The Fama-French 'three-factor model' as it came to be known, and the subsequent four-factor model that included momentum, weren't really economic theories. They were exercises in data mining, with dubious explanations tacked on after the fact. What's more, they were exercises in data mining that revealed several time-honored Wall Street strategies--dismissed by finance scholars since the 1960s as folklore or worse--to be consistent money makers.
I delve deeper into the whence and why of the factors. It isn't obvious these are data mined because they are prominently used by academics and practitioners for benchmarking, so either everyone's an idiot, or we have some explaining to do. I go over the anomalie soup that gave rise to size and value. How the size effect was initially >15%, now <1%. What did they initially represent? Financial distress. That didn't work. Measure distress independently and you get a very strong, and anomalous negative relationship. Hmmm. Fama and French's factors--the market, size, and value--are the most dominant risk factors applied outside the Capital Asset pricing Model's singular 'beta' model. They are ubiquitous in the academic literature and also in fund style guides.
See this brief (2:37 second) discussion of how size was essential in uncovering that CAPM betas does not work.
see my longer videos for more, or buy the book.
1 comment:
Can one conceptually reduce the three and four factor models to ABT? If so, are not F&F models are data mining exercises?
Post a Comment