Barra_CNE5

Provide risk forecasts by Barra China Equity Model

Code Usage

1. data.py Extract data from Wind database.

2. style_factor.py Build style factors.

3. factor_exposure.py Prepare factor exposures data for regression: truncate, winsorize and normalize style factors, build industry factors.Return a dataframe with hierarchy index (datetime, code) and columns containing: industry factors, 10 style factors, daily return and weight.

4. regression.py Calculate factor returns by weighted linear regression. Predict risk by factor returns.

5. utility.py Some utility functions.