Bourse Consult Reports

Bourse Consult has produced several reports over the years for their clients. Here are the non-confidential reports that have been published for our client the City of London Corporation.

JUST PUBLISHED!

The Post-Trade Infrastructure for Carbon Emissions Trading

- – - – - -

The Competitive Impact of London’s Financial Market Infrastructure

Current Issues Affecting the OTC Derivatives Market and its Importance to London

The Future of Clearing & Settlement

1. Trade and Post Trade Infrastructures in Europe

2. The ECB Steps in….

3. US Exchanges and Clearing Houses

4. European Exchanges and Clearing Houses

Remarks prepared by Hugh Simpson of Bourse Consult

For the 9^th Annual Clearing and Settlement Conference in London on 7 September 2006

One of the fundamental questions faced by European securities markets is whether competition or integration will be more effective at delivering an efficient structure.

My starting point is a bias in favour of competition. Historically, this has proved itself again and again to be the best way of stimulating innovation and driving efficiency. There need to be very good reasons not to allow competition to shape the industry.

Potential for competition

Let’s look at the different components of the transaction life-cycle – trading, clearing and settlement – to assess how open they are to competition.

What they all have in common is that they are network businesses – the more people use a particular service, the more attractive it becomes. And this tends to result in an “all or nothing” tipping effect. It is difficult to sustain a market where many competitors all enjoy equal shares. There is likely to be just one winner and all the rest are losers. The antidote to this tipping effect is interoperability between the different providers. A good example is mobile phone networks: interoperability is why there are multiple competing mobile phone companies. However, mobile phone companies grew up in a world where standards for interoperability existed from the beginning. In clearing and settlement these standards have to be reverse-engineered into businesses that already exist.

Beyond the similarity that trading, clearing and settlement are all network industries, differences start to emerge.

Trading is a business that is very open to competition. The barriers to entry are low. Once the investment has been made in establishing a trading system, in technical terms, the marginal cost of adding additional securities is low. There may be some regulatory and legal costs, but these are not insuperable. The relationship with customers is essentially transaction by transaction. Once customers have established connectivity to different trading platforms the cost of switching between them is negligible and the decision can be made trade by trade. Other things being equal (an important condition! We’ll discuss later what this means) the fact that my last trade was on platform A does not give me any reason for my next trade to be on platform A.

Clearing is different. Switching clearing houses trade by trade is unattractive: I would lose the ability to net trades and offset risks on my new trades against my old. There would be a cost in maintaining collateral in two places. Thus the fact that my last trade was cleared through clearing house A gives me a strong incentive to clear my next trade through the same clearing house.

However, provided interoperability is possible between different clearing houses (another important condition) it is realistic for traders to make a periodic choice between providers and switch all their business to a different clearing house that offers a preferable service.

Settlement is probably the least open to competition of the three layers. As with clearing there are significant switching costs. If I settled my last trade in settlement system A, there would be costs of repositioning my holding if I wanted to settle my next trade elsewhere.

There are two ways in which CSDs differ from other parts of the process.  CSDs have a “home country advantage” – in many countries, the CSD of the country where a security is issued offers a more direct link to the issuer – important for corporate actions – and a more secure legal framework than a CSD in another country. In addition, CSDs – unlike trading and clearing – serve a number of purposes not connected with trading on an exchange: processing corporate actions between issuers and end-investors, providing collateral for central bank operations … For these purposes, many investors have a preference for using the home country CSD. Since traders need to interact with these investors, it would not be possible for a trader to make a clean break and move all its business to another CSD.

So there seems to be a spectrum, with trading most open to competition and settlement least open.

The outcome suggested by this analysis is in fact what we see developing. There is some competition between exchanges – provided by virt-x for blue chip stocks and by the LSE for Dutch stocks – and from specialist new entrants.

In clearing there is more limited competition, with both virt-x and the LSE offering a choice of CCP. But this tends to be a periodic choice for users, not trade-by-trade, and is possible because there is interoperability between the CCPs. There is another form of competition – contestable franchises. The LSE conducted an RFP for its CCP and the ISE chose from existing providers.  This competition for the franchise reflects the network effects: it is easier to move the whole market than for individual firms to make the shift. There is still a question about who should make the decision and in whose interest it is made.

