All events
Fund Blowups
LTCM Collapse
Related
For educational purposes only. Not financial advice. Higher returns come with higher risk. Never risk more than you can afford to lose.
Trading StatementREF: TS-2026-03
Trading StatementFor educational purposes only. Not financial advice. Higher returns come with higher risk. Never risk more than you can afford to lose.
Long-Term Capital Management was founded in 1994 by John Meriwether, a legendary bond trader who had previously run the enormously profitable arbitrage group at Salomon Brothers. Meriwether's track record was extraordinary: his group at Salomon had generated billions in profits during the late 1980s and early 1990s, making it the most profitable division of one of Wall Street's most powerful firms. When he left Salomon in 1991 amid a Treasury bond bidding scandal (in which he was not directly implicated but bore supervisory responsibility), the financial world expected he would resurface with something ambitious. LTCM exceeded those expectations, assembling perhaps the most intellectually formidable team ever gathered at a single hedge fund.
The fund's advisory board included Myron Scholes and Robert Merton, who would win the Nobel Prize in Economics in 1997 for their work on options pricing theory. Their Black- Scholes model had revolutionized derivatives markets and underpinned trillions of dollars in financial transactions worldwide. The team also included David Mullins, a former Vice Chairman of the Federal Reserve, and Eric Rosenfeld, a former Harvard Business School professor and Salomon arbitrageur. The combination of academic brilliance and practical trading experience was unprecedented. Investors clamored to get in, and LTCM was able to charge a 2% management fee and 25% performance fee, higher than the industry standard, while also requiring a three-year lock-up of capital with a minimum investment of $10 million.
LTCM raised $1.25 billion at its launch, one of the largest initial capital raises for a hedge fund at that time. Its investor base read like a directory of the global financial elite: major banks, central banks, university endowments, and wealthy individuals. Merrill Lynch invested $15 million from its own balance sheet. The Bank of Italy invested $100 million. Even executives at the major banks that served as LTCM's counterparties invested personal funds, blurring the line between the firms that were supposed to be monitoring their exposure to the fund and the individuals who stood to profit from its success. This intermingling of interests would become critically important when the crisis hit.
LTCM's core strategy was convergence trading, which exploited small pricing discrepancies between related securities that mathematical models predicted would eventually converge. The simplest version involved buying cheap bonds and selling expensive ones that were nearly identical in every way except liquidity. For example, the fund might buy recently issued 29-and-a-half-year Treasury bonds (which were slightly cheaper because they were off-the-run) while simultaneously shorting the 30-year on-the-run benchmark Treasury (which commanded a slight premium because of its greater liquidity). The spread between these two bonds was tiny, often just a few basis points, but LTCM's models showed with near mathematical certainty that the gap would close as the bonds aged and became essentially interchangeable.
Because the individual profit on each trade was minuscule, LTCM needed enormous leverage to generate meaningful returns. The fund employed leverage ratios of approximately 25 to 1, meaning that for every dollar of investor capital, it controlled roughly $25 in assets. By early 1998, LTCM had approximately $5 billion in investor equity supporting $125 billion in on-balance-sheet assets. But even this staggering figure understated the true risk, because the fund also held off-balance-sheet derivatives positions with a notional value of approximately $1.25 trillion. These positions included interest rate swaps, equity derivatives, and credit instruments spread across virtually every major market in the world. The fund's models indicated that the risk of significant loss was vanishingly small, with the probability of losing more than 20% of capital in a single year estimated at less than one in several billion.
The early results seemed to validate the strategy brilliantly. LTCM returned 21% in its first partial year of 1994, 43% in 1995, 41% in 1996, and 17% in 1997 (after returning $2.7 billion in capital to investors because the fund was too large to find enough profitable trades). By 1997, Meriwether and his partners were so confident in their models that they returned capital to outside investors, increasing the partners' own share of the fund and the effective leverage ratio. The fund's partners had approximately $1.9 billion of their own money invested, representing the vast majority of their personal wealth. This concentration of personal capital would transform the fund's eventual losses from a professional setback into a personal catastrophe.
The first tremors arrived in May 1998, when the fund lost 6.4% as spreads widened in response to growing concerns about emerging market economies. The Asian financial crisis of 1997 had already rattled global markets, and investors were becoming increasingly nervous about extending credit to developing countries. LTCM's models registered the losses but indicated they were within normal parameters and that spreads would reconverge. In June, the fund lost another 10%. Meriwether and his partners were concerned but not alarmed; their models told them that the probability of spreads widening further was extremely low. They maintained their positions and even added to some, viewing the wider spreads as an opportunity to increase expected returns.
