issues. Hedge funds, like venture capital and private equity funds, are financed by large institutional investors and getting a fund off the ground depends on the reputation of the founding partners and their ability to convincingly communicate the new fund’s strategies. Meriwether, despite the scandal, had preserved his reputation, and by using Salomon’s publicly available disclosure documents he could demonstrate the success of his arbitrage group. But he was modest and somewhat withdrawn and did not make much of an impression during LTCM’s road show.
Scholes, however, was all the things that Meriwether was not. He was confident, at times to the point of arrogance, and after years of practice in the classroom, articulate and often entertaining. He characterized LTCM’s planned strategy of making many trades that produce small profits with minimal risks as “vacuuming up nickels”—and would then pull a nickel from the air in the manner of a magician. Prospective investors generally saw Scholes as the smartest guy in the room, and he helped raise much of the initial $1.25 billion of investor capital.
Vacuuming Up Nickels
After the eleven general partners and their trading staff settled into LTCM’s headquarters in Greenwich, Connecticut, they got down to the business of vacuuming up nickels through a combination of investment strategies. In the spirit of the first hedge funds, LTCM set off to balance long positions with short positions, thus setting up a hedge. (In shorting a stock or other security, the investor is betting that its price will fall; going long is betting it will rise.) Meriwether and his partners engaged in paired bond investments that involved simultaneous long and short positions of similar investments, betting most often on changes in “spreads.”
When bond traders refer to a “spread,” they mean the difference in the interest rates between two bonds. Long-term bonds, such as twenty-year U.S. Treasury bonds, will almost always have a higher interest rate than a medium-term bond, such as a five-year U.S. Treasury note. The difference between the two interest rates is the spread, and the amount of that spread tends to conform to quantifiable factors that can be analyzed with mathematical modeling techniques. If LTCM’s models calculated that the existing spread wasn’t reasonable given economic and market forces, a trader would structure a “convergence trade,” betting that the anomaly would soon right itself—that one interest rate would fall and the other would rise to reestablish the “correct” spread. Since the expected change in interest rates would also change the bond prices, the trader sets up the bet by buying one bond long and selling the other short.
LTCM made other convergence bets with pairs that were not as similar as two U.S. Treasury obligations but that had some rational and historical relationship: spreads between investment-grade bonds and junk bonds; between U.K. bonds and German bonds; between Australian bonds and Canadian bonds; and between far-riskier bonds issued by countries in emerging markets. Another part of their convergence strategy was to diversify by type of bond and the patterns and frequency of their spreads, so that bets do not all go awry at the same time. During its brief life, LTCM made thousands of such investments based on the assumption that interest rates would realign according to their models, resulting in more nickels for the fund.
Through the end of 1997, the nickels were flowing in. Concentrating mainly on convergence plays in the bond markets, LTCM made money month after month. For one remarkable stretch, the hedge fund recorded a profit for nineteen consecutive months, an accomplishment comparable to a sports team enjoying a nineteen-game winning streak. Further, the fund never experienced two consecutive months of losses, and those few losses were relatively small and well within the expectations for a hedge fund.
All of those profits were plowed back into LTCM’s business, and the fund’s equity capital account soared from $1.25 billion in 1994 to $7.4 billion at year-end 1997. Even better, the seemingly prudent risk-management techniques were keeping the investors’ loss exposure to a minimum. Based on value-at-risk—a technique pioneered in the early 1990s by traders at J. P. Morgan that estimates the maximum loss in a trading portfolio within the normal bounds of probability—LTCM looked like a financial fortress. Most of its trades—around seven thousand of them at one point—were of the off-setting, long-short variety; they were diverse, and the correlation between them was low. As a result, LTCM’s “VaR” tended to fluctuate between $30 million and $50 million on any given day. That’s certainly a lot of money, but still less than 1 percent of LTCM’s equity cushion of over $7 billion—that is, with the heroic assumption that the underlying statistical assumptions would hold.
LTCM’s record of profitability, combined with its tiny VaR estimate, emboldened its lenders. By the end of 1997, a consortium of banks—a veritable Who’s Who of Wall Street financial institutions—had advanced about $125 billion of debt to LTCM, even though the fund had only $7.4 billion in equity capital. Using that high degree of leverage, LTCM was able to greatly extend its search for additional nickels. (Some observers modified Scholes’s nickel analogy to take into account the risk of its extraordinary leverage, saying that LTCM was picking up nickels in front of an advancing steamroller.)
