Emotions are your worst enemy, says this successful fund manager. He inveighs against trying to time the market. By Anne Kates Smith, Executive Editor May 31, 2007 John Montgomery's 11 Bridgeway funds select stocks through computer-driven models. He explains the benefits of this approach.KIPLINGER'S: Why do you favor number crunching over other methods? MONTGOMERY: The biggest enemy in investing is emotion. Quantitative methods remove the emotion completely. Our process looks exactly the same in a bull or bear market. Here's a rule of thumb: If something doesn't affect the decision you make, don't look at it! Most days, I don't know whether the stock market is up or down. None of the computer screens in my office has a ticker tape. RELATED STORIES Bridgeway Bounces Back 25 Best Mutual Funds Mutual Fund Center Should we follow your example by reacting less to daily gyrations? We try to court long-term investors. It's unacceptable to me when shareholders buy after a great run-up, then sell when a fund underperforms. That happened in our fund Aggressive Investors 2. During four years of market-beating returns, the average shareholder's return was roughly the same as the fund's. Last year, when the fund didn't do as well, the average shareholder's return was less than the fund's, because they bought high and sold low. That drives me nuts. Chasing returns is a huge problem -- it's killing results. Long-term stock-market returns are about 10% a year. Investors probably lose two percentage points to costs and another two points to poor timing. Advertisement So how do you adjust to changing market conditions? By paying attention to your asset allocation. For example, this is the longest small-company-dominated market in eight decades. We've been saying for three years that you shouldn't have more money in small caps than you originally meant to put there. Our Ultra-Small Company fund was up 22% last year alone. If you started out years ago with 10% of your portfolio in small companies, you might be close to 35% now. Time to trim back because those gains could go the other way. What mathematical models do you find most useful? We're excited (I confess to emotion on this) about a new area of research into contrarian investing. We're looking at ways to measure how much a stock is loved and identify stocks that are mercilessly pounded beyond what's reasonable. The model we're developing appears to dampen the risk for stocks in a big way.