Portfolio Risk Analysis

Inheriting all valuation problems of individual funds and single asset classes, multi-asset investment portfolios, especially with a high concentration of hedge funds (hedge fund of funds – HFoF), are prone to additional levels of risk.

First, HFoFs tend to exhibit higher negative skewness and kurtosis than individual funds or traditional asset classes. Therefore, the inappropriateness of the mean-variance theory, which assumes a normal distribution, applies to HFoF to a higher extent. Put simply, relying on standard deviation and its related risk metrics to gauge market risk can lead to either overestimating or underestimating the actual risk levels due to the non-normal distribution of hedge funds.

Second, the diversification of portfolio constituents leads to a higher portfolio correlation with major markets compared to individual funds. Therefore, increasing a portfolio’s number of assets doesn’t necessarily improve its diversification and market neutrality.

Third, complex multi-asset portfolios with diverse asset allocations require measuring risks across all risk categories: market risk, liquidity risk, currency risk, inflation risk, concentration risk, volatility risk, strategy risk, operational risk, legal risk, and fraud risk.

Finally, high cross-correlations across underlying funds and indices make portfolio attribution and factor analysis more difficult - often classic regression models do not work and more advanced techniques should be applied.

Hedge Fund Of Funds and Multi-asset Portfolios Valuation Challenges

• A greater degree of non-normality of return distributions compared to individual funds, which makes the mean-variance framework and standard deviation derived risk statistics hardly applicable. Consequently, alternative statistics capable of analyzing various return distributions, such as Value-at-Risk, Omega, Kappa, Conditional Value-at-Risk (CVaR), and Expected Tail Loss (ETL), are employed. However, leveraging advanced statistics introduces additional challenges and computational complexities.
• High portfolio correlation with indices. While diversifying across various asset classes is a typical strategy to mitigate this, it may not be as effective with alternative investments like hedge funds. In fact, adding more hedge funds to the portfolio often leads to increased correlation with major market indices.
• Excessive cross-correlations among underlying funds can significantly degrade portfolio diversification. Our findings regarding the optimal number of hedge funds in a portfolio are consistent with François-Serge Lhabitant's research in "Diversification, Rational Expectations, and the Sector Risk Model." The ideal number of assets in a multi-asset portfolio or a hedge fund of funds falls within the range of 15 to 18.

What We Can Do For You

Our on-demand portfolio risk assessment services encompass the following offerings:

• Advanced hedge fund portfolio factor analysis (Kalman filters, Elastic-net, LASSO, the Akaike information criterion etc.)
• Portfolio stochastic simulation
• Marginal risk contribution analysis
• Portfolio Sensitivity analysis
• VaR and tail-risk analysis through different Value-at-Risk models, e.g. Hybrid or Modified VaR
• Principal Component analysis
• Extreme event Stress testing including correlation stressing
• Portfolio What-if analysis
• Distribution adjustment analysis
• Structured products' analysis (on-demand)
• Portfolio Cross-correlation and diversification analysis