Essays on Investor Trading Activity in a Limit Order Book Market

Transcription

1 Essays on Investor Trading Activity in a Limit Order Book Market A thesis submitted to The University of Manchester for the degree of Doctor of Philosophy in the Faculty of Humanities 2013 Adeola Deji-Olowe. Manchester Business School, Accounting and Finance Division 0

2 TABLE OF CONTENTS LIST OF TABLES... 5 LIST OF FIGURES... 6 Abstract... 8 Declaration... 9 Copyright Statement Dedication Acknowledgement CHAPTER 1: INTRODUCTION AND OVERVIEW Introduction Thesis Chapter Overview and Contribution Price Discovery and Profitability - Analysis of Broker Activities in a Nascent Financial Market How Informative Are Investor Trades? An Analysis of the Price Impact of Investor Trades Analysing the Order Submission Strategy in a Limit Order Market CHAPTER 2: INSTITUTIONAL DETAILS MALTA STOCK EXCHANGE Institutional Details and Data Basic Features of the Malta Stock Exchange Market Historic Performance and Stock Index MSE Trading Procedures and Regulations Trading Sessions Membership - Brokers MSE Trading Procedures Order Submission Time in Force Restrictions (TIF) Volume Disclosure Order Submission Special Terms Order Alteration and Cancellation MSE Trading Procedures Order Execution Priority MSE Trading Procedures Trading Period The Pre-Opening Period

3 2.6.2 At The Open Period The Trading Period Data Description Stock Demand and Trading Activity Broker Activity and Trading Patterns Investor Activity and Trading Patterns Stock Order Submissions, Revisions and Execution Broker Order Submissions, Revisions and Execution Regulatory Framework and Environment TABLES CHAPTER 3: PRICE DISCOVERY AND PROFITABILITY - ANALYSIS OF BROKER ACTIVITIES IN A NASCENT FINANCIAL MARKET Introduction Literature Review Price Discovery Trading Profitability Institutional Details Trading Activity on the Malta Stock Exchange Data Methodology Analysis of Price Discovery Error Correction Model Weighted Price Contribution Empirical Results Price Discovery Does Investor Volume Drive Broker Share of Price Discovery? Profitability Conclusion TABLES CHAPTER 4: HOW INFORMATIVE ARE INVESTOR TRADES? AN ANALYSIS OF THE PRICE IMPACT OF INVESTOR TRADES Introduction

4 4.1 Related Research The Vector Autoregression (VAR) Model Institutional Details Structure, Trading Activities, Market Demand and Supply Data Data Clean Up Variables Empirical Results Quote Revision Equation Investor Class Price Impact Impulse Response Function Conclusion CHAPTER 5: ANALYSING THE ORDER SUBMISSION STRATEGY OF INVESTORS IN A LIMIT ORDER MARKET Introduction Literature Review Research Methodology Control Variables and Testable Hypotheses Relative Bid-ask Spread Investor Informational Asymmetry Methodology Ordered Probit Regression Ordered Probit Regression - Marginal Effects Institutional Details and Descriptive Statistics Order Submission, Order Execution Priority and Trading period Descriptive Statistics Order Aggressiveness Order Aggressiveness and Time of Day Investor Class Order Submission Pattern Empirical Results Investor Class Order Aggressiveness - Bid-ask Spread Investor Class Order Aggressiveness - Depth Investor Class Order Aggressiveness Order Size

5 5.8.4 Investor Class Order Aggressiveness Volatility Time of Trade and Order Submission Strategies Order Submission Strategies and Order Book Height Investor Class Information Asymmetry and Order aggressiveness Conclusion CHAPTER 6: Conclusion and Future Research Conclusion Directions for Future Research BIBLIOGRAPHY APPENDIX Total Word Count 50,001 4

