Yet, there is very little data on the actual risk management practices of hedge fund advisers. What are prevailing hedge fund adviser risk management practices? How do they compare with best practices? What should investors be wary of? What steps can hedge fund advisers take to better meet, and anticipate, the expectations of investors in the marketplace?
To answer these questions, we recently conducted a survey of the valuation and risk management practices of 60 hedge fund advisers from across the globe.
Many hedge funds are small operations with relatively few employees. Yet, they are increasingly investing in complex and illiquid assets and follow complex trading strategies that are the focus of much larger financial institutions with sophisticated risk management systems. The question for hedge fund advisers and investors alike is whether the risk management practices employed by the hedge fund industry are appropriate for the risks being incurred.
Our survey queried respondents on their use of 10 risk metrics for both position risk and portfolio risk, as well as risk governance. The results show considerable variation: some funds appear to have in place a risk management structure that is appropriate for their strategies and investments but other funds may be falling short. The sophistication of their risk management policies and procedures may not be keeping pace with the complexities of their strategies and investments, raising the concern that they are exposing themselves and their investors to risks that they do not fully understand.
Based on an in-depth analysis of the survey results, we identified nine ‘red flags’ where a sizable number of hedge fund advisers are not following industry risk management best practices – trading limits, stress testing, liquidity analysis, back testing, and an understanding of leverage are a few. Hedge fund advisers that raise one or more of these flags need to examine whether their risk management is appropriate for the risks they are taking. Hedge fund investors need to watch out for these red flags, and they must perform extra due diligence accordingly.
A starting point in measuring risk is measuring the degree to which the portfolio is diversified amongst positions. Position limits serve to limit exposure to a given position and industry concentration limits serve to limit exposure to a given industry with most firms using these limits to track and ensure diversification. But, these figures are not at 100% (industry concentration limits are not applicable to a sector fund, but even with sector funds factored out, only about 70% of respondents use industry concentration limits). Not using position and industry concentration limits is a disaster waiting to happen. There have been a number of large hedge funds that were forced into liquidation due, in part, to large investments in a single position or industry. Long Term Capital Management in 1998 (Russian government bonds) and Marin Capital in 2004 (General Motors debt) are two notable examples.
While concentration limits are a starting point for tracking diversification, they do not take in to account the correlation between positions that cross different concentration buckets. One model that does attempt to take account of such correlations is value at risk (VaR). It can be a good starting point for measuring market risk and credit risk embedded in a firm’s investments. Overall, 55% of respondents reported using VaR to analyse the risk of individual positions while 69% reported using VaR to analyse portfolio risk.
VaR is not an appropriate risk measure for all strategies – for example it would not be appropriate for a distressed debt portfolio – thus it is not surprising that not all respondents are using VaR. Given the complexity of some stochastic or parametric VaR models, the use of VaR is correlated with firm size, with larger firms more likely to use VaR than smaller ones. When smaller firms do use VaR, the analysis is often done by the fund’s prime broker or an external third party provider.
Stress testing, correlation testing and back testing VaR provides a snapshot of risk, based on ‘normal’ market conditions. It does not provide information on how a position or portfolio willbehave under extreme market conditions such as increased price volatility, decreased liquidity, or changes in asset correlations. In order to capture these low probability, but potentially high impact ‘tail events’, it is important to supplement VaR by conducting stress testing and correlation testing.
Stress testing should include shocks to market prices, volatility, the holding period to liquidate assets, leverage, and interests/credit spreads (eg flight to quality). In order to enable a firm to be more proactive in monitoring and mitigating risks in the portfolio, stress testing should include not only extreme movements but also changes that are only slightly greater than the norm in order to determine where thresholds or trigger points may be that would signal an adjustment in the portfolio. In addition, stress testing should also include scenario analysis where multiple factors are stressed at once. Traditionally, scenario analysis is often done by simulating past market events such as the October 1987 stock market decline, the 1997 Asian financial crisis, and the bursting of the dot com bubble in 2000. However, when reviewing these events it is important to note that the markets were different than they are today and that these events occurred only once. Therefore, scenario analyses should also include events that are relevant to the firm’s strategy and may occur given today’s market environment.
Correlations between market instruments are often not stable, and often tend towards 1 or -1 in times of stress. Correlation testing should include changes in correlations that occur in normal times as well as extreme movements to 1 or -1, similar to the stress testing approach. Only 60% of our respondents that used VaR for portfolio risk did both stress and correlation testing. Using VaR without doing both stress and correlation testing definitely raises a red flag, since without them VaR does not give a complete picture of risk – only a starting point.
Just as important is back testing the predictions of any model – whether risk measurements or valuation. Without back testing, firms run a greater risk of using flawed models to monitor and mitigate risk. For example, when back testing VaR one would expect occasional changes in the value of the portfolio that exceed VaR. However, in back testing VaR, if there are discrepancies that exceed the statistical nature of the model, the sources of the discrepancies need to be identified and adjustments made to the model or its inputs. Of those firms that used VaR for portfolio risk, 36% did not engage in back testing, running the risk that they are making decisions based on flawed models.
Adequate liquidity is critical to a hedge fund being able to continue trading in times of stress. It depends not only on cash on hand and the availability of credit, but on the ability to liquidate assets rapidly without incurring significant loss. Around 70% of our respondents track the liquidity of their positions.
