With all the real-life stress and volatility that markets experienced over recent months, the European Securities and Markets Authority's (ESMA's) 2019 stress test of clearinghouses (CCPs), released in mid-July 2020, might now seem like little more than a footnote. While in S&P Global Ratings' view this exercise might not have brought any transformational insights, we believe it fulfils a specific role and its findings demand further exploration.

Now in its third iteration, the quality and depth of the testing exercise has steadily improved. Furthermore, ESMA is rare among global regulators in conducting such a public test for CCPs. We believe this is necessary because, while the European Market Infrastructure Regulation (EMIR) sets a common standard across the EU in law, in practice the supervision of CCPs is fragmented among many national competent authorities. Furthermore, even with the overlay of common technical standards, EMIR provides a framework rather than a straightjacket for CCP risk management. While the ESMA exercise is no substitution for CCPs' own stress testing and fire-drills, it provides key insights and comparison. It also brings a timely examination of CCPs' management of concentration risks, which even now remains a lively topic after the September 2018 Nasdaq Clearing member default event.

As in previous years, market attention has focused on CCPs that showed modelled deficits in prefunded resources or other weaknesses. For the counterparty credit stress, this time LMEClear finds itself in the spotlight; last time, it was BME Clearing. But there are other findings to ponder besides this--on many aspects of the study.

Like others before it, the outcome of this stress test was heavily determined by the inputs--chief among them, the market stress assumptions. The scenario was, at first sight, stressful but far from extreme--for example, equity markets were assumed to fall around 15% over five days. However, with its de facto assumption of a "bolt from the blue" episode of severe market price adjustment amid previous sustained low market volatility, it did spur divergent outcomes across CCPs. The choice of a five-day stress and dislocation was also key, since some CCPs assume that they can complete their default management within shorter windows.

For investors, banks, counterparty credit risk analysts, and supervisors alike, these exercises will likely remain insightful. Future editions might well explore other market scenarios, and given the interconnectedness of European markets, would be most useful if they cover the broad swathe of European CCPs, not only those in the EU.

Transparency Improves As The Test Evolves

This is ESMA's third CCP stress test exercise, and with each iteration the scope, sophistication, and transparency of the exercise has grown. The first test was reported in 2016 on a no-name basis, and was concerned only narrowly with counterparty credit risk. ESMA introduced a liquidity test into the second exercise, but reported on a no-name basis. This time, ESMA tested and reported each CCP's credit and liquidity risk outcome, and also introduced a new consideration--concentration risk--that was again reported on a no-name basis.

The quality and sophistication of the exercise is also steadily improving, in our view. For example, the ESRB parameters were more granular than before, the concentration risk assessment was a new and quite thorough analysis, and wrong-way risk was picked up in the assessment of collateral.

As an aside to the main findings, ESMA also includes incidental data on margin collateral and concentration by member. These data confirm our existing view on two matters:

• Collateral quality is generally high in Europe, being centered on cash and government securities with de minimis unsecured investments and non-prefunded resources.
• Member concentration varies significantly across the 16 CCPs. Indeed, while ESMA examined the concentration of the default funds, CCPs' quarterly disclosures on margin concentrations (see chart) tell the same story in a global context--LCH and Eurex Clearing (the main component of the 'DB1' bubble in the chart) have particularly unconcentrated memberships.

Chart 1

Figure 1

Table 1

Table 2

The Test Assumed Stressful, But Not Extreme Market Movements

The market stress scenario set by the ESRB is at the core of the stress test. These parameters guide the valuation of clearing member and client portfolios and collateral held at the CCPs. This in turn is a critical input to the magnitude of any shortfall in financial and liquidity resources, when combined with the quantum of defaulting members. The severity of the market parameter assumptions is critical--ESMA found that the resilience of CCPs' resources is more sensitive to market value assumptions than to the assumption on the number of member defaults.

As ESMA acknowledged, these shocks were below the maximum five-day historical movement observed for a number of asset classes, and in some cases even below the magnitude of the subsequent March 2020 bout of pandemic-induced volatility. Nevertheless, the ESRB scenario did make the tough assumption of a "bolt from the blue" stress amid a period of sustained low volatility through late 2018 and early 2019, whereas in March 2020 volatility spiked, but after a short ramp-up that would have led to an escalation in margin resources at many CCPs.

We note also that the ESRB scenario was designed to reflect a reasonable adverse scenario on a forward-looking (not historic) basis--a plausible market correction that would not necessarily be felt equally across all asset classes (see charts 2a and 2b). As a result, it might not be an equal test across all CCPs (many of which focus on only one or two asset classes).

