After a brief literature overview, the paper will deal with the transformations in the landscape of trading venues by discussing the impact of ECNs on competition and market transparency in the US and in Europe. It will focus on the technological advances in execution concepts – their functionalities and capabilities (regarding improved transparency, order routing and best execution) will be described in section III. Section IV will introduce the upcoming regulatory regimes in the US and Europe, RegNMS and MiFID respectively. Both new legislations try to catch up with recent years' technological advances and the consequential changes in the landscape of trading venues by creating a level playing field for the different types of trading venues that have emerged. The requirements of both legislations concerning market trans­parency, order routing and best execution will be highlighted. Section V details the impact of the MiFID on MTFs and develops scenarios on their prospects based on recent regulatory discussions in Europe. Section VI concludes by providing a summary of the findings.

II Literature overview

Competition for order flow between exchanges and ECNs and within these groups of trading venues is discussed intensively in the academic literature. Barclay, Hendershott & McCor­mick (2003) as well as Bessembinder (2003) examine the competition among different trad­ing venues in the US. The former show that ECNs attract more informed orders than NASDAQ market makers. The later examines quotations, order routing, and trade execution costs for seven markets that compete for orders in large-capitalization NYSE-listed stocks and finds evidence for substantial quote-based competition. A survey of competition among trading venues in Europe has been conducted by Gresse (2006) by investigating the relationship of trading activity on the crossing network POSIT and the liquidity of LSE's quote-driven SEAQ. Conrad, Johnson & Wahal (2003) investigate the execution costs of trades sent to traditional and alternative trading systems. They conclude "that orders sent to tradi­tional brokers have higher execution costs than those executed by alternative trading sys­tems" ⁵.Korhonen (2001) presents general policy topics raised by ECNs and provides an analysis of selected issues related to the "Investment Services Directive (ISD)"⁶, the prede­cessor directive of the MiFID.

Best execution is a key component of both the MiFID and RegNMS. Already Macey & O'Hara (1996) mentioned the dual problem of defining and enforcing best execution. Amongst others one of their recommendations how to deal best with the issue of best execu­tion is increasing competition among trading venues. McCleskey (2004) argues that best execution should be regulated, as investors are faced with an information asymmetry, which prevents them from assessing the quality of their executions themselves. The data they would need to assess whether an execution was good or poor is not available for them. Schwartz & Wood (2003) on the other hand offer several caveats against more regulation.

The impact of market transparency on market quality is investigated in several papers. Bes­sembinder, Maxwell & Venkataraman (2005) assess the impact of transaction reporting on execution costs for corporate bonds. The results indicate a significant reduction of execution costs by the introduction of transaction reporting. Aygouleas & Devgiannakis (2005) address the impact of pre-trade transparency on market volume by using trading volume data before and after the introduction of a central order book at the LSE. The authors conclude that in­creased transparency does not lead to an increase in trading activity and transfer these find­ings to the transparency regime with the upcoming application of the MiFID.

III Development and status quo of the ECN/MTF landscape

In the last decade, the financial trading industry faced a dramatic revolution in the way trad­ing is conducted on international markets. More and more stages of the trading process have been automated by substituting human activities with electronic systems. Starting with quote machines, technology bit by bit conquered other stages of the trading process and the trad­ing value chain. In the late 1990s, electronic trading systems came up in Europe implement­ing new market mechanisms like the open order book model. As a key step forward, these market models extended the technological reach beyond pure order routing via electronic means into fully electronic matching and price determination.