6 LIST OF TABLES Chapter 2 Table 2.1 Equity Securities Listed on the MSE as at 30 November Table 2.2 Name of Member Brokers of the Malta Stock Exchange 35 Table 2.3 Listed Stocks and Volume Traded ( ) 36 Table 2.4 Summary of Stock Trade Days, Trade Count and Volume Traded 37 Table 2.5 Order Submission - Pre-opening and Continuous Trading Period 38 Table 2.6 Summary Broker Volume Traded and Number of Trades 39 Table 2.7 Summary Broker Trading Activity 40 Table 2.8 Summary Broker Trades - Pre-opening and Continuous Trading Period 41 Table 2.9 Summary Investor Class Trading - Analysis by Volume Traded 42 Table 2.10 Summary Top 10 Investor - Market Share 43 Table 2.11 Stock Order Submission, Revision and Cancellation 44 Table 2.12 Stock Bid - Ask Volume at Trading and Pre-opening 45 Table 2.13 Broker Order Submission, Revision and Cancellation 46 Table 2.14 Broker Bid - Ask Volume at Trading and Pre-opening Period 47 Chapter 3 Table 3.1 Broker Market Share - Total Volume Traded and Number of Trades 68 Table 3.2 Summary of Stock Trading Activity Order Submission 69 Table 3.3 Broker Trading Activity Order Submission 70 Table 3.4 Error Correction Model Results 71 Table 3.5 Weighted Price Contributions Results 72 Table 3.6 Weighted price Contribution Result by Broker Type 73 Table 3.7 Summary of Proportion of Volume Traded by Distinct Investor Class and Broker Client Structure 74 Table 3.8 Empirical Analysis of Broker Profitability 75 Table 3.9 Broker Profitability Analysis 75 Chapter 4 Table 4.1 Listed Stocks and Volume Traded ( ) 98 Table 4.2 Investor Class - Analysis by Volume Traded 98 Table 4.3 Summary of Proportion of Stock Volume Traded - Distinct Investor Class 99 Table 4.4a Estimated Coefficients for Quote Revision per Stock Return 100 Table 4.4b Estimated Coefficients Trade Equation - Trade Direction 101 Table 4.5a Estimated Coefficients Quote Revision - Return (Signed Volume) 102 Table 4.5b Estimated Coefficients Trade Equation - Signed Volume 103 5

7 Table 4.6a Estimated Coefficients Quote Revision - Return (Signed Volume Sq.) 104 Table 4.6b Estimated Coefficients Trade Equation - (Signed Volume Sq.) 105 Table 4.7 Impact of Investor Trades on Stock Price - Return and Trade Equation 106 Table 4.8 Estimated Coefficients for Multivariate VAR per Stock 107 Chapter 5 Table 5.1 Summary Stock Trading Activity 148 Table 5.2 Summary Investor Trading Activity 148 Table 5.3 Distribution of Order Aggressiveness level during Trading Day per Stock 149 Table 5.4 Distribution of Order Aggressiveness level during Trading Period 150 Table 5.5 Order Aggressiveness per Investor Class 151 Table 5.6 Distribution of Order Aggressiveness (Investor Class) 152 Table 5.7 Ordered Probit - Investor Class Order Aggressiveness 153 Table 5.8 Marginal Probabilities - Investor Class Order Aggressiveness 158 Table 5.9 Investor Class Order Submission and Order Aggressiveness 159 Table 5.10 Diagnostic Result for Ordered Probit Regressions 160 LIST OF FIGURES Chapter 2 Figure 2.1 Malta Stock Index Index 36 Chapter 4 Figure 4.1 Impulse Responses to Shocks for Stock BOV 109 Figure 4.2 Impulse Responses to Shocks for Stock MLC 110 Figure 4.3 Impulse Responses to Shocks for Stock HSB 111 Figure 4.4 Impulse Responses to Shocks for Stock MSI 112 Figure 4.5 Impulse Responses to Shocks for Stock MIA 113 Figure 4.6 Impulse Responses to Shocks for Stock IHI 114 Chapter 5 Figure 5.1 Distribution of Order Aggressiveness during Trading Day 161 APPENDIX Table A1 Global Competitiveness Report Table A2 Market Concentration Market Capitalization 176 Table A3 Percentage Trade Count - Trade per Broker per Stock 177 Table A4 Summary Broker/Broker Relationship - Sell Trades 178 Table A5 Summary Broker/Broker Relationship - Buy Trades 179 6

8 Table A6 Summary Listed Stocks Volume Traded 180 Table A7 Summary Investor Class - Analysis by Volume Traded 180 Table A8 Description of Distinct Investor Class 181 Table A9 Summary of Stock Activity Analysis by Stock 182 Table A10 Summary of Stock Activity Analysis by Investor Class 183 7

9 Abstract The University of Manchester Adeola Deji-Olowe Doctor of Philosophy (PhD) Essays on the Impact of Investor Trades in a Limit Order Book Market November 2013 This thesis consists of three essays examining the impact and consequences of the trading behaviour of a finely disaggregated category of investors in an electronic limit order book equity market, the Malta Stock Exchange (MSE). The three essays in market microstructure are closely related and examine how investor heterogeneity impacts the informational content of the limit order book, the informational content of individual trades, the price impact of investor trades, the aggressiveness of order submission strategies and the price discovery process within such a market. The first essay investigates the role of the financial intermediary in the price discovery process in a limit order market. We address this issue by analysing the trades of brokers within the Malta Stock Exchange by comparing the profitability of their individual trades and the impact of these trades on the price discovery process. The results of a Weighted Price Contribution methodology indicate that more active brokers that dominate the market in terms of volume and amount traded account for a significant portion of the price discovery process. We also find that the level of profitability of these brokers is directly proportional to the amount of volume traded and their relative share in the price discovery process. This appears to rule out the possibility of manipulative trades by these brokers in order to influence profitability The second essay examines the price impact of the order flow emanating from finely disaggregated classes of investor with the aim of determining whether detectable differences exist in the extent to which orders emanating from particular groups of investors impact on the evolution of stock prices. On the aggregate stock level, results indicate price impact is inversely related to liquidity and as such the price impact of trades is of a higher magnitude and significance in stocks that are less liquid. Significantly, we find that stocks with higher liquidity and trading volume adjust quickly to price changes and the cumulative impact is realised earlier for these stocks. Similarly, for investor classes, our results show that the magnitude and significance of individual price impact increases as liquidity of the stocks declines, showing that as liquidity increases in the order book, the impact of information asymmetry begins to diminish. Institutional investors consistently have the highest significant impact on the evolution of prices across all the stocks. The final essay examines how investors structure their order submission choice in response to changes in the limit order book and market conditions (such as order depth, volatility, returns, and height of the limit order book). We identify 7 distinct investor classes who differ in their trading requirement and the information set available to them, and as such we expect that these investors will adopt different strategies to maximise their trades. The results show variability in the submission strategies adopted by investors as trade sides changes from buy to sell trades. It also indicate that investors have to balance between execution risk, the timely use of private information and the risk of being picked-off by other informed investors. In analysing the varied responses of these investors, we find that the order submission strategies adopted is most responsive to the risk of non-execution. 8