Chart 2

Chart 2b

If the shocks were significantly less severe than a CCP's own hypothetical or historical stress testing assumptions, the CCP is likely to have shown only modest consumption of its default funds under the ESMA exercise. And the reverse is also true. So, for example, ESMA's assumption of a market stress that lasts five days is likely to have been demanding for CCPs that, when sizing margins and default funds, assume only the minimum permissible close-out period under EMIR--which is just two days for cash equities, bonds, and exchange-traded derivatives contracts (itself more than the one-day assumption used by some U.S. CCPs).

A Useful Adjunct To Actual Stress In March 2020

The ESMA stress test and the period of market volatility in March 2020 both constitute a theoretical assessment of resource adequacy for European CCPs because there was no real-life member default. As we discussed in our April comment, there was one prominent member default recorded by the U.S. CCPs, and none in Europe. CCPs managed the U.S. default without resorting to pre-funded resources other than those of the defaulting member. (See So Far, So Good For Clearinghouses Despite Oil And COVID-19 Market Volatility, April 16, 2020.) As long as clearing members continue to post their margin collateral when called, extreme market volatility does not pose an immediate threat to a CCP.

While the ESMA exercise is undoubtedly useful and the level of assumed market stress was similar to the actual one in March 2020, we see the March episode as a truer test of CCP risk management. Market moves in March led to huge margin swings, demanded the reinforcement of default funds (in some cases), and heavily tested CCPs' operational capacity at a time when many had activated their business continuity plans. It also provided a wealth of data to inform CCPs' future scenario analysis.

Counterparty Credit Risk: Comfortably Resilient, For Most

For a CCP, credit risk arises when a member defaults and there is a simultaneous severe adverse change in market prices (which could be up or down, depending on the member's positions). Under EMIR standards (which are among the toughest globally), this is assessed as a "cover 2" level--where the two largest (that is, the most impactful) clearing members for a CCP are assumed to default, amid a range of hypothetical or historical market stress scenarios. ESMA ran its stress test assuming cover 2, in conjunction with its market stress scenario (see above), selecting:

• The largest members at individual CCPs; and
• The largest members across the entirety of the EU and U.K.

Faced with a defaulted clearing member, the CCP will use available resources in a prescribed order (see figure 2). If all of these paid-in resources are exhausted, most CCPs have the right to use powers of assessment (PoA) to call additional resources to absorb residual losses and to act as a buffer against future losses.

Figure 2

The member default and market risk factor stresses were sufficiently stressful that most, if not all, of the CCPs would have needed to use not only the margins and default fund contributions of the defaulting members, but also some of the default fund contributions of the nondefaulting clearing members. This is unsurprising--after all, these are exactly the sort of extreme events that these pooled resources are intended to cover.

We agree with ESMA's view that the results showed the CCPs to be resilient (albeit subject to the above caveat on the levels of assumed market stress). However, as in ESMA's last exercise, one CCP stood out negatively (see figure 3). This time, the modelling showed deep losses of over €700 million at U.K.-based LMEClear that would have required it to use its PoA (i.e. non-prefunded resources) of €206 million.

Figure 3

We agree with ESMA's view that the CCP could have used its contractual rulebook powers to try to absorb these losses. However, these powers have rarely, if ever, been used by any CCP globally. In addition, given the hugely stressful market conditions under which they would be called, it seems reasonable to be cautious on their efficacy--some members might choose to renege on their contractual commitment if they believe that a CCP is not viable. In our analysis of CCPs, we therefore focus on prefunded resources when we examine the adequacy of a CCP's financial safeguards package--that is, the default waterfall and associated risk management measures.

While most other CCPs show usage of less than half of their default fund, there are two implicit implications.

• First, as the 2018 Nasdaq Clearing default event also showed, clearing members face a real risk of losing some or all of their default fund contributions.
• Second, before the CCP uses its default fund, it will have consumed its skin in the game. This raises the question of how CCPs' recovery plans envisage that they would rebuild this buffer. We would expect this to be more straightforward for CCPs that routinely carry significant excess capital or those that are owned by highly profitable FMI groups, and harder for smaller CCPs that have marginal profitability.

Fresh Insights On Concentration Risk Deserve Further Examination

ESMA introduced the concentration test in response to the NASDAQ Clearing event in September 2018 (see Risk Management Takeaways From The EUR114 Million Nasdaq Clearing Loss, Sept. 27, 2018) where the CCP had to risk-manage the default of a highly concentrated, individual clearing member in its commodities clearing service. This led to a €114 million loss that consumed Nasdaq's skin in the game, and €107 million of mutualized default fund resources (roughly two-thirds of the total default fund).

ESMA's analysis shows in stark light the concentrated nature of clearing in Europe's CCPs--a natural consequence of the aggregation benefit presented to clearing members from channeling their activity through a small number of CCPs, and also the willingness of some members and their end-clients to build concentrated positions. ESMA also found that these concentrations exist in all asset classes.

