European Harmonisation of Cross-Border Settlement
Conference 21/22 March 2002
Clearing and Settling European Securities: Where Competition Works (and Where it Doesn't)
Don Cruickshank
Chairman of the London Stock Exchange
Thursday 21 March
11.05am
Click here to see the press release
Introduction
Good morning ladies and gentlemen. Here I am again addressing the vexed question of the future of clearing and settlement in Europe.
Why bother, some people say, you only run a trading platform. Don't bother, others say, it's all safe in our hands.
So what's my interest in this subject?
The first is the public interest of an integrated capital market for Europe. Reform of the clearing and settlement infrastructure is essential if we are going to deliver that integrated market and capture the benefits, in terms of a lower cost of capital for European firms and competitively priced financial products for European citizens. As an aside, I note that the EU Heads of Government made specific reference to the need for more efficient clearing and settlement "at the European level" in the conclusions of last weekend's Barcelona Summit. I couldn't agree more.
My second interest is as chairman of the London Stock Exchange. The reality - from the perspective of our customers - is that we sell a joint product in economic terms - trading, clearing and settlement - they don't draw a distinction - and increasingly our customers are demanding access to a wider range of financial products. We fully intend to compete to serve those customers over the range of equities that are available across Europe. It follows then that we need reform of the clearing and settlement layer to allow us to compete to provide exchange services and cheaper access to these products.
To those - increasingly few I'm glad to say - who have suggested that Europe is well-served by the existing infrastructure, I would ask, why then does it cost so much more to settle an inter system as opposed to an intra system transaction? In the case of Crest settling a security held in Germany between 17 and 65 times the cost. For a French security even more.
And to those who have responded that "You would say that wouldn't you - it's only because you don't own Crest" - my response is that we have no interest in owning a CSD and that any short-term benefit to our shareholders if we did, would be just that - short term. Our view - which is beginning to gain currency - is that the vertical silo model, replicated in Member States across Europe, acts against the interests of our ultimate customers - who are, just in case we forget, firms across the economy and all investors, that's who we are here to serve. Existing structures are inherently anti-competitive and therefore unsustainable in economic terms given the well-established principles of EU competition law and the well-understood powers to enforce them.
My task today is to try to persuade you of the strength of these arguments and to reinforce our position on a few key points:
First, that there are parts of the securities industry the relevant market in economic in which competitive pressure can lead to better and more efficient outcomes. They need rigorous application of competition law. There are, however, other parts of the industry - other relevant markets in which a competitive outcome is an illusion. They need public policy intervention.
Second, the idea that rules providing access and enforcing interoperability will deliver effective competition at the CSD layer is well intentioned, but misguided.
Third, that the public policy intervention to deliver the necessary structural solution has to come from the institutions of the EU from a European perspective.
Finally, that though it's a difficult problem to crack, it's not insurmountable. But crucially, we shouldn't and can't expect the solution to come from the producers - or their shareholders.
Agenda for Today
This morning, I'll summarise the position of the London Stock Exchange as it has developed since I first commented on this subject last year. Then I'll draw out some themes from the various recent contributions to what is becoming a complex public debate. And in particular, I'll take the main criticisms of our position head on and attempt to rebut them. I will also be re-emphasising the importance of a single CCP as an essential part of the solution.
Summary of LSE position
So first, the summary of the London Stock Exchange position.
One: the costs are high, at least 1.6 billion. That's just the direct costs. The indirect costs are much higher. And these costs end up in the costs of capital of European firms - to the tune of 10 basis points probably more -damaging their competitiveness in world markets.
Two: these excess costs flow from fragmentation of the clearing and settlement system, that in turn, fragments the whole capital market. There is no effective competition between exchanges. Dominant providers in domestic markets are the order of the day and in particular, exchanges that own or aspire to own clearing and settlement systems extract rents - super normal profits -from their customers and bundle charges for the joint services they offer.
Three: the economic characteristics of clearing and settlement mean that CSD services are more efficiently provided to society at large by a single system than multiple systems, given the strong positive network effects that arise from having everyone connected to the same system.
