As we wrote in August, hedge funds seemed to be repositioning themselves for a down market. According to the Hennessee Group, average net (dollar) exposure had fallen from a high of around 55% in mid 2007 (the market peak) to 35% by mid 2008. Regardless of whether this means hedge funds can tel the future or not, shifting net exposure like that is obviously a useful skill. The mere fact that funds kept pushing down their net exposure does suggest some ability to read the tea leaves.
Now a new study seems to add credence to the argument that hedge funds have a statistically significant ability to time markets. Researchers from Citigroup and Athens University of Economics & Business found that there was a significant correlation between hedge fund market beta and market performance itself. In other words, hedge funds (specifically, “equity hedge” and “equity market neutral” hedge funds in the HFR database) became slightly more correlated with the market right before the market actually rose.
The authors set out to examine the behavior of three factors on hedge fund returns: value, momentum and “market”. But the market factor was the only one to show a significant relationship to future hedge fund returns.
As the table below shows, the correlation between the average market beta of hedge funds and the market’s returns is insignificant using the last month’s (“t-1”) market returns, is slightly higher using this month’s market returns and is relatively large (0.17) with next month’s returns. In other words, equity hedge funds seem to modestly ratchet up their market beta concurrent with rises in the market, but clearly ratchet up their beta in anticipation of a good month to come for the markets. Equity market neutral funds – although market neutral” also seem able to anticipate coming market “up months” (red circles below).
Quant strategies seemed to adjust their betas so quickly that they were able to ratchet-up their market exposure in the same month that the markets actually rose. Curiously, this market timing ability didn’t seem to exist for short sellers – arguably, the strategy most dependent on picking market directions.
To test the economic significance of their results, the authors calculated the trailing 24 month market betas for hedge funds since the turn of the century. Then they created a passive strategy that would invest in the market when this beta value ticks upward from one month to the next and sell the market when it ticks downwards. This passive strategy delivered a Sharpe ratio that blew the pants off that of the market itself.
The authors stop short of trying to explain the causality behind this correlation, but their conclusion is clear:
“…we find evidence that hedge fund managers adjust their ‘market’ betas in a way that exploits subsequent equity market return…According to the rolling correlation coefficients it appears that on average ‘Equity Hedge’, ‘Equity Market Neutral’ and ‘Quant Directional’ managers shift their betas at the – expost correct – expectation of a favourable market return in the next period.”
Time to take up tasseography.