The FSB published “Progress in Reforming Major Interest Rate Benchmarks Interim Report” on July 9th. We get into some of the detail.
The FSB, through the Official Sector Steering Group (OSSG), is primarily focused on larger markets: EURIBOR, LIBOR and TIBOR (collectively known as the major IBORs) but has not forgotten markets like Australia, Hong Kong, Mexico, South Africa, etc. The FSB’s underlying theme on benchmarks has been that they should be backed up with actual transactions data. Transaction based data is termed “IBOR+”. Collecting transaction data is no easy task.
What has been accomplished between the FSB’s initial report in July 2014 “Reforming Major Interest Rate Benchmarks” and now?
“…reviews of respective benchmark methodologies and definitions, data collection exercises and feasibility studies, consideration of transitional and legal issues, and broad consultations with submitting banks, users and other stakeholders; this work has also typically being undertaken in liaison with relevant international authorities. In recent years authorities in all three jurisdictions where the IBOR administrators are located have taken action to more formally regulate these benchmarks…”
“…The authorities in all three IBOR jurisdictions have taken action to regulate these benchmarks, given their systemic importance. The LIBOR administrator and 20 submitting banks have been regulated in the UK since 1 April 2013, the TIBOR administrator was designated under the new legislative framework in Japan in May 2015, and EURIBOR will be regulated by the new EU Regulation (which will also capture LIBOR and replace the existing UK regime)…”
Behind the IBORs are baseline risk free rates (RFR). There certainly is a need for robust RFR – or at least get as close as possible by isolating out counterparty risk. (Think LIBOR, with its exchange of notional and counterparty risk versus centrally cleared derivatives like OIS.)
“…OSSG members have also made concrete progress in identifying potential RFRs. In particular, detailed data collection exercises have been undertaken in key markets, and work is now underway to identify potential RFRs, where these do not currently exist…”
Repo was mentioned several times as a possible source for RFR rates. For example,
“Although euro-area authorities already consider EONIA to be a viable RFR, they are also working with market participants to develop a new RFR based on general collateral (GC) repo markets…”
The section on RFRs in the recent report used the word “encourage” five times. Reading between the lines, this effort is often about twisting the bank’s arms to contribute data. No surprise that the banks are a bit shy here.
One interesting part of the report was on how the data going into LIBOR would be sourced. It could include not only inter-bank deposits but wholesale funding as well. Repo was on the list in a supporting role.
“…Eligible transactions. Submitters’ unsecured wholesale funding deposits, Commercial Paper (CP) and primary issuance Certificates of Deposit (CD) would be directly included as an input to the LIBOR submission with other transactions such as overnight indexed swaps (OIS), repurchase agreements, or floating rate notes included only when necessary to support expert judgement. Historical transactions and inter/extrapolation would be used when (i) a benchmark submitter has no available transactions for a particular tenor but it does have transaction-derived anchor points for other tenors of that currency, and (ii) if the submitter’s aggregate volume of eligible transactions is less than a minimum level specified by IBA…”
We know that isn’t always possible find a traded price given the lack of liquidity in unsecured deposit trading. OTC trading, unlike exchange based trading, often has a “this is where I think it will trade” component that is more about trader triangulation and instinct than it is about actual data points. Unfortunately this process has often been opaque and was easily gamed. Certainly more actual trade data is better than less – but the depth of the data is important too. Is the sample big enough to be meaningful? CP risk is different from repo risk and should not be jammed into a basket together. The FSB and those tasked to get this right are appropriately focused on data sufficiency. But will the data reported include some statistics to assess exactly what kind of risk is being reported, how deep the markets are and how that changes over time? We hope so.