Once equity trading went electronic, the smart money knew it was only a matter of time before machines began making decisions that were once the domain of the trader. For the buyside, it has always been essential to know which broker to choose, based on past performance and a sellside firm’s expertise.
Today, the same rules apply—except that man’s best friend on the buyside isn’t the person on the other end of the phone or IM. It’s becoming more and more the computer on the desk.
Traders today at Batterymarch Financial Management are the beneficiaries of an overhaul of their trading system that makes skilled traders better. The firm’s new technology uses computers to analyze historical trading information, and then provides them not only with a trading strategy, but also the best brokers or algos to choose from for a particular stock. It’s called profiling a stock, and many predict this is where trading on the buyside is headed.
At Batterymarch, traders are still important to the process, as they can override the suggested strategy. They also still need to decide which algo or desk to use from a short list of suggested options. For traders there, the technology introduced in September systematizes a process that was once manual. But more important, it puts them in tighter sync with portfolio managers and gives them greater control over their orders.
That’s the payoff from two years ago, when Batterymarch, a Legg Mason company based in Boston, hired a quant analyst to review its trading and examine how portfolio managers interact with the desk. In roughly a year, the new hire was promoted to director of trading. But not before he began introducing new ideas at a firm that has been an iconoclast in its own right for much of its 42 years.
"We did this to empower our traders to be more efficient in the current market structure," said the new hire, Dragan Skoko, now Batterymarch’s head of trading. Skoko comes with a heavy math and statistics background. "We believe it is necessary to give them the proper tools, so when they reach a fork in the road, they can make the important decisions that add a lot of value."
It starts with a new proprietary pre-trade platform, which drives the process and tells traders the best way to trade a stock. Capturing historic data from a variety of internal and external sources, the pre-trade platform profiles stocks based on a number of characteristics, including liquidity, volatility, trend and momentum.
In everyday American life, profiling carries a much different connotation and is highly controversial—its use in law enforcement would be one example. As it relates to investing, however, many predict that the profiling of stocks and the using of a particular execution strategy based on past performance will usher in a new way of trading on the buyside.
Batterymarch’s pre-trade platform then spits out an execution strategy for the trader, based on a portfolio manager’s urgency and the expected level of alpha—as well as the historical data. Each strategy offers a subset of execution options to the trader: He can choose from a menu of upstairs desks or algorithms, depending on if he wants high-touch or low-touch. Or the trader can override the strategy. But every decision—and data point—along the way is captured.
Luke Lyons, a former prop trader who heads the program and derivatives desk at AllianceBernstein, estimates that maybe 10 buyside firms, including his own, are profiling stocks today. A number of buyside shops, including some of the biggest and most sophisticated, are looking to get into the game, though. The reason is simple: Better trading that adds to performance "could be the difference between winning and losing a pension fund mandate," Lyons said.
For Batterymarch, the pre-trade also includes a richer flow of information between the investment side and the trading desk. Today, the firm’s portfolio managers and traders are more closely aligned in the investment process, speaking the same quant language, if you will. Skoko, the new director of trading, is a 16-year veteran who’s spent half his career on the buyside. The upshot of his work is that Batterymarch’s new trading process gives traders a better view into the market-one that allows them to react faster-while portfolio managers get better feedback.
This pre-trade strategy selection gets to the heart of choosing the best algorithm or desk for each stock, said Jamie Benincasa, senior vice president of sales at FlexTrade, which sold Batterymarch its execution management system. "How many algorithms are there to choose from? This winnows down the options and presents a trader with the best choices. It is where the buyside needs to go."
Benincasa pointed out that more firms are moving toward this strategy, which essentially turns the buyside’s current algo selection process "on its head." The firms that profile stocks look at historical comparables: For instance, when trading a small-cap stock with an arrival price benchmark, traders usually prefer to look at which algo performed the best with stocks of similar characteristics. "They don’t start with the broker and work their way down," he said. "They start with how they want to execute and work their way back to the broker."
According to William Elcock, chief executive officer at Batterymarch, which manages $23.5 billion in equities, the trading environment had changed drastically in just a few years. Batterymarch realized this, determined it was time to re-evaluate its trading and hired Skoko.
