Implementing hedge funds in a structured way
Before making its first hedge fund investment, Bayerische Versorgungskammer (BVK) used a four-step analysis and structuring process to help determine its goals and establish an appropriate vehicle for hedge fund investing. The first step was to consider whether the addition of hedge funds into the portfolio made sense. The analysis showed that, despite moderate yield expectations, the inclusion of hedge funds in the portfolio could have a positive effect on the entire portfolio by reducing risk.
The asset-liability analysis of the individual pension schemes was then integrated. BVK concluded that the asset class hedge funds should deliver a premium to reach the revenue goals. Therefore a conservative absolute revenue goal of 5% per annum was established. It was also decided that a risk reduction effect in the entire portfolio should be achieved through low volatility hedge fund investments and a lower correlation with traditional asset classes. A low shortfall risk was intended as a secondary goal.
BVK analysed the specific risks associated with employing different hedge fund styles on the basis of the CFSB/Tremont hedge fund sub-indices. Both yield distributions and their volatilities were analysed. All strategies were checked using the Jarque-Bera tests (which took into consideration the skewness and kurtosis of the yield distributions), when yields were normally distributed. Only equity market neutral and managed futures styles fulfilled the normal distribution assumption. BVK checked the degree of volatility by calculating the autocorrelation. As the volatility was determined to be too low, it was adjusted through the Geltner correction process. The loss potential, including the maximum drawdown, minimal monthly yield and the empirical VaR, were calculated and analysed for all strategies.
The correlation of different hedge fund strategies with the existing BVK asset classes was then examined. To do this, multivariate risk analysis (standard regression as well as the style analysis regression according to Sharpe) was performed by regressing the yields of the hedge fund styles on the yields of the BVK benchmarks. This analysis revealed that non-directional strategies tend to produce more positive alphas than directional strategies, whose yields tend towards beta risks. Both emerging market hedge fund strategies and short bias strategies were ruled out as emerging market strategies are explained to 80% through the beta risks of the emerging market bonds and shares while short bias strategies result in little alpha with only the negative beta risks remunerated.
BVK also took market efficiency risks into consideration. The average correlation of the yields of the various funds within a style category was considered to be an appropriate statistical parameter. (The higher the correlation, the more likely the individual hedge fund within the strategy is likely to take advantage of the same efficiencies.) BVK found that, measured on the average correlation, hedge funds from the event driven and convertible strategies are most similar in their investment styles. The more funds there are, the smaller the inefficiencies will be, and therefore the smaller the yields (according to tendency). The outlook for the fixed income arbitrage and equity market neutral strategies is obviously better.
Two mandate suggestions for a multi-strategy fund of funds and a relative value fund of funds were developed from this preliminary analysis. A multi-strategy fund of funds offers the manager a large degree of freedom in terms of the yield potential and the risk diversification balance between directional and non-directional strategies. It also helps to avoid unwanted style concentration. A broad and high-quality style coverage was therefore an important consideration in manager selection.
In the relative value fund of funds the yield realisation results from focusing on non-directional strategies, which focus on alpha and avoid beta risk. The fund is heavily weighted towards equity market neutral strategies, since a low shortfall risk, a very good Sharpe ratio and a low probability for an obvious reduction in market efficiencies are present. There is no preponderance of fixed income arbitrage, event driven or convertible arbitrage strategies in this fund since higher shortfall risks are present for these styles. There is also a high probability that market inefficiencies will decrease in these strategies.
The last stage in the process was to search for appropriate managers. The pitching was started through IPE Quest and a total of 86 fund of fund managers applied. At the end of the process three fund of funds were selected and funded.
Highlights and achievements
BVK sought to create an investment level for hedge funds that guaranteed a high measure of efficiency and flexibility similar to its master funds structure. “Masterschuldscheine”, “Schuldscheinplattform” or German Spezialfonds were examined regarding practicality and costs. But these were found to be either too expensive or not flexible enough. The ideal solution for BVK appeared to be the issue of a Luxembourg FCP fund, which allowed the scheme, after approval through the regulatory board, to use the FCP funds as a master fund for hedge fund investment. This solution turned out to be by far the most cost-efficient variant.
BVK’s investments in hedge funds take place either via a direct participation in single funds or fund of funds, through certificates or swaps. This is the first special funds investment ever to be established via an FCP in Luxembourg.