A Preliminary Study of Stock Picking in the Construction Industry of Taiwan: Efficient Market Hypothesis Revisited

Abstract

One of the trading techniques that become popular among many traders is the technical analysis, which prospers through the first half of the 20th century. In recent years, this method is used less frequently and works only for less liquid securities. However, Aldridge (2009) finds that the technical analysis can still generate profitable trading signals based on intra-day data sampled at hourly intervals. In this study, we utilize VAR, Granger causality test and co integration test to construct models with sound theoretic basis and take the sample at daily intervals. In addition, we adopt an experimental design to exam these models and find supporting evidence for the technical analysis. This study uses High wealth Construction in the Taiwan stock market as a sample case to compare the program trading profits when foreign institutional investors’ (FINI) trading information is used and when domestic industrial information is used. The results show that co integration exists for High wealth Construction’s stock. At the first stage (2007.1.2-2010.8.2), the profit is $29.3, which increases to $35.5 at the second stage (2007.1.2-2013.2.1). In addition, we show that utilizing domestic information, specifically, the historical price of Taiwan 50 and Huaku, rather than the securities lending information of FINI, investors are able to make higher profits. Therefore, this study provides evidence that the construction industry in Taiwan during the period August 2010 and January 2013 does not support the weak-form market efficiency.