Liquidity and best execution in the UK: A comparison of SETS and Tradepoint

This paper examines the relative attractiveness of the execution currently available on the two UK stock exchanges, by considering the extent to which better prices are, or appear to be, available on one or the other. Although the explicit focus is on the extent to which brokers' obligation to provide best execution for investors is being met in a world of multiple trading venues, the results have a direct implication for professional investors' choice of venue. Best execution, the requirement that agents obtain fair prices for their clients, is not simple to define when there are many trading methods and venues. Regulators in the UK and US have adopted different approaches to the definition of best execution: in the US, the approach is to leave the definition of best execution ambiguous and to require the participants to demonstrate compliance, while in the UK the regulator specifies exactly the conditions under which this obligation is discharged. We examine the impact of the UK's definition on the trading on SETS and Tradepoint, the two competing order books in the UK. As part of this, the paper also examines the relative prices available on the two systems, both in terms of the best offered prices and trade prices. During the period analysed, SETS was vastly more active than Tradepoint. SETS had 1.4 million trades, while Tradepoint had only 1,500. There were similar, but less marked, differences in the numbers of orders (2.9 million on LSE and 755,000 on Tradepoint). The level of activity on Tradepoint reduced dramatically when one participant withdrew from the market, which suggests that Tradepoint has a limited depth of investor participation as well as trading activity. Tradepoint volumes are equivalent to approximately 0.2 per cent of LSE trading. SETS had a very high availability (generally, over 95 per cent of the period) of bid and ask prices for almost all of the 135 stocks analysed. Tradepoint had high availability for fewer than 30 stocks for part of the period and very restricted availability of a two-way market at other times and for other stocks. The best bid-ask spreads on SETS were generally between 0.4 and 0.8 per cent, while those on Tradepoint were generally twice as wide (between 1.0 and 1.25 per cent). Tradepoint offered more attractive prices, on either the bid or ask, on average for between 45 and 90 minutes per day, at volumes that were broadly comparable with those available on SETS. In terms of best execution, approximately 30 per cent of trades on SETS were at prices that could apparently have been bettered on Tradepoint. There was also a very small number of trades (30) on Tradepoint that could have been bettered by trading through SETS. In both cases, a lack of depth might explain why the trade was not done through the market quoting the better price or the explanation may reflect settlement preferences (the two exchanges use different settlement procedures). In addition, the bulk of these trades were of institutional size and, therefore, likely to have been quite closely monitored by the participants. Overall, this research suggests that there is no strong evidence of mis-allocation of trading between SETS and Tradepoint. This situation might change, however, as more competitive alternatives to SETS open in the UK.