These three funds have put up impressive numbers by using computer models to pick stocks. By Andrew Tanzer, Senior Associate Editor December 18, 2006 Human emotion and behavior are too often the enemy of sound investment hygiene. You probably should be buying when the herd is selling and selling when the herd is buying. Many investors fall deeply in love with the stocks they own. Fund managers and Wall Street security analysts often behave like cheerleaders for the companies they're supposed to be dispassionately analyzing.All this emotion partly explains the growing interest in so-called quantitative funds, mutual funds largely managed by soulless computers that crunch the numbers. Here are three quant funds, each operating on different principles, that have posted good results. Charles Schwab has been making a name for itself in quant funds such as Schwab Core Equity (symbol SWANX), which beat Standard Poor's 500-stock index by an average of two percentage points over the past five years through mid December. Co-manager Jeff Mortimer says that Schwab plugs 18 factors, such as short-selling interest and stock buybacks, into the computer. Out pops a portfolio that includes Hewlett Packard, American Express and Lockheed Martin. Mortimer says the computer model ignores charts and technical analysis and "looks purely at the numbers." Peter Hill, portfolio manager of HighMark International Opportunities (HIOMX), says quant-driven fund management is particularly well suited to investing internationally. Market inefficiencies are greater outside the U.S.; the quantity and quality of numbers from overseas have improved dramatically in recent years; and the sheer immensity of global markets makes it difficult for a team of humans to keep up. Hill says that investment behavior is pretty much the same everywhere: "People buy when the market is hot, sell when it's cold.... In good times they follow the herd and look for confirming evidence. When people are nervous and panicky, they shun risk for safety stocks." Advertisement HighMark has beaten its international bogey by three percentage points per year, on average, over the past five years. Hill estimates that three-fourths of outracing the index stems from country selection, one-fourth from stock picking. Two of the numerical inputs he particularly favors are price-to-book and price-to-sales ratios. Currently, his computer advises him that Europe and Brazil are relatively attractive markets. Mark Coffelt utilizes an eclectic combination of quantitative and qualitative methods in running Texas Capital Value Growth (TCVGX). The computer points him in the direction of "sweet spots in the market," such as large-company value stocks or international shares (the current advice). The computer ranks securities in the different sectors. Then Coffelt puts on his qualitative hat and decides if these are actually smart buys. He figures that the overall process is 80% quant, 20% qualitative. Sometimes he differs with the computer's conclusion. For instance, real estate investment trusts score well, but when he studied the sector he couldn't find any REITs that appeared undervalued to him. This fund is hard to categorize because it invests in stocks of all sizes at home and abroad -- and Coffelt himself expresses nothing but disdain for style-box investing. But over the past five years, his fund has returned 16% annualized, an average of seven points per year better than the SP index. Give those computers a Christmas bonus.