The deeper implications for companies and markets of off-exchange trading
Steve Wunsch, CEO of the upstart Arizona Stock Exchange (AZX), likens the spectrum of trading alternatives facing investors to a choice of restaurants. The fare at Nasdaq boasts depth, speed and liquidity at a rather high price; the crossing networks' black bargain boxes offer low transaction costs and sizeable deals, but entail a wait and a degree of uncertainty about what will turn up on the plate; and the traditional exchanges go on serving a reliable meal of meat and potatoes.
As the battle for market share is waged by these three groups, listed companies can rest assured there will be enough food to satisfy investors, whose appetites grow and grow. Last year, US exchanges saw over 152 bn shares traded for nearly $4 trillion. US institutions have an estimated $4.7 trillion at their disposal, which they can nowadays invest with hands-on, real-time order control, often without market intermediation.
It may all be a bit daunting for stock issuers, but staying abreast of market developments could be a boon to the proactive IRO who wants to know how and why stocks change hands - or why they don't. According to one futuristic market model, IROs might even find themselves brokering stock on the trading floor, dining at the same table as the largest investors.
'The potential is limitless. IPOs could sell their own new issues through home shopping networks,' says Frank Baxter, president of LA-based Jefferies & Co, a brokerage group which transacts some 15 mn shares a day. 'Companies already ship prospectuses on the Internet. There has even been a Ponzi scheme. All the trading pieces are out there, but we will need to go beyond existing mechanical structures. We should be thinking organically, not mechanically. All our exchanges and trading systems are based on a Newtonian trading vision. Future structures will be based more on fuzzy logic, chaos and organic paradigms.'
Baxter, Wunsch and a handful of other pioneers are tackling the future of equity trading with perseverance and creativity. The one-size-fits-all markets of the past have been replaced by a myriad of niche opportunities to suit different investors. Perhaps the greatest limitation is the scope of imagination: many regulators, exchanges and brokers are still stuck in an outmoded mindset, and have yet to grasp the nature of evolving capital markets.
Take AZX as an example of the new paradigm: an end-of-the-day cross with elements of price discovery described by one observer as 'years ahead of its time'. It rarely trades more than a million shares a day, and on a bad day there might only be four. But what if there were 4 mn?
The AZX profile is growing as investors grasp the concept and alliances are hammered out with bigger systems like ITG/Posit, paving the way for an integrated market. As a trading model, AZX, with its price discovery twist, has market makers worried and the SEC has placed a volume limit on the system, beyond which it may have to register as an exchange.
The bottom line of an electronic market is that stock quotes are available to investors over distance. Telecom developments have already brought improved speed and efficiency and, along with other technological advances, will continue to do so.
The NYSE and Amex have exploited technology, too, for electronic order routing and so on, but quotes must still pass through their physical trading floors and they are losing market share to screen-based competitors like Nasdaq, Instinet and London's Seaq International. Twelve years ago Nasdaq's dollar volume was less than 20 per cent of the NYSE's, but by 1994 it had risen to well over half. Seaq-I, meanwhile, is said to trade nearly 8 mn NYSE shares a day. And all markets are facing an exodus of volume to third market brokers, as well as proprietary trading systems (PTSs) like Instinet and crossing networks. Shares traded on PTSs are now estimated to represent around a sixth of NYSE value.
The drive to technology and institutional investor control has also produced non-regulated off-exchanges, which derive stock prices from the home exchange. Exchanges may deride these networks as parasitic, but selling quotes and information to them generates useful revenue: the NASD gets over half its revenues from sales to third party vendors like Quotron, Instinet and others.
For now, neither the NYSE nor Amex plans to go screen-based but the global trend is certainly away from trading floors. In 1988 just nine of the world's 22 leading exchanges were screen-based; today it's the opposite. But the NYSE insists that nothing can replace the feel of the floor, where brokers meet to gauge market sentiment; and both the Big Board and Amex stress that their spreads are narrower than Nasdaq's.
'This is an era of unbundling trading costs,' says Alden Adkins, VP in charge of strategic planning at the NASD. Specifically, the development of crossing systems means patient traders no longer pay the cost of impatient ones. The end result is that difficult trades are getting more expensive, while easier ones are getting even cheaper.'
