With stocks struggling through a terrible bear market over the past two years, investors have taken huge sums of money out of stocks and poured money into bonds — a seemingly “safe” decision that will end up costing them dearly, says fund manager and author James O’Shaughnessy.

In his latest market commentary on O’Shaughnessy Asset Management’s web site, O’Shaughnessy says that from Jan. 1, 2008 to Aug. 26, 2009, investors pulled $41.8 billion from open-end U.S. equity funds while adding more than $310 billion to open-end U.S. bond funds. They’re doing so at a time when stocks have been hammered — Research Affiliates’ Rob Arnott noted back in February that over the preceding 40 years, bonds actually beat stocks.

But O’Shaughnessy says those who are dumping money into bonds are falling prey to recency bias, the phenomenon in which we give more weight to recent history than to the longer-term picture. “We’ve seen other 40-year periods where bonds have beaten stocks and we are far more interested in what happened afterwards,” he writes. “I believe that all of these changes to asset allocation are coming at exactly the wrong time and that investors who follow this herd-like move to overweight bonds versus stocks will be bitterly disappointed over the next ten years.”

To support his view, O’Shaughnessy looked at how two different types of portfolios would have fared following recessions. The first uses the typical 60% stocks, 40% bonds allocation. (The stock portion is 50% in large-cap value stocks; 35% in small-cap stocks; and 15% in large-cap growth stocks. The bond side is 45% in long-term U.S. Treasuries; 45% in intermediate-term Treasuries; and 10% in T-bills.) The second portfolio is made up of what O’Shaughnessy says many advisors are recommending today: 15% in equities (the S&P 500) and 85% in intermediate and long-term Treasuries and T-bills.

O’Shaughnessy looked at how those two types of portfolios would have performed following every recession since 1926. The results after:

One year: The 60/40 portfolio returned 28.82%, on average, in the first post-recession year vs. just 7.25% for the 15/85 allocation;

Three years: The 60/40 portfolio returned an average of 14.98% annualized in three years following a recession, vs. 6.88% for the 15/85 allocation;

Five years: The 60/40 portfolio returned an average of 12.71% annualized in the five years following a recession, vs. 6.25% for the 15/85 allocation;

Ten years: The 60/40 portfolio returned an average of 12.84% annualized in the ten years following a recession, vs. 6.40% for the 15/85 allocation.

“That means that if you currently have $250,000 in your retirement account and follow the 15-85 route being advocated by so many advisors today, ten years from now your account would be worth $464,897 versus your account growing to $836,702 following the 60-40 portfolio and getting the average results we have seen from this study — nearly double the value of the bond-centric portfolio!” O’Shaughnessy writes.

The bottom line, according to O’Shaughnessy: If investors “have made it through the worst drop in the equity market with a portfolio that was dominated by equities, it makes no sense at all to now move them to a portfolio dominated by bonds. That would have been a prudent move before the plunge in stocks but seems downright foolish to do so afterward.”