The pace is slowing, but the volume of institutional order flow bypassing block desks at U.S. brokers continues to grow. That's the upshot of a new TowerGroup report about trends in direct market access. Tower estimates that one-third of institutional volume was traded via front-end systems last year. That's up from 11 percent in the first year of the decade.
That's not good news for brokers. They've been able to fashion DMA as a service by virtue of their memberships on Nasdaq and the various exchanges and by providing the software. But commissions are low. DMA trades earn them only a penny per share versus a nickel for traditional worked orders. The numbers look even worse when algorithmic trades are included. TowerGroup estimates that another seven percent of share volume is flowing self-directed through brokers' algorithmic trading servers. Tower counts those trades separately because the commission structure is different. Self-directed algorithmic trades earn brokers two cents per share.
The Tower report does predict a respite for brokers from the rapid run-up in direct market access. The consultancy forecasts growth in DMA this year at nine percent, slowing next year to 5.5 percent. Those numbers compare to double-digit growth in each of the first five years of the century. That silver lining for brokers comes with a caveat. The trades that do land on brokers' block desks are likely to be the tougher ones, according to the report. That may not be healthy for desks' profitability.
The Tower report was authored by Dushyant Shahrawat, a senior analyst in Tower's securities and capital markets practice. Shahrawat has been with the consultancy since 1998, covering buyside order management systems, offshore outsourcing, Linux and Web services and other areas.
Traders Magazine technology editor Peter Chapman interviewed Shahrawat.
Traders: One-third is a lot of order flow.
Shahrawat: It took a lot of time to put that number together. We did that very gingerly. Nobody seemed to have the breakdown although we talked to some very large buyside firms. Two of the top five money managers in the world didn't even have a very good idea as to how total buyside order volume breaks up by strategy. They had back-of-the-envelope analysis. So aggregating was quite a challenge. It could be off, but no more than five to 10 percent.
Traders: How much of that is hedge funds and how much is traditional money managers?
Shahrawat: About 55 percent is hedge funds and 45 percent is long equity. But that is changing. For two reasons. First, most of the upswing in trading volume from hedge funds is behind us. The hedge fund market will not grow by leaps and bounds as it has over the last three to four years. Also, long equity management firms are continuously automating. Their percentage of the total DMA market will probably start to inch up. I wouldn't be surprised if that 55/45 split is reversed.
Traders: You report that Morgan Stanley's split is 60/40 hedge fund/traditional.
Shahrawat: Right. That comes directly from them.
Traders: Within the long equity category, where do quants stand?
Shahrawat: DMA is about 30 percent to 40 percent of their trading. It's not tremendously high.
Traders: But still higher than that of a mutual fund?
Shahrawat: No question.
Traders: So within that long equity segment, quants are the most aggressive users of DMA?
Shahrawat: Correct.
Traders: The buyside wants DMA primarily to reduce commissions and not for best execution?
Shahrawat: Yes. Most buyside firms haven't done the analysis to understand the implicit and explicit costs of their trading. They simply look at commission expenses. They aren't motivated to think in terms of fully-loaded expenses – just purely explicit costs. Looking at the implicit trading costs requires a different set of tools and transaction cost analysis. We're not there yet.
Traders: In the past, the buyside was overpaying for some executions?
Shahrawat: Yes. And now they are getting regulatory scrutiny. Unbundling isn't taking place because buyside firms have discovered religion overnight. Regulators and their largest customers are saying: You aren't keeping our best interests in mind. You charge me a percentage and you couldn't care less what you pay in commissions.' Because all trade commissions are just charged back to the client. So there is pressure on the buyside to be more diligent about ensuring they are getting the best price possible. There is no conclusive analysis [that the buyside has been overpaying], but clearly that is the consensus. They don't spend adequate time and attention in analyzing the quality of their executions. A few firms such as Fidelity do, but not the majority.
Traders: Are DMA trades the easier ones?
Shahrawat: There's no simple answer. By definition, DMA is a straight-forward strategy so some of the simpler trades are put through that pipe. But it is difficult to discern patterns in that volume. Are they the more liquid or less liquid stocks? What is the trading volume over the course of a day? Is it seasonal? Are there certain sectors that tend to be traded more through DMA? That requires access to actual trading volume from the investment managers. We were not privy to that. So, I can't conclusively say that DMA trades are easier. But my instinct is that it is the more liquid stocks, simpler trades, that are pushed through DMA.
Traders: You report DMA has become a commodity. That means all systems are the same?
Shahrawat: Yes. A pipe is a pipe is a pipe. Five years ago just adding a pipe was value added. Not anymore. Everyone does it. That's why people are moving into algorithmic trading.
Traders: Firms are catching up in algorithms also.
Shahrawat: That's right. Not every broker has the capability. But by the end of 2005 or into 2006, everyone will have virtually the same algorithms. So that will no longer be a differentiating factor.
Traders: Then firms are trying to differentiate within DMA?
Shahrawat: Well, you shouldn't think about DMA as an animal in and of itself. The buyside doesn't get excited about doing DMA trades. They get excited about getting the best price possible and lowering their execution costs. DMA is just one element of that overall strategy. DMA should be considered as it relates to other trading strategies. As one part of a larger group of services. Too many people miss that point. And that point is very critical.
Traders: So the sellside is more concerned with providing a range of services, and not devoting too much attention to a single product?
Shahrawat: Having the best DMA product in the market means nothing. Large buyside firms want brokers to offer them a range of services for their different trading needs. So that is why individual brokers have not over-invested in DMA. They recognize the importance in being able to spread the dollars.
Traders: That would suggest that second tier institutional brokers just offering DMA are at a disadvantage? They were leaders in DMA at one point, but then the bulge bracket caught up?
Shahrawat: Brokerages that don't have a block trading desk, that provide purely electronic trading, are at a disadvantage versus the Morgans and the Goldmans.
Traders: You report the bulge bracket has wrested control of DMA from second string brokers and vendors?
Shahrawat: Yes. That is very important. The brokers are where the action is. The brokers have been buying up all the independent players. So you have to focus on the larger guys. The market cannot sustain the number of brokerage firms pushing agency trading. Eventually there will be consolidation in the market.
Traders: Lava was bought. Sonic was bought. Instaquote was bought. Where does that leave the second tier brokers without direct access?
Shahrawat: They're up the creek. Either you use a white label product from someone like Citi and take the risk they won't look at your order flow… They promise they won't, but can they resist that urge? And you cannot not offer these products and services. They are almost required.
Traders: Doesn't sound good. What's new on the DMA horizon?
Shahrawat: Options. At Tower, we've been approached to investigate the potential for DMA in options trading. Expansion out of equities and into other product areas is a crucial issue.
TM: O.K. Thanks, Dushyant.
(c) 2005 Traders Magazine and SourceMedia, Inc. All Rights Reserved.
