Dark Pools and Flash Trading. New trends in Equity Trading?

Contents

List of Figures

List of Tables

List of Abbreviations

1. Introduction

2. Equity Trading and its new phenomenons - Definitions and Characteristics
2.1 What is Equity Trading?
2.2 Dark Pools - Definition
2.3 Dark Pools - Rationale
2.4 The Trading Framework
2.4.1 Pre-trade phase
2.4.2 Trade phase
2.4.3 Post-trade phase
2.5 Orders
2.5.1 Market Orders
2.5.2 Limit Orders
2.5.3 Peg Orders
2.5.4 Hybrid and Complex Orders
2.5.5 Order Parameters
2.5.5.1 Display Parameters
2.5.5.2 Quantity Parameters
2.5.5.3 Time in force Parameters

3. Different types of Market Structures and Market Liquidity
3.1 Physical and electronic markets
3.2 Continuous Markets
3.3 Quote-driven and order-driven markets
3.4 Displayed and nondisplayed markets
3.5 Market Liquidity
3.5.1 Block liquidity
3.5.2 Supply and demand for liquidity

4. Pricing in the dark pool sector
4.1 Price discovery
4.2 Price derivation

5. Regulatory Framework and Control in Europe
5.1 Regulatory Framework in Europe - MiFID
5.2 Financial Regulation and Dark Pools
5.3 Reporting and transparency

6. The structure of dark pools
6.1 Exchange orders and brokers as sources of dark liquidity
6.2 Multilateral Trading Facilities (MTFs) as sources of dark liquidity
6.2.1 Electronic limit order books
6.2.2 Crossing Networks/Price Reference Systems
6.3 Broker desks as sources of dark liquidity
6.4 Direct market access (DMA) as source of dark liquidity
6.5 Hybrid business models as sources of dark liquidity
6.6 Market overview
6.7 Dark sector evolution

7. Trading in the dark
7.1 Execution issues
7.2 Trading Strategies
7.2.1 Block Trading
7.2.2 Program Trading
7.2.3 Algorithms and Algorithmic Trading
7.2.4 High Frequency Trading
7.2.5 Gaming
7.3 Aspects of Technology
7.3.1 Order Management Systems and Execution Management Systems
7.3.2 Routing Engines
7.3.3 Matching and Pricing Engine
7.3.4 The FIX Protocol

8. Conclusion - The future of dark pools and flash trading

List of References

List of Figures

Figure 1: Market Clearing Price

Figure 2: Daily market share equities (FT, Trading Room, 2012)

Figure 3: The Thirty-Millisecond Advantage, "The New York Times”, 23.07.2009

Figure 4: Broadcom’s Performance on the 15th July 2009, FAZ, 06.08.2009

List of Tables

Table 1: Fidessa Fragmentation Index, report for week ending 27th July 2012, Fidessa Group plc

List of Abbreviations

illustration not visible in this excerpt

1. Introduction

The history of equity trading began hundreds of years ago, when the first companies needed money for their projects and asked private investors instead of banks for equity. Later, in 1969, when the first electronic stock trading was introduced in the U.S., namely Instinet or Institutional Networks, trading became electronic. Shortly thereafter, in 1971, the first fully automated exchange, the so called NASDAQ (National Association of Securities Dealers Automated Quotations), was created. The NYSE, which was founded as a physical, order-driven, auction-based market, entered the electronic sphere in the early 1970s. In 1977 the first electronic trading montage screen, showing quotes on NYSE stocks, went live through Instinet’s efforts. Another pioneering platform was Investment Technology Group’s (ITG) POSIT (Portfolio System for Institutional Trading), which was introduced in the late 1970s. POSIT was already crossing block trades electronically away from the exchanges on a scheduled basis (Domowitz et al, 2008: 1).

From 1980 on brokers started to use their own electronic proprietary trading systems to cross trades for clients. Further pioneering move was the introduction of the first after­hours crossing platform in 1986 by Instinet. In Europe it was the Paris Bourse (today part of NYSE Euronext), which developed a leading edge electronic trading platform for the French stock market as early as 1989. Although electronic trading developed during the late 1980s and early 1990s, it was only in the late 1990s and beginning of the 21st century that technology, communications and networking reached a state that ATSs became useable (ibid).

