Stock Return Synchronicity and Technical Trading Rules

Abstract

We explore the potential source of returns from technical trading rules at the firm-level by examining the cross-sectional relationship between technical trading returns and stock return synchronicity. Inspired by Roll (1988) and Morck, Yeung and Yu (2000), we use R-squared of a regression of individual stock returns on the market return as our measure of synchronicity. If a low R-squared is largely attributable to noise trading, stocks will have lower synchronicity with market factors and lower R-squared. Low R-squared stocks earn higher expected returns, according to De Long, Shleifer, Summers, and Waldmann (1989, 1990), or due to limits of arbitrage (Shleifer and Vishny, 1997), a predicted relationship which we termed as the Noise Hypothesis. Overwhelming support in prior literature is in favor of lower expected returns in low R2 stocks, or the Price-Informativeness Hypothesis according to Morck, Yeung and Zarowin (2003). China is our context for investigation; it is the second most synchronous market in the study by Morck et al (2000) which has grown to become the world’s third largest stock market by market capitalization. We find evidence of a negative relationship between returns from technical trading rules and R-squared over 1991-2009, in favor of the Noise Hypothesis. Our results remain robust after controlling for firm-specific characteristics which include market-to-book, size, leverage, dividend payout ratio, turnover and firm age. Thus, an additional simple yet practical statistics - the R-squared - can guide trading decisions using technical trading rules. Do technical trading rules work? Possibly only when the R-squared is low, and for larger and younger stocks with lower turnover. However, sub-period analysis reveal that when there is an improvement in the information environment after the punctuation by an economically significant fundamental shock - the Non-Tradable Share (NTS) reform in China in April 2005 - technical analysis work better post-NTS reform during 2005-09 for stocks with higher R-squared, consistent with the Price-Informativeness Hypothesis, and generally for older and bigger stocks with lower turnover and higher market-to-book ratio. Thus, without the guide of R-squared, investors should take the market prognosis by all these “alchemists” with their “voodoo” charts with a heavy dose of salt. We also reconciled the lively debate and extremely mixed evidence on the interpretation of R-squared and its relationship with the cross-sectional returns of stocks.

Keywords: Stock return synchronicity, R-squared, idiosyncratic volatility, technical trading rules, price informativeness, noise

JEL Classification: G11, G12, G14, G15, C1, O16, O53