10 Declaration I, Adeola Deji-Olowe, declare that no portion of the work referred to in this thesis has been submitted in support of an application for another degree or qualification of this or any other university or other institute of learning. 9

11 Copyright Statement i. The author of this thesis (including any appendices and/or schedules to this thesis) owns any copyright in it (the Copyright ) and s/he has given The University of Manchester certain rights to use such Copyright, including for any administrative, ii. Copies of this thesis, either in full or in extracts and whether in hard or electronic copy, may be made only in accordance with the Copyright, Designs and Patents Act 1988 (as amended) and regulations issued under it or, where appropriate, in accordance with licensing agreements which the University has from time to time. This page must form part of any such copies made. iii. The ownership of certain Copyright, patents, designs, trademarks and other intellectual property (the Intellectual Property ) and any reproductions of copyright works in the thesis, for example graphs and tables ( Reproductions ), which may be described in this thesis, may not be owned by the author and may be owned by third parties. Such Intellectual Property Rights and Reproductions cannot and must not be made available for use without the prior written permission of the owner(s) of the relevant Intellectual Property Rights and/or Reproductions. iv. Further information on the conditions under which disclosure, publication and commercialisation of this thesis, the Copyright and any Intellectual Property Rights and/or Reproductions described in it may take place is available in the University IP Policy (see in any relevant Thesis restriction declarations deposited in the University Library. The University Library s regulations (see ) and in The University s policy on Presentation of Theses.. 10

12 Dedication I dedicate this work to my two beautiful children, Dara and Dabiolu 11

13 Acknowledgement The completion of this thesis has been one of the most challenging yet rewarding achievements of my life and I have received guidance, support and encouragement from several people along the way. I would like to thank God for saying a thing and performing it to the utmost. Through time and the many challenges, the unshakeable presence of God, the confidence in His grace and His profound love helped me keep it together and navigate the waters. I would like to express my sincere and heart felt gratitude to my supervisors, Professor Mike Bowe and Professor Stuart Hyde, for their persistent support and encouragement. Their insightful knowledge in the study of market microstructure and their constant patience were critical to the completion of this study. During the times when I did not believe this work could be done, they provided the necessary counsel and motivation to ensure that the work was done. I am eternally grateful and indeed very blessed to have these two outstanding people in my corner. I would like to thank the members of the doctoral committee of the Manchester Business School who over time have provided valuable suggestions on the direction of the study. This thesis has benefited from several comments and suggestions made by the participants at the Eastern Finance Association (2012) and Mid-Western Finance Association (2012) meetings and I am thankful for the platform provided to discuss my study with other participants. I would like to thank my friends and my family for the part they played in ensuring this dream came to pass. Without the support both financially and emotionally, this PhD would not have happened and I am eternally grateful. Finally, I would like to thank my children Dara and Dabiolu for putting up with me in the years it took to complete this journey. This work is indeed dedicated to them. 12

