Working Paper: NBER ID: w2568
Authors: John Y. Campbell; Robert J. Shiller
Abstract: Error-correction models for cointegrated economic variables are commonly interpreted as reflecting partial adjustment of one variable to another. We show that error-correction models may also arise because one variable forecasts another. Reduced-form estimates of error-correction models cannot be used to distinguish these interpretations. In an application, we show that the estimated coefficients in the Marsh-Merton [I9871 error-correction model of dividend behavior in the stock market are roughly implied by a near-rational expectations model wherein dividends are persistent and prices are disturbed by some persistent random noise. Their results thus do not demonstrate partial adjustment or "smoothing" by managers, but may reflect little more than the persistence of dividends and the noisiness of prices.
Keywords: cointegration; error-correction models; financial markets
JEL Codes: C32; E44
Edges that are evidenced by causal inference methods are in orange, and the rest are in light blue.
| Cause | Effect |
|---|---|
| one variable (C29) | another variable (C39) |
| stationary linear combination of levels (C32) | changes in at least one of the cointegrated variables (C32) |
| yield spread in the term structure of interest rates (E43) | changes in short-term rates (E43) |
| equilibrium error (D50) | changes in short-term rates (E43) |
| agents' forecasts (G17) | equilibrium error (D50) |