We are looking forward to discussing LTCM this Thursday. To facilitate the discussion, we have provided a short narrative below based on the infamous book written about LTCM: When Genius Failed. At the end of the narrative are links to short, recent articles about LTCM in the context of the current credit crisis. Enjoy!
Asha, Drew, Karen, and Molly
Asha, Drew, Karen, and Molly
Long-Term Capital Management (LTCM) was founded by John Meriwether in 1994 with a group of partners that included two noble prize winners (Myron Scholes of the Black-Scholes model for option pricing, and Robert Merton who shared the Nobel with Scholes for their work on option pricing), a Harvard Business School professor, and other accomplished academic minds from the top finance and economics schools around the globe. Institutional Investor described the founders as “the best finance faculty in the world.” As a hedge fund, LTCM would build upon the experience of the partners, many of whom prior to starting the Fund, had run the arbitrage desk at Salomon Brothers bank, where they had earned over $500 million a year in profits. The founders planned to replicate the success they had achieved at Salomon, where they had used complicated mathematical models to understand and predict market movements. The models allowed them to identify inefficiencies in the bond markets, which they could exploit to earn significant profits off of mispriced fixed-income assets.
Due to the academic prestige of the founders, and their background in arbitrage trading at Salomon Brothers, LTCM quickly developed an aura of mystique and invincibility that was unrivalled on Wall Street. Although some investors were put off by the secrecy of the partners, who were unwilling to reveal their investment strategies, many potential investors clamored to join the Fund, and it opened for business in February of 1994 with $1.25 billion in capital.
The investors, with their high expectations for LTCM, were not disappointed. LTCM earned close to 40 percent returns during its first few years in operation, and grew to an enormous size. The fund amassed assets in excess of $128 billion, and up to $7 billion in capital. Of course, such a meteoric rise is often followed by a painful crash, and LTCM was no exception. In 1998, after four years of astounding success, LTCM’s fortunes changed. In just five weeks, the fund would lose virtually all of its outstanding capital, and become dangerously leveraged—at the height of the calamity, Long Term was leveraged 100 to 1.
To understand the scope of LTCM’s collapse, and its potential impact on the broader markets, it is important to grasp the implications of the fund’s investment strategies. LTCM focused on earning money off of inefficient bond prices and pairs trading. The bond strategy, which involved short selling over-priced bonds, only allowed the fund to earn a very small margin on each bond—as little as 1 percent, or $10 off of every $1,000 bond. Thus, in order to achieve real profits, LTCM had to engage in extremely high volume trades. However, to facilitate these large volume trades, LTCM would have to borrow heavily, resulting in highly leveraged positions. LTCM used computer programs and mathematical models to identify asset pairs whose price spread diverged from historical norms, and then they would bet on the spread returning to the model baseline. Myron Scholes referred to this strategy as “using a vacuum cleaner to pick up nickels that no one else could see,” but later could be characterized as “picking up nickels in front of a bulldozer.”
Given the academic background of the partners, one can be sure that these strategies were not undertaken capriciously. Rather, the partners developed complex models that were specifically geared towards cabining the risks that the Firm would take. Specifically, “Long Term was an experiment in managing risk by the numbers.” Supplanting mathematical concepts from the field of physics, the models built upon the law of large numbers—that random events will distribute themselves into the familiar bell curve. The partners believed that price changes in financial markets were inherently random events. Thus, while no one could predict any particular change, the overall distribution of prices in the market would follow a bell curve distribution.
Given this understanding, and sufficient data regarding past market activity, the partners believed that they could calculate with precision the exact risk to which their portfolio was exposed. For example, in a letter sent to investors in late 1994, the partners announced that the fund would lose 5 percent of its money 12 percent of the time. In addition, their models told them that the maximum amount they were likely to lose in any one day of trading was $45 million, and that the probability of the firm losing all of its capital in one year was only one in ten to the twenty-fourth power!
However, these models relied on certain critical assumptions. First, they assumed that each market transaction was random—that all events occur independently of one another. Second, the models presumed that past data was sufficient for predicting the future activity of the markets. Finally, although LTCM relied heavily on past data, their models were based on an insufficient history of data. They did not extend back far enough to capture the extreme volatility that had racked the market during historic events like the great depression. Thus, their sample size provided an inadequate picture of the potential range of market movements.