In order to address associated market risks, CCPs usually apply additional margin requirements on these positions to compensate for the disproportionately higher risk of loss that they bring. ESMA's analysis indeed confirms that the vast majority of CCPs do apply related add-ons. However, there seems to be a significant dispersion in practices and assumptions, with one or more CCPs in each asset class collecting far less concentration margin than ESMA calculated would be necessary. In the worst case, one CCP's small commodities service, the modelling showed that the market impact of liquidating concentrated positions could consume 145% of its total pre-funded margins (see chart 3).

Chart 3

Despite the complexity of the task, we believe that the results provide a good estimate of the concentration risk profile of the tested CCPs. We see the analysis as an important starting point in this evolving area of CCP risk management, not least given the calculated inadequacies even before one factors in any auction risk component.

Satisfactory Liquidity Coverage, But Questions Arise Around Central Bank Access And FX

CCPs' primary liquidity risk arises when the CCP has insufficient cash to meet its payment obligations under its clearing contracts. As with market risk under the counterparty credit stress, until a clearing member defaults, CCPs typically have very limited liquidity risk. The risk analysis therefore starts with the assumption of the simultaneous default of the two most impactful members, and a concurrent market stress that affects the related cashflows. CCPs then model their projected cash in- and outflows through the multi-day default management period, notably assessing not only the adequacy of resources in total, but also the adequacy on each day, and even intraday. This is particularly demanding for multi-currency CCPs and those that clear cash products (such as equities, bonds, or repurchase agreements).

In addition to their basic calculation, ESMA added four stressed assumptions: the removal of excess margins, a market access delay, settlement lag, and no access to spot FX markets. Among them, we consider excess margins to be particularly unreliable in a stress, so focused on the result after this overlay (see Liquidity Matters For Clearinghouses, published March 16, 2017). This is a key assumption for many CCPs since, as ESMA's data shows, excess margins are often 15% or more of total collected margins.

ESMA's analysis nevertheless showed no aggregate liquidity shortfalls over the seven-day stress horizon on an aggregated (fully fungible currencies) basis.

We note, however, two important takeaways from ESMA's analysis:

• A number of CCPs would have met potential shortfalls by accessing committed or uncommitted lines with central banks. This highlights the regulatory advantage that many European CCPs have over their peers in the U.S. and elsewhere that cannot rely on central bank access. However, the terms and availability of central bank access diverge across Europe. Harmonized access to intraday and overnight borrowing and deposit placements would further support the creditworthiness of European CCPs.
• European CCPs sometimes struggle to access foreign currency liquidity resources, in particular reliable access to dollars. These multi-currency CCPs cannot rely on dollar swap line access through their central banks and often do not have direct access to foreign central banks. It is not entirely surprising therefore that ESMA's analysis shows a $320 million deficiency for Eurex Clearing. We consider this deficit noteworthy, though it does not affect our overall view of the CCP's management of clearing and settlement risk--the shortfall is modest, easily addressed through the FX spot market, and it has rulebook powers to deliver alternative currencies (such as euros) if it had to do so.

Where Next For The Exercise?

Given the important insights on concentration risk, we expect that ESMA will continue to assess it, and may feel able to increase the transparency of associated reporting of outcomes. The market shock assumptions are also critical to the exercise, and future scenarios will likely differ from those used this time. They could also explore the effects of irrational or dislocated markets--extreme but plausible scenarios.

The next consideration is Brexit. More than half of the initial margin collected by the 16 CCPs sits in the three U.K.-domiciled CCPs--LCH Ltd., ICE Clear Europe, and LME Clear. The U.K. has already left the EU, but these CCPs will remain highly relevant to EU clients and so of significant interest to EU regulators. The reverse is also true of the major EU-based CCPs. Furthermore, LCH Ltd. maintains interoperable relationships with EU-based EuroCCP and Switzerland-based SIX x-clear. For investors and counterparty credit risk analysts, coordinated exercises across these three jurisdictions would provide further insights into the resilience and interconnectedness of EU and non-EU clearing infrastructure.

ESMA already assesses the potential systemic impact of the failure of the largest banking groups, but this is a purely computational exercise, and it excludes the global exposures of those groups--they likely appear among the largest members of many U.S. and other global CCPs. There are two implications of this:

• The stress test exercise is no substitution for CCPs' own rigorous mathematical stress-testing of resources and their "fire-drills" that seek to test the operational resilience and practical execution of each CCPs' close-out processes; and
• Regulators' coordinated, global "fire-drills" involving the leading CCPs will likely continue to provide important insights into the resilience of interconnected global markets.

Related Research

• So Far, So Good For Clearinghouses Despite Oil And COVID-19 Market Volatility, April 16, 2020
• Clearinghouses Continue To Up Their Risk Management Game, Jan. 29, 2020
• Risk Management Takeaways From The EUR114 Million Nasdaq Clearing Loss, Sept. 27, 2018
• Liquidity Matters For Clearinghouses, March 16, 2017
• Gaining A Deeper View Of Clearinghouse Risk Management, May 20, 2016