George Athanassakos is a professor of finance and holds the Ben Graham Chair in Value Investing at the Richard Ivey School of Business, University of Western Ontario.
I know what I report in this article will not make value investors particularly happy, but it will definitely make growth investors downright depressed.
There is plenty and unequivocal evidence from academic research in Canada and around the world that, on average, value stocks (defined as stocks with low price-to-earnings or price-to-book ratios) beat growth stocks (those with a high P/E or P/B). For example, in recent research I carried out using U.S. data, I found that, on average, value stocks beat growth stocks by about 6 per cent over the 1982-2013 period. The results are similar in Canada and in international markets. But do value stocks really deserve an award for being such an outstandingly performing group of stocks? My research suggests not, at least the way academics define value stocks. This is because value stocks beat growth not because value stocks produce an outstanding long-run performance, but rather because growth stocks earn terrible long-term returns.
The P/E (or P/B) multiple is a function of the growth rate of earnings going forward. Companies have low multiples because markets expect low earnings growth. Companies have high multiples because markets expect high earnings growth. The markets tend to be overoptimistic about growth for high-multiple firms and overpessimistic about growth for low-multiple firms. As a result, in theory, investors bid up (overvalue) high-multiple firms and bid down low-multiple firms.
What does the evidence show? Researchers at the Darden School of Business, rather than examining firms based on sorting by P/E or P/B (a proxy for growth expectations), looked at the actual asset growth of U.S. firms over a 40-year period and compared the stock performance of high-growth firms with that of low-growth firms. They found that low-growth firms outperformed the high-growth firms by a whopping 22 per cent, on average per year, over the four decades. But that was primarily because high growth firms experienced very low returns: These stocks tend to attract a lot of attention and a lot of trading by investors, and thus tend to become overvalued, leading to lower returns going forward.
This is consistent with findings in my recent research, which looked at this question from the P/B angle. I examined one year ahead buy-and-hold returns for value (low P/B quintile) and growth (high P/B quintile) stocks across different earnings quality quintiles, defined by net income volatility over a five-year period – the higher the volatility, the lower the earnings quality. I found that while a value premium (defined as the difference between value and growth stock returns) was evident in the total sample (6 per cent), the premium appears to be driven primarily by firms in poorer earnings quality quintiles.
The intuition for this finding is that as growth stocks, on average, tend to be bid up by investors, the less visible ones – which tend also to have poorer earnings quality – are bid up the most and end up having lower returns than the better quality growth stocks. For value stocks, there is no evidence that poor earnings quality exerts a discernible effect.
Having said that, however, there is a silver lining to the story as far as value investors are concerned. It has to do with the fact that academics do not look at (as they do not know) the actual stocks that value investors buy, they only look at stocks value investors consider a potential buy. Value investors sort stocks by P/E or P/B to find potentially undervalued stocks. They then value these stocks to determine their intrinsic value and only then make a decision using the concept of margin of safety.
Value stocks as defined by academics may not be worth buying if they do not meet that margin of safety requirement. In other words, the stocks actually chosen by value investors may have a great performance even if, on average, low P/E or P/B stocks do not.