WARREN BUFFETT had it all wrong, Mark Carhart told a packed New York conference last summer.

The fortysomething from Goldman Sachs cited study after study showing big-name companies with high price-earning multiples or rapid growth rates make poor investments.

Traditional stock pickers, including Buffett, a fabled raconteur, may “tell great stories,” said Carhart, but betting on big names like Coca-Cola and Gillette was so old-fashioned and obviously no match for Carhart and his complex box of tricks.

Well, now Carhart has a pretty good story to tell. Alongside Raymond Iwanowski, Carhart heads Global Alpha, a $10 billion (€7 billion) hedge fund for clients and employees of Goldman Sachs. Global Alpha is a quantitative fund, meaning that its trades are determined by computers and mathematical models devised by its egghead managers.

If computers can beat chess champions, why not Buffett? And Global Alpha’s computer-powered arithmetic has paid off handsomely for investors – when it works. In 2005, the pair notched a 51% gross return.

Money magazine ran an article last year with the headline “Could a computer be the new Buffett?” Apparently not.

Last year, Global Alpha lost 9%, its first deficit since 1999. All eyes have been on Alpha’s performance this year. It doesn’t look good.

Last month was the worst in its 12-year history. The fund was down 22.7% after making a series of costly – and presumably computer-generated – mistakes. Goldman also had to inject $2 billion into Global Equity Opportunities, another quantitative fund run by the two, after it lost 28% in the first eight trading days of August.

Usually nobody, apart from Goldman, knows the exact details of Global Alpha’s trades. But in a letter to shareholders last week Global Alpha set out details of some of its missteps.

August was a horrific month for the stock markets and the hedge funds in particular. Alpha’s woes were exacerbated by massive selling as the subprime mortgage market collapsed, credit suddenly dried up and investors in some funds started asking for their money to be returned.

The fund also wrongly bet that the Japanese yen would fall and the Australian dollar would rise. Global Alpha went bullish in Norway and bearish in Finland, only to lose money on both. Earlier this year, Alpha bet against the Canadian dollar, now at a 30-year high. It took a negative view on Japanese government bonds; they rallied.

The problem for Global Alpha, and its rivals, is that there are now thousands of hedge funds chasing the same ideas. What were once unique strategies are now widely imitated and returns are falling.

The average hedge fund gained 13% last year, according to Hedge Fund Research. A mutual fund tracking the S&P 500 returned 15.8%.

When he was at school, Carhart had a number of nicknames. At high school, he was reportedly called Marcus Aurelius Piscarus Carhartus, until he helped invent a machine that could fling an egg off the top of the school without it cracking. Then he was called The Inventor. Considering Alpha’s past record, it’s still a bit early to be writing it off. But it does seem the Inventor and pal are not all they were cracked up to be.

Buffett (nickname “the Sage of Omaha”) has been here before. You may remember when a lot of people said he’d gone senile for not pouring his money into dotcom stocks. He’s not a big fan of quants either.

Times Recommends

• Lies and cover-up at MG Rover exposed by report
• High oil prices push manufacturing costs up
• Primark founder steps down as chief executive