Outlook: Order-driven anomalies

The Stock Exchange is sticking to the script on the alleged shortcomings of its new order-driven trading system, but worries are plainly mounting that Sets is not the unqualified success we have been led to believe. Yesterday's move to allow execution only brokers to warn their clients to trade later in the day when they might get a better price is an embarrassing admission that the new system is potentially flawed.

In the run up to its launch four weeks ago, it was claimed that Sets would be the investor's friend, making it cheaper for everyone to buy and sell shares by slashing the difference between the bid and offer prices that used to represent the market maker's cut. It hasn't worked out that way. Far from becoming extinct, jobbers are for the time being making hay. They are still conducting two thirds of all business and making a better turn on it to boot.

According to the Exchange, the average spread has fallen below the old system's benchmark of just over 0.6 per cent. For most of the day it hovers around 0.5 per cent, but that disguises periods at both ends of the trading day when the spread is much higher. At the opening, it can be four times as much and does not tend to fall until about 10 o'clock when the market has enough depth to make it a sensible reflection of buying and selling intentions.

What that means in practice is that some pretty silly prices get quoted, and because the computer just does what it's told, some pretty silly deals get done. This is fine if you are close to the market and able to get a feel for what is the right price by watching other deals go through the system. It is not much good to Auntie Flo telling her no-frills broker to sell her building society shares "at best price".

Every day a glaring anomaly is thrown up. For instance, a string of trades in Williams yesterday at 351p was followed by one insignificant deal of 1,000 shares at 390p. Moreover, because it was the last order book transaction of the day in the stock, the higher price was booked as the closing price and is what appears in today's newspapers. This despite bearing no relation to the vast bulk of yesterday's dealings.

The relationship between the Stock Exchange and its computers being what it is, Gavin Casey is taking great care not to seem complacent, but he appears worryingly content to wait and see if higher volumes will solve the problem. They may do, but just in case the Exchange reverts to form, make sure you deal after lunch and set a limit.

Comments