Boom to bust in five years

Jonathan Davis: The demise of Long Term Capital Management shocked many

If the history of gold is a morality tale of timeless proportions, as Peter Bernstein's argument suggests, what significance can one read into a more recent specific event, the dramatic collapse of the ultimate egghead hedge fund, Long Term Capital Management? Two years ago the ineffably superior fund, founded by a legendary Salomon Brothers trader John Meriwether, and boasting two Nobel economics laureates on its payroll, was ignominiously bailed out by the main Wall Street banks, with the tacit blessing of the US Federal Reserve.

If the history of gold is a morality tale of timeless proportions, as Peter Bernstein's argument suggests, what significance can one read into a more recent specific event, the dramatic collapse of the ultimate egghead hedge fund, Long Term Capital Management? Two years ago the ineffably superior fund, founded by a legendary Salomon Brothers trader John Meriwether, and boasting two Nobel economics laureates on its payroll, was ignominiously bailed out by the main Wall Street banks, with the tacit blessing of the US Federal Reserve.

In five short but spectacular years, LTCM had run a course from being the hottest thing to have hit the investment world in years to one of the most spectacular collapses in the modern financial era. In April 1998, after four years of profit, based on the successful implementation of awesomely complex mathematical trading strategies, the fund had $5bn of equity capital, and a client list that included many of the biggest names on Wall Street, among them the bosses of such blue-chip names as Goldman Sachs and Merrill Lynch.

Five months later, LTCM was effectively bust, kept afloat only by the considered opinion of the Federal Reserve chairman Alan Greenspan and the Wall Street mighty that it was too dangerous to let fail, given the loss of confidence sweeping through the world's financial markets at the time. In those five months, the supposedly risk-free trading strategies in which LTCM specialised had succeeded in running up losses of nearly $5bn, including two 24-hour periods when this one small fund, numbering no more than two hundred employees, managed to lose in excess of $500m of its shareholders' money.

How could it have happened? And what were the factors that brought it so dramatically to the brink? The outline of the story has been known almost from the day the fund imploded, but only now are the full details coming to light, and nowhere more graphically than in Roger Lowenstein's soon-to-be-published full-length account of the great drama.

Mr Lowenstein's book on the LTCM saga, When Genius Failed, is to be published by Fourth Estate in the New Year, after it has finally worked its way through the inevitable thickets of legal scrutiny.

What emerges from reading an advance proof of this fantastic story is a remarkable insight into human elements - fear, greed, pride and arrogance - that lurk not far below the surface of financial markets.

Those who like romans-á-clef will enjoy the drama of the denouement, as the big animals of the largest Wall Street firms, bristling with ego and self-importance, gather to try to reach an agreed settlement on how to bail out the stricken hedge fund (whose downfall they had all indirectly helped to bring about by their imprudent lending to the fund in its heyday). More interesting are the nuggets of detail illuminating how the fund set about making its mark, and the fatal flaws in its seemingly cast-iron strategy, which brought the fund to its knees, and seriously destabilised the global financial markets, at a cost felt by ordinary investors everywhere.

In effect, what LTCM did was to harness the greed and drive of some of the pushiest traders on Wall Street to an academic view of the way the financial world should work, rightly hailed as a breakthrough when it was proposed, but which had grave practical shortcomings.

Most of the initial LTCM trading was straightforward by Wall Street standards, bets on convergence and bets on movements in relative value. The first were based on the assumption that two securities which had diverged from their historic relationship would eventually revert to their original relationship. For example, LTCM would buy huge amounts of a six-month-old, 30-year Treasury bond, and sell equally huge amounts of the just-issued but more expensive 30-year Treasury bond, aiming to profit from the expected convergence in the two bonds' yields.

Relative value bets were similar but subtly different. They were positions that stood to produce a profit if LTCM was able to determine correctly that one set of securities were overpriced relative to another. In both cases the fund's managers deliberately avoided taking a simple position on which way a particular variable (such as interest rates or bond yields) was likely to move. Trying to forecast short-term movements in such things was something which academic theory had taught them was a mug's game. Where LTCM differed from the rest of Wall Street was, first, that its investment strategy was backed by detailed and complex computerised analysis of historic patterns of behaviour, which allowed it to make robust (but, as it turned out, fallible) assumptions about the probability and risk of future outcomes; second, that its traders had massive confidence in their own cleverness and ability to crunch the numbers better than anyone else; and third, that they employed breathtaking amounts of leverage to try to maximise the profitability of their trading activity, so confident were they that they had a better mousetrap than the rest of Wall Street.

Thus, while the long-term government bond trade mentioned above yielded only a minuscule profit per transaction, it was turned into a genuine moneyspinner by virtue of being undertaken in huge volumes - LTCM bought and sold $1bn of each security - and being financed almost entirely by borrowed money. One of the firm's founders said the fund was less an investment firm than a highly sophisticated financing operation.

Mr Lowenstein has unearthed the startling fact that in the fund's second year, when it could still do no wrong, it made a gross return of 59 per cent, of which the partners took 25 per cent as their performance fee. Yet the actual cash on cash return of the fund itself was just 1 per cent. The rest of the performance - 58 per cent - was accounted for by leverage. LTCM routinely borrowed 30 times its equity capital, most at virtually no cost, thanks to its cult status and godlike reputation, and the wilful gullibility of the banks.

Eventually, the mighty edifice crumbled. The mathematical models proved to be flawed. Real life does not follow a normal bell-shaped probability distribution. The pattern of events in the summer of 1998, when all LTCM's bets went wrong at the same time, should have been a once-in-a-millennium event, according to the fund's probability wizards. But it happened, mainly because markets in times of stress cease to behave like well-ordered computer models, and start to behave like emotionally-charged herds, a lesson that has to be painfully relearnt every few years.

Even the real world, dramatic though it was in 1998, was not itself enough to bring about LTCM's downfall. What ultimately did for the fund was the arrogance and greed of some key principals, which led them not only into ever-riskier strategies (such as betting on the volatility of futures and options contracts) but persuaded them to load up the fund with so much debt.

Less than a year before the collapse, LTCM actually forced most of its investors to take back half their capital to gear up the performance of the fund even further. Some of the partners also borrowed heavily personally to turn their own shareholdings - already worth a combined $1.9bn in April 1998 - into something even greater. When the crunch came, the fund simply ran out of liquidity, other traders ganged up on them and the partners found themselves squeezed out of all (or nearly all) the wealth they had so effortlessly acquired through their early fabulous success. Robert Merton, one of the two academics who earned the Nobel prize for their efforts in modelling option prices, provided the fund's epitaph when he told his cheering students at Harvard on the day the prize was awarded: "It is a wrong perception to believe you can eliminate risk just because you can measure it."

But make no mistake: it was the gearing, and the human failings it represented, which ultimately did for LTCM.

* davisbiz@aol.com