CEO ● EuroCCP
Editor’s note: This story originally appeared in the Fall 2008 edition of Global Custodian magazine.
The European securities services industry was never more interesting than it is today. New trading platforms are proliferating, settlement systems are allying themselves with one another, new and old clearing houses are scrambling for position, agent banks are pondering the mismatch between their historical structure and the changing shape of the markets, and regulators and central banks have become actors as well as observers and supervisors. Amid the most contentious, exciting and unpredictable developments in European market infrastructure for years, the London-based CEO of DTCC-owned EuroCCP is having the time of her life.
In July this year, Diana Chan was expectant. If all went according to plan, Turquoise, the first of two European equities trading platforms that the organization she runs will supply with clearing services, soft launched for the second half of August, prior to going live across 1,300 stocks in 13 markets from September 5. Later in the same month SmartPool, the “dark pool” trading platform created by NYSE Euronext, BNP Paribas and HSBC and also supported by EuroCCP, was planning to launch. With the London market spawning what the scriveners of the European Commission have taught us to call multilateral trading facilities (MTFs), these are certainly propitious times in which to be in the central counterparty clearing (CCP) business. Trading firms can now choose not just between Turquoise and SmartPool, but also from BATS, ChiX, Liquidnet, Nyfix and Pipeline as well. All are challenging the established stock exchanges of Europe – and, of course, each other – to command a sufficient share of equity market liquidity to survive, and all need a CCP to support them. In this highly competitive arena, even the normally torpid London Stock Exchange is worried enough to have teamed up with Lehman Brothers to build a “dark pool” of its own and begun to consider how it might add an equity CCP to its repertoire. So it is not hard to see why Diana Chan thought being CEO of EuroCCP, the European clearing subsidiary of The Depository Trust & Clearing Corporation (DTCC), might be an interesting job.
She took the position in the very week in November last year when the Markets in Financial Instruments Directive (MiFID) took effect in Europe. It is to the price transparency initiated by this measure, introduced by the European Commission, that the explosion of competition for liquidity in Europe is usually traced. This has in turn sparked excitement in the clearing area as well, and who better to play a leading part in it than an American CSD and CCP group that have seen it all before? “Market participants realized that competition in the trading space is not much use unless they also have meaningful choice over where they clear their trades,” says Chan. “As it happens, I have always worked for American firms, and the European debate over clearing and settlement has always tended to refer back to the lower costs of clearing and settlement in America. One reason this job appealed to me was because I could see EuroCCP has a valuable business proposition for Europe, and can make a substantial contribution to lowering the costs of clearing in Europe to US levels.” Certainly, no one can dispute her experience or understanding of the actual and prospective clearing and settlement infrastructure of Europe. There is probably nobody on earth who has read and learned more about the subject than Diana Chan, who has followed its expected and unexpected twists and turns extremely closely for much of the last 5 years, initially as a product manager with Citi in New York (1999-2004) and later Paris (2004-2005), where she covered the Euronext markets for the giant American bank. Throughout 2006 and 2007, Chan effectively ran the relationship between Citi and the European regulators – including the European Central Bank (ECB), for which she coauthored a paper on the custody industry in Europe – on securities services issues. As events conspired to persuade the ECB to launch T2S in July 2006, and the European Commission to publish its Code of Conduct for European Clearing and Settlement in November the same year, Chan found herself itching to become a player rather than a spectator.
