We created the system this way because in 2001, you may recall, volatility was off the charts, and it was hard to persuade people to take advantage of the plunge in values and buy stocks. We aimed to be the first rating system to embed a measure of volatility in our rankings. Rather than just focus on typical rating factors such as earnings growth, price momentum, estimate revisions and insider buying, we wanted to help people find stocks that they could hold comfortably. This proved to be an awesome idea and has accounted for much of the success of the system over the years.

Picking surprise winners

There's another secret to the power of the StockScouter system that doesn't directly show up in the ratings: Recognizing that much of an individual company's price movements stem from the popularity of its market-capitalization, industrial sector and the growth or value styles, the system systematically recommends stocks characterized by the greatest number of these "market preferences."

If the system determines the market has exhibited a hankering for large-cap value stocks in the transportation and capital-goods sectors, for instance, then all of the companies chosen for MSN Money's StockScouter portfolio will have market caps of more than $10 billion, be relatively cheap and slow-growing, and do their work in industries like trucking, railroads or defense contracting.

This built-in flexibility, backed by hundreds of hours of testing, has helped StockScouter make many seemingly odd choices over the years that have turned out brilliantly.

As an example, during the bear market years of 2001 and 2002, the system churned out one profitable portfolio after another because it was focused on small-cap regional banks and real estate investment trusts at a time when most people continued to be fixated on the possibility that their rapidly deteriorating large-cap tech stocks would turn around. Those regional banks were huge winners, while most of the techs never recovered.

The strategy of focusing on strong-trending sectors and market-cap groups kicked into high gear during 2003, though, as the strategy produced a 83% gain, more than three times the excellent advance of the broad market.

What you may not recall about that year is how difficult it was to buy stocks in January 2003 amid a steep decline in both public confidence and share values over the looming crisis in Iraq. It's times like that which make a systematic approach so valuable.

The courage to rebalance

Indeed, I'm not worried about explaining how to exploit StockScouter portfolios because I know that most people won't have the discipline to execute the strategy and will kick out stocks with which they are unfamiliar. It's just human nature. If you decide that you want to give it a try anyway, remember that success in the market often demands that you do things that feel uncomfortable. Consider the courage that a re-balance might have required July 1, 2006, during a scary plunge in prices. Many pundits were predicting a return of the bear market. Yet StockScouter's top 10 produced a return of around 16.9%, led by FTD Group (FTD, news)and Claire's Stores (CLE, news)up 30% apiece over the next six months.

Start with the current top 10 and then pick up with the half-yearly re-balance schedule on January 2 or July 2, whichever comes first. Decide how much you are going to invest in the strategy and then divide that amount equally into the top 10 names. On the last trading day of the half-year or thereabouts, sell them all and reinvest the proceeds in the new top 10. Repeat until rich.

Ratings can change fast. Don't worry about that. Every current crop of ratings is good for six months.

Fine print

The 50-stock Scouter portfolio, re-balanced every six months, has returned 16.9% on average per year since inception in 2001, or nearly 6 percentage points less than the 10-stock version of the model.

This column was originally published in February 2007. At that time, Jon Markman did not own or control any shares of companies mentioned.