Data-Mining Bootstrap Procedure with Potential Predictors in Forecasting Models: Evidence from Eight Countries in the Asia-Pacific Stock Markets

Abstract

We use a data-mining bootstrap procedure to investigate the predictability test in the eight Asia-Pacific regional stock markets using in-sample and out-of-sample forecasting models. We address ourselves to the data-mining bias issues by using the data-mining bootstrap procedure proposed by Inoue and Kilian and applied to the US stock market data by Rapach and Wohar. The empirical findings show that stock returns are predictable not only in-sample but out-of-sample in Hong Kong, Malaysia, Singapore, and Korea with a few exceptions for some forecasting horizons. However, we find some significant disparity between in-sample and out-of-sample predictability in the Korean stock market. For Hong Kong, Malaysia, and Singapore, stock returns have predictable components both in-sample and out-of-sample. For the US, Australia, and Canada, we do not find any evidence of return predictability in-sample and out-of-sample with a few exceptions. For Japan, stock returns have a predictable component with price-earnings ratio as a forecasting variable for some out-of-sample forecasting horizons.

Keywords: Data-Mining, Bootstrap, In-Sample Predictability, Out-of-Sample Predictability, Forecasting

JEL Classification: C13, C14, C15, C22