Finally, settlement is the area with least competition.

What needs to be done?

So, the structure is developing pretty much as expected, but is not delivering what the customers want. What’s the problem?

In trading, although competition is possible, it is too limited. Clearing and settlement are too fragmented. This has two consequences – in themselves, they are costly and inefficient for traders and investors who operate across multiple markets. But also, they fail to support – or in some cases, actually hinder – competition between exchanges.

Thus, when looking at European clearing and settlement, we have two objectives: that they should become more integrated and that they should support competition among trading venues. These two objectives are not always clearly distinguished in the minds of commentators. While they are not in conflict with each other, achieving one does not necessarily deliver the other. In this presentation I am addressing mainly the second objective, of supporting competition at the trading level, but the two are closely related.

Dutch trading

The experience of competition in providing a trading facility for Dutch shares is instructive.

I am sure most of you are familiar with this story, but I will briefly outline what happened, as it provides a rare experiment enabling us to compare the effect of different market structures and the conclusions are important.

In 2003 the Dutch market became discontented that their integration within Euronext was moving, except for fees, which were rapidly harmonised upwards. Deutche Börse was the first to respond, offering a trading service for Dutch shares, with clearing and settlement through their separate clearing and settlement infrastructure. This achieved less than 1% market share.

The LSE came later, but offered a trading service that used the same post-trade services as Euronext, meeting the condition we discussed earlier. For post-trade processing, trades were fully fungible between the LSE and Euronext. Initially this offering was successful, achieving a 10-15% market share. However, in response, Euronext cut its prices by an overall 30%, according to UK Competition Commission – and the LSE’s market share fell back

There are two important lessons. First, a neutral post-trade environment is a necessary condition for competition in trading. Second, it is not sufficient.

A very interesting analysis by Foucault and Menkveld[1] throws light on what else is required. They focus on traders using “smart routing” technology – technology that consolidates quotes in different markets and routes orders accordingly, so as to minimise trading costs. They suggest that liquidity on the new entrant exchange depends on the proportion of smart routers in the trading community. In the case of the LSE’s DTS, the proportion of smart routers was not sufficient for it to achieve a critical mass. As a result, DTS became stuck in a vicious circle of low intensity competition: the low proportion of smart routers means there is limited liquidity on the new entrant system and the limited liquidity means there is little incentive for traders to adopt smart routing technology.

This does not mean that competition between trading platforms will never happen. The threat of competition does work: Euronext prices were cut. Smart routing technology does exist and is used in other markets – especially the US, where there are competing exchanges and other trading platforms, like ECNs. A comment that I have heard from firms on the DTS is that the Dutch market was too small to justify implementing smart routing for that market alone. If there was competition for trading in one or more larger markets, this could lead to an alternative virtuous circle, where a high proportion of smart routers creates a critical mass of liquidity on competing platforms and this in turn creates an incentive for more firms to adopt the technology.

We should prepare for increasing competition between trading platforms. There are likely to be new entrants to the market, such as NYSE or NASDAQ. MiFID will stimulate competition by removing concentration rules, creating new categories of trading venues, such as Multilateral Trading Facilities and Systematic Internalisers, and by setting new standards for best execution. While adoption of smart routing technology at the front end is one of the requirements to make this work, so is the availability of a neutral post-trade process at the back end.

Regulatory intervention

Against this background it is not surprising that the European authorities have been taking a closer and closer interest in the structure of trading, clearing and settlement. Their interest in increasing the efficiency of clearing and settlement came first. In the European Commission, the Internal Market Directorate General has been developing its policy on increasing efficiency since at least 2004. However, in the past year they have been joined by DG Competition, with a specific focus on promoting competition. DG Competition have commissioned and published a detailed survey of vertical structures in all 25 EU countries, have collected voluminous data on major markets and in May published an issues paper on competition aspects of securities trading and post-trading activities. Finally, of course, competition authorities get involved when there is a specific proposal to be reviewed, as with the UK Competition Commission’s review of possible merger proposals for the LSE.

In the past year, Commissioner Charlie McCreevy has given the industry an ultimatum – two ultimatums to get its house in order. In July he announced his wish for the industry to develop a code of conduct addressing price transparency, separate accounting, price unbundling and open access. Of these the roadmap for effective rights of access on a fair, transparent and non-discriminatory basis is the most important and the most challenging.

