StatPro Revolution Alpha uses many complementary methodologies in our risk analysis including Revaluations, Monte Carlo, Variance-Covariance, Factor Modeling, Historical Simulations, Relative Risk and Marginal Risk.
Revaluations: Revaluing each security in the portfolio due to changes in underlying markets provides a robust multi-dimensional stress test. Revolution Alpha's revaluation methodology considers changes in equity and commodity prices, yield curves, volatility surfaces, credit spreads and other markets and then calculates the new value of every security in the portfolio. This revaluation is done in incremental steps for each market at every level in the portfolio for a complete picture of how the portfolio would react to each scenario.
Monte Carlo: A Monte Carlo simulation is similar to the revaluation approach above, except that instead of choosing how markets move in steps, it randomly samples those movements from probability distributions and then revalues the portfolio. Revolution Alpha's Monte Carlo simulations focus on the most relevant areas of those distributions using a technique called “importance sampling” so that the simulations provide the most stable and accurate results available.
Variance-Covariance: This approach uses exposures, volatilities and correlations to estimate the amount of risk in every aggregation level of the portfolio. Revolution Alpha's implementation provides correlations between managers, strategies, sectors, securities and benchmarks and allows for "what if" style analysis where changes to portfolio allocations and volatility assumptions can be made. An important benefit of this methodology is simulating what happens when all correlations “go to one” as happened during the financial crisis.
Factor Modeling:Revolution Alpha uses a stepwise regression approach with over 300 factors to determine which macro factors best describe the portfolio. Whether based on manager returns or current holdings, this approach allows a large and complex portfolio to be distilled into a handful of meaningful risk drivers. It also can reveal surprises – so called "hidden bets" – that explain actual portfolio sensitivities.
Historical Simulations: Replaying the worst periods in market history is a popular way to test a portfolio’s robustness. Revolution Alpha uses our factor modeling to simulate any time period by considering how today’s portfolio would fare through the historical event.
Relative Risk: For portfolios that follow one or more benchmarks, relative risk measures like tracking error, beta, and relative VaR provide a point of comparison between the risks taken by the portfolio and the risks taken by the benchmark.
Risk Attribution: Also known as marginal risk, this approach signals how much risk is taken by each portion of the portfolio. Risk attribution is measured in a number of ways and answers questions like “which parts of my portfolio naturally hedge the other parts?” and “are the risks taken by my managers commensurate with the returns they generate?”
How We Can Help You
StatPro offers specialized services for a variety of investors, managers and asset owners.