Leverage can be accessed in a number of ways. There is explicit leverage, such as margin, short selling, and repurchase agreements, that shows up on the fund’s balance sheet; there is also off-balance sheet leverage, such as futures, forwards, swaps, and other derivative contracts, where either all or part of the notional value of the contract is off-balance sheet. The survey shows that approximately 60% of respondents monitor balance sheet leverage and approximately 50% of respondents monitor off-balance sheet leverage for both portfolio and position risk. Most hedge funds use at least some leverage.
One group of respondents that should certainly track off-balance sheet risk are those that hold off-balance sheet derivatives with embedded risk, such as forwards, futures, swaps, and other derivatives. Yet, of those respondents that reported holding some type of derivative contract, the percentage measuring off-balance sheet leverage for portfolio risk was no different than for the sample as a whole; only50%.
As is the case with VaR, firms should do stress testing and correlation testing of their liquidity and leverage measures. How do minor market movements affect liquidity? Will credit still be available? Will there be margin calls and will the fund be able to meet them? In times of stress many market participants may seek to liquidate a position at once, and this scenario should be tested when conducting liquidity and leverage analyses. Again, tests for minor movements and general trend analyses should also be conducted so that trigger points can be established and related courses of action determined. While the data are somewhat better than for VaR, there are still a sizable number of firms that are tracking liquidity and leverage without stress testing and correlation testing. In particular, 21% of respondents that reported tracking liquidity and 26% of respondents that reported measuring off-balance sheet leverage are not conducting both stress testing and correlation testing.
Aside from using risk metrics, all funds should have a written risk management policy. For the Fund, it is a crucial tool in communicating to all employees how risk is going to be managed. For investors, it is a first step in identifying red flags. Not having a written risk management policy is itself a red flag.
The survey revealed that approximately 80% of respondents had a written risk management policy. However, our experience is that such policies often lack sufficient breadth and detail. At a minimum, a written risk management policy should include acceptable levels of risk, how risk exposures will be identified, and how risks will be mitigated. It should be reviewed by the firm’s general partner or board of directors. Here the industry is falling short. Of funds with boards of directors, 43% had no input into the formulation of risk management policy, and 20% received no information on the policy whatsoever.
While risk management is a firm-wide responsibility, firms should have an independent risk management function reporting directly to senior management. In smaller hedge funds where there is often no senior management apart from the portfolio management, complete independence is difficult to achieve, but, to the extent practicable, a segregation of duties should be maintained.
An increasingly common position in the financial services industry is that of Chief Risk Officer (CRO). Hedge funds are much less likely to have a CRO than most other financial services firms. Our 2004 Global Risk Management Survey, which surveyed a broad array of financial institutions of all sizes, found that 81% had CROs. By contrast, just 47% of hedge fund survey respondents had CROs.
Interestingly, a hedge fund is much more likely to have a chief compliance officer (CCO) than a CRO. 88% of respondents reported having a CCO, and for firms with CCOs, the position constituted the individual’s full time or primary duty about two-thirds of the time. One likely explanation for the high percentage of CCOs is that registered hedge fund advisers are required to have CCOs in most jurisdictions.
Also of paramount importance to investors, regulators, and hedge funds themselves is controlling operational risk. Indeed, a study published in 2003 found that half of all hedge fund failures could be attributed to operational risk. Key to any control environment is a segregation of duties. This includes senior professionals independent of the investment team, who are able to take ownership of the operations of the fund. Over three quarters had a chief financial officer (CFO). Over 85% of the time, these were full time or primary duties. Only 14% of respondents had neither a COO nor a CFO.
Of course, simply having senior professionals to oversee the operational aspects of running a hedge fund is not sufficient. Policies need to be established and controls need to be in place to make sure those policies are being followed. Over 85% of respondents had written policies thatcovered record retention, third-party service level agreements, business continuity plans, and administration functions. While this is an impressive percentage, best practice would be to review and update these plans at least annually. There is some variation by policy, but in all cases these policies are being updated at least annually less than half the time. These should be reviewed by the fund’s general partner or board of directors. Here, as was the case for other risk management policies, the industry is falling short. Of respondents with boards of directors, over 45% of boards played no role in setting internal controls or setting policies to monitor those controls and over one-third of boards of directors did not receive any information on internal control policies or monitoring.
Policies need to be followed. A leading practice is to have an operational oversight committee to review and update operational risk policies, ensure that policies are being adhered to, and that any exceptions are discussed and resolved. The survey found that 42% of respondents had operational oversight committees. The survey also queried respondents on specific aspects of operational risk management. The data revealed that 58% of respondents included operation of due diligence, 55% of respondents included a full account reconciliation, and 40% included a review of third party providers in line with service level agreements as part of their operational risk management. While the absence of one of the aforementioned practices is not necessarily a reason to panic, there is cause to evaluate why that policy isn’t being followed. Additionally, if multiple practices are omitted from firm risk management procedures, a re-evaluation of the way your business is run or how your money is invested should be a high priority.
The hedge fund industry is certainly not a mature industry, but it is an industry entering middle age – an age where alpha will be harder to generate and where risk management will be of greater importance. As hedge funds grow in size and complexity, risk management practices are becoming increasingly important for hedge fund advisers, investors, and regulators alike.
Rob Mirsky is Head of the UK hedge fund practice at Deloitte