Electronic trading: portal to the future? Technological advances are moving markets toward an all-electronic environment, thanks in part to the growth of electronic exchanges. But the specialist system is far from dead, and some experts see a continuing need for at least occasional human intervention.
In January, 2004, for example, New York Stock Exchange (NYSE)-listed Cadence Design Systems announced that it was joining the new dual listing program of the Nasdaq Market System. Cadence CFO William Porter explains the company's decision to dual-list as a shot across the bow of an outmoded market structure. "We do support electronic markets, and we think the movement to more electronic markets would be better for all trading. That was one of the reasons for us to try this experiment--to get more of a focus on electronic trading."
Cadence wasn't alone. Five other companies simultaneously announced their decision to dual-list on the Nasdaq in January 2004. A year later, insurance behemoth American International Group (AIG) announced its own dual listing decision--not on the Nasdaq but on the upstart Archipelago Exchange, founded as an electronic communications network (ECN) in 1996. (In addition to AIG, Metropolitian Healthworks, an American Stock Exchange-listed company, and options brokerage optionsExpress, a Nasdaq-listed firm, have dual-listed on Archipelago this year, according to a spokesperson.)
What's going on here? Of course, you could count on your fingers the number of corporations that have opted to make a statement with their listing decision--and you wouldn't even need all of your fingers to do it. Moreover, the decision to dual-list makes little difference in how stocks trade: It is more of a political statement than a financial tactic.
"Other markets have long traded NYSE stocks, and most of these stocks were already traded on the Nasdaq, so I take the dual listing to be a statement of a degree of dissatisfaction," says Hendrik Bessembinder, Professor of Finance and A. Blaine Huntsman Presidential Chair in Finance at the University of Utah's David Eccles School of Business. "These companies were saying they were not totally satisfied with how things were being handled in New York."
Indeed, Porter says that in the year since the dual listing, he has seen little change in how Cadence's stock trades, but quickly explains that's at least in part because Securities and Exchange (SEC) regulations continue to give the NYSE an advantage in the trading of listed stocks. The "trade-through" rule requires that trades go to the market with the best price, but Porter says that 70 percent of the holders of Cadence stock are institutions.
"They're not as interested in best price as in speed and anonymity of execution," he explains. "But as long as the trade-through rule exists, if you have a large institution wanting to sell 500,000 shares, and Mom & Pop want to trade a hundred shares for a penny better, the whole trade gets kicked over to the specialist."
Every NYSE stock is assigned to a specialist whose job is to ensure an "orderly market," providing liquidity by stepping in where angels fear to tread, buying in falling markets and selling in rising ones. Specialists get compensated with privileged access to information and can step in ahead of other investors to "price improve" bids and offers.
ECNs and electronic exchanges, by contrast, have no specialists. Bids and offers meet directly and automatically. Critics have long charged that the NYSE specialists often are less a help than a hindrance, an unnecessary and costly drag on the market. "I think the specialist system has some faults," says M.R. "Hank" Greenberg, chairman and CEO of AIG. "For some companies--sizable companies--it may be okay. But for large-cap companies, it may need some adjustment."
Adjustment is just what it's getting. First scandal, then reform seems to be the rule of thumb in the U.S. markets. In the 1960s, a deluge of paperwork led to paralysis in back offices and ushered in automation on the stock exchanges. In the mid-1990s, scandal at the Nasdaq led to a change in order-handling rules and ushered in the era of the ECN.
Now, in the wake of problems at the NYSE, specialists there find themselves confronting the wolf at the door: ECN-like automatic execution. The NYSE's own hybrid market proposal would offer investors the option to choose between automatic execution at posted prices or the tender ministrations of a specialist.
Noreen Culhane, executive vice president of the NYSE's Global Client Group, says, "The investor can opt for a normal order execution model, with opportunity for price improvement, or for the new automatic execution model, which, in effect, will reduce turnaround time to less than a second and provide for anonymity. But, you are giving up the opportunity for price improvement." Meanwhile, regulators are preparing to change the trade-through rule that has has worked in favor of the specialist system.
It bears mentioning that companies haven't left the NYSE for Nasdaq, merely chosen to dual-list; on the other hand, Nasdaq statistics show that nine companies left Nasdaq to list on the NYSE last year. And NYSE still has the biggest share. A recent study showed that the NYSE handles 80 percent and Nasdaq 55 percent, respectively, of the shares traded of the companies listed on their exchanges.