"He’s not a traditional trader by training, but his skill set is so important these days," said Elcock, explaining that Skoko’s mathematical background is the right fit to analyze the mountains of data that drive the firm throughout its investment process.
The other key part of the plan was to more closely link portfolio management and trading. "The investment team and trading have to be connected at the hip," Elcock said. "We really believe you have to measure all parts of the investment process. Transaction cost and their analysis is a big part of what we do, and we learn from it, ultimately creating alpha for our clients."
What impressed Skoko most was Batterymarch’s seriousness about moving forward to upgrade its trading. "During one of the greatest financial crises of our time, they were willing to invest in trading," Skoko said. "Most firms were scaling back."
Everything was built in-house, including the pre-trade analytics software and its trade-cost analysis system. That necessitated customizing Batterymarch’s EMS, FlexTrade, which acts as a trader’s "dashboard" throughout the trading process.
The importance of data capture in the process can’t be overemphasized, Skoko said. It is the backbone of the firm’s pre- and post-trade, including its TCA. Results from TCA are input into the pre-trade analysis to "recalibrate" execution strategies and the subset of selections within a strategy. "Once we have enough data, we can find out where we need to improve," Skoko said. "What’s neat is that we feel the best is yet to come, as far as improving our process and getting smarter in how we trade."
Digging deep into the data from the trading and portfolio manager side is what drives the firm, said deputy chief investment officer Stephen Lanzendorf. Lanzendorf, who oversees trading, said the firm has a simple mantra: If you can measure it, you can improve it.
"We have a huge amount of data that goes from the portfolio management team to the trading desk and then back again," he said. "We sit together as teams and analyze the data, and talk about the implications and what we might be doing differently as portfolio managers and as traders. There are invariably takeaways for both sides."
Marybeth Forbes, a longtime senior trader on the desk, said the new pre-trade analysis has streamlined her old manual process and gets her trading much faster. "It used to be time-consuming to put together each trading strategy," Forbes said. "Now I don’t have to look at all the factors individually to decide what my strategy might be. We’ve already done that systematically."
Additionally, Forbes said, having real-time information to monitor an order against various benchmarks gives her more insight and control in accessing the market. "With all this data at our fingertips, we can dig deeper into what’s happening with our order," she said.
AllianceBernstein’s Lyons, a quant himself, said he wasn’t surprised Batterymarch had created its own pre-trade analytics and was profiling stocks. "They’re really the godfathers of quantitative investing, considering who’s worked there and what they’ve done," he said.
Since 1969, Batterymarch Financial Management has looked at the investment world a little differently from most. Its founder, Dean LeBaron, was among the first quant managers on the Street, using computers and passive strategies when only a handful of firms dared.
On the trading side, LeBaron brought in Evan Schulman as head trader in 1975. Schulman executed the first program trade while at Keystone Funds. He not only helped computerize the firm-front, middle and back office-but also introduced an innovative trading system. Schulman let brokers access its orders via computer to trade. He would later leave the firm to found Lattice Trading, which offered an electronic trading product that was a forerunner to today’s algorithmic trading.
Wall Street sources say Batterymarch over the last 25 years has been either contrarian or progressive, or both. When the Nasdaq market was hot, institutions enjoyed better pricing on Instinet, trading inside the wide spreads offered by market makers. Batterymarch, however, would use Instinet to place outsize bids and offers a dollar or so away from the market. "You’d see 658,000 shares for sale and you knew it was Batterymarch," said one longtime buyside trader. "And you’d watch the brokers look to move the market in that direction so they could fill that order."
More recently, after penny trading was introduced, displayed liquidity dried up. Buyside trading desks were struggling to trade block-size orders. It was a huge issue for the industry, but not for Batterymarch. Its strategy was to blend in and slice up its orders well before algorithmic trading, sources said. During that same period, Batterymarch used an incentivized volume-weighted average price payment schedule to get brokers to pay attention to its orders: Beat VWAP, get 3 cents a share; match it, get 2 cents; underperform, and the commission could be as little as a penny.
