Program trading is booming, but the share of flow reaching the sellside is in decline. With business up so strong overall, though, brokers don’t seem to mind. Armed with increasingly sophisticated technology, the buyside is doing more trades in-house, rather than shipping them out to brokers.
The steady change over the past year or so has begun to alter the relationship between the buyside and their brokers. The trend has also changed how execution management system providers design the buyside’s trading tools.
And when the buyside does send its orders to the sellside, it is increasingly bypassing the program trading desk and using the firm’s algorithmic trading services. “The buyside has grown substantially as more money has flown into their investment vehicles,” says Andrew Silverman, Morgan Stanley’s head of sales and distribution for electronic trading. “In order to handle the flow, we see many traditional money managers now utilizing program trading strategies.”
Do-It-Yourself
The trend of more do-it-yourself trading of single stocks is now expanding into the program business. Seven or eight years ago, industry pros say, buyside traders directed 100 percent of their program trades to sellside desks. Today that figure stands at about 75 percent.
Data gathered between November 2006 and February 2007 for the latest Greenwich Associates portfolio trading study showed that the buyside will trade 25 percent of all program trading volume itself this year, compared with 21 percent in the same period a year earlier. Although the buyside increased its level of self-trading for programs by 4 percent in a year’s time, which might not seem like much, that still represents a 19 percent increase.
The increase occurred at a time of exploding program trading volume. According to the Greenwich study, the average buyside shop saw a 50 percent increase in the dollar amount of program trading it executed in 2007, compared to the previous year.
Marketwide, total portfolio trading volume jumped 52 percent to $2.1 trillion, from $1.38 trillion the previous year, the study reported. Greenwich attributes these enormous increases to cost, and how institutions viewed the program business as a way to lessen their average cost of trading, or “blended rate.”
And according to the 135 institutional equity buyside desks the Greenwich study surveyed, program trades represented roughly half of their total equity trading volume.
Million-Dollar Trade
Program, or portfolio, trades are defined as those involving 15 or more stocks valued at more than $1 million. There are two kinds: agency, where brokers work baskets for a commission, or the buyside handles the order itself; and principal, where a broker commits capital and takes on the client’s risk at a predetermined price. Agency trades are the most popular, constituting roughly 75 percent of all executed programs.
The buyside’s move to handle more of its own trading-whether single-stock or program-is happening for a number of reasons, says David Liles, global head of electronic trading at agency broker Sanford C. Bernstein.
For one, the quality of the trading technology the buyside has at its fingertips in some cases now rivals that of the sellside. Another reason is anonymity. “The buyside wants to avoid having anyone else know what they are doing, and algorithmic trading has definite advantages over other execution methods in this area,” Liles says. Third, the desire for more control over a trade drove the buyside to use algorithms themselves. Finally, trading talent has been moving from the sellside to the buyside, Liles notes, so a lot of buyside traders have sellside experience. “If a buyside trader has the right tools and the right experience, he’s in a good position to take greater control of his order flow,” Liles says. And of course, cost is always important, he says in agreement with Greenwich’s findings. If a buyside trader takes on the responsibility for a trade, it should cost less.
Jeff Galloway, CFA, head of derivative and program trading at Houston-based AIM Investments, says his firm’s agency business in three years has shifted from being 90 percent handled by the sellside to being 90 percent self-directed. Program trades account for roughly 20 percent of the firm’s overall flow.
“By sending a particular program or order to a broker-dealer, you don’t have the capabilities for real-time data,” he says. “A portfolio manager would ask, How are we doing on XYZ program? What percentage are we done on? How are we doing versus arrival price?’ You’d have to call a broker-dealer and gather that information. Now, that information and data are available real-time at my fingertips.”
But many of the types of agency program trades the buyside is executing itself are either the very regimented, such as trading volume-weighted-average price every day, or the very routine, such as market-on-close trades, says Brian Fagen, head of electronic trading and program trading sales at Lehman Brothers.
“You still don’t see a lot of buyside firms handling more complex program trading situations,” he says. “Those are the agency trades that tend to end up on the sellside desks.”
Do It My Way
When the buyside hands a trade over to a broker to execute electronically, he doesn’t do it perfunctorily, according to Vijay Kedia, president and chief executive of EMS provider FlexTrade, in Great Neck, N.Y.
“They’re still using all the broker algorithms, but they control the delivery of the orders to the brokers much more actively than they used to,” he says. “It’s a combination of the buyside using algorithms to define the trajectories of when they want a program’s stocks to be executed, but at the same time, leveraging the right broker algorithm for each chunk at different times of the day.”
