As you probably know, most of SMI's investing strategies take advantage of the observable market phenomenon known as momentum.

Momentum — or "performance persistence" — is just what it sounds like. It means an investment that has been performing well in the recent past tends to keep performing well into the near future.

(There's also negative momentum. If an investment hasn't been a good performer lately, that not-so-good performance is likely to persist too.)

A couple of weeks ago, Corey Hoffstein and Justin Sibears at Newfound Research published a study aimed at trying to discover the "optimal holding length" for a momentum strategy. To put it in form of a question, "How long will it be, on average, before returns hit their peak and start to decay?"

If you're academically inclined, you can click the link above and wade into the details, but the bottom line is this — as summarized by Mark Hulbert at MarketWatch:

[Hoffstein and Sibears] found that the optimum length is a function of both the period over which momentum stocks are identified and the length of time you hold the stocks that have been identified. The sum of those two periods should be between 12 and 18 months.... (emphasis added)

In other words, the longer it takes identify momentum, the shorter your holding period is likely to be before momentum wanes.

The SMI approach

By its very nature, it is impossible to identify momentum right away. It takes time for a stock or a fund to "break out," just as it takes a little while to notice that a baseball player is on a genuine hitting streak.

For SMI's Upgrading strategy, we examine past performance (for mutual funds) over three time periods — the most recent 3 months, the most recent 6 months, and the most recent 12 months. We give the 3-month figure more weight than the 6-month figure and the 6-month figure more weight than the 12-month figure.

This unequal weighting approach reflects our belief that the more current a set of performance data is, the more relevant it is. By giving more weight to the most recent three months, we're able to identify momentum more quickly than if we equally weighted a fund's performance across the past 12 months. At the same time, including more than just three months helps to eliminate the "false positives" that could occur from looking exclusively at a short time period. 

The wisdom of our approach appears to be backed up by the Hoffstein and Siebears research (and another study [PDF] which they reference published in May by HIMCO). On average, our Upgrading funds are held about nine months, which when added to the "identify-the-momentum" period, puts us squarely in that 12-to-18 month sweet spot that minimizes fund turnover and uses momentum to its maximum (or close-to-maximum) advantage.

Market observers still are trying to figure out what causes momentum and why it works as it does. Perhaps someday they will. In the meantime, we can just be glad that it does work — most of the time. As we have stressed before, a momentum approach can go through periods of underperformance, especially when trends are not apparent. Even so, the overall performance evidence for momentum-based investing is compelling.

And our latest Upgrading report card is certainly encouraging.