Mirroring the trend in baseball popularized by Michael Lewis's book, "Moneyball," today's extremely volatile, high-speed marketplace is forcing the buy side to squeeze more out of analytics in order to achieve best execution. Similar to the way in which Major League Baseball teams now use advanced analytics to evaluate talent, in recent years buy-side trading desks increasingly have turned to transaction cost analysis (TCA) tools to get a grip on the costs they incur at various points throughout the trade life cycle and to enable firms to grade themselves against their peers.
The buy side also has been driven to use TCA by institutional clients, including pension funds and endowments, that want transparency into their executions. Even buy-side firms' internal compliance departments find TCA useful, since it can help determine whether the brokers through which they transact are providing best execution. And while only a handful of firms utilized sophisticated techniques such as predictive analytics and pre-trade TCA just five years ago, now they are industry standards.
[For more on the spread of TCA on Wall Street, see "TCA Grows Beyond the Equity Trading Desk."]
But since buy-side firms face a growing mountain of data, often referred to as Big Data, that must be analyzed and boiled down into actionable intelligence, trading desks may not always know how to pinpoint the information that actually can help them get the most out of their trades, according to sources. And since the hidden costs of trading can vastly exceed the commissions the buy side earns, these missed signals can be especially damaging.
"It's incredibly hard to do correctly," Matthew Celebuski, senior managing director and head of quantitative research at institutional brokerage SJ Levinson & Sons, says of TCA. "You have to be able to perform a broad range of analysis on that data. And you must be able to display that analysis and provide a research tool so a trader who understands the order flow can look through it -- not for the purpose of giving himself a report card, but to improve trading."
Tapping the Right Data
But the problem may be even more deep rooted than that. A high-ranking official at an electronic marketplace says buy-side firms are analyzing the wrong data with TCA, a mistake that can cost them and their investors dearly in the long run. "Everyone knows we're measuring the wrong things," says the official, who would speak with Advanced Trading only on the condition of anonymity.
For example, if a trader receives instructions from a portfolio manager regarding the volume of a particular stock to buy, or when to buy it and how fast to execute the trade, the portfolio manager's ability -- not the trader's -- should be measured on that transaction. "The inputs are absolutely wrong," the source contends.
For the most part, the buy side leans on TCA to grade its trading practices and measure a desk's performance against expected costs. But TCA usually doesn't arm firms with the information they need to actually improve their trading, SJ Levinson's Celebuski contends. In order to give buy-side clients the clearest view possible of the quality of their executions, Celebuski says, SJ Levinson breaks down trades from a number of perspectives, including the separate vantage points taken on a transaction by a portfolio manager and the trading desk.
According to Celebuski, the firm also analyzes the effect timing has on an execution, measures the trade against a wide range of benchmarks, and gauges the effects the venue on which it was traded may have had on the execution. SJ Levinson, he says, even measures the venue against a set of benchmarks that are designed to help a trader determine whether it's a good place to trade. In addition, the firm's TCA enables traders to take a critical look at their algorithms.
"Once you understand all the aspects of the trading process, you see things that could be improved," Celebuski says. "Whether that's timing, where you trade, the use of limit orders -- all of these things come into play when you're analyzing the trades."
But the challenge facing the buy side when it comes to sifting through the data and homing in on what's most pertinent is difficult to overcome, acknowledges Robert Kay, the head of analytics at TCA provider TradingScreen. "There are a vast number of statistics, and they're not all necessarily measured in the same way. It's fair to say the smarter traders have figured that out and make use of that information," he says. But there's no doubt that some buy-side traders "are probably not using the right metrics or possibly -- perhaps even more dangerous -- they're not drawing appropriate conclusions from the metrics they're using."
Defining Trade Context in Real Time
Meanwhile, traders say one of the biggest challenges they face in optimizing their executions is defining the contextual aspects of the market before firing off a transaction. That includes figuring out how liquid a particular stock might be at a certain point in a trading session, how volatile the market is overall and what the general market movement is like on a moment-by-moment basis.
"We as human traders cannot effectively process that information across more than maybe 10 stocks at any given point in time," says Mark Kuzminskas, director of equity trading at Robeco Investment Management. "So we need real-time analytics to help us manage that information flow. That's where TCA comes in."
The most effective TCA tools, Kuzminskas adds, are migrating away from basic functions such as determining how a trade stacked up against a volume-weighted average price (VWAP) benchmark, or how well a desk may have done against rivals; instead, next-generation TCA tools are zeroing in on what's happening in the market in real time. In fact, Kuzminskas reports, Robeco is looking to make the leap from post-trade analysis to real-time TCA integrated into the decision-making process, a move he says will allow traders to know when to speed up or slow down their executions and that ultimately will beef up the bottom line.