14 CHAPTER 1: INTRODUCTION AND OVERVIEW 1.1 Introduction Understanding the different equilibrium reaction of asset prices to trading activity initiated by the various classes of market participants posits one of the fundamental quests of the market microstructure literature; specifically, which trades move prices and does the trading activity of certain classes of traders move prices more than others?. The trading activities of different classes of investors and the subsequent impact of their trades on price movements in financial markets has been the subject of extensive theoretical and empirical research in the recent market microstructure literature. This thesis investigates the order flow emanating from finely disaggregated classes of investor within an electronic limit order book market to provide insights into the trading strategies adopted and the overall impact of these activities. Investors trade for a variety of reasons which may incorporate liquidity requirements, speculative trading to portfolio rebalancing. Focusing on the execution of investor trades, we study the order submission strategy and the subsequent price impact of these trades. Our objective is to determine whether detectable differences exist in the extent to which orders emanating from particular groups of investors impact on the evolution of stock prices. Empirical studies on investor order flow, price impact and order submission strategy have proliferated in recent times due to the availability of high frequency intraday data. The availability of high frequency data has enabled the study of the informational content of the limit order book, the price discovery process, the price impact of individual trades and investor order submission strategy. Whilst these studies use order flow data from limit order markets, very few of these studies utilise data that identify the particular investor/agent that is responsible for the trade. The data used in this thesis is more extensive and more finely detailed than in the vast majority of prior studies. Not only does it contain the tick by tick activity of the broker, time stamped prices, the trader type, the counterparty on the buy and sell side; it also contains the numbered identity of the investor on whose behalf the broker is trading. The content of this data allows for a unique study of investor trading activity and the aim of this thesis is to explore the interaction between heterogeneous investors, the limit order book and the price dynamics within a financial market. 13

15 1.2 Thesis Chapter Overview and Contribution In this thesis, we examine the three related research questions using order flow data emanating from finely disaggregated classes of investor within an electronic, order driven financial market, the Malta Stock Exchange. The MSE data set is unique and extensive because it contains details about the actual investor trading, the direction of the trade, the investor class of the initiating trader, the counterparty to the trade, time stamped price of quotes and actual trades that occurred. This structure helps facilitate in-depth analysis of investor behaviour and the subsequent effect of the trades on the evolution of price in the market. Chapter 2 presents an in-depth descriptive overview of the MSE, laying out the microstructure details and trading rules enforced in the market. Chapter 3 examines the role of market intermediaries by analysing broker trading activity and its impact on the price discovery process and profitability. Chapter 4 investigates the trading activity of investors and the subsequent effect of their trades on the evolution of price in the market. Chapter 5 investigates the dynamic reaction of a finely disaggregated class of investors to the limit order book. Chapter 6 concludes and provides some directions for future research. A brief discussion of the novel research contributions which constitute the subject matter of chapters 3 through 5 is now highlighted Price Discovery and Profitability - Analysis of Broker Activities in a Nascent Financial Market The role of financial intermediaries in financial markets, the implication of their trades on profitability, return predictability and the impact of their informational advantage in the price discovery process remains a largely unresolved question. In nascent markets where the role of financial intermediaries is intensified and necessary for market development and growth, it is potentially more important to confront the issue of asymmetric information whose impact may inhibit market efficiency and growth. This research aims to analyse the impact of the broker activity and any informational advantages they possess on the price discovery process within one such market, the MSE. We analyse the informational content implied in investor trades in order to ascertain if certain brokers have access to price relevant information by virtue of their trading activity, physical location and ability to observe order flow that they can translate into trading strategies to earn profits. This paper seeks to advance the literature on the informational content on broker trades, profitability and price discovery within equities market and makes three significant 14

16 contributions to existing literature. First, prior empirical studies on broker profitability and price discovery in emerging markets (such as Kwaja and Mian, 2005; Uppal and Mangla, 2006; Aggarwal and Wu, 2006; and Felixson and Pelli, 1999) use daily aggregated trade data. And as far as we are aware, this study is the first to use high frequency data (tick by tick intraday data) to analyse broker informational content and activity in an emerging market. Second, we extend the literature on price discovery within the same equities market. As to our knowledge, this is the first paper to look at price discovery amongst brokers within the same market without having to compartmentalise the brokers into different groups (for example, local versus foreign brokers, market intermediary versus client). Third, using this unique data set, we test for broker profitability in the context of individual stocks and relative to the investor base of the broker. The dataset contains precise information (including the account number and investor type) of the investor and as the trades originated within the MSE are initiated at the behest of the investor; we are able to directly identify profitability as a reflection of investor trading activity. The market dominance by a few brokers and limited competition creates a potential avenue for brokers to carry out manipulative trades in order to earn profits. Hammad (2007) in a related study suggest that increasing the level of broker competition will reduce the competitive edge of dominant brokers and their propensity to carry out such trades. This paper contributes to enhancing the development and microstructure design of nascent markets and the strengthening of market regulation How Informative Are Investor Trades? An Analysis of the Price Impact of Investor Trades The trading activities of different classes of investors and the subsequent impact of their trades on price movements in financial markets have been the subject of recent extensive theoretical and empirical research in market microstructure literature. In this paper, we study the price impact of the order flow emanating from finely disaggregated classes of investor within an electronic, order driven financial market. Our objective is to determine whether detectable differences exist in the extent to which orders emanating from particular groups of investors impact on the evolution of stock prices. Earlier research examining the price impact of individual trades concludes that the difference in the price impact of these trades is due to the informational content of trades. Whilst these previous studies largely disaggregate investors into individual and institutional classes, a distinctive contribution of 15

