Ladies and gentlemen, it is a pleasure to be invited back to speak at the Chicago Federal Reserve third annual conference on CCP Risk management.
Last year, I was humbled, as a mere politician, in a room full of experts, to be ask to contribute on CCP resolution policies
- and to be asked back again means I can’t have done that bad a job.
I certainly learned a lot, sitting listening for the rest of the day, last year, and I hope to do the same this year.
However, this year’s conference on margin models for CCPs is one that the politicians in Europe have been less deeply involved in.
The level one text of EMIR on margin requirements is surprisingly short - only 5 paragraphs which broadly restate the PFMI, and include a mandate for ESMA to specify further details on “appropriate percentage and time horizons for the liquidation period and the calculation of historical volatility... taking into account the objective to limit procyclicality”
Even the recitals of the text where you can normally expect to find stronger motivation on objectives and outcomes required by the politicians are surprisingly brief.
“A CCP should have a sound risk-management framework to manage credit risks, liquidity risks, operational and other risks, including the risks that it bears or poses to other entities as a result of interdependencies. A CCP should have adequate procedures and mechanisms in place to deal with the default of a clearing member. In order to minimise the contagion risk of such a default, the CCP should have in place stringent participation requirements, collect appropriate initial margins, maintain a default fund and other financial resources to cover potential losses.”
It amounts to a list of outcomes that margin models should seek to achieve.
Yet leaves a lot of scope for CCPs to use different models and different methods to achieve those outcomes.
I would argue that this is the right approach.
There has been a political discussion in the European Parliament recently over how much over sight and approval financial supervisors should have over margin models.
Admittedly, it has been more in the context of bilateral contracts, however the same arguments are relevant here.
It is the primary function of a CCP to correctly calculate margin for managing the risk in derivative products,
If regulators become too prescriptive over those margin models then what function are the CCP's management serving?
Yet, we need to balance that with the public policy role that CCPs have been assigned since policy makers came up with the idea of mandating central clearing.
We have given CCPs the status of a Public Good, which means the politics has changed.
Everything they now do becomes a matter of public scrutiny
Including their core function.
In the 2011 film “Margin Call”, we see the drama of Zachery Quinto, at 1 am in the morning, running models on a fictional bank’s balance sheet and discovering that they are entering the limits of their risk models and will not be able to cover their margin calls in the near future.
The idea made popular in the film - which BTW most policy makers will have watched, is that the rest of the banking sector could be brought down by margin calls which were aimed at making the system safer, and the one that motivates the concept of using a CCP in the first place.
Yet if the CCPs models don’t hold up under stress, or exacerbate the crisis in the way that Zachery Quinto was so shocked by, then we are in no better a situation than before the clearing obligation mandated the use of CCPs.
It is hard politically, therefore, not to argue that higher margin requirements and larger default funds, accompanied by detailed rules on the quality of collateral are not the best accompaniment to the clearing obligation.
From a policy perspective why not try to Create CCPs that are fortress- like in their approach to everything.
Yet, clearly there is a balance to be reached.
Trapping liquidity and high quality collateral within a CCP where it cannot work for the wider economy, in order to mitigate a once in a million year scenario also does not seem prudent.
The need to facilitate growth in Europe - and globally, is well known by politicians everywhere.
So this is the balance that we are asking CCPs to maintain.
And part of the purpose of today’s conference is to ask how well they are doing it.
If we don’t ask these questions now and ask experts to scrutinise current practice, we won’t know whether CCPs are successfully walking the balance beam until they fall off.
We are also subjecting them to commercial pressure from their members and the wider market place to reduce margin levels if we are not able to explain empirically, why they should be kept at higher levels.
The ongoing debate between central banks and regulators in the US and EU over anti-procyclicality measures matters.
The level of credibility both sides have is not in dispute - all are trying to act in the public interest as part of their founding mandates, and trying to find the best way to achieve the balance that is so clearly in everyone’s interest.
I look forwards to hearing more from David Murphy later on, concerning his paper, evaluating the five tools he sees as able to mitigate margin procyclicality.
In addition, Given the international work released by the CPSS IOSCO in this area over the summer I will be interested to hear how other international regulators reacted to these different tools.
Beyond that, I hope that the CCPs themselves are reading the different papers that are coming out of the global regulatory community and using them to inform their evolving thinking on more sophisticated margin models.
I am not going to talk about the technical aspects of margin setting, and Margin Periods of Risk, close out periods, confidence levels, or how to calculation Specific Wrong Way Risk.
There are many experts in the room who are much better placed than I to discuss these issues later today.
However, the less technical concerns I have are more around governance and how to keep checks and balances over different margin models going forwards.
Transparency with respect to margin models was hotly discussed when we worked on EMIR.
Some CCPs argued that this was the essence of their business model, so to disclose details about how they margined products to anyone at all would destroy their competitive edge.
They also made a compelling argument that should market participants become aware of too much detail of how margins were calculated they would be able to game the system in order to reduce margin payments.
