Model averaging and persistent disagreement

In-Koo Cho, Kenneth Kasa

Research output: Contribution to journalArticle

Abstract

The authors consider the following scenario: Two agents construct models of an endogenous price process. One agent thinks the data are stationary, the other thinks the data are nonstationary. A policymaker combines forecasts from the two models using a recursive Bayesian model averaging procedure. The actual (but unknown) price process depends on the policymaker’s forecasts. The authors find that if the policymaker has complete faith in the stationary model, then beliefs and outcomes converge to the stationary rational expectations equilibrium. However, even a grain of doubt about stationarity will cause beliefs to settle on the nonstationary model, where prices experience large self-confirming deviations away from the stationary equilibrium. The authors show that it would take centuries of data before agents were able to detect their model misspecifications.

Original languageEnglish (US)
Pages (from-to)279-294
Number of pages16
JournalFederal Reserve Bank of St. Louis Review
Volume99
Issue number3
DOIs
StatePublished - Jan 1 2017

Fingerprint

Model averaging
Politicians
Model misspecification
Deviation
Rational expectations equilibrium
Faith
Stationary equilibria
Bayesian model averaging
Stationarity
Scenarios

ASJC Scopus subject areas

  • Business and International Management

Cite this

Model averaging and persistent disagreement. / Cho, In-Koo; Kasa, Kenneth.

In: Federal Reserve Bank of St. Louis Review, Vol. 99, No. 3, 01.01.2017, p. 279-294.

Research output: Contribution to journalArticle

Cho, In-Koo ; Kasa, Kenneth. / Model averaging and persistent disagreement. In: Federal Reserve Bank of St. Louis Review. 2017 ; Vol. 99, No. 3. pp. 279-294.
@article{67558b62d7a44c4e906957d3becea01e,
title = "Model averaging and persistent disagreement",
abstract = "The authors consider the following scenario: Two agents construct models of an endogenous price process. One agent thinks the data are stationary, the other thinks the data are nonstationary. A policymaker combines forecasts from the two models using a recursive Bayesian model averaging procedure. The actual (but unknown) price process depends on the policymaker’s forecasts. The authors find that if the policymaker has complete faith in the stationary model, then beliefs and outcomes converge to the stationary rational expectations equilibrium. However, even a grain of doubt about stationarity will cause beliefs to settle on the nonstationary model, where prices experience large self-confirming deviations away from the stationary equilibrium. The authors show that it would take centuries of data before agents were able to detect their model misspecifications.",
author = "In-Koo Cho and Kenneth Kasa",
year = "2017",
month = "1",
day = "1",
doi = "10.20955/r.2017.279-294",
language = "English (US)",
volume = "99",
pages = "279--294",
journal = "Federal Reserve Bank of St. Louis Review",
issn = "0014-9187",
publisher = "Federal Reserve Bank of St.Louis",
number = "3",

}

TY - JOUR

T1 - Model averaging and persistent disagreement

AU - Cho, In-Koo

AU - Kasa, Kenneth

PY - 2017/1/1

Y1 - 2017/1/1

N2 - The authors consider the following scenario: Two agents construct models of an endogenous price process. One agent thinks the data are stationary, the other thinks the data are nonstationary. A policymaker combines forecasts from the two models using a recursive Bayesian model averaging procedure. The actual (but unknown) price process depends on the policymaker’s forecasts. The authors find that if the policymaker has complete faith in the stationary model, then beliefs and outcomes converge to the stationary rational expectations equilibrium. However, even a grain of doubt about stationarity will cause beliefs to settle on the nonstationary model, where prices experience large self-confirming deviations away from the stationary equilibrium. The authors show that it would take centuries of data before agents were able to detect their model misspecifications.

AB - The authors consider the following scenario: Two agents construct models of an endogenous price process. One agent thinks the data are stationary, the other thinks the data are nonstationary. A policymaker combines forecasts from the two models using a recursive Bayesian model averaging procedure. The actual (but unknown) price process depends on the policymaker’s forecasts. The authors find that if the policymaker has complete faith in the stationary model, then beliefs and outcomes converge to the stationary rational expectations equilibrium. However, even a grain of doubt about stationarity will cause beliefs to settle on the nonstationary model, where prices experience large self-confirming deviations away from the stationary equilibrium. The authors show that it would take centuries of data before agents were able to detect their model misspecifications.

UR - http://www.scopus.com/inward/record.url?scp=85022231012&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85022231012&partnerID=8YFLogxK

U2 - 10.20955/r.2017.279-294

DO - 10.20955/r.2017.279-294

M3 - Article

VL - 99

SP - 279

EP - 294

JO - Federal Reserve Bank of St. Louis Review

JF - Federal Reserve Bank of St. Louis Review

SN - 0014-9187

IS - 3

ER -