17 this study is that we investigate a finer range of investors without the need for this generalisation. Consequently, we examine the order flow of 7 different investor groups using data from an order-driven market to test the private information implicit in their individual trades. These 7 investor groups are the resident individual, resident collective insurance scheme, non-financial private entity, resident credit institutions, resident insurance companies, auxiliaries and other European countries credit institutions. Our paper makes several contributions to the literature on the price impact of investors in the equity markets. The first relates to an accurate and definitive classification of investor classification. Academic research on the informational content of investor trades has depended on a variety of indicators, such as trade size, geographical location, trading platform utilised and order submission pattern in order to determine the categorisation of an investor as either institutional or individual; informed or uninformed (Lee [1992]; Anand et al [2005];Griffin et al [2003]; Chakravarty [2001]; Linnainmaa [2010]). However, while at times persuasive, these generalisations are never definitive in determining investor classification. In contrast, the data available for utilisation in this study enables a more definitive classification to occur without resorting to such generalisations. The MSE data set is unique because it contains details about the actual investor trading, the direction of the trade, the investor class of the initiating trader, the counterparty to the trade, time stamped price of quotes and actual trades that occurred. This structure helps facilitate in-depth analysis of investor behaviour and the subsequent effect of the trades on the evolution of price in the market. The study of investor behaviour is particularly important in new, emerging equity markets with relatively low liquidity such as the MSE. This is because the price impact of trades emanating from certain investor class has the potential to impact the overall direction of the market. The results emanating from such analyses can be used by regulators in emerging markets to create market structures and design regulatory policies that effectively protect all investor classes, reducing the crowding out effect of trading by large institutions that tend to account for an important share of trading in these markets. Second, by sorting all investors in the market into specific investor classes, we are able to test and analyse whether the price impact of investor trades varies across investor classes in relation to the liquidity and other characteristics of the specific equity being traded. We begin by estimating the price impact of trades for each stock included in our sample and then we analyse the impact of the trades of each investor class on the evolution of the price. In addition, we study the variation in the 16

18 price impact of investor class trade across stocks with different characteristics. This is important and interesting to the extent it may reveal a level of asymmetry in the price impact of the different investor classes across stocks of varying liquidity (in terms of volume traded and number of trades) Analysing the Order Submission Strategy of Investors in a Limit Order Market In limit order markets, investors are faced with two major trading decisions; the decisions to demand immediacy by placing market orders or to supply liquidity to the market by placing limit orders. Along with the information on the fundamental value of a stock, an investor s decision to place a limit or market order is a complex relationship between the state of the limit order book and the order placement strategies adopted by other investors. These order submission strategies are diverse and in this chapter we are interested in understanding the relationship between the state of the limit order book and the marginal reaction of investors to the state of the order book. The strategic behaviour of investors in placing limit or market orders is fundamental to the study of market dynamics. The difference between placing a limit order and a market order in a limit order market is captured in the price at which the order is placed and the immediacy with which it is required. The investor faced with a continuous trade-off between price risk, order execution, and adverse selection risk has a constant need to review their optimal trading strategy in order to trade profitably. In modelling the investor decision making process, several empirical and analytical studies have been carried out to offer potential explanations for the price aggressiveness of investor trades in response to the extant state of the limit order book when they submit their orders. By employing a classification based on price aggressiveness to measure the strategic order submission process adopted by different investor classes, we intend to determine if these strategies can be interpreted as a reaction to the state of the limit order book. The state of the order book (e.g. spread, depth), market related variables (e.g. volatility, time of trade) and the endogenous relationship between price aggressiveness and quantity will be analysed to determine the relationship between strategic investor order submission behaviour, the market trading mechanism and order flow. Very few studies examine the marginal response of investors to changes in the limit order book and market conditions in the light of order submission (see for example Beber and Caglio [2005] and Menkhoff et al [2010], Duong et al [2009]). To our knowledge, this is the 17

19 first paper to analyse and compare the marginal response of a broad range of investors (representative of an entire financial market) to changes in the limit order book and other market conditions. Whilst the above papers distinguish between informed and uninformed traders, only Duong et al (2009) is able to present an in-depth investigation of order aggressiveness taking into account a narrower categories of traders (individual and institutional). Our paper is able to extend the results of these papers by further taking into account investor heterogeneity without the usual broad categorisation of informed vs. uninformed and individual vs. institutional traders. Also, whilst these papers are not able to differentiate between broker trades and client trades, our data is robust on client level information and our analysis is carried out only for actual investors (in which case, broker trades are not in any way implied or reflected in our study). Our analysis reveals that institutional investors alter their trading strategies to improve the probability of execution. These investors are able to receive timely information and reflect their sensitivity to the risk of non-execution by placing aggressive trades when volatility increases. The resident individual on the other hand appears to protect itself from these traders by placing limit orders in a bid to reduce the risk of being picked-off by these investors. 18