Although the concept of hiding is not a new concept, as it has been around for several years, in the form of hidden orders, the dark pool phenomenon and crossing network structure came into focus only in 2002, when the INET, a product of a merger between the Island ECN and Instinet, announced that it would stop displaying order book limit prices to avoid connecting to the relatively slow Intermarket Trading System (ITS). With this action INET effectively "went dark”, as limit orders were no longer visible to market participants. A formal dark pool platform, Instinet CBX, followed in 2003 (Domowitz et al, 2008: 1).

While in 2003, 7 crossing networks has been established as providers of nondisplayed liquidity, only 5 years later their number had surpassed 40. Although estimations about the current dark volumes are not easy, as these volumes appear mixed with all OTC trades, it can be assumed that approximately 15 % of all traded volume in the U.S. and about 10 % of all European trades are executed in the dark. With this fast growth, in only one decade, electronic trading and especially dark trading became a very important market mechanism. This becomes even more apparent when looking at the projections which suggest that at least half of European and U.S. markets will trade in the dark within the next 5 years (Grant, 16.12.2009).

This paper is supposed to give the reader a better understanding of equity trading in general, with a focus on new trading phenomenon’s, namely dark pools and flash trading. It provides a definition of equity trading in general and dark pools in special in chapter 2 where these terms as well as other trading related terms are explained and an overview of the trading framework is given. Moreover the most important types of orders and order parameters are explained to the reader.

In order to better understand the concept of the new mechanisms, chapter 3 gives an overview of the different types of market structures and explains the importance of market liquidity. The chapter is concluded with a differentiation between liquidity suppliers and liquidity demanders.

In chapter 4 the pricing in the dark pool sector is explained. The importance of price discovery and price derivation is highlighted.

Chapter 5 deals with the important topic of regulation and control. Due to the multiplicity of different national regulations, the paper focuses only on the European regulatory framework, which is MiFID.

Finally in chapter 6 the structure of dark pools is explained. All different sources of dark liquidity are highlighted and a market overview of the dark sector is given. The chapter is concluded with a regard on the dark sector evolution. Chapter 7 continues with the dark pool sector and focuses on the different trading strategies in this sector. In addition the technological aspects of dark trading are explained in non-technical terms.

Chapter 8 will summarize all the important information about dark pools and flash trading.

2. Equity Trading and its new phenomenons - Definitions and Characteristics

In the following chapter a basis about equity trading and dark pools should be given to the reader. The most important terms will be explained and an overview of the trading phases and the different types of orders and order parameters will be given.

2.1 What is Equity Trading?

In general equity trading can be described as the buying and selling of securities, which can take place on a regulated market, for instance at one of the major stock exchanges like the New York Stock Exchange (NYSE) or the London Stock Exchange (LSE), or off­exchange, bilaterally, on the so-called Over-The-Counter (OTC) markets.

Equity trading can be performed by the owner of the shares directly or by an agent authorized to buy and sell on behalf of the share's owner. Proprietary trading or principal trading is buying and selling for the trader's own profit or loss. In this case, the principal is the owner of the shares. While agency trading is buying and selling by an agent, usually a stock broker, on behalf of a client. Agents are paid a commission to the broker for performing the trade and for supplying the investors with research on shares.

Stock exchanges and brokers have so called market makers who help to limit price variations (volatility) by buying and selling a particular company's shares on their own behalf or on behalf of others clients. They are called market makers as they influence with their trades, usually of huge blocks of shares, the price of a share. They buy certain shares because they assume that according to researches, news flows or the market situation, these shares give a very interesting buy opportunity and so they can sell them to investors. The blocks of shares are only kept for a short period of time, usually of maximum an hour (wikinvest).

2.2 Dark Pools - Definition

Having define equity trading, and before getting on a more detailed discussion on the subject, a common definition of the term dark pool should be given to the reader. Dark pools can be defined as a type of alternative trading systems (ATS) that do not display prices to the public, unlike on exchanges and other platforms with a "public order book”, and which are often owned by big banks like Goldman Sachs and Credit Suisse (SEC Fact Sheet, 21.10.2009).

Another definition is given by Erik Banks, who defines a dark pool as "a venue or mechanism containing anonymous, non-displayed trading liquidity that is available for execution" (Erik Banks, 2010: 3).