Four: it is not possible to introduce effective - note effective - competition into the clearing and settlement layer of activities. There is little substitutability and the majority of services offered by the different clearing and settlement systems are not in effective competition with each other.
Five: this analysis leads us to conclude that Europe is best served by a combination of:
- one, a well-governed single clearing and settlement system that would minimise market distortions and maximise the expected economies of scale and network externalities; and
- two, we must have rigorous enforcement of competition policy in the separate markets of trading services and all the various value-added services currently dominated by the owners of CSDs.
Themes from Commentators in the Debate
In the last few months, various commentators and agencies, including the European Commission, the G30, the Giovannini Group and the European Central Bank - and including producers in the shape of Deutsche Brse - have all piled into the debate. Most important perhaps for the prospects of a structural solution is the current investigation by DG Competition of the European Commission. Our analysis in response to the Commission's requests for information identified a whole number of apparent breaches of well-established principles of EU competition law and when he spoke on this subject in July, Commissioner Monti certainly appeared to confirm the position.
He said that the Commission have identified - and I quote - "a number of possible competition concerns relating to the infrastructure for clearing and settlement. These concerns relate to access, vertical silos, excessive prices, exclusivity and consolidation."
Also the G30 Group have flagged up some interesting issues that require further analysis, including:
- the possible mandating of a regional monopoly supplier based on the DTCC model as a not-for-profit, user-owned utility.
- The possible forcible segregation of monopoly services from other services-ie, trading platforms could be barred from owning or controlling clearing and settlement systems in a structural solution.
- organisations could be compelled to allow users to select only those services that they require at a 'fair' price "unbundling" and to allow other organisations to compete for the provision of other complementary services.
From these and other recent contributions - for example from the Giovannini Group who recently issued their first report - I would suggest that, despite the apparent wide spread of views and policy prescriptions, there is a welcome degree of consensus beginning to emerge.
First there is an acceptance that the markets for clearing and settlement aren't working, that this is costly and that "something needs to be done" to deliver effective solutions.
On the other hand, there is still disagreement on the extent to which competition can be introduced and particularly on the potential for competition between CSDs. This in part is dependent on whether one accepts the definition of a CSD as a natural monopoly. It also depends on what faith you place in interoperability. Even those who strongly recommend competition see consolidation to a small number of CSDs - say 3 or 4 - as inevitable and welcome. So, even within this optimistic scenario we could find the market just shifting from national dominance to European joint dominance which could still stifle effective competition in other parts of the capital market.
There is also a welcome recognition that the scope for market pressure by direct users of the system is limited. This is partly a consequence of the power of the vertical silos to bundle costs but is also explained by the real lack of incentives on intermediaries. The fact that only two of Cedel's shareholders stood out against the proposal for Deutsche Brse to buy Clearstream is illuminating.
And finally and this is where our analysis is beginning to gain support, there is broad agreement on the need for some from of intervention on public policy grounds. Not yet in favour of the structural solution that we have recommended, but at least to try to limit and compensate for the market power of the vertical silo. Either, by regulation of access and prices, or, more radically, to require exchanges to divest themselves of clearing and settlement organisations.
So we're getting somewhere. However, the debate has also thrown up a number of criticisms of our position along the lines of:
- We can't interfere in market structures: competitive solutions are possible.
- Anyway interoperability will work, it's like telecoms.
- If we go for concentration, we'll lose innovation and increase systemic risk.
- It's all too difficult, there's too much resistance (with an underlying nationalistic perspective).
So, I want to address these main challenges to our position. Briefly today. In more detail in the further analysis we will be producing in the near future.
What is a competitive market?
Let me start with "Competition will give us the answer".
I trust that I don't need to establish my credentials as a promoter of competition. During my time at OFTEL, the objective was consistent: where possible, offer companies the chance to develop, to innovate and to provide the services that customers want.
Moving on to the banking sector, in 2000 I carried out a competition review for the Treasury only to find that in many different ways a regulatory contract between the banks, government and regulator had written the rules. I therefore recommended improved competition scrutiny, prevention of anti-competitive mergers and increased transparency in banking supervision.
