FactorToWeightBaseClass

class sigtech.framework.strategies.equity_factor_basket.FactorToWeightBaseClass

Baseclasses: ABC

Subclasses: CapWeightedFactorToWeight

Class containing a factor_to_weight method to calculate stock weights from factor exposures.

This allows us to group together this method with the ability to add factors to the FactorExposures object.

static combine_factors(factors: DataFrame, factor_weights: Optional[dict] = None, z_score: bool = False, proportion: Optional[float] = None, long_only: bool = False) → Series

Sum over the supplied stock-factor loadings to return stock weights.

Parameters:

• factors – Factor indexed pd.DataFrame of stock loadings. Stocks as columns.

• factor_weights – Optional dictionary of factor weights. Keys should match factors index.

• z_score – Boolean of whether to normalise stock-factor loadings before combination. This allows different factors to be put onto the same scale. Default is False.

• proportion – Optional, proportion for long/short allocations.

• long_only – If proportion is passed with long_only=True, only the top proportion % are returned with positive values and the negative mirror image is removed. If long_only=False both the positive and negative weights are returned. If proportion is not passed this parameter has no effect.

Returns:

Series of stock weights. Index will be (possibly subset of) columns of factors.

static summed_weighted_factors(factors: DataFrame, factor_weights: Optional[dict[str, float]] = None, z_score: bool = False) → Series

Sum the stock-factor loadings over factors to return stock weightings. If weights is passed the factors are appropriately weighted, otherwise they are equally weighted.

Parameters:

• factors – Factor indexed pd.DataFrame of stock loadings. Stocks as columns.

• factor_weights – Optional dictionary of factor weights. Keys should match factors index.

• z_score – Boolean of whether to normalise stock-factor loadings before combination. This allows different factors to be put onto the same scale. Default is False.

Returns:

Series of stock weights.