He defines further that anonymous, non-displayed trading liquidity means order flow that is not visible in public order books, like those operated by exchanges, and that this fact leads to the name "dark" liquidity. Furthermore a venue is any electronic platform and a mechanism is any structure within an exchange or any participant in the market that offers non-displayed liquidity. Execution is, according to Erik Banks, the capability to trade an asset through the submission of an order (ibid).

Thus it can be summarized that a dark pool is an accumulation of orders to buy or sell stocks (or other assets), but whose existence is not publicly known or advertised. According to his definition a dark pool resembles a traditional visible market in terms of structure, function and executing according to market rules, but differs as it indicates not the market depth.

2.3 Dark Pools - Rationale

Having said this in the following a look should be taken on why dark liquidity exists and why the number of dark pools and other mechanisms is growing in recent years.

In general new financial products or markets are always introduced in order to bring market participants advantages, like costs savings, higher returns or faster execution times. In the case of dark pools the primary drivers are confidentiality, reduced market impact, cost savings and price improvements. To explain these drivers a simple example can be used: If, for instance, an investor wants to buy a large number of a certain stock, he will try to do so by being as quietly and confidentially as possible, as if his intention becomes public there is a risk that other investors might try to jump ahead of the investor to buy the same stock. The price of the stock would increase in such a case. This would create an unfavorable price movement for the investor. Of course, this would not happen if the investor would try to buy only a small number of the stock, as a small order would not generate the same interest in the market. Thus only large trades, so called block trades, which are defined as those in excess of several thousand shares per trade, are central to dark liquidity (Banks, 2010: 5).

Another driver for trading in the dark is cost saving. In general it can be assumed, that all electronic trading generates cost savings, as all execution that is done off-exchange avoids the payment of exchange fees. In addition there is the possibility of price improvements, as active sell-side and sophisticated buy-side institutions use advanced technologies and analytics for different strategies, like high frequency trading or algorithmic trading. These strategies are designed to take advantage of electronic trading and to increase short-term returns. Venues that are able to take or provide liquidity away from conventional exchanges are important in this process and can attract buy- and sell-side investors. It can be considered that any mechanism or venue that brings together sellers and buyers in a confidential manner, reduces market impact, generates fees savings and creates the possibility for price improvements, will succeed. In the case of dark pools all four advantages are given, which helps to explain why their market share has increased in recent years (ibid).

Indeed various catalysts have led to the development of new venues in the last years. First of all technological innovation plays the most important role in the evolution of the off-exchange sector. Without the development of communications networks and the processing of speed and power, the creation of efficient and reliable platforms and sophisticated routers and algorithms, as well as rapid pricing and matching routines would have not been possible. Although a closer look at the technical aspects will follow later in the paper, a correlation of the development of new venues with the rise of increasingly sophisticated technologies is obvious.

Further to the technological innovation regulatory changes have been fundamental to the development of the dark sector. Those changes came in various forms and across various jurisdictions, but the most important to mention include the Regulation on Order Handling Rules (Regulation OHR) in the U.S., the Regulation Alternative Trading Systems (Regulation ATS) in the U.S. and the Markets in Financial Instruments Directive (MiFID) in Europe. Another catalyst for the growth of the off-exchange market was the so-called decimalization, so the moving of the minimum quoted price to 0.01 from some larger amount, which had as consequence that smaller price increments led to lower spreads, which in consequence led to lower profit opportunities and this in return led to the reduced willingness to risk capital. Indeed the consequences of decimalization were quite dramatically as it has caused average trade size executed on exchanges to decrease by 70 % in the U.S. and by 50 % in Europe. As it became more difficult to cross block trades without being noticed, more business was routed to the non-visible markets (ibid: 15).