Unfortunately, market events often are not completely independent of one another. Market events are driven by “the human spirit,” and “markets have memories.” Thus, unlike random coin tosses, if the market experiences three bad ‘tosses’ in a row, the result of the fourth toss may not be entirely random. Some investors will be influenced by the results of the earlier tosses, and change their behavior accordingly. Furthermore, a sampling of past events may not be a good predictor in a dynamic market environment, where the patterns in the market are constantly changing and adapting to new conditions. A static model composed of past data cannot adequately account for the new possibilities, new problems, and new products constantly emerging in an evolving financial system.
These flawed assumptions ultimately led to the firm’s downfall. The firm, with capital in excess of $4.7 billion, began looking into emerging markets and riskier trades to provide investment outlets for all of its money. The partners widened the scope of their activities, betting in the equities markets, engaging in more directional bets, and buying bonds in emerging markets, most notably Russia.
In mid-1998, banks and investors became increasingly wary of perceived risk in the wider market, particularly given the wake of the financial crisis in Asia. Thus, the markets faced a “flight to quality,” as investors rushed to buy Treasury bonds and shed riskier and less liquid investments. Therefore, the spread between treasury bonds and other bonds (such as corporate bonds and mortgage-backed securities) climbed to unprecedented levels.
Such a climb was unanticipated by LTCM’s models, which were predicated on the assumption that as markets became more efficient, spreads between various bonds would close, rather than widen. Based on this assumption, LTCM had invested heavily in riskier, less liquid bonds and gone short on Treasuries. Unfortunately, investors now began demanding a high premium for the less liquid instruments, due to the growing aversion to risk in the market.
In particular, LTCM had invested heavily in the Russian bond market, where concerns of currency devaluation had pushed the yield on short term bonds to a 120 percent annual rate. The partners were convinced that this perceived risk in the market was unjustified, as the thought that Russia simply would not allow its currency to fail. However, on August 17, 1998, the fears of all the investors in the market were borne out—Russia, contrary to its many public promises to the contrary, devalued its currency and defaulted on its government debt. This announcement sparked a frenzy in the marketplace, as investors simultaneously began pulling out of all emerging markets, and any risky investments around the globe. Stock markets in Venezuela, Eastern Europe, and Asia plunged. Most deleteriously for LTCM, investors in every market swiftly sold riskier bonds and clamored to purchase the safest bonds, pushing spreads wider and wider.
Simultaneously, LTCM began losing money in every one of its investments. Contrary to its fine-tuned models, which told the firm that it could never lose more than $45 million in one day, LTCM lost $553 million, or 15 percent of its capital, the Friday after Russia defaulted.
Although LTCM remained confident that if they could ride out the unrest, their models would eventually be borne out and spreads between bonds would eventually converge, the firm continued to hemorrhage capital. Spreads continued to soar, well beyond reasonable levels. The market was gripped by fear, leading to panicked sales that LTCM’s rational models were incapable of predicting. LTCM’s losses forced the highly leveraged firm to sell some of its positions at very unfavorable prices, further exacerbating their difficulties.
Within five weeks, the firm had lost virtually all of its capital—it was down over 80 percent from starting capital of $4 billion to only $555 million. However, the fund’s impending collapse did not just spell disaster for the fund’s partners or private investors. To obtain financing for its highly leveraged positions, the fund had borrowed extensively from banks around the world, including Merrill Lynch, Bears Sterns, and The Bank of Switzerland. Furthermore, the bank had a portfolio of derivative contracts with many of these banks worth close to $1 trillion. The firm’s extremely large portfolio had induced the banks to loan the firm large amounts of money at extremely low rates, and the banks now stood to lose a sizeable amount if LTCM collapsed.
In fact, the firm’s outstanding obligations to the investment banks were so large that the Federal Reserve became nervous that LTCM’s collapse would create a systemic shock in the markets. LTCM’s reach was so great, and the exposure of the banks so large, that the Fed feared a downward spiral that would destroy the fragile economy persisting in the wake of the financial crises that shook the market in the late 1990s. Upon reviewing the Firm’s positions, the Fed realized that the exposure of many of Wall Street’s top investment banks meant that if Long Term collapsed, the markets would seize up. “Losses in so many markets and to so many players would spark a vicious circle of liquidation, extreme fluctuations in interest rates, and then still further losses: Markets would . . . possibly cease to function for a period of one or more days and maybe longer.”
Thus, the Fed scrambled to organize additional investment for LTCM. The Fed hoped that by pouring additional capital into the firm, LTCM’s liquidity could be increased, enabling it to weather the storm created by the unpredicted shifts in the bond market. Throughout the firm’s demise, the partners maintained that they were not suffering a collapse, rather they were simply facing liquidity problems. Regardless of the characterization of the trouble, it was clear that without an immediate capital infusion, the firm would be out of money by the end of the week, and experts suggested the effect on the wider market could be catastrophic.
