WHAT WE’VE BUILT
Aapryl is an online database and analytic platform which gathers, analyzes and leverages manager, market and client data to help clients more cost-effectively and optimally manage their portfolios while simultaneously gaining more useful market insights.
- Search engine for data on entrepreneurial managers.
- Insightful analytics and peer comparisons.
- Information trend analysis and sharing.
- “Skill scores” based on skill vs. luck methodology.
- Qualitative scores calculated mechanistically.
THE AAPRYL ADVANTAGE
Aapryl is a suite of investment manager diagnostic analytical tools based on non-traditional approaches. Through the use of patent pending proprietary models, we offer asset allocators and other participants of the market the ability to more systematically screen and evaluate investment managers with a higher degree of confidence. The outcome from using this system is complementary to traditional due diligence processes.
AAPRYL MODAL COMPONENTS
Manager diagnostics analysis: Managers’ returns regressed against MSCI and Russell factors, style based indices and sector groupings to determine the effects of factor exposures on performance. Manager excess returns are then decomposed into a factor “clone” and stock selection returns.Consistency score: a modified batting average methodology to measure consistency of and duration of potential for outperformance. Manager Edge: a modified measure of benchmark outperformance over different periods.Active Share: the deviation of the manager’s holdings from the market benchmark.
AAPRYL METHODOLOGIES
Logistic regression, which measures the relationship between a categorical dependent variable (i.e. membership in a particular performance quartile over the subsequent three year period) and various independent variables.
Cross sectional recursive regression, which allowed us to test multiple independent factors,and select the best predictor(s) for the dependent variable. This method was used to generate forecasts, with the first regression including initial 24 observations. To avoid look-ahead bias in the dependent variable, we multiply regression betas with out of sample independent variables to forecast 3 year forward returns.
