Successful investors and traders who pursue market-beating returns in sector and other specialized funds have learned how to exploiting trends. In doing so, most of them had to shed some very natural but wrong-headed notions about how stock prices behave and, by extension, how stock fund values behave.

From childhood, we're raised on pat truths such as what goes up must come down. That maxim seems to fit our experience in the physical world. A ball tossed in midair, of course, comes back to earth. And we seem to see the same thing outside the world of gravity. The cost of gasoline spikes higher at times, then falls back, only to rise again. Hemlines rise and fall with fashions. Even our moods and personal fortunes seem to sway between good times and bad. 

And when we watch the day-to-day volatility of the stock market, it's easy enough to apply this simple notion to stock prices.

But in fact, while price movements in one direction don't continue forever, they can continue for longer than you suspect. And this leads to a very important statistical fact.

The more a stock or sector outperforms the market, the more likely it will continue to do so. Success usually begets more success. 

A number of studies bear this out. In one, William O'Neil & Co. looked at thousands of the biggest winning stocks over a 40-year period. The average stock in this elite group had already appreciated more than 87% of its peers over the prior year just before beginning its big advance.

Market analysts coined the term "relative strength" for this concept. Relative strength is simply a measure of a stock or fund or sector's share price performance relative to its peers. 

The corollary exists on the downside as well. The more a stock or sector underperforms, the more likely it will lag its peers the future.

Fund manager James O'Shaughnessy demonstrated both sides of the relative strength equation in his landmark book on indexing strategies, What Works on Wall Street (McGraw-Hill). Using the CompuStat database, O'Shaughnessy ran simulations of a variety of indexing strategies applied across a 45-year time span. One of the most powerful selection criteria turned out to be relative strength. 

One simulation created a portfolio of the 50 stocks with the highest one-year relative strength ratings. The portfolio held those stocks until rebalancing the following year. This approach produced an average annual return of 18.1% vs. 15.1% for O'Shaughnessy's all-stocks universe.

Another simulation looked at the approach of investing in the worst-performing stocks over the prior year, then holding until rebalancing on an annual basis. That portfolio returned 6.7% a year on average. Let that serve as a warning to bottom fishers!

There's nothing really mysterious about why price trends tend to be self-reinforcing. When buyers begin to bid up a market, or a sector, or an individual stock, it catches the attention of other potential buyers. These traders don't want to be left behind, so they, too, begin buying, which in turn attracts even more longs. This buying momentum begets more buying momentum, as greed pushes an increasing number of traders into the market. The effect is especially pronounced when a market or a market, sector or stock is making a series of new highs.

The same self-feeding phenomenon occurs on the downside as well. The more a stock, sector or index heads south, the more the affected shareholders fell pain and fear. As more shareholders sell, the stock, sector or index falls further under the same selling pressure, turning still more shareholders into sellers.

Obviously, trends cannot last forever, but they often continue for much longer than many people expect, making top and bottom picking difficult, if not impossible. Trending markets repeatedly pause or pullback, leading traders to think the trend has ended-only to be trapped when the market continues in its previous direction. As a result, it's much more profitable (and easier) to go with the flow-with the trend-than it is to fight the market.

Dave Landry contributed to this tutorial.