Is TCA a Crock?

Ask any buy side trader and they'll give you the party line that's become standard throughout the industry: transaction cost analysis or TCA is now part of their daily workflow.

The buy side is widely using TCA these days either because they're required to do so by institutional clients like pension funds or endowments, or they're turning to it simply to gain a better grasp of the trading lifecycle and the costs that come with it.

Advanced TCA techniques like predictive analytics and pre-trade TCA - which were only being deployed by a handful of buy-side firms just five years ago - have become the norm on buy-side desks as traders look to get a better sense of their pre-trade expenses and other factors that may impact their orders.

There's also no shortage of product offerings these days, with vendors like Elkins McSherry, Abel Noser and Investment Technology Group delivering TCA that includes some sort of pre-trade, predictive capabilities.

I've been told that TCA ultimately helps institutions find better, more cost-efficient liquidity sources as well.

But just like Michael Lewis' Moneyball illustrated how professional baseball clubs were using the wrong information to analyze talent and construct championship teams, are buy-side firms measuring the wrong things with TCA?

A high-ranking official at an electronic marketplace contends they are, and that such misuse has cost firms and ultimately their investors millions in the process.

"Everyone knows we're measuring the wrong things," the official said, arguing that if a trader gets any instructions from a portfolio manager on how much volume to buy of a particular stock, or when to buy it and how fast to execute the trade, the portfolio manager's ability should be measured on that transaction, not the trader's.

"The inputs are absolutely wrong," said the source, who asked to not be identified.

The practice of measuring a trading desk on implementation shortfall - which is the difference in price from when a portfolio manager decides to buy or sell, and the final price at which the trader executes that decision - is also a key issue, the source explained.

"If the trading desk is measured on implementation shortfall and through no fault of the trader the stock moves away from the (chosen) price, the trader is not incented to buy it even though the portfolio manager might want to buy boatloads of the stock," the official said. "That puts the trading desk at odds with the portfolio manager's goals."

Forcing a trader to make trades based on volume-weighted average price benchmarks (VWAP) also skews the way in which trade costs are analyzed over the long-term, the source added.

"Let's say the portfolio manager had to buy 10 million shares of something and if he traded according to VWAP, that means you might only be able to buy 100,000 shares a day," the official said. "Maybe in the 100 days it takes you to fill that order, the stock price moves and generally it moves against the portfolio manager's wishes.

The source continues: "It costs that much more to buy that stock over time, in fact so much so that it's not even worth buying it anymore."

Another big problem with TCA is that while it will rate a trading desk against its peers, a user will have no way of knowing what sort of "garbage" data rival firms are feeding into the service, the official said.

"My point is that if the data is not correct, then it's wrong," the source added.

"Everyone knows it's wrong and yet they're still trying to use it in different ways, shapes and forms." As the Senior Editor of Advanced Trading, Justin Grant plays a key role in steering the magazine's coverage of the latest issues affecting the buy-side trading community. Since joining Advanced Trading in 2010, Grant's news analysis has touched on everything from the latest ... View Full Bio