In the quantitative world of finance, evaluating the performance of traders has long been an important but elusive task. Now, as the buyside does more self-directed trading, it’s becoming even more important to assess how well traders do their job and how much money this puts in the pockets of fund owners.
Clear answers aren’t easy to come by. But Morgan Stanley Investment Management is currently implementing a framework it believes will quantify the value a trader delivers to the firm’s investment process. MSIM, which plans to compensate traders based, in part, on the numbers that emerge from its calculations, hopes other firms will embrace a similar methodology to bring more rigor to the trading discipline.
“The goal is to align the performance evaluation of traders with the way investment managers are evaluated, and the behavior of traders with the objective of the firm,” said Robert Shapiro, an executive focused on trading and execution analysis at MSIM. The objective, he said, is to “preserve, protect and grow client returns.”
That’s a laudable ambition. But the impetus for all the mental effort is simpler. “Traders believe in their heart of hearts that they add value,” Shapiro said. “We’re hoping to quantify the extent to which traders contribute to the objectives of the investment process.”
Buyside firms have wrestled with these issues for a long time. In this, MSIM has company. Fidelity Investments, Franklin Templeton Investments, Wellington Management and AllianceBernstein Investments have focused intensely on trading metrics. Franklin Templeton, for instance, analyzes two-dozen benchmarks that help it evaluate different aspects of the trading function for the large orders it executes, according to Bill Stephenson, head of global trading strategy at the $416 billion investment firm.
Transaction cost analysis, of course, has existed since the 1980s. Its focus has been on using benchmarks and analytics to try to ensure the best trading outcomes. But figuring out exactly how much better one trader did than another is complicated by a range of factors, including the orders the trader is given to trade, the time horizon, the stock’s liquidity and the market volatility.
MSIM’s effort to quantify the value traders provide got its marching orders from Ray Tierney, global head of equity trading at the firm. When the 26-year sellside veteran moved to the buyside in 2006, he decided more metrics were needed to excavate the value traders bring to the investment process. “In our business, we’re paid for the executions we get every day, so we need to know exactly what we’re getting,” he said. MSIM has just over $200 billion in equity assets.
The PM Perspective
MSIM set itself a straightforward task: to lay out the metrics needed to evaluate traders the way portfolio managers are evaluated. PMs are judged on their absolute returns, their returns relative to a passive benchmark such as the Standard & Poor’s 500 Index, and their returns relative to their peer universe, MSIM’s Shapiro said. So traders should be assessed, the firm decided, on absolute performance, relative performance and their ranking in their peer universe. Shapiro is a former head trader at Iridian Asset Management who was schooled in trade cost analysis during a stint at Abel/Noser Corp.
In the MSIM universe, three benchmarks will now be used to gauge a trader’s performance. The firm calls this the “PM Performance Parallel.” It has also come up with a metric that represents the actual value a trader added to the fund and its clients in AUM dollars. “All this gives PMs perspective on how well traders are implementing their orders,” Shapiro said. “And traders tend to improve the moment they think they’re being graded.”
MSIM hopes its efforts will overhaul the way buyside firms look at the trading function. “What makes TCA and the evaluation of traders difficult is that there are no uniform standards,” Shapiro said. “This construct can be applied by a lot of firms that feel our way of viewing performance reflects their goals as well.” He added that MSIM is willing to share its research and metrics with other firms “to help legitimize the trading discipline as part of the investment process.”
The first of the three metrics is absolute trading performance. This hails from the arrival price, or implementation shortfall, benchmark. The arrival price is the price of the stock when the trader received the order. The more slippage a trader has from that price, the larger the absolute cost of trading that order.
The second metric, relative cost, comes from something called PWP, or participation-weighted price. That’s ITG’s name for its version of a benchmark that’s been around since the 1990s, when Abel/Noser conceived of a “20 percent test.” The original test, according to Peter Weiler, executive vice president of global sales at Abel/Noser, “compares a trader’s execution price to a what-if scenario in which 20 percent of each minute’s volume is traded at the volume-weighted average price until the order is completed.”
ITG acknowledges the benchmark’s history. “The PWP concept has been around for years, but had to overcome the challenge of its computational intensity,” said Jon Fatica, a director at the firm. The benefit of the benchmark, he explained, is that a trader’s result can be compared to a passive participation strategy that an algorithm could achieve with minimal market impact. That provides a measure of a trader’s skill at taking advantage of liquidity in the market. “It’s man against a machine,” he said.
ITG had the mechanics of PWP in place for about eight years and did the calculation for a large client for many years. But it was Franklin Templeton that put PWP on the map. The global investment management firm had used a proprietary participation-based measure of trading costs for nine years but switched to PWP a year ago, after ITG showed that it functioned similarly to the other benchmark but was easier to calculate.
MSIM’s third trader-performance metric involves the trader’s ranking in a peer universe. For this, MSIM wanted to rank the arrival-cost result of a trader’s orders to those costs for other similar orders. MSIM worked with ITG on this. “We wanted to be able to compare a domestic large-cap, high-ADV, high-momentum, low-volatility trade to other similar trades during a similar period, rather than to trades from institutions of comparable size,” Shapiro said. The attributes MSIM considers critical for these comparisons are trading region, market cap, trade size, volatility and momentum. ITG runs the comparisons off data in its universe of trades.
MSIM also decided it needed to round out its trading-evaluation schema by tying a trader’s performance to compensation. MSIM knew that a fund’s AUM had to be part of the equation, so a trader’s performance had to be expressed in AUM dollars.
The firm’s quant team worked with George Sofianos, an execution-quality expert at Goldman Sachs, and his group to devise what MSIM calls a “trade performance metric,” or TPM. The TPM is based in part on a methodology Sofianos developed to quantify how well traders take advantage of trading opportunities given a stock’s momentum. As Shapiro put it, TPM offers a way to gauge “how effectively the trader sourced liquidity while exploiting the trend in the stock he was trading.” The metric incorporates order difficulty and data that quantifies the momentum of the stock while the trader was in the market and shortly afterwards. This is expressed in AUM dollars and therefore can be converted into a revenue line for trading, Shapiro said.
However, this metric alone isn’t enough. While TPM conveys the absolute amount of value a trader contributed to a fund’s AUM, it doesn’t quantify how well a trader did compared to a passive benchmark or his peers. In MSIM’s view, the analytical framework that replicates the way PMs are assessed is therefore necessary to evaluate the trader’s performance in a broader trading context.
(c) 2009 Traders Magazine and SourceMedia, Inc. All Rights Reserved.