The Role of Pre‐Existing Liquidity in Determining Pricing Efficiency and Liquidity Gains Following the Introduction of SETSmm

• Published date: 01 March 2015
• Author: Patricia L. Chelley‐Steeley

On 3 November 2003 the London Stock Exchange (LSE) introduced a new trading system called SETSmm for mid‐cap securities. SETSmm is a hybrid trading system that combines the characteristics of its auction trading system SETS (Stock Exchange Electronic Trading System) with market maker participation to ensure the continuous provision of liquidity. This was an important event for the LSE as there is growing evidence to show that the trading mechanism can have a profound effect on the behaviour of security returns. In this paper we examine the impact that the introduction of SETSmm had on firm value, liquidity and efficiency. Our paper adds to the growing literature that examines the impact of the trading mechanism but is distinctive as it is the first paper to examine the effects associated with the introduction of SETSmm.

During recent years there has been an increasing trend towards the adoption of automated auction systems which tend to provide trading opportunities at lower costs that offered by dealer systems. Pioneering work by Amihud and Mendelson () has shown that the trading system can exert a strong influence over market quality. In their comparisons of auction and dealer markets they showed how the trading system can influence the amount of noise in security prices and their volatility. Pagano and Roell () and Pagano et al. () in their examination of securities cross listed on the Paris auction market and the London dealer market showed that bid‐ask spreads were higher in dealer markets. The influence of the trading mechanism on liquidity was also shown to be important by Huang and Stoll () who concluded that spreads for a group of matched securities on the NASDAQ were considerably higher than for NYSE securities. Easley and O'Hara () suggest that changes in information quality are important as poor information quality increases the role of private information creating a form of systematic risk.

The move from one trading system to another has also been linked to changes in trading volume. Amihud et al. () and Muscarella and Piwowar () documented that a move from call to continuous trading at the Tel‐Aviv Stock Exchange and the Paris Bourse respectively led to higher levels of trading activity.

The influence of the trading system on liquidity is important as there is growing evidence to suggest that liquidity influences the pricing of financial assets. For example, Amihud and Mendelson () and Elswarapu (1997) have documented a positive relationship between bid‐ask spreads and risk adjusted stock returns. Amihud () has shown that illiquidity is positively related to excess returns while Liu () and Marcelo and Quiros () have shown that illiquidity is an important component of priced risk. Changes to a trading system that alters liquidity can therefore influence firm value. Empirical proof of this relationship has been provided by Amihud et al. (), Muscarella and Piwowar () and Henke and Lauterbach () who all show that transfers from call trading to continuous trading are associated with a significant return premium.

The potential beneficiaries of changes in market architecture that are successful are wide ranging. Traders may benefit from lower trading costs but will also benefit from being able to make better decisions because information they receive is of a better quality. Shareholders may also benefit from a migration premium which increases the value of shareholdings. Firms also benefit because changes to the trading system can lower the cost of capital and lead to an increase in the success of marginal investment opportunities. A sufficient increase in marginal investment opportunities may even be able to act as a stimulus for economic growth.

We use the event study approach to examine abnormal returns around the time it was announced that securities were to migrate to SETSmm. We show that at the time SETSmm migration announcements were made there was a modest rise in firm value that is related to liquidity changes. Motivated by the studies of Chordia, Roll and Subrahmanyam (), that show that increased liquidity can improve efficiency, we examine whether liquidity and efficiency improves for migrating securities. Using the partial adjustment model with noise of Amihud and Mendelson () we examine the changes in pricing efficiency that arise in the post‐SETSmm period. We find that securities adjust more quickly to new information after SETSmm is introduced and find evidence of a decline in pricing inefficiency. Moreover, the magnitude of the efficiency gain is related to the SETSmm premium which accrues at the time of the announcement. We find that these changes are stronger for the most liquid securities in the sample. These changes are not discernible in a group of control securities.