T2S is still some way off – the scheduled completion date is 2013 – but when it arrives, EuroCCP will hope to find itself operating in an environment that’s closer to its American parent: multiple trading platforms supported by a single CCP and a central securities settlement platform. But until then, it is the Code that is having a more powerful impact on the daily life of Diana Chan and EuroCCP. In common with all the stock exchanges, CSDs, international CSDs (ICSDs) and CCPs active in Europe, EuroCCP signed the document, committing it to price transparency and “interoperability” with other service providers. In theory, the Code is a godsend for a newcomer such as EuroCCP, since it aims to give market participants a choice in where to clear their trades. With full interoperability between CCPs, users of EuroCCP would be free to trade with counterparties using other CCPs. In practice, however, CCPs are finding it hard to persuade trading platforms to give them access to the trades and even harder to convince incumbent CCPs to put an interoperability link in place. Although a number of links are in place between CCPs in Europe, they are cooperative in nature, and aim only to enable members of one CCP to net trades against members of another without the expense of full membership. Incumbent CCPs have endorsed links of this kind because they can continue to intermediate the trades of their own members, and even attract some new business, while market participants like the opportunity to widen their range of trading counterparties without having to pledge collateral in more than one place. The goal set by the Code of Conduct – namely, to transform links into a means of competition for business between CCPs by obliging them to “interoperate” with each other – has proved much less popular.
“The cooperative links are the infrastructural equivalent of a potluck party, where everybody brings a dish, and you all have a good time because there is something for everybody to enjoy,” jokes Chan. “Competitive links are like inviting yourself to eat someone else’s lunch.” Yet the demand for interoperability was not the invention of officials at the European Commission. It was driven by complaints from market participants that the cost of clearing was too high, especially in vertically integrated markets where the trading platform also owns the CCP.
“The whole interoperability discussion began because trading firms wanted a lower cost of clearing,” explains Chan. “Where a trading venue has only one provider, which is usually owned by the trading venue, the clearing costs are found to be too high by the trading firms. They want choice, and that was embedded in the Code of Conduct, one of whose objectives is to enable trading firms to select a CCP that can provide the same or a better service at a cheaper price.” It seems that objective is not being achieved. In a recent report to the Economic and Financial Affairs Council (EcoFin) of the European Union, European Commission officials noted that no less than 82 requests for interoperability had been filed by various parts of the European securities trading, clearing and settlement infrastructure. Of these, just one had resulted in an agreement to interoperate. In March, LCH.Clearnet agreed to interoperate with SIS xclear in the clearing of trades on the London Stock Exchange (LSE). Even that grudging success is now in doubt, following the merger between the LSE and Borsa Italiana, which comes with its own CCP, in the shape of CC&G.
This is the nub of the problem. Most trading venues in Europe either own their own CCP, or have close links with a particular CCP. The reason why incumbents resist the idea of interoperability is the same reason as challengers request it: Both believe it will lead to an outflow of business from the incumbent. “A highpriced CCP has a limited incentive to open up its client base and its business to a lowpriced competitor,” says Chan. The fact that doing so also imposes a cost, in the shape of the necessary technology investment to build the interface, makes the idea even less attractive. So far, this negative incentive has proved more powerful than the voluntary obligations established by the Code. Even European Internal Market Commissioner Charlie McCreevy, who introduced the Code, has criticized the “endless footdragging” that has greeted its provisions. Some observers have concluded that the Code has failed already. It is not necessary to agree with them to ask whether interoperability is easier to conceive than to implement. While the theory fulfills the dream of trading houses to trade and clear wherever they wish, while CCPs net their trades into a single collateral call and settlement instruction, the commercial barrier to making a reality of it – why would any CCP open its business up to a competitor? – is high.
Chan adds that the practicalities are also complex. If an interoperability agreement was in place between EuroCCP and Eurex Clearing, for example, a brokerdealer could in theory trade on Deutsche Boerse as well as Turquoise and SmartPool, and still clear all its trades through EuroCCP, even if its counterparty chose to clear through Eurex Clearing. But that is where the posttrade problems begin. “Each CCP would have only one side of the trade,” explains Chan. “Each CCP would guarantee to its own participants that the trade will actually settle at the price agreed, and each CCP would guarantee the same to the other CCP. In other words, each CCP would look to the other CCP to make good if its participant defaulted. The question then is: What happens if the other CCP defaults? Ultimately, one CCP is exposed to the possible failure of another CCP, and how one CCP can assess the financial condition of the other CCPs it interoperates with is not an easy question to answer.” These bilateral issues are multiplied in line with the number of CCPs used, entailing as many counterparties and collateral pledges and settlement instructions as there are CCPs involved. With eight CCPs in Western Europe alone, such intricacies are likely to make nobody but a lawyer happy. It is also hard to envisage CCPs cheerfully sharing detailed financial information with one another, or being prepared to rely on assurances from foreign regulators.
