Print this
The Future of Trading

Will electronic trading systems mean the end of market-making and liquidity as we know them?

By Nina Mehta

One day in the future, maybe five years from now, the idea of a financial exchange may seem like a throwback to the nineteenth century. How quaint, some will say, a place—a piece of real estate!—where buyers and sellers came together to transact business.

Exchanges as we now know them aren’t likely to survive too far into the next millennium. The ramp-up of communications technology, faster executions, and the wildfire proliferation of electronic trading, real-time quotes and information promise to deliver an electronic Nirvana based on leaner, more efficient markets. At the same time, electronic equity trading systems such as OptiMark, Posit and Datek Online’s Island ECN are loosening the grip of historical monopolies such as the New York Stock Exchange and Nasdaq. Other electronic trading systems are already taking over the foreign exchange markets, and a number of new ventures have their sights set on the bond markets and the over-the-counter derivatives markets.

The pressing debate is no longer about whether the markets will become more electronic, or whether they are being liberalized and globalized. The debate is about whether, for all the advantages of electronic trading, the road kill will take the form of impaired liquidity and market-making. So far, two things have become pretty clear. The first is that the structure of the cash and derivatives markets is being radically transformed, and the second is that these technology-fueled changes are eliminating the need for intermediaries—to date, a cornerstone of the markets.

The future of market-making

For centuries, exchanges served investors by centralizing trading in one location. Brokers executed orders with market-makers, who provided two-way quotes and continuous markets, ensuring transparency and price discovery. Professional traders on the floor and market-makers made money through spreads and through their ability to read order flow. Access to order flow was market-makers’ information advantage.

Technology leveled that playing field, eroding the advantage of those who have traditionally provided liquidity to the markets. “We now live in an era in which almost all news is digested at the same time by most of the people in the marketplace,” says Hunt Taylor, senior vice president at Tremont Advisers, an investment services firm that specializes in hedge funds. “This has made the futures markets, for example, where contracts are proxies for an industry and a tremendous amount of the trading world’s intelligentsia is concentrated, much more efficient.”

The efficiency democratized the markets. Until recently, for instance, only dealers could trade directly on Nasdaq. But the rise of electronic communication networks, which allow investors to transfer risk directly and avoid the transaction costs associated with brokers and intermediaries, narrowed spreads and dramatically reconfigured the game for all players. “If you’re a buy-side guy, life is great now,” says Maureen O’Hara, a professor of finance at Cornell and an expert on market structure and regulation. “These little spreads have made your life wonderful. The ability to use alternative trading systems has made you able to compete with dealers in ways you couldn’t before.”

“There are tremendous economies of scale in regulation. it may well make sense to have a private firm that provides regulatory services.”
—Maureen O’Hara

Just last month, Nasdaq announced that it had embedded OptiMark’s fuzzy-logic technology in its trading system, and that broker-dealers could begin executing trades via OptiMark. Founded two years ago by Bill Lupien, the creator of Instinet (now owned by Reuters), OptiMark is an electronic order-matching system that allows buyers and sellers with different priorities and levels of urgency to find each other without impacting the price of what they’re trying to trade. In the OptiMark system, traders enter their size and price preferences for a trade onto a screen; the trade details sit in the dark until the trade is matched by other orders also sitting in the dark, waiting for the right counterparty or counterparties to show up.

The Nasdaq/OptiMark arrangement could well be the harbinger of a new financial world order, where investors determine their optimal trading strategies and execute them without the aid of market-makers.

Living on nickels

So is there any kind of viable future for market-makers? “Someone once said market-making is the business of picking up nickels in front of a steamroller,” says Tremont’s Taylor. Market-makers earn a steady income based on tiny profits day after day, until they get hammered by the markets, after which they have to crawl back and pick up enough nickels to keep going. In another era, the principal providers of liquidity could act on information they had before the whole world found out about it. But that’s no longer the case. “If I’m a principal provider of liquidity,” says Taylor, “the minute I find out about something, all my customers found out about it too. My ability to avoid that steamroller is diminished because technology is publishing pictures of the steamroller on the Internet.” Indeed, one of the powerful impulses behind the Securities and Exchange Commission’s encouragement of equity trading systems is that these systems will enable other entities to compete with market-makers, thus reducing the costs associated with trading in listed markets.

