In a few thousandths of a second, computer systems used by high-frequency traders can anticipate large purchases by pension and mutual funds, driving up prices and sharply reducing their returns. And the problem is widespread; such high-frequency transactions account for about 60 percent of modern stock-trading.
UC Irvine researchers are studying ways to outsmart these high-speed highwaymen. Mutual fund managers might, for instance, randomize their orders of stocks in a piecemeal fashion to conceal them from the lightning-speed analysis of high-frequency trading algorithms. Or they might time their orders to hit many markets simultaneously, leaving would-be scalpers no time to anticipate the investments and attack.