Editor's note: This is part 2 in a two-part series on electronic stock trading and how the relationship between the buy and sell sides is at a turning point. As the buy side takes greater control of its trades, the sell side is looking for new ways to add value. Last week (E-Tools Energize Buy-Side Trading), we looked at the buy side's perspective; this week, we look at the sell side's.
After unleashing a multitude of algorithms into the buy-side trading community--which empowered buy-side clients to take control of their own trading but often left them confused about which algorithms to use--the sell side is fine-tuning its quantitative tools and retraining its sales traders to offer continued value to buy-side clients. Many bulge-bracket firms--the major brokerage houses that underwrite and distribute securities as well as produce research--are taking on a consulting role, helping buy-side customers choose algorithms. Brokers say they'll advise buy-side firms on which electronic strategies to apply for particular trading styles and develop customized algorithms, as well as pre- and post-trade analysis tools, for clients.
Algorithms enhance the sales trader's ability to provide market color, says Andrew Silverman, head of U.S. algorithmic trading at Goldman Sachs.
The brokerage firm uses the Cube to advise buy-side clients on "how to segment their flows scientifically, based on characteristics of stocks," says Jana Hale, managing director and head of global algorithmic trading at Goldman Sachs. Some of the standard advice can be automated and incorporated into order-management systems. Other times, it's provided by a broker, Hale says.
Sell-side traders are touching less order flow these days, as users gravitate toward algorithms and other low-cost trading venues.
Streetwide estimates suggest that 50% of total buy- and sell-side order flow is passing through electronic channels--a combination of program trading, direct market access, algorithms, and crossing networks. Where the high-touch broker can add value is on the remaining 50%, although that portion "is going to decrease as more users go on board with algorithms," Hale says.
The increase in model-driven trading appears to be forcing brokers to retrain their traders and sales traders. While predicting that algorithmic trading wouldn't replace human traders, Sang Lee, managing partner at Aite Group, wrote in a March report, "We are witnessing the birth of a new type of trader: the Execution Consultant who is part [quantitative]/execution expert capable of providing high-touch services to maximize the potential benefits that can be gained through electronic trading."
The recognition by brokers that the buy side isn't adopting algorithmic trading as quickly as brokers expected may help drive the consultative trend. A buy-side algorithmic trading study conducted by Financial Insights and released in May found that only 5% of total buy-side order flow is being pushed through algorithms. The study surveyed 60 of the top 477 investment managers, including hedge funds. While large firms and quantitative firms are the heaviest users of algorithms, Financial Insights contends that actual usage among most firms is quite modest. (Editor's note: A recent Tabb Group study reported that 11% of buy-side order flow is captured by algorithms. Financial Insights used median calculations, which reduce the impact of outlying firms, the small number of firms that use algorithms substantially more than everyone else. Tabb Group used average figures.)
"Algorithmic trading is everywhere on the Street now, but the problem is, it's undifferentiated. It's hard to know one provider's VWAP [volume-weighted average price] from another's," says Randy Grossman, research manager of institutional trading and investment management at Financial Insights. "The fact that there are so many providers pushing their algorithms and their product has led to confusion about what algorithmic trading is about and what it can provide."
So far, the buy side mainly is using basic algorithms, such as volume-weighted average price and time slicing, predominantly on large-cap stocks, Grossman says. "They haven't ventured too far in the realm of creating their own algorithms or customizing algorithms, even though those services exist at most major brokers," he says. Also, the buy side is focused on the productivity gains and the anonymity as benefits but hasn't looked at algorithms as a tool for performance improvement, Grossman notes.
"It's probably the early stages of the art of algorithms," says James Leman, head of execution trading for the Americas at HSBC Securities (USA). "As the use expands, the buy side will want the brokers to counsel them more and probably provide more innovative uses of algorithms."
But if the buy side is relying more on algorithms and direct market access to execute orders, what's the broker's value proposition going forward, and how is the interaction with the buy side going to change? And what does the future hold for the sales trader?
"The broker's value proposition is as technology provider," says Robert Iati, a Tabb Group partner. In 2004, large brokers went on a direct-market-access buying spree: Banc of America Securities bought Direct Access Financial; Citigroup acquired Lava Trading; Bank of New York purchased Sonic Financial Software; and Piper Jaffray acquired Vie Systems. Additionally, in 2000, Goldman Sachs obtained REDIPlus through its acquisition of market maker Spear, Leeds & Kellogg. Within three years, direct-market-access technology has become a commodity at a faster rate than ever before, Iati says. As a result, the sell side still is searching for the value proposition for the next generation of its business, he adds.