While venturing out of the US to Canada and the UK - Nasdaq is facing a backlash from the US exchanges. Amex levelled complaints in 1993 of $3 bn 'excess' trading costs paid yearly by Nasdaq customers. Last year, following an academic study, over a dozen class action suits charged Nasdaq market makers with collusion, price fixing and anti-trust violations: the charges are denied; a consolidated case is under way.
When the Justice Department launched its antitrust probe into Nasdaq, it demanded truckloads of records from every Wall Street market maker. Instinet was asked to supply wide-ranging information about its dealings with Nasdaq - a sensitive issue, since on the one hand, Nasdaq market makers are among its biggest customers; on the other hand, Nasdaq and Instinet compete for trading volume in OTC-listed stocks.
Instinet is the most widely used PTS and provides a useful working model for what a future global capital market might look like. A member of over a dozen exchanges worldwide, it is a broker that looks like an exchange; and the fact that only customers have access to its prices perturbs traditional market makers.
Instinet's real-time trading has made inroads into US markets, especially Nasdaq's. Traders like its anonymity, particularly for stocks traded on Nasdaq, where the identity of the buyer or seller can sway the price. They also like the fact that Instinet allows them to trade between the bid and the ask. However, around 45 per cent of Nasdaq trading is in blocks of 10,000 shares or more, and that business is mainly done by phone rather than on Instinet.
Instinet was first to institute an after hours call market match in 1986. ITG/Posit, AZX, Chicago Stock Exchange's Match, Morgan Stanley's Match Plus and CS First Boston's Lattice have all followed. And even the NYSE offers an after-hours cross, though it is little used. Electronic off-exchange crossing now accounts for around 6 per cent of US trading volume. Traders submit anonymous orders and, at the given time, the system sorts through and matches orders using the price from the primary exchange.
Crossing systems are used mostly for NYSE stocks, attracting little activity in Nasdaq stocks. This is partly because the NYSE has a lead over Nasdaq in terms of institutional holdings but, according to Adkins, institutions are a growing constituency for Nasdaq, and institution-friendly Instinet will keep growing too.
Evan Schulman, president of Lattice, CS First Boston's 'execution engine' which provides sophisticated order management tools to traders, sees equity markets as illiquid and inefficient without electronic access for institutions. Schulman, describes Lattice as 'built by the buyside for the buyside.'
'Right now there are market imperfections,' Schulman says. 'Think of the poor specialist. He can bring $3 bn to the market to provide liquidity. The upstairs market can bring $20 bn. If you add pension funds like CalPERS and CREF, you soon get to trillions. The market maker cushion of liquidity is no longer there, and as a result we have wide spreads and volatile markets. To get liquid markets, funds need electronic trading and real-time control over orders.'
With systems like Lattice, traders are not restricted solely to buy, sell or limit orders; they can fashion strategies with orders contingent on events. 'Call market crossing networks are one response to illiquid markets,' says Schulman. 'Continuous market and electronic trading will win out as the mainstay of institutional liquidity.' Lattice currently trades around 1mn shares a day for 20 users. About 10 per cent of that is in continuous off-exchange matches. If an order doesn't match, then it will be filled on Nasdaq or on the floor in Boston or New York. As Schulman says, volume begets volume, and the hit rate of any matching system is low until critical mass is reached.
In Schulman's future vision, all exchanges, dealer markets and PTSs will be nodes in a unifying Lattice-like network. Institutions will be there providing liquidity, and market makers will be sitting inside the institutional bid-ask spread servicing retail clients. 'I view it as a dynamic flow of orders, with good traders providing retail customers with narrower spreads and better pricing,' he says.
Bruce Weber, assistant professor of financial technology at New York University and visiting at Wharton this year, wrote a study for the London Stock Exchange in 1994 in which he concluded that competition provided by alternative trading mechanisms could benefit issuers by fostering more attractive and liquid markets. 'However, there are questions of fairness, and whether crossing systems reduce trading costs,' Weber said. 'People forget about opportunity cost. PTSs are good at advertising how much they save for investors when orders execute, but they don't tell you about the added cost of filling the orders that don't.'
Weber believes exchanges are passing up important opportunities in avoiding crossing networks. But the new methods cut out exchange members and other traditional intermediaries. 'They don't want the exchanges to break their rice bowl,' he says.
Weber sees recent innovations in price flexibility on systems like Chicago Match and the UK's Tradepoint Financial Networks plc as positive, but says regulators must catch up. There are also problems with impaired price discovery and prices being restricted to certain users, as on Instinet. And PTSs and third market dealers are soaking up the low-risk profitable trades and leaving the exchanges with 'toxic waste'.
