Return to search

Unspanned Macro Risks in the Term Structure of Interest Rates.

This dissertation advances the theoretical and empirical understanding of unspanned macro risks in the term structure of interest rates. The first chapter shows that a significant portion of bond risk premia is unspanned by the first three principal components of the bond yield covariance matrix, which together explain over 99.5 percent of the cross-sectional variation in bond yields. Fluctuations in unspanned factors induce movements in expected excess bond returns and expected future short-term interest rates in opposite directions, and are linked to both business cycle and higher-frequency variation in the level of macroeconomic activity. The second chapter examines the plausibility of unspanned macro risks in the term structure of interest rates within a simple New-Keynesian macroeconomic model augmented to allow for time-varying risk prices. In a calibrated version of the model, shocks to the natural rate of output less autonomous spending (so-called 'demand shocks') drive a large percentage of time-variation in bond risk premia, yet the importance of these shocks in explaining contemporaneous yield variation is significantly more subdued. The model demonstrates that within a plausibly calibrated macro-finance model of the term structure there may exist a significant wedge between factors driving cross-sectional variation in the term structure and those driving yield dynamics. The third chapter analyzes the ability of two leading nonlinear consumption-based term structure models to match salient features of bond risk premia. Both an external habit formation model (Campbell and Cochrane 1999) and a long run risk model (Bansal and Yaron 2004) are able to replicate the central finding of Cochrane and Piazzesi (2005): a single return forecasting factor---a tent-shaped linear combination of forward rates---captures all information in the term structure relevant in predicting excess bond returns. Both models fail to replicate an important feature of bond risk premia observed in the data, however. Neither model is able to account for the existence of unspanned macroeconomic factors in the term structure. In both models, yields summarize virtually all information in state variables useful in forecasting excess bond returns.

Identiferoai:union.ndltd.org:CHENGCHI/U0003482859
CreatorsBrooks, Jordan.
PublisherNew York University.
Source SetsNational Chengchi University Libraries
Detected LanguageEnglish
Typetext
RightsCopyright © nccu library on behalf of the copyright holders

Page generated in 0.0019 seconds