Are CCPs Increasing Risks? Part II

Yesterday's post looked at the big risks in the CCP landscape from simultaneous requirements for mandatory margin, novated contract terms, complex portfolio and collateral valuations, reference data reforms, and default funds.  Today we'll look at the competitive differentiators that potential CCP members, sencond tier swaps dealers, and even end users should evaluate when deciding where to clear their OTC derivative transactions.  Note that each of these three categories will end up with a very different risk profile on these factors because of the influence of intermediation and waterfall effects.

Swaps Clearing Eligibility:  Each CCP will have a different range of eligible derivative transactions that they can clear and margin.  Since the real benefit of clearing derives from netting efficiency, it is important to choose a CCP that matches the scope of the member/end user portfolio as nearly as possible.  In a bilateral netting, the netting efficiency will be 100 per cent between the parties (excluding any independent amounts).  As CCPs will only accept "standardised" swaps for clearing, the netting benefit will be less - perhaps much less if there are more exotics in the portfolio.  Some of the key implications are:

•  The demand for initial margin will be much greater, especially from end users.  The CCP will require initial margin on the swaps being cleared at the CCP, and the swap dealer will also require initial margin for swaps which remain bilateral.  This will squeeze end users who are not natural holders of margin-eligible asset classes.  Indeed, corporates don't tend to hold large, deep pools of securities at all.  Those end users who do hold deep pools of high quality securities (that's you pension funds and insurance companies) may find themselves targets of clearing members trying to gain mandates over their assets for their own proprietary liquidity and collateral transformation purposes (e.g., see Lehman Brothers International and MF Global litigation).
  
  
• You could end up posting variation margin to the CCP and/or to the swaps dealer even when you are in the money, increasing your exposure to default and other risks.  Valuation of a bialteral swaps portfolio always produces one net number, positive or negative.  Split that portfolio into two - a CCP cleared portfolio and a residual bilateral cleared portfolio - and you are going to produce two net numbers - one for each subsidiary portfolio.  As you can be in the money on one, and out of the money on another, you can end up paying variation margin that increases your net default credit exposure.

Collateral Eligibility and Haircuts:   Swaps dealers and end users will hold assets which are attractive to them given their individual geography, investment objectives, liquidity needs and business models.  Hedge funds, pensions and insurance companies may have a pool of high quality, liquid securities that will be eligible for initial margin collateral.  Corporates may have just what they need for treasury management.  In the nature of things, end users will hold a lot of less liquid, more difficult to value, equities, corporate debt, municipal debt, money market instruments, etc.  This means there is going to be arbitrage in collateral transformation, with a lot of scope for things to go wrong - badly wrong.  Collateral transformation will also cost fees, require operations for monitoring, etc.  And the top tier firms don't just want your margin assets, they want all your assets to be subject to a pooling and margin arrangement, subject to a mandate allowing them to lend, repo and margin those assets at their discretion, profoundly changing custody expectations and protections.  As a result, members and end users will generally want a CCP that will take the assets they have as margin, minimising transformation costs and risks.

From Betting the Business: Once the dealer gets the securities it has the incentive to re-use them in many different ways to boost its returns beyond the commissions paid by the corporation for the collateral transformation services. As the bottom section of the figure shows, upon receiving the securities from the corporation, the dealer can exchange them into higher yield assets through swaps, sell them outright and invest on a cheaper synthetic through derivatives and get additional leverage. In all these transactions the dealer is taking higher risks, as well as counterparty risk. And remember that when the corporation decides to close the derivatives trade in the CCP, it returns the cash to the dealer, which in turn is obligated to give the securities back to the corporation. Because the investments made by the dealer do not match the securities that have to be returned to the corporation, the dealer is exposing itself to additional risks and can face big problems.

Valuation Models:  Each CCP will use a proprietary methodology for marking cleared swaps to market for purposes of calculating variation margin payments.  As these may be different than the valuation methodology used by swaps dealers as counterparties, and different still from the methodology used by end user hedge funds, institutional investors and corporates (who have different business model objectives than traders), there can be a risk that variation margin payments arise when unexpected.  This will naturally pose liquidity stress and risk as variation margin payments must be met daily, and intial margin demands may be higher than expected.  Swaps dealers and end users will want to evaluate CCP valuation models to ensure that they broadly agree with the principles and methods being applied, and that they can handle the projected volatility and variation that may arise under stress scenarios.  It is worth remembering that CCPs will not exercise the discretion and tolerance that are common in the OTC market when a disputed valuation arises.  The CCP will insist on payment of variation margin when due, and delivery of initial margin when demanded.  This makes the system overall much more fragile and difficult to predict.