With the encroachment of US-based systems like Instinet and Global Posit on Seaq business, US investors may turn to a familiar screen to trade foreign stocks. 'Alternative systems are spreading out globally, opening local markets to their customers and cutting out intermediaries,' Weber says. 'However, worldwide market developments have borne out the continuing need for regional exchanges. The 1980s idea of 24-hour, global trading has been debunked.'
Off-exchange trading systems boast that they cut investor costs in more ways than just through broker commissions, citing market impact and dealer spread, for example. Santa Monica-based Plexus Group specialises in improving trading performance, mainly by analysing transaction costs. It says that off-exchanges are burdened with hidden costs, such as that of waiting for a match. This comes as some surprise, given Plexus's history: founded in 1986, an early niche was in marketing Instinet's Crossing Network to investors. After that, in 1990, Plexus began investment management consulting for money managers, sponsors and brokers.
'I care deeply about a central auction market,' says Plexus partner Mark Edwards. 'There is a feel, and the market maker standing there has a better pricing gauge based on the swelling of underlying demand and supply. When a trader needs to trade, he has to leave the cross and trade the old fashioned way by locating or provoking liquidity. Brokers are still the fastest mechanism for true price discovery.'
Edwards sees PTSs as a neat accessory to exchanges, but warns that by fragmenting the market they hamper price discovery and market depth. He says PTSs have drawn the cream off exchanges by taking away index, quant and liquidity traders, leaving market makers without their insurance policy against hungry information players.
In Plexus's 1994 ranking of non-full service brokers' trading costs, off-exchange PTSs Crossing Network and ITG/Posit took first and fourth place for exchange trading. In Nasdaq trading, the top four were Instinet, ITG/Posit, Crossing Network and AZX. Plexus determines the rankings by analysing over 800,000 institutional trades executed in the previous year and comparing each trade to a historical benchmark.
Within Plexus' core client base of over 40 money managers, ITG/Posit accounts for about 3 per cent of the volume, the Crossing Network 1 per cent. For OTC stocks, Instinet has 7 per cent, Posit 2 per cent and Crossing Network less than 1 per cent. Edwards calls the figures 'substantial', adding that some day all market segments will be integrated electronically. 'The best mechanism for price discovery is yet to be discovered,' he says. 'If it was, markets would already be using it. We don't know the right questions to ask yet, and traders still think as they did 50 years ago. Systems like AZX are years ahead.'
Off-exchange giant Jefferies & Co was ranked number one by Plexus in exchange trading for full-service brokerage firms in 1994. It topped the list in Nasdaq trading, ahead of Salomon Brothers; and came first in Nasdaq trading in small-caps, large trades and illiquid stocks. With 13 offices worldwide, Jefferies relies on what Frank Baxter calls the most 'subtle, sophisticated and elegant software' - human software.
Jefferies is one of an elite half dozen brokerages operating a worldwide off-exchange market, aggressively putting together huge blocks and taking commissions on both sides of the trade. Baxter is a firm believer in a mix of different trading mechanisms. 'They complement one another,' he says. 'Although they compete viciously, the results are good for buyers and issuers.'
Jefferies created Posit back in 1987 to stay competitive, electronically crossing lists of stocks with index traders. Investment Technologies Group, parent of Posit which went public in 1994 with Jefferies holding 80 per cent, trades around 9 mn of the group's total turnover of 15 mn shares per day. 'With electronics, the physical place is no longer important,' says Baxter.
Baxter estimates alternative markets to be taking some 9 per cent of volume away from stock exchanges but he says price discovery is not in danger. 'In every market, a few transactions set the price,' he says. 'As long as there is agreement that it represents fair value, you have a market place. Theoretically, when so much volume leaves the exchange that the price loses credibility, traders will migrate back to adjust prices to the rest of the market. An invisible hand maintains the equilibrium.'
As for the poor broker-dealer in all this, Baxter argues that as information is liberated, the agents in all businesses have a harder time justifying their contribution. 'Issuers should be thinking about being more proactive and undertaking the specialist function - the buyer and seller of last resort for their stock,' Baxter suggests. 'Companies should take their place as a central market player.' As for institutions, Baxter says that those which are permanent holders can add to the market making function. 'At the moment, specialists are not providing the same trading cushion as in the past,' he adds. 'That has made markets more volatile. For that reason, a new structure has to emerge.'