20 CHAPTER 2: INSTITUTIONAL DETAILS MALTA STOCK EXCHANGE 2.0 Institutional Details and Data This study analyses trading activity on the Malta Stock Exchange (MSE). Malta is a nascent market which through the adoption of an advance legislative and regulatory framework is becoming increasingly popular as an active financial hub and fund jurisdiction in Europe. A member of the European Union, Malta is a high income country with a GDP of about $8.1 billion (2009 estimate) with services accounting for 80.6%, agriculture 1.% and industry 18% of this figure. Though a small market, it is an active financial market organized in accordance with practices in advance economies based on international best practices. Appendix A1 shows the Global Competiveness Index report ( ) provided by The World Economic Forum on financial markets worldwide and their ranking. The report shows the ranking of all aspects of the Maltose financial market, ranging from market sophistication to the strength of auditing and reporting standards. Constructed following an analysis of 134 countries, the positive review of the index affirms the status of Malta as an international financial services country and also highlights the fact that though it is a nascent market, it is fairly sophisticated in its operations. This section highlights relevant features of the market that facilitate this study. 2.1 Basic Features of the Malta Stock Exchange Established following the enactment of the Malta Stock Exchange Act of 1990, the Malta Stock Exchange (Borsa Malta), the only market licensed to operate in Malta, commenced trading operations in The Malta Stock Exchange (MSE) was established to act as operator and regulator of the capital markets where its responsibilities included the listing of new shares, trading of existing shares, licensing of approved stock brokers, monitoring of market participants, enforcing regulations and also acting as the operator of the Central Securities Depository. Following its start-up operations in 1992, stocks were traded on a weekly basis until 1998 when the exchange started to meet to trade on a daily basis..organised as an order driven market with about 10 licensed stockbrokers, trades were initially undertaken manually using a call over system where the names of the stocks are called out for bids and offers. In 1999, 19

21 an electronic trading system was introduced, and by the end of 2001 the exchange introduced remote trading where members were able to trade located away from the floor of the exchange. At this time, MSE introduced the electronic trading order book and member brokers could trade remotely from their respective office using a computerised on-screen trading system. Currently, the exchange uses the Malta Automated Trading System (MATS), which is a remote computerised system, as its trading platform. The financial instruments trading on the exchange include equities, corporate bonds, collective investment schemes, Maltese government bonds and treasury bills. As at the end of 2007, the total market capitalisation was 7.7billion, a figure which excludes collective investment schemes. The equity market contributed 3.7billion to the total market capitalisation whilst the corporate bond market, Malta Government bonds, and the treasury bills market contribute 484million, 2.9billion and 352 million respectively. With 16 listed stocks as at 2007 and 13 member brokers acting on behalf of their clients, the equity market represents 37% of total market capitalisation. Table 2.1 shows the list of all shares which were listed and traded on the floor of the exchange and Table 2.2 shows a list of registered and authorised member brokers allowed to trade on the exchange. During our sample period (January 2003 June 2007), the equities markets listing remained remarkably stable with an addition of only 1 stock (Medserv listed in 2006) within that period. However, the exchange witnessed several admittance to listing for existing equities mostly for issues subsequent to bonus issues, right issues and pursuant to conversion of shares into euros. 2.2 Market Historic Performance and Stock Index The aggregate measure of historical investment performance produced by the exchange is provided by the MSE index. The index, which is the only official index compiled on the MSE, includes all the shares listed on the exchange and it is weighted by the capitalisation of all component stocks and collated using the last closing price of these stocks. The MSE index started with a base of 1000 on December 27, 1995 and as at December 27, 2007 the index stood at The share index highlights two peaks in 1999 and In 1999, the privatisation of government holdings and the growth of collective investment schemes were the two major driving forces in the stock market. Due to heavy demand and low supply of traded equities, share prices which had shown an increase in 1998 rose steeply in 1999 with the index trebling in the last quarter of this financial year. Following the announcement that Malta was joining 20

22 the European Union in 2004, the capital market started to witness an upward trend in 2003 and this continued to gather momentum in Monetary growth also accelerated in 2005 and this expansion resulted in an increase in the turnover of the equities market. Therefore, the index which had experienced a recovery in 2003 continued an upward trend and as prices of the stocks increased, the market witnessed a sharp rise in the share index in Following a rigorous year of compliance with its eligibility criteria, the FTSE Index Company, in September 2009, confirmed the inclusion of the Malta s equity markets within its Global Equity Index Series (GIES). The FTSE GIES, which includes 50 countries, covers over 8,000 equities and 98% of the world financial market capitalisation, is a much sought after index. The inclusion of the MSE in this index is seen as a positive development, which is bound to increase the functionality, presence, infrastructural development and attract potential clients to the exchange. 2.3 MSE Trading Procedures and Regulations Trading on the exchange is open to authorised brokers, officials of the exchange and other person authorised by the exchange. All trades transacted on the floor of the exchange are required to be undertaken during official trading hours through the Malta Automated Trading System (MATS), and any violation to this rule must be authorised by the Exchange Trading Sessions The timing of the market trading session is determined by the Board of the Exchange periodically, and the opening and closing period may vary for the different instruments trading on the exchange. This is done as deemed appropriate by the board, and the market is subsequently notified. The equities market is traded in two daily sessions, the pre-opening period and the continuous trading session. During the period under review, the pre- opening period starts at 8.30 am and finishes at 10.am while the continuous trading session opens at am to close at 12.30pm. All trades are transacted within the stipulated hours, and any trade that occurs outside of these hours must have the express permission of the Exchange Chairman, who must judge the trade to have little impact on the market. 21

