Working Paper: CEPR ID: DP2849
Authors: Carlo A. Favero
Abstract: The expectations model of the term structure states that the yields to maturity of long-term bonds are equal to the average of expected future short-term bond yields. This venerable model has been subjected to numerous empirical tests and almost invariably rejected. The empirical failure is generally attributed either to systematic expectations errors, or to shifts in the risk premia. In fact, the empirical tests, based on the estimation of single-equation models, are not able to discriminate between these two hypotheses. A recent strand of the macro-economic literature has analysed monetary policy by including the central bank reaction function is small empirical macro models. By simulating these models forward it is possible to derive the full forward path of short-term interest rates and hence to construct any long-term interest rate consistent with the expectations model. A direct test of the theory, based on full information, can then be immediately constructed by comparing observed long-term rates with the simulated ones and the associated 95% confidence interval. This is what we do in this Paper. Our results shed new light on the empirical validity of the expectations model.
Keywords: expectations model; small macro-models; term structure of interest rates
JEL Codes: E44; E52; F41
Edges that are evidenced by causal inference methods are in orange, and the rest are in light blue.
| Cause | Effect |
|---|---|
| Expectations model (C51) | Long-term bond yields (E43) |
| Central bank's reaction function (E52) | Future short-term interest rates (E43) |
| Simulated rates (E47) | Comparison with observed long-term rates (E43) |
| Presence of risk premia (G19) | Measurement of risk premia (G19) |
| Expectations theory (D84) | Consistency with term structure behavior (E43) |
| Systematic expectations errors (C51) | Empirical failure of expectations model (C59) |
| Shifts in risk premia (G19) | Empirical failure of expectations model (C59) |