Yet the obvious alternative to interoperability – consolidation of CCPs – is not an appealing one either. It would take the debate back to where it was in 2003 when LCH merged with Clearnet in what many saw at the time as the first step toward a single European CCP akin to the DTCCowned National Securities Clearing Corporation (NSCC) in the United States. If being the NSCC for Europe is where EuroCCP would like to end up, it is certainly not where it is starting from now. “We are speaking to all the trading platforms that are being set up now, and being a CCP to MTFs has to be our initial target,” says Chan. Unlike the longerestablished CCPs linked to traditional stock exchanges, most of the postMiFID crop of trading platforms do not have an equity stake in the CCP they have appointed. This has freed them to make economic decisions in a way that is not open to vertically integrated silos. “Those MTFs that do not have equity stakes in other CCPs are more likely to look seriously at us, and we believe we have an excellent proposition for them,” says Chan. “Once we have the MTFs as clients, we will see which of the established exchanges will find it beneficial for them to focus on trading, and let common, low cost clearing be one of their strategies to succeed in the competition with MTFs to win trading liquidity.” This is an intriguing idea: using commercial incentives to break up the vertical silos that are presently inhibiting interoperability.
“If a platform which owns a CCP was to look seriously at its top and bottom lines, it should come to the conclusion that, if they cannot clear at the same cost as a thirdparty provider, their bottom line would improve if they did not do clearing,” explains Chan. “It was the view that the trading platforms in the US, albeit at a different time and in a different space, came to. They concluded that their competence lay in trading, and they were happy to let someone else handle the clearing.” In her view, the established exchanges are mistaken to cling to clearing. They are already coming under pressure from the MTFs at the trading level, and exiting the clearing business is one way for them to cut costs. As she points out, in the US market, none of the trading platforms – large or small – is in the clearing business anymore. It is handled by an organization (NSCC) that is now clearing at a price equivalent to onethird of one eurocent for each “side” of a trade.
Since EuroCCP can make use of the NSCC investment in its mainframe computers, and its business continuity plans and risk management methodologies, it should be able to drive its price toward the US benchmark once its startup costs have been recovered. This is why Chan is at pains to point out that, although EuroCCP was originally set up in 2001 to support Nasdaq Europe, it would be quite wrong to assume the technology has simply been taken out of mothballs. She says the EuroCCP technology platform had to be completely redone to accommodate seven currencies and 13 markets, and to match the proposed modus operandi of Turquoise and its investment banking backers. “The technology developed to underpin the overseas ambitions of Nasdaq 7 or 8 years ago and the technology needed to meet the requirements of Turquoise today were vastly different,” says Chan.
At the same time, DTCC has used the intervening years to enhance the CCP services it supplies in the US, not least in terms of capacity enlargement and risk management controls. The EuroCCP platform has drawn on that investment. Chan says the risk management applications had to be adapted to the European regulatory environment, which insists the client, and “house” funds available to cover a default are not commingled, but otherwise the requirements remain the same as those used in the US today. “The fundamentals of our approach to risk management are the same in Europe as in the US,” says Chan. “We mark open and failed positions to market, assess the volatility of stocks, monitor illiquid stocks, and take into account continuously the financial situation of individual clearing participants. We also conduct stresstests, running all the open positions through 50 extreme and volatile market scenarios every night, and calculate whether we have sufficient collateral to cover the impact of the worst case scenario. If we do not, we call for additional collateral from the clearing members the following morning.” In fact, Chan says virtually the only aspect of the EuroCCP platform of 2008 that is the same as the EuroCCP platform of 2001 is the name. The technology had to be rethought for the Turquoise RFP process and, after the contract was won, rethought again on the back of extensive input from potential users. “A Turquoise steering committee met regularly with EuroCCP to design not only the trading features but also the operational functionality of the platform, such as processing corporate actions on unsettled positions held overnight inside EuroCCP,” says Chan. The company even had to reapply for its regulatory license, which DTCC had al lowed to lapse. Reauthorization by the Financial Services Authority (FSA) was granted in March this year.