The fact that market-making is increasingly difficult to justify financially has led many in the field to wonder why they should take on the risk of having to provide pricing during market downturns without the guarantee of a spread, however small. The markets, however, haven’t changed overnight. Profits from market-making have been decreasing for some time, although it is now much tougher to earn a living. “Specialists, for instance, have had to shift from getting a fairly high percentage of their earnings from commissions to making their livelihood from trading,” says Mitchel Abolafia, a sociologist at the State University of New York at Albany who has studied trading environments at a number of exchanges. “Consequently, there has been a washing out among specialist firms because some of them weren’t that good at trading and didn’t want to take on the level of risk it required.”

“Anybody with cash and a computer can supply liquidity on a computerized exchange.”
—Craig Pirrong
University of Washington

The increasing disintermediation of the markets has inspired some to look for technological solutions. Since computers can now recalculate gargantuan portfolios in a few seconds and execute block sales according to complicated hierarchical rules, the expectation is that market-making can be partly automated. At its core, though, market-making is about risk-taking—and risk-taking cannot be left to computers, no matter how intelligently programmed they are. Still, some believe that making markets will be drastically different in the future from what it is today. The difference: computational algorithms will be incorporated into the process.

Five years ago, Andrew Lo, a professor of finance at MIT’s Sloan School of Management, and Tomaso Poggio, a professor in the brain sciences department and the artificial intelligence lab at MIT, created a brouhaha by publishing a paper arguing that computational learning algorithms, which draw on neural networks, could have uncovered the Black-Scholes-Merton formula on their own, if given a large enough set of daily options data. Their point was that it is possible to use statistical methods of estimation to reveal certain nonlinear relationships among financial securities.

Market-making, of course, is a balancing act. Market-makers are required to provide price continuity, liquidity and orderly markets—but they’re also supposed to earn a profit. “There’s no doubt that having humans involved in a market-making capacity enhances market liquidity,” acknowledges Lo. “The question remains, however, as to how much of this activity can be automated and how much cannot be—in other words, what’s the art vs. the science, or rather, what’s the art vs. the engineering?” A few years ago, he and Poggio began developing a way to train algorithms to mimic the behavior of specialists on the NYSE. After slogging through many generations of algorithms, Lo and Poggio figured out how to balance some of the competing goals of market-making in an optimal fashion. MIT recently filed for a patent on their behalf.

Trouble ahead

Not everyone, however, thinks market-making will survive the onslaught of electronic markets. Steve Wunsch, president of the Arizona Stock Exchange, a fixed-time, single-price call market for NYSE and Nasdaq stocks that last month received approval from the SEC to operate during regular trading hours, is withering in his condemnation of the changes wrought by the democratization of the markets. The increase in on-line retail trading has altered the way institutions trade, he says, with disastrous consequences for the markets.

Risk Management to the Rescue?
It could be said that modern risk management was born in the 1970s with the introduction of financial futures—and beefed up in the 1980s as the increase in computational power enabled huge portfolios to be analyzed quickly. But another, less-grand view of risk management suggests that the tightening of spreads has been one of the key catalysts behind the growth of risk management as a trading discipline.

According to this theory, risk management has taken off because it’s simply more difficult to make money than it used to be. The more efficient the markets, the more spreads migrate south. “You may make a lot less money per trade,” says Junius Peake, a professor of finance at the University of Northern Colorado and a former governor and vice-chairman of the National Association of Securities Dealers, “but if you do 50 times more trades, you can more than make up for it in volume. But if you do 50 times more trades, you’ve got to have a risk management system to make sure you control that risk.” Put another way, risk management is not the brainy child of advances in computational power, but the inevitable stepchild of diversification and narrower spreads.

Risk management, of course, is already a mantra in the trading world. And as floor skills—such as knowing who is trading and who is holding what—disappear in electronic environments, having the discipline to evaluate trades objectively and get out of losing situations efficiently will be the way traders make money.

Nonetheless, everyone doesn’t see risk management as a magic bullet. Attention to risk is good when it’s focused on reining in the likelihood of fraud or rogue trading, notes Paul Wilmott, a principal of Wilmott Associates and author of Derivatives: The Theory and Practice of Financial Engineering. “The basic value-at-risk methodology, however, is about what happens on a typical day. But about what happens on a typical day, who cares? It’s what happens on those atypical days that banks go down. VAR gives people a false sense of security.”

