ECN: A new force in the stock markets? (Global Securities).

Editor's note: Traditional stock exchanges, led by the New York Stock Exchange, remain the most visible and prestigious channels for trading securities in this country. Electronic communications networks (ECNs), however, represent an emerging band of competitors. This article looks at a number of the major entrants, assesses their strategies and attempts to understand why this movement developed and what it could mean to corporate finance executives

David Whitcomb joined the faculty of New Jersey's Rutgers University in 1975, pioneering the study of market microstructure and tinkering around with financial models on the side. He rose through the ranks to full professor, but it was his move out of academia that has really put him on the map.

Whitcomb's tinkering at Rutgers allowed him to benefit in a big way from the regulatory and technological sea change that is reshaping the Nasdaq securities market. A few years ago, he founded a company called Automated Trading Desk that applied his models to the markets; the models help him predict short-term stock price shifts. His company trades about 50 million shares of stock a day, and he estimates profits for the year 2001 at around $50 million, on revenues of less than $100 million. Most of those trades are in Nasdaq stocks.

That's not bad. But the big game, Whitcomb says, has yet to get underway. He's licking his chops at the prospect of applying his technology and models to the listed stock market. "That marketplace needs reform," he says, "Spreads are now wider in the [listed market] than in the Nasdaq for comparable stocks, an absolute flip-flop from two years ago. I'd like to see some ECN (electronic communications network) or ECNs change the way business is done in the listed market."

He points to one spot of hope. This March, the Pacific Stock Exchange closed its trading floor and shifted its listings to Archipelago, an ECN that had become an official exchange. Through its relationship with the Pacific Stock Exchange, Archipelago gained the right to trade not only Pacific but also New York Stock Exchange listings.

A spokesperson for the NYSE declined to comment on this development. "We don't talk about our competitors," she said. But others in the industry expect that the stage may be set for a new wave of change at the very heart of the stock markets.

"If you look at what will be the key goal for ECNs in the next year or two, it will be delving into NYSE types of stocks. We're late in it, but all of us are focusing on it," says Joel Steinmetz, senior vice president of Instinet, a New York-based ECN.

Chris Keith, now president and CEO of Exchange Lab, invented and implemented the New York Stock Exchange's electronic trading system, called DOT. After retiring from the job of chief technology officer at the NYSE, he set himself to solving a riddle. Given that electronic trading was more efficient in processing transactions, why hadn't ECNs been able to take more market share away from the New York floor?

His answer was that ECNs weren't able to duplicate the flexibility of human traders. But that's changing, he says. "Look at what an ECN does now!" he says. "Rather than one or two different kinds of orders, an average ECN might have 15 or 20. I think the [big exchanges are] going to begin to lose market share at this point, simply because of the fact that the advantages of their systems are being replicated by electronics -- and were never replicated before."

At Exchange Lab, Keith has been working to design and implement new technologies that will bring trading up to date. He isn't alone. The past few years have seen explosive innovation in the development of markets and trading systems. It began with ECNs, but is rapidly moving beyond them to a new frontier.

The implications for corporate finance are far-reaching. Most immediately, Archipelago's debut as an exchange means more competition for listings. Earlier this year, Archipelago lured away Nasdaq's former senior counsel for listings, David Strandberg. As Archipelago's new director for issuer services, Strandberg will spearhead efforts not only to attract initial public offerings, but also to persuade companies listed elsewhere to switch to Archipelago.

Competition promises good things for corporations facing a decision on where to list. Archipelago plans to offer CEOs and CFOs an array of electronic information services unparalleled by any other exchange. "We're able to offer quite detailed information products that would be of very significant strategic value for management of our listed companies," Strandberg says. "If I'm a CFO racing to the airport and I get an email alert that my or my competition's stock price has gone up or down, or that there's a big news announcement affecting my company -- that would be a very useful tool." Companies that list on Archipelago might also look forward to real-time reports on how their stock is trading or has traded -- even reports on who is buying and selling. No other exchange offers this kind of strategic information.

Strandberg's feet are on the ground, but he's thinking big. "It's unrealistic to expect IBM or Microsoft to transfer their listings to us in the short term, but we're going to be a very viable, credible alternative, and think that over time we'd be competitive in that regard," he says. Companies that want Archipelago's enhanced services but are unwilling to abandon their NYSE or Nasdaq listings might, he suggests, consider a dual listing that treats Archipelago as "a second electronic home."