Since the adoption of Regulation ATS by the SEC in 1998⁷, ECNs are required to either reg­ister themselves as a broker-dealer or to become a self-regulated registered securities ex­change. Around the year 2000, a significant number of ECNs were competing in the market for markets in the US. But in the following years, a first wave of consolidation among ECNs with a peak in 2002 decreased their number: Archipelago acquired REDIbook in 2001 and the GlobeNet ECN in 2002, Instinet and Island ECN merged to INET ATS in 2002. In Octo­ber 2001, Archipelago was the first ECN to become a registered securities exchange. The impact of the change from broker-dealer status to exchange status on execution quality and market share has been examined by Nguyen, Van Ness & Van Ness (2005). With the acqui­sition of Brut LLC by NASDAQ in 2004, a second wave of consolidation among ECNs and exchanges started that culminated in the merger of NYSE and Archipelago as well as the acquisition of INET ATS by NASDAQ, both announced in April 2005 and approved^8,9 in De­cember 2005. ECNs that came out as winners of the consolidation process have gained noteworthy market share, e.g. Archipelago's total trading market share in January 2006 in NYSE-listed equities accounted for 5.0%, in AMEX-listed equities for 31.0% and in NASDAQ-listed equities for 22.5%¹⁰. ECNs have been successful in the US as they intro­duced new trading concepts, e.g. the concept of order-driven markets with the open order book approach. Open order books increased market transparency as they show competing and firm bids/offers that are immediately accessible for market participants. Hendershott & Jones (2005) confirmed that "... more transparency is associated with better market quality"¹¹and market quality based on increased transparency has been a crucial competitive advan­tage for ECNs in the US.

The attempt to copycat the success of ECNs on European markets failed. This can be ex­plained by three main reasons - these are market models, liquidity and the regulatory envi­ronment:

″ In Europe, ECNs did not benefit from the transparency inherent to their market mod­els as the trading concept of the ECNs' open order books was not new for markets in Europe. Leading European exchanges already had modernised their market models/systems and turned from pure floor based trading mechanisms or quote-driven markets to electronic open order book trading systems in the late 1990's¹². From a functional, non-institutional view, one could argue that the European order driven markets/exchanges already were ECNs themselves.

Another reason for the failed attempt to conquer European markets by ECNs was the different structure of liquidity in Europe compared to the US. In the US, liquidity is fragmented per security across a number of markets¹³. On European markets, trading is fragmented among more than 20 stock exchanges. Though eyed per traded in­strument, liquidity in European equity trading in nearly all instruments is highly con­centrated on just one market – typically the predominant market of the instrument's country of origin, i.e. German equities are most actively traded at Deutsche Börse, French equities at Euronext, and British equities at the London Stock Exchange¹⁴. As "networks are by their nature self-rein forcing'¹⁵and "liquidity is also self-rein forcing'¹⁶, those markets that are the predominant market for a specific share realize huge net­work externalities. Trading shares on a market other than the predominant market is less attractive due to higher implicit trading costs. Accordingly, it is very hard for new entrants to attract liquidity away from the established markets in Europe. Network ex­ternalities among European exchanges have been investigated, e.g. by Hasan & Schmiedel (2003): "The evidence shows that adopting a network strategy is signifi­cantly associated with higher liquidity, growth and efficiency...'¹⁷.

Up to now, the regulatory environment of securities trading in Europe is defined by the "Investment Services Directive (ISD)" established in 1993. The ISD defines framework legislation; its implementation in the EU member states differs signifi­cantly. This holds especially for the possibility to execute orders outside Regulated Markets: In some EU member states, e.g. France¹⁸, there is a "concentration rule" which forces transactions up to a certain transaction size to be conducted on a Regu­lated Market. In other member states, e.g. Germany¹⁹, there is a "default rule", which requires banks/brokers to execute orders on an exchange unless an investor opts-out on a per order basis. Yet another group of member states, e.g. the UK, has neither a concentration nor a default rule, i.e. executing orders outside a Regulated Market or internalising them is generally possible. In countries that have a concentration rule or default rule, it is difficult for ECNs to make a stand against the incumbent exchanges. Furthermore, the different national implementations of the ISD complicate the situa­tion for ECNs targeting a pan-European approach.

Some entrants to the European market for markets recognized the regulatory obstacles due to concentration and default rules and tried to enter the market by setting up Regulated Mar­kets themselves. But, most of these attempts, e.g. Jiway, NASDAQ Europe or NASDAQ Germany, failed²⁰. They were not able to break the existing network externalities of liquidity in European markets. Only some special types of ECNs had success in Europe, as they fo­cused on niches and special needs of investors. Crossing systems and systems for block trading²¹, such as POSIT²² or Liquidnet²³, or electronic systems for bond trading, such as EuroMTS²⁴ and Eurex Bonds²⁵, achieved to gain some market share. With the new regula­tory regime of MiFID applicable in Europe in 2007, some of the aforementioned obstacles for ECNs' success in Europe may cease to exist which will be discussed in section V of this pa­per.

