American Century: Getting jazzy in KC
It’s the most visible market and one of the most talked about. But the US equities market also is one of the most inefficient in the developed world. What makes it so fascinating is that this inefficiency is often combined with innovation and the combination has left the market susceptible to transformation by the internet.
The explosive growth in the use of the internet in the US for general retail services as well as for on-line retail equity trading would suggest that the US is yet again leading the development in the use of technology. A much larger proportion of the US public has access to a computer and cheap internet services than in other countries. Consumers are more used to TV shopping, making the leap to internet shopping relatively painless; and retail finance on the internet has been pushed along by the vibrant equity culture of the US, guided to a great degree by the need for individuals to take responsibility for their own pensions.
Yet this has not found its way through to the institutional market. While Europe might lag in the retail sector, it is ahead in the institutional market. Its various exchanges are still hampered to a greater or lesser extent by their antiquated membership structures but they at least have long since introduced various forms of electronic trading. Yet the major US equities exchange, the New York Stock Exchange, still operates on open outcry and each side to a trade usually has to route an order through five or six different levels of intermediation – people – to get that trade done.
Nasdaq is an electronic exchange but its inefficiencies have been highlighted in the last two years by the rapid growth of electronic commissions networks (ECNs) which now account for at least 30% of its business. ECNs came into being two years ago after the SEC allowed commissions to be charged for providing liquidity, something the brokers cannot do. The majority of ECNs simply match trades, with no price discovery.
The main impact of ECNs has been on Nasdaq, because the “floor” of the exchange – a series of 100 or more electronic market-makers – is not as efficient as it could be at providing liquidity. Brokers, for example, use ECNs as inter-dealer brokers on Nasdaq, whereas on the New York Stock Exchange they can use the floor, which historically has been better at sourcing liquidity (although that is rapidly changing). ECNs have had only minimal impact on the NYSE, taking at best 10% of the market
There are now 10 ECNs and several have financial backing from major investment banking and management companies. The SEC has ruled that they can file to become exchanges. Archipelago has filed, others will follow. Their resounding success on Nasdaq has led the exchange to conclude a venture with Optimark (see below), in an attempt to stem the flow away from the floor. It also has been a reason behind both Nasdaq and the NYSE deciding to demutualize so that they can respond more effectively to challenges and even use their capital to buy an ECN.
The growth of ECNs has made the market even more inefficient: with all the national and regional exchanges and the ECNs there are now more than 20 providers of liquidity. Not that they all cater to the same market. Of the ECNs, Island is popular with day traders; Strike was initially set up as a clearing network by Bear Stearns and now has most of the major investment banks as equity partners; Brut is more of a dealer network ECN whereas Archipelago has a broad range of clients. And Instinet doesn’t even want to be an ECN. It’s been in business for 30 years, has been owned by Reuters since 1988 and would rather be seen as a global agency broker.
The ECNs may have made the market more inefficient by further fragmenting liquidity but they have made the inefficiencies of the incumbents appear all the more stark. The dynamics of the market are such that the old system cannot last much longer without fundamental change. There is the growth of the retail market as direct investors, for a start. If, as the Securities and Investment Authority predicts, retail will account for 50% of volume trading by 2002, the institution-dominated system will have to adapt and after-hours trading is just part of that.
Decimal revolution
Other reasons all prove the need for greater automation, speed and accuracy: first is the move towards T+1 settlement, which should be here by 2001, with T+0 not far behind. Second, the move towards decimalization in the equities markets to replace quoting in fractions will create a broader range of price quotes and almost certainly lead to an immediate increase in trades done. Jim Lehman, managing director at Salomon Smith Barney, suggests an increase of 300% as a conservative estimate. A comparison with the increase in orders following the SEC rule changes in 1997 is instructive. In 1996 Merrill Lynch, one of the largest players, was processing 50,000 quotes a day; a year later it was 500,000. This year as many as one million trades a day have been done by Merrill.
Of those firms which have bought stakes in ECNs, some are doing it purely as a way of putting pressure on the incumbents. One of the major investors is Goldman Sachs. Its head of electronic trading, managing director Duncan Niederauer, says the investments have been made “because we would like to see the market structure in the US substantially altered”. The ECNs are a hedge for him. “The ECNs might just be a footnote in history. I’m confident the NYSE and Nasdaq can both adapt, but we need to have other options if the organizational challenges prove insurmountable for them. ECNs could become less relevant if the primary exchanges innovate, but ECNs could also substantially alter the course of market structure if the exchanges choose not to.”
Bill Harts, managing director and head of portfolio trading strategies at Salomon Smith Barney, says ECNs can be used as a way of diversifying where the banks put their order flows. “One of the reasons we invested in Strike is that we wanted to have an outlet other than Instinet. We were and still are a big customer and we usually get good service. But we don’t get the control.” By implication, the banks can use the ECNs as a bargaining chip with the more established competitors, which until a couple of years ago meant mainly Instinet.
Another reason for Salomon’s investment has to do with a loophole in the SEC regulation allowing the creation of ECNs: if the bid price for a customer order going through an ECN is met by an offer from an account not a client of the ECN, then the ECN can charge a commission on both sides of the trade. Bloomberg started charging almost immediately. Strike offers an alternative.
A trader at Morgan Stanley Dean Witter says:. “We’re not buying into ECNs to be in the ECN game per se as we’d rather see the exchanges adapt. We’re in Brut because we like the technology and want to learn more about ECNs. And our investment in Eclipse is because we like the idea of 24/7 [markets open 24 hours a day, seven days a week] and Eclipse is pushing to extend trading hours.”
Optimark and Archipelago are two order-routing systems which are attempting to provide better options for price discovery and order matching – Archipelago is even looking at ways of incorporating research into its product.
We examine Optimark and another new company, internet investment bank Wit Capital in the following pages. But first the views of one of the more technologically friendly – and outspoken – asset managers in the US, American Century’s Harold Bradley.
