Press Releases
HEDGE FUND RETURNS: SHOULD RED OCTOBER REALLY HAVE BEEN A SURPRISE?
November 21, 2005
In his speech at GAIM entitled “How can we anticipate which market scenarios can hurt hedge fund performance?”, Dr Raphael Douady, Research Director of Riskdata discussed the lessons of October’s significant losses.
Riskdata’s risk profile analysis of aggregated HF & FoF indices, using data up to September 2005, shows a significant exposure to emerging market Asia, small caps, energy and high yield. If one were to use this risk profile to extrapolate what would be the behaviour of these aggregated indices from the 1st to the 20th October, one would find predicted losses of 0.9%, very close to the actual loss. This “out of the sample” estimate is strong evidence supporting the view that interprets ‘Red October’ as a direct consequence of a significant concentration of exposure on these markets, combined with the fact that a large number of funds liquidated their positions during the month to secure their cumulated returns.
Dr Douady said:
“The key question for FOHFs is whether October’s performance will lead to the loss of institutional money. The logical answer is that if the performance was a surprise, institutional money may be at risk. But we don’t think October should have been a complete surprise."
“If the portfolio is simply a black box from the perspective of institutional investors, then they can only form the conclusion that a 1.5% monthly loss is due to bad manager selection. If, on the other hand, it can be explained by market exposure, which the FOHF manager and his clients have understood and accepted, then there is much less chance of money being pulled out.”
“Effective risk transparency is the key to preventing redemptions. Risk profile analysis is precisely providing that – it means, as Emanuel Derman clearly formulated ‘…estimating the effects on our P&L of major market moves, thus providing effective risk transparency for our portfolios’.”
“There are three criteria for reliable scenario analysis – ie risk transparency:
“Firstly, it must be in line with the time horizon: as investors are locked in for one to three months, the risk model must integrate the portfolio dynamics: a static snapshot view of the portfolio is only useful to estimate overnight risk, not for monthly risk."
“Secondly, it needs to capture the distribution specificities of hedge fund returns: the fact that the upside is not the symmetrical of the downside, and that the impact of crisis events cannot necessarily be extrapolated from ‘business as usual behaviour."
“And thirdly, it needs to find the efficient frontier between two types of errors: important ‘type one’ errors that miss systematic risk, because the factor set is too small or the model inadequate, and ‘type two’ errors that merely highlight spurious relationships."
“The message is that appropriate models can and do work, giving FOHFs a better control of their exposure. There is a significant systematic component in most hedge funds, and this systematic component cannot be captured by over-simple models. Red October seems to be a very strong confirmation of the value of effective risk transparency for all market participants”.
For further information please contact:
Sales
Riskdata - London
Tel: + 44 (0)20 7659 0433
info@riskdata.com
Riskdata – New York
Finn Christensen
Tel: +1 212 931 5794
Media Fishburn Hedges
Florencia Panizza/Andrew Marshall
Tel: +44 (0)20 7839 4321
Florencia.panizza@fishburn-hedges.co.uk
“The challenge for us was to get an aggregated view of the fund risk profile while tracking the specific risk of each underlying strategy. We think Riskdata is able to address both of these issues simultaneously. Moreover, we can monitor the impact of any portfolio allocation on the fund risk profile.”
Philippe Uzan,
CDC IXIS Multi-Strategy Fund
Latest Case Study
Texas Treasury Safekeeping Trust Company Risk Management
Endowment Adopts Risk Budgeting Strategy and Selects Riskdata To Implement New Approach