Further to the catalysts described above one other reason for the increasing importance of off-exchange trading rest to mention which is capital accumulation and mobility. Actually capital is a raw material of every economy, and includes the equity and debt obligations firms raise to fund their operations. Therefore as the global demand for goods and services has risen in the past, the demand for capital financing has risen as well. As industrialized and emerging nations continue to build on their economic bases it is quite likely that the amount of outstanding capital will continue to rise. And as capital is mainly supplied by institutional investors, like mutual funds, hedge funds, insurance companies, and other asset managers, this has radically reshaped capital mobilization, allocation and trading. The most important of these investors, namely hedge fund investors, are especially interested in sophisticated, high-volume trading strategies, like high frequency trading or program trading, and these are critically important to the electronic markets and dark trading. Of course, capital needs to be centralized in order to give interested parties the possibility to trade it in an efficient and effective way. Therefore both exchanges as well as over-the-counter venues are crucial for capital trading. But as the supply of capital has increased enormous in the last decades, any single venue is able to handle the large amount of outstanding capital on its own. Therefore there is a need for a multiplicity of venues, and this made it possible for new venues, including dark venues, to develop and expand. Further to the accumulation of capital also the ability to move capital quickly across markets had helped in the development of new platforms. Both traders and investors ask for rapid possibilities to buy, sell, hedge or transfer capital, and as this demand will probably continue, one can assume that venues, which offer such services, will continue to benefit (ibid).

2.4 The Trading Framework

Investors trade for different purposes and over different time horizons. Therefore their demand for and supply of liquidity are very different. But although different trading strategies exist, the process of trading typically follows a logical sequence that represents an entire cycle. The key elements of any trading framework include three phases, which will be described in the following. These phases comprise together the "lifecycle” of a trade. Some of the steps described in the following, like the entry of a trade are very short, lasting no more than a few seconds or even milliseconds, while other steps such as the execution can last for milliseconds, seconds, minutes, hours or even days, depending on the specifics of a transaction. And still others such as clearing and settlement can take up to several days to conclude.

2.4.1 Pre-trade phase

The first step in the pre-trade phase is the analysis of potential trading opportunities. This analysis can occur through fundamental analysis, technical analysis, benchmarking, or other models. The output of this analysis process is the identification of a specific security that should be bought or sold. The second step of the pre-trade phase includes the analysis of the opportunity in relation to the current trading portfolio. As the addition of a single security can radically change the characteristics of a portfolio, it is critical to understand the effects before the execution of a trade. The last step in the pre-trade phase is the identification of specific transaction strategies. The trader must decide which venues should be triggered, when the trade should be executed, what prices are acceptable and so forth. Further the trader must also define a strategy for the case that an order cannot be filled immediately. He must decide whether the order should be canceled or rerouted in such a case. Another decision regards whether to display an order or to leave it in the dark. Each one of these points contributes to the development of a trading strategy (Bhowmik, 2012).

2.4.2 Trade phase

The first step in executing a strategy that has been developed by a trader during the first three stages of the trading framework is the order entry. Order entry is a positive action that must be taken by a trader. The trader can chose between various order entry mechanisms, like verbal/telephonic communication to a broker, or via direct input into some type of order application, like electronic communication networks (ECN) entry via interface (e.g. application program interface), broker algorithm via interface, and so forth. All these mechanisms are valid but whereas voice via broker is less common in today’s marketplace, algorithms are increasingly common. In fact investors send less order flow to brokers because they can reduce costs and minimize the potential for information leakage by using flexible technologies for order entry by themselves. The order entry stage gives the trader the opportunity to define precisely what should be done with regard to the purchase or sale of securities. Numerous types of orders can be selected with parameters that define price, time, venue and so forth. While certain clients rely on voice-only orders which are typically taped in order to ensure accurate transcription of details and which may be followed by a confirming message from the broker, many other clients prefer to directly input their own orders electronically by supplying relevant details in the entry fields provided by the application. The more complex the order details, the greater the flexibility needed within the application. Of course, if a large number of orders should be entered, as in the case of high frequency trading for instance, the process must be automated. In such case batches of orders may be generated automatically by the trader’s model and submitted either directly into the market or via a router to one or several brokers or venues (ibid).

The second step in the trade phase is order routing. Even though an order has been entered into an order management system or conveyed verbally to a broker it must still find its way to a marketplace. This step is heavily dependent on technology. The only practical way for tens of thousands of individual orders to enter into a marketplace at any one time is by placing them into an electronic line and leading them to a destination.

This line is a network that connects the client terminals with the servers that feed the engines of a particular exchange. The faster this network connection and the more robust the server architecture and the closer to the physical proximity of an exchange, the quicker is the delivery and the execution. The client or the broker gives the instructions where the order should go. While some of the routing is very standardized and can be viewed as a simple set of instructions stating that the order should be moved from the order entry terminal and delivered to a particular destination, also more complex routers exist. These routers are known as smart order routers (SOR) or routing algorithms and they ensure that the order is treated in a very specific way if it is not filled immediately. In such cases the order can be moved sequentially through other random or defined dark of light venues in search of best execution opportunities (ibid).

