Since we’ve released our new book, Quantitative Momentum, we’ve received a handful of basic questions related to momentum–specifically as it relates to stock selection. At this point, the so-called “momentum effect” has occupied academic researchers for several decades. Researchers have found that, on average, stocks with strong recent performance relative to other stocks in the cross section of returns tend to outperform in the future (see Levy 1967 for an old example and JT 1993 for a newer version). The effect has been well-documented by numerous follow-on researchers and the theory of “why” momentum works has been extensively explored (although we still don’t completely understand why it works). So if an investor wants to harness momentum and implement it in the real world, a common question arises:
What is the best way to measure momentum for stock picking purposes?The academic research response is to focus on so-called, “12_2 momentum,” which measures the total return to a stock over the past twelve months, but ignores the previous month. (e.g., Ken French data) But why use 12_2 momentum? Why shouldn’t we use the 3-month momentum, or the 6-month momentum? Why 12-months? And why drop the most recent month’s returns? Let’s take these questions one at a time.