bayesline.api.equity.FactorCovarianceSettings

bayesline.api.equity.FactorCovarianceSettings#

model FactorCovarianceSettings#

Settings for factor covariance matrix forecasts.

Fields#
Field	Type	Required	Default
`combine_standardized`	`bool`	No	`False`
`halflife_cor`	`float`	No	`120.0`
`halflife_vol`	`float`	No	`60.0`
`halflife_vra`	`float` \| `None`	No	`None`
`nw_lags_cor`	`int`	No	`0`
`nw_lags_vol`	`int`	No	`0`
`nw_lags_vol_halflife_override`	`float` \| `None`	No	`None`
`shrink_cor_length`	`int` \| `None`	No	`None`
`shrink_cor_method`	`Literal`[‘LIS’, ‘QIS’, ‘GIS’] \| `None`	No	`None`

field combine_standardized: bool [Optional]#: If True, the correlation matrix is combined with an estimate based on the standardized returns. The halflife_factor_vra is used for the weights.

field halflife_cor: float [Optional]#: The half-life for the factor correlation.

field halflife_vol: float [Optional]#: The half-life for the factor volatility.

field halflife_vra: float | None [Optional]#: The half-life for the factor volatility regime adjustment. If None, no adjustment is applied.

field nw_lags_cor: int [Optional]#: The Newey-West lags of the factor correlation.

field nw_lags_vol: int [Optional]#: The Newey-West lags of the factor volatility.

field nw_lags_vol_halflife_override: float | None [Optional]#: The Newey-West lags override for the factor volatility for lagged returns.

field shrink_cor_length: int | None [Optional]#: If a shrinkage method is provided, the length of the shrinkage for the factor correlation.

field shrink_cor_method: Literal['LIS', 'QIS', 'GIS'] | None [Optional]#: The method to use for the shrinkage of the factor correlation. If None, don’t shrink.