Whatever the risk, however, institutional investors in LTCM indeed enjoyed outsized returns on their investment. Even after subtracting the steep fees paid to the managing partners, the investors realized a 43 percent return on their invested capital in 1995, the first full year of LTCM’s operations, followed by a 41 percent return in 1996. By comparison, their return in 1997 was a “disappointing” 17 percent. Insiders like Meriwether and Scholes did even better in those initial years, since they were beneficiaries of the management fees. The insiders saw their personal accounts grow to a combined $1.9 billion.8
But by the end of 1997, LTCM’s very success made the possibility for continuing success uncertain. By scouring the world for profitable convergence plays with their $100 billion-plus war chest, they were picking the market clean of profitable opportunities. Moreover, a host of imitator hedge funds began to adopt LTCM’s strategies, further diminishing attractive prospects. And much more alarming, the unsettled Asian markets of the late 1990s produced an “Asian flu” that was growing more virulent and widespread by the day, causing currency and securities prices to plummet across the world and upsetting the market stability that LTCM needed to execute its closely calibrated convergence trades.
If anyone could understand LTCM’s growing challenges, it was Scholes, and—tellingly—he refrained from putting his Nobel Prize money into LTCM, despite saying in an off-the-cuff interview in late October of 1997 that he would “most likely” do so.9
Changing Strategies
In response to the falloff in the investors’ returns during 1997, LTCM embarked on two strategies intended to restore its profitability levels to those of its earlier years: diversifying well beyond bonds and interest-related convergence plays and increasing the fund’s use of financial leverage. Both strategies proved disastrous.
The diversification initiatives were various but focused initially and most aggressively on “equity vols,” shorthand for a financial instrument whose price is based on expected volatility in the stock market. Although the Chicago Board Options Exchange was developing its own volatility index in the early 1990s—today regularly traded under the name and symbol “VIX”—it did not begin active trading the index until 2004. LTCM was unmistakably one of the first movers in the 1990s when it came to trading volatility—and that made sense, because the equity vol is tied directly to the Black-Scholes model.
Given that the main determinant of an option price is the volatility of the price of the underlying stock, it follows that movements in option prices themselves give investors an idea about the expected volatility of the overall market. Sparing the reader the sophisticated mathematics underlying an equity vol, suffice it to say that the instrument provides a means for investors to hedge or speculate on the amount of change—upward and downward—in stock market volatility. Who would be affected by a greater or lesser amount of future volatility? Option traders for one. Whether puts or calls, the greater the volatility of the underlying stock, the more valuable the option. So betting on future volatility provides a way to speculate on—or hedge against—the direction of the overall options markets. For that reason, some view option investing as little more than bets on future volatility. But whoever the buyers and sellers—and whatever their motives—if some of them have an opinion that the market is going to be significantly more volatile than expected in the coming months, they can buy equity vols; other investors, seeing a calmer environment in the future, will want to sell them.
Based on their long-term studies of volatility throughout the world’s markets, and with Scholes and Merton assisting in the decision making, LTCM’s traders dove headlong into the equity vol market. They traded not just in the United States but in any country where there was enough historical data on market volatility to price this new type of derivative-related product. By one account, LTCM was responsible for around a quarter of the total trading in market volatility, and it reached the point that each 1 percent change in the price of volatility led to a profit or loss of several tens of millions of dollars for the fund. LTCM became known within the hedge fund community as the “Central Bank of Volatility.”10
In the spirit of further fund diversification—and the continued willingness of banks to lend money on the fund’s bets—LTCM expanded its signature investments in convergence trades and equity volatility to interest-rate swaps and merger arbitrage. Toward the end of its run, the fund (much to Scholes’s dismay it should be noted) began making last-ditch “directional”—meaning unhedged—bets on things like the Japanese bond market and even took a short position on Warren Buffett’s Berkshire Hathaway holding company.11
The second and even more disastrous decision made by Meriwether and his partners was to increase the proportion of debt supplied by the banks as a way of further leveraging LTCM’s returns. LTCM partners saw the firm’s seventeen-to-one ratio of debt to equity at the beginning of 1998 as too conservative. So they returned $2.7 billion of capital to the outside investors—most of whom were predictably upset by being kicked out of the highly profitable fund—resulting in a capital base of just $4.7 billion. The LTCM managers did not make any commensurate reduction in the fund’s borrowings, so the new debt-to-equity ratio increased to about twenty-seven to one. Less than 4 percent of LTCM’s total assets were funded by owner’s equity—the rest was coming from what Wall Street’s traders call “other people’s money.”