Regulation National Market System (NMS), a package of reforms proposed by the SEC in 2004, addresses several issues, but none has stirred more controversy than the proposed change in the trade-through rule. While complex in its details, the overall effect of the regulatory reform is simple, says former SEC Chief Economist Lawrence E. Harris, now Fred V. Keenan Chair in Finance Professor of Finance and Business Economics at the University of Southern California's Marshall School of Business.
"The old trade-through rule required that fast-trading markets would have to halt if they would trade at prices inferior to those seemingly available at floor-based markets," Harris explains. "But, in practice, many of those seemingly better prices are not actually available, so the rule prevented and prevents electronic markets from arranging trades."
Regulation NMS will get rid of this rule and substitute an alternative. The alternative will require brokers and exchanges to trade their orders wherever the best price is exposed, if the best price can be accessed instantly. The rule thus reorients the yield sign that governs traffic between the electronic markets and the floor-based markets. Now, the electronic markets have to yield to the floor. If the proposed regulation is adopted, the floor will have to yield to the electronic markets. (In remarks before an FEI audience last month, Nasdaq CFO David Warren said he believes that trade-though revisions "will force the NYSE to go electronic.")
Harris says that there is no doubt the new trade-through rule will promote competition. But some question whether a trade-through rule in any form is necessary. Says Bessembinder: "Most markets in our economy work without the equivalent of a trade-through rule. No law or regulation says that a particular car dealer has to always match the deals being offered by every other car dealer, and yet the car market works reasonably well despite that."
Those who oppose the trade-through rule claim it makes price the most important factor in competition among markets. But some investors, like the institutional investors in Cadence Design, would happily pay a bit more to buy or accept a bit less to sell if they could get other advantages, such as speed or liquidity. Harris dismisses the criticism, noting that investors already go to markets that offer the best prices when the prices are immediately available, and avers that some form of trade-through rule is necessary to protect small investors from unscrupulous brokers.
But even in its modified form, the trade-through rule makes it clear that the fully automated execution of trades is sounding the knell for human intervention in stock trading. And some wonder whether that is entirely for the better.
The March of Electronic Trading
In order to understand where the market is going, it's important to understand where electronic trading has come from, and how fast it has grown. Electronic trading of stocks essentially began with the development of Instinet, the brainchild of an aerospace analyst who wanted to replace the middlemen in stock transactions by using a fully computerized exchange.
A specialist on the Pacific Stock Exchange, William Lupien, saw Instinet demonstrated in 1971 and was one of the first to recognize that it could be the best friend brokers ever had. He talked the Pacific Stock Exchange into allowing him to use it at his trading post, and it worked so well that eventually he got a group of investors together to buy the company. Lupien became Instinet's president in 1983, and turned his attention to automating trading in Nasdaq stocks. At the time, Nasdaq used computers only to display quotes.
Under Lupien, Instinet came to be the unofficial insider's marketplace for the Nasdaq market-makers. In the mid-1990s, it also became the centerpiece of a corruption scandal, when two researchers proved that while Nasdaq market-makers competed aggressively on the Instinet system, the prices they made available to the public on the Nasdaq system were collusive.
Subsequent investigations by the SEC and the Justice Department resulted in new order-handling rules that made ECNs a part of the market vocabulary. The first of these was a trade-matching system called Island, the trading venue of choice for day traders fascinated with then-hot Internet stocks. Island gave traders fast, fully automated execution. With Island, Instinet and the Nasdaq offering liquidity on separate platforms, concerns about market fragmentation were growing. Entrepreneurs in Chicago spotted the potential to start an ECN that would essentially link various pools of liquidity, and called it Archipelago.
Archipelago's early backers included some of the biggest names on Wall Street--Goldman Sachs, J.P. Morgan & Co. and the mutual fund company American Century, among others. In 2000, Archipelago reached an agreement with the Pacific Stock Exchange that aimed to vault it out of the ranks of mere ECNs and make it a full-fledged exchange. In 2005, after an IPO in 2004, Archipelago announced that it was purchasing the Pacific Stock Exchange. Nelson Chai, CFO of Archipelago Holdings, explains that exchange status gave Archipelago some important economic advantages, most notably listing fees, a share in tape revenue and a reduction in clearing costs.
In 2002, Instinet announced that it was purchasing Island, combining both liquidity pools in Inet by 2003. In 2004, the new entity applied for exchange status. Inet CEO Alex Goor sees fully automated electronic trading as the face of the future. "It is inevitable that marketplaces will look like ECNs. It's very hard to see any example in the world where technology comes out and people ignore it. I haven't seen people throwing televisions out and having a piano in the parlor for entertainment. They might add a piano, but they aren't giving away TVs." Archipelago's Nelson Chai agrees. "This is the only place in the world where stocks are traded as they are on the NYSE."
