Traders on the floor of the New York Stock Exchange have long scoffed at the “black boxes” that electronically match buy and sell orders on almost every other stock market in the world.
By Justin Schack
June 2002
Institutional Investor Magazine
Now they’ve built one of their own.
Next month a system backed by three of the floor’s most powerful members will begin allowing institutional investors, major brokerage firms and floor traders to match block trades of NYSE stocks in an electronic crossing network - just what the folks who run around in trading jackets have derided for decades. Their opposition has prevented the exchange from moving more swiftly to embrace technology that investors have long believed would make the market more efficient.
The crossing system will be run by Linx, a company founded in November 1999 by Richard Rosenblatt, who has spent 33 years as an NYSE floor broker. Linx has financial backing from LaBranche & Co., the exchange’s largest specialist firm; CEO Mike LaBranche sits on the Linx board, along with NYSE vice chairman Robert Murphy, who headed specialist firm Robb Peck McCooey before selling it to LaBranche last year.
“There’s a lot of inward thinking on the floor,” concedes Linx CEO Joshua Rose. “But these guys are not afraid to think outside that box.”
In the past five years, upstart automated trading networks like Island and Archipelago have roiled the Nasdaq Stock Market, capturing 50 percent of its volume. So far the electronic communications networks, as they’re called, haven’t put much of a dent in the NYSE’s market share. Lately, though, they have redoubled their efforts, posing a bigger threat to floor-based exchanges than ever before.
Some 100 institutional investors trade about 4 million NYSE shares a day - without an exchange or a traditional broker - on Liquidnet, a new electronic matching system. The Nasdaq InterMarket, a service that allows Nasdaq member firms to make over-the-counter markets in NYSE stocks, has garnered about 10 percent of NYSE volume since its launch two years ago. Nasdaq is going after listed order flow with its Primex system, which CEO Hardwick Simmons describes as “essentially the New York Stock Exchange in a box"; Nasdaq is expected to eventually trade NYSE-listed stocks on the SuperMontage execution system it is launching this summer.
“Listed trading is going to be the next big battleground,” says Glenn Schorr, an analyst who follows electronic brokerages and trading systems at Deutsche Bank Securities. “The ECNs already have half the Nasdaq market, and this is the next logical place for them to go to grow their business.”
Linx represents an attempt by Rosenblatt and his allies on the NYSE floor to meet these threats head-on by molding automated trading to the exchange’s needs. Unlike other crossing networks, it is designed to include floor brokers and specialists as well as institutional investors and brokers. All block trades matched on Linx will be executed, or “printed,” on the NYSE. The electronic engine will match buyers and sellers strictly according to the number of shares of a particular stock they want to trade. Once a match occurs, the prevailing price on the floor is attached to the trade, and the transaction is routed electronically to a floor broker who executes it as a “cross” at the specialist post for that stock. The process is not unlike an automated version of the crosses of NYSE-listed shares that are performed every day by institutional brokers on upstairs trading desks and then processed on the floor.
This hybrid structure - which essentially integrates a crossing network with the continuous, human auction - is meant to appease a diverse group of market players. Institutional investors have complained for years that retail investors and floor traders dealing in 100-share lots on the exchange floor make it impossible for them to efficiently execute their big orders. Those complaints have only grown louder with the advent of decimal pricing, which allows quotes to fluctuate far more frequently than they did when shares were priced in fractions. Linx’s network gets around this problem by connecting buyers and sellers of large blocks according to the volume they seek to trade. The anonymous cross makes it less likely that institutions will telegraph their intentions to the rest of the market and cause price movements that increase implicit trading costs. But because prices are determined by the NYSE auction, institutions don’t run the risk of missing out on the best price by executing away from the exchange’s central pool of liquidity, as they do in a traditional crossing network.
The Big Board doesn’t own or operate Linx, but it supports the experiment. It stands to benefit substantially if Linx is successful, because institutions will be less likely to seek liquidity off-exchange, on ECNs or other alternative systems. Any volume lost to competitors means a decline in fees from executing trades and selling market data; these now make up 35 percent of the NYSE’s revenues.
Linx represents a major departure for the NYSE floor, where change comes slowly. Rosenblatt first dreamed up the idea for the network more than a decade ago and promptly ran it by NYSE chairman and CEO Dick Grasso, who at the time was president under chairman Bill Donaldson. “Dick was the first guy I showed it to,” says Rosenblatt. “He thought it was a good idea.” So good, in fact, that Grasso asked if the exchange could develop and operate the system itself. Rosenblatt demurred, saying he would rather build it privately, but at the time he found few others interested in helping him.
Recently, with ECNs luring away Nasdaq volume and institutional investors complaining more loudly about the costs associated with executing large trades (Institutional Investor, April 2002), Rosenblatt found backers in Murphy and LaBranche. Early last year Rosenblatt hired Rose, a former Goldman Sachs block trader and one of the founders of Investment Technology Group’s popular Posit crossing network, as chief executive.
The system is set to begin beta testing with a small group of customers next month, while Rose works to line up more participants. He would prefer to bring major institutional brokerage firms on as strategic investors that could enter orders into Linx on behalf of their customers, but he is prepared to market directly to institutions if the sell-side firms balk.
Rose has held regular meetings about the new system with Grasso and other senior NYSE officials, whom he describes as “supportive, but in an arm’s-length sort of way. We will have to integrate some of our technology with theirs, but they are not doing anything for us that they wouldn’t do for another member firm.”
To be sure, plenty of ambitious trading systems aimed at institutional investors have appeared and folded in recent years. One big unanswered question regarding Linx is whether institutions will shun the system because the orders it matches electronically are briefly exposed to the “crowd” on the exchange floor before being processed, giving other traders the opportunity to improve the price for either the buyer or the seller. Although such price improvement has long been the reason so many believe that the NYSE’s human auction is superior to fully automated systems, it may turn off users who want quicker turnaround times and less information leakage. Linx may also face a perception problem: Institutions are livid about floor traders exploiting penny price increments to trump large institutional orders; they may recoil at the prospect of a trading system conceived, owned and operated by floor personnel.
But regardless of whether Linx revolutionizes the NYSE, the simple fact that the network exists represents a major step forward in the thinking of Big Board floor traders, many of whom would rather wait another few decades for change.
