Imperfect Macroeconomic Expectations: Evidence and Theory

Abstract

We document a new fact about survey expectations: in response to the main shocks driving the business cycle, expectations of unemployment and inflation under-react initially but overshoot later on. We show how previous, seemingly conflicting, evidence can be understood as different facets of this fact. We finally explain what the cumulated evidence means for macroeconomic theory. There is little support for theories emphasizing under-extrapolation or two close cousins of it, cognitive discounting and level-K thinking. Instead, the evidence favors the combination of dispersed, noisy information and over-extrapolation. * This paper was prepared for the 2020 NBER Macroeconomics Annual. We owe special thanks to our discussants, Ricardo Reis and Jessica Watcher, and the organizers, Martin Eichenbaum and Erik Hurst. We also thank Fabrice Collard for help with some of the material in Section 5. We finally thank Jeffrey Campbell, Olivier Coibion, Nicola Gennaioli, Cosmin Ilut, Yueran Ma, and Alp Simsek for useful comments. Angeletos acknowledges the financial support of the National Science Foundation (Award #1757198).

Dynamic Information Aggregation: Learning from the Past

“…The extent to which endogenous outcomes help aggregate information is increasing in the precision of private signals about economic fundamentals and is decreasing in the strength of general equilibrium feedback effects. With imperfect information aggregation, the equilibrium dynamics oscillate around the economic fundamental, and the forecast errors display an initial underreaction and a delayed overreaction which is consistent with the empirical evidence in Angeletos, Huo, and Sastry (2020). Together with the solution to the exogenous information environment, our results provide a second benchmark to characterize properties of equilibrium under incomplete information.…”

“…Moreover, the inflation dynamics in this confounding regime can be empirically relevant. The impulse response function of inflation forecasts displays an initial underreaction and a delayed overreaction, which is consistent with the evidence on expectations documented in Angeletos, Huo, and Sastry (2020) and Kucinskas and Peters (2018).3 The model also predicts waves of under-and overshooting of the inflation itself, similar to the identified responses from Christiano, Eichenbaum, and Evans (2005) and Altig, Christiano, Eichenbaum, and Linde (2011).…”

“…It is positive initially and turns to negative later on, implying that agents' average forecasts underreact initially and overreact later on. This pattern is an empirical regularity which is emphasized in Angeletos, Huo, and Sastry (2020) and also identified by Kucinskas and Peters (2018). 16 Angeletos, Huo, and Sastry (2020) rationalize this pattern by combining dispersed information with over-extrapolation, and here we find that it can be the result of imperfect information aggregation under rational expectations.…”

“…When the central bank employs a Taylor-type monetary policy, a more aggressive response of 3In Angeletos, Huo, and Sastry (2020), this pattern is rationalized by imposing a behavioral bias-over-extrapolationand we show it can be a natural result of imperfect information aggregation.…”

“…16This oscillatory pattern of forecast errors could help reconcile the seemingly conflicting evidence documented by Coibion and Gorodnichenko (2015) and Kohlhas and Walther (2019)-the former finds underreaction of average forecasts while the latter finds overreaction, as explained in Angeletos, Huo, and Sastry (2020).…”