The third step in the trade phase is finally the execution. This step is in fact completely technology-driven and based on pricing and matching routines that run in an automated way. As the loss of even some milliseconds in the process can be harmful to a trader’s position, the most important factor in order execution is time. Therefore advanced technologies are employed, in order to route orders into venues, and to flow them through pricing and matching engines, which match and execute, or reroute orders, with a minimum of latency (ibid).

2.4.3 Post-trade phase

After an order has been executed, details must be reported to all stakeholders. This comprises the first step of the post-trade phase. Reports are sent to the trader, but also to regulatory bodies and to the market at large. But of course the levels of detail which are reported are not the same to all parties and also the time horizons over which reporting must occur are different. While deal confirmations are send out immediately, via an order management system or through electronic messaging, the reporting to the public or to regulators may occur at the end of the trading day or at some later date. Same as the execution of an order, also the reporting is largely driven by technology (ibid).

The last step in the trading lifecycle is the dual process of clearing and settlement. Once the trade has been executed and reported, it must pass through the clearing process, which is usually managed by an independent clearinghouse or the clearing department of the venue. Clearing includes the confirmation of all relevant details of the executed trade, including the counterparties, the price, the quantity and so forth. If any of these details do not match then the process is diverted to a resolution department where further investigation must be undertaken. After the clearing process is concluded the final stage can begin, namely the settlement. This step includes the delivery of cash for shares and vice versa, between the two parties to the trade through the clearing and settlement agent. This step is extremely automated and usually involves electronic debits and credits to the cash and securities accounts of buyers and sellers. Once this stage is completed the trade lifecycle is concluded (ibid).

2.5 Orders

In trading instructions are given how and when to buy and sell securities. These fundamental instructions for the transfer of liquidity are orders, and they are essential in creating and transferring liquidity. In the following the most common types of orders will be explained. Within the broad classes of orders certain additional parameters may be attached, to specify for example the speed of execution, to achieve price improvements, to limit the risk, or to determine the method of display.

Before describing in the following the major types of orders, some important terminologies in relation to order types should be defined:

- Bid: the highest price at which one party is willing to buy a security (representing the demand side of the transaction) (Deutsche Börse Glossary)
- Offer: the price at which one party will sell a security (Banks, 2010: 34)
- Best bid: the best buy offer. The highest price at which one party is willing to buy (www.centralact.com)
- Best offer: the best sell offer. The lowest price at which one party is willing to sell (ibid)
- Inside spread: the difference between the best bid and best offer (Banks, 2010: 34)

2.5.1 Market Orders

Market orders, alias unpriced orders, are the classic form of orders. They are not-limited buy or sell orders which should be executed as soon as possible, at the next price in the market.

Within the class of market orders several subtypes exist, including stop orders, trailing stop orders, market-to-limit orders, market-if-touched orders, market-on-close orders, market-on-open orders, uptick/downtick orders and sweep-to-fill orders.

The most used of these order types are stop orders, trailing stop orders, market-to-limit orders and sweep-to-fill orders (Deutsche Börse Group, 2009:11).

Stop orders are similar to limit orders: securities are sold or bought if a trigger price is attained. In the case of insufficient volume a fill cannot be guaranteed even if the stop is triggered and the order converts into a market order.

Trailing stop orders are similar to stop orders except that a trailing amount is attached that moves with the market price.

Market-to-limit orders are not limited buy or sell orders which should be executed at an action price or in the continuous trading at the best limit price in the order book. A Market-to-limit order is only accepted if there are only limit orders at the opposite site in the order book. If only a part execution of a market-to-limit order is possible, the rest of the order will be placed in the order book with the limit of the part execution (ibid).

Sweep-to-fill orders are orders that should be executed as fast as possible at the best available price, regardless of venue. That means that these orders are submitted to the first venue with the best price and are filled to the extent possible. Then the remaining portion sweeps to the next venue with the next best price, and so forth until the order is completed. Typically with each sweep the price becomes less favorable (Banks, 2010: 36).

