[Skip to content]

Sign up for our daily newsletter
The Actuary The magazine of the Institute & Faculty of Actuaries
.

When genius failed…

I have been working on a new theory of capital markets. It goes like this: if you are smart enough – and I mean really, really, smart; not just smart enough to program your VCR, but smart enough to understand stochastic calculus and partial differential equations and quantum physics – then you can lose a really spectacular amount of money.

My theory is only logical. If you are dumb, you probably don’t have much money, so you can’t do too much harm. But if everyone thinks you are a genius, they will throw money at you – they will beg for the privilege of investing in your company, with no questions asked. At least, that’s what happened to the geniuses at Long Term Capital Management (LTCM) – just before they lost $4.6bn and brought the world’s financial markets to the brink of disaster. When genius failed, by Richard Lowenstein, recounts the whole catastrophe.

Market efficiency

It starts off simply enough: John Meriwether, a quietly spoken young bond trader at Salomon Brothers, was a man who believed in the efficient markets hypothesis. He knew that markets might be slightly inefficient from time to time, creating small price discrepancies – but he was confident that in the long run, any such discrepancies would disappear. A smart trader would identify the inefficiencies, set up an arbitrage deal, and simply wait for the market to get more efficient.

Meriwether tried it, and it worked. He set up a bond arbitrage department at Salomon, and he hired the smartest people he could find: PhDs from MIT and professors from Harvard. They used sophisticated financial mathematics to ferret out anomalies in prices, and then they set up arbitrage deals – and they made a packet. In 1990, they made $1bn.They weren’t exactly modest about their achievements. They demanded – and got – higher and higher bonuses. One trader walked away with a $23m bonus for one year’s work.

Then, in 1991, it all fell apart. One of Salomon’s traders was a bit too greedy, and he broke the rules: he lied to the US Treasury. When the Treasury found out, it cost Salomon $200m in fines and compensation payments.

Meriwether was not involved in the scam – but he did know about it, and he had failed to report it to the regulatory authorities. Facing prosecution, with his name tarnished by the scandal, he was forced to resign. (The case was later settled out of court.)

Salomon (re)united

Meriwether decided he would start working for himself. In 1994, he set up his own hedge fund, LTCM, to do arbitrage deals. He persuaded a few of his old friends from Salomon to join him. There was just one hitch: since the market is pretty efficient, price discrepancies are rather small. An arbitrage deal involving $100 might only make a profit of a few cents, so if you want to be seriously rich, you have to invest a lot of money. But where do you get it?

Meriwether decided to get his capital – a mere $2.5bn – from rich investors. It was difficult at first, because he was charging much higher fees than normal, and he wouldn’t tell the investors exactly what he was going to do with their money. But he got most of the money and borrowed the rest from the banks which were only too happy to oblige: they lent billions and billions of dollars, at the most favourable interest rates, and with minimal collateral requirements. As Lowenstein says: ‘LTCM was as fetching as a debutante on prom night, and all the banks wanted to dance.’

High-risk venture?

After LTCM collapsed, a lot of people asked ‘Why did the banks lend so much money to such a high-risk venture?’ The banks believed Meriwether and his colleagues when they assured them that the risk position would be strictly monitored and managed using the most sophisticated tools – so it did not appear to be a high-risk venture.

LTCM refused to provide the banks with any details of their derivatives positions or liabilities, but they were willing to go in blind (after all, these people were geniuses and must have known what they were doing). The banks were eager to lend money to LTCM because they wanted to do business with them: LTCM would be doing big deals – and generating a lot of commissions and fees. And maybe, by watching, they could learn how to make the same kind of marvellous money machine.

  • In 1994, LTCM earned 28%.
  • In 1995, LTCM earned 59%.
  • In 1996, LTCM earned 57%.
  • In 1997, LTCM earned 25%.

By the end of 1997, LTCM had more money than it wanted. Investors were begging to be allowed to put more money into the fund – but the fund was unable to find new trades quickly enough to keep pace with the growth of its capital.

Things were starting to get a bit more difficult. Several new hedge funds were set up, specialising in arbitrage deals. Big players like Goldman Sachs were successfully copying LTCM’s arbitrage strategies, using the same mathematical models. More and more arbitrage investment leads to greater and greater market efficiency – so it was harder and harder to find any price anomalies in the market.

