# arch.unitroot.cointegration.CointegrationAnalysisResults.residual_variance¶

property CointegrationAnalysisResults.residual_variance

The variance of the regression residual.

Returns

The estimated residual variance.

Return type

float

Notes

The residual variance only accounts for the short-run variance of the residual and does not account for any autocorrelation. It is defined as

$\hat{\sigma}^2 = T^{-1} \sum _{t=p}^{T-q} \hat{\epsilon}_t^2$

If df_adjust is True, then the estimator is rescaled by T/(T-m) where m is the number of regressors in the model.