NEW YORK — Algorithmic trading is eight basis points cheaper than other means of trading, a new report shows. When the difficulty of the trade is taken into account, that difference grows to 11 basis points.
The report was written by Ian Domowitz, managing director, and Henry Yegerman, director, of ITG Inc., New York, an equity trading and transaction analysis firm that offers algorithmic trading models to institutional investors.
"Beyond any productivity enhancements that (institutions) might obtain in using model-based trading engines, algorithmic trading is a cost-effective technique," Mr. Domowitz said in an interview. "The cost-effectiveness holds regardless of trade difficulty."
Randy Grossman, capital markets research manager at Financial Insights, Framingham, Mass., said performance data for algorithmic trading has been lacking as broker-dealers and technology vendors flood buy-side institutions with product offerings.
"This is precisely the type of information that I've been prompting people — particularly buy-side firms — to ask for when they're investigating who to trade with," Mr. Grossman said. "People are starting to get it. Papers like this are helping make people aware that all algorithms are not created equal."
In fact, the ITG researchers found that algorithmic trading performance was consistent among the six brokers whose trading was analyzed — until order size increased; then it began to vary. Messrs. Domowitz and Yegerman declined to name the brokers they studied.