The Sub Prime Crisis: Hedge funds in trouble

(Ref: The Economist, dt. 18 August; The Financial Times, August 20)

One of the first groups of market participants to find themselves in trouble when the sub prime crisis unfolded was hedge funds. Some of the hedge funds which have found themselves badly mauled by the market turmoil include the global equity fund of Goldman Sachs, Renaissance, a successful quant fund and various hedge funds in Japan. Basin Capital of Australia has also lost heavily.

Both human factors and computer driven models have contributed to the sad plight of these hedge funds. Banks started putting pressure on hedge funds to increase their collateral forcing them to take desperate measures to generate liquidity. At the same time, long-short equity neutral funds which assume that some stocks will rise while others fall, found that underlying assumptions behind their computer based trading strategies were faulty.

On August 13th, Goldman announced its leading global equity fund had lost more than 30% of its value within a week. The bank had to put in $2 billion of its own money and $1 billion contributed by investors. Renaissance, founded by James Simons, a prize wining mathematician also suffered big losses. Quantitative hedge funds in Japan seem to be among the worst affected. Whether hedge funds will stage a comeback in the foreseeable future depends on a big question. Will the pension funds, endowments and rich individuals investing in hedge funds hold their nerve?

As hedge funds find themselves in serious trouble, quantitative models are again coming under close scrutiny. Recall that they were the undoing of the celebrated Long Term Capital Management (LTCM) in 1998. Quantitative models try to find minute market inefficiencies and exploit them. Computers help in finding these inefficiencies quickly so that the traders can take advantage of them before they disappear. As many people start using similar models, such opportunities disappear.The only way to get ahead is to come up with more complex and sophisticated models. Long-short funds, for example generated profits as recently as February/March 2007, using this approach. But to get really good returns, leverage is needed. And as we have seen, leverage can be a risky proposition.

One major lesson which seems to be emerging from the crisis is that quantitative models also cannot overlook the behavioural factors involved in trading. As John Authers has mentioned in the FT (Aug 20), “ …human judgement when it comes to investment is flawed in predictable ways that lead to predictable mis-pricings in the market. A quantitative model that will follow rules set for it by humans, without the risk of human judgment subsequently messing things up, is needed to take advantage of those mispricings.”

In recent weeks, these models have come to naught. The models wanted the funds to hold certain positions. But the need to generate liquidity forced funds to sell their good investments. As many quants followed suit, what we saw were 25 standard deviation events which in normal circumstances, would happen only once in 100,000 years. Leverage amplified these losses. In other words, the models failed to account for the “fat tails.”

Authers mentions that mathematical models will need to improve significantly in the months to come. The quants must take into account:
 The herding effects, i.e., other funds taking similar positions
 Impact of their own actions on the market
 Need to use leverage to magnify returns.

If the regular steam of profits based on the models, comes up with huge losses occasionally, the combination of rigid quantitative strategies with leverage may not be all that appealing.

In short it looks as though there may not be that many leveraged active quant funds we will see going forward. Quite likely the ones that remain will belong to the large well capitalized investment banks.