The Real Cost of High Frequency Trading

Michael Lewis’ new book, “Flash Boys,” relates a real-life techno thriller in which a trader who identifies and ultimately thwarts a scheme deployed by piratical “High Frequency Traders” to squeeze a relatively small amount out of many stock transactions being executed electronically. As our hero’s trade bounced around hyperspace looking for the best price, the HFTs would detect its path, use their speed to get in front of the trade, and buy up the inventory so that the price of our hero’s trade could be dictated. Of course, all of this would happen in time scales measured in milliseconds, or thousandths of a second, and the value extracted would be small, but because of the number of transactions the pirate’s booty would be large.

The Lewis book seems to have captured the imagination of much of the public, stirring policy makers and regulators to act, or rather to talk about acting. That is heartening to the few of us who have been urging action on HFT for years. Unfortunately, it has also confused the issue because the book has become emblematic of the “bad thing” about HFT. If a policy maker or regulator cannot find this “bad thing” in a given market sector, he or she just might decide that the sector has no HFT problem.

The use of stacked super computers, high powered information pathways and complex algorithms to create super-powered trading robots does not manifest in a single type of behavior. HFTs are not simply pirates in pinstripes roaming the internet sea in black-sailed computers. The process of automated trading accounts for well over 90 percent of the derivatives traded on exchanges and between 50 and 70 percent of all stock market trading. It is everywhere and high tech is used by big name firms as well as relative small HFTs who day trade in massive quantities and ultra-high speeds.

HFT has almost nothing to do with the concept of an investment in which one buys in anticipation of receiving a good return based on the performance of a company or government. Commencing the investment a millisecond sooner or later is not particularly important to that result. Nor does it effectively provide “liquidity” to the markets through “market making.” That is the business of providing continuous offers to buy and sell for a modest but relatively risk free-profit so that real investors can see what prices are available. While many have observed that real market making is exceedingly rare nowadays, the proponents of HFT and certain academics using curiously outdated and inapt methodologies claim that HFT provides more valuable market liquidity. That claim is simply wrong.

There is a fundamental and inescapable truth: the speeds that have been achieved are useful only to extract value from the modern process of the financial markets, not from investment. This is true whether or not HFT is used to effectuate a particular scheme that can be described in a successful nonfiction best seller. I had the opportunity to ask a staffer from the CFTC (the agency regulating derivatives) at a public forum what social benefit is derived from decreasing the average speed of a transaction from 125 milliseconds to 25, and I still await the answer. The staffer’s boss said that they would address that later, but it appears that he meant much later, after the public advisory committee meeting.

The real meaning of HFT is described best in a recent University of Chicago study that describes how HFT thrives at time scales in which price relationships that should be tautological break down, as if fundamental laws of nature had been repealed. It has to do with the inevitable small technology variances among trading venues (exchanges) that create tiny delays. These price anomalies would be rectified without HFT since they relate to the same underlying share prices, but the HFTs snipe true market makers by pre-positioning themselves. The study shows how this sniping reduces the liquidity in the market costing true investors money all day long, every day.

This is not a war game that only affects fat cat bankers, leaving the public unharmed. For one thing, the investors are often retirement funds and other savings pools. But perhaps more important, the investors know that they are being damaged and the cost has to be passed on through increased returns on investments. While the cost of doing a transaction may have been driven down by technology, the price itself must be higher to compensate for the uncertainty and the reduced reliable liquidity caused by HFT. This is a burden that is ultimately borne by the businesses, governments, and households that rely on the capital markets for needed funding. That means that it is a burden on the activities that create jobs and enhance the well-being of the 99% that depend on salaries.

When policy makers or regulators say that they see no evidence of “Flash Boy” type schemes in their particular regulatory patch, they miss the point. They also miss the point when they tout prophylactic measures that shut down the market when the HFT computers spin out of control because of errors or unintended consequences. The technology arms race is simply bad for the public. And the government is supposed to be looking out for the interests of the people, not the financial sector.