The increased use of financial leverage was likely fueled by the desire of LTCM’s general partners to reestablish a record of extraordinary returns. The return on the fund’s investment was calculated by dividing its profits by its total equity capital; so if those profits were divided by the new and lower equity base, the returns would suddenly look better. Furthermore, with the number of outside investors reduced, a greater share of those profits would go to LTCM’s inside partners and portfolio managers. Such use of financial leverage may not have been a new ploy for Meriwether and his traders. A former Salomon Brothers partner alleged that the Meriwether group, while it operated at Salomon, “was never really profitable, but created the illusion of profitability by creating the illusion it wasn’t using much capital.” The Salomon partner, further commenting on LTCM’s use of capital, said, “If they had been running the operation with $8 billion of equity, which they needed, they wouldn’t have been earning 41 percent returns, but more like 18 percent or 19 percent, which would have meant that they did about half as well as the stock market.”12
Adopting an extreme leverage position meant that a loss of just 1 percent of the $130 billion of LTCM’s assets would wipe out more than a quarter of the fund’s equity; three additional months of those kinds of losses would throw the fund into bankruptcy. But if this vulnerability came up in their discussions with banks (who continued to lend money despite a dangerously high leverage ratio), LTCM managers could point out that their finely calibrated models still produced a miniscule VaR.
A Quick Demise
Everything fell apart in 1998. A perfect storm created a chain of unforeseeable disruptions in the world’s currency and securities markets, throwing LTCM’s models into confusion. The contagion of Asia’s financial flu spread throughout the world and roiled financial markets, including the U.S. stock market, and LTCM took big losses on equity volatility trades that were based on more quiescent market conditions. But the event that sealed LTCM’s fate was the sudden declaration by Russia, on August 17, that it would default on much of its debt and, furthermore, that it would no longer support the ruble in the foreign currency markets. The firm’s models could not account for such unprecedented market developments.
Panicky investors throughout the world moved their investments to the safe harbors of U.S. dollars and Treasury bonds. With that flight to safety, interest rate spreads between U.S. bonds and foreign bonds diverged dramatically; the sudden rise in demand for riskless U.S. bonds pushed down the yields on U.S. securities, while the lack of demand for the riskier bonds of other countries led to much higher yields. The worldwide movement of funds to U.S. securities may have been seen by some as a gratifying show of confidence in the American economy, but for LTCM, with its portfolio of convergence bets, it meant disaster. On August 21, four days after the Russian default, LTCM lost $551 million—despite a VaR that specified a probable worst-case loss for the fund of only $35 million.13 For the entire month of August, LTCM lost $1.85 billion. During the first three weeks of September, the fund continued to bleed, with another half-billion dollars added to the biggest derivative trading losses in history. It took a Federal Reserve–engineered rescue, in which a reluctant consortium of fourteen major banks provided an emergency infusion of capital, to stem the damage.
In his chronicle of the fall of LTCM, author Roger Lowenstein identified swaps, equity volatility, and a miscellaneous basket of derivative and arbitrage positions as each being responsible for roughly a third of LTCM’s 1998 losses.14 The combined loss for 1998, about $4.5 billion, effectively wiped out the fund’s $4.7 billion capital position at the beginning of that year. The burden was shared between the equity owned by the managing partners and that owned by the outside investors, with Scholes, Meriwether, and the other LTCM insiders absorbing more than 40 percent of the losses.15
The $3.6 billion rescue funds provided by the banks gave them a 90 percent ownership, with LTCM’s key managers allocated a 10 percent share to entice them to stay while the fund was wound down in an orderly fashion. If the banks hadn’t “willingly” stepped in to avert an LTCM bankruptcy—they were, in truth, strong-armed by the Federal Reserve Bank of New York—there would likely have been a major financial crisis. In August the fund had well over $100 billion in assets, but that number greatly understated the magnitude of the potential problems; the notional value of LTCM’s derivative trades—the amount of assets those derivatives controlled—was around $1 trillion. Although many of the positions would have been offsetting, the Fed simply could not stand by and let a single counterparty to a trillion dollars in trades default on its obligations. LTCM, therefore, had the unwelcome distinction of leading the coming parade of SIFIs—systemically important financial institutions—that were too big to fail.
