Dynamic term structure models: The best way to enforce the zero lower bound

This paper studies whether dynamic term structure models for US nominal bond yields should enforce the zero lower bound by a quadratic policy rate or a shadow rate specification. We address the question by estimating quadratic term structure models (QTSMs) and shadow rate models with at most four pricing factors using the sequential regression approach. Our findings suggest that the two models largely provide the same in-sample .t, but loadings from ordinary and risk-adjusted Campbell-Shiller regressions are generally best matched by the shadow rate models. We also find that the shadow rate models perform better than the QTSMs when forecasting bond yields out of sample.