Segregation Models:  There are three different segregation models emerging, and even these three are not entirely clear as to how they will work to protect CCP members and end users as the law keeps shifting.  The three models are (a) Net Omnibus Segregation (prevalent UK model with shared client risk); (b) Gross Omnibus Segregation or "Legally Segregated Operationally Comingled" (incoming US model); and (c) Individual Client Account segregation (only being offered from Q2 2012 by Eurex).  I'll go into the details of segregation options in a further post to follow, as it is important to address segregation risks arising in the clearing member as well as in the CCP.

Generalising here, pooling and netting are in the interests of dealers (minimising initial margin at the CCP and so optimising asset liquidity and re-use opportunities) and account level segregation is in the interest of end users (better protection and portability in a default).  Each potential clearing member, second tier swaps dealer and client is going to have make choices between cost, convenience and protection.  MF Global's collapse has left a lot of end users more wary and looking for better protection, so we are likely to see some differentiation as intermediaries self-sort with some offering more old-fashioned fiduciary protections.

Default Guaranty Fund Exposures:  Clearing members will have to contribute to the Default Guaranty Fund of a CCP.  Because losses that deplete the Fund should be uncommon if the CCP does its risk management and margining job right, the risk of a loss that hits members should be seen as a tail risk.  In that case, capitalising that risk will be very expensive and lock up quality capital, depleting market liquidity.  Supervisors are looking at other options which incorporate mutualising the contingent risk among the clearing members in ways that would only hit the members for contributions if losses occur.  How the Default Guarantee Fund is structured, funded, capitalised and administered therefore has serious implications for clearing members' credit, reserve, liquidity and liability management.  It's worth remembering that - just like deposit insurance - the time when a Deposit Guarantee Fund needs top-up from the members is likely to be a time of high stress and low systemic liquidity, complicating any draw on members.  Supervisors are likely to take different views on what is best for the CCP, best for the clearing members, and best for systemic financial stability.

That's enough for today.  Some of the points require much more complex analysis, such as segregation and default guarantee funds, and I'll address this in later posts.  More to follow!

Are CCPs Increasing Risks? Part I

The title of this post is the title of a debate being hosted Post-Trade Forum next month:

Post Trade operational risk: Are Clearing Houses increasing risks?

Gary Wright of B.I.S.S. Research has kindly asked me to give the key note.  I'm also on a panel next week at the RISK Annual Summit on Collateral Trends for Corporates which will cover some of the same ground.

In reflecting on the topic I realised that I have been studying the risk mitigation implications of swaps clearing for nearly 25 years, since the very first swap clearing house was proposed by the Board of Trade Clearing Corporation in the late 1980s while I was on the Settlement Systems Studies Group at the Federal Reserve Bank of New York.  I was co-inventor of the first global, real-time, ISDA-CSA compliant OTC derivatives margin system for Clearstream in the late 1990s.  I've written and thought about the clearing and margining of OTC derivatives for longer than almost any of my contemporaries.

So here's Part 1 of what I'm going to say about the risks of CCPs being rolled out globally later this year:

1. As someone who has supervised and built clearing and settlement systems for a long, long time, let me first observe that the single surest sign of system failure is pre-mature specification before the requirements and regulations are complete.  This always leads to cost overruns, delays and often complete abandonment of a half-completed system.  Since complete regulations have yet to issue in US, UK, EU and Asia, and since all these regulations will be inconsistent to some degree, the risks of project problems for CCPs are very high.  (Disclosure: Granularity has a great track record of problem project remediation so I consider this potentially good for business.)
   