Our results suggests that both potential traders and existing shareholders have benefited from the introduction of SETSmm. Traders have benefited because SETSmm prices have been of a better quality than SEAQ prices as they adjust to their true value more quickly providing more accurate information to the market. This is important as it allows investors to make more informed investment decisions. Existing shareholders also benefit from the introduction of SETSmm as they benefit from a migration premium which raises their investment return. The discovery of such a premium suggests that the cost of capital for migrating firms has fallen which will allow firms to undertake a greater range of capital investment projects at lower costs. Therefore the move from one trading system to another can have real effects on an economy.

The remainder of this paper is set out as follows. Section 2 describes the SETSmm trading system, Section 3 describes the methodology. Section 4 describes the data and provides some summary information. Section 5 provides the results. Section 6 examines whether the liquidity and efficiency changes are related to the migration premium. Section 7 provides the conclusions.

In its search for an optimal trading mechanism, during the last decade the London Stock Exchange made several changes to the way securities were traded on its exchange. Prior to October 1997 all trading on the London Stock Exchange was conducted through its dealer system called SEAQ(Stock Exchange Electronic Quotation System). This system required market makers to provide firm two way quotes during the official trading day. Moreover, market makers were prevented from quoting better prices on competing systems, although, they could improve upon quoted prices by engaging in telephone negotiations with clients. Board and Sutcliffe () showed that over half of all trades took place in this way during the period April 1992 to March 1994.

On 20 October 1997 the trading of FTSE 100 constituent stocks was transferred to a new auction trading system called SETS. In their analysis of intraday spread patterns, Cai et al. () compared SETS and SEAQ spreads and found that over some periods SETS spreads were about a quarter of those on SEAQ. At the time SETS was introduced a call auction opened the market and continuous trading took place until the market closed. In 1999 the most liquid FTSE 250 stocks were also placed on SETS, Lai () reports a rise in adverse selection following migration for these firms as evidenced by a rise in the realised and effective spread. To improve liquidity at the close a closing auction was introduced in May 2000, Ellul et al. () show that the closing call provides more efficient price discovery than the alternative dealer system but despite this significant amounts of order flow was still directed through dealers, especially if firms were less liquid.

On 3 November 2003 the London Stock Exchange introduced SETSmm. This new system combined the electronic order book of SETS with market maker participation and was directed at mid‐cap stocks with lower levels of liquidity than securities trading on SETS. SETSmm allows traders to place orders on an electronic order book which displays and executes them automatically according to price then time priority. As well as market and limit orders, at best fill or kill and iceberg orders are allowed to be placed. Every stock trading on SETSmm allows market makers to trade alongside the order book quoting two way prices, to ensure the continuous provision of liquidity. Designated market makers are able to submit orders (committed principal orders) either during continuous trading or during the call auctions but are not compelled to do so. The participation of dealers within the system rather than off market may provide greater stability in pricing. As shown by Amihud and Mendelson () dealers tend to smooth prices by ironing out temporary imbalances in order flow. By comparing price formation for stocks cross‐listed on SEAQ and on European auction systems Ellul () also finds that dealers have a stabilising effect. More recently Friederich and Payne () show that even off market dealer participation in London has led to more stable pricing.

A market maker can submit an order anonymously as a committed principal or they can reveal their identify. Member firms that engage in anonymous trading are restricted to one buy and one sell order per security at any time. There are no limits on named orders. Large trades made through designated market makers are governed by exclusive publication rules. Transactions of at least 4 × NMS (Normal Market Size) need not be reported until either 80% of the trade is offset or the end of the trading day is reached, whichever comes first. While, trades greater than 75× NMS have an extended publication delay of three days or 90% of the trade. Gemmill () discovers that publication delays following block trades at the London Stock Exchange do not influence liquidity or the speed that prices adjust to a block trade while Ellul () shows for a sample of cross‐listed stocks that information from a block trade has a significant...