Interestingly, though, recent academic research suggests that there is a legitimate role for human intervention in the markets. In fact, where specialists do not exist, it has recently proven necessary to invent them.
Towards a More Liquid Market
Much of the criticism of New York Stock Exchange specialists focuses on the money that a specialist makes. Says Hendrik Bessembinder, "But to those who are unhappy with a specialist's ability to make money, I would pose a question: 'What alternative method do you suggest for compensating a specialist?'" Indeed, recent research suggests human intervention is important, even indispensable to some stock trades--and not just trades in small stocks, either.
For example, although the Paris Bourse is fully automatic, investors seeking to buy or sell large blocks of shares in French blue chips often prefer to put the trades together away from the automated systems. Says Kumar Venkataraman, a professor at Southern Methodist University who coauthored a study of the Paris Bourse with Bessembinder, "Institutional traders seem to be a little worried about electronic trading systems. A very high percentage of large orders, block trading, goes to upstairs broker-dealer trading desks, where guess who is playing an important role? Human intermediaries!"
For smaller stocks, human intervention can be even more critical. In the early '90s, a number of exchanges in Europe and Canada installed sophisticated electronic systems that allowed for fully automatic trade execution. "There were no intermediaries," Venkataraman says. "They let public traders send limit orders to provide liquidity. But once they started trading like this, they found it was not working, especially for illiquid stocks."
For these stocks, exchanges found it necessary to introduce liquidity providers who functioned something like specialists. By removing the risk of illiquidity, the liquidity providers actually increased the value of the firms whose stocks they traded. Venkataraman found that firm value rose as much as 5 percent as a result.
"ECNs do very well for liquid securities, but for less-liquid stocks, it may be important to designate somebody to provide liquidity," he says. However, there's an important difference in how these liquidity providers are paid. In the European models, since the specialist firm provides a service to the listing company, the company pays the specialist directly for the service.
The work of Venkataraman and others suggests that liquid and efficient markets can have a direct impact on the value of firms. That alone should make market structure an important issue for CFOs. But electronic trading of bonds, now in its infancy, could prove to have a much more direct impact on financial strategy than anything happening in the stock market.
Electronic bond trading systems have not gotten much traction in the markets, in part because of the complexity of bonds and the difficulty of getting timely information about prices. But the NASD TRACE (for Trade Reporting and Compliance Engine) recently began to report bond trading information with a 15-minute lag. Says Harris, "The publication of such information, which we've taken for granted for a century in the equity markets, will lead to substantial growth in electronic bond trading systems, substantial reduction in the cost of trading these bonds--especially for retail investors--and with that reduction, a very substantial increase in trading volume."
Harris predicts that this will lead to a fall in the cost of capital for corporate borrowers. It may also influence the kinds of bonds issued. A recent SEC study of transaction costs in the bond market found that such complex features as sinking funds, calls and puts made bonds costlier to trade. "Since investors avoid bonds that are expensive to trade, they will pay more for simple bonds, so financial executives can lower their cost of capital by issuing simpler bonds," Harris says.
Although they differ in the details, most market observers have little doubt that the listed market will bow to the trading efficiencies brought to stocks by electronic trading systems. Even the New York Stock Exchange's own hybrid market proposal adopts the essentials of fast and fully automated execution. It's not so much a question of whether automatic execution of trades will substantially replace specialists, but how. Says David Whitcomb, chairman of Automated Trading Desk, a fully automated Nasdaq market maker, "The NYSE has been dragged kicking and screaming into the 20th century--if not the 21st."
Warren, Nasdaq's CFO, sees the new market structure offering a new set of challenges to executives of public companies. "There are big differences between markets in terms of benefits a company's investors receive, and for companies as well. It's not just a decision of where you list, but also where you decide to remain listed."
But experience in Europe suggests that it is premature to predict the demise of human intervention in stock trades. No one has yet invented a fully automated trading system that can tap hidden liquidity--those big buy or sell orders that institutions would be crazy to post publicly for fear of moving the market. That means that, notwithstanding scandals and criticisms, there will probably be a role in the market for specialists for some time to come.
Gregory J. Millman (email@example.com) is a New Jersey-based freelance writer and frequent contributor to Financial Executive. His book about the Internet-triggered trading bubble, The Day Traders, was published in 1999.
|Printer friendly Cite/link Email Feedback|
|Author:||Millman, Gregory J.|
|Article Type:||Cover Story|
|Date:||Apr 1, 2005|
|Previous Article:||CFOs to tech: 'I'll spend for the right technology'.|
|Next Article:||Summit 2005.|