Conclusions
We set out to learn what had happened to the small-beats-large anomaly in the days since it was first revealed by Rolf Banz and discovered that it may never have existed at all. If anything, it appears that large stocks beat small. Our findings are counter to those of earlier researchers.
Four factors that may explain the difference in our findings: First, we used the CRSP 9-10 index to represent small stocks and the S&P 500 index to represent large stocks. The earlier researchers constructed their own indexes using individual securities data from CRSP.
Another factor is that the CRSP data is constantly revised. We assume it is revised to make it more accurate, so if this accounts for the differences, we would expect our findings to be more accurate too.
Third, we used annualized returns in our analysis. Again, we believe this is a better approach because it is more relevant to investors and it avoids overstating the benefits of the more volatile asset class.
Finally, our conclusions rely, in part, on our use of "lag-adjusted" monthly returns. Since Fama and French also used them, such returns should not account for the difference in our findings and theirs. However, Banz used raw monthly returns, and this may explain differences in our findings and his.
If the small-beats-large anomaly relies for its existence on the use of certain indexes or the use of monthly instead of annualized returns, it is a highly illusory concept, at best. An advantage either exists or it doesn't. Investors should not put their money at risk using a mere mathematical chimera.
Our research suggests that it is time to rethink the idea that small stocks outperform large stocks on a risk-adjusted basis. On the contrary, it appears there may actually be a large company stock advantage.