Yet if clearing members and their clients are unable to see how margins are calculated, how can they fulfil their fiduciary duties to their shareholders and investors to ensure they are not interacting with a more risky entity than they were aware of?
Would they change their behaviour and use different products or combinations of products if they found that they were not comfortable with the margin methods used by a particular CCP?
Beyond that, are certain market participants choosing to leave positions unhedged because they consider the margins currently being required by a CCP are simply too high in comparison to the risk they were hedging?
Without insight into the margin models it’s hard to say.
Perhaps one answer would be for the risk committee of the CCP to have access to more information on margin models from the staff of the CCP.
In Europe, the composition of the risk committees is mandated under law.
There must be representatives of clients and clearing members on those committees and national competent authorities have the right to attend all risk committee meetings.
Perhaps if they had granular access to those margin models and a binding opinion over whether to accept or reject them in the case of new products or material changes to existing products,
then the rest of the market would be able to take this as indicative of confidence in the modelling technique.
Or perhaps they would see it as an unfair advantage for those on the risk committees to gain confidential insight that they are unable or unwilling to take on.
I am wary of too much involvement of supervisors in margin setting
A regulatory stamp of approval should not replace conservative and prudent risk management by CCPs staff.
All of the work we have done on CCP recovery and resolution has cast light upon the incentive structures that cause CCPs to do their jobs so well.
The tensions between the differing interests of clearing members and CCPs management is fundamental to how a CCP functions.
Too much involvement of outside forces could upset this balance.
However, we have to take into account the different business models of CCPs when thinking about these incentive structures.
In my opinion that is why certain areas of the CPSS IOSCO work leaves areas for discretion, and options for regulators and supervisors to take into account.
Not because they can’t reach agreement -
But genuinely because the answer to the question may differ in different circumstances
The different business models of public utility, mutual or shareholder owned entity have all found different ways of keeping then tensions between different parties acting in a positively reinforcing manner.
That diversity between models is of benefit to the market and the wider economy.
Partly because should one model prove less than reliable the whole system will not collapse all at the same time
but also because they may cause different behaviour when reacting to similar outside stimulus.
Taking into account the procyclical, anti-procyclical debate - and Zachary Quinto of course - if the different modelling of CCPs causes a less coordinated reaction to large events then this is a positive thing.
Competition and innovation in margin models is horribly controversial.
EMIR explicitly prohibits CCPs competing on risk grounds
This was the compromise between those who thought we should mandate a public utility model for CCPs within the EU, versus those who thought the commercial entities would provide a better solution.
For those who saw central clearing as a public good, it was hard to accept that profit should form a part of that
But for those who had formed parts of the CESAME group prior to the financial crisis
Or had worked on the best way to break down the post trade Giovanini barriers to a single market in the EU in clearing and settlement,
Any way not based on competition was impossible to accept.
Particularly when we already had several globally successful CCPs operating in Europe.
I think nationalisation would have come up against EU state aid rules - something I'm sure LCH and Eurex would be glad to hear.
In seriousness, setting up a model of based on competing entities came with its own problem-
What were CCPs supposed to compete on the basis of?
Is innovation in margin models competing on risk?
When we negotiated the text, we meant that manipulating the models so as to require less margin n order to steal business from another CCP should not provide the basis of offering a competing product
However, if a more sophisticated margin model is devised that requires less margin for different reasons, how are policy makers to know the difference?
Again this is where more transparency around margin models is of benefit.
The innovation mentioned briefly in the paper produced for this conference, that really interested me, was the idea of looking at margin tailored to individual counterparties.
A client specific margining regime
In an environment where the clearing obligation is bringing more market participants into central clearing
and the backdrop of the leverage ratio under discussion at Basel,
It makes sense that CCPs are looking to how they can meet a broader range of client needs
If they were to be able to use a more granular approach to margining that one assumes would take into account specific credit risk in a more individual manner, would this allow an option for more entities to become direct members of CCPs?
Or would it simply allow clearing members to also provide a more tailor made service to their clients?
Naturally there would be level play field concerns and I'm sure both clearing members as well as supervisors would need to be sure that discrimination wasn't taking place
But this could be the kind of innovation that those of us who favoured competitive models for CCPs would have envisioned.
CCPs offering better customer service gaining more market share - not just the big getting bigger because of an inbuilt size advantage when it comes to managing liquidity.
Last year at this conference, as we discussed different resolution tools for dealing with a major clearing member default, or a cyber attack upon a globally systemically important CCP,
The issue I raised that I wasn't sure had been sufficiently addressed was whether a CCP sufficiently active in one product could be brought down if it was found to have mis-calculated the margin for that product.
As some CCPs have taken a Titanic-like, or fire wall approach to different product lines, with separate default waterfalls for different products,
They are attempting to create separate lines of failure that can be isolated should this happen
they are obviously concerned about this concept too - and are putting in place measures to mitigate against it.
This is why this conference comes at such an important time.
I still believe that innovation is a good thing
And I believe that CCPs are better placed to do the jobs themselves rather than if the public sector took them over.
But given their new function in society,
We have a responsibility to keep casting light and transparency into what they are doing.