Despite fewer trades being handed off to brokers, the sellside says it is fine with all of this. The program trading business is evolving, they say, and their desks are evolving along with it. Some brokerages even say they are still seeing increased program volume at their desks.
Technology Improvements
At Morgan Stanley, among the three largest program trading houses on the Street, combined risk and agency volume was up80 percent last year from 2006, says James Harris, the firm’s head of program trading. A proportion of that growth resulted from significant improvements to its trading technology. In addition, the firm has seen a similar increase in list trading going directly to the algos, according to Silverman. “The ratio of what we’re doing in program trading-electronically versus our overall number-has risen twofold,” he says.
Morgan Stanley is not alone. With increased buyside algo usage, much of the agency program business has been leaving program desks to become the algo business at sellside firms. As the buyside started using electronic tools and executing trades long ago, sellside firms everywhere adapted and provided algorithms for baskets.
For its part, AIM Investments will always need the sellside for some agency program trades, Galloway says-but mostly for technical support, not execution. Galloway chooses between several strategies and more than 70 algos to execute orders. He needs the sellside to ensure that AIM’s algorithms work properly, that he has a thorough understanding of what each algo does and that he knows what algos different firms offer. The relationship changes, he says, but doesn’t disappear.
“At the end of the day, from my perspective, it is still a relationship business,” Galloway says. “People still like to talk to people.”
Vendors Respond
EMS builders also note that the increase of program trading on the buyside is affecting their product development. Although they serve both sides of the Street, EMS firms have gone to great lengths to take the buyside’s needs into account, Kedia says.
At FlexTrade, the buyside now represents almost two-thirds of its client roster, according to Kedia. His firm’s EMS development is more focused on buyside-specific requirements-such as trade-cost analysis and the ability to design customized algorithms, as well as portfolio-level algorithms using various models.
“In terms of [FlexTrade’s] research efforts spending,” he adds, “it’s more on the buyside now, relative to the sellside.”
Buyside traders have been looking for advances in portfolio algorithms to take even more control of their orders. According to a 2007 TABB Group report on institutional equity trading, 65 percent of large-and 60 percent of midsize-buyside firms say they are using portfolio algorithms, either themselves or, more commonly, through their brokers.
Algorithms trace their origins to program and proprietary trading desks. Fittingly, a majority of bulge bracket and other large brokerages have been busy developing better portfolio algos-the next level in the algo space.
According to TABB, they are designed to execute “a strategy based upon selected correlations of multiple assets-assets relative to one another, to benchmarks or to hedging instruments.”
Portfolio algos that balance buys and sells or excel at maintaining specific neutralities, like balancing sectors, potentially could push more agency volume into the electronic camp, says Lehman’s Fagen. But because many portfolio algos are relatively difficult to use, he adds, complex agency orders will still reach sellside desks to tap their high-touch proficiency.
SIDEBAR to Main Story:
Joined at the Hip
Large brokerages have seen increased algorithm use in the portfolio business push their firms’ electronic and program trading desks into closer alignment. At those firms, the trend has led to gains in efficiency and reduced costs.
Programs and algorithms are closely intertwined on the electronic trading desk at Sanford C. Bernstein, says David Liles, its global head of electronic trading, a department which encompasses both. This two-pronged coverage has allowed his desk, which was formed almost five years ago, to integrate its pre-and post-trade analytics tightly with its algos and other quantitative trading tools.
“It’s a little more of an integrated effort than you’ll find at some other firms, and it works to our advantage,” Liles says. “It’s not two competing groups with different agendas trying to pull clients in multiple directions.” Both desks were growing in 2007, he notes. But Bernstein’s algo business has been growing noticeably faster than its program business.
Merrill Lynch has also seen the two desks grow together, says Bill Stush, a director within portfolio and electronic trading. The firm runs both a portfolio and an electronic desk, but considers them to be one functioning unit with the same product and sales management team. And when a program order comes to Merrill Lynch, there’s no fighting between the two for it. “The order and client direction should decide the execution venue, not a particular coverage group,” Stush says.
And algorithmic trading has been filtering down to the mid-size, lower-volume brokerages that do programs, because it’s proven more cost efficient, says Jay Bennett, Greenwich Associates’ managing director. Accordingly, he’s seen more mid-sized firms combine their program and electronic desks. By James Ramage
(c) 2008 Traders Magazine and SourceMedia, Inc. All Rights Reserved.
http://www.tradersmagazine.com http://www.sourcemedia.com