Access rights

Let us now consider the many different kinds of access rights, as it is important to focus on the right kind, if we are to achieve our objective.

First, there is the question of access by market participants. In general, I do not think this is a problem. If a new bank or broker wants to become a participant in a CCP or CSD they are generally welcomed and no obstacles are put in their path. In any case, this is covered by standard number 14 of the ESCB/CESR standards and rights of access will be strengthened under MiFID next year, so I do not think this area needs to be the focus of attention.

More significant are the rights of access between market infrastructure providers – between exchanges, CCPs and CSDs. These can be considered along the two dimensions, vertical and horizontal.

We do not need to spend long on vertical access rights – making connections between an exchange, CCP and CSD in the same country. These generally work very efficiently. And if they do not, it is a matter for that market to sort out. It does not require European intervention.

Horizontal access rights making connections two CCPs in different countries or between two CSDs – are a much more complex matter. A few of these exist, but they are immensely complicated to establish. The difficulties are mostly technical and legal. The Clearstream case, which was investigated by DG Competition, is the only case I am aware of where there was found to be deliberate obstruction of access by another infrastructure provider.

In itself, horizontal access is not the answer to stimulating competition between trading platforms, but it has a contribution to make. It is a means to that end, when taken in conjunction with the last kind of access right I want to discuss.

To my mind the most important kind of access is neither horizontal nor vertical. I would like to call it “diagonal access” – this is my contribution to industry jargon. By this I mean access between different layers of the infrastructure in different countries. The ability of the LSE to use the Dutch CCP and the Dutch CSD for its DTS is an example of what I mean by diagonal access.

It is this diagonal access that will support competition between trading platforms and between clearing houses. If Euronext, the LSE or any new entrant wants to offer a competing trading platform for Italian equities, for example, they will need to have access to CCG and Monte Titoli on the same terms as those enjoyed by Borsa Italiana.

Diagonal access exists already and works between consenting organisations. What happens if they don’t agree?

Enabling one institution to obtain diagonal access to an unwilling partner requires many more questions to be answered: who can make the decision to create diagonal access? Is it the exchange or CCP providing the flow of data or the CCP or CSD wishing to receive it? Or maybe it should be the users, who actually pay the bills? Are there legitimate grounds for refusing to open up access – who should be allowed to weigh up the costs and benefits?

I can’t give you answers to these questions. But if the industry roadmap fails to address them, it will not achieve its purpose.

Assessing proposed mergers

At this point, let’s look at the various exchange mergers that have been mooted and consider how they match up to these standards.

The least problematic, it seems to me, are the possible mergers involving an American exchange, either NYSE or NASDAQ. They do not arrive with any clearing or settlement baggage either in Europe or in the US and come from an environment where they no longer expect to own clearing or settlement assets. These mergers seem unlikely to create new barriers, but their ability to stimulate new competition depends on their ability to get access to critical bits of infrastructure.

More problematic to my mind is a possible merger involving Deutsche Börse, which (through Clearstream) owns the German CSD and through Eurex controls the CCP for the German market. Given the intense interest in post-trade activities, Deutsche Börse’s statements on the shape of clearing and settlement after a merger have been quite limited. They have said (I quote)

“the combined group would commit to not extending the integrated model beyond the markets in which it currently operates, namely the Germany cash equities market. If and when a European solution emerges for cash equities clearing, Deutsche Börse would carve out its German cash equities clearing activity, which currently forms part of Eurex Clearing, and contribute it to a European solution.”[2]

No mention of opening up access to German clearing and settlement for other exchanges. In other words, a merged Deutsche Börse-Euronext could dust off the plans that Euronext was working on for offering a trading service in UK equities using LCH.Clearnet and CREST, but there would be no commitment to enable the LSE to offer a similar service in German equities using Eurex Clearing and Clearstream Frankfurt. Maybe Deutsche Börse would willingly provide such access but I have failed to find any statement confirming that they would.