2.5.2 Limit Orders

Further to market orders a limit can be added to an order. Limit orders are buy or sell orders with a price limit. They should be executed at a certain price or better. If the price limit is not reached, limit orders are not executed and depending on the investors instruction these orders are than deleted or the investor gives also a time limit, which indicates how long a limit order is valid. An unfilled limit order is placed in the exchange’s limit order book for future execution. While in the process of being filled, a limit order is considered to be a working order.

Subtypes of limit orders are: stop limit orders, trailing stop limit orders, limit-if-touched orders, limit-on-close orders, limit-on-open orders, discretionary orders and intermarket sweep orders (Deutsche Börse Group, 2009: 11).

2.5.3 Peg Orders

Peg orders are based on a price that is pegged to a recognized base reference, such as the NBBO or the EBBO. Within the class of peg orders the following subtypes can be considered:

- Primary peg orders: are peg orders that peg to the same side of the base reference
- Market peg orders: are peg orders that peg to the opposite side of the base reference
- Midpoint peg orders: with midpoint orders the execution of an order takes place at the precise midpoint course between ask and bid price of the base reference. Midpoint orders interact only with other midpoint orders and not with the other orders in the order book.
- Alternative midpoint peg orders: are peg orders that peg to the "less aggressive” side of the midpoint (Interaktivbrokers.com).

2.5.4 Hybrid and Complex Orders

Further to the order types noted above, many other orders can be created, by combining different order types with each other. An example is for instance the cross order, which is an order to buy and an order to sell the same stock at a specific price. Sometimes broker receive an order from one customer to buy and from another customer an order to sell the same security; this is then a cross order, which is not fictitious.

The most complicated order types are called algorithms, which contain many different parameters that define precisely how an order is to be filled (Banks, 2010: 40).

2.5.5 Order Parameters

Many of the orders from the four main order classes’ described above can be further defined through additional parameters, namely display parameters, quantity parameters and time in force parameters.

2.5.5.1 Display Parameters

Display parameters need to be added specifically, with a "do not display” instruction. If such an instruction is not given, an order is assumed to be visible to the public in most marketplaces. A "do not display” inscription can be applied to all or only part of the order. If it applies to all of the order, the order is a hidden one, if it applies only to a part of the order; the order is an iceberg order. It should be noted at this at this point that both hidden and iceberg orders are critical to dark liquidity formation. Therefore a closer look at these two order types will be taken in the following.

2.5.5.1.1 Hidden Orders

Hidden orders are non-visible limit orders, which are embedded in a venue’s dark book. Generally, they are excluded from the MiFID pre-trade transparency regulations, because of their huge volume compared with normal market size. In order to comply with MiFID requirements for orders which are large in scale the minimum size for a hidden order is specified in Table 2 in Annex II of the MiFID Implementing Regulation 1287/2006, titled "Orders large in scale compared with normal market size” (CESR, 2010: 18).

Hidden orders rest hidden, even if the rest volume of a part execution is smaller than the requested minimum size. Generally hidden orders are executed in the order book as limit orders, which mean that the execution follows the price-/time priority. However if there are hidden and visible orders at the same price, visible orders are executed before hidden orders (Deutsche Börse Group, 2009: 15).

2.5.5.1.2 Iceberg Orders

Iceberg orders (also known as reserve orders) are orders which make it possible to an investor to place orders with huge volume into the order book without making the whole volume visible to the public. Iceberg orders are limit orders with a defined total volume as well as a defined visible part of the order, the so called peak-volume. Both, the total volume as well as the peak-volume, have to match a so called round lot, which means that the volume of an iceberg order has to be a normal unit of trading for a security, which are generally 100 shares of a stock.

In the continuous trading when the first peak has been fully executed, another peak is automatically displayed in the order book. The hidden part of the order is then reduced by the corresponding number of shares. When an iceberg order’s displayed part is fully executed, and the next peak converts from hidden to displayed status in the order book, the newly displayed peak also receives a new time stamp which determines its time priority (Banks, 2010: 42).

