Portfolio Conversion of Libor to RFR trades

The closer we get to year end 2021, the more important the question of what will happen to existing LIBOR Swaps when and if LIBOR is no longer published or declared a non-compliant benchmark by the regulator.

One approach is ISDA’s work on new Fallback language in the 2006 Definitions (see Libor Fallbacks: What will the GBP Spread Be?). Another approach is to replace Libor Swaps with RFR Swaps; SONIA in GBP and SOFR in USD. We recently covered a Compression Auctions for RFRs approach proposed by Darrel Duffie and today I will look at TriOptima’s proposed Benchmark Conversion offering.

Background

My first reaction to reading about the Compression Auction proposal was:

  • The proposal is well thought out from a risk & trading point of view but we lack a volunteer host (any takers SEFs/MTFs/CCPs?).
  • One-off auctions are demanding e.g. CCP member default (though that is a very different circumstance)
  • Can’t we arrange a deterministic and automated conversion instead?

TriOptima’s proposal is the first proposal I’ve seen along these lines.

TriOptima Benchmark Conversion for CCP portfolios

The new service converts legacy IBOR trades (e.g. GBP LIBOR Swaps) to RFR trades (e.g. SONIA swaps) within participants defined risk tolerances.  The service is based on and embedded in the existing triReduce line-item compression service.  This enables reuse of the mature triReduce processes for:

  • Sourcing of trades by counterparty from participants
  • Reconciliation of the trades and exclusion of breaking trades from compression
  • Calculation of the portfolio and trade level risk (IR delta grid)
  • Line-item compression in which an algorithm proposes to tear up some of the trades and replace with fewer trades with less gross notional but similar risk
  • Setting of risk tolerances per portfolio risk factor sensitivity (combination of currency, index and tenor point) to control how precisely similar is the risk of replacement trades

The Benchmark Conversion service extends this so that:

  • LIBOR trades than cannot be compressed in the regular compression run can be compressed and converted
  • Replacement trades are created in the RFR index instead of the LIBOR index
  • Risk tolerances instead constrain the outright risk change in LIBOR and RFR index as well as the basis risk between each LIBOR term rate tenor and the alternative RFR.

TriOptima has clarified a few further points:

  • triReduce is already running at the largest CCPs including LCH, CME, JSCC, Eurex, NASDAQ OMX.
  • Non-bank participants have started to participate via 11 of the 12 largest FCM/clearing broker members providing client portfolios (though compression is at client discretion).
  • Several aspects of the Benchmark Compression service have been built, tested and deployed in production.
  • The service does not require sign up to the ISDA Benchmarks Supplement protocol as a prerequisite.
  • The implementation of Benchmark Conversion can be separated by currency e.g. GBP SONIA and USD SOFR are likely to transition sooner than other currencies.

Observations on the CCP portfolio proposal

This new service presents as an extension to an established set of infrastructure with wide existing participation and relatively low development and adoption cost. 

The pipes and plumbing are the same as for regular compression runs, but I would expect participants to face some changes to internal risk management, operations and IT to handle booking RFR trades from triReduce runs and monitoring the cross-risk / basis risk impact between RFR and LIBOR trades.

I wonder whether CCPs will prefer to tackle this conversion themselves.   This can be a big project and could delay their product pipeline.  

Extension to uncleared portfolios

TriOptima expect that the service will also gets used in the uncleared portfolio space.   Of course – along with arranging for conversions – participants can continue to compress uncleared portfolios with associated leverage and operational cost savings.  

Challenges of uncleared conversion

IBOR to RFR conversion in the uncleared world brings several further challenges over the cleared world:

  • Many more portfolios given two parties per portfolio rather than one for a given CCP.
  • Much more diverse participants and participant infrastructure with many more golden copies of the trades (CCPs and leading banks provide all the trades for the CCP version).
  • Some buy side firms will find the cost and project delivery challenge more than they want considering they only accrue leverage benefits (unless banks find another way to incentivize them).
  • Greater product complexity / greater likelihood of triReduce calculated risk being different from the participants view given more complex linear and non-linear product types (e.g. non-clearable IRS, swaptions/caps/floors/exotics) in uncleared portfolios.
  • For uncleared portfolios, BCBS has given a steer that clearing / UMR exempt legacy trades should remain clearing / UMR exempt despite being terminated and replaced by new trades in such a conversion.   However, national regulators still need to confirm this.

All in all, while TriOptima’s approach is a decent start, the above challenges look large indeed.

Conversion vs Auction

In the Compression Auction approach Duffie suggests that CCPs would want to run this as it is in their interest.   This suggestion serves also to make the point that someone needs to volunteer to do it.   No one seems to have stepped forward so far, at least not that I a aware.

One advantage of an auction is that the risk change is explicitly priced into the trades executed.   However, TriOptima assert that anything that can be achieved in the auction can also be achieved by setting risk tolerances along with each firm’s own valuations in their proposed approach.  

In the coming year we should start to see further signs on how the market will adopt these approaches. 

Overall what’s realistic?

An optimist can expect all CCP IRD trades to be converted through one of the approaches above but surely the uncleared world is too complicated to uniformly be converted?  This implies there will be portfolios which will need special handling after the transition for daily PnL, risk management, margining, reporting etc.   This unconverted population is likely to heavily feature dealer to client portfolios and more complex products.  

It is a massive challenge for the industry.

Perhaps this challenge is so grim that conversion is mandated by regulators or becomes the norm once the practicalities fully dawn.   I doubt this will be uniform even if it does happen partially.    Therefore, muddling through might be the order of the day.  

This means that a solution is required in each organization for handling unconverted uncleared trades post transition, which will only add further to the cost and inconvenience penalty associated with uncleared OTC trading.

How should unconverted trades be handled?   

This is a topic I plan to look at in future. 

Conclusion

TriOptima’s proposal seems to be a neat approach to converting CCP trades and may also help with uncleared.

It will compete with auction proposals, CCPs’ own conversion proposals and possibly other vendors and there may well prove to be utility in multiple approaches.

Inevitably conversion won’t be complete and exhaustive.   Read this blog for future posts on how unconverted trades might get managed post transition.

Stay informed with our FREE newsletter, subscribe here.