It is interesting to compare this approach with the public policy of another vertically integrated exchange that has also been mentioned as possibly being involved in merger discussions – Borsa Italiana. Their public statement says

“Borsa Italiana Group promotes and offers the post-trading services of Cassa di Compensazione e Garanzia S.p.A. and Monte Titoli S.p.A. in an equitable, transparent and non-discriminatory manner and on the basis of criteria and procedure aimed at assuring interoperability, security and equal treatment among market infrastructures…[3]”

This is a very different approach. And they can point to examples where they have provided such access.

Of course, unlike Deutsche Börse, Borsa Italiana is not yet a publicly quoted company and it would be interesting to see how these principles would survive after a listing or a combination with Deutsche Börse.

Personally, I find it very hard to see how a closed vertical structure can survive in the environment that is taking shape. When Europe was a collection of self-contained markets, a vertical silo was a source of strength. When Europe is a continent of consolidating infrastructure, it becomes a constraint. As we saw in the case of Dutch trading, when Deutsche Börse wanted to offer a competing trading venue, but use its own clearing and settlement infrastructure, it was less successful than the LSE, which used the same post-trade infrastructure as the incumbent. A vertically integrated structure is a handicap there.

I doubt whether any of the vertically integrated exchanges could take part in merger activity with another exchange without attracting the attention of the competition authorities. The issues paper from DG Competition says “vertical integration may result in foreclosure at all levels of the value chain”. This seems to me to put down a pretty clear marker. Intervention from the competition authorities would put a vertically integrated exchange at a disadvantage compared with an unencumbered competitor. In addition, when there is a bank within a group, it results in a commitment of capital which could otherwise be returned to shareholders. It is likely that vertically integrated groups will become subject to such stringent rules requiring the different parts to operate at arm’s length from each other that there will be little benefit left in the vertical structure. I would not be surprised to see them being dismantled in the near future.

But when that happens, don’t cheer too quickly. Remember that quoted companies have an obligation to realise the maximum value for their shareholders. If Clearstream was  offered for sale, would the users come up with €3 billion to buy it? I’m sure the money they received for selling half of Cedel to Deutsche Börse in 2002 has all been spent by now. It is more likely that the highest bid will come from a commercial operator – Macquarie Bank perhaps, a private equity house, or maybe Citibank.

Remember that we identified two objectives for a future clearing and settlement industry structure. Dismantling the vertical structures would address one of them – creating an infrastructure that supports competition between trading venues. But it would not necessarily achieve the other – creating a more integrated clearing and settlement structure.

Conclusions

It is easy to say that we want more competition in trading and more integration in post-trading. In an ideal world this is what we would get. But we may not get everything we want. If we have to choose, I would stick my neck out and say that I would push for tough, vigorous competition in trading, even if this meant we got less integration in clearing and settlement. Yes, of course, there are inefficiencies in fragmentation. Yes, these inefficiencies impose a cost. But I believe these costs would be worth paying for the cost reductions and innovation that would be released by stirring up real competition in trading.

http://ssrn.com/abstract=686141

[2] Deutsche Börse revised proposal for combination with Euronext, June 2006

http://deutsche-boerse.com/dbag/dispatch/en/listcontent/gdb_navigation/investor_relations/60_News/10_Press_Releases/Content_Files/13_press/June_2006/pm_news_euronext_190606.htm

[3] Borsa Italiana statement on post-trading and custody

http://www.borsaitaliana.it/prodottieservizi/postnegoziazioneecustodia/postnegoziazione.en.htm

September 2006

2. The ECB steps in ….
On 7 July the ECB issued a press release[1] stating that it was evaluating opportunities to provide efficient settlement services for securities transactions in central bank money, leading to the processing of both securities and cash settlements on a single platform through common procedures. It had decided to explore the setting up of a new service – which may be called TARGET2-Securities – for securities settlement in the euro area.

This announcement and subsequent developments seem to us to raise many important questions. In this commentary we aim to identify the key issues that we believe need to be clarified.

1 The process for reaching a decision

The ECB expects to make a decision by “early 2007” – about six months from now. The proposal has apparently been discussed with a small, but undisclosed group of banks. For the financial community at large, there is nothing concrete to evaluate other than the press release and a briefing given to journalists.