So in promising to help trading platforms trim their costs, EuroCCP is launching itself off a completely refurbished base, in regulatory as well as technological and risk management terms. In fact, its main challenge is not the credibility of its own claims, but the fact that the ultimate beneficiaries of those savings in costs – market participants – struggle to make themselves heard, and in some cases do not even want to make themselves heard. “Ours is a peculiar industry because it is the trading platforms which determine which entity is their CCP,” says Chan. “Yet it is the trading firms which pay. Trading platforms may be inclined toward a CCP that is in a position to offer them a share of the profits. Trading firms are also a heterogeneous group. They all have different business models and business priorities. Some of them might have an equity stake in a CCP, and so they would prefer to have that CCP over another one which is cheaper for everybody else.”
Confronted by this melange of perverse incentives, Chan decided to inject some arithmetic into the debate. EuroCCP took the annual reports of its rivals and the pricing information released under the Code of Conduct – that much maligned document has in fact had great success in forcing exchanges, CCPs and CSDs to publish their pricing schedules – and analyzed both sets of information to see if it could work out which European CCP is the cheapest of them all. That EuroCCP found it was the cheapest provider, with an anticipated average cost of 2.9 eurocents per “side” against an average of 26 eurocents per “side” currently paid by trading firms to the incumbent providers of clearing services in Europe, was not a surprise. But the scale of the variations among providers, and the implied cost of the vertical arrangements, was. When the differentials are multiplied by an estimated average of 3 million equity trades a day in European markets, it suggests that trading firms are paying nearly €1.5 million a day more for clearing than they need to, or an aggregate of €350 million extra per year. The analysis, albeit based on assumptions that rivals have questioned, also yielded massive differentials in the costs of using different CCPs in Europe (see Table 1). Those who dispute the figures are invited by Chan to check the assumptions – and they are all published, in immense detail, in a 30page document entitled The Clearing Industry in Europe: Cost Comparison, which has no less than 47 source citations – and come up with different numbers. “Since we have made all our assumptions very explicit, we would like CCPs who disagree with the unit costs we have devised to check our assumptions and advise us on the right methodology to derive a more accurate number,” says Chan. So far, no CCP has taken up her offer. If the document is undeniably among the first fruits of the Code of Conduct, it also illuminates the limitations of transparency as a guide to action. “We have shown that transparency does not mean clarity,” says Chan. “Nor does transparency mean comparability.”
Such boldness – effectively pointing out that incumbents are profiteering and especially those in vertical silos – is way beyond the controversy threshold of most people working in most lines of business. But then Diana Chan is not like most people. Born in Guangzhou to an educated family persecuted as “class enemies” throughout the Maoist period, she saw her father jailed during the Cultural Revolution.
She was a toddler when her parents arranged to smuggle her to Hong Kong in a fishing boat. Her brother – who was exiled to the countryside at the age of 14 to “learn from the peasants” – later made the same trip by swimming for five hours across the South China Sea. “I hit the jackpot with my ancestors – intellectuals, capitalists and landowners,” says Chan. “That was why my parents had me smuggled out because they could not see a future for me in China.” Seeking a better life in Hong Kong meant leaving her family behind, and visiting them only when it was safe to do so, but she has rewarded her parents richly for the sacrifice they made. Diana Chan attended school and university in Hong Kong. Upon graduation with a degree in social sciences, her first job was at a computer service bureau, selling multimedia training applications for computer programmers. It taught her only to recognize that the next 2 years would be better spent at the Harvard Business School. Armed with an MBA, she began to look for work on Main Street as well as Wall Street. Turned down by General Foods for a marketing role, Chan joined Morgan Guaranty Trust Company of New York on its graduate trainee program. She moved on quickly. After just 18 months in New York, she was posted to Euroclear, where JPMorgan then still held the operating contract at the Brusselsbased ICSD. “I thought I would be there for a short time, but it ended up being 11 years,” says Chan. Starting as a financial controller, she moved into strategy and product management, before returning to JPMorgan proper as global custody network manager for Asia, based in Singapore. When JPMorgan sold its global custody business to Bank of New York the following year, Chan did not stay long. After a year, she joined Citi in Singapore in 1997. The rest is recent history. But it is a life story which shows why Diana Chan is not afraid of the truth, even if it is painful for some.