Branching out

In order to find profitable opportunities, LTCM had to branch out into new areas such as equities and commercial mortgages, but they didn’t really have much expertise in these areas, and the mathematical models did not work so well in these markets. The risks were quantifiable, and since risk-free arbitrage trades were vanishing, they had to start taking more chances by betting on unhedged deals (these were called ‘directional trades’). So investment risks were going up.

At the same time, LTCM pushed up its financial risk. Since they couldn’t find enough profitable investments for so much capital, they decided to give some of it back. They returned $2.7bn in capital to their investors (at the time, many of the investors complained bitterly, but soon afterwards they were thanking their lucky stars). Equity was reduced, the borrowings remained high, so the fund’s leverage ratio zoomed up to about 28 to 1. In 1998, LTCM started to lose money. All along, they had been betting that market efficiency would increase: although markets would fluctuate up and down, if you waited long enough, price anomalies would disappear (they called this ‘convergence’) – anything else would be irrational.

They knew, of course, that problems could arise in the short term, but weren’t worried. They had diversified their investments across every corner of the globe – in Russia, Brazil, Italy, Japan, Germany; based on historical correlations, it was highly unlikely that everything would go wrong in every market, all around the world. But that is exactly what happened.

Russian roulette

On 17 August 1998, Russia defaulted on its debt, and all hell broke loose. No one had believed that the IMF would let Russia go down the drain – but it did. Suddenly, the world looked a whole lot riskier and no one wanted to buy any risky securities – and if you owned them (and LTCM owned a lot), you could not sell them. The price of low-risk securities jumped while the price of high-risk securities plummeted. Instead of converging, the ‘spread’ in prices increased sharply and it happened in just about every market. Diversification did not help, because the international markets suddenly became much more highly correlated than anyone could have expected, based on historical data. After the Russian default, the value of LTCM’s holdings fell through the floor – they lost $550m in just one day (21 August). They lost 44% of their capital in the month of August.

Their creditors were holding those high-risk securities as collateral – and they demanded that LTCM top up the collateral to cover the amount of the debts. So LTCM had to sell whatever they could, at a loss, to meet the margin calls.

Circling sharks

LTCM knew that this panic was only temporary. In the long term, the market would bounce back and they would win back their money, but they couldn’t wait. The investment banks knew that LTCM was in trouble and would be forced to sell, at any price. LTCM was caught in a squeeze, and the predators were ready to make a killing. Employees would later complain bitterly about the predatory tactics of some traders.

  • On 21 September, LTCM lost $553m in one day.
  • On 22 September, LTCM lost $152m in one day.

On 23 September, the Federal Reserve stepped in to organise a bail-out for LTCM. As Lowenstein describes the negotiations, it was a lot like a high-stakes poker game, complete with desperate bluffs, last-minute bids, and under-the-table chicanery. Finally, a consortium of banks put up $3.6bn and LTCM was saved, but many of the LTCM partners were ruined.Of course, Meriwether was right. In the long run, the markets did bounce back into shape, the prices did converge. The banks who bailed out LTCM got all their money back. John Meriwether is now looking for investors for a new hedge fund.

Human perspective

When Genius Fails is a highly readable account: it focuses on the human factors of the disaster. Although it is ‘unauthorised’, the author has clearly had extensive discussions with many of the main protagonists, who described what they thought and felt, and why they behaved as they did. The financial theories are described clearly so that the lay reader can understand what went wrong, but without too much detail.

Nicholas Dunbar’s book, Inventing money: the story of LTCM and the legends behind it, is much harder going. Dunbar is primarily interested in the ideas, not the people. He starts with a history of derivatives trading, and he follows through the development of the options pricing theory (Black-Scholes formula) and risk management techniques (value at risk). His bibliography includes references to Daniel Bernoulli, Louis Bachelier, Richard Feynman, Oldrich Vasicek, Benoit Mandelbrot, Stephen Ross, and Fischer Black. The mechanics of the arbitrage deals are explained in much greater detail than in Lowenstein’s book and while it is interesting, it requires a fair amount of concentration and should be recommended only for people interested in financial mathematics. Some people might be interested in the impact of LTCM on the regulation of financial markets.

The collapse of LTCM raised a number of issues. The Bank for International Settlements has been trying to work out how banks can assess the risks associated with lending to hedge funds. The US government also conducted an enquiry into the bailout: was it right for the Federal Reserve to intervene? The transcript is available at www.commdocs.house.gov/committees/bank/hba51526.000/hba51526_0.htm.

cover

To purchase this book, please click on the cover image

02_09_07.pdf