The scenario sounds simple. You’re an institutional buyside trader who just got an order to sell 100,000 shares of an illiquid name at a specific price as soon as possible. Where do you begin?
Maybe 10 years ago you’d have shipped it to the New York Stock Exchange floor, where for decades 80 percent of all of its shares traded. But because the market’s now far more fragmented, just under 40 percent of listed stocks trade there. So, do you execute it yourself or call a broker?
If you handle it yourself, as more of the buyside is, you’re faced with a plethora of choices. You could look in the giant buyside-only dark pool, Liquidnet, and try to find a counterparty there. You could cut the order into pieces and use algorithms, smart order-routing and direct-market-access tools to send them around to the 40-odd dark pools to search for an anonymous match, or to the ECNs, or to Nasdaq or the NYSE. Through your algos you can check some of those dark venues simultaneously, but not others. And what if you call a broker? Do you think he can do a better job sourcing liquidity? Do you ask him to use capital? You have a wealth of liquidity options all offering the potentially perfect solution for executing your block order, but you’re finding that it’s still not easy to get it done. And the whole time, the clock is ticking.
Big Board Blues
Call it Fragmentation 2008. This is a re-creation of a previous theme. The first market that subdivided itself into multiple competing venues was the Nasdaq market in the 1990s. As a result of the Nasdaq price-fixing scandal, the number of ECNs, with the blessing of regulators, mushroomed.
Today, it’s the market for NYSE-listed securities that has fragmented. As a result of the Securities and Exchange Commission’s Regulation NMS, half the trading in NYSE names now occurs in venues other than the Big Board.
And because institutional investors are much more likely to own NYSE-listed names than Nasdaq names, this new bout of fragmentation is especially acute. By some estimates, three-quarters of institutional money managers’ orders are in NYSE names; the balance is mostly in Nasdaq stocks.
Making fragmentation different this time is the presence of so many crossing systems, or dark pools, which affect liquidity for both Nasdaq and NYSE-listed names. For many traders, this is the real problem: So much liquidity is hidden in systems operated by upward of 40 broker-dealers and exchanges.
Fragmentation affects traders in two ways: by (1) reducing the amount of information available and (2) making it harder to trade blocks.
In a recent survey conducted by Traders Magazine covering 126 buyside firms, the major trading problems associated with fragmentation are slower fills and bad prices.
It’s a Problem
In the survey, 80 percent of the respondents call fragmentation a problem. Not all agree on the magnitude of the problem, though-34 percent call fragmentation a “big” problem. Another 46 percent say it is a “small” problem. The remainder say it is not a problem.
When asked if fragmentation affects their portfolios’ performances, the numerical breakdown is similar. Three-quarters say fragmentation affects performance and, consequently, returns to investors-from pension funds down to mom-and-pops.
Still, some say fragmentation has benefits-that fragmentation helps keep trade costs low, spurs innovation, facilitates anonymity and brings flexibility (see sidebar).
Yet while not everyone agrees fragmentation is a problem, just about everyone on the sellside and at the market center level is working on a “solution.” Most of these moves-collectively called aggregation-involve connectivity, routing and algorithms.
“Fragmentation is a huge issue,” says Kevin Connellan, director of equity trading at Northern Trust Global Investments. “It’s a major concern, primarily because you have to use an algorithm; you can’t physically do this yourself anymore. There’s no successful methodology available. And there’s no master key that unlocks everything.”
Reg NMS Lights Fuse
Fragmentation in the NYSE-listed market can trace its roots to dramatic changes at the Big Board itself. The NYSE didn’t cede the market share monopoly it had in its names accidentally or voluntarily-it was more or less forced to when the SEC enacted regulation to generate more competition. The SEC’s Regulation NMS trade-through rule, which went into effect last year, forced the Big Board to go electronic by allowing traders to disregard NYSE quotes that weren’t automatically accessible. The move opened up the New York to more automated executions. In doing so, it created more competition, which allowed liquidity to move to other exchanges and to alternative trading systems such as ECNs-like BATS and Direct Edge-and a growing legion of dark pools.