23 2.3.2 Membership - Brokers To become a member and trader on the floor of the exchange, a person must (i) be licensed by the exchange, under the Investment Services Act (Cap 370 of the Laws of the exchange), (ii) be an European Investment firm, and (iii) must be assessed by the Exchange to have adequate trading skills, financial resources and organisational structure. Prior to becoming a member, an application must be sent to the exchange, with an adherence to the compliance requirements. Following the review of an application and before the issuance of a membership, the applying member must provide evidence of cleared financial assets pledged to the exchange. Subsequent to submitting an application form to the exchange and the application being approved by the Board, all traders are expected to undergo a training course and undertake a test prior to authorisation being granted. Membership will be granted only if at least one trader on the application form passes this test. This is a requirement to show the level of competency and trading abilities of the prospective traders. Trading on the MSE is done remotely through the MATS and the number of trading platforms of each member firm is not to exceed the number of dealers licensed to trade for that firm 2.4 MSE Trading Procedures Order Submission The transaction of financial assets, buy and sell, on the floor of the exchange is carried out on the MATS platform in a manner prescribed by the Exchange. For a transaction to be deemed valid, it must follow the rules set by the exchange with respect to all aspects of the trading; timing of the trade, specification of the order, tick size, order submission, trade ranges, to mention a few. The MSE is an order driven market and it operates an open order limit book, where all members are able to view the order submission and specification, including the order price, disclosed volume, time of trade and the identity of the member submitting the order. Given by Section of MSE byelaws (Trading Procedures and Regulations) the MATS allow the submission of three types of order submission. A Limit order specifies the price of the trade when it is posted into the system and the price limit is not violated by trading above or below the price given at the point of posting into the system. A Market order on the other hand when posted does not specify a price and the trade is expected to transact promptly and at the best price currently available. In the MSE, a Market order has priority over limit orders at the same price levels and is permitted to trade through a range of prices starting from the best available prices. However, the range of prices 22

24 through which the market order is permitted to walk through is set and may be limited by the Exchange. Lastly, a Range order is a limit order that allows for a trade to be executed within a range of prices. What this implies is that the order is able walk-through the market going beyond the best market price on the other side of the market into the depth of that market side, with the specified price range acting as its restriction. All orders, be it limit, market or range orders are allowed to have special features as allowed by the MSE. These special features include: Time in Force Restrictions (TIF) As given by Section of the MSE Byelaws, orders are allowed to have special time restrictions which limit the time given for that particular trade to be executed. A trade is allowed to have a TIF restriction of a day, open, good to date, good for week, good for month or Fill or Kill restriction. Under the TIF of fill or kill (FOK), the order is executed either in whole or in part, or it is cancelled within a period after receiving the order for execution. As soon as a FOK is posted, the portion of the order that can be filled is executed and any remaining portion is cancelled, and a FOK order that is not likely to be filled given market price is not posted. For all orders that have TIF restrictions, the order is automatically cancelled from the order book once the stipulated time elapses Volume Disclosure Orders submitted are permitted the flexibility of volume disclosure and are allowed to disclose total volume or specify any amount lower than the total volume of the order. Section to of the MSE Byelaws specifies the rules guiding the disclosure of volumes during order submission, and specifies that a disclosed volume can be entered for an order at the point of posting for execution. The disclosed volume must not exceed the total volume and where the order does not disclose its volume, the undisclosed volume is assumed to be the total volume. Since the MATS operates an open order limit book, once an undisclosed amount is posted, members can only see the member posting the order, but the volume of the order remains hidden. However, an indication will be given to members that an undisclosed amount is present. In the case of a partial disclosure, the disclosed amount acts as a roll-in quantity for the total order submitted. Following a partial fill, the disclosed volume is only rolled over if there are no orders behind at the current price level. However, if volume of an order can be 23

