statsmodels.tsa.ardl.UECMResults.fpe# UECMResults.fpe# Final prediction error using Lütkepohl’s definition \(((nobs+df_{model})/(nobs-df_{model})) \sigma^2\)