Another note of caution is sounded by Seleyman Basak and Alex Shapiro, both professors at the University of Pennsylvania’s Wharton School, who recently argued in a paper that risk managers who use VAR to manage their market-risk exposure “often optimally choose a larger exposure to risky assets than non-risk managers, and consequently incur larger losses when losses occur.”

Still, the popularity of VAR is now helping fuel structural changes in the trading world. Indeed, risk management may hasten the demise of floor-based exchanges, says the University of Northern Colorado’s Peake. He expects derivatives exchanges—especially in the United States, where floors are still strong—to metamorphose into for-profit, independent trading systems hooked up to the guarantor of the derivatives instrument (something equivalent to, say, an Options Clearing Corp. or a Board of Trade Clearing Corp.). Meanwhile, the physical exchanges will die off. “If you’re at JP Morgan,” he adds, “it’s one thing to be able to control your inventory and trading activities through a computer system you control on the front end, and another thing to have a bunch of people standing in a room yelling at each other. Even though they work for you, it’s much harder to make sure you have second-by-second risk management.”


“It used to be the case,” he argues, “that an institution would go to a dealer or market-maker and say, ‘I’ve got to sell 1 million shares, can you help me?’ But now the institution has to fit its trades into an essentially retail-dominated flow, which means it breaks them down into tiny pieces, and the market-making community doesn’t find out about it.” This creates less liquidity in the markets at any given point in time, but even that’s not the crucial problem. The real problem with cutting market-makers out of the loop: prices can’t be discovered accurately. The change in trading style has thus made it impossible for dealers to make markets in size to institutions and has made it impossible for institutions to give them trades to make markets for—because, Wunsch says, “the institutions are too afraid they’ll miss the volume-weighted average price or something like that.”

What’s at the root of this change in the markets, in his view, is antitrust law—which has yanked the trading world in a “mistaken socialist direction.” The old network of market-making—which the dealing communities created with upstairs information, based mainly on institutional order flow—“was able to find its way to a good price in the market that was relatively stable,” he says. “But the antitrust settlement with Nasdaq and the new order-handling rules [in 1997] essentially banned all of the upstairs dealings that the network entailed.” As a result, the markets are far more volatile, prices fly around without much rhyme or reason, and nobody can read the flow anymore.

“If i’m a market-maker, the minute i find out about something, all my customers found out about it too.”
—Hunt Taylor
Tremont Advisers

Adding fat to the fire is what Wunsch refers to as the “entirely flawed theory” that terminal-based trading displays all relevant information to everybody. The problem, he says, is that screens display no information about the direction and timing of order flow; moreover, there are no AI programs that can determine when the flows will move up or down on any stock or market. “That’s why market-making as a profession is going to die,” he says, “and the floors are just one form of it that’s going to die. The on-line trading phenomenon will take over—certainly in all stock and derivatives markets, and probably in bonds too, although they’re a few years behind.”

The future of liquidity

Market-making as we now know it may be headed for the dustheap, but the survivors will have to cope with an even more pressing concern: What will happen to liquidity? Will it fracture as electronic trading systems draw investors away from the traditional markets? Or will new types of liquidity develop as electronic markets take shape?

For as long as exchanges have existed, liquidity has been their raison d’etre and their measuring stick. If a market isn’t deep enough, spreads will be wide and traders likely to pack up and move their business elsewhere. New markets, however, do not automatically drain liquidity from existing markets or exchanges. They can enlarge the pie by generating more volume. “As long as the big traders stay with the Instinets and the NYSE, market liquidity is not going to fracture that much,” says Cornell’s O’Hara. “It’s not obvious to me that liquidity has really suffered in the last six months or one year. In fact, [the growth of ECNs] seems to have induced greater participation by more traders.”

Some in the markets make a distinction between short-term and long-term liquidity, and between retail and institutional liquidity. An argument gaining legs is that short-term liquidity may suffer as ECNs gather strength and divide the market, but that this will pass as consolidation takes place or as systems are developed to link information about liquidity and prices on various ECNs and trading platforms.

