Industry confronts hard choices in CCP recovery debate

It is often said that learning to recover from mistakes is more important than trying not to make any at all, but critics of swap market clearing houses claim the sector has got it the wrong way round – investing heavily in the infrastructure required to clear swaps, while the resources and practices that would set them back on their feet after a heavy loss remain scarily inadequate.

At the core of the issue is how CCPs would recover from the default of one or more of their clearing members without becoming insolvent. With an eye on that doomsday scenario, regulators also want robust resolution plans in place should a CCP fail altogether.

On October 15, the Committee on Payments and Market Infrastructures and the International Organization of Securities Commissions (CPMI-Iosco) published a final report on the recovery of financial market infrastructures (FMIs). On the same day, the Financial Stability Board (FSB) issued a more wide-ranging report on resolution planning.

Both papers largely take the form of suggested mechanisms rather than definitive standards, which caused frustration among some participants, while also highlighting the fact that CCP risk is still a topic for consideration and debate rather than final conclusions.

The European Commission is understood to have delayed its consultation on recovery and resolution planning for FMIs until next year, while no timeline has yet been set in the US.

In my opinion, resolution is not really necessary

As regulators prepare to tackle the issue, dealers and CCPs have been airing strong opinions about the range of options, as well as the central question of whether they should be pursued in recovery or resolution mode.

"We would suggest that closure of a large, systemic CCP would be more disruptive and costly than recovery. In my opinion, resolution is not really necessary. If it becomes obvious that a CCP's standard default management procedures aren't working – that is, an auction has failed, or it can't find hedges – even then, we would like to see last resort options like forced tear-ups pursued within the recovery phase, as CME already has in its rule book," says Ulrich Karl, director of CCPs at HSBC in London.

JP Morgan, however, counters that measures undertaken with recovery in mind cannot be relied upon alone. "Up until now, the best resolution options have been seen as partial or full tear-ups, complete liquidation, or porting to a new CCP. Those aren't bad ideas in their own right, but we believe CCPs won't be able to liquidate or port positions in 24–48 hours without causing extreme further chaos in the markets," says Emily Portney, global head of agency clearing and collateral management and execution at the bank in New York.

The best way of doing that, argues Portney, is to ensure the CCP has a fully funded default fund with which to continue operating – an idea that appears in a JP Morgan white paper on CCP resolution, co-authored by Portney and published on September 11. It suggests a CCP should hold a separate, ring-fenced recapitalisation fund, which it and its members would contribute to on top of their default fund contributions. CCPs responded that the proposal would make clearing too expensive.

But few dispute that the failure of a large clearing house would lead to a crisis even bloodier than that of 2008, so any suggestion that CCPs need more resources to withstand a member default should be taken seriously. As market participants continue to debate the issue, Risk looks at the options available.

1. Assessment powers

For most CCPs, the simplest way to recover from a member default that consumes the default fund is to replenish it, through top-ups from surviving members. So-called assessment powers give CCPs the right to call on members to provide additional resources in a crisis situation, but the idea is not universally popular.

Some critics insist CCPs should not be able to count non-funded contributions as part of their default waterfalls, saying the mechanics around the exercise of CCP assessment powers renders them inherently unreliable. "Assessments will be called upon at the worst possible time, probably simultaneously with other CCPs. It's a fallacy they will be available," says JP Morgan's Portney.

The use of assessment powers varies between CCPs, but almost all rely on them to some degree. LCH.Clearnet, for example, requires the replenishment of default fund contributions by members to cover a maximum of three defaults in six months, with the last top-up being voluntary. But if members fail to meet any mandatory contributions, the CCP may declare them to be in default.

Unlimited top-ups to the default fund are ruled out under the Basel III framework, since for a CCP to receive qualifying (QCCP) status, dealers cannot have unlimited exposures to it. If they do, they lose the crucial right to apply a 2% risk-weighting to all exposures to that CCP. That has been acknowledged in both the US and European versions of Basel III, and the European Market Infrastructure Regulation (Emir).

Some CCPs have already taken note, hardwiring top-up caps into their rule books in recent months.

"We actually pointed out to several Asian CCP operators that they wouldn't be able to achieve QCCP recognition from Esma unless they added a cap to their default fund top-ups," says a senior risk manager at one US bank.

2. Pre-funded recapitalisation

The most controversial idea in JP Morgan's white paper, the creation of a ring-fenced fund to recapitalise a CCP, could effectively require a doubling of current contributions from clearing members and CCPs themselves.

According to the paper, the size of the recap fund should relate to the results of rigorous, regulatory-driven stress-testing, which JP Morgan argues should dictate the amount of capital CCPs should hold. "As with default fund contributions, we think recapitalisation funding should be in place upfront and that the CCPs themselves should contribute more," says Portney.