Skoko began his undertaking by first learning how the investment side operated—a quantitative approach that screens stocks based on fundamentals. Then he reviewed where the market structure was headed. Finally, he went to work with the trading desk, learning how they approached trading. This led to the roughly 10 preset trading strategies—variations of scheduling, implementation shortfall and liquidity seeking.
Skoko found that the trading desk was a huge ally. He cited a couple reasons for the success both during the evaluation process and after. First, the culture at the firm encourages interaction and is transparent. Second, the desk was supportive because he included them early in the process and earned their trust. Ideas from the trading desk led to creation of the execution strategies, as well as the design of the trading platform itself.
"Their feedback resulted in a number of enhancements, including the customization of broker algorithms," Skoko said, adding that the culmination of their work "highlights the strategic role today’s traders need to play on the buyside."
Batterymarch, even though it does its own proprietary pre- and post-trade analysis, still relies on its brokers and wants strong relationships, Skoko said.
For one, investment managers today need the sellside’s expertise to better understand the complexities of regulation, technology and the rules of engagement in markets, particularly for global investors like his firm that trade in roughly 50 markets. Batterymarch uses 50 brokers worldwide and trades with 24 domestically.
"We need access to field experts, and they reside on the sellside," Skoko said. "Market structure research is important to us, because it helps to shape our sourcing of liquidity and execution strategies. Strategic thinking is becoming critical to the buyside trader."
The sellside provides market structure research through both written reports and meetings. But there are also everyday issues, like how a desk services the client, which is important, Skoko said. For example, suppose a trader on the desk is not satisfied with its participation level or pricing on a trade. The trader might pause the order and call the broker to find out what happened. Maybe routing logic has changed or there was leakage somewhere.
"We’re pretty good at picking up the phone to alert them that something didn’t feel right," Skoko said. "We put a lot of thought into how we trade." But rather than being quick to cut off a venue, he prefers to alter how the algorithm accesses liquidity. "We have a lot of faith in our process."
Now that Batterymarch’s system has been implemented, Skoko can look back and see the fruits of the 15-month project. But he also understands that markets change, and because of that, he realizes that future adjustments will come from trading performance and what the data tells him.
"Internally, on our end, there is so much to analyze and so much to learn. And externally, from the sellside and the exchanges, you have another whole set of data to analyze both during and after the trade," he said. "It’s a great environment to be in, and TCA is a big help."
Red Star Loves Skoko
At the age of 16, Dragan Skoko stood 6-foot-8. His height and basketball skills helped capture the attention of one of the best semi-pro teams in Yugoslavia, Belgrade Red Star. It wasn’t bad, he said, going to high school, playing ball against top competition, while earning enough to pay for lodging and other necessities. Later, however, Skoko suffered several major knee injuries. "It changed my game drastically," he said.
Fortunately, his game tapes found their way into the hands of college basketball scouts in the U.S. Skoko visited his brother Goran, who played basketball at Fordham University in New York and was taking his SATs to become eligible to enroll at Manhattan College. During his stay, Skoko also received a phone call from the staff at Radford University in Virginia. He ended up going there, instead.
Shortly after Skoko left for college in 1990, civil war broke out in his country; it continued during his four years at Radford, where he studied math. "Those were incredibly tough times," said Skoko, who always wondered about his parents’ safety in Yugoslavia. Friends on campus helped him get through the uncertain times. But the rigor of being a student-athlete was his best diversion. His parents survived and moved to New York City in 1996.
After graduation, Skoko worked at ILX Systems in New York as an analyst for nearly eight years, and then moved to Boston to work for PHZ Capital Partners for six years. He then spent two years at Wellington Management. As an escape from the quantitative grind, Skoko today finds a more personal outlet. He works with youth as a coach in an Amateur Athletic Union (AAU) basketball league. It’s his way to give back to a game that gave him his start in this country—plus, he finds great enjoyment from working with kids.
"I’ve got this connection with them that is hard to explain," Skoko said. "Because I am so much taller than they are, I think it gives us an immediate conversation point. And they get a big kick out of seeing me dunk for them in practice once or twice a year." Outside of work and coaching, Skoko spends as much time as he can with his wife, Nikole, and their three children.