Momentum refers to the market anomaly where stocks that have had the greatest return continue to outperform those with weak returns. One of the most famous papers on the topic was written by Narasimhan Jegadeesh and Sheridan Titman (’93). In their paper, ”Returns to Buying Winners and Selling Losers: Implications for Stock Market Efficiency”, the authors estimated that the excess return for buying past winners vs. losers amounted to 1% a month. To construct the portfolio, the authors looked at stock return over the past 12 months, excluding prior month returns.

Why is Momentum Considered an Anomaly?

Under the Efficient Market Hypothesis, investors are expected to behave rationally and all new information is expected to be priced into the market instantly. If markets are efficient, we should not see momentum. A number of papers have explored aspects of momentum. Some have argued that the anomaly has weakened in recent years, but most studies acknowledge that momentum effects continue.

What Causes Momentum?

Many behavioral economists believe that momentum is caused by cognitive biases. A cognitive bias is an error in reasoning. Conservatism is the cognitive bias where original beliefs are over-weighted even in the face of new information. For example, the buyer of a stock may have a current view on a company’s prospects and earnings. If the company exceeds earnings expectations the stock market participant may focus on their original views and may not give enough weight to new information. This would cause the market participants to under-react to new information. The view that momentum is the result of cognitive biases has come under recent scrutiny. For example, one recent paper argues that in the face of noisy market information, momentum effects may be deemed to be rational.

Is There Any Alternative Explanation?

A recent paper by Robert Novy-Marx sheds some interesting light on momentum. In his paper titled, ” Fundamentally, Momentum is Fundamental Momentum”, Novy-Marx explores the relationship between price momentum and earnings momentum.

To test the relationship, Novy-Marx looked at excess earnings as measured by SUE and CAR3. Without getting too far into the details, SUE is a measurement of the companies earning surprise relative to analyst’s expected returns. CAR3 is a measure of excess returns over 3 days around the stock’s earning announcement date.

In the paper, Novy-Marx first showed that the momentum effect, after accounting for company size, market value of equity to book value of equity and gross profitability, results in .6% monthly excess returns. When he used regression analysis of returns and included the effects of SUE and CAR3, the excess returns were reduced to .15% and were not statistically significant

The result is very interesting. What it says is that price momentum in stocks is almost entirely explained by earnings surprises and the 3 day reaction to those returns.

So is Momentum Really an Anomaly?

The efficient market hypothesis states that new information is reflected in the price of stocks very quickly. Before we can judge whether an effect is an ‘anomaly’ we need to judge the effect against some model. Typically, the aggregate stock market return is used to judge risk and return. However, a number of researches have argued that using market return is not necessarily the right model to measure against. The efficient market hypothesis doesn’t imply that surprises in earnings should be priced into the market in advance. Given the Novy-Marx paper there is an interesting argument to be made that momentum is driven by real changes in information and therefore it doesn’t directly violate the efficient market hypothesis.

What is the Practical Implication of This?

Aside from the theoretical discussion of market efficiency, it seems to me that the momentum effect may be the result of a combination of behavioral factors as well as new information coming into the market. When trading individual stocks, having awareness for the momentum of the stock and what causes this momentum gives us another piece of information we can use in a mosaic of information that may help investors to make timing decisions on buying and selling securities. If nothing else, I’d take a close look at earnings releases vs. expectations and dig further into earnings prospects for those stocks with extremely high (or low) momentum.