Before diving into today’s post, I must make it clear that I am not an anti-value investor. I started my career as a value investor and I still think that the value factor is real and has historically been great. In fact, I believe that value and momentum are the two most reliable factors that investors should try to exploit for superior returns. However, the world is changing, and over the years I have come to believe that the value factor has decreased in size over time and is probably much smaller today than it was 30 or 50 years ago, something I have analyzed. here.
And now, Mathias Hasler of Boston College has presented some findings that make me wonder if the value factor was historically as large as we first thought. Returned to the original publication of Fama and French which systematically described and popularized the value factor as the difference in performance between the 30% stocks with the highest book-to-market versus the 30% stocks with the lowest book-to-market ratio.
Hasler noted that to calculate their results, Fama and French had to make several seemingly innocuous decisions about how to calculate the book-to-market ratio and how to classify companies into deciles. He then changed these options slightly to create 96 alternative ways to calculate the original value factor and see how these alternatives compare to Fama and French’s results:
In terms of the reported book value, Fama and French used the reported book value at the end of a company’s fiscal year, subject to the restriction that the fiscal year ended at least 6 months earlier to ensure that the data was available to the public. in real time. As an alternative, Hasler chose the book value of a company reported in the last intermediate or final results as long as that result was at least six months earlier. So, in the case of Fama and French, the book value used could be six months at times and up to 17 months if the company’s fiscal year coincidentally ended in the month after the formation of their portfolios. In the case of Hasler’s alternate specification, the book value was typically 6-9 months because it also supported quarterly results.
In terms of market value, Fama and French used the market value of the stock at the end of the last fiscal year (same as the book value), while Hasler alternatively used the most recent market value of the company or the value of market that is a month. old.
Fama and French excluded from their study all companies with a negative book value, while Hasler alternatively allowed companies with a negative book value to be included.
Fama and French included finance companies, but in later studies on the profitability factor they excluded them because finance companies (especially banks) have much higher financial leverage than other companies due to the nature of their business and therefore have much higher financial leverage than other companies. a systematically distorted book-market relationship. . Hasler alternatively excluded finance companies from the calculations.
Finally, Fama and French used the top and bottom 30% of companies each year to build their portfolios, while Hasler allowed this threshold to vary to 20% or 40%.
Different combinations of all these original and alternative ways of calculating the original value factor create 96 different results. The graph below shows the monthly excess return of value stocks versus growth for the period originally investigated by Fama and French from 1963 to 1991. The dot marked ‘HML’ shows the original result for Fama and French, while the AHML point market is the average value premium of the 96 different specifications.
The value premium in the original specification compared to other specifications
Source: Hasler (2021)
The difference between the original value premium and the average of the alternative specifications is 0.09% per month or about 1.1% per year. This difference is both statistically and economically significant and indicates that by sheer chance, Fama and French may have chosen a specification that accidentally ended up giving one of the largest value premiums they could have found. To be clear here, I do not accuse Fama and French of any crime or p-piracy. When we do research, we all make tons of these decisions, and while we are careful to make sensible decisions and check for potential distortions, no researcher can verify all possibilities or simply stop considering them.
Hasler’s off-sample tests can also see that the historical value premium may have been reported as too high. If you examine the value premium for the time period after Fama and French’s original publication, you find that the average value premium is statistically indistinguishable from zero. This indicates that the original Fama and French post may have been biased upwards by chance and that the decline in the value factor that we have seen over the last 30 years was simply a return to more normal premiums that were observable all the time if we had measured in alternative ways.