Uncategorised posts

Correlations and the lesson, finally, learned

US-based quant shop AQR Capital has pioneered the notion of hedge fund beta as an investable product. With first-year performance numbers now in, Greg Bright spoke with the firm’s managing and founding principal, Cliff Asness.

Two years ago Cliff Asness weathered a media storm when he went public, almost daily for a month or so around August 2007, defending quantitative investment managers.

Most quant strategies had suffered a sharp fall during the very early stage of the global financial crisis. Some pundits believed that this signalled an end to quant investing because there was too much money investing in a similar fashion based on the same handful of factors.

Asness agreed with the reasoning, that too many investors were using the same or similar strategies, but disagreed vehemently with the conclusion that quant strategies in general had had their day.

It might be more difficult for quant managers to add value in the future, he said at the time, but you could say that about other strategies as well. This proved to be prescient and Asness appears to have been vindicated subsequently.

At least in part that vindication has been helped by the performance of one of AQR’s own funds, the DELTA strategy (a collection of hedge fund betas) which was launched in October 2008. Importantly, the fund made money in both up and down markets, including the cataclysmic December quarter last year, when almost all hedge fund strategies suffered.

The DELTA strategy is not hedge fund replication, though it shares many of its goals. It invests in the same underlying instruments that the hedge funds represented in the main indices invest in themselves, rather than replicators who use a mix of high level investments, largely futures and a lot of cash, which happen to correlate to the indices (but not capturing the spirit of the actual strategies as DELTA does). The full volatility DELTA fund, which is also available in a lower octane version, returned 26 per cent in the 12 months to end of September net of fees, with a realized volatility of 5 per cent.

Asness has been talking about hedge fund beta for some time. His long-held view is that most hedge funds have been too closely correlated with the markets. With DELTA, the market exposure is taken out, leaving the return from systematic trading strategies which are designed to match all-but two of the Tremont hedge fund index’s categories. This was clearly very helpful over the first few month’s of the fund’s life. AQR employs more than 65 underlying strategies for the fund.

Asness says that investors should not pay as much for hedge fund beta as they would expect to pay for other hedge fund investments, but the fee should be greater than a traditional index fund fee because of the greater complexity of the strategy and trading costs.

Most investors are still reeling from the various impacts of the financial crisis, one of which is the way correlations spiked to one. But Asness says that when he asks investors what sort of correlation they would be happy with from their alternatives, the answer is usually about 0.5. With DELTA, he targets zero. While he is the first to acknowledge the short-term is more a case study than a proof, through some very rough markets over the first 12 months the correlation has indeed been approximately zero.

One of the main lessons from last year, Asness says, was the level of market exposure within most hedge funds. Asness with colleagues wrote a paper in 2001 entitled ‘Do Hedge Funds Hedge’ which concluded that most hedge funds were correlated with the market to a factor of about 0.5 or 0.6. The paper, in his own words, made him wildly unpopular among some of his hedge fund colleagues at the time.

Over the past five years, he says, if anything the correlations have probably gone up, although he believes that they are likely to go back to their average of 10 years ago, but unlike the DELTA Fund, they are still way too high.

While Asness says that investors should learn some lessons from the events of 2008, they should be careful not to “over-learn”. By that he means that investors generally become too risk averse after bad events, at a time when there are often more opportunities.

“It’s an almost immutable law of the market that when fewer people want to do something it offers greater opportunity,” he says.

Join the discussion