For instance, if the total volume of an iceberg order is 100,000 shares and the peak volume is defined as 10,000 shares, the market would see only 10,000 shares on the ask or bid side of the order book, depending whether the order would be a buy or a sell order. When the display quantity of 10,000 shares would be executed, the next peak volume would be automatically visible to the market, and wait for its next opportunity to get a fill. The trading system automatically displays new peaks after additional executions, until the final peak is displayed or the order is cancelled. As iceberg orders are not specially marked in the order book, it is not visible from the outside whether an order has a rest volume or not. Only after the full execution the total volume is visible. But the immediate automatic display of a new peak after the currently displayed peak is executed creates a distinct pattern in the order book updates that is observable to any trader who monitors the order book closely. A trader who detects an iceberg order cannot determine the iceberg order’s size although its peak size may provide a signal. But as peaks are executed and new ones are displayed, the trader can form more precise forecasts of the order’s total size. So rather than disclose the information about the order size right away, iceberg orders sequentially reveal more information as peaks are executed. Iceberg orders are therefore often justified and promoted as an order type that facilitates the execution of large orders with minimal price impact, a property that should appeal to large liquidity traders. In the order book iceberg orders, including their hidden volume, have a higher priority than hidden orders (Deutsche Börse Group, 2009: 15)

Although both hidden and iceberg orders assume lower priority than visible orders that have the same price level, these order types are, according to recent empirical evidence about traders' order submission strategies on electronic limit order books, more and more used. The growing importance of these special types of orders can be seen in the fact that they make up recently around 44 percent of all Euronext volume (Buti et al, 2008: 3).

2.5.5.2 Quantity Parameters

In addition to display parameters, many of the orders described above may also contain quantity parameters. These parameters specify the action to be taken if the full order cannot be executed instantly. The most common quantity parameters are:

- Fill-or-kill (FOK): with this designation the order should be immediately and entirely executed. If this is not possible the order should be cancelled (Private Trader Club; Trader Lexikon).
- All-or-none (AON): with this designation the order should be immediately and entirely executed. If the execution of the entire amount is not possible the order remains in force pending future execution unless it is specifically cancelled. An all-or-none order is similar to a fill-or-kill order, except that it stays in force until it is withdrawn (Interactivebrokers.de: 2012)
- Immediate-or-cancel (IOC): with this designation the order should be immediately and entirely executed. Any unfilled portion of the order is to be cancelled shortly after the order has been submitted (deifin.de: 2012).
- Minimum acceptable quantity (MAQ): with this designation at least a minimum amount out of the full order size must be filled. When an MAQ order is partially filled and the residual order size is lower than the MAQ, the MAQ will be reset to the residual order size (BATS Europe, Market Guide, 2012: 2).

2.5.5.3 Time in force Parameters

Further to display and quantity parameters some orders can also contain time in force parameters, which define certain duration limitations or order extensions. One of the most common of these time in force order limitations is the “Good-for-day” instruction, which indicates that the order is only valid for the current trading day, so until exchange closing. Other common time in force parameters are “Good till date”, “Good till cancelled” and “Good after order” (Banks, 2010: 43).

The list of order types and order parameters described above is not complete. In fact many other subtypes of orders can be created. The main point to highlight at this point is that orders are key for market activities.

3. Different types of Market Structures and Market Liquidity

Further to the different order types described before also different types of markets exist. These apply to all kind of securities and assets, but this paper will only concentrate on market structures which apply to stocks. The most common differentiations in market structures for stocks are: physical and electronic markets, continuous markets, quote- driven and order-driven markets as well as displayed and non-displayed markets.

The success of all of these markets relates on market liquidity and on the way in which that liquidity is accessed (dark or light). Therefore in the following also some aspects of liquidity should be explained to the reader.

3.1 Physical and electronic markets

Although the operation of a physical and an electronic market may be identical, the interactions of participants are quite different. While in an physical market traders, brokers, specialists and other exchange personnel negotiate on the physical trading floor in order to buy and sell stocks, in an electronic market these parties are in front of their monitors, usually in different places, and negotiate via advanced technologies.

Both structures have their advantages: while physical trading offers speed and immediacy that is simply not available via an electronic marketplace, which inevitably has a degree of latency, electronic trading gives access to participants from various locations, removes the human element from the trading process, and let therefore less room for errors. Finally electronic trading also offers possibilities for participation to a much larger number of buyers and sellers (Banks, 2010: 48).

The advantages of electronic markets explain why there are more and more of these.

[...]