The ECB proposal has implications for all the users and owners of the European security market infrastructure – that is, virtually every financial institution in the EU. So far the consultation process seems not to have started, as there has been no presentation of what is proposed, no public timetable and no explanation of the criteria on which a final decision will be made. Presumably the ECB intends to undertake a full consultation so that it can understand the ramifications of its proposal and ensure that whatever decision it makes will command public support, but the details are not yet known.

2 What is actually being proposed? What benefits will it bring?

The background to the proposal is a fairly technical debate about the best way to synchronise delivery of securities with payment when settling a transaction in securities. There is general agreement that the most efficient approach is for both security and cash movements to be managed by the same platform.

In some countries – for example, the UK and France – this process is managed by the securities settlement system (also known as CSD), which determines when settlement takes place. As a result, the CSD effectively “controls” some payments across the books of the central bank: when the CSD determines that a transaction has settled, this causes the money to move on the books of the central bank.

In other countries, however, the central bank is unwilling to “outsource” control of central bank payments to another organisation. So, to maintain an integrated system, if the CSD cannot manage the money, the central bank must manage the securities.  The result is TARGET2-Securities (known for short as “T2S”).

The proposal appears to be that, at the start of every day, participating CSDs transfer their securities balances and outstanding transactions to T2S. During the day, T2S settles these transactions and at the end of the day reports to the CSDs on what has happened.

One consequence of this model is that it separates the operation of settlement from the other functions performed by CSDs, such as asset servicing, asset financing and provision of collateral. These other functions require access to real-time, intraday information on the securities balances held by participants in the systems and the ability to control those balances. Making this possible will require a sophisticated linkage between T2S and CSDs and, even so, will introduce processing delays while information has to be exchanged.

It is not at all clear how the introduction of T2S will result in any savings for the industry. CSDs will face the one-off costs of adapting their systems to transfer settlement functions to T2S; they will still need to maintain and update all the information they hold now; and in addition will have to maintain an interface to T2S. CSDs will not be able to reduce their running costs by much, but, on top of those current costs, the industry will have to bear the new costs of developing and operating T2S.

Developing T2S would be a major project. Using data from 2004 (the latest year for which consistent data[2] are available), there were over 500,000 securities settlements per day in eurozone settlement systems. This compares with a daily average of 270,000 TARGET payments. In other words, TARGET2 would not be a payment system that also settled securities. It would be a securities settlement system that also handled payments.

3 How does it relate to settlement in other currencies?

By linking its proposal to TARGET 2, the ECB implies that it will provide settlement only against euros. However, several systems provide settlement in both euros and other currencies, notably VPC in Sweden, providing settlement in euros and kroner, and CREST, providing settlement in euros, pounds and dollars for both the UK and Ireland. Both VPC and Euroclear are undertaking major projects to integrate settlement platforms between countries that outside and inside the eurozone (Sweden and Finland in the case of VPC; the UK, Belgium, France, Ireland and the Netherlands in the case of Euroclear).

If the ECB proposal requires settlement of euro transactions to be handled separately from those in other currencies, it would actually be a backward step, as it would undo the integration that currently exists. It would also remove some of the flexibility that currently exists for UK securities to be settled against payment in euros or eurozone securities to be settled against payment in pounds.

4 Governance

The ECB has said that it intends to own and operate T2S. This goes against the trend of recent years, which has seen central banks in France, Italy, Spain and the UK transfer their securities settlement activities to private sector operators.

The governance model remains unclear. How will the ultimate users of the system be able to influence its operation? Will the ECB put in place any user committees? How will their membership be selected? Will the ECB bind itself to follow their recommendations?

Similarly, the commercial model for the operation of T2S has not been explained. Will the ECB undertake that funds will be available for future investment in the system when this is required? What will happen to any surplus that it generates?

Conclusion

The project risk in a development that has to interface with 12 different legal jurisdictions is significant. But there is also the risk that announcing the T2S project will stall other projects already under way aimed at integrating European markets. The ECB needs to do much more to explain its proposal, which at face value seems to increase risk, cost and fragmentation in European securities markets and goes against the trend in other developed countries.

[2] Source: ECB Blue Book http://www.ecb.int/paym/market/blue/html/index.en.html

3. US Exchanges and Clearing Houses

4. European Exchanges and Clearing Houses