During the years around Reg NMS’s creation and implementation, pools of non-displayed liquidity began to proliferate and be seen as viable options. They varied from independent dark pools, such as Liquidnet and Pipeline, to broker-consortia such as the block-trading BIDS, to internal dark pools that some describe as little more than opportunities for brokers to market and brand electronic access to their existing flows.
The changes at the NYSE have produced a dramatically different trading environment. Ultimately, two-thirds of survey respondents said this environment has led to either speed or price issues when trading blocks. The trends show a new world where the average order size in the displayed markets has shrunk to under 250 shares.
And because money managers have ever more assets under management, block sizes that buyside traders are seeing have been growing, says Jeff Wecker, chief executive of Townsend Analytics, makers of the execution management system RealTick and other buyside services. “The problem [of executing blocks] has existed for a very long time,” he says, “and one would argue it’s getting bigger, because the number of blocks that need to be executed is getting bigger.”
How Blocks Trade
Typically, traders execute a block electronically by breaking it up into smaller pieces to minimize its impact on the marketplace. A trader searches for counterparties among the displayed and dark venues, as well as natural indications from his trading relationships-brokers, if he’s a buyside trader, and his other customers, if he’s on the sellside. But in today’s fragmented marketplace, this is becoming harder to do at the ever-quickening pace traders demand.
“The complexity continuing to be introduced by market fragmentation and all of these different ways of executing order flow creates an uncertain amount of latency in the placement of all the order fragments necessary to create a block trade,” Wecker notes. “There’re a lot of considerations, and as time is ticking, it does affect the execution quality.”
Less Color From the Floor
Fragmentation isn’t just about execution, though. It’s also about that aggregate market picture the sellside helped supply. For decades, upstairs brokers provided the buyside with crucial information on stocks: where natural liquidity was, how particular stocks were trading and news. As the sellside sees less order flow cross its desks, its ability to provide that level of market color is diminished.
“Quite frankly, because of the fragmentation and the changes on the NYSE, there’s not a whole lot of information there anymore,” says John Russell, senior vice president and director of U.S. Trading at Franklin Templeton Investments, a San Mateo, Calif.-based firm with $600 billion under management as of March 1. “Everyone seems to be hiding in some of these dark pools.”
The lack of flow-driven information hurts brokers when the buyside calls on them to execute orders-particularly in difficult names. Before, it was a narrow world where brokers saw where the flow originated and how it got filled, says Dave Leone, special consultant to NYFIX, which owns the dark pool Millennium. But as the buyside increasingly uses broker algos, DMA tools and dark pools, the broker can get left out of the information cycle he used to inhabit.
Making Connections
The traditional upstairs group doesn’t have as much visibility into that flow as they used to,” Leone says. “Their ability to discern the market is minimized. So their ability to execute on a block basis is a little bit diminished.”
Buyside traders have noticed this when they ship orders to brokers. According to the Traders Magazine survey, 52 percent said a broker’s liquidity sourcing abilities are worse than they previously were.
A major reason is that Reg NMS’s trade-through rule has complicated the block-trading process for brokers. Now, traders must take out top-of-book quotes. In the past, this wasn’t the case for institutional traders and their 100,000-share block trades, says Alfred Eskandar, head of corporate strategy at Liquidnet.
“To get that trade done in a marketplace that’s protecting the best price, irrespective of size, forces a lot of inefficiencies,” he says. “This is because you need to be hunting to source all the different venues to aggregate enough orders to satisfy your large order.”
And because each venue in which one hunts has varying amounts of liquidity, a trader who could access all of them quickly stands to have the biggest advantage when it comes to trading blocks in a fragmented world. In fact, 49 percent of respondents to the buyside survey agreed. They said better connectivity between dark pools and exchanges and ECNs was the best answer to fragmentation. And it’s in this direction that solutions from many brokers, technology vendors and exchanges have been moving.
