Institutional Investor magazine has published its annual Elkins/McSherry survey of trading costs, and naturally this year the context is the impact of high-frequency trading in markets around the world.
“Fortunately, despite all the controversy about high frequency trading, the evidence suggests that these high-speed, high-volume players continue to put downward pressure on equity transaction costs, not raise them, according to our 14th annual survey of trading costs, conducted for Institutional Investor by New York–based Elkins/McSherry, a subsidiary of Boston’s State Street Corp,” the article says.
For the last five years, I’ve been following the survey for our blog, and each year New York Stock Exchange-listed stocks have had the lowest transaction costs in the world. This year, the article says the U.S. lost its lowest-cost crown:
“The average overall cost of equity trading in the U.S. edged up during the 12 months ended June 30, but that increase largely reflected lower trading volumes and higher volatility, says James Bryson, president of Elkins/McSherry. The underlying trend in costs is still headed lower, he adds. And costs fell significantly in most international markets.
. . .
“Average transaction costs for New York Stock Exchange–listed stocks rose 10 percent in the latest period, while costs on the Nasdaq Stock Market were up 16 percent, according to the survey. The U.S., with an overall average cost of equity trading of 19.63 basis points in the period, is no longer the world’s low-cost trading venue. Japan and Sweden share that crown, with an average cost of 18.34 basis points in the period, followed closely by France, at 18.49 basis points. Average transaction costs globally declined 8.1 percent over the past year, to 38.02 basis points.”
I hope my CIPs (colleagues in Paris) take note of that uptick in France’s ranking.
The survey coverage always includes a country-by-country table of trading costs, but this year it’s missing. Here’s hoping someone at the magazine or E/M will forward it to me, or at least consider this a request that it be restored next year.
The other notable change is that for the first time I can remember, the article includes a table showing average execution time since 2006-2007. To no one’s surprise, following an uptick in 2007-2008 it’s been falling sharply.
The uptick in speed may be expected, but for the execution time of large trades to drop by more than half in four years is nonetheless remarkable.
How will today’s level of volatility and trading volumes impact next year’s survey? Would love to hear your comments.