Multi-factor manager performance attribution

Client

A Large US based Insurance firm

• The client, a leading asset management company, wanted to identity parameters that can help in analyzing and assessing the security selection skill of a portfolio manager in a portfolio

• Employed multi-factor performance attribution model to study active returns breakdown to identify skill-based returns.
• Aggregate time-series data for all funds to create panel data and in turn perform analysis to assess manager’s skill and also highlight its persistence over a long timeframe
• Performed regression analysis along with various other descriptive analysis using MATLAB. Used independent variables like risk, duration, volatility and the dependent variable was active returns.
• Wrote efficient codes for running regressions & performing various robustness tests. Automated entire process starting from loading new data feeds on a monthly basis to running the regression model
• Results were significant with factor exposures explaining excess returns for portfolio to a great extent
  

• Client published the findings from the study as part of research paper, and attributed Crisil GR&RS associate for the excellent work
• This study helped in analyzing new strategies to generate active returns