After days of intense negotiations facilitated by the Federal Reserve, a consortium of fourteen banks agreed to invest $3.65 billion into LTCM, stabilizing the firm and enabling it to endure the unrest in the markets. The partners were wiped out, and had to secede control to a supervisory board. Nonetheless, the influx of capital did sustain the firm through the continuing turbulence in the markets, and LTCM was able to sell off its positions at profitable margins. By 2000, the fund had paid back the conglomerate of investors, and even garnered them a small profit. Thus, the organized rescue forestalled a calamity in the marketplace, and rewarded the investors for the risk they took in pouring money into the hemorrhaging firm.
However, regardless of the effectiveness of the bailout in addressing the immediate concerns of the Federal Reserve and LTCM, many remained nervous about the long-range implications of the Fed’s actions in facilitating the rescue. Although no public funds were invested in the firm, the government’s presence was a powerful force behind the bailout. Thus, people feared that the government-orchestrated rescue would encourage others to make similarly risky investments, under the implicit assumption that the Fed would bail them out if they ever got in over their heads.
As for Long Term’s partners, they lost everything in the bailout. Most of the partners had invested their salaries, their bonuses, and even their wives’ and kids’ money back into the Fund. Combined, they lost a total of $1.9 billion dollars, and some even faced personal bankruptcy. Nonetheless, in December of 1999, many of the partners teamed up again, and raised $250 million to open their next hedge fund. It, too, faces impending failure. Meriwether is currently considering creating a new hedge fund—third time’s the charm.
Suggested Readings:
Professor Fischel and Professor Picker commented on LTCM in a WSJ op-ed piece shortly after the collapse and facilitated rescue of the fund, calling it a “firm that failed well.” Recently, LTCM has returned to the editorial pages in the context of the current credit crisis. The NYT criticized the short-term memory with regards to Long-Term Capital, arguing that once again, Wall Street has made the mistake of placing “complacent trust in financial models” and went even further, calling for regulation of obscure derivatives. The Economist has called LTCM the test case for the credit bubble, but rather than prick the bubble the experience with LTCM redoubled the market's faith in its own model. The editorialists at WSJ’s DealJournal note that many people (and especially Meriwether and the other original partners of LTCM) do not believe that the rescue of LTCM deserves the label of “bailout” because of its small size in comparison to bailouts like TARP and the lack of government funds injected into the rescue. Whatever its label, there are those who believe the Fed’s involvement in preventing the collapse of LTCM in the late 90s contributed to the moral hazard of enabling (and implicitly insuring) riskier investments by banks such as Bear Stearns.
Due to the academic prestige of the founders, and their background in arbitrage trading at Salomon Brothers, LTCM quickly developed an aura of mystique and invincibility that was unrivalled on Wall Street. Although some investors were put off by the secrecy of the partners, who were unwilling to reveal their investment strategies, many potential investors clamored to join the Fund, and it opened for business in February of 1994 with $1.25 billion in capital.
The investors, with their high expectations for LTCM, were not disappointed. LTCM earned close to 40 percent returns during its first few years in operation, and grew to an enormous size. The fund amassed assets in excess of $128 billion, and up to $7 billion in capital. Of course, such a meteoric rise is often followed by a painful crash, and LTCM was no exception. In 1998, after four years of astounding success, LTCM’s fortunes changed. In just five weeks, the fund would lose virtually all of its outstanding capital, and become dangerously leveraged—at the height of the calamity, Long Term was leveraged 100 to 1.
To understand the scope of LTCM’s collapse, and its potential impact on the broader markets, it is important to grasp the implications of the fund’s investment strategies. LTCM focused on earning money off of inefficient bond prices and pairs trading. The bond strategy, which involved short selling over-priced bonds, only allowed the fund to earn a very small margin on each bond—as little as 1 percent, or $10 off of every $1,000 bond. Thus, in order to achieve real profits, LTCM had to engage in extremely high volume trades. However, to facilitate these large volume trades, LTCM would have to borrow heavily, resulting in highly leveraged positions. LTCM used computer programs and mathematical models to identify asset pairs whose price spread diverged from historical norms, and then they would bet on the spread returning to the model baseline. Myron Scholes referred to this strategy as “using a vacuum cleaner to pick up nickels that no one else could see,” but later could be characterized as “picking up nickels in front of a bulldozer.”
