As soon as Regulation NMS began to look like a reality, exchanges were off to the races, merging with the predominant electronic communications networks (ECNs) and alternative trading systems (ATSs) to build up market share. The ECN and ATS operators needed the exchanges in order to acquire an air of legitimacy as well as to participate fully in the generation of market data and the collection of market data revenue. Caught by surprise, brokers soon followed with the next wave of acquisitions.
The last two surviving ECNs from the initial wave of acquisitions, Attain and OnTrade, part of NexTrade Holdings, were acquired, respectively, by Knight Capital Group in October 2005 and Citigroup earlier this year. The motivations seem fairly clear. First and foremost, firms want to strengthen their positions against the consolidated New York Stock Exchange and Nasdaq, both of which - through the NYSE's merger with Archipelago Exchange and Nasdaq's acquisition of INET - essentially have recreated the duopoly of transaction venues that spawned the first wave of ECN creation in the late 1990s. As a second order of business, brokers are interested in competing with some of the block-crossing networks, such as LiquidNet and Pipeline. Finally, another attraction of buying an ECN that already has ATS status with the Securities and Exchange Commission (SEC) is that it allows brokers to profit not only from executing trades, but also from reporting those trades to the consolidated tape system (CTS).
Although the buy side's increasing use of algorithms and the concomitant interest in direct feeds for data-hungry trading engines would seem to make the production of market data a prime goal for a recent ECN acquirer or builder, the uncertainty surrounding the post-Reg NMS world means that it is difficult to project market-data revenue accurately, and it seems unlikely that a brokerage would start an ECN solely for this purpose, brokerage executives say. The consensus view among brokerage executives interviewed by Advanced Trading for this article seems to be that the approximately $500 million pool of annual market-data revenue potential does not, on its own, justify the $30 million to $50 million it would take to buy or build an ECN.
However, because Reg NMS obligates orders to be routed to the venue with the best price, a small ECN can be the target of all order traffic in the market for a given security at a given point in time, according to Will Sterling, head of institutional electronic trading at UBS. "Reg NMS says you have to route when you have the best price, period," Sterling says. "So it creates a very strong support system for smaller exchanges and ECNs. A single order is justification enough for you to connect to them."
The Economics of ECNs
"Market-data revenue is icing on the cake," says Tom Richardson, managing director in global equities at Citigroup, who is heading up the ECN technology and eight staff members Citi acquired from NexTrade. "It is by no means a given. If you look at the economics of owning and operating an ECN, clearly the ability to charge access fees is the primary motivating factor, and, second to that, keeping transaction fees down for the client benefits everyone, including us."
It had become common practice for ECNs to strike partnerships with exchanges, particularly regional venues such as the National and Chicago exchanges, which would rebate ECNs their transaction fees for printing under the identifier of the exchange on the tape. The added benefit to those brokers such as Citi, which recently invested in both the Philadelphia and Boston exchanges, becomes fairly obvious: Market data revenue benefits the bottom line of the broker both as an operator of an ECN and as an exchange shareholder. Throw in the fact that large brokers now own front-end order-routing and trading tools - for example, Citi has Lava and Lehman Brothers has Townsend Analytics - and it's easy to visualize brokers' response to declining commissions: If you can't get higher commissions or more commissions, then collect a fee at every step of the trade.
Initially, OnTrade under Citi, which is expected to begin operation sometime this summer, will report to the Alternative Display Facility (ADF) operated by NASD, Richardson says. Ultimately, however, it may seek a partnership with an exchange for reporting, he acknowledges, adding that the new OnTrade will be a revamped version of itself that takes advantage of Citigroup's global reach, liquidity and distribution, as well as its ownership of Lava Trading.
As with all broker incursions into formerly independent businesses, pains must be taken to assure the buy side that these new vertical-integration ventures are not simply nickelodeons through which brokers' proprietary desks view client trades. Richardson asserts that the books and records of both OnTrade and Lava are housed separately from any of Citigroup's brokerage businesses.
But skepticism still must be overcome. Even an agency broker-dealer with no proprietary desk "monitors clients to make sure the order flow covers the commission by some multiplier," says Tom Price, an analyst with TowerGroup. "I would imagine the same thing happens with these ECNs. Does the buy side want to go there, and do you, as a broker, want to keep them?" Still, if Reg NMS operates as it should, any liquidity source that offers the best price must execute at that price, no matter who owns it, Price points out.
Other brokers on The Street are circumspect about their ECN plans. Knight executives were not available to comment for this story on their plans for Attain, which now will be called Direct Edge. Goldman Sachs, one of the primary original investors in Archipelago, declines comment.