Finance
Macroeconomics
Econometrics

Financial Economics
Macroeconomics
Econometrics
International Finance

May 1998, (Oral) Macroeconomics, Econometrics
May 1997, (Written) Microeconomic and Macroeconomic Theory

Macro Factors and the Affine Term Structure of Interest Rates

Professor Christopher Sims
Professor Stefan Krieger
Professor Hua He

May 2001

M.Phil., Yale University, 1998
M.A., Yale University, 1997
B.A. Wuhan University, 1992

Carl A. Anderson Dissertation Fellowship, 2000-2001
Yale University Dissertation Fellowship, 2000-2001
Yale University Fellowship, 1996-2000
Ford Foundation Fellowship, Yale University, 1996
Graduate Summer Research Grant, Yale University, 1996

Teaching Fellow, Corporate Finance, Yale, 1999-2000
Teaching Fellow, Intermediate Macroeconomics (honor course), Yale 1999
Teaching Fellow, Introductory Macroeconomics, Yale, 1998-1999

Research Assistant, Professor Christopher Sims, Yale University and Princeton University, 1998-2000
Evaluated the real interest burden of government liabilities in the U.S. and Latin American countries using data on term structure and public debts.
Designed and estimated a Bayesian VAR model to analyze the government fiscal effects on the economic growth in the U.S. and other OECD countries.

Graduate Research Intern, World Resource Institute, Washington, D.C., 1997
Designed and estimated a panel data model to evaluate the health effects of air pollution in China.

"Macro Factors and the Affine Term Structure of Interest Rates", Yale University, 2000 (job market paper).

"Monetary Policy and the Slope Factor in Empirical Term Structure Estimations", YaleUniversity, 2000 (job market paper).

"Cyclical Behavior of Risk Premia in the Term Structure of Interest Rates", Yale University, 2000 (job market paper).

"Financial Constraints and Corporate Investment: A Second Look", mimeo, Yale University, 1998.

In this dissertation I construct a general equilibrium model to analyze the macroeconomic determinants of the term structure of interest rates. From this model, I derive an affine term structure in which the factors are observable macroeconomic variables. By providing a framework in which the term structure can be explicitly developed from a large class of underlying structural monetary business-cycle models, this dissertation connects the term structure literature with the business-cycle literature.

In the first essay I formulate and solve a general equilibrium business-cycle model characterized by monopolistic competition and sticky prices. An affine term structure model is derived from the model solution, with observable macro variables as the factors. The factor representing monetary policy is strongly mean-reverting, and the factor representing technology is more persistent. The impacts of macro shocks implied by the model on the yield curve are consistent with the empirical patterns in the data: i) monetary-policy innovations have large but transitory effect on the nominal bond yields, primarily by changing the slope of the yield curve; ii) technology shocks have more persistent effect in shifting the level of the yield curve, but have little effect on the slope. There are striking similarities in the pattern of factor loadings between my macro-factor model and the latent-factor models that are widely used in the empirical finance literature: from simulation studies of the macro-factor model I can also extract the "level" and "slope" factors, similar to the ones extracted from the latent-factor models. Another finding of interest from simulation studies is that most of the movement of the "slope" factor can be explained by monetary-policy innovations. This provides a theoretical explanation for the conjecture that the "slope" factor found in the empirical term structure studies represents the impact of monetary policy (Knez, Litterman and Scheinkman (1994)). This work extends that of Evans and Marshall (1998) who have studied the reaction of term structure to monetary policy in a dynamic stochastic general equilibrium (DSGE) model, but without relating their results to the empirical term structure models of the finance literature, and of Piazzesi (1999) who has studied the relation of empirical term structure models to macroeconomic variables, but without using an explicit DSGE model.

In the second essay the empirical relationship between the "slope" factor and monetary policy shocks is further studied. I first use the term structure data in the United States after 1982 to extract the "slope" factor implied by different term structure models: a two-factor Vasicek (1977) model, a two-factor Cox-Ingersoll-Ross (1985) model, and the macro-factor model developed and calibrated in the first essay. I then estimate a structural VAR model to identify the time series of the exogenous impulses to monetary policy during the same period of time. Finally, the extracted "slope" factor is regressed on the empirical monetary policy innovations, and a close relationship is revealed between them. This paper is an empirical confirmation of Knez, Litterman and Scheinkman (1994) 's conjecture as well as a validation of the macro-factor model developed in the first essay.

One of the implications of the macro-factor model developed in the first essay is constant risk premia in the Treasury market. However, empirical studies by Fama (1986) and others demonstrate that risk premia vary over business cycles, yet their results do not agree with each other. The third essay addresses this problem. A general equilibrium business-cycle model is solved with an algorithm that yields solutions accurate to the second-order, and numerical methods are adopted to produce time-varying risk premia. This work is still in progress. I will then explore the model’s implications of the behavior of risk premia over business cycles. I expect the need to incorporate financial market frictions into the model to match the empirical pattern in the data.