Assessing predictability of the tendencies in economic activity using the yield curve

Abstract

We study the power of the yield curve to predict changes in economic activity in countries at different stages of economic development. Using the yield spread, we assess the relationship between the slope of the yield curve and cumulative and marginal real GDP growth in highly developed (U.S., Norway) and emerging (Russia, Ukraine) countries. Within- and out-of-sample models are constructed to evaluate the explanatory and predictive power of this relation respectively. We find the overall significance of the yield spread in explaining subsequent economic growth across all the countries contingent on a specific time horizon. The results of forecasts evaluation for the U.S., Norway and Russia demonstrate the rationality of using the yield spread models in highly developed countries as opposed to emerging ones where naïve autoregressive models are preferred. The thesis highlights the potential of further investigation in the post-soviet region.