Macro-Finance Essays on the Term Structure of Interest Rates

This thesis contributes to the literature that analyses the term structure of interest rates from a macro-finance perspective.

Chapter 1 of this thesis provides a structural interpretation behind the decline in the US term spread's predictive power with regards to future real output growth. Our analysis is conducted through use of a Dynamic Stochastic General Equilibrium New-Keynesian model that is estimated on both macroeconomic and financial data. Our findings indicate that it is changes to the composition of shocks hitting the US economy that has caused the term spread, through the endogenous monetary policy response, to cease being a useful indicator of future output growth.

Chapter 2 examines the importance of shifts in the expectations of agents in the form of "news shocks" in explaining the variation in the slope of the term structure of interest rates. The methodology employed in this chapter is a medium-scale Dynamic Stochastic General Equilibrium model that has been augmented to permit a role for both anticipated and unanticipated components in the usual array of structural shocks. In order to quantify the relative importance of each structural shock, the model is estimated via Bayesian methods on US data. We find the anticipated Total Factor Productivity shock to be quantitatively unimportant in driving US term spread fluctuations since, conditional on this shock, our model is unable to generate the observed leading procyclical movement of the spread found in the data. We do, however, find a limited role for the anticipated wage mark-up shock in that it accounts for a small share of the variation in consumption, hours and real wages.

However, it is the unanticipated shocks that account for the major share of variation in the term spread as well as other key macro aggregates.

The third and final chapter of this thesis examines the ability of the industry-standard Dynamic Stochastic General Equilibrium model to jointly explain both macroeconomic and financial data. We compute a second-order solution to our model in order to derive predic- tions for risk premia on equities and real, nominal and corporate bonds. Our central result is that by appending the Smets and Wouters (2007) model with Epstein-Zin preferences, long-run nominal risk and a credit market friction, we are able to generate realistic moments for the financial series under consideration without distorting the fit of our business cycle statistics.