Always I adopted a users' perspective and never expected established, especially dominant or monopolistic producers to agree with me.
And it should also be remembered that every time I have gone to the competition authorities for regulation or structural solutions this was only because competition wouldn't deliver.
These experiences have taught me that there are activities in the economy where competition works and areas where it doesn't, and especially where networks are involved, the relationship is complex and needs careful analysis to get it right. I think it is worth examining what conditions are necessary for competition to work.
The first thing to say is that competition is itself a process not an outcome that, when working effectively, delivers significant benefits to end users. That process also delivers benefits to some, but only some, producers. The private interests of other producers are harmed by competition - they fail to make a profit, they go out of business, they get taken over etc. Taken together, producer interests are, at least in the short term, harmed by competition. Hence the need for tough competition law to ensure that producers do not thwart the process of competition. And I have to say that DG Competition on limited resources have always acted with great rigour and responsibility.
Broadly speaking, to be effectively competitive, a market needs end-consumers to have a real choice of suppliers for any particular service, these end-consumers need to be informed by appropriate price signals and their actions need to translate into the success or failure of individual producers.
Unfortunately, in the real world, having a number of different suppliers providing the same thing does not mean that competition always works this way. In some markets, companies have market power which they may abuse so as to impair competition-for example, by restricting access to essential inputs. Something a vertical silo might be expected to do. Or price signals may be absent or distorted, in which case consumers cannot make informed choices. To those who claim this isn't the case in the market for European securities, I would ask them to look at the token price list for settling German equity.
Or, and this is a real problem in the clearing and settlement markets, while the output of a number of firms is apparently the same, it is not the same as far as consumers are concerned, so there is no real choice. What economists call a lack of demand side substitution. The current market in the core services provided by CSDs is a case in point - there are lots of CSDs in Europe (and indeed around the world) but if I, as a end-consumer, want to trade, or hold electronically, a particular equity I seem to be offered no real choice.
Now, in many industries involving networks, some activities in the supply chain may be candidates for competition, while others tend towards natural monopoly. An obvious example is energy, where generation and supply are competitive but the national transmission grid is a natural monopoly.
The transacting of securities is such an industry. The challenge is to identify which activities can be competitive and which cannot. In other words, to find the optimal mix of competition and regulation so as to deliver the best result to end-users.
There are lots of potential services to look at. The obvious ones are:
- Clearing
- Settlement
- Custody - in the form of holding electronic share certificates
- Custody - in the form of acting as an agent for someone further up or down the chain
- Stock lending
- Risk managing
Some of these markets look potentially much more competitive than others - 100s, or at least 10s, of suppliers, low switching costs for customers, reasonably transparent prices.
If I just focus on the markets in which CSDs operate. These CSDs have market power over most, if not all, trades and electronic holdings of their home equities. And here, perhaps, it is worth pausing just to see what one might do with that market power. Are there related services - say stock lending or asset servicing - which customers want, and could be provided by third parties as long as they had access to the customers of the CSD? And if there are such services, do third parties get non-discriminatory access? Or is it something that can only be supplied by the CSD itself? Are there artificial barriers placed in the way of competitive supply, or is it really the case that somehow these services are intrinsically cheaper if supplied by the CSD itself? This is what a competition authority would take a long hard look at.
But what about competition between CSDs? For as long as it remains expensive to move equities from one CSD to another, it is difficult to see how CSDs can compete to hold the same equity. Traders want access to the biggest pool of potential buyers and sellers of a particular equity. Generally that equity has a home CSD where the electronic shares have been created. That is where the pool exists and, unless there is a really big cost differential - or other disaster - that is where the equities will stay. But there are other problems. The layers of intermediaries between the end users and the CSDs mean that those who might want to find a cheaper, or better, alternative may not even know what costs are being incurred on their behalf. Price signalling is weak. And the intermediaries? As long as they all face the same costs, then do they really care that much if the costs are higher than necessary? And if they have sunk investments which are valuable only because they overcome one or more inefficiency, then they have a positive incentive to keep that inefficiency.
