In efficient markets, stock prices should reflect all available information, leaving few opportunities for investors to earn excess returns. However, over the past 50 years academics have uncovered dozens of “anomalies,” or predictable patterns in returns. For example, simple portfolios based on asset growth, profitability, or changes in accruals have historically earned high risk-adjusted returns. While the existence of so many anomalies calls into question the efficiency of the stock market, some researchers question the validity of the anomalies themselves, as many of the portfolios that earned high returns in the 1980s, 1990s, and 2000s have failed to produce high returns in recent years. In a new study forthcoming in the Journal of Finance, Boone Bowles, an assistant professor of finance at Mays Business School, and his colleagues carefully focus on the timing of important information releases to examine the changes in anomaly returns over time. Their research provides some intriguing results.

Decoding Anomalies: Patterns That Challenge Market Efficiency

Anomalies refer to patterns in stock returns that contradict theories of efficient markets. These patterns suggest that investors could historically earn excess returns by constructing portfolios based on simple characteristics. The fact that hundreds of anomalies have been documented creates a predicament for academics: are these findings reliable, or just spurious results that found their way into top journals because of publication biases? Practitioners also care whether trading on anomalies can really generate durable excess returns.

To answer these questions, this study makes a key innovation: anomaly portfolios should be constructed based on the most up-to-date information possible. The previous convention in the literature was to construct portfolios annually, typically at the end June.  However, most anomaly-relevant information is released in February and March. Thus, Dr. Bowles and his colleagues consider rebalancing anomaly portfolios during these months, precisely as new information is released.

Using these up-to-date anomaly portfolios, Dr. Bowles and his team then examine how quickly anomaly returns decay after new data is released. Consistent with theories of costly information processing, they find that anomaly returns are largest immediately after data releases, and then deteriorate as arbitrageurs trade on the new information. For example, the asset turnover anomaly generates returns that are almost three times higher in the first month after information is released compared to a few months later.

Implications for Academia and Practice

This has important implications both for academics testing anomalies and practitioners trading on them. First, it suggests many anomalies are real and not just the result of data mining. Further, their temporary nature, and overall diminishment in recent years, is consistent with theories of costly information processing. Second, the common academic practice of forming portfolios in June results in forming portfolios long after data is released. This practice underestimates the magnitude of many effects and completely hides the dynamics of returns in the weeks following information releases.

For practitioners, the results demonstrate the importance of rapid trading close to new data releases. Anomaly returns are often short-lived, so continuously monitoring data releases and quickly trading is imperative. This can be challenging for individual investors, suggesting that sophisticated quantitative funds may be best positioned to profit from anomalies.

Key Findings

The study provides guidance for both academics and practitioners in unlocking the potential excess returns from anomalies. Key implications include:

• Form portfolios right after new data is released to accurately gauge the magnitude of anomaly returns.
• Anomaly returns demonstrate real market inefficiencies, not just data mining biases.
• However, returns decay quickly as arbitrageurs trade on new information.
• For practitioners, continuously monitor data releases and rapidly trade to capture the anomaly premium.

Concluding Insights: The Race Against Time

While markets have become more efficient over time, some anomalies still offer potential excess returns to savvy investors. But capturing these returns requires identifying and trading quickly after new information releases. As information processing becomes less costly, the lifespan of these anomalies is likely to shorten further. By closely monitoring information flows and acting swiftly, academics and practitioners alike can continue uncovering market inefficiencies.