III.2 Recent technological improvements and competitive dynamics

Intermediation in trading processes changed significantly: Increased communication capaci­ties and increased computing power enabled brokers in the late 1990s to implement new business models for retail customers (online brokerage) where orders are not touched by the broker anymore but are forwarded directly to the markets. This model of a virtual "Direct Market Access" (DMA) is also provided to professional customers. Brokers (the Sell Side) offer the execution concept DMA to both customer groups at a lower fee than traditional ser­vices as there is no need for cost-intensive trader involvement at their desks. Increased cost consciousness²⁶ and technological capabilities on the Buy Side lead to a wide usage of DMA. Simultaneously, the Buy Side is endowed with the ability to implement its own execution concepts on top of the virtual market access provided by DMA. Execution concepts that use DMA as a technological basis are Algorithmic Trading, Smart Order Routing and Liquidity Aggregation:

Algorithmic Trading emulates a broker's core competence of slicing a big order into a multi­plicity of smaller orders and of timing these orders to minimise market impact via electronic means. Based on mathematical models and considering historical and real-time market data, algorithms determine ex ante or continuously the optimum size of the (next) slice and its time of submission to the market. A variety of principles²⁷ are used for these algorithms that aim at reaching or beating an implicit or explicit benchmark: e.g. a volume weighted average price (VWAP) algorithm targets at slicing and timing orders in a way that the resulting VWAP of its own transactions is close to or better than the VWAP of all transactions of the respective se­curity throughout the trading day or during a specified period of time. Up to now, the concept of Algorithmic Trading is primarily used to work "low-touch orders", i.e. plain-vanilla orders in liquid stocks, to unburden human traders and enable them to concentrate on "high-touch orders", i.e. orders in less liquid stocks or high volume orders that need cautious handling in order to minimise market impact. Increasing sophistication of the Algorithmic Trading solu­tions is likely to shift these boundaries into more complicated transactions in the near future.

Smart Order Routing concepts – also known as Best Execution Engines – imply forwarding orders to the "best" out of a set of alternative venues while taking into account the different attributes of each venue. What is "best" can be evaluated considering different dimensions – either specified by the customer or by the regulatory regime – e.g. price, liquidity, costs, speed and likelihood of execution or any combination of these dimensions. In order to be able to assess which venue is "best", a prerequisite is market transparency and immediate access to visible and firm bids and offers across the different trading venues. The open order book approach, e.g. of ECNs, endows Smart Order Routing systems with this kind of market transparency and enables them to perform order routing exploiting the increased connectivity of electronic trading systems, e.g. based on the FIX protocol²⁸.

Third-party software providers, such as Belzberg²⁹, Firefly Capital³⁰ or Lava Trading³¹, offer DMA in combination with on-top services as e.g. Algorithmic Trading, Smart Order Routing or Liquidity Aggregation. With the Liquidity Aggregation concept they offer their customers a virtually consolidated order book across different trading venues. More sophisticated ver­sions include functionality for "liquidity discovery", which means sending e.g. an immediate­or-cancel (IOC) limit order to different venues in order to unveil hidden liquidity. Hidden liquidity refers to volume that is not displayed in the limit order book, e.g. the reservation size of an iceberg order.

Algorithmic Trading as a concept focuses on the splitting and timing of orders on a single market with the goal to minimise market impact. Smart Order Routing and Liquidity Aggrega­tion are concepts that – per definition – are useful for traders in a fragmented market envi­ronment. As noted above, this fragmentation of liquidity per share on equity markets is pri­marily a US phenomenon, therefore Smart Order Routing and Liquidity Aggregation tech­nologies are primarily provided by US service providers and primarily used in the US market as of today whereas in Europe, liquidity concentration per share did not lead to a wide usage of these concepts yet³².

