One of the quandaries of finance research has been why the value premium (the tendency for low P/B or P/CF or P/E stocks to outperform the market) has not been reflected in the performance of ‘value’ active mutual funds. A new paper by Ludavic Phalippou in the Financial Analysts’ Journal argues that the reason for this is that only the small-caps and micro-caps exhibit a large value premium. What this means is that an investor should either focus on buying small cap value index funds or should buy individual micro-cap value stocks.
The article is not available free online. Following is an excerpt from the paper’s conclusion:
The premise of this article is that if the value premium is a result of both pricing errors and limited arbitrage, then the value premium should be concentrated in stocks that are both held by relatively less sophisticated investors and expensive to arbitrage. Such a concentration is suggested in the literature but has not been quantified. In this article, I show that, indeed, at least 93 percent of market capitalization is free of a value premium. Using institutional ownership (IO) as a parsimonious way to classify stocks by their mispricing likelihood, I show that the value premium monotonically decreases from a high 185 bps for low-IO stocks to a negligible 13 bps for high-IO stocks. This result also holds when returns are value weighted and, importantly, is driven mainly by the long side. Low-IO value stocks are those with the most abnormal returns. The anomaly is a value premium, not a growth discount, as is sometimes argued …
The extreme concentration of the value premium has important practical implications. First, arbitrageurs can expect to face substantial costs when trying to arbitrage the value premium, and those focusing on the stocks most held by institutional investors (the larger, more liquid stocks) will have difficulties generating arbitrage profits. The value premium concentrates where arbitrageurs usually do not go. This reason is also why studies have found that value and growth mutual funds perform the same. Second, studies that select a subsample of stocks that, for instance, either have at least two to five analysts following the stocks or are traded on the NYSE end up with a sample that is almost free of the value anomaly. Such a fact is important to bear in mind when interpreting the results found in such samples.