2. CCP clearing will become mandatory for all "standardised" OTC derivatives.  One of the things that isn't yet defined is what "standardised" means in OTC derivatives.  In the broad array of financial instruments where two parties exchange cash flows based on fluctuations in reference indicators, the contract terms which must be common or configurable to be standard has never been completely described.  While standardisation has progressed quite a bit in the last decade with ISDA standard agreements and FpML, especially for products that are designed to clear through Swapclear and other facilities, there is a long way to go in establishing certainty as to which products must and can't be CCP cleared.  A side note to make here is that the regulators will try to force more standardisation by requiring bilateral exchange of intial margin for all swaps dealers and unilateral initial margin for end users for all instruments not cleared through a CCP.  The pressures both ways should create some interesting chaos as the industry grapples with a changed cost base, collateral demands, CSA amendments, tri-party collateral facilities, and CCP integration.  It should be noted that an end user will have increased exposure to his bank/counterparty for any initial margin provided in the event of the bank/counterparty's default, unless a tri-party agent is used for custody.
   
3. CCPs will have to provide daily valuations of cleared derivative instruments and portfolios in calculating variation margin.  This is not as straightforward as it sounds as there is a diversity across the industry in the modelling of valuations for OTC derivatives.  There is scope for some serious game playing in seeking advantage in steering the portfolio valuation methodology toward more accommodative models, especially as liquidity tightens or volatility increases.  Given that there has been tremendous concentration in the ownership of demutualised clearing houses in the past decade, there is a serious risk that owners/incumbents that dominate OTC dealing and clearing house risk committees could rig the system to their benefit.  Perhaps I'm being cynical in thinking this would ever occur to them, but let's face it, fiduciaries are harder to find than predators in the current market climate.
   
   On the operational side, daily valuations of OTC derivatives portfolios are standard for all financial institutions regulated under Basel Capital Accord principles, and for many investment banks who need to tightly manage their treasury operations.  Daily valuations are not standard for a very broad swathe of hedge fund, institutional and corporate counterparties, some of whom have very modest means and very few OTC derivatives.  Many businesses were forced by their banks to take interest rate swaps as part of a commercial loan origination.  These end-users are not going to benefit from CCP clearing as there will be little scope for netting gains and/or little improvement in their risk profile from CCP credit interposition relative to their bank. 
   
4. Besides the portfolio, the CCPs will have to provide daily valuation of collateral.  One of the challenges in building the Global Credit Support Service in the late 1990s was providing the valuation methodology for 54 currencies and over 180,000 securities in a manner which would allow the users to have real-time transparency of margin positions.  While systems have come a long way since then, collateral valuation remains a non-trivial challenge, not least because few models accurately capture the volatility and liquidity conditions which follow a default.  Because valuation is difficult, particularly for less liquid securities and currencies, the CCPs will tend to require high quality assets as initial margin.  Even there, we have seen huge volatility for sovereign debt in the past few years, and there is little sign that this will diminish going forward.  If a CCP is too restrictive on initial margin assets, it will be difficult to attract users, so there will be pressure to take lower quality assets at more generous valuations. 
   
5. Variation margin must be paid in the currency of the underlying swap obligation.  This sounds straightforward, but will prove an operational nightmare.  The practical implication is many times more payment confirmations to be reconciled in many more currencies.  Instead of one net payment in one currency, counterparties will have to track, confirm and reconcile payments in a range of currencies.  The costs and complexity will probably drive renegotiation of many existing derivatives to reduce the operations costs and burden.
   
6. The CCP drive comes in parallel with an initiative to standardise reference data, and in particular Legal Entity Identifiers (LEIs).  The standard has yet to issue though expected in June, so the CFTC is now proposing that US-based CCPs use a CFTC Interim Counterparty Identifier (CICI).  This may sound a small thing, but believe me the standardisation of reference data is a major headache. If it were an easy thing to do, it would have been done decades ago.  While I'm generally in favour of LEIs, the tight time frame for implementing such a radical reform which must then be integrated into multiple existing and new systems to accurately ensure mapping across internal and external reference data is a major technology and operations challenge. 
   
7. Finally (for this installment), CCPs will be subject to supervisor stress testing and scrutiny of Default Guarantee Fund arrangements.  It is expected that owners/members of the CCP will have to absorb some contingent liability for CCP losses as an alternative to the massive capitalisation that would otherwise be required to cover difficult to measure losses on OTC derivative portfolio defaults under stress conditions.  Again this is an area where regulators have yet to provide a detailed picture of what they want or what they will expect or what they will accept as sufficient.  The capital and liability-sharing arrangements will have implications for members as well as CCPs, so could become an element in CCP evaluation and competition. 

More to follow!