A Brief History of UMR Phase 1
Phase 1 firms had spent 2 – 3 years preparing for the regulation, including the creation of custodial arrangements, legal documentation, margin models, and reconciliation processes. However, what probably surprised the phase 1 firms the most was by how much and how quickly their initial margin requirements grew after go-live.
The dealer community reacted quickly where it could, and within days, nearly all interdealer inflation swaps were being traded on a CCP.
Also, within weeks, interdealer liquidity for Emerging Market NDF’s (Non-Deliverable Forwards) started leaving the bilateral space and also started to move to CCP’s.
The clear (excuse the pun) advantage was the netting effect that clearing provided across counterparties for less directional portfolios. To better understand this effect, consider as an example a simple portfolio of 10y USD Interest Swaps traded bilateral (margin calculated using the ISDA SIMM model) and the same portfolio cleared at CME:
A simple $2bn notional offsetting position can result in nearly 200m Initial Margin whereas the same position netted down at a CCP naturally results in no margin.
This rapid large-scale shift into voluntary clearing of inflation swaps and NDF’s was possible because dealers were already fully set up set up and clearing other products at these CCP’s.
Even with these risk mitigating actions, just 6 months of trading volume resulted in nearly $50bn* of IM posted (remember trades pre phase 1 were ringfenced into legacy CSA’s) across these 20 counterparties.
It was not until Q1 2017, that the first post-trade multilateral optimization run was executed, but the biggest hurdle that most risk managers faced was access to the right data and analytics:
Fast forwarding to Phase 5
Phase 5 has a lot in common with Phase 1 as we are again in uncharted waters – not only because of the number and scale of the counterparties that need to be onboarded by custodians and papered, but also because we are dealing a new risk management and operational paradigm for many of these counterparties.
Whilst the foundations for UMR have been laid by Phase 1 in terms of documentation, margin calculation, and reconciliation processes, with just over 6 months to go, have phase 5 counterparties asked themselves the right questions to avoid massive margin growth after go-live with the potential risk that they lose trading lines with some of their dealers as a result?
What is the $75bn question?
There is no single $75bn question (and the number will be much greater after Phase 5), but as the former head of front office risk management for a phase 1 firm, the questions below are the ones that I believe not enough people are asking:
Have you got agreements in place to shift voluntarily into clearing e.g. Clearing broker arrangements, and Cleared Derivative Execution Agreements (CDEA’s)? Likewise, have you analysed whether the netting effect for certain portfolios makes it more economical to clear as opposed to trade bilaterally?
Are you operationally set up to participate in post trade risk/margin optimisation runs and have you implemented the appropriate data models and messaging infrastructure to do this without manual intervention?
How will your traders determine pre-trade which counterparty to execute with in order to mitigate risk/optimise IM and how will this be captured and documented for MIFID II best execution?
Is your infrastructure comfortable in handling multiple netting sets to ensure ringfenced legacy trades can be unwound when required to avoid a massive IM penalty of putting on an offsetting trade in the new (UMR compliant) netting set?
Other considerations
Initial Margin requirements are dynamic and this can result in liquidity requirements if portfolios move. The 50m initial margin threshold only exacerbates this, making margin requirements non-linear and creating a potentially liquidity cliff edge.
Consider the following example where a counterparty executes USD 4bn ATM-100 1y10y Receiver Swaptions with a Dealer, and the impact of the same trade after rates rally 100bp (something we saw take place in 6 months from the back end of 2018):
The IM computed has been simplified to only include the “delta” component of the ISDA SIMM model, but the cliff-edge effect is clear to see. Delta hedging the option would mitigate the initial margin requirements ONLY if the delta hedge is executed with the same counterparty as the option trade (leading us back to our pre-trade question before).
Finally, consider also that many real money hedge portfolios look very similar and a large market move could result in a large liquidity squeeze. In periods of relative calm, it is often easy to forget the impact this can have, but if you cast your mind back to 2011 when LCH increased haircuts on Italian government bonds (which were one of the cheapest to deliver liquid eligible collateral at the time), you will recall magnitude of collateral calls and the panic that ensued the market at the time.
Wrapping up
Whilst compliance has to be the number one priority, and it is essential to decide on elections such as Triparty Vs Third Party, and “allocated margin flow” vs “greater of margin flow”, the cost of funding excess margin and contingent liquidity requirements will creep up faster than most are imagining.
Try to answer the “right” questions, and if that is not possible due to time, resource, or other constraints, ensure that you consider them in your implementation approach to ensure you can be nimble and not have to “throw away work”.
Finally, seek to ensure that the “dynamic” nature of Initial Margin is well understood, and captured in any liquidity risk analysis conducted on your portfolios.
Please contact us if you would like to understand more