In fact, brokers may have no choice but to reinvent themselves since the traditional role of the sales trader--to offer market information and analysis--has diminished as a result of electronic trading. "Broker-dealers don't have as much market color on individual stocks because orders are going down fragmented information channels," Goldman Sachs' Silverman said at the trading technology conference. To address these information silos, the sell side is giving the buy side tools, such as algorithmic trading, to analyze the data, he added.
Because stocks are trading in smaller quantities, creating more trades, and across more venues, the job of the trader is that much more difficult. It's challenging for a human to assimilate and analyze information on thousands of stocks, says Silverman, in an interview with Wall Street & Technology, InformationWeek's sister publication. But an algorithm can sort through thousands of issues and give information on individual stock or sector trends in real time, he says.
So does this mean that algorithms are replacing the sales traders in their role of providing market analysis?
"It's not replacing the sales trader or the information," Silverman says. "I would say it's enhancing it. The sales trader is still there, giving you the high-touch service, but now there are a lot of tools [that] can add additional types of color that it would not have been possible to push to the customer probably three or four years ago."
For example, Goldman Sachs and other brokers are creating tools that will choose the best algorithm for a particular execution.
Similarly, Deutsche Bank is "going to give customers the ability to send their orders to an algorithm that will choose an algorithm based on liquidity measures and urgency," said Gregg Sharenow, director of execution marketing group at Deutsche Bank, at the same trading technology conference.
The changes brought about by electronic trading aren't "the demise [of the sales trader], it is reinvention," Goldman Sachs' Hale says. The firm has expended a tremendous amount of effort over the past five years reinventing the role of the trader and sales trader, she says. While this person focused on single shares 10 years ago, today the trader services the client on multiple products and has deeper knowledge of options and options volatility, how to trade portfolio trades, and when one strategy is better than another, Hale says.
Institutional sales staffs will be more computer savvy and quantitatively oriented, brokers say. "People spend a big chunk of their day interacting extensively with these tools, and they're getting their arms around the newer tools--algorithms in particular--[learning] what their strengths and weaknesses are," says Dave Cushing, managing director of execution services analytics in Lehman Brothers' equities division.
After spending tens, if not hundreds, of millions of dollars on building algorithms and the underlying market-data infrastructure to support the tools, brokers plan to apply the same disciplines to futures and options trading via the same platforms. "Algorithms are going to warp across asset classes into [foreign exchange] and fixed income," says Raymond Killian, president, chairman, and CEO of Investment Technology Group, an agency broker that also provides equity-trading services and transaction research.
Unlike bulge-bracket firms that have investment banking and mergers-and-acquisitions activity to generate revenue, ITG has to focus on the electronic side. "Our skills are the intellectual-property content of our technology. We spend 23% of our budget on technology--it's our only weapon," Killian says.
In their role as execution consultants, brokers will be in a race to roll out new functionality. "Brokers will become aggregators of financial technology," predicts Robert Flatley, managing director of electronic trading services at Banc of America Securities. "Because we're now the owners of those applications, there's interest in having [direct market access] interoperate with program trading, algorithmic trading, FIX [Financial Information Exchange] order routing, and the existing blotter on the desktop," he says.
To add value, Tabb Group's Iati says brokers need to integrate the various components of electronic trading. Say the buy side is using Lava Trading, Credit Suisse First Boston's algorithms, transaction-cost analysis from Plexus Group, and research from Lehman Brothers. "The buy side has all these tools on their desk. It's confusing for them and their workflow isn't easily manageable," Iati says. "Wouldn't it be good to have a seamlessly integrated order-management system, direct market access, algorithmic trading, research, pre-trade and post-trade analysis?"
Financial Insights' Grossman points to the need for more product differentiation, education, and documentation on the performance of algorithms from brokers. "The battle is going to be fought on the transaction-cost-analysis front," he predicts. Transaction-cost analysis "is part of the education equation. You need to show people when they should be using algorithms, which algorithms they should be using, and the way to show that is through the pre-trade and post-trade transaction analysis."
As these quantitative tools continue to evolve and the buy side becomes more comfortable using algorithms and transaction-cost models, will the consultative role of the broker still hold? "I don't think that technology is going to lead to disintermediation or job destruction, not by a long shot," says Lehman Brothers' Cushing. Right now, he adds, "the buy side is feeling a little overwhelmed by having to assimilate all this technology and is looking to the sell side to provide solutions. So our role has been enhanced to the extent that we have been asked to help navigate this new world order."
Image by Gerald Bustamante/Veer