Forecast comparison of exchange rate models with the Kalman filter

Pin T Ng, Asdollah Heidari

Research output: Contribution to journalArticle

Abstract

We show that the structural models' out-of-sample predictive performance of the dollar/pound, dollar/mark, dollar/yen, and trade-weighted dollar exchange rates is inferior to that of the simple random walk. However, the application of a Kalman filter to the structural models, which corrects the time-varying properties of coefficients over time, improves the predictive performance of exchange rate models. Our findings suggest that the coefficients of time-series models and structural models which use time-series data have shifted over time. Thus, a model that utilizes this information may increase its predictive performance in in-sample and out-of-sample forecasts.

Original languageEnglish (US)
Pages (from-to)435-443
Number of pages9
JournalTechnological Forecasting and Social Change
Volume41
Issue number4
DOIs
StatePublished - 1992
Externally publishedYes

Fingerprint

Kalman filters
Structural Models
Time series
Structural model
Kalman filter
Forecast comparisons
Exchange rates
Coefficients
Time series models
Time-varying
Out-of-sample forecasting
Time series data
Random walk

ASJC Scopus subject areas

  • Business and International Management
  • Management of Technology and Innovation
  • Applied Psychology

Cite this

Forecast comparison of exchange rate models with the Kalman filter. / Ng, Pin T; Heidari, Asdollah.

In: Technological Forecasting and Social Change, Vol. 41, No. 4, 1992, p. 435-443.

Research output: Contribution to journalArticle

@article{6db8a35099dd4e1f82f38af9caba35a5,
title = "Forecast comparison of exchange rate models with the Kalman filter",
abstract = "We show that the structural models' out-of-sample predictive performance of the dollar/pound, dollar/mark, dollar/yen, and trade-weighted dollar exchange rates is inferior to that of the simple random walk. However, the application of a Kalman filter to the structural models, which corrects the time-varying properties of coefficients over time, improves the predictive performance of exchange rate models. Our findings suggest that the coefficients of time-series models and structural models which use time-series data have shifted over time. Thus, a model that utilizes this information may increase its predictive performance in in-sample and out-of-sample forecasts.",
author = "Ng, {Pin T} and Asdollah Heidari",
year = "1992",
doi = "10.1016/0040-1625(92)90048-X",
language = "English (US)",
volume = "41",
pages = "435--443",
journal = "Technological Forecasting and Social Change",
issn = "0040-1625",
publisher = "Elsevier Inc.",
number = "4",

}

TY - JOUR

T1 - Forecast comparison of exchange rate models with the Kalman filter

AU - Ng, Pin T

AU - Heidari, Asdollah

PY - 1992

Y1 - 1992

N2 - We show that the structural models' out-of-sample predictive performance of the dollar/pound, dollar/mark, dollar/yen, and trade-weighted dollar exchange rates is inferior to that of the simple random walk. However, the application of a Kalman filter to the structural models, which corrects the time-varying properties of coefficients over time, improves the predictive performance of exchange rate models. Our findings suggest that the coefficients of time-series models and structural models which use time-series data have shifted over time. Thus, a model that utilizes this information may increase its predictive performance in in-sample and out-of-sample forecasts.

AB - We show that the structural models' out-of-sample predictive performance of the dollar/pound, dollar/mark, dollar/yen, and trade-weighted dollar exchange rates is inferior to that of the simple random walk. However, the application of a Kalman filter to the structural models, which corrects the time-varying properties of coefficients over time, improves the predictive performance of exchange rate models. Our findings suggest that the coefficients of time-series models and structural models which use time-series data have shifted over time. Thus, a model that utilizes this information may increase its predictive performance in in-sample and out-of-sample forecasts.

UR - http://www.scopus.com/inward/record.url?scp=50749132322&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=50749132322&partnerID=8YFLogxK

U2 - 10.1016/0040-1625(92)90048-X

DO - 10.1016/0040-1625(92)90048-X

M3 - Article

VL - 41

SP - 435

EP - 443

JO - Technological Forecasting and Social Change

JF - Technological Forecasting and Social Change

SN - 0040-1625

IS - 4

ER -