Evolution of Stock Exchange Trading Systems
Until the mid-nineties, almost all stock exchanges functioned as open out-cry systems. A broker received customer orders, directly from customers or routed via their back-offices, by phone. He would usually then make a note of the order in a little notebook, make some strange finger movements, and call out the name of the stock to indicate that he wanted to sell it. Alternatively, he could look around for somebody doing a different set of finger movements and crying out the name of the same stock, indicating he wanted to buy. Once the buying and selling brokers met, they decided whether they could make a deal at the price offered (for the sale) and desired (for the buy) and if the deal came through, it would be inputted at the end of the day into the exchange’s system, which was traditionally paper-based and now electronic.
Though there was price discovery through the movements in prices that various brokers offered and desired for the stock, price dissemination was imperfect. The result was that individual brokers could get their clients (mostly retail clients) very different prices, depending on their skill and, sometimes depending on their integrity, brokers would hoodwink gullible clients by giving them the worst price of the day, keeping the rest for themselves.
An electronic trade matching system does the same thing the brokers did earlier – get buyers and sellers together so that they can deal at a price acceptable to both of them and publish these prices in real-time. At the centre of this system is something called an order book which stores the buy and sell orders that have been received for the various listed stocks. There are usually different order books for different types of trades, in that sense, they represent different markets administered by the same authority. For example, the option trading market will have a different order book from the equity cash market.
Essentially, an order book has to match the buy and sell orders for each stock so that the buyer gets the best price offered from his point of view (the lowest price to sell) and the seller also gets the best price (the highest price to buy).
The basic rule of priority for matching trades is first price and then time. Matching always takes place at the passive price (the price already present on the exchange if a better offer is received, e.g. if a new offer to sell at £10 matches comes in with an existing offer to buy at £12, the deal will happen at £12).
The biggest advantage of electronic matching is precise price discovery and fair trading. All orders are immediately recorded on the system and brokers can see the prevailing price quotes in the order book so clients can demand that information is shared with them and accordingly modify their orders. Each order is uniquely numbered and time stamped. There is also immediate execution of market orders and limit orders, which have an existing matching price on the system.
Also, in an automated exchange a broker’s customer can actually track his order on the trading system so there is absolute transparency. India’s National Stock Exchange (NSE) and the Bombay Stock Exchange (BSE) do not allow orders to be bunched so the broker cannot cheat customers on the price. The system also maintains an automatic audit trail so surveillance becomes easier
The input of all prices into electronic systems has also enabled trade through rule, the national best bid offer (NBBO), the concept in the US whereby the investor/broker has the option to buy/sell at whichever exchange they get the best price. The reduced role of the broker and the automatic execution of trades have reduced both time and cost of execution. This has also led to the growth of program and algorithmic trading, which has further pushed down costs.
The London Stock Exchange, with 300 member firms and 15,000 securities, was one of the first prominent exchanges to use an electronic order book in 1986, with the introduction of the Securities Exchange Trading System for a relatively small group of actively traded large-cap stocks.
India joined the bandwagon with the advent of the National Stock Exchange (NSE). It arrived with a fully computerized order book in 1994 called NEAT (National Exchange for Automated Trading). This enabled it to spread across to various towns and cities in India by setting up terminals connected to the central system through VSAT. Currently, it enables trading in 1363 securities through 2856 VSAT terminals (servicing 9000 users) spread across 354 cities. It managed to take over a large portion of trades done on the old but well-established Bombay Stock Exchange (BSE) (the oldest stock exchange in Asia that started operation in 1875). The competition from this new technology compelled the BSE to finally change to a computerized matching system in March 1995 called BOLT (Bombay Online Trading).
BOLT is one of the few stock trading systems in the world that handles hybrid/mixed modes of trading: both order-driven (where orders are received directly from customers and matched) and quote driven (where broker-dealers give two-way quotes for bid and offer which are then matched with customer orders). It supports the normal segment, the auction segment, the odd-lot segment and continuous net settlement. There are more than 6,000 BSE trading terminals installed across the country.
The Tokyo Stock Exchange moved to electronic trading on 30 April 1999.
The Nasdaq (once an acronym for the National Association of Securities Dealers Automated Quotation system) transformed the capital markets landscape with a market maker set-up that published quotes on an electronic matching system. In the market maker set up, there are at least two market makers on each stock who issue bid and offer prices. The Nasdaq does not have a physical trading floor that brings together buyers and sellers. Instead, all trading on the Nasdaq exchange is done over a network of computers and telephones. In November 1998 the parent company of the Nasdaq purchased the Amex although the two continue to operate separately. In the UK, the counterpart of the NASDAQ is the SEAQ.
