(Important)
For further reading
https://www.investopedia.com/terms/m/multifactor-model.asp https://corporatefinanceinstitute.com/resources/knowledge/other/multi-factor-model/
Statistical Risk Model (Steps)
The function fit_pca fits the PCA model with returns.
The function factor_betas gets the factor betas from the PCA model.
The function factor_returns gets the factor returns from the PCA model.
The function factor_cov_matrix gets the factor covariance matrix.
The function idiosyncratic_var_matrix gets the idiosyncratic variance matrix.
The function idiosyncratic_var_vector gets the idiosyncratic variance vector.
The function predict_portfolio_risk gets the predicted portfolio risk.
Create Alpha Factors
The function mean_reversion_5day_sector_neutral generates the mean reversion 5 day sector neutral factor.
The function mean_reversion_5day_sector_neutral_smoothed generates the mean reversion 5 day sector neutral smoothed factor.
Evaluate Alpha Factors
- The function sharpe_ratio gets the sharpe ratio for each factor for the entire period. The smoothed factors have a lower sharpe ratio
There is no expectation of any major improvement in the factor since the factor is already stable.
The function OptimalHoldings._get_obj returns the correct objective function.
The function OptimalHoldings._get_constraints returns the correct list of constraints.
The function OptimalHoldingsRegualization._get_obj returns the correct objective function.