Reaction from many market participants has focused on the trade-off between staving off comprehensive recovery methods, such as forced tear-ups, versus the inevitable rise in clearing fees that would result if CCPs were required to double the amount of capital they contribute to cover losses.

But even where reactions have been critical, many dealers believe the concept of a CCP pre-funding its default waterfall commitments is a sound one.

"For a CCP that has sustained losses, replenishing the skin-in-the-game capital contribution relies on them being able to raise equity finance. I'm not sure how easy that would be in a crisis situation. If a CCP were in trouble, the operator's share price would already be falling sharply; raising new capital would also be difficult, since investors and underwriters would be extremely stressed too," says Eric Litvack, head of regulatory strategy at Societe Generale Corporate & Investment Banking in Paris.

For their part, some CCPs argue pre-funded recapitalisation would alter the economics of clearing so drastically that it could compromise the stability of the system.

"Recapitalisation funds would essentially be an inefficient use of capital for clearing members in an environment where capital is already increasingly constrained, especially for clearing members that face a large number of CCPs. This type of requirement would likely result in a decrease in the number of clearing members and an increase in concentration," says Suzanne Sprague, executive director of collateral and risk at CME Clearing in Chicago.

Both CPMI-Iosco and the FSB omitted any discussion of a recapitalisation fund from their proposals, but the idea of transferring a stable CCP to a new ownership structure after a critical failure is sketched out by the FSB.

"Resolution authorities should have the power, subject to legal safeguards for counterparties relating, in particular, to netting sets and collateral arrangements, to transfer to a third-party purchaser or bridge institution the ownership of an FMI or all or part of an FMI's critical operations (for example, clearing in one specific product)," the paper states.

3. Variation margin haircuts

Variation margin haircutting involves covering losses that exceed the size of a CCP's default fund using the cash collateral of non-defaulting members.

Proponents argue the method mirrors what would happen were the CCP to enter into insolvency – when members would face the bankrupt CCP as creditors – but that it does so in a compressed timeframe and with the involvement of fewer lawyers.

"The good thing about variation margin gains haircutting is it's comprehensive. Even if the default fund was bigger, or it was underwritten, none of those solutions are comprehensive. There's a cap on all of them. The total losses on cleared derivatives positions could exceed all the resources you can throw at them. You still need a tool that's comprehensive, and can cover all potential losses. The other solutions available to CCPs don't mean they avoid using variation margin haircutting, in the very worst instance; they just push it further into the tail, and at a cost," says HSBC's Karl.

But detractors argue the method is both arbitrary and too big a capital exposure to account for. "Variation margin haircutting is practical, but it's effectively an unlimited exposure, which Emir was supposed to get rid of. Equally, why should the winning side of the trade be the one to suffer the haircut? Their position at that CCP may be offsetting a losing position somewhere else. I would rather see further optional recapitalisations than an exhaustive variation margin haircut, which would be arbitrary and unfair," says the US bank's senior risk manager.

But CPMI-Iosco appears to support this approach, noting in the October paper that "variation margin haircutting represents a measurable and controllable exposure within statistical confidence levels".

"You're very, very unlikely to exhaust variation margin haircutting, even if it's difficult to prove mathematically," agrees a regulatory source at one European dealer. "If a defaulter stops meeting variation margin calls, all it would mean is everybody on the other side of the market getting, in effect, a haircut of 100%. Obviously, you would then need some additional funds to move the defaulter's portfolio, but that should be provided for via initial margin."

As an additional benefit, proponents argue, dealers that know they are likely to face a sizeable haircut are likely to attempt to close out positions and reduce their risk – arming the CCP with precisely the trades it needs to hedge the defaulter's portfolio.

But CPMI-Iosco acknowledges the flaws in variation margin haircutting, concluding that it does not necessarily allocate losses to those best able to cope with them. "The burden of variation margin haircutting may fall more heavily on those with directional positions, which may tend to be end-users, than on those with balanced positions," the paper states.

4. Initial margin haircuts

If variation margin haircutting is accused of being arbitrary, then applying a haircut to a CCP's entire margin pool, which is by far its largest financial resource, would seem beyond the pale. But such an approach has found support in some quarters.

"We think it would be fairer to take variation margin and initial margin and haircut the total across the board, and have a uniform distribution of losses that way. Using total equity is consistent with futures and general bankruptcy treatment. If, when you're in the money, you have at the back of your mind that you get those winning positions haircut, you're much less likely to see it as an incentive to keep your trades on in a CCP," says Tracey Jordal, senior counsel for trading and markets at Pimco in New York.

Others argue that, in the case of either variation or initial margin haircutting, doing the latter alone is likely to be impractical. "For some products, initial margin haircutting would be difficult to implement, and may not give enough proceeds. I also think there's an incentive question: if there was a problem at the CCP, everybody has the risk of getting haircut. So they will reduce their risk, and those facing potential defaulters will be competing with the CCP for hedges," says the European bank's senior risk manager.

