- Brinson performance decomposition model for all such funds
- Asset allocation for all such funds, asset allocation persistence, proportion of significantly positive asset allocation
- Fund benchmark performance, appropriateness of benchmark as baseline for performance
- Regression-based fund performance measures
- Model sensitivity to time window choice, third-party data sources credibility, model complexity, estimated interim parameters
- Performance persistence
- Performance association with fund asset
Note: Codes are primarily SAS and R programs. Data are mainly from Wind and CSMAR.
Code 0.R - XVIII.sas include whole process of building Brinson decomposition model for mutual fund in China, from data downloading, cleaning, organizing to model building and data analysis.