The NYSE realised that there were investors who would prefer an automated matching system and introduced the Direct+ automated trading platform of the NYSE, which limits the size of orders to 1099 shares, in December 2000 in phases and fully implemented it in April 2001. Due to this quantitative limitation, this has remained a bonsai market with about 10 per cent of the trades on NYSE. However, as we shall see shortly, all this is supposed to change in the near future.
The one titan holding the relentless progress of a fully automated system from wiping out all other trading systems is the New York Stock Exchange (NYSE). With a 212-year history and a specialized trading system that has served it well for 133 years, the NYSE was in no hurry to change.
Currently, in the NYSE, trades of less than 1099 shares can be routed to the Direct+ system at the discretion of the customer and auto-matched. As per data published in the NYSE website, trade execution on this system takes just 0.6 seconds. Other trades go to one of the 800 brokers on the floor of the NYSE who route the trades through a specialist for that stock. The NYSE is an auction market and the specialist, in most cases, simply matches the best bid with a matching offer. However, he does more than just that. He works for a specialist firm appointed by the company to act as the specialist for their stock. If the orders become unbalanced, the specialist tries to make the market and bring it back into order even if it means buying or selling against the market out of his own account to achieve some balance to the buy and sell orders and curb volatility. The specialist makes money by pocketing the spread – the difference between the bids and ask price on market orders. A specialist can also give vital inputs on abnormal movements in stocks, caused by short selling and persons acting in concert.
However, this system has its detractors among both the investor community, and of course, the competitors. They are sometimes accused of trading for their own accounts ahead of customers, leaving clients with inferior prices and sometimes unfilled orders. Exchanges have an obligation to supervise the activities of specialists. The worst fears came true when the five largest floor-trading specialist firms at the NYSE in March 2005agreed to pay $241.8m to settle charges that, from 1999 to 2003, the firms profited from by improperly intervening in customers’ trades.
To match the competition from automated exchanges like the International Securities Exchange and NASDAQ, and some say, to adapt to the competitive forces unleashed by the requirements of the new Regulation NMS (Regulation National Market System), the NYSE has planned to introduce its own hybrid system, for which it has filed for permission with the SEC. The Regulation NMS improves on the current trade through the rule (whereby an exchange cannot give an investor a deal on its own system if there is a better price elsewhere) by covering NASDAQ listed securities and excluding manual quotes (i.e. floor quotes). As per timelines announced by the SEC for the implementation of Reg NMS, the first 250 NMS stocks would be covered on 9 June 2006 and the rest of them would be covered after 9 more weeks.
The NYSE also faces stiff competition from the emergence of ECNs (Electronic Communication Networks) through which investors can get access to multiple exchange order books, with virtually no manual intervention, which pushes down the cost of a trade substantially and also helps the investor to get the best price in the market. That was to some extent offset by the charges levied by these networks for access to these quotes. Reg NMS also grants uniform market access at harmonised charges for quotes displayed through SROs (Self Regulatory Organizations) by ECNs thereby increasing competition for the registered exchanges.
In the NYSE’s hybrid system, its automated matching system (Direct+) and its specialist managed trade floor would compete with one another, which means that the quantitative restriction of 1099 shares as maximum order size on the Direct+ would have to go. But that isn’t the only major change. The NYSE is making its trading floor more system savvy with a new trading platform developed for the Big Board by the Big Blue (IBM), called ‘Trade works’. The system will have custom-designed hand-helds running a version of Windows for the floor brokers and specialists. The handhelds will directly connect to customers, or clerk’s workstations, where orders are received to cut down on execution time. The handhelds will also provide real-time market information and a ticker tape of orders so a broker can intervene when he wants to handle something manually, e.g. input a trade to balance the position or to control speculative volatility.
The NYSE took another step in this direction by announcing the take over of the fully automated Archipelago exchange in May 2005.
Which is better? The open outcry, the hybrid or the electronic matching system? The electronic matching system is definitely quicker, less expensive and perhaps fair to everybody. But this advantage itself can be viewed as a disadvantage: often large players want to exploit the dynamics of the market to make a killing. That apart, an investor may want to buy/sell a strategic stake at a special price. Recently, the NYMEX (the New York Mercantile Exchange) set up an open outcry pit in Dublin.
The NYSE has definitely proffered some very cogent arguments in favour of keeping the specialists alive in the market. Its detractors are not convinced and point out delays in trade execution and the possibility of the specialists manipulating the market (as happened not so long ago). They also say that all the information and surveillance that the specialist provides can be tracked just as effectively through electronic surveillance, which can then feed into price bands and trade breakers to curb excessive volatility.
The answer is to let the market decide. As the NYSE is at pains to reiterate, the hybrid system will give the customer a fair choice between the auto-matching Direct+ system and the manual system. It makes no sense to destroy this choice before giving it a fair chance. Let us see what the SEC has to say on this, and how flawlessly the NYSE pulls it off.
The author would like to acknowledge Satish Swaminathan and Shiv Kumar for their input.