There is another area where this type of analysis is illuminating. One of the arguments we have used is that, compared to the DTCC, European investors must pay considerably more to the European CSDs (taken together) than they would to the DTCC for the same volume of transactions. We have been doing more work on this problem and, along with CEPS, have highlighted one of the main differences between the two systems - namely the absence of significant netting of transactions. This, in turn, seems to be largely as a result of a lack of a single central counter-party.
Our analysis suggests that to net efficiently across an integrated European capital market would require a single central counter party. Notwithstanding the emergence of a number of different CCPs in the existing fragmented world, if Europe is to benefit from efficient netting, efficient capital requirements for intermediaries and other benefits, what is required is a single central counter party. This is a function that really doesn't seem to be a candidate for competitive supply. And it looks like most commentators agree.
But will it emerge? In a fragmented market, with not many trades across CSD systems, there will not be many trades to net across those boundaries, so multiple CCPs do not seem to be the cause of much inefficiency. But their mere existence reinforces the fragmented nature of the market, as cross system trades are even more expensive, requiring - some say - inter-linkage of the CCPs. So the costs mount, and so it goes on..
Interoperability will do the trick
Faced with this analysis, and sensing that DG Competition might share these conclusions, producers say that interoperability will do the trick. Again the analogy with telecoms and other networks is made: just ensure interconnection and open access between those networks and then you get competition.
So let's test whether competition will deliver the best results to end-users under full or partial conditions of interoperability.
First let me disabuse anyone who thinks it's like telecomms. Otherwise there is a danger that the EU will draw the wrong lesson from the liberalisation of that sector. Careful analysis of that experience indicates that a dematerialised security held in a particular CSD is most like a telephone connected to a particular, local, network. Trading requires access to that security - just like a call to a particular customer requires access to the network to which they are connected. Liberalisation of the telecommunications sector has dramatically increased the competitive provision of telecommunications services, except for - a big exception - the provision of call termination services. In lay terms, the last bit of network connection to you, the receiver of a telephone call.
This analogy should not be taken too far - but I am afraid that we have a situation in which the best fit for the clearing and settlement problem is precisely the one part of the telecommunications market that has not become particularly competitive. So relying on a telecommunications model to provide the solution is misguided - no, let me say, plain wrong.
Setting aside the comparisons with telecomms, now let's turn to what interoperability between CSDs actually means - since it is used to describe very different things. At a minimum, interoperability means there are efficient electronic links between CSDs allowing DVP settlement. This means users of one CSD can settle securities transactions in another CSD. The way this normally works is that a CSD acts as accountholder in the foreign CSD on behalf of its own accountholders. It also requires the creation of some form of 'shadow' securities - called depository instruments here in the UK. This means local custodians can be bypassed for cross-border settlement purposes.
Going a step further, interoperability would also allow the provision of asset servicing via the linked CSDs. This business model, which is being pursued by CREST and the Swiss CSD, would allow the bypass of local custodians also for asset servicing. However, the reality in the marketplace is that, without full harmonisation of legal and fiscal rules in each jurisdiction, it will be virtually impossible to bypass these local custodians for these value added services post-settlement.
This immediately points to one of the reasons why interoperability may well not reduce costs to end users: local custodian services are still required. This is illustrated by the fact that many of the current 60 or so links that exist between CSDs in fact attract little or no business; they simply don't offer the full range of services that end-users require. Some CSDs, such as Euroclear, follow the business model of internalising the costs of local custodians and offering end-users the full range of services. However, this is clearly still costly: from our previous calculations it turned out that Euroclear's costs are significantly higher than those of the national CSDs. Now don't get me wrong: we know these higher costs are due to the greater level of value added services provided-and Euroclear may well be a lot cheaper, and more efficient, in doing this than the custodians are-but it does illustrate the high cost of current fragmentation in Europe.