On August 17, 1998, Russia defaulted on its domestic government bonds and devalued the ruble. This event, while significant, should not in theory have been catastrophic for LTCM, whose direct Russian exposure was relatively limited. But the Russian default triggered a global flight to quality that was unprecedented in its speed and severity. Investors around the world simultaneously rushed to sell risky assets and buy the safest government bonds they could find, particularly US Treasuries and German Bunds. This caused every spread trade that LTCM held to move against them simultaneously. The correlations between their supposedly diversified positions surged toward one, meaning that positions designed to hedge each other instead compounded losses. In a single day on August 21, LTCM lost $553 million, approximately 15% of its remaining capital.
The fund's models had assigned a probability of essentially zero to the losses LTCM was experiencing. According to their value-at-risk calculations, the losses of August 1998 were a 10-sigma event, something that should not occur once in the entire history of the universe. The problem was that the models assumed market participants would behave rationally and independently, but in a panic, everyone acts the same way. Making matters worse, other banks and hedge funds had copied LTCM's strategies, so when LTCM tried to reduce positions, their selling drove prices further against them, which forced more selling in a vicious feedback loop. The fund that had been built on the assumption that markets would behave according to mathematical models was being destroyed by the very human reality of panic and herd behavior.
By mid-September, LTCM's capital had shrunk to approximately $600 million, but its on-balance-sheet assets still totaled around $100 billion, meaning leverage had ballooned to nearly 167 to 1. The fund was hemorrhaging money daily and was running out of margin to post to its counterparties. Meriwether made a desperate attempt to raise new capital, approaching Warren Buffett, George Soros, and others. Buffett, along with Goldman Sachs and AIG, offered to buy the fund's entire portfolio for $250 million and inject $3.75 billion in new capital, but the offer came with conditions that would have wiped out the existing partners entirely, and Meriwether rejected it. Time was running out, and every day of delay increased the risk that counterparties would begin seizing collateral, triggering a disorderly liquidation that could destabilize the entire financial system.
John Meriwether was the driving force behind LTCM and bore ultimate responsibility for its risk management failures. A former bond trader who had built Salomon Brothers' legendary arbitrage desk, Meriwether combined genuine mathematical sophistication with a trader's instinct for identifying mispricings. His fatal flaw was an unwavering faith in the models his team had built. When the models said the fund's positions would converge, he believed them even as real-world evidence mounted that something had fundamentally changed. After LTCM's collapse, Meriwether went on to found two more hedge funds, JWM Associates in 1999 and JM Advisors Management in 2010, both of which also suffered significant losses during market crises, suggesting that the lessons of 1998 were never fully internalized.
Myron Scholes and Robert Merton lent LTCM an aura of intellectual invincibility that attracted investors and encouraged the fund's counterparties to extend extraordinary amounts of leverage. The irony of two Nobel laureates who had developed the theoretical framework for pricing risk being involved in one of history's greatest risk management failures was not lost on the financial world. Their models, rooted in the assumption of normally distributed returns and continuous, liquid markets, simply could not account for the fat-tailed, correlated, liquidity-driven moves that occurred during the crisis. Scholes later said that the fundamental lesson was that risk is not the same as uncertainty: you can model risk, but you cannot model the unknown unknowns that emerge during a true crisis.
William McDonough, President of the Federal Reserve Bank of New York, orchestrated the rescue that prevented LTCM's failure from becoming a systemic catastrophe. On September 23, 1998, McDonough convened the heads of 14 major banks and investment firms at the New York Fed's offices and informed them that LTCM's disorderly failure could trigger a cascade of defaults that would destabilize global markets. Over two days of tense negotiations, a consortium of 14 firms including Goldman Sachs, Merrill Lynch, JPMorgan, Morgan Stanley, and others agreed to inject $3.625 billion into LTCM in exchange for 90% of the fund's equity. The Fed did not provide any public money, but its role in organizing and pressuring the bailout raised important questions about moral hazard and the implicit government backstop that the largest financial institutions enjoyed.
The immediate market impact of the LTCM crisis extended far beyond the fund itself. Credit spreads widened to levels not seen since the 1987 crash, as banks and investors rushed to reduce risk. The spread between corporate bonds and Treasuries blew out, making it dramatically more expensive for companies to borrow. Emerging market bonds collapsed as investors fled anything perceived as risky. The S&P 500 fell approximately 19% from its July 1998 peak to its October low, though it would recover by year-end. More importantly, the crisis revealed the extent to which major banks were interconnected through their derivatives exposures, a lesson that would be forgotten by the time of the 2008 crisis but was deeply alarming in the moment.
The Federal Reserve cut interest rates three times in quick succession during the fall of 1998 to prevent the crisis from spreading to the broader economy. These rate cuts, while effective at calming markets, were criticized by some as reinforcing moral hazard by bailing out reckless speculators. The cuts also provided additional fuel for the dot-com bubble that was already inflating in technology stocks. Some economists have argued that the Fed's response to LTCM established a pattern, later termed the "Greenspan put," in which market participants came to expect that the central bank would intervene to support asset prices during any significant downturn, encouraging ever-greater risk-taking.