In addition to the worry that an electronic world of competitive, decentralized markets will fracture liquidity, there is continuing concern about institutions being the key drivers of liquidity. In the equity options market, for instance, says Sheldon Natenberg, a floor trader at the Chicago Board of Trade and author of Option Volatility & Pricing, retail customers tended to disappear after 1987. At the same time, institutions began trading more against large portfolios. The combination made trading more difficult for options traders. “Think of what’s happened in the mutual fund industry in the last 10 to 12 years,” he says. “If a mutual fund wants to do a buy-write, where it owns stock and wants to sell options against it, the fund wants to do a lot of them and pretty much all at one time. In other words, there’s more and more trading being concentrated in fewer and fewer hands.”

Similarly, while the anonymity and flexibility needs of institutional investors are being met by new equity trading systems, retail customers are likely to lose out, since alternative trading systems are closed to them unless they trade through a brokerage. The new systems make trading easier for all traders, but only “as long as the exchange or system is open to anyone who wants to participate,” says Natenberg.

Market veterans confident about the staying power of small investors argue that changes in market structure and the continuing reduction in transaction costs will redress liquidity imbalances by enticing new providers of liquidity into the markets. What this holds in store for the capital markets, however, is a matter of dispute. There is now a simultaneous movement toward electronic trading and the for-profit, public ownership of exchanges—“because electronic trading allows the entry of liquidity suppliers from anywhere,” says Craig Pirrong, assistant professor of finance at the University of Washington. “We don’t have to rely on a specialized group of locals on the floor of any exchange.” For exchanges, this could lead to more interexchange competition. Knock-down drag-out fights on the same product have been extremely rare among open outcry exchanges, “but in an electronic era it will be easier to add a new contract and compete with an existing one, since exchanges can draw on liquidity suppliers from around the world,” notes Pirrong. “Anybody with cash and a computer can supply liquidity on a computerized exchange.”

New providers

Another way to view the changes taking place is to consider the utter reversal in the markets that some are predicting: that buy-side customers who used to be the beneficiaries of greater liquidity may be the ones providing that liquidity in the future. Those in the hedge fund industry note that hedge funds are turning into the liquidity-providing proprietary-trading desks of yesterday, as proprietary trading firms and the dealing rooms of banks reassess their risks and shift to markets perceived as less efficient, or that have less-erratic income streams attached to them. Others say that tens of millions of ordinary people around the world trading on their own accounts will provide liquidity in the future.

Throughout history, notes Tremont’s Taylor, “providing liquidity and speculative capital has been the private domain of financial institutions.” But now hedge funds and other entities can be on both sides of a transaction. “They can take a directional view on interest rates or equities,” he says, “and they can have a stake in the income stream that comes to those who facilitate that trade by providing liquidity.”

The movement toward direct access to the markets is the most startling change in the markets for another reason as well, points out Michael Adam, chairman of Inventure America, a software firm, and founder of Adam, Harding & Lueck, a hedge fund. The market establishment knows that order flow produces profit, he notes. This means that dealers and brokers make their money off the hidden spread they charge customers who have inferior access to the markets. However, the great rope trick, in his view, is that the dealing rooms and those with direct access to order flow have obscured the relationship between access to order flow and the ability to make money. The assumption, he says, is that traders make money because of superior trading skills, when in fact they generally make money because of their superior access to the markets. This assumption is then reinforced and perpetuated by the way traders are paid, to the detriment of clients.

So even though spreads have narrowed, the newfound ability of hedge funds and other investors to access order flow directly will change the business of who provides liquidity. There is currently not enough liquidity on ECNs, Adam concedes, but once ECNs are connected to one another electronically, hedge funds will be able to compete with banks to provide order flow—and will win out, he says, because they have lower overhead costs than banks have.

“Having humans involved in a market-making capacity enhances liquidity. but how much of this activity can be automated and how much cannot be?”
—Andrew Lo

Not everyone, however, thinks that a battalion of decentralized liquidity providers around the world is the answer to concerns about liquidity. Hedge funds trade for themselves and have no enduring—that is, obligatory—commitment to the markets in which they trade. Retail investors and day traders, meanwhile, are not seen as reliable providers of liquidity. They may make a contribution to the markets, but they don’t add sufficient long-term liquidity. “If you bring 100 people who were former dentists, construction workers or waiters into the markets, 99 percent of them will wind up trading for a year, losing as much as they can afford to lose, and then going away,” says an independent trader at Nymex. “You’ll still end up with a net loss of liquidity. There’s a real difference between the professional trader who’s on the floor and committed vs. somebody who’s trying a change of career because he couldn’t make as much money waiting tables and has dreams about striking it rich as a commodity trader.”