But lack of harmonisation is not the only reason why interoperability will not deliver the solution to the European clearing and settlement problem. The other reasons are that, first, interoperability is costly in itself, and, second, interoperability will not lead to effective competition between CSDs.
As to the costs of interoperability, there are first of all the one-off costs of creating the links. These costs can be huge. They are not only about IT infrastructure, but in fact to a major extent about costs of finding out about, and complying with, legal and fiscal requirements in the jurisdiction of the CSD that is being linked in to.
In addition, there is the loss of direct contact between listing companies and their shareholders. Under the "ideal" interoperability model, in which custodians are bypassed, the foreign CSD will be registered as the holder of the original certificate in its account in the home CSD. However, it is no longer directly visible to the share issuer who the actual owner of the shadow security is in the foreign CSD, or when and how often that shadow security changes hands.
Further, having shadow securities floating around across several CSDs may significantly affect the possibilities of efficient netting. This is because, for netting purposes, an original certificate of a certain security may not necessarily be considered identical to a shadow certificate of that security. Two shadow certificates of the same security may not be considered identical either. This netting problem could only be overcome through a single CCP that links into the different CSDs; a solution that I will touch upon later on.
I have not seen any description or analysis by any of the advocates of competition among CSDs of how such competition actually would take place. My own experience, and the analysis we have undertaken and made public, clearly suggests that the claim that competition will work once networks are interconnected is made too lightly. The London Stock Exchange is currently preparing further analysis for the European Commission in which we discuss why competition between CSDs or CCPs doesn't work, not even under full interoperability - a wonderland in which the necessary laws have been harmonised.
I know that was just a quick canter through an abstruse and geeky problem. But it seems that life is full of such problems if you really want to deliver economic reform. Declaratory public policy - Stockholm, Lisbon, Barcelona -is easy. Real reform is excruciatingly difficult. Those of you interested in this vital decision: "Is interoperability enough?" may wish to read our imminent submission in more detail.
A Single CSD won't Innovate
Another loose criticism - from the economics text book - is that a single provider, even well-governed, just wouldn't innovate. Of all the criticisms this is the one that is possibly least well-grounded in fact or logic.
Let's be clear, each national market is already served by a monopoly provider of CSD services. These CSDs all innovate - certainly in the case of Crest they provide leading edge technology solutions. So no lack of innovation there.
And I think a proper examination of the DTCC would demonstrate to any critic that such a single provider offering services to several different competing exchanges innovates consistently and successfully. Indeed, its business model is geared towards generating new revenues from new products and services at quite a fast rate. That is the best and acceptable measure of innovation performance. And it's a measure against which DTCC and many European CSDs have performed well.
Systemic Risk
Another spectre that has been raised particularly since the events of 11/9 is the suggestion that a single provider somehow increases the level of systemic risk. There are two possible ways in which this objection can be interpreted. Neither is convincing. First, that there is somehow a greater operational risk to the market from relying on a single provider. To which I would say the following. The market is no more protected now from disaster because there are lots of different providers of services given that they are not able to substitute for one another. Put crudely, if for example Crest were to be out action could its customers switch to say Clearstream? No.
And would this operational risk go away in an interconnected and interoperable system? No, if anything it would be worse, since an interconnected system is usually only as strong as its weakest link and this link is potentially difficult for other participants to identify. A single provider on the other hand takes responsibility for the integrity of the system as a whole. Accountability is clear. Its standards on connectivity to customers are better founded, and is legitimised to dig deeper into customers' systems and network decisions. Think bank payment systems as the model. And it's cheaper.
And on the management of credit risk. It's an issue that is dealt with successfully in the US and by every domestic CSD in Europe. There is no reason why a single provider is more vulnerable - and many reasons why it would be likely to succeed - in managing that risk. We welcome in particular the consultation launched last week by the ECB and CESR to determine appropriate standards and we expect its conclusions finally to lay this spectre to rest.
It's all too difficult
And then we have the "It's all too difficult" argument.
Here, let me start with how important it is that DG Internal Market takes the lead and deploys sufficient resources in tackling the difficulties.