For the banks in the rescue consortium, the bailout ultimately proved profitable. The positions that had been moving against LTCM did eventually converge, just as the fund's models had predicted. The consortium members gradually unwound the portfolio over the next year and recovered their investment plus a modest return. This outcome reinforced Meriwether's belief that he had been right all along and merely needed more time and capital, a conviction that led him to start a new fund employing similar strategies. For the partners of LTCM, however, the losses were devastating: their $1.9 billion in personal capital was almost entirely wiped out, and several partners were forced to sell homes and other assets to cover personal debts.
The LTCM debacle prompted a wave of soul-searching about the regulation of hedge funds and the risks of excessive leverage. The President's Working Group on Financial Markets issued a report in 1999 recommending greater disclosure requirements for hedge funds and improved risk management by their bank counterparties. However, the political momentum for meaningful reform dissipated as the economy boomed and the crisis faded from memory. The Commodity Futures Modernization Act of 2000 actually moved in the opposite direction, explicitly exempting over-the-counter derivatives from regulation and preventing the CFTC from imposing oversight. This legislative choice would prove catastrophic a decade later when the unregulated credit default swap market nearly destroyed the global financial system in 2008.
The crisis fundamentally challenged the dominant risk management paradigm that had prevailed in quantitative finance. LTCM's models relied on assumptions borrowed from physics: that market returns were normally distributed, that correlations were stable, and that liquidity would always be available. The 1998 crisis showed that financial markets are fundamentally different from physical systems because they are populated by human beings who panic, herd, and act irrationally. Nassim Nicholas Taleb, who had been critical of LTCM's approach, used the episode as a central example in his later works on black swan events and the limitations of statistical models in finance. The concept of fat-tailed distributions and regime changes in correlation structures entered mainstream financial thinking, though the degree to which these insights were actually incorporated into practice remains debatable.
The LTCM collapse also exposed how concentrated risk had become in the global financial system. The fund's $1.25 trillion in derivatives positions meant that its failure would have forced its 14 major bank counterparties to recognize enormous losses and attempt to replace hedges simultaneously, almost certainly triggering a chain of defaults. This concentration of systemic risk in a single, lightly regulated entity shocked regulators and policymakers. Yet the fundamental dynamic, where a small number of highly leveraged institutions are interconnected through complex derivatives in ways that create systemic fragility, was not meaningfully addressed before it recurred on a vastly larger scale in 2008. The lesson of LTCM was learned, discussed, written about extensively, and then largely ignored by the very institutions that should have known better.
The most fundamental lesson of LTCM is that leverage transforms manageable losses into existential ones. LTCM's trades were, for the most part, correct in their directional thesis. The spreads did eventually converge, just as the models predicted. But that correctness was irrelevant because the fund did not survive long enough to see it happen. At 25-to-1 leverage, a 4% adverse move wipes out all equity. For individual traders, this lesson translates directly: no matter how confident you are in a trade, the amount of leverage you employ determines whether a temporary drawdown is a nuisance or a catastrophe. Survival must always take priority over returns.
The LTCM story illustrates that diversification can be an illusion during crises. The fund held thousands of positions across dozens of markets and asset classes, and its risk models showed that these positions had low correlations under normal conditions. But during the Russian debt crisis, correlations surged as every market participant simultaneously tried to reduce risk. Positions that had never moved together before suddenly all moved in the same direction. For traders, this means that diversification based on historical correlations can provide a false sense of security. True stress testing requires asking what happens when all your positions go wrong at the same time, because in a genuine crisis, they often will.
LTCM's collapse also demonstrates the critical importance of liquidity as a risk factor. The fund's models valued its positions based on mid-market prices, but during the crisis, bid-ask spreads widened dramatically and in some markets liquidity simply vanished. A position that showed a small profit at mid-market could produce a massive loss if you actually tried to sell it. For traders of any size, this means that the ability to exit a position is as important as the position's theoretical value. Illiquid positions should be sized conservatively, and the cost of exiting under stress should be incorporated into every risk assessment. LTCM's partners learned this lesson at the cost of their personal fortunes.
Finally, LTCM teaches that intellectual brilliance and past success are not protection against catastrophic loss. The fund was run by some of the smartest people in finance, including two Nobel laureates, and it had produced spectacular returns for four consecutive years. Yet all of that intellect and experience could not prevent the fund from blowing up when market conditions moved outside the parameters their models had been built to handle. Humility before the markets is not merely a platitude but a survival strategy. The traders who endure over long careers are those who recognize that they cannot predict every possible outcome and who size their positions and manage their risk accordingly. Markets have a way of finding and punishing overconfidence, no matter how well credentialed its source.