Take It Inside
Broker solutions to fragmentation have usually involved internalizing order flow and providing the buyside algos to reach other non-displayed liquidity. And brokers like fragmentation because it gives them a tactical role in the trading process, according to Diego Perfumo, an analyst at the Greenwich, Conn.-based investment advisory firm Equity Research Desk.
“Brokers want fragmentation,” Perfumo says. “And they will support as many venues as possible to fragment the market so that they have a strategic purpose, which is to perpetuate their role as an intermediary. If everything were centralized in one place, there’s no reason why the buyside would use an intermediary to reduce searching costs.”
Regardless, finding contra flow internally helps reduce transaction costs and market impact, says Joseph Mazzella, managing director of global trading for Knight Equity Markets, a Knight Capital Group subsidiary. Connectivity is still the theme with order flow internalization, he adds. Once in-house, it can be sent to as many venues as Knight’s system can reach. Currently, Knight is joined to 16 dark pools, Mazzella says. “That’s going to be the key,” he notes. “Whoever builds the better mousetrap will be extremely successful.”
Deutsche Bank agrees with the importance of internalization and connectivity in helping the buyside trade blocks. It employs a strategy like Knight’s, but adds a relationship component, says Robert Flatley, global head of Autobahn Equity, Deutsche Bank’s equities electronic trading platform.
“You’ve got to be really close to those venues, as well,” he says. “You’ve got to sit on the board and try to influence them, in terms of the types of orders, features and interconnection between those dark pools. We’ve acquired a minority stake in the ECN BATS Trading. It is our strategy to not only be a client of those venues, but to influence them and drive them.”
Deutsche Bank has also addressed the market color drought by opening up its internal analytics to clients. Whatever their trading strategy, clients can use Deutsche Bank’s information on one or many orders, including proposed ones.
Centralizing Flow
At Merrill Lynch, the thinking is that buyside clients need a seamless mechanism to access liquidity from numerous pools, including exchanges, ECNs and dark pools, says Clarke Roberts, a director in the portfolio and automated trading group at Merrill.”Merrill hasbuilt sophisticated smart order-routing and crossing functionality that leverages botha continually growing number of external venuesand Merrill’s own rich internal pool,” he says.
The brokerage has brought more efficiency to trading blocks bycentralizing itsglobal orderflow. As a result, Merrill now has a centralrepository forcash, program trading, derivatives,DMAand algorithmicorders to matchanonymously and electronically.
Aggregation Game
Technology vendors are also playing the aggregation game. Townsend Analytics uses its buyside and sellside partnerships to provide tools to join as many displayed and non-displayed venues as possible, Jeff Wecker says. It continues to build into its main product, RealTick EMS, tools to connect to those venues for both its buyside and sellside clients.
In the same vein, vendor Pragma Financial Systems and broker-dealer Weeden & Co. have built OnePipe, which distributes an order among 25 “passive” liquidity sources from a central position on the user’s OMS. OnePipe search dark pools, hidden liquidity and internalization engines, looking at historic fill rates within each venue to determine where among each, and without preference, the order should be sent. Subsequent routes are based on real-time fill analysis. OnePipe will soon be able to search non-displayed liquidity in public markets as well, says Doug Rivelli, a managing director in Weeden’s program trading group. Pragma and Weeden say that they routinely see crossing rates of 40 percent with OnePipe.
Exchanges and ECNs are doing their part, too. Nasdaq will launch its own solution sometime this month, Brian Hyndman, senior vice president for transaction services at Nasdaq OMX Group, told Traders Magazine. The exchange will pool electronic liquidity from a “handful” of non-displayed execution venues. After orders pass through Nasdaq’s book, they will look for electronic indications before routing to the protected quotes at other market centers.