In their initial paper on the ISD, DG Internal Market recognise many of the current problems - noting that transactions either must be finalised through a complex chain of clearing and settlement systems before ownership rights are ultimately transferred; or the purchaser is required to be a member of multiple systems with all the associated costs.
However, of more concern is DG Internal Market's apparent reliance on strengthening competition between CSDs - essentially by ensuring non-discriminatory access to CCP clearing and settlement. Article 15.1 of the existing ISD stipulates that authorised firms can have remote access to clearing and settlement systems. The Commission proposes to strengthen this by allowing market participants to designate the place of settlement, and regulated markets to use CCP services in other Member States.
In other words, the measures proposed by the Commission do not call into question existing ownership or management arrangements for regulated markets/trading platforms, CCPs/clearing houses or settlement systems. The provisions seek to ensure that joint or linked ownership of regulated markets and CCPs or clearing and settlement systems (eg, vertical silos) does not impinge on any decision by the operator of CCP or clearing and settlement systems to make access available to unaffiliated regulated markets. The ISD seeks to provide access to relevant systems where this will assist a regulated market or its participants to secure more efficient finalisation of transactions.
Part of the reason for this of course is that DG Internal Market has a stated policy of taking no position on market structures. My comment on that is to say: That's ok where there is the prospect of effective competition ie where there is a "market". But where there is no such prospect - as in this case - those responsible for public policy are obliged to take a view and act on market structure. This is something I hope that the Commission will address in its Communication.
In addition to predictable criticism of our position from some producers of the services in question - "we are already efficient" - "comparisons with the US are unfair" - "it will all be resolved by the market" - came the suggestion that what we have proposed is fine in theory but not in practice.
A number of reasons have been advanced:
First: that shareholders won't willingly give up valuable profit streams. I agree that this is an important potential barrier which won't be helped by the current process of exchanges becoming for-profit organisations, some of them building silos, with shareholders who are value investors. In reality, value investors invest in such exchanges in part because they hope to enjoy the returns derived from their ability to extract economic rents: first, from exploiting their monopoly position in a domestic market; and, secondly, from their vertical ownership of clearing and settlement and credit functions.
So having value investors own exchanges could even have the opposite effect: to delay reform. But this is not an argument against public policy intervention. These are sophisticated investors who know the risks and deserve no compensation for the loss of such revenue streams.
Then we get: Intermediaries will resist change, either because they have perverse incentives - to continue to exploit the current inefficiency - or because they have commercial constraints on how much they can push for national reforms in Europe. Intermediaries have all built long-running, profitable relationships with governments or other issuers - relationships, understandably, they are reluctant to upset. And as intermediaries, they usually have the ability to pass on costs to end users, protecting them from any serious negative effect on their bottom line.
However, to this I would say that they are also paying additional costs of the additional capital that they require as a result of fragmentation and particular underdeveloped central counterparty systems and, hence, the ability to net exposure positions. Our assessment is that the cost of this inefficiency is of similar order to the costs to firms and frankly it should be sufficient incentive for change.
There are also jurisdictional problems to contend with in considering where a single provider of clearing and settlement services would be located and what might be the relevant law. I agree, these are serious questions and would likely be the origin of a similar set of political trade-offs as were generated by the creation of European Central Bank. But none of this is insurmountable, given the political will and the sheer scale of the prize - a more competitive European economy.
Single CCP
It would be an oversight to conclude my speech today if I didn't mention our support for the creation of a single pan-European central counterparty. Currently, with the exception of LCH in London, all domestic CCPs in Europe provide services solely to national exchanges. Often, there are also ownership or governance ties between these CCPs and the national exchanges and/or their respective national CSDs.
Before even talking about the prospects for a pan-European body, I should observe that the LSE's experience is that much advantage can come simply from the introduction of the trade anonymity and credit intermediation that a CCP provides. Liquidity on the LSE's order book, SETS, has grown markedly since the CCP started operation in February 2001. A further development of our CCP service, the introduction of settlement netting, is scheduled to go live in July and is now being tested. The benefit to the market of settlement netting is a reduction in the number of settlements. This leads to reductions in settlement costs, settlement fails, intra-day credit and liquidity requirements and delivery risk.
