naked capitalism

Two major types of hedge fund strategies, namely event-driven (Newspeak for risk arbitrage) and statistical arbitrage (typically, very high volume trading to capture and correct anomalies in prices relationships in various markets, such as among stocks bonds, or derivatives, or across markets) are having trouble.

It isn’t yet clear how far reaching these problems are. MarketWatch reported that event-driven funds as a whole lost roughly 1% to 2% in July (not surprising given the recent turmoil in the deal market. But it remains to be seen how the losses were distributed (ie, whether there are some funds that really got it wrong and will face investor redemptions) and whether, given the sea change in the credit markets, there is now too much capital committed to this style:

Merger arbitrage hedge funds, which bet on the outcome of mergers and acquisitions, lost money in July over concern that turmoil in the credit market could derail some deals.

Merger arbitrage managers tracked by Hedge Fund Research lost 2.04% in July on average. Another index of merger arbitrage funds compiled by Hennessee Group fell 0.71% last month. Hedgefund.net’s merger arbitrage index fell 0.86% in July, based on early reports from managers.

Merger arbitrage hedge funds buy shares of target companies and bet against the stock of acquiring firms. As a deal nears completion, the share prices of the two firms usually converge, generating returns for managers. That’s known as the spread.
A surge in leveraged buyouts and other debt-fueling acquisitions by companies in recent years helped generate big gains for these types of hedge funds. Managers make the most money when the value of acquisitions is raised higher than the original offer price. See story on merger arbitrage.

Merger arbitrage funds generated returns of 12.93% on average in 2006, the best result for at least four years, according to Hedgefund.net.

For the stat arbs, DealBreaker, in “Blood Bath for the Quants?” says readers are reporting that some funds are in distress:

“The worst day ever,” one reader tells DealBreaker. Another describes it as “a bloodbath.”….

A number of DealBreaker readers have written in to say that the StatArbs are in trouble.

“The trading volumes are going thru roof, to the point that the program trading desks servers are crashing,” one reader writes. “For what it’s worth, I think it’s part of the unwinding of leverage from June and July. A number of multi-strategy shops are looking at losses and can’t even get a bid on certain swaps, fixed income and derivatives. So, what’s a risk manager/CIO going to do? Sell the most liquid part of their book.”

Note that many (most? all?) statistical arbitrage players rely heavily on computers not simply to identify possible trades, but also to execute them automatically. Speed is of the essence, since the anomalies are often fleeting. Needless to say, just like program trading of the 1980s, a computer-driven trading process is impossible to override when markets behave in unanticipated ways. You either have to halt the trading entirely or let it run. There is no way to apply human judgement except via modifying the models.