Working Paper: NBER ID: w3451
Authors: Martin D. Evans; Karen K. Lewis
Abstract: Most studies of the expectations theory of the term structure reject the model. However, the significance of the rejections depend strongly upon the form of the test. In this paper, we use the pattern of rejection across maturities to back out the implied behavior of time-varying risk premia and/or market forecasts. We then use a new technique to test whether stationary risk premia alone can be responsible for these rejections. Surprisirj1y, this test is rejected for short maturities up to 6 months, suggesting that time-varying risk premia do not explain it all. We also describe hew this method can be used to test other asset pricing relationships.
Keywords: Risk Premia; Expectations Theory; Term Structure; Interest Rates
JEL Codes: E43; G12
Edges that are evidenced by causal inference methods are in orange, and the rest are in light blue.
| Cause | Effect |
|---|---|
| rejections of the expectations theory (D84) | time-varying risk premia (C22) |
| time-varying risk premia (C22) | rejections of the expectations theory (D84) |
| bond maturities increase (G12) | variance of risk premia increases (G40) |
| upward-sloping yield curve (E43) | market forecasts predict higher future interest rates than those that occur ex post (E47) |
| ex post biased forecast errors (C51) | nonstationary excess bond returns (G12) |
| time-varying risk premia alone (C22) | rejections of the expectations theory (D84) |
| short maturities (up to 6 months) (G19) | hypothesis that time-varying risk premia are solely responsible for rejections is rejected (C22) |