The European Securities Forum were the first to raise the idea of a pan-European CCP in 2000. The founding principles, they suggested, should be that it should be user-governed and owned, be located in a single jurisdiction and be capable of netting across competing exchanges. Well, we are all for that, I must say. Since that time though, despite some very significant efforts by Pen Kent, ESF's outgoing chairman, at brokering a merger between LCH (which already conforms to the ESF user-governed principle) and Clearnet, the French CCP, or Eurex Clearing, the German CCP, progress has been obstructed by the unwillingness of the national exchanges that own the latter two bodies to give up the sizeable revenue streams they receive from them.
Further integration at the CCP level certainly makes sense with increasing amounts of cross-border trades. The more trades that are cleared by the same CCP, the greater the potential for settlement and risk netting efficiencies, the more efficient the use of collateral, the greater the economies of scale that can be drawn from a single set of operating systems.
Also, if legal 'equality' can be created between a trade in a security on exchange X and a trade in the same security on exchange Y, inter-exchange netting will occur. If there were equivalent access routes to settlement for exchanges X and Y (back to the CSD layer again), all other things being equal, more competitive tension will exist between the exchanges, forcing them to pay even greater attention to delivering value for money to the end user.
However, there appear to be a fundamental problem with the single CCP model. Because the single CCP would provide interoperability between CSDs by virtue of having a DVP account in each one, end-users would be spared the direct costs of cross-border settlement - all end user settlements would be with the CCP account in their home CSD. But the single CCP itself would incur all of the cross-border charges in this case. These would be passed back to end-customers in a very indirect and "smeared" fashion. One would believe though that if enough critical mass can be achieved in the netting process (reducing the number of settlements) the total costs of the system will be reduced. It is not clear though at this stage what this 'break even point' needs to be.
A single European CCP could work without CSD restructuring but we think probably not. In any event it would require difficult decisions and compromises to create: which company law? which regulator? which bankruptcy law? which location? which technology? which business model? All difficult choices and trade-offs to be made.
Certainly, exchanges will have a major role to play in pushing for the creation of a single CCP. If enough of the bigger ones agree to clear their business through the same CCP, it is likely that will 'tip' the others into following suit.
A single CCP would be significant advance towards an integrated capital market: it would put competitive pressure on exchanges; and make it easier to achieve the desirable CSD structure.
Conclusions and next steps
The theme of today's conference is of course, "European harmonisation of cross-border settlement." I've tried to make the case for real harmonisation and my belief that Europe would be best served by a single well-governed monopoly provider of settlement services. I have answered the objections that have been advanced since we first made this call and set out some practical steps for how we might get from here to there, including the crucial role of the single central counterparty.
To summarise:
First and most important, to gain a proper understanding of the market in Europe we need a rigorous de-layering of the current fragmented market or set of markets: separating out core services from the value-added services that CSDs currently provide: and analysing where competition can be introduced, and where natural monopoly characteristics eliminate the scope for introducing effective competition.
This analysis would, I am confident, lead to the conclusion that we have reached, that interoperability cannot provide an efficient solution.
That solutions can only come from the EU institutions. We know it's difficult, and don't expect producers to make it any easier.
In the meantime, we look forward to the forthcoming output by DG Internal Market. I can assure the audience that we will be actively responding to its consultation. We will also be responding to the consultation launched last week by the ECB and CESR. I am pleased that they have decided to launch their investigation and in particular to look at the scope for standard setting to mitigate risk.
We also look forward to the second report of the Giovannini Group and forthcoming conclusions of the G30 in the firm belief that they will move the debate forward.
On a final note, I said a moment or two ago, that this conference was about progress towards cross-border settlement for Europe. I suggest that the whole concept of "cross-border" within the EU is antediluvian and that we will know that we've got there, when cross-border means from Europe to the US and not from Portugal to Spain or Italy to Greece.
Thank you.