Figure 1: Order flow allocation from Buy Side Trading Desks. Source: TabbGroup (2005)

With the advent of these new execution concepts, the responsibility of order execution quality is shifted from brokers to the Buy Side trading desks and the brokers' traditional role as in­termediaries changes to become primarily technology providers: Sophisticated Buy Side trading desks compete with traditional brokers and Independent Software Vendors (ISV) compete with brokers in offering advanced trading strategies. Figure 1 gives an impression of the relevance of this change, as it exhibits data on the shift of the order flow allocation from the phone to new concepts like DMA and Algorithms.

IV New regulatory regimes on both sides of the Atlantic

The aforementioned new execution concepts have significant effects on order handling and the structure of the order flow, as e.g. the average size of trades is shrinking at the major exchanges (see figure 2) due to increased order splitting and timing. With RegNMS in the US and MiFID in the EU coming into effect in 2006 and 2007 respectively, regulators on both sides of the Atlantic respond to these changes. Both new legislations try to catch up with re­cent years' technological advances and the consequential changes in the landscape of trad­ing venues. They intend to create a level playing field between the different types of trading venues and a harmonisation in the order execution process. In the following subsections IV.1 and IV.2, the paper outlines the two regulatory environments and their primary differences in comparison to the prevailing regulations in the US and in Europe with a focus on the key as­pects of market transparency, order routing requirements and best execution. This lays the foundation for section V that will focus on the impact of the MiFID on MTFs and the future European landscape of trading venues.

Figure 2: Average value of trades (data provided by the World Federation of Exchanges, http://www.world-exchanges.org)

IV.1 RegNMS in the US

In 1975, the "National Market Systems" (NMS) with its components "Intermarket Trading Sys­tems" (ITS), "Computer Assisted Execution System" (CAES) and "Consolidated Quotation System" (CQS) was introduced. The core component, the ITS, links all US markets trading

exchange-listed securities. Markets exclusively trading NASDAQ securities do not have to participate in the ITS. There is no comparable centralised Intermarket Trading System for NASDAQ securities yet.

The requirement for best execution in the ITS Plan³³ forces markets to route their orders for exchange-listed securities to the market which is offering the best price with receiving mar­kets generally having up to 30 seconds to respond³⁴. This is set forth in the so-called "Trade-Through Provision" in the ITS Plan. A Trade-Through happens if an order is executed on a market despite there is a better price available on another market. The existing Trade-Through rule protects markets like the NYSE: NYSE's share of trading accounts for 73.7%³⁵ of the US equity market, as it provides the most competitive quotes in its listed securities creating the National Best Bid and Offer (NBBO) more than 89%³⁶ of the time.

RegNMS was initially scheduled to come into effect in June 2006; however, on May 18^th, 2006 the SEC³⁷ extended compliance dates from June 29^th , 2006 to a series of five dates, beginning on October 16^th , 2006. RegNMS will update the existing rules that do not reflect the status of technological progress on the one hand and the requirements of (especially) Institutional Investors on the other hand. Its goal is to level the competitive playing field by specifying equal regulation for all kind of stocks as well as different kinds of markets. The scope of the regulation will be extended to NASDAQ stocks.

Based on RegNMS, trading venues will be classified in so-called "fast markets" and "slow markets". The new "Order Protection Rule"³⁸ changes regulation for Trade-Throughs, as only orders on fast markets will be protected against Trade-Throughs, i.e. an order can be exe­cuted on a "fast market" despite of a better price being available on a "slow market". The "Order Protection Rule" privileges "fast markets", which support automated and immediate execution. As floor-based markets are likely to be categorised as "slow markets", they face the risk of significant loss of order flow. Therefore, they will have to modernise their trading systems or merge with other markets providing electronic matching facilities. The merger of NYSE and Archipelago as well as the recent announcements of the Philadelphia Stock Ex­change³⁹ to close its floor and the Boston Stock Exchange⁴⁰ to become a fully electronic market can be seen as direct consequences of RegNMS. Furthermore, in early 2006 the new ECN BATS⁴¹ was launched⁴². Citigroup⁴³ announced in early 2006 to launch another new ECN as well. This shows that RegNMS not only forces "slow markets" to adapt but it also provides prospects for new ECNs to successfully enter the market after years of ECN shut­downs and consolidation.