The First ECNs

Although the subject of market structure may seem arcane and technical, the consequences of change will impact anyone who buys, sells or issues stock.

"I think the clearest case for [caring] is the collapse of the bid-asked spread," Whitcomb declares. The spread is the difference between the buying and selling price of a stock. When spreads are high, buying and selling stock is relatively costly. When spreads fall, investors and companies that issue stock both stand to benefit. "Individuals can really hold stock at almost no cost, other than the financial capital invested, because it doesn't cost much to buy and sell," Whitcomb says. This translates into more demand for stock, which can translate into lower stock prices and cost of capital for corporations.

The NYSE and the Nasdaq are very different markets. The NYSE's system of trading floor and specialists dates back more than two centuries. Despite some impressive technological innovations, it's essentially the same sort of auction market it's always been.

The Nasdaq, in contrast, is a newer model. The Nasdaq market was created by an association of traders dealing in what used to be called over-the-counter stocks. In the 1970s, these dealers started posting the prices at which they would buy or sell on a system called "National Association of Securities Dealers Automated Quotes" -- Nasdaq for short.

But a scandal erupted in the late 1990s when the public discovered that the Nasdaq game was rigged. Only small investors paid the prices publicized on the Nasdaq, where market-makers colluded to guarantee themselves high spreads. The dealers traded among themselves at better prices and narrower spreads on an electronic trading network called Instinet. A barrage of lawsuits followed this discovery. The dealers ended up settling a class-action lawsuit for $1 billion, but if that were the only cost they'd faced, they'd have gotten off easy.

In order to ensure that collusion didn't happen again, the SEC mandated changes in order-handling rules. The result of the changes was to open the Nasdaq market to new competitors -- ECNs and alternative trading systems (ATS). These new competitors will completely change the nature of the Nasdaq market.

Instinet has actually been around since 1969, and has been owned by Reuters Group PLC since 1987. The first alternative ECN to emerge was Island, the creation of a group of so-called "SOES bandits" who had used the Nasdaq's Small Order Execution System (SOES) to profit from the mistakes of market-makers. Later, SOES bandits would be more widely and popularly known as day traders. Pressured by its big institutional clients, Instinet had refused to allow these traders to use its network. So the traders built their own network, and called it Island.

When the SEC's new order-handling rules became operational in 1997, Island found itself on a level regulatory playing field with Instinet, and soon eclipsed its rival. By January 2002, Nasdaq statistics identified Island as the biggest ECN, with 11.3 percent of share volume and 23 percent of trades. Instinet, by contrast, stood at 9.8 percent of share volume, and 10 percent of trades. In early April, Instinet CEO Douglas Atkin resigned, and published reports said Reuters was actively seeking to sell the company or merge it with another trading operation.

Island has changed with time. Andrew Goldman, its executive vice president, says, "Day traders are now eclipsed by a broad range of market participants - buy-side firms, quant [quantitative] shops, proprietary trading shops." On some days, he says, Island does 40 percent of the volume in the QQQ security, designed to mimic the performance of the Nasdaq 100.

So far, however, the upstart ECN hasn't been able to seize the brass ring it has been reaching for since 1999 -- official status as a full-fledged stock exchange. Recently, the SEC bounced Island's application back with requests for revisions.

Meanwhile, though, some of Island's most formidable competitors have managed to seize that prize.

Archipelago

Archipelago began almost by chance. Stuart and MarrGwen Townsend had met at the University of Chicago in the late 1970s, where they were studying economics. They worked together programming computers to run complex economic models, starting with mainframes and following the curve of technology through minicomputers and beyond. They founded their own company, Townsend Analytic Systems, and during the 1980s and 1990s took on assignments designing information and trading systems for exchanges in the U.S. and Europe.

In August 1996, when the SEC announced a change in order-handling rules that would be effective in January 1997, the Townsends didn't see the implications immediately. MarrGwen recalls it was only in November that they realized that they could create their own ECN using technology they had developed. Moreover, they could offer something that no other trading venue offered -- access to all pools of liquidity.

Island, as the name implies, was somewhat cut off. Although it was a relatively deep and liquid market in its own right, sometimes orders sent there did not find a match. When this happened, Island routed them on to the Nasdaq, where they might or might not find a match. But if a match happened to be sitting in another ECN, it was inaccessible.

