News: Which transaction based benchmarks might replace LIBOR?02 May 2013 | Jon Skinner
The IOSCO panel is nearing a recommendation which looks set to push hard for transaction based benchmarks to replace those currently set by panels of dealers / banks. Some press is here:
- Bloomberg: Gensler-Wheatley Panel Aims to End Conflicts in Benchmark Rates,
- Bloomberg: Benchmark Rates Should Be Anchored in Real Data,
- Iosco Says
- FT: Regulators call to tie benchmarks to data
What might those transaction based benchmarks be for USD interest rates? There seem to be two candidates:
Candidate 1: Unsecured funding overnight rates
- Overnight: USD Fed Funds rate - weighted average of end of day interbank close out transactions
- Futures: CME Fed Funds or OIS futures
- Swap rates: OIS are already cleared and e.g. LCH publish USD OIS swap rates out to 30 years
- Pros and cons: OIS has the appeal of already having a swap market established. However, the futures are relatively limited in depth and liquidity. Worse still the overnight rate itself is based on a small market in terms of actual transactions - the end of day cash shorts and longs being settled out between the banks. These transactions can be sizable they are a tiny tip of the iceberg of all interbank funding which comes from two main sources: interbank repo and various central bank "temporary" programs (which presumably noone wants to base a benchmark on).
Candidate 2: Secured funding overnight rates
- Overnight: USD DTCC GCF Repo rate - weighted average rate from all transactions matched and settled via DTCC's FICC (not broker repo rate quote based indexes)
- Futures: NYSE LIFFE US GCF repo rate
- Swap rates: GCIS (swaps based on GCF repo rate) are already traded but only out to 2-3 years and are not yet clearing. So no CCP can produce a transaction based swap rate.
- Pros and cons: The premise is that there exists already for USD the DTCC GCF repo rate which is approximately 40% of the interbank overnight repo market (the triparty subset). GC stands for general collateral and excludes securities on special and therefore subject to a dislocated repo rate. This market is very deep and very liquid and therefore hopefully less susceptible to rate manipulation. Seemingly a good starting point? The disadvantage for GCIS is that there are no published transaction based swap yields for these swaps (they are not cleared so SwapClear cannot provide them). This can be solved either by SDRs publishing them (if they remain uncleared) or by the clearing of the interbank GCIS and having the CCPs add to their published settlement swap yields. Also the swaps market for this interest rate (GCIS) is not yet very developed whereas OIS is quite well developed. GCIS is being traded but only out to 2-3 years not 30+ like OIS.
How will the decision be made? It is unclear to me whether this decision will be made by the market or regulators or a combination of the two. I doubt regulators won't want to be seen to pick the winner directly but rather agree principles and criteria. Once their views are finalized it feels like the market will push this towards a conclusion through the growth of liquidity in the swaps and futures for the chosen solution.
As well as the new swaps, what happens to IR discounting? The liquid benchmark would presumably be applied to discount all IRS. In the same way as OIS has been / is being adopted as the discount rate for both OIS and LIBOR swaps, the logical progression would be that GCF repo rate would be adopted as the new discounting factor for GCIS, OIS and LIBOR swaps both at CCPs and for bilateral trades. LIBOR swaps markets are still finishing the transition from discounting on LIBOR to discounting on OIS. Though recalibrating both LIBOR and OIS swaps to discounting off GCF repo rates would be a further major transition, the LIBOR to OIS discounting transition may well have paved a lot of the structural path already e.g. by allowing all trades to be discounted on a different index from the floating rate index on the trade.
Will USD LIBOR IRS and OIS disappear immediately? No! It's unclear how this part of the transition might happen given 30+ year tenors. See my other post.
What to watch for: for USD LIBOR replacement keep an eye on the trends in futures volumes / open interest (e.g. CME OIS/Fed Funds futures and NYSE LIFFE GCF repo futures) and swap volume / open notional ($ OIS/GCIS swaps via SDRs). (For EUR and other currencies a similar but separate set of details will apply).