Next month marks the 25th anniversary of the collapse of Long-Term Capital Management. In the four-and-a-half short years it was alive, the firm established a name for itself on Wall Street while remaining largely anonymous beyond. In death, it became well-known the world over.¹

The rise and fall of LTCM has all the hallmarks of a great story: brilliant minds (including two Nobel laureates); dazzling success; enormous wealth; a sudden, perilous downfall; risk of devastating collateral damage; and a thrilling finale. The story burrowed its way into public consciousness as the finance industry’s Titanic. One book about it, When Genius Failed, became an international bestseller and is still required reading for anyone who works or invests in markets. 

Even today, LTCM’s shadow endures. In a paper released last month discussing regulatory approaches to investment funds, the Central Bank of Ireland cites LTCM as a cautionary tale. Having sustained significant losses, “the failure of LTCM had the potential to generate broader contagion effects to the financial sector,” it reminds us. “Subsequently, the Federal Reserve Bank of New York brokered an injection of $3.6 billion from some of the largest creditors of LTCM… Without this, LTCM would not have been able to meet its payment obligations by the end of September 1998.”

The official anniversary of LTCM’s downfall is September 23. That was the day William McDonough, President of the Federal Reserve Bank of New York, gathered top officials from sixteen of the world’s most powerful banks and investment houses into a room and told them to find a solution to the mess LTCM was in. LTCM was big – it had over 60,000 trades on its books comprising over $125 billion of assets and $1.4 trillion notional value of derivatives. McDonough was worried that a disorderly unwind of the firm’s positions would snowball through markets leading to catastrophic losses that would “pose unacceptable risks to the American economy.” Fourteen of the sixteen firms agreed to inject capital into LTCM to keep it afloat.²

But it was back in August that the downfall began. On August 21, 1998, Eric Rosenfeld, one of the partners of LTCM, was about to tee-off on a golf course in Sun Valley, Idaho, where he was vacationing with his family. Russia had defaulted on its debt earlier that week but the firm didn’t have much Russian exposure and so he wasn’t too concerned. At 9am, he called the office. “It was bedlam,” he recalls.  

The firm’s core business was arbitrage, principally bond market arbitrage. Rosenfeld and his colleagues would buy undervalued bonds and sell short overvalued bonds affected by similar factors in order to capture price differentials without taking directional market risk. A textbook trade involved “on-the-run” and “off-the-run” bonds. LTCM would buy “off-the-run” bonds with 29 years left until maturity and sell “on-the-run” bonds – the most recently issued US Treasury bonds with a 30-year maturity. Because the freshly issued bonds tended to be more liquid, they would trade at a premium, reflected in a lower yield. LTCM could buy the 29-year bond at a yield of 5.62% and sell the 30-year bond at a yield of 5.50% – its objective being to pick up the 0.12% yield differential over time. Convergence would normally occur within six months, since by then the US Treasury would have issued another 30-year bond sucking the liquidity (and the premium) away from the one on LTCM’s books.

By pairing both legs of the trade, LTCM aimed to eliminate directional risk. A 30-year bond may normally trade in a range of 0.85 percentage points around its 5.50% yield. The combination of both trades reduced that variance by 25 times. 

LTCM did lots of trades like this. It had as many as 10 distinct categories of trades on its books, across interest rate swaps, fixed rate residential mortgages, European and Japanese government bonds, stock market volatility, merger-and-acquisition activity, dual-listed companies and more. Diversification added to the pairing strategy to reduce risk. Before August, LTCM had lost no more than $125 million on a single day across an asset base that peaked at almost $140 billion. 

But August 21, 1998 was different. One of LTCM’s strategies was risk arbitrage, where it would bet on the outcome of merger-and-acquisition deals by buying the stock of a merger target and selling short the stock of the acquirer. That morning, one of the deals on which it had a position was called into doubt. The target’s stock fell by 50% and LTCM was in the hole for $160 million. One by one, other trades started to fail, trades that should not have been correlated. A single bet tied to the US bond market lost $100 million; another $100 million evaporated in the UK. By the end of the day, LTCM was down $550 million – the single biggest daily loss in the firm’s history and five times more than anything its models anticipated. 

Rosenfeld flew back home to consult with his colleagues. On Sunday August 23, they gathered at the firm’s headquarters in Greenwich, Connecticut to analyse what was happening. “One after another, LTCM’s partners, calling in from Tokyo and London, reported that their markets had dried up,” recorded the Wall Street Journal. “There were no buyers, no sellers. It was all but impossible to manoeuvre out of large trading bets. They had seen nothing like it.”

It didn’t let up. By the end of August, LTCM was down 44% for the month; year to date, the fund was down 52%.

