When Genius Failed: The Rise and Fall of Long Term Capital Management

As the title suggests, this book encapsulates the entire journey of Long-Term Capital Management (LTCM). LTCM was a highly leveraged hedge fund, founded by John Meriwether, which collapsed in 1998 and had to be rescued by a consortium of 14 banks orchestrated by Federal Reserve Bank of New York. Roger Lowenstein has not just detailed LTCM’s life, but also captured very beautifully the nature and inner workings of its founders and the key team.

Setting up LTCM:

John Meriwether (JM) started his career at Salomon Bros, where he founded Arbitrage group, which was a pre-cursor of the future LTCM. He hired a team of quant-oriented traders to trade in bonds – this team used mathematical models to predict the price behaviour of bonds based on the historical relationships. Arbitrage Group was extremely successful, and it quickly became the most profitable unit of Salomon Bros, where they traded the proprietary capital of the firm.
JM founded LTCM in 1993, after he was forced to quit Salomon Bros in 1991. His core team from Salomon joined him in the next few months. In fact, the core team at LTCM was considered one of the best teams in any hedge fund ever –Lawrence Hillebrand ((had 2 degrees from MIT), Eric Rosenfeld (Assistant Professor at Harvard Business School), Greg Hawkins (PhD in financial economics from MIT). In addition to these, the team included two professors – Robert C Merton and Myron Scholes – who would later share the Noble Prize in economics in 1997 for inventing the Black-Scholes model of valuation of derivatives. The hedge fund also roped in David Mullins, Vice-Chairman of the Federal Reserve and second in line to Alan Greenspan, as a partner. With this star-studded team, LTCM raised US$1.25bn, probably the largest amount by any fund in its initial round. The major investors were large financial institutions like UBS, Merrill Lynch, PaineWebber and a few quasi-Govt. funds like HK Land & Development Authority, GIC, Bank of Taiwan, Pension fund in Kuwait, Sumitomo Bank of Japan. The fund also included ~US$140mn of partners’ contribution as well.

The Rise of LTCM:

LTCM had a good start – it returned 28% and 59% in the first and second years respectively, and on the back of it, was able to raise another US$1bn in 1995. At the start of 1996, the fund’s capital was $3.6bn and AUM was US$102bn implying a leverage of 28x. In 1996, it delivered another 57% return; with this the partners’ money in the fund stood at US$1.4bn. The main feature of LTCM was convergence trade – finding bonds which were mispriced vs. their futures in US, Europe, and Japan. Similarly, it also traded convergence among European sovereign bonds (e.g. Italy, LTCM had a large position and made significant profits in 1995). However, by now, competition was increasing and LTCM was looking for more opportunities for investment. It started to invest in equity markets – paired shares, merger arbitrage. After raising another round of fresh money and 25% returns in 1997, its capital account was about US$7bn. Struggling to find new avenues to invest, the partners in 1997 decided to return around half the money to initial outside investors (they returned ~US$2.7bn). In hindsight, this was a godsend moment for those investors who got the money back with 82% returns over 4 years. At this moment, the partners’ capital was ~US$1.9bn, 40% of the total capital of LTCM (US$4.7bn).

The Fall of LTCM:

The basic assumption of Black-Scholes model is consistent volatility. However, volatility never remains the same or consistent. This was the major fallacy of the model which LTCM didn’t capture. 1998 saw many issues come to the fore globally – Asian currency crisis & Russian debt default. These two events created massive scare among investors and started the flight to safety which invariably meant a move to treasury bonds. LTCM was badly hit, as investors rushed to safe assets and liquidity dried up in risky bonds. Moreover, before the Russian crisis, volatility was increasing and LTCM bet heavily on shorting volatility and believed that this volatility would normalise. With volatility increasing, LTCM was facing losses in almost all its trades now. The common theme among its trades was reversion to mean; the fund could have survived over a long term had it not been highly leveraged. It needed liquidity to survive the interim period of massive losses and since it was heavily leveraged, it couldn’t sell its bad/underperforming assets. Basically, in times of low liquidity, it’s not the asset one want to sell, rather the asset which one is able to sell.
JM tried desperately to raise fresh money, and, in this process, it allowed prospective buyers to check its books. LTCM partners claimed that this process was abused by bankers and in particular, Goldman Sachs, which traded heavily on its proprietary book. Gradually, LTCM trades which were heavily guarded by its partners were now an open secret. Almost every bank knew its positions, which hurt LTCM significantly. In hindsight, its decision to force original investors take out half of the money proved to be ominous. In September 1998, LTCM’s capital had dropped below US$1bn and AUM was still ~US$100bn, implying a leverage of an astronomical 100x.
Finally, the Federal Reserve of New York managed to get 14 banks on board, and they invested US$3.65bn for 90% equity and managed to keep the partners on board to run the fund. By mid-October when the Fed cut rates, LTCM started to stabilise. Thus, in a matter of just 6 months, LTCM, the star hedge fund was surviving on a lifeline. Just to give a perspective – the value of US$1 invested in LTCM at the start was US$4.11 in April 1998 and this had dropped to US$0.33 by the time the bailout was completed. However, internal conflicts among partners started to emerge and people started to leave the year after. The partners had so much of trust (or rather overconfidence) in their models and systems that they went on a roadshow in late 1998/early 1999 to investors explaining what really happened! In the end, the fund was liquidated in early 2000 after redeeming the investment from the consortium.

Conclusion

The key message from the book for me is that the mathematical models which were the bedrock of LTCM work perfectly well on back-tests and enormous historical data points but fall short when dealing in live markets. The partners remained steadfast to their models and rather than reviewing them critically, they doubled their bets. While conviction in one’s strategy is important, there can be no substitute for checks and balances, or for regular critical reviews. Excessive leverage kills; noted economist JM Keynes had said: “Markets can remain irrational longer than you can remain solvent”. This applies quite aptly to LTCM – losses on its fund led to the leverage increasing exponentially, finally resulting in a bailout and ultimately liquidation.

Hope you enjoyed reading my book review just as much as I enjoyed reading this book.

Saurabh Chugh

April 2023