25 executed upon submission, then both the disclosed and the undisclosed volume will be filled to the extent of the total volume of the order Order Submission Special Terms The priority and processing of an order can be affected by the special terms attached to the order at submission to the MSE, these terms are such that once an order has a special term attached to it, it is not logged into the regular order book but into a special terms order book, which can be viewed separately by members. Section of the MSE Byelaws specifies what is permitted as a special term and the rule of the Exchange is that while special term orders can be posted during the pre-opening period, they are not traded until the market is opened. The allowable special term orders are the All or None, Minimum Fill and the Minimum Block orders. For the All or None (AON) special term order, the total volume of the order is either filled in its entirety or the order is not executed at all. The Minimum Fill (MF) order must specify a minimum volume that must be totally filled and any outstanding volume of the order may be traded in partial sizes after the specified minimum volume is filled. The Minimum Block (MB) on the other hand, also specifies a minimum volume that must be filled in entirety as in the MF, but subsequent trades will be disclosed in blocks using the lesser of the initial minimum block or the remaining volume of the order. At a single price limit, all limit order within the regular order book must be filled before any special term order is transacted Order Alteration and Cancellation Section and of the MSE Byelaws allows for a broker changing the terms of an order that has already been posted for execution. The MSE specifies that once an order is posted, a broker can still change the terms of the order which may include the order type, the financial instrument being traded, the price, volume, underlying client and any special term that may apply. However, once an alteration is made, the time priority is changed or the order maybe cancelled and then re-entered. If the broker wishes to change the financial instrument being traded or the order type (buy or sell) then they are required to cancel and re-enter his order. If there is a change in price, increase in disclosed volume, removal of a special term or a change in the underlying client, then a new effective time stamp is given to the order which immediately affects its time 24

26 priority. However, if the alteration is a decrease in disclosed volume, changes in disclosed volume, changes in special terms or changes in the time in force restriction, then the order will keep its initial time stamp and hence its time priority. Any other change done to an order that has not an been specified will have no effect on it time priority and an order can be cancelled as long as it has not been matched 2.5 MSE Trading Procedures Order Execution Priority Section gives the procedure for queue priority and orders that cannot be executed immediately upon submission are queued in the limit order book in a specific manner based on a queue priority. In the order of importance and priority, the first factor is that determines queue priority in the MATS is price. A market or a limit order at the point of posting for execution will specify a price and for both buy and sell, this price will determine its priority. Therefore, the higher priority price will be deemed the better price and it will command a higher queue priority. This implies that a sell order that has the lowest price will have a higher queue priority and a buy order with the highest price will maintain a higher queue priority. Since orders with special terms can be viewed separately from a special terms order book, all special terms order; All or none, Minimum Fill and Minimum Block, are treated with a lower queue priority from those without special terms. Within a single limit price, orders with fewer restrictions are given a higher queue priority and when multiple special term orders are posted with the same price limit, the time priority moves into force. Once an order is entered into the MATS, it is assigned a time stamp to reflect the time of entry and date posted, the time of entry determines its queue priority on a first- in- first- out basis. Most importantly, when multiple entries are posted with the same price limit, the order with the earliest time stamp takes queue priority. Orders submitted during a previous trading session are given queue priority over those posted during the current trading session. For any two orders received at the same time, priority will be assigned by the MATS using a random factor. 2.6 MSE Trading Procedures Trading Period The MATS operates two trading session within a daily period; the pre-opening and the continuous trading period. The pre-opening period is held prior to the start of the regular 25

27 trading period and here brokers are able to submit orders but these orders are not processed immediately. However, though not processed, these orders are queued in accordance with the queue priority rule as stated above, ready for the at the open period. The opening period is a special period that happens immediately after the pre-opening period and this period initiates the trading session The Pre-Opening Period During the pre-opening period, brokers are permitted to enter orders, cancel previously placed orders, make changes to orders and carry out inquiries on orders already posted into the system and view the order book. The time of the pre-opening is determined by the Board of the exchange and during the period under review, the pre-opening period starts at 8.30am and ends at 10.00am. Although orders can be posted during the pre-opening, they are not immediately executed and once an order is posted, the system will automatically verify if that particular instrument is expect to open and calculates the expected opening price in accordance with the Opening Algorithm. For an instrument to open there must be at least two orders on opposite sides of the order book with prices that can facilitate a deal. In the case where an imbalance is created during the pre-opening where the best bid price is higher that the best offer price, then the opening price will be chosen based on the maximum shares to be traded, minimum imbalance in share volume, least change from the previous trading session and the highest share price; all given in order of priority. At the pre-opening period, if a broker submits an order that has a price that is better than the current price predicted for the opening on the system, the information will remain private to the member submitting the order and the Exchange. Although all members can view the order book, they will not have access to the exact price and this order will be treated as if it has been posted at the current predicted opening price. Only the member that posted the order and the exchange will know that the order has created a market imbalance. However, since the opening price is predicted by the lowest ask, the highest bid and the volume of orders posted, the opening price can be expected to change as bids are continuously posted At The Open Period At the open period and following the end of the pre-opening period, the opening price is calculated based on the Opening Algorithm with the aim of maximising the volume that can be executed at that price. The open period signifies the start of the trading period and once an opening price is allocated; all orders that can be executed at that price are transacted. 26