Non-Linear Predictability in Stock and Bond Returns: When and Where is it Exploitable?

Federal Reserve Bank of St. Louis Working Paper No. 2008-010A

74 Pages Posted: 30 Apr 2008 Last revised: 15 Jan 2009

See all articles by Massimo Guidolin

Massimo Guidolin

Bocconi University, Dept. of Finance; Bocconi University - CAREFIN - Centre for Applied Research in Finance

Stuart Hyde

Alliance Manchester Business School - University of Manchester

David G. McMillan

University of Stirling

Sadayuki Ono

Hiroshima University

Date Written: January 2009

Abstract

We systematically examine the comparative predictive performance of a number of linear and non-linear models for stock and bond returns in the G7 countries. Besides Markov switching, threshold autoregressive (TAR), and smooth transition autoregressive (STAR) regime switching models, we also estimate univariate models in which conditional heteroskedasticity is captured by a GARCH and in which predicted volatilities appear in the conditional mean function. Although we fail to find a consistent winner/out performer across all countries and markets, it turns out that capturing non-linear effects may be key to improve forecasting. U.S. and U.K. asset return data are "special" in the sense that good predictive performance seems to require that non-linear dynamics be modeled, especially using a Markov switching framework. Although occasionally stock and bond return forecasts for other G7 countries also appear to benefit from non-linear modeling (especially of TAR and STAR type), data from France, Germany, and Italy imply that the best predictive model is often one of the simple benchmarks, such as the random walk and univariate auto-regressions. U.S. and U.K. markets also provide the only data for which we find statistically significant differences between forecasting models. Results appear to be remarkably stable over time, robust to changes in the loss function used in statistical evaluations as well as to the methodology employed to perform pairwise comparisons.

Keywords: Non-linearities, regime switching, threshold predictive regressions, forecasting, predictability in financial returns

JEL Classification: C53, E44, G12, C32

Suggested Citation

Guidolin, Massimo and Hyde, Stuart and McMillan, David G. and Ono, Sadayuki, Non-Linear Predictability in Stock and Bond Returns: When and Where is it Exploitable? (January 2009). Federal Reserve Bank of St. Louis Working Paper No. 2008-010A, Available at SSRN: https://ssrn.com/abstract=1126727 or http://dx.doi.org/10.2139/ssrn.1126727

Massimo Guidolin (Contact Author)

Bocconi University, Dept. of Finance ( email )

Via Roentgen, 1
2nd floor
Milan, MI 20136
Italy

Bocconi University - CAREFIN - Centre for Applied Research in Finance

Via Sarfatti 25
Milan, 20136
Italy

Stuart Hyde

Alliance Manchester Business School - University of Manchester ( email )

Booth Street West
Mezzanine Floor, Crawford House
Manchester M15 6PB
United Kingdom
44 (0) 161 275 4017 (Phone)
44 (0) 161 275 4023 (Fax)

David G. McMillan

University of Stirling ( email )

Stirling, Scotland FK9 4LA
United Kingdom

Sadayuki Ono

Hiroshima University ( email )

1-1-89 Higashi-Senda
Naka-ku
Hiroshima, 730-0053
Japan
81-82-542-7009 (Phone)

Do you have negative results from your research you’d like to share?

Paper statistics

Downloads
637
Abstract Views
3,128
Rank
75,838
PlumX Metrics