Price momentum is a well known "market anomaly",defying efficient market theorists. In general, the effect is stocks that have been outperforming in the past tend to continue to do so, and vice versa. Whether there is some information leakage that allows the better informed to bid up stocks over time, or just a variant on the "greater fool theory", buy high in hopes to sell higher, price momentum [PM] is an empirically justified market phenomenon (at least without trading costs!).

Yet, PM doesn't work the same in all markets and all months. I ran a simple analysis using trailing 6 month performance PM measure, ranking the largest 1500 stocks monthly over the last 25 years into deciles, with the best performers ranked 1, worst 10.

There are better PM measures than just T6M, but it does encapsulate the effect. Over the period, the top ranked stocks outperformed the worst ranked by about 1% per month (again without trading costs, which can substantially diminish returns). I further looked at the PM spread results in up/down markets and by month of the year.

In up markets, which seem kinda scarce around here these days, the average monthly spread was only 40bps, and with the variability, that positive spread wasn't statistically significant. However, down markets are where at least this measure of PM really shined, averaging nearly a 2% monthly best/worst spread consistently enough to be significant. So, in falling markets, investors will dump losers more so than relative winners - no big surprise.

Next, I looked at month of the year, and again the results varied considerably. The net effect is that PM doesn't work too well from Jan - May, generating no positive spread, but from June-Sep the strategy blazes, in up and down markets, averaging a 3% monthly spread! Then things reverse course and PM fails in Oct & Nov, only to revive in Dec, and then back to failure the first part of the next year.

Sure, it would be nice if the strategy worked consistently every month, but it doesn't, though it does show this temporal consistency. The guess is that institutional investors start looking at their portfolios relative to their benchmarks and peers about mid-year. Those that are lagging have to start playing "catch-up", and so they load up with winning stocks (good PM), and dump their losers (bad PM). This gets the spread going in earnest. Near the end of their fiscal years, usually Oct and Nov, they swap out of the winners, after grabbing some of the excess returns and move back to the "cheap" names their research suggest are good values, and not driven by hype and fools, only to see the effect come back in December, due to tax loss selling. Then the cycle begins anew the next year.

Now given the variability of the effect, PM can be a tough factor to implement. Yet, being aware of PM, on margin, might help when deciding to make an investment. During this time of the year though, you are going against the odds if you buy stocks that have lagged the market over the last 6 months. It appears that at least until October, now is not the time to be a contrarian.

• Have We Learned Anything Yet?  »
• Thought Experiment: Busting the Euro  »
• Big Troubles for the Euro  »
• Asset Securitization Crisis: The Butterfly Effect   »
• TARP: Alternative Acronyms Edition  »

Dan Knight

• Bio & more articles
• Visit author site