statsmodels.tsa.vector_ar.var_model.VARResults.forecast_cov

VARResults.forecast_cov(steps=1, method='mse') [source]

Compute forecast covariance matrices for desired number of steps

Parameters:	steps (int) –

Notes

\[\Sigma_{\hat y}(h) = \Sigma_y(h) + \Omega(h) / T\]

Ref: Lütkepohl pp. 96-97

Returns:	covs
Return type:	ndarray (steps x k x k)