Sun, 11/05/2008 - 19:07
A non-linear, factor-based model could have reduced the proportion of hedge fund 'time bombs' - managers who have simply been lucky rather than skilled in their past performance - exposed by the fallout from the credit crunch, according to a Riskdata study of hedge fund performance.
Using this approach would have enabled an investor in a broad hedge fund portfolio to achieve a 4 per cent return over the past nine months, compared with a flat return for the hedge fund universe as a whole during that period, Riskdata says
Since last summer, massive drawdowns have been experienced by a number of hedge fund managers, even though nothing in their track records prior to June 2007 appeared to have indicated that investors were potentially running a high risk.
The Riskdata study examines whether it is possible to detect funds that would not fare well under extreme conditions before their failings are exposed by market stress, examining the performance between last July and March this year of 3,200 hedge funds and funds of hedge funds that report returns to the HFR Database and had a track record dating back at least to December 2004.
The benchmark portfolio on which the study is based is iso-weighted on these 3,200 funds, which were broken down into three categories. For a group of 389 funds (12 per cent), the crisis was business as usual; they earned an average return of 8.7 per cent over the period.
A second group of 2,098 funds (65 per cent) experienced very high drawdowns to a degree that would normally occur only once every eight years under normal distribution standards, but this was in line with their experience with extreme risk before the crisis. Given the market conditions, investors had no reason to be surprised with this group's performance of 1.7 per cent.
The third group of 729 funds (23 per cent) experienced drawdowns that were not only high compared with their normal distribution models but were more than twice previous maximum drawdowns - that is, nothing in these funds' track records could have alerted an investor to the potential for a high level of loss. This group recorded an average decline of 9.4 per cent.
Unsurprisingly, Riskdata says, the highest proportion of 'time bomb' funds was among credit-related and relative value strategies, which accounted for 40 per cent of the total, while the proportion was much lower among managed futures, short bias and macro funds.
According to Riskdata, there are two principal approaches to measuring such risks. One is to use a sophisticated return-based risk model, which involves avoiding funds seen to have an abnormal return distribution, which have a high extreme risk compared with their normal risk.
Using return-based risk models results in the effective elimination of extreme risk-takers from the portfolio, but the benefit of this is largely offset because it also eliminates successful risk-takers, and completely fails to detect time bombs with hidden risk. This approach would have delivered a 0.4 per cent return over the test period, not much improvement on the flat benchmark portfolio return.
The alternative is a non-linear approach that focuses on modelling the relationship between market factors and manager returns and seeks to detect time bombs by identifying funds whose predicted risk, based on the long-term risk of underlying factors, is significantly higher than the observed maximum drawdown.
Using this criterion, an investor can eliminate funds where predicted extreme risk, using all factor history, is more than twice than the observed past maximum drawdown, or 2.3 times the fund's volatility, which corresponds to the 99th percentile of a normal distribution.
A non-linear, factor-based model would have avoided most potential time bombs while keeping successful risk-takers in the portfolio, and the investor would have outperformed the benchmark by 4 per cent.
'This study demonstrates that pure return-based models, however sophisticated, are insufficient to support sound risk budgeting,' says Riskdata research director Raphael Douady. 'Such models can help reduce risk levels, but do not reduce 'hidden' risk nor help find 'good' risk.
'An efficient non-linear, factor-based model is the only approach that can help investors discriminate between lucky managers and talented ones. The major advantage of factor-based models versus return-based ones is that they draw on the long-term history of market factors, including crises, even if the funds have a short history. Non-linearity is a key feature to capture the correlation breaks that occur during crises and liquidity traps.'
Riskdata chief executive Olivier Le Marois adds: 'The crisis of hedge fund performance over recent months has clearly demonstrated the value of an effective risk transparency, that is, a way to anticipate the most adverse market scenarios.
'Delivering 4 per cent of excess return versus an iso-allocated benchmark of more than 3,000 hedge funds is an incredible result, the more so as the selection was made at the end of June 2007 and kept frozen for the whole nine-month test period. This type of risk management, which recognises the specificities of hedge fund risk through factors, can be used in practice by investors to improve returns.
'The last few months have been tough for hedge funds - and the reason in large part is the 23 per cent of 'time bomb' funds that averaged a loss of 9.4 per cent, dragging the hedge fund universe as a whole down to an average zero return. Our study suggests that to continue to attract and retain institutional money, parts of the industry need to take the opportunity to rethink their approach to risk management.'
Thu 23/04/2015 - 06:00
Wed 22/04/2015 - 11:00
Mon 20/04/2015 - 08:03
Fri 17/04/2015 - 14:00
Thu 30/04/2015 - 06:00
Wed 29/04/2015 - 19:35
Tue 28/04/2015 - 18:30
Wed 22/04/2015 - 11:00
Fri, 01/May/2015 - 19:00
Fri, 01/May/2015 - 17:00
Fri, 01/May/2015 - 12:00
Fri, 01/May/2015 - 09:00
Fri, 01/May/2015 - 06:00
Thu, 30/Apr/2015 - 16:00
Sun, 03 May 2015 00:00:00 GMTSenior Analyst-Global Hedgefund-NY-CT
Sun, 03 May 2015 00:00:00 GMTAlgorithmic Trading Developer - C++, Linux, Python
Fri, 01 May 2015 00:00:00 GMT