Aftermath
In his 1997 Nobel Prize acceptance speech, a year before the LTCM meltdown, Scholes enumerated prior market losses that could be attributed to derivative trading, including the bankruptcies of Metallgesellschaft and Barings Bank. But he minimized the economic consequences of the losses. He argued, rather, that another group of investors was on the other side of the trades and therefore realizing offsetting gains. By that logic, Scholes maintained that there is no overall “deadweight cost” to society, however painful and consequential those losses might be to the losing party.
Scholes wound up as one of those losing parties. The loss he incurred as an LTCM manager amounted to most of his net worth. To make matters worse, he also wound up owing millions of dollars to the U.S. Internal Revenue Service in connection with an LTCM tax shelter that the government successfully challenged. The haven that Scholes devised rivaled some of LTCM’s derivative trades in complexity—encompassing tax-favored London investors, below-rate interest agreements, complex leasing agreements, and, most damaging in court, a special deal for Scholes under which he received bonus shares in LTCM for designing the shelter.16
Yet Scholes continues to remain philosophical, if not detached, regarding the LTCM experience. In a February 2000 television interview he said only, “I felt quite badly for investors and for others who had worked with us because it was the case that we had a great idea, and a great franchise, and a great application for these ideas for problem solving, and realizing eventually that it was very difficult to effect.”17 And the 1998 LTCM failure did not deter him from founding, the very next year, the limited partnership Platinum Grove Asset Management—a derivative-based hedge fund. He served as chairman of Platinum Grove and brought on other Stanford and MIT PhDs to merge the worlds of business and academia: “Although we are in business, hoping to end up with a profit,” he explained, “we replicate the university setting. We conduct research; we discuss it and improve it; and we build models and empirically test them.”18 Platinum Grove managed close to $10 billion in assets at its height, with what Scholes described as a “liquidity and risk transfer service” for investors and business firms. It was LTCM redux, with Scholes alleging that Platinum Grove enjoyed low volatility and “almost zero correlation with bond markets or stock markets.”19
But when the securities markets plummeted in the 2008 financial crisis, there appeared to be correlation aplenty. In a reprise of the 1998 experience at LTCM, Platinum Grove lost 29 percent of its investors’ money in the first two weeks of October of 2008 and 38 percent through the rest of that month. In the wake of investor demands for the return of their remaining funds, Platinum Grove suspended withdrawals.20 Little information is available today on the ongoing business of Platinum Grove, but Scholes is no longer listed as a partner on the management company’s website and the goals of the fund are now much more straightforward and appear to be based on prosaic investments in stocks, bonds, and currencies.
Warren Buffett, one of the masters of such prosaic investments, is among the many Wall Street veterans who remain fearful of an unchecked use and growth of derivatives. In contrast to Scholes’s view that offsetting gains and losses in derivative trades minimizes any harm to the economy, Buffett continues to sound the siren about their dangers. In his chairman’s letter accompanying the Berkshire Hathaway 2002 report to shareholders, he stated, “Charlie [Munger] and I are of one mind in how we feel about derivatives and the trading activities that go with them: We view them as time bombs, both for the parties that deal in them and the economic system.”21
In the meantime, the derivative markets continue to grow at a feverish pace. In his 1997 Nobel Prize acceptance speech, Scholes noted that the notional value of the derivatives market had grown to about $45 trillion—rivaling the gross national product of the world—but more recent estimates indicate that the notional value of derivatives has increased tenfold since then. For the most part, derivatives are accepted as useful tools for financial institutions—even the insurance companies that are part of Buffett’s Berkshire Hathaway operation use them. When employed properly, derivatives reduce financial risk and foster greater efficiency in the financial markets. Yet when used improperly, excessively, or for purely speculative purposes, they can in fact be time bombs. In any case, there is certainly a commonsense argument for more effective oversight and regulation over a sector of the economy that is volatile enough to create widespread financial emergencies.
