Washington has already been holding high-frequency traders’ feet over a fire of scrutiny. Now two recent studies may turn up the heat.
The first, by brokerage Investment Technology Group, said high-frequency trading strategies hurt best execution in the dark pools they frequent. The second, by Quantitative Services Group, a research provider, said these strategies feed on orders run through such passive algorithms as volume-weighted average price and lead to higher impact costs.
Both studies attempt to provide material evidence on how high-frequency trading strategies affect the equities markets. They also fall on the heels of a study the electronic brokerage Pipeline Trading Systems and the buyside firm AllianceBernstein conducted earlier this year which examined dark pool aggregators and adverse selection–and involved high-frequency traders. All three state that interacting with high-frequency traders in certain scenarios can be risky.
Despite these studies, there are still trading pros who are not convinced of their conclusions–including a large percentage of buysiders who said in an October Greenwich Associates study that they need more information on the benefits and pitfalls of high-frequency trading before forming an opinion on the practice.
Whether these studies will be enough to convince them remains to be seen. Still, the studies’ authors appear unwavering in their findings.
"We see clear signs that the dark pools that have more high-frequency flow have a lot more adverse selection than the dark pools that have lower high-frequency flow," said Hitesh Mittal, who manages ITG’s crossing and algo products.
As many dark pools allow unfettered access, regardless of flow type, high-frequency trading activity therein has surged, according to the ITG study. And as high-frequency traders use short-term alpha models, incidences of adverse selection have also risen, ITG reported. For its part, ITG runs a dark pool that does not let in high-frequency traders. Adverse selection refers to trading with the wrong counterparty, such as a more informed trader.
For the QSG study, VWAP algo orders incur significantly higher impact costs, particularly in liquid, low-priced names, according to Tim Sargent, the firm’s president and chief executive. This is because high-frequency traders find it easier to pick off such orders.
Orders using a VWAP algorithm are broken into many "child" orders and then sent out into the market in chunks at scheduled intervals throughout the day. High-frequency traders, meanwhile, read the tape to try to identify flow patterns within a stock. When they do, Sargent said, they can establish positions in front of the child orders.
"They would identify this price pattern and jump in and scrape away liquidity," Sargent said. "This increases the likelihood that the client using the VWAP is going to have to cross the spread for the volume that they need at each one of those set points along the path."
The QSG study finding came as no surprise to those in the industry who follow transaction costs trends. VWAP algos have developed a reputation for being costly strategies, said Alex Hagmeyer, a quantitative analyst at QSG.
"One of the major steps we’ve taken is to be able to attribute those costs to the algorithm level, or the strategy level," he said. "It wasn’t a huge surprise to me, from an internal perspective, because I’m looking at costs by algos all the time, and trying to help our clients understand which algos have increasing trends in high costs and which ones don’t. And we see that VWAP algos over time can be quite costly."
And even though VWAP’s susceptibility to gaming has long been apparent to experienced traders, many in the industry continue to use the algos carelessly, Sargent said. The QSG study professes to quantify and document how high-frequency trading strategies expose VWAP users to even greater costs, he added.
“So, if it wasn’t good before, it’s even worse now,” Sargent said.
Some in the industry, though, say more time and information are required before a definitive value judgment can be ascribed to high-frequency trading. Back in October, Greenwich Associates released a survey that showed how institutions weren’t sure whether high-frequency traders benefit or detract from the quality of liquidity in the marketplace. And almost 90 percent of the Greenwich survey’s participants said there’s no hard data available currently to show conclusively how high-frequency trading affects trading costs.
Pipeline, ITG and QSG have attempted to supply just that. But others, starting with Goldman Sachs Electronic Trading, have hard data showing that implicit costs have actually been falling since 2003.
Ultimately, the industry needs more time to reach a degree of clarity on high-frequency trading, said Amit Manwani, who heads quantitative and analytics products in the U.S. for Nomura Securities International. It took the industry time to process the impact program trading had on the equities marketplace after most blamed the practice for the Crash of 1987, he said. And high-frequency trading has fallen under similar scrutiny during the most recent dislocations in the marketplace.
"We’re in that period where there’s a huge gap in understanding of various market practitioners of what this form of trading means," Manwani said, "and what are the advantages and what are some of the potential abuses that can take place, and how to make sure that we capitalize on all of the good that can come out and are able to prevent some of the damage that can happen from high-frequency trading."
