تجزیه و تحلیل تجربی تامین نقدینگی توسط مردم محلی در بازارهای آتی: شواهدی از بورس آینده سیدنی

Contrary to the received view of market makers in theoretical literature, this study provides direct evidence that locals on the Sydney Futures Exchange (SFE) do not trade exclusively as passive market participants. In fact, rather than act purely as market makers, locals as a group are almost as likely to demand as supply liquidity. Further, locals trading on the floor of the SFE are less likely to supply liquidity when bid–ask spreads, trading frequency and price volatility are high, as well as around information announcements. These findings are consistent with aggressive trading by locals on the basis of a short-lived information advantage. This study also documents considerable diversity in the propensity of locals to supply liquidity, finding that it is related to the quantity, frequency and average size of their trading activity.

Trade on most of the world's leading futures exchanges takes place on a floor by open outcry. The Chicago Board of Trade (CBOT), Chicago Mercantile Exchange (CME) and Sydney Futures Exchange (SFE) are examples of leading international futures exchanges which operate floor-trading systems. In the typical futures floor-trading environment, transactions are executed by floor members and local members (herein locals). Floor members are typically institutions who trade both on their own account and on behalf of clients, whereas locals are individuals who trade on their own account. Exchanges have traditionally assumed that trading by locals enhances market liquidity.1 This study (1) examines the extent to which locals supply liquidity in futures markets and (2) determines the conditions under which they are most likely to supply liquidity. This study is important for several reasons. The findings have a number of policy implications. Locals were introduced by the SFE for the explicit purpose of enhancing market liquidity. This paper provides evidence of whether locals fulfil this role. This issue also extends to the ongoing screen trading debate surrounding futures markets as locals (and therefore any liquidity they supply) are unlikely to migrate from open outcry to a screen trading environment.2 In this respect, the paper is particularly timely given that the SFE and London International Financial Futures and Options Exchange (LIFFE) intend to move to fully automated trading during 1999.3 This study also has implications for theoretical literature on the economics of market making. The dominant theoretical model describing how market makers set bid and ask quotes is Stoll (1976), which predicts that the bid–ask spread is set to recover order processing, inventory holding and asymmetric information costs. A later version of this model by Ho and Stoll (1983), developed in a competitive framework, predicts that market makers seek to manage inventory levels around an optimal level, and recoup these costs, through offering a price concession (or discount) for inventory decreasing trades.4 In effect, long (short) inventory positions are managed by inducing sales (buys) through shifting both buy and sell quotes downward (upward). In apparent contradiction to Ho and Stoll (1983), Manaster and Mann (1996) find that inventory reducing trades by locals in futures markets are typically realised at better prices relative to other traders, than inventory-increasing trades. Such behaviour is inconsistent with the prediction that price concessions and discounts are used to induce inventory-reducing trades. An important assumption underlying the Ho and Stoll (1983) model is that market makers are purely passive and supply quotes (liquidity). This study tests whether local members participate as suppliers or demanders of liquidity. Resolution of this issue is important, since it permits evaluation of a fundamental assumption underlying existing literature. To our knowledge, this is the first study in futures markets to explicitly analyse the supply of liquidity by specific market participants. 5 It is related to a number of prior studies that examine participation by various market participants in other markets. For example, Madhavan and Sofianos (1998) analyse the trading activity of specialists on the New York Stock Exchange, Sofianos and Werner (1997) examine the activities of floor brokers on the New York Stock Exchange and Reiss and Werner (1997) examine trading between dealers on the London Stock Exchange. This study extends this research to locals in futures markets. The remainder of this paper is organised as follows. Section 2 describes the institutional detail, including the organisation of trading on the SFE and role of local traders. Section 3 provides some theoretical rationale for examining the supply of liquidity by locals under different market conditions. Section 4 describes the data and method used to document and analyse the trading behaviour of locals, while Section 5 reports the results. The final section provides a summary and suggestions for future research.

This study documents that local traders on the SFE participate in a significant proportion of total trading activity. However, contrary to received theoretical inventory control models, this study provides direct evidence that locals do not participate exclusively as passive market participants. Rather, locals as a group are almost as likely to demand liquidity as they are to supply it. This study provides strong evidence supporting Manaster and Mann 1996., which asserts that local traders are ‘‘active’’ position takers. When considered alongside Manaster and Mann 1996. and Duffy et al. 1999., this study paints an interesting picture of local member behaviour and suggests locals generally manage inventory levels without the use of premiums or concessions in bid and ask quotes. This implies they are not passive market participants. The study then considers the market conditions under which locals supply and demand liquidity. Specifically, locals are less likely to supply liquidity when bid–ask spreads are wide, when trading volume and price volatility are high, and around information releases. In contrast, locals are more likely to supply liquidity just before the market close. The study also finds considerable evidence that the propensity for locals to supply liquidity is related to their trading behaviour, in particular, the volume, frequency and size of their trades. This study has important policy implications. Locals currently receive conces- sions on exchange fees effectively a wealth transfer from non-locals. in order to encourage them to provide liquidity. This is despite the fact that a significant portion of their trades consumes liquidity. Further, locals are less likely to supply liquidity in times when it is most valued, and locals that trade more frequently and in larger size are less likely to supply liquidity. At the very least, these findings question the basis for providing locals with concessions on exchange fees. Further research could examine local trading in screen-traded environments such as SYCOMw, and may provide an interesting comparison of local behaviour in a screen-trading environment.