Efficiency or statistical illusion? The case for the Japanese Stock Market

Abstract

The recent proliferation of hedge funds suggests that capital markets present windows of opportunity to realise substantial arbitrage profits thus violating the 'no arbitrage’ condition of efficient markets. This thesis examines several observed return patterns that have raised questions about the efficiency of capital markets and/or the validity of the asset pricing models used to analyse them. The study focuses on the Japanese stock market which is under-analysed despite being the second largest in the world. We first look at three stock attributes that can arguably differentiate between future winners and losers. These are size, price and book value to market. In contrast to older studies, we find no significant evidence of a size effect. The price and book value to market effects however are statistically significant although both appear to be cyclical in nature suggesting that they are at least partially driven by macroeconomic risk factors and so are not pure anomalies. The short term reversal of stock returns is investigated next. Unlike previous studies, a strategy that utilises optimal investment portfolios is simulated. By avoiding previously documented methodological problems, it is shown that contrarian profits are statistically and economically significant and that they are overwhelmingly attributed to investor over-reaction to firm-specific events, implying that significant short term inefficiencies occur in the Japanese stock market. Finally the effectiveness of the law of one price and its implications for the relative pricing of assets is examined. It is shown that the returns of securities with similar systematic risk are highly correlated and their relative prices oscillate around an equilibrium value. Large deviations from that value can be exploited by a trading strategy known as pairs trading. Simulations of several strategy variants generate statistically and economically significant profits which are not attributable to systematic risk. It is concluded that relative stock prices are not always efficient in the short term. Such inefficiencies can be profitably exploited as prices are eventually driven to equilibrium.