5. Forced tear-ups

Widely seen as the nuclear option for a CCP seeking to re-establish a matched book, the forced cancellation of a contract is not a scenario any CCP takes lightly. It is, as CPMI-Iosco notes, "equivalent in its effect to closure or wind-down of the CCP, albeit in a manner which would allow a restart should participants wish".

CME still has the option for tear-ups as a last-resort recovery mechanism in its rule book, and rival LCH.Clearnet confirms it is able to close out contracts at the bottom of the waterfall.

The idea of cancelling some or all of a CCP's contracts and forcing counterparties to face one another in the bilateral market is designed primarily to stave off bankruptcy, wherein clearing participants could face the prospect of having their margin and default fund contributions tied up in lengthy insolvency proceedings.

But full or partial tear-ups would cause significant disruption to the products or markets where contracts are terminated, CPMI-Iosco warns. "Participants may consider termination and wind-down to be the least bad option in some cases, but they might also consider it unacceptable for the trades not to proceed to the original termination or settlement date, particularly if their risk management relied upon derivatives positions," the paper adds.

6. Insurance

When the topic of CCP insurance is raised, the reaction from some market participants is predictably sceptical. "Have you ever known an insurance claim to be paid up promptly, without a fuss, when you need it?" asks a senior derivatives executive at one large bank.

David Hardy, former chief executive of LCH.Clearnet and now chairman of GCSA Capital, has heard these criticisms before. GCSA is spearheading a consortium of insurers offering an insurance underwriting service to major FMIs, either to help cover losses from member defaults, replenish a depleted default fund, or to cover non-default losses.

Collectively, the insurers have claims-paying ability in excess of $300 billion, according to a filing with Iosco from September 2013. The policies can be tailored to a CCP's demands, with the insurers' cover sitting at the front or the back of the waterfall as required. No FMI has yet taken out a policy, however.

"We're aiming to offer diversification of loss absorption. Default funds around the world currently tie up around $35 billion in capital, with the aim of covering a catastrophic tail risk. That's precisely the kind of cover insurers are suited to augment or replace. Right now, CCPs are underwritten by their members – and the CCP has the ultimate insurance policy of being able to, potentially, tear-up all of its members' obligations or haircut margins. That somewhat negates the point of moving products into clearing in the first place, from the perspective of end-clients," says Hardy.

GCSA is understood to have pitched its offering to all of the largest global CCPs, but Hardy declines to comment on customer discussions, citing non-disclosure agreements. According to market sources, however, the firm is close to agreeing a deal with one major operator, which could spur other CCPs to follow suit.

7. State bail-out

For all the FSB's post-crisis work on ending too-big-to-fail across every part of the financial system, many participants believe that, when push comes to shove, systemically important CCPs would still never be allowed to default, even if it involved taxpayers' money.

"I personally believe the nuclear option is the state stepping in. I can't seriously imagine the Fed letting the CME go under, or the Bank of England with LCH.Clearnet," says a senior risk manager at one global bank.

Some lawyers even claim some kind of state intervention is mooted in the FSB report, which says: "Jurisdictions should have in place appropriate arrangements and powers to provide temporary funding to facilitate resolution and to recover any resulting losses to public funds from the FMI, unsecured creditors (including FMI participants) or, if necessary, participants in the financial system more widely."

8. Rely on existing contributions

Increasingly widespread calls for CCPs to contribute more of their own capital to the default waterfall are based on the belief that CCPs don't have sufficient resources to cope during a crisis. It's a premise not everyone accepts.

"If one is being fair about things, CCPs actually have plenty of capital. The resources available to deal with a default are a CCP's equity, plus the entire guarantee fund," says a senior risk manager at one large hedge fund. "I think the banks are being very disingenuous about this but I'm not sure why."

Others go further, arguing dealers push for CCPs to put more of their own capital into the default waterfall has more to do with lowering the capital burden applied to their own default fund exposures than it does with ensuring a CCP's incentives are aligned with its members.

Speaking at the International Swaps and Derivatives Association's European conference in London in September, the group's chairman, Stephen O'Connor, countered industry concerns by saying LCH.Clearnet and CME "probably" have enough of their own capital in the default waterfall. That assurance may not be enough to assuage the concerns of many dealers.

"Clearing houses are operated as highly leveraged institutions because, frankly, they have very little capital at risk. That means they can compete on risk terms - be it lower margins, accepting a broad range of collateral, or clearing for products that worry us - because they're underwritten by us. Our strong feeling is they need to have as much skin in the game as the largest clearing member, to ensure their incentives are aligned," says the US bank's senior risk manager.