Direct Edge ECN’s answer is to connect other dark pools to its routing algorithm and its matching engine, says chief executive Bill O’Brien. The firm, like its competitors, constantly searches for the best ways to integrate its own liquidity with that of others.
The Big Board Responds
For its part, the NYSE has several initiatives to bring back flow. To begin with, it will open up its reserve-book functionality to outside brokers later this year. The ability to post reserve orders, non-displayed priced orders resting on the NYSE book, has been limited to floor brokers. Insiders acknowledge the move should return some of the Big Board’s lost market share. Also, the NYSE’s dark pool joint venture with the parent company of BIDS Trading should be operational by summer. The planned exchange facility, which requires SEC approval, will try to match non-displayed blocks sitting in the crossing venue, posted liquidity on the NYSE’s display book and floor brokers’ reserve orders.
Finally, NYSE Euronext announced last month that it will give NYSE Arca clients access to the dark liquidity in 29 destinations, including broker-dealers’ pools and ATSs. Arca is offering participants non-displayed quotes and aggregated liquidity, which should improve the potential for price improvement. At the time of the announcement, the exchange hadn’t named the participating venues.
None of those interviewed expects any of these strategies to consolidate liquidity into one market center-or return the NYSE to its glory days of 80 percent market share. Nor does anyone think that should happen. “I don’t think any one venue’s going to solve it for everybody and for every order,” Liquidnet’s Eskandar says. “I don’t think it’s realistic.”
Fragmentation’s Future
Sometime before the end of next year, industry experts predict the number of execution venues will shrink. Some expect drastic consolidation. Northern Trust’s Kevin Connellan anticipates a sizable culling of the marketplace down to five or six venues, including the NYSE and several bulge bracket firms. Others say the number of venues will continue to grow before consolidation begins. This is because there isn’t yet a shortage of innovative ways to deliver liquidity, says Townsend’s Wecker.
Knight’s Mazzella argues that as long as dark pools continue to link up, there will be no need to consolidate. However, he adds, only those that have volume and order flow will do well.
To Weeden’s Rivelli, each venue today has a certain nuance to it-some let traders negotiate, others have algorithmic liquidity flowing through them, while some others offer various types of liquidity. And no single destination is going to provide all of those different nuances of order flow, he says. “I think there’s always going to be room for multiple providers.”
Fragmentation’s Upside
The New York Stock Exchange is no longer the dominant place to trade blocks. And that isn’t necessarily a bad thing, many industry insiders say. Having more than 45 venues to trade blocks provides flexibility and lowers trading costs, they say. And, they add, the anonymity that dark pools provide brings obvious advantages. For a small firm, fragmentation can allow for better control of one’s order, says Kevin Chapman, managing director of trading at Nicholas-Applegate Capital Management, a San Diego-based firm with $14 billion in assets. The firm has traded small- and mid-cap names for many years, Chapman says, so it’s always had liquidity issues. Long-term manager Nicholas-Applegate likes to hide in anonymous systems and wait for liquidity to appear.
Others point out that competition breeds innovation. If all orders traveled to one exchange, there’d be less incentive to create new order types and supportive technology, says Will Geyer, president of JonesTrading. Additionally, having “a distributed model” reduces the risk of the market shutting down from a technological failure or physical disturbance that occurs at any exchange or venue, he adds.
Since Regulation NMS went into effect last year, the securities market has improved as well, according to Daniel Gray, market structure counsel in the SEC’s Division of Trading and Markets. Speaking at January’s Security Traders Association of Chicago’s midwinter meeting-and not for the commission-Gray downplayed fragmentation when he said speed and market execution quality had improved since NMS. He added that, for institutional investors, average trade size may have shrunk in the public markets but trading overall has become more efficient.
(c) 2008 Traders Magazine and SourceMedia, Inc. All Rights Reserved.
http://www.tradersmagazine.com http://www.sourcemedia.com
