numeraire.core.stats.alpha_regression#

numeraire.core.stats.alpha_regression(portfolio: Series, factors: DataFrame, *, nw_lags: int = 0) → AlphaResult[source]#

OLS of portfolio (excess) returns on factor returns; HAC (Bartlett) coefficient errors.

nw_lags=0 gives White heteroskedasticity-robust standard errors; positive lags add the Newey-West autocorrelation correction. Rows are inner-joined on the index.