IV.2 MiFID in Europe

The "Markets in Financial Instruments Directive", which is the follower directive of the ISD of 1993, came into effect in April 2004 and will have to be applied by Investment Firms and Regulated Markets in November 2007. The goal of the MiFID is to harmonise regulation on a European level, to increase transparency and accessibility of markets, to ensure price forma­tion and increase investor protection. A level playing field among different types of trading venues shall assure competition and foster innovation. To achieve these goals, MiFID – like RegNMS – regulates market transparency, order routing requirements and best execution. MiFID reaches further into the securities trading industry than RegNMS, as it also addresses other topics such as record-keeping, code of conduct, organisational requirements and oper­ating conditions for Investment Firms and Regulated Markets.

The regulatory process of the MiFID is based on the so called "Lamfalussy Process" that is subdivided into four successive steps named "Levels": The MiFID itself provides framework legislation ("Level 1"). Implementing measures ("Level 2") in the form of concrete parameters and further details define how the MiFID will work in practice. In "Level 3" the Committee of European Securities Regulators (CESR) supports the EU member states to implement the "Level 1" and "Level 2" results in national law and thereby fosters harmonised implementa­tion; in "Level 4" the consistent application of these laws is supervised. Currently (as of May 2006) the process is in the finalisation of "Level 2".

"Level 2" measures can be specified either in the form of directives and/or in the form of regulations. Regulations are binding in their entirety and are directly applicable in all EU member states. Directives have to be transposed into the national legal framework and thus leave some flexibility as to the form and means of implementation and adoption to existing national laws and market environments. Yet, directives bind the member states as to the re­sults to be achieved. Draft versions of both a "Level 2" directive⁴⁴ and a complementary regu­lation⁴⁵ have been released by the EU Commission on February 6^th,2006. Therefore, the following consideration of MiFID's impact on MTFs will be based on these documents. Ac­cording to the "Lamfalussy Process", the European Parliament and the European Securities Committee (ESC) will inspect the drafts for three months. After approval by the ESC, the European Parliament will have an additional month to verify whether the framework set by the "Level 1" directive has been maintained in the implementing measures. The "Level 2" documents are scheduled for final adoption by the EU Commission in summer 2006.

Trading venues are classified by the MiFID into "Regulated Markets", "Multilateral Trading Facilities (MTF)" or "Systematic Internalisers". Regulated Markets were already defined⁴⁶ in the ISD of 1993 and correspond to the existing exchanges' trading set-ups. In MiFID they are defined as "a multilateral system operated and/or managed by a market operator, which brings together or facilitates the bringing together of multiple third-party buying and selling interest in financial instruments – in the systems and in accordance with its non-discretionary rules – in a way that results in a contract, in respect of the financial instruments admitted to trading under its rules and/or systems, and which is authorised and functions regularly ..."⁴⁷.MTFs represent a new category in European securities legislation: an MTF is defined as "a multilateral system, operated by an investment firm or a market operator, which brings to­gether multiple third-party buying and selling interests in financial instruments - in the system and in accordance with non-discretionary rules - in a way that results in a contract ..."⁴⁸,which makes them an analogue to the systems known as ECNs in the US. A Systematic In­ternaliser is defined as an "investment firm which, on an organised, frequent and systematic basis, deals on own account by executing client orders outside a Regulated Market or an MTF"⁴⁹.

MTFs can either be operated by an Investment Firm or by an operator of a Regulated Market and Systematic Internalisers are per definition Investment Firms, i.e. there are two institu­tional forms that are regulated by MiFID: Regulated Markets (respectively their operators) and Investment Firms.

To sum up chapters IV.1 and IV.2, the following table gives a comprehensive overview on the commonalities and differences of RegNMS and MiFID: