The survey of head traders at 60 institutions conducted by Financial Insights, found that the 93% of respondents use algorithms, primarily to fill larger liquid orders. Sixty percent of participants are most likely to use the technology to trade large cap stocks as compared with 25 percent for mid caps and 10 percent for small caps.

Randy Grossman, research manager for the capital markets practice of Financial Insights suggests that traders are using algorithms as productivity tools, feeding their easiest orders through the computer while leaving the difficult ones to be executed by a human trader.

"A decade ago traders managed trade blotters by passively handing out discretionary orders to brokers and taking trade reports throughout the day," he says. "Today, many actively work their orders in the market to ensure lower commission rates, greater control and less information leakage, all of which are benefits of algorithmic trading, therefore pointing to a high potential for future growth in users."

Recent comparative research by ITG estimated that the use of algorithmic trading can shave up to 11 basis points from investor costs. Average performance differences across providers for very small orders were few, noted ITG, but gaps between providers grew as order size increased.

Financial Insights found that the most basic VWAP (volume-weighted average price) algorithms remain the most popular with buy side users. Fewer than half (40 percent) report that they use customised strategies, while a 51% majority expressed some confusion over the intricacies of algorithmic trading, suggesting an educational role for brokers.

Though the buy-side still remains in the early stages of adoption, Financial Insights estimates that currently 30-35% of trading volume at bulge bracket firms is executed via algorithms.

As the sell-side continues to aggressively push algorithmic trading to contain its costs, buy-side adoption is certain to accelerate.

"The initial phase of the arms race is over," says Financial Insights' Grossman. "However, the next challenge for algorithmic trading providers will be to differentiate themselves and better educate their customers."