Most people are focussed on the need to move to clearing, but the elephant in the room is the change in margin requirements for trades which won’t be eligible for clearing. Dividing Rates products into the possible and improbable we begin to see the challenge ahead.
Vanilla trades (clearable now, or maybe in the short to medium term)
- Rate Swaps – cleared already with wide scope at SwapClear, CME, SGX and others
- FRAs – SwapClear since end of 2011
- Swaptions – not cleared yet, but European swaptions may be on someone’s RADAR (although I have no inside information)
- Caps, Floors – not cleared yet, but would make sense to launch alongside Swaptions
- Cross Currency Swaps – should be clearable, but for the difficulty of managing the settlement of the principle exchanges
Non-vanilla (less likely to clear)
- Total Return Swaps
- Range accruals
- IRS in-arrears
- Constant maturity swaps
- Quanto (float float) rate swaps
- Caps or Floors with ratchets or other path dependant options
- American or Bermudan Swaptions
- Target Redemption Notes (TARNs)
- Cross currency swaptions
The reason that IRS and FRAs are easily clearable is they meet the standards at CCPs for risk management, including liquidity & independent pricing, and suitability for default management. Products such as Swaptions take clearing into uncharted territory as they include non-linear risk from their optionality, and so need greater consideration for margining, to ensure the waterfall of protection at a CCP is still safe from the non-linear behaviour of those trades.
As we move down the list, the products become lower volume, and contain complex risk or settlement characteristics which are unlikely to meet any CCPs criteria for being accepted, much less that of ESMA or the CFTC.
The guidance from the CFTC and ESMA is that the margin on non-cleared trades should be “comparable” to that of cleared products. As things stand, for rates products CCPs such as SwapClear and CME have adopted a historic simulation VaR to derive the Initial Margin, and may move to using a Monte Carlo VaR for options products if it can be demonstrated to produce suitable margin figures for IM and the Default / Guarantee fund.
VaR on exotics
For non-clearable trades in the current market, most are subject to an ISDA Credit Support Annex, requiring a daily mark-to-market and margin call on the net MTM of the portfolio in-scope of the CSA. Most Collateral Management teams have had years of practice at gathering, calculating and calling the portfolio valuation (or variation margin in CCP speak), and moving cash or bonds to cover the liabilities. ISDA are busy working towards a standardised CSA to align the currency of OTC trades with the cash used to fund the margin deposits – reducing the number of currencies for margin deposits down to the four majors, USD, GBP, EUR, JPY.
The new regulations will require firms to update their CSAs (a bulk legal challenge in itself), to include an add-on, referred to as an Independent Amount (IA) equivalent to the Initial Margin called by a CCP.The interesting part, is that the portfolio of not-clearable trades are the illiquid and unusual products, when regarded as a standalone portfolio will present a risk profile of non-linear payouts. (If these products are hedged using vanilla trades, they will be outside the scope of the CSA as they will be cleared). The output of a VaR model on these trades alone will challenge the assumptions in each firms VaR models on the distribution of returns, by being less evenly distributed, so leading to fat tails, black swans, or even black swans with fat tails.
A typical bank VaR model will be calibrated for a 95% confidence level with a one day holding period (for example see page 87 of this Goldmans 10k filing here: https://www.goldmansachs.com/investor-relations/financials/current/10k/2010-10-k.pdf). By comparison, the same model at SwapClear is a 100% confidence level, with a 5 day holding period, and for client accounts a 7 day holding period – dramatically more cautious.
CCPs avoid one problem with margin agreements – you can’t dispute the CCP margin call (or you can, but they’ll put you in default). The reason being is that the CCP needs to see guaranteed settlement on trades and margin calls, as another indication of the health of a firm – and given the liquid nature of products, don’t expect there to be wide variance between themselves and their members on trade valuations.In the bilateral world, the ISDA CSA contains procedures to help guide towards the resolution of a dispute, but doesn’t prevent any firm rejecting the valuation of trades, the population of a portfolio, and in the case of complex VaR based IA the underlying calculations to that amount. With Collateral Management teams in the future being responsible for communicating margin calls based on VaR, their ability to explain the background to a margin call – and if necessary reconcile the inputs, scenarios and modelling of their VaR implementation becomes a real challenge. I can see the market risk and collateral management teams spending long hours together in the coming months to provide their firms with the right tools, to support client facing communication on the new margin calculations.
At the moment we don’t have any clear rules or specifications from the US or EU regulators, only indications that to level the OTC playing field between cleared and not-clearable products, the margin levels should be “comparable” – this can only mean a huge amount of work to implement the rules once they are made, and obviously a gigantic sucking sound as firms hoover up cash or eligible securities to post against their new margin requirements.
And to make things worse – most CSAs are cross-asset, so will include the exotics from rates, credit equities, fx and commodities products into the mix.
As the Scouts say “be prepared”.
Bill Hodgson is a thought leader in the capital markets space, initially working on developing software to process OTC derivatives at Merrill Lynch, in the days when paper trade tickets were still the norm. Subsequently he has worked with major banks to improve their OTC processing capabilities, and then with LCH.Clearnet and DTCC to develop the market infrastructure for OTC products. Bill’s website and blog are a great source of further information please visit.