Regardless of how quickly the winners and losers fall into place in the new electronic environment, no one doubts that the current changes in market structure are creating new uncertainties along with greater efficiencies. Moreover, none of these new trading systems has been tested in a market crisis. The hope is that if these new systems are not much better, they will at least be no worse than what we have now.

Regulation For a New Marketplace
Under the best of circumstances, technology helps people envision the future. Regulation, however, is often a retrospective industry, based on the assumption that the future will be similar to the past and that good regulatory solutions from the past will be related to those of the future.

As electronic trading changes the structure of the financial markets, the worry is that the regulators won’t keep pace—and that inefficiencies will flourish and competitive markets won’t. The development of new trading systems is “up there with the Big Bang and the French Revolution,” says Joe Grundfest, a professor of law and business at Stanford Law School and a Reagan-era SEC commissioner. “The implications for market structure are extremely large—and, if anything, the policy-making apparatus is only beginning to appreciate the magnitude of the changes that will be forced on regulatory structures.”

Supervising the Screens

Chief among the concerns is who will regulate the new markets and what the SEC refers to as “alternative” trading systems. The commission’s view of the markets has always relied on intermediaries—the broker-dealers who have the responsibility to protect investors. But as intermediaries disappear and people meet one-on-one, the regulatory tools must be changed to identify fraud, the misuse of information and so forth.

Another issue is that all computerized trading systems are currently registered as broker-dealers and regulated by the National Association of Securities Dealers. The problem: many operate as exchanges, not as traditional broker-dealers, and are therefore not subjected to the kind of oversight accorded traditional marketplaces. The plot has gotten even thicker, however, since some trading systems will now be able to register as exchanges, although how they will be regulated is still uncertain.

Then there’s the conflict-of-interest issue. For decades, exchanges have functioned as self-regulatory organizations. But now tension is growing as exchanges begin to compete with electronic trading systems—some of which they will be responsible for regulating. “With the proliferation of these systems, the whole concept of self-regulation becomes problematic,” says Maureen O’Hara, a professor of finance at Cornell and an expert in securities-industry regulation. “It made sense when you had the New York Stock Exchange and the NASD—they could self-regulate and it wasn’t a big deal.” But with the field widening and the dominant marketplaces competing for order flow, the situation has gotten dicey. “It may get a little better if Nasdaq is spun off as a private corporation,” she adds, “but then that just highlights the difficulty of self-regulation in a world in which the competitors are all for-profit institutions.”

An issue related to this is who picks up the tab for SRO oversight. The NYSE and the NASD currently spend hundreds of millions of dollars each year on regulatory compliance and surveillance, notes Stanford’s Grundfest, and the alternative trading systems aren’t bearing any of those costs. “Some costs—such as the compliance of floor traders—don’t apply to them, but insider trading can occur as easily on Archipelago as on the NYSE or Nasdaq,” he says. “As long as the overall market was stable under the old structure, that was a viable regime. But if, in the future, 50 percent of traffic goes through ECNs or other electronic systems, then there’s an argument that at least some of the regulatory costs should be borne by the ECN-type structures.”

The SEC clearly recognizes these problems. Late last month, chairman Arthur Levitt floated the idea of creating a single, independent regulator to oversee the securities markets. The exchanges would police their own trading, but the super-regulator would be responsible for overseeing intermarket trading and the activities of exchanges and other trading systems.

The announcement was greeted by support from many, including the NASD. The NYSE, however, criticized the proposal, complaining that it would “weaken investor protections and do irreparable harm to the NYSE brand.” Richard Grasso, the Big Board’s chairman, also panned the idea of separating the exchange’s regulatory function from its trading operation, which Levitt said would likely be necessary for SEC approval of exchanges’ plans to shift to for-profit entitites. Grasso argued that this sort of “severence...would be both culturally and practically impossible.” This back-and-forth suggests a new depth in the budding conflict between the public interest of SROs and the trading interests of exchanges.