The Townsends developed a system that would treat all of the ECNs and Nasdaq as if they were simply one vast market, tying together the various islands of liquidity into an archipelago of trading opportunities -- hence the name.

The idea caught the attention of major investors. Merrill Lynch & Co., Goldman, Sachs and Co., E*Trade -- even Instinet itself -- all backed the Archipelago concept with capital.

Although Archipelago didn't get off the blocks as fast as Island, its slow-and-steady approach is paying off.

Archipelago bought the equities trading arm of the Pacific Stock Exchange in the fall of 2000. A year later, it got regulatory approval to become an exchange. And a month later, it announced a merger with RediBook, an ECN established by Spear Leeds Kellogg. Goldman Sachs had bought Spear Leeds Kellogg a few months earlier, and the roster of equity holders in Redi includes Fidelity Investments, Lehman Brothers and Credit Suisse First Boston.

Archipelago and RediBook had been the third- and fourth-largest ECNs in the Nasdaq market. Their merger made them the largest in total share volume.

Status as an exchange brings considerable advantages, Stuart Town-send says. As an ECN, Archipelago was in the anomalous position of being regulated by its competitors in the NASD. As an exchange, it will be regulated by a self-governing organization. It has access to the other perquisites of exchange status -- most notably, the intermarket trading system that links the regionals and New York, and tape revenue (the money exchanges are paid every time they put a trade on the data feed). As an exchange, Archipelago will also save millions of dollars in clearing costs.

Nasdaq's Supermontage

Although Archipelago has been the first, and so far the only, ECN to achieve exchange status, another applicant is waiting in the wings -- Nasdaq itself.

"People have said we're too closely tied to the broker-dealers," notes Scott Petersen, a Nasdaq spokesman. "We're moving toward complete independence." Petersen expects the SEC to approve Nasdaq's application for exchange status in the first half of this year. "Exchange status will allow Nasdaq to tap the capital markets and strengthen our competitive position," he said, "It's an essential step on the road to an IPO."

At the core of its new exchange system will be Supermontage, a system that -- if it lives up to its billing -- will display the five best bids and offers for every stock, allow users to specify three different kinds of orders and allow for anonymity for those placing orders, an important consideration for major traders.

Last year's implementation of decimal pricing was a factor in creating this new opportunity. Ever since the 16th century, brokerage firms, specialists and market-makers had locked in some level of income with fractional spreads based on one-eighth increments. The move to decimal pricing collapsed spreads. "Decimalization has had a negative effect for profit margins of major brokerage firms. They're getting squeezed by ECNs to a penny," says Kevin Connellan, director of equity trading at Northern Trust Global Investments in Chicago, "So they've gone to agency trading -- they buy for you and then charge you a commission, so the actual prices you get will be a lot more transparent. ECNs continue to gain market share partly because there's continuing pressure on buy-side houses to search for the best execution."

But the ECN business itself is getting tougher, more competitive and more commodity-like. "My sense is that a lot of the ECNs in one way or another are all coming down to the same thing, starting to look alike. What you'll be facing as a trader is whether you're going to work on Supermontage or work the ECNs yourself," Connellan predicts. Working them by yourself presents a challenge, though -- there are so many. "A trader only has two arms and two eyes," he notes.

Beyond ECNs

But the same kind of technological innovation that brought us ECNs is bringing a solution to ECN proliferation and the market fragmentation that has ensued. Connellan uses a system from Lava Trading Inc. to aggregate all quotes from all ECNs. Others are also preparing technology for a market that is rapidly moving beyond ECNs. Exchange Lab's Keith is working on a new system, called PDQ, which aims to overcome the proliferation of ECNs, trading systems and exchanges.

"Suppose you had a TV set, and every time someone had a new TV program they had to drop a different line to your living room, put a different box in your closet, and give you a different remote with a different methodology?" he asks. The analogy isn't too far-fetched; different trading systems and ECNs do require their own software and often their own hardware. "We're not trying to tout one method over another," Keith says, "We're trying to create an open platform. We're in the Time-Warner cable business -- not the program business."

As Yogi Berra famously observed, predicting is tough -- especially when it's about the future. But there's little doubt that the re-shaping of the Nasdaq markets by ECNs was merely the first phase in an overall market reform that has plenty of wind left.

Gregory J. Millman is a freelance business writer in New Jersey and a frequent contributor to Financial Executive.