What if it never happened?

Finance suffers from a replication crisis and it is well known that once a market anomaly has been described in the academic literature it tends to disappear in financial markets. But what if factor performance never disappeared? What if it is just hiding behind a corner? That at least is what a study from CUNY says.

Typically, when academics describe a market anomaly like the value, momentum, or size effect, they form portfolios based on some past observations, then let the portfolio run for a month and do the entire exercise again. The hope is that this monthly rolling process will systematically create outperformance over the market as time passes.

Unfortunately, with the proliferation of smart beta investing and other quantitative methods in finance, factors that are described in the literature are increasingly exploited by systematic investors to the extent that these factors seem to disappear. The chart below shows the average performance of 260 factors described in the literature in the US stock market over different time periods. Looking at the green bars shows that since 2009, the average outperformance of these factors has almost halved from about 0.6% to 0.35%.

Factor performance in US equity markets

Source: Dong and Yang (2023)

Something interesting happens if one looks at the average factor performance in months 2 to 24 as is done with the purple diamonds in the chart. Essentially, instead of reconstituting factor portfolios every month, the researchers let each factor portfolio run for two years and then did the exercise again. Leaving portfolios untouched shows no decay of outperformance after 2009.

The key to understanding what is going on here seems to be the behaviour of retail investors. Retail investors display some well-known biases, and they tend to trade with some anomalies (for example, retail traders trade with experts when it comes to momentum stocks) but they tend to trade against some other anomalies. For example, retail investors prefer growth stocks over value stocks, low quality, unprofitable stocks over high quality, profitable companies or they prefer lottery stocks with an unlikely but very large positive return vs. more likely small negative returns. In total, of the 260 anomalies examined in the study, retail investors trade against some 60% of the anomalies and with about 40% of the anomalies.

Look at the outperformance of anomalies where retail traders trade on the same side as the professionals. The decline in outperformance is much more pronounced and even over the next 24 months, the outperformance completely disappears in the years after 2009.

Factor performance in US equity markets when retail investors trade with anomaly returns

Source: Dong and Yang (2023)

Compare this with the anomalies where retail investors trade against the anomaly.

Factor performance in US equity markets when retail investors trade against anomaly returns

Source: Dong and Yang (2023)

Here, the size of the outperformance remains more stable in the first month and even increases in the months 2 to 24.

This indicates that it is the growing influence of retail traders that influences the performance of systematic factors. When retail investors trade with professionals, they quickly destroy the factor performance because the combined flows of systematic traders and retail investors lead to a rapid adjustment of share prices. But when retail investors trade against systematic factors, the flows of retail investors push against the factor, share prices correct much slower and mispricing survives for longer. And that means that patient factor investors can still exploit these factors if they leave their factor portfolios untouched for longer.