Disorganized Markets

Other serious regulatory concerns relate to the national market system, which, for more than two decades, has been the organizational backbone of the U.S. securities market. In 1975, Congress passed the amendments to the Securities Exchange Act of 1934 that authorized the development of the NMS and strengthened the SEC’s ability to promote it. The goal was for markets to share information so investors could get the best execution on trades. The main components of the NMS are the Intermarket Trading System, a kind of bulletin board communications system to share trade information, the Composite Quotations System and the Consolidated Tape Association. (The latter two disseminate quotes from market-makers and sell consolidated last-sale information.)

“The intermarket trading system should have been dead and buried with a stake through its heart years ago.”
—Junius Peake
University of Northern Colorado

One of the main gripes is that the technology behind the NMS’s main systems is controlled by the Securities Industry Automation Corp., which is two-thirds owned by the NYSE and one-third owned by Amex-Nasdaq. Nasdaq and the NYSE—the two dominant marketplaces—make pots of money off the fees they get from selling their trade information through the CTA to mutual funds, investment houses and other firms. Nasdaq’s tape revenues, for instance, are reputed to be in the neighborhood of $100 million a year. This may have helped inspire Datek Online’s Island ECN to apply for exchange status, says one market participant. If it registers as an exchange, it will presumably share in the revenues that come from selling its trade data.

Right now, these alternative trading systems are not yet incorporated into the NMS. So even though they fall under the NASD’s jurisdiction, they are not surveiled and investigated effectively. Moreover, while quotations on some computerized equity systems are incorporated into the NMS information infrastructure, most are not. This leads to issues of price discovery and fairness, particularly since retail investors cannot—at least for the time being—execute trades directly on these trading systems.

An example of what could go wrong is the market-making debacle of the mid-1990s, when Nasdaq market-makers quoted tighter spreads on Instinet than they did to public investors on Nasdaq. “The SEC is going to require that these new trading systems be somehow linked,” speculates a market professional, “since it has never been particularly enthusiastic about people transacting at different prices.” One possibility often cited is an open limit-order book system that enables investors to see all the bids and offers posted on exchanges and alternative systems.

Addressing this issue in recent years, the commission has noted that as the markets develop and more trading moves to electronic trading systems, registered broker-dealers could conceivably form their own SRO, or, alternatively, that a tiered definition of exchanges might make sense. But a more innovative option, suggests Cornell’s O’Hara, may be the development of regulatory providers that offer services along the lines of the audit functions provided by large accounting firms. “There are tremendous economies of scale in regulation,” she says. “It may well make sense to have a private firm that provides regulatory services.”

SOS for the SEC?

Not surprisingly, some in the industry are less than spectacularly confident about the SEC’s ability to organize the markets of the future. “The Intermarket Trading System should have been dead and buried with a stake through its heart years ago,” says Junius Peake, a professor of finance at the University of Northern Colorado and a former governor and vice-chairman of the NASD. A long-time, outspoken advocate of the need for a national market system, he argues that the tripartite NMS was a disaster from day one. The problem was that the systems were not linked, resulting in a market structure even Rube Goldberg couldn’t have envisioned—and that this has stifled technological improvements in the markets. By not divorcing the needs of the market from the desires of its largest exchanges, he says, the SEC has been “fighting progress.” The commission has been in the business of protecting marketplaces for more than two decades, he adds, when all along it should have focused its attention on ensuring competitive prices so traders and investors could get the best execution.

Ultimately, Peake notes, the SEC could find the most unavoidable pressure to reconfigure the markets coming from abroad—as is happening in the futures market. “If somebody in Europe can build a better mousetrap and start trading U.S. stocks more favorably than we can,” he says, “we could wake up some morning and find that most of our markets are being traded overseas.”

The SEC is now insisting that self-regulation cannot afford to be jeopardized if privatization occurs. Peake praises the commission’s stand in this area, noting that if the NYSE or the NASD privatize, “it’s crucial that all the regulatory functions be removed from any market-center operation.” Grundfest, who adds that so far the SEC has done a decent job of responding to shifts in technology and allowing the markets to evolve, agrees that the commission’s regulatory responsibility needs to be carefully managed. “My hope,” he says, “is that the [SEC’s] political process does not slow down the great benefits that innovation can generate.”


Was this information valuable?
Subscribe to Derivatives Strategy by clicking here!