The impact of the term spread in US monetary policy from 1870 to 2013

Abstract

In this paper, we apply a novel econometric approach joint with an exhaustive revision of the main events in the history of US monetary policy in order to check the effectiveness of monetary policy focused on interest rates. Unlike the traditional cointegration approach, this new methodology allows us to break with the rigidity of traditional approaches in favour of letting the series be cointegrated, and the spread is able to follow a long-memory process; i.e., it does not necessarily need to be I(0) and also rejects the assumption that interest rates could follow the dichotomy I(0)/I(1). To the best of our knowledge, this is one of the first applications of the Fractionally Cointegrated Vector Autoregressive (FCVAR) model (Johansen and Nielsen (2012) and Nielsen and Popiel (2016)). Aiming to achieve this goal, we use two databases, i.e., the Jordà-Schularick-Taylor Macrohistory Database and Shiller’s database. Our results cannot reject the Expectations Hypothesis of Term Structure in this time period, and more importantly, we also find that the long-term rate drives the long-run relationship, contributing to the total proportion to the common trend; the persistence of the spread shows us effective control power over interest rates by the Fed.