“Our vision is to be the CCP of choice for Europe,” says Chan. “We believe that every trading platform has to make strategic decisions about its combined cost of trading and clearing, and those which opt for the lowest cost clearing solution that is compatible with being safe, robust and scalable are making a wise decision. If we are able to deliver, and be responsive to the needs of market participants and the trading platforms that we clear for, we offer a way for Europe to achieve what it needs to achieve without going through the pains of either interoperability or consolidation. All Europe has to do is choose the most robust and reasonably priced provider.” EuroCCPparent DTCC has currently set the benchmark for “reasonably priced” CCP services at an average of 2.9 cents a side, when its own calculations suggest its most expensive competitor is charging 55 cents a side, and its cheapest, 7 cents a side. EuroCCP has committed itself to a ceiling of 6 cents a side. Like its parent, EuroCCP is run “at cost,” which means excess revenues will be returned to participants in the form of refunds or fee reductions, as determined by its board of directors that is composed of mainly participants and several public interest directors. As coprovider with LCH.Clearnet to SmartPool – EuroCCP covers the nonEuronext markets, and LCH.Clearnet the Euronext markets – the company has a chance to show market participants how it can compete sidebyside with an incumbent. There are of course plenty of hidden costs in clearing, too. Some CCPs charge annual membership fees. Acceptable forms of collateral vary widely, as do the rates of interest earned on cash collateral, and some CCPs charge connectivity fees on top of transaction charges. EuroCCP suffers from some of these limitations too, but unlike its European rivals it can point to an alternative it owns already: NSCC. In fact, if EuroCCP is a success, there is obvious scope for setting up transatlantic clearing arrangements between EuroCCP in London and NSCC in New York. Chan says such arrangements would be subject to regulatory approval on both sides of the Atlantic, but there are two major transatlantic stock exchange groupings in existence already, and discussions are in hand about crosslistings and other forms of cooperation. “If regulators can reach agreement on the trading layer, the clearing layer comes immediately after,” says Chan. “Discussion at the policy level has not reached down to the clearing layer yet.”
Besides, Chan faces challenges enough to create workable models for the fragmented European marketplace of today. “In the US, trades cleared and netted through NSCC drop directly into DTCC’s depository subsidiary, whereas in Europe we have to drop them into one of the 13 national CSDs,” explains Chan. “In Europe, the number of CSDs we have to interface with may eventually rise to more than 20.” To service Turquoise, EuroCCP came up with the ingenious solution of working with an agent bank. After running an RFP process of its own, EuroCCP appointed Citi to act as settlement and assetservicing agent, recognizing that working with an intermediary present in all 13 markets would facilitate access to multiple CSDs. EuroCCP is working with Citi to support the SmartPool platform as well. EuroCCP’s use of Citi is on a nonexclusive basis that permits trading and clearing firms to use other settlement agents if they choose. It is a model capable of even wider application, in which agent banks can handle the procedural and scheduling complexities of settling trades in multiple markets, while EuroCCP fulfills its aim of cutting settlement as well as clearing costs by netting all transactions on multiple trading platforms into one settlement per stock traded per member firm every day. But at present the economies created by this are inevitably limited to the trading platforms supported by EuroCCP, or indeed any other CCP. Unless Europe makes a reality of interoperability, or the CCPs of Europe consolidate by merger or competitive attrition, market participants are likely to continue to be denied for some time the benefits of single margin calls and single net settlements. But if Diana Chan has anything to do with it, they will certainly not have to wait forever.