How is momentum justified as a common risk factor?

Question

Momentum as a common risk factor?

This question is partly a follow-up to another question found here. In this other question it was noted in momentum is difficult to explain as a common risk factor in factor pricing models like the intertemportal capital asset pricing model (I-CAPM) or arbitrage pricing theory (APT). In these models, it is assumed that exposure to one of these factors represent exposure to some sort of undesirable risk. In this question, I'm trying to understand how to interpret exposure to momentum as exposure to some form of common risk. In particular, I'd like to know

1. Who was the firm to include momentum as a risk factor? What was the explanation?
2. It seems like momentum is often attributed to behavioral over- or under-reaction. (This could be irrational or maybe even rational-overreaction, I suppose---right?) Is there an interpretation that rationalizes momentum? (I mean one that gives an explanation where exposure to momentum is a bad thing.)

For some reference:

Jegadeesh and Titman (1993) review some explanations of momentum, including overreaction to information, relation to size effect and systematic risk, short-term price pressure, lack of liquidity, delayed stock price reaction to common factors

The paper argues that the "relative strengths" premium (the strategy of buying past winners) is not due to exposure to systematic risk, cannot be attributed to "lead-lag effects that result from delayed stock price reactions to common factors," but that the evidence seems to be consistent with delayed price reactions to firm-specific information.

Stocks in the winners portfolio realize significantly higher returns than the stocks in the losers portfolio around the quarterly earnings announcements that are made in the first few months following the formation date. However, the announcement date returns in the 8 to 20 months following the formation date are significantly higher for the stocks in the losers portfolio than for the stocks in the winners portfolio.

The evidence of initial positive and later negative relative strength returns suggests that common interpretations of return reversals as evidence of overreactlon and return persistence (i.e., past winners achieving positive returns in the future) as evidence of underreaction are probably overly simplistic.

Answer 1

The Jegadeesh and Titman (1993) paper is usually considered Original Source, though I'm sure you could find something earlier that looks similar if you looked hard enough.

I don't think there is a satisfactory explanation. Momentum is not correlated with macroeconomic variables, it does not seem to reflect persistent exposure to other (known) sources of risk, and is driven almost entirely by the 7 to 12 months before stocks are chosen as part of a momentum portfolio. This all makes it extremely difficult to put a solid theoretical model to it that doesn't have a behavioral aspect.

Recently (mid 2009), there was a momentum crash in the market. Daniel and Moskowitz argue that momentum has rare "panics" periods like this, where crashes occur for momentum. So during normal periods investors may be being compensated for this risk. The question becomes "what are these crashes caused by?" The authors make the case that following a period of declining stock values, momentum portfolios will be long a lot of low market beta stocks, and short a lot of high beta stocks. If there is a sudden rebound in the market, momentum strategies will experience a crash. So this is at least suggestive of a direction for more formal modeling, but they ultimately suggest a behavioral interpretation may be necessary as well.

If you have not seen it, Jegadeesh and Titman (2011) review the evidence for momentum in more detail, and explore behavioral explanations for momentum, of which I have little knowledge.

Answer 2

Just to add to the discussion of momentum, there is a new paper by Novy Marx on this topic: http://rnm.simon.rochester.edu/research/FMFM.pdf. In the paper, he argues that the price momentum that we observe is actually earning momentum. So this pushes the puzzle to a more fundamental variable rather than just pure price. However, earning momentum itself is also hard to justify under efficient market hypothesis. The underreaction story seems to be very compelling.