Given the academic background of the partners, one can be sure that these strategies were not undertaken capriciously. Rather, the partners developed complex models that were specifically geared towards cabining the risks that the Firm would take. Specifically, “Long Term was an experiment in managing risk by the numbers.” Supplanting mathematical concepts from the field of physics, the models built upon the law of large numbers—that random events will distribute themselves into the familiar bell curve. The partners believed that price changes in financial markets were inherently random events. Thus, while no one could predict any particular change, the overall distribution of prices in the market would follow a bell curve distribution.
Given this understanding, and sufficient data regarding past market activity, the partners believed that they could calculate with precision the exact risk to which their portfolio was exposed. For example, in a letter sent to investors in late 1994, the partners announced that the fund would lose 5 percent of its money 12 percent of the time. In addition, their models told them that the maximum amount they were likely to lose in any one day of trading was $45 million, and that the probability of the firm losing all of its capital in one year was only one in ten to the twenty-fourth power!
However, these models relied on certain critical assumptions. First, they assumed that each market transaction was random—that all events occur independently of one another. Second, the models presumed that past data was sufficient for predicting the future activity of the markets. Finally, although LTCM relied heavily on past data, their models were based on an insufficient history of data. They did not extend back far enough to capture the extreme volatility that had racked the market during historic events like the great depression. Thus, their sample size provided an inadequate picture of the potential range of market movements.
Unfortunately, market events often are not completely independent of one another. Market events are driven by “the human spirit,” and “markets have memories.” Thus, unlike random coin tosses, if the market experiences three bad ‘tosses’ in a row, the result of the fourth toss may not be entirely random. Some investors will be influenced by the results of the earlier tosses, and change their behavior accordingly. Furthermore, a sampling of past events may not be a good predictor in a dynamic market environment, where the patterns in the market are constantly changing and adapting to new conditions. A static model composed of past data cannot adequately account for the new possibilities, new problems, and new products constantly emerging in an evolving financial system.
These flawed assumptions ultimately led to the firm’s downfall. The firm, with capital in excess of $4.7 billion, began looking into emerging markets and riskier trades to provide investment outlets for all of its money. The partners widened the scope of their activities, betting in the equities markets, engaging in more directional bets, and buying bonds in emerging markets, most notably Russia.
In mid-1998, banks and investors became increasingly wary of perceived risk in the wider market, particularly given the wake of the financial crisis in Asia. Thus, the markets faced a “flight to quality,” as investors rushed to buy Treasury bonds and shed riskier and less liquid investments. Therefore, the spread between treasury bonds and other bonds (such as corporate bonds and mortgage-backed securities) climbed to unprecedented levels.
Such a climb was unanticipated by LTCM’s models, which were predicated on the assumption that as markets became more efficient, spreads between various bonds would close, rather than widen. Based on this assumption, LTCM had invested heavily in riskier, less liquid bonds and gone short on Treasuries. Unfortunately, investors now began demanding a high premium for the less liquid instruments, due to the growing aversion to risk in the market.
In particular, LTCM had invested heavily in the Russian bond market, where concerns of currency devaluation had pushed the yield on short term bonds to a 120 percent annual rate. The partners were convinced that this perceived risk in the market was unjustified, as the thought that Russia simply would not allow its currency to fail. However, on August 17, 1998, the fears of all the investors in the market were borne out—Russia, contrary to its many public promises to the contrary, devalued its currency and defaulted on its government debt. This announcement sparked a frenzy in the marketplace, as investors simultaneously began pulling out of all emerging markets, and any risky investments around the globe. Stock markets in Venezuela, Eastern Europe, and Asia plunged. Most deleteriously for LTCM, investors in every market swiftly sold riskier bonds and clamored to purchase the safest bonds, pushing spreads wider and wider.
Simultaneously, LTCM began losing money in every one of its investments. Contrary to its fine-tuned models, which told the firm that it could never lose more than $45 million in one day, LTCM lost $553 million, or 15 percent of its capital, the Friday after Russia defaulted.
Although LTCM remained confident that if they could ride out the unrest, their models would eventually be borne out and spreads between bonds would eventually converge, the firm continued to hemorrhage capital. Spreads continued to soar, well beyond reasonable levels. The market was gripped by fear, leading to panicked sales that LTCM’s rational models were incapable of predicting. LTCM’s losses forced the highly leveraged firm to sell some of its positions at very unfavorable prices, further exacerbating their difficulties.
Within five weeks, the firm had lost virtually all of its capital—it was down over 80 percent from starting capital of $4 billion to only $555 million. However, the fund’s impending collapse did not just spell disaster for the fund’s partners or private investors. To obtain financing for its highly leveraged positions, the fund had borrowed extensively from banks around the world, including Merrill Lynch, Bears Sterns, and The Bank of Switzerland. Furthermore, the bank had a portfolio of derivative contracts with many of these banks worth close to $1 trillion. The firm’s extremely large portfolio had induced the banks to loan the firm large amounts of money at extremely low rates, and the banks now stood to lose a sizeable amount if LTCM collapsed.