John Meriwether, the firm’s founder, called an old market contact, Vinny Mattone, for advice. “You’re finished,” Mattone told him. “When you’re down by half, people figure you can go down all the way. They’re going to push the market against you. They’re not going to roll [refinance] your trades. You’re finished.”

Initially, Meriwether, Rosenfeld and the team thought they could soldier on. They faxed a letter to investors on September 2 blaming losses on a major increase in volatility and flight to liquidity caused by the crisis in Russia, magnified by seasonally thin markets. “Many of the Fund’s investment strategies involve providing liquidity to the market. Hence, our losses across strategies were correlated after-the-fact from a sharp increase in the liquidity premium,” they explained.

At the time, the team considered the market conditions temporary; they invited clients to contribute more capital to the fund, reminding them that many had asked to add to their investment over a period when the fund had been closed. “[W]e see great opportunities in a number of our best strategies, and these are being held by the Fund,” they wrote. “[T]he opportunity set in these trades at this time is believed to be among the best that LTCM has ever seen.”

But Mattone was right – the firm was finished. Having run through $1.8 billion of its $4.1 billion of capital in August, LTCM ran through $1.9 billion more in September. 

What Went Wrong: Take One

Over the years, commentators have looked back at this episode to analyse what the team did wrong, how genius of this magnitude could fail.

“If you take the 16 of them, they have about as high an IQ as any 16 people working together in one business in the country, including Microsoft,” Warren Buffett told a group of students a few weeks after the collapse. “An incredible amount of intellect in one room. Now you combine that with the fact that those people had extensive experience in the field they were operating in. These were not a bunch of guys who had made their money selling men’s clothing and all of a sudden went into the securities business. They had in aggregate, the 16, had 300 or 400 years of experience doing exactly what they were doing and then you throw in the third factor that most of them had most of their very substantial net worth in the business. Hundreds and hundreds of millions of their own money up (at risk), super high intellect and working in a field that they knew. Essentially they went broke. That to me is absolutely fascinating.”

Various reasons have been given for the downfall. One is that the LTCM partners were overly beholden to their models. Rosenfeld was a finance PhD; he had spent several years as an assistant professor at Harvard before joining up with Meriwether and, like many of his fellow partners, was fluent in mathematical modelling.

“LTCM’s experts start with an academic research framework and quantitative orientation to build and refine models and to put concepts into practice on a large scale,” the firm’s marketing materials note. Its brochure describes LTCM not as a fund but as a “financial technology company.” (A red flag if ever there was one.)³

But many of the pricing anomalies that LTCM sought to exploit could be identified without sophisticated modelling at all. Although models were important to how trades were implemented and risks assessed, partners knew that their models were approximations to reality: They were far too experienced to enter into positions without an understanding of why a trading opportunity had opened up. The firm’s overall risk model even incorporated judgement-based buffers. Thus, investors were told to expect annual volatility of 20%, higher than both the volatility the risk model anticipated (14.5%) and the volatility the fund historically realised (11%).⁴

Another reason presented for the firm’s demise is that it employed too much leverage. “At LTCM the best minds were destroyed by the oldest and most famously addictive drug in finance,” wrote Carol Loomis for Fortune magazine. “Had the fund not grievously overextended itself, it might still be trucking along, doing its thing, working those brain cells.” Indeed, curtailing excessive leverage in the finance industry became a key policy conclusion in the aftermath of the collapse (albeit not one that was implemented in time for the 2008 financial crisis). 

LTCM did leverage up, with unfortunate timing, going into 1998, when it returned $2.7 billion of capital to investors without reducing its asset base commensurately. But prior to that, its leverage had been managed tightly. Before the capital distribution, the firm had an equity base of $7.5 billion and a balance sheet of $129 billion, equivalent to a 17x leverage ratio. Yet 80% of its positions were in government bonds of G-7 countries, traditionally lower risk assets (and reflected as much in bank risk weighting methodologies). And compared with banks, LTCM ran a conservative balance sheet. At the end of 1996, Morgan Stanley, whose equity base was similarly sized at $7.4 billion, ran a 27x leverage ratio; Salomon Inc ran a leverage ratio as high as 42x.⁵

As Donald MacKenzie, a professor of sociology, notes, “Blaming LTCM’s crisis on leverage is rather like attributing a plane crash to the fact that the aircraft was no longer safely in contact with the ground: it identifies a necessary, but in no sense a sufficient, cause. Leverage at best explains the fund’s vulnerability to the events of August and September 1998, but does not itself explain those events.”

So if it was neither seduction of models nor leverage that caused LTCM to implode, what was it? Join me over the paywall to explore…...