Abstract
In this article, we study whether the behaviour of oil prices can be used as a reliable predictor for the disaggregated industry-level stock market indices. We find strong evidence for the relevance of changes in oil price as a predictor for the returns of UK industry portfolios, while this relevance is heterogeneous across industries. In an out-of-sample framework, we find that both the contemporaneous and lagged oil price changes do predict UK industry stock market returns. The predictive power is more transient for the latter case, and mostly appearing after allowing for time variation in the relative performance. In addition, we find some evidence of asymmetry in the oil–stock price relationships.
Acknowledgement
We thank the editor and an anonymous referee their helpful comments and suggestions in improving the overall quality of this article.
Notes
1 Note that the performance of the out-of-sample forecasts can differ on how a given data set is split into estimation and forecasting periods; we use various window sizes to test the robustness of the results.
2 We use the FTSE All-Share Index as it represents 98–99% of UK market capitalization; it is the aggregate of the FTSE100, FTSE250 and FTSE Small Cap indices. FTSE calculates the disaggregate-level indices followed by the Industry Classification Benchmark (ICB) in a 4-tier hierarchy (10 industries, 18 super-sectors, 39 sectors and 104 subsectors). We use 10 FTSE All-Share industry indices. We use both UK pound and US dollar returns, but as results tend to be similar we only report results for the local currency returns.
3 It is according to which the returns are forecasted to be zero.
4 To save space, we did not report these results.
5 Giacomini and Rossi (Citation2010) provide critical value for various significance levels, window and sample sizes.