An article in the June 2006 edition of the European Central Bank’s Financial Stability Review (FSR) claims that hedge fund activities pose considerable risk to the financial system. We disagree with this conclusion, which is based on mere speculation. We outline the fallacies in the reasoning of the FSR article and makes some propositions on how to assess the welfare impacts of hedge funds.

In particular, we argue that it would be worthwhile for financial regulators to work towards obtaining data on hedge fund leverage and counterparty credit risk. Such data would allow a reliable assessment of the question of systemic risk. In addition, we argue that besides evaluating potential systemic risk, it should be recognised that hedge funds play an important role as “providers of liquidity and diversification.”

The conclusion in the FSR article is based on the following observations:

1. The article argues that the returns of different hedge fund strategies have become more correlated over time. Likewise, correlations between individual hedge funds classified in the same strategy have increased. This is interpreted as evidence of similar trades across different funds.

2. According to the article, hedge funds are investing in increasingly illiquid positions. This asset illiquidity exposes them to the risk of sudden redemptions from investors leading to a forced sell-off of illiquid assets.

3. Finally, the article shows that in certain market conditions (poor fund returns, increasing risk aversion, increasing interest rates), high investor redemptions are to be expected.

The article concludes that given these observations, there is a risk of "adverse effects of disorderly exits from crowded trades". However, these three observations can be contested, as detailed below.

The increasing correlation is derived using a correlation coefficient based on 12 monthly observations. With such a small sample, it becomes impossible to draw reliable inference. In addition, the increasing correlation of returns of single hedge funds in the same strategy is based on comparing data for 2004 with data for 2005. Just by way of random influences, it becomes quite likely to observe fluctuations between the two years, even if the true economic mechanism has remained constant. In other words, the results are indistinguishable from statistical noise. It should also be noted that the correlation actually decreases for four out of 11 strategies. It is unclear why the article interprets this result as “the fact that correlations are trending higher”. Furthermore, an increasing correlation between funds in the same strategy is by no means equivalent to similar positions. The academic literature interprets high correlation across funds in the same strategy as evidence of similar risk factor exposure. This risk factor exposure may well be obtained by holding different positions. This can be shown by taking the example of volatility risk exposure. Exposure to volatility risk may be obtained by trading index straddles, trading straddles on individual stock options, delta neutral trading in stocks or in the index, or trading volatility futures and options or even by holding stocks that have a high exposure to changes in market volatility.

An increasing correlation does not necessarily mean that the risk factor exposure is increasingly similar across strategies. It may simply be the case that the correlation between risk factors themselves is increasing. Take the example of two hedge funds, one is exposed only to volatility risk (short position in volatility), one is exposed only to stock market risk (long position in the stock market). If volatility has a tendency to increase when stock markets decline (and vice versa), then the two funds will have a high correlation. This has nothing to do with similar positions. In fact, there is zero overlap in the positions of these hypothetical funds.

The FSR article’s claim that hedge funds take positions in more and more illiquid securities is not substantiated by any data. The article merely states that “the liquidity of hedge fund investments may be decreasing, as recently hedge funds have reportedly been acquiring less liquid assets”. While it is obviously difficult to obtain data on this topic, it is revealing for the article’s methodology that a mere guess suffices to support the statement.

Most importantly, even if the claim of increasingly similar positions in illiquid securities corresponded to reality, this does not allow for the conclusion that adverse affects are likely to occur if investors start to redeem their shares in a massive way. In fact, hedge funds have a number of mechanisms that aim at matching funding liquidity and asset liquidity. In particular, lock-up periods - ranging from a month to more than a year - protect hedge funds from situations where they are obliged to sell assets due to decreasing funding liquidity. The FSR article itself asserts “a tendency […] to offer longer lock-up periods”. Therefore, even if hedge funds are investing into more illiquid securities, sudden redemptions by investors may simply be impossible. In addition to lock-up periods, hedge funds typically require a notice period prior to redemption, further increasing the time to redemption for investors, and they charge redemption fees that give additional disincentives for sudden redemptions.

For a full copy of EDHEC's document, please refer to the pdf below.