In fact, the firm’s outstanding obligations to the investment banks were so large that the Federal Reserve became nervous that LTCM’s collapse would create a systemic shock in the markets. LTCM’s reach was so great, and the exposure of the banks so large, that the Fed feared a downward spiral that would destroy the fragile economy persisting in the wake of the financial crises that shook the market in the late 1990s. Upon reviewing the Firm’s positions, the Fed realized that the exposure of many of Wall Street’s top investment banks meant that if Long Term collapsed, the markets would seize up. “Losses in so many markets and to so many players would spark a vicious circle of liquidation, extreme fluctuations in interest rates, and then still further losses: Markets would . . . possibly cease to function for a period of one or more days and maybe longer.”
Thus, the Fed scrambled to organize additional investment for LTCM. The Fed hoped that by pouring additional capital into the firm, LTCM’s liquidity could be increased, enabling it to weather the storm created by the unpredicted shifts in the bond market. Throughout the firm’s demise, the partners maintained that they were not suffering a collapse, rather they were simply facing liquidity problems. Regardless of the characterization of the trouble, it was clear that without an immediate capital infusion, the firm would be out of money by the end of the week, and experts suggested the effect on the wider market could be catastrophic.
After days of intense negotiations facilitated by the Federal Reserve, a consortium of fourteen banks agreed to invest $3.65 billion into LTCM, stabilizing the firm and enabling it to endure the unrest in the markets. The partners were wiped out, and had to secede control to a supervisory board. Nonetheless, the influx of capital did sustain the firm through the continuing turbulence in the markets, and LTCM was able to sell off its positions at profitable margins. By 2000, the fund had paid back the conglomerate of investors, and even garnered them a small profit. Thus, the organized rescue forestalled a calamity in the marketplace, and rewarded the investors for the risk they took in pouring money into the hemorrhaging firm.
However, regardless of the effectiveness of the bailout in addressing the immediate concerns of the Federal Reserve and LTCM, many remained nervous about the long-range implications of the Fed’s actions in facilitating the rescue. Although no public funds were invested in the firm, the government’s presence was a powerful force behind the bailout. Thus, people feared that the government-orchestrated rescue would encourage others to make similarly risky investments, under the implicit assumption that the Fed would bail them out if they ever got in over their heads.
As for Long Term’s partners, they lost everything in the bailout. Most of the partners had invested their salaries, their bonuses, and even their wives’ and kids’ money back into the Fund. Combined, they lost a total of $1.9 billion dollars, and some even faced personal bankruptcy. Nonetheless, in December of 1999, many of the partners teamed up again, and raised $250 million to open their next hedge fund. It, too, faces impending failure. Meriwether is currently considering creating a new hedge fund—third time’s the charm.
Suggested Readings:
Professor Fischel and Professor Picker commented on LTCM in a WSJ op-ed piece shortly after the collapse and facilitated rescue of the fund, calling it a “firm that failed well.” Recently, LTCM has returned to the editorial pages in the context of the current credit crisis. The NYT criticized the short-term memory with regards to Long-Term Capital, arguing that once again, Wall Street has made the mistake of placing “complacent trust in financial models” and went even further, calling for regulation of obscure derivatives. The Economist has called LTCM the test case for the credit bubble, but rather than prick the bubble the experience with LTCM redoubled the market's faith in its own model. The editorialists at WSJ’s DealJournal note that many people (and especially Meriwether and the other original partners of LTCM) do not believe that the rescue of LTCM deserves the label of “bailout” because of its small size in comparison to bailouts like TARP and the lack of government funds injected into the rescue. Whatever its label, there are those who believe the Fed’s involvement in preventing the collapse of LTCM in the late 90s contributed to the moral hazard of enabling (and implicitly insuring) riskier investments by banks such as Bear Stearns.
Absolutely riveting!
Posted by: Drew Navikas | 02/24/2009 at 09:39 PM
Someone did a nice tv special on LTCM; I think it was PBS. I don't know what year. Now that I think about it, NOVA may have put it out. Might be interesting if you could find some clips.
-cd
Posted by: Cleve | 02/25/2009 at 06:18 AM
I actually read the "infamous book" written about LTCM and it was very interesting. Sometimes you can learn more from failure then from success.
Posted by: digital options | 02/23/2012 at 02:42 PM