Treliant Industry Insights
January 8, 2019
Graham Broyd, David Wagner, and Cathy Cullen
Note: This update is part of a continuing series on the transition away from LIBOR, one of the world’s most widely used interest rate benchmarks. You can read Part I, Part II, and Part III of the series, and take our LIBOR Risk Assessment online.
During the current transition in interest rate benchmarks from the London Interbank Offered Rate (LIBOR) to alternative benchmarks, there is one pivotal difference that cannot be stressed enough: All of the alternatives, including the Secured Overnight Financing Rate (SOFR) in the U.S., are benchmarks that reference overnight rates, not term rates. The new term structures for these rates will be based on averages of overnight rates, not on a single rate, such as a three-month or 12-month LIBOR rate that reflects term lending risk. This becomes critical when one has to convert a contract to a SOFR reset from LIBOR and a value adjustment is necessary. It is especially important for loans, or mortgages, that have traditionally had a term premium built into the resetting benchmark rate.
At the end of last year, Treliant continued to walk banks through the risks and processes of LIBOR transition—this time at a LIBOR Transition Roundtable in London that we co-hosted with our U.K.-based consulting partners New Link, and at a Global Financial Markets Intelligence conference in New York. In both instances, it became clear that most traders, lenders, and other affected parties are still not grasping that the alternative benchmarks for LIBOR will, at their core, be overnight rates reflecting overnight risk, with no term and almost no credit risk. This is a fundamental change from the currently used LIBOR benchmark—one that has profound people, process, and system impacts across trading, valuations, risk, modelling, finance, accounting, and client perceptions.
To recap the basics: The LIBOR transition is a complex change program. It will impact about $300 trillion of debt and derivatives worldwide, affecting mortgage banks, credit card issuers, corporate lenders, traders in interest rate derivatives and futures, and others. Questions related to the transition continue to mount, since this benchmark changes so much of the financial infrastructure that a generation has been accustomed to. Open questions include: how to identify where you have risk related to the LIBOR benchmark; how traders manage and hedge that risk; how contracts have to be changed; how to execute the “value transfer” from legacy LIBOR to new SOFR contracts; how to communicate with clients; and how to manage the internal reporting and pricing risks, as well as the external conduct and client perception risks.
Constructing New Rates
The Federal Reserve’s Alternative Reference Rates Committee (ARRC) has recommended SOFR as the U.S. replacement benchmark, and the other LIBOR committees in Frankfurt, London, Tokyo, and Zurich have all recommended overnight replacement benchmarks—ESTER, SONIA, TONAR, and SARON, respectively. SOFR is a newly constructed rate, first published in July 2018, and administered by the New York Fed; it represents the value of the following three types of overnight U.S. government securities repurchasing (repo) transactions:
- Tri-Party General Collateral rate—based on trade-level tri-party data.
- Broad General Collateral Rate—TCR plus GCF repo.
- Secured Overnight Financing Rate—BCGR plus FICC cleared bilateral repo.
These are overnight transactions, with average daily volume of approximately $800 billion. There is some activity in the one-week, one-month, and even three-month government securities repo market. Most of these “term” transactions are bilateral and uncleared, with sporadic activity. LIBOR is being replaced because its underlying activity was nontransparent and sporadic, with extremely low numbers of transactions. And the only term at which SOFR can be truly representative as a benchmark is the overnight term.
To put this into context, consider that corporate loans today are most frequently based off the three-month LIBOR. Adjustable rate mortgages are most frequently reset off the 12-month LIBOR.
What will three-month and 12-month SOFR “terms” look like? They will be averages, over 90-day and 360-day periods, of the overnight repo rates. They will not reflect the rates for a three-month repo or a 12-month repo of a U.S. government security. There will be no “term” premium built into the price of SOFR. The one-month SOFR benchmark will be based on the average SOFR rate for 30 days. The “O” in SOFR (and SONIA, TONAR, and SARON) is for “overnight.”
The industry is still really grappling with this fundamental change. We have all grown up in a LIBOR world where the lending costs from one bank to another would reflect the risk of six months of lending being fundamentally different than the risk of overnight lending—and priced as such, with the risks to other (interbank) counterparties being priced as a credit spread. That construct is dead. The transition to its replacement is a matter of how quickly SOFR (and ESTER, SONIA, TONAR, and SARON) will be adopted and implemented by market participants as the alternative benchmark. But again, there is no LIBOR-equivalent term rate around the next corner.
Calculating New Rates
How will SOFR “term” rates be calculated? The most likely process is that a three-month rate will be set on the first date of the accrual period (say December 15, 2018, looking forward to March 15, 2019). Then each overnight rate will be recorded for the 90-day time period, accumulating to a 90-day average of the overnight rate therefore not observable or complete as a three-month rate until the end of the period, March 15.
There are certain idiosyncrasies around these calculations, including the industry grappling with providing a single rate on the first day of the accrual period that is to be paid on the last day. But protocols for the calculations will be announced, probably around mid-year 2019, by the International Swaps and Derivatives Association (ISDA) for all derivatives contracts subject to an ISDA agreement. We have been advised that there will be a consultation period, starting in January, for industry participants to provide feedback to the ISDA in the U.S. on these issues. However, many participants already provided feedback to the U.K. Financial Conduct Authority’s (FCA’s) consultation paper on the USD alternative rate structure, and the comments were released on November 27, 2018.¹ The protocols for “term” rates for SONIA in the U.K. are well-established already, and SOFR calculations are likely to be similar in methodology to SONIA. There would certainly be benefits of consistency across markets and benchmarks wherever possible.
Early Issues Flagged
Some early SOFR issues have been flagged, including the following:
- Twelve-month Rates. Banks won’t lend for 12 months at the underlying SOFR rate. That’s right, banks will have to add their own spread for term and credit to the SOFR rate, in order to get it to the equivalent reference value as previously established by LIBOR. This creates a significant value transfer risk. When one transfers legacy LIBOR contracts to SOFR, there has to be an agreement (multi- or bilateral) on the price of that term and credit premium and how is it calculated. There will be winners and losers of that value transfer. Do clients feel the bank has been clear, fair, and not misleading, or do they sense misconduct? And how do you protect yourself from any such challenge or law suit?
- Internal Arbitrage. As traders establish these new rates and rate curves, there is a risk of internal arbitrage or internal trader manipulation, since finance, risk, clients, markets, and systems may have different calculations, pricing, or systems for the rates and a different schedule for transition from one benchmark to another, depending on product and currency.
- Transfer Pricing. A bank’s asset liability management process is dependent on a logical and consistently applied pricing methodology for the cost and use of funds. All debt products—bonds, notes, securitizations, and structured products—could be subject to different schedules for transition from LIBOR to the alternative benchmark depending on the underlying product and currency and the accepted market practice. This will add a real burden to the Treasury function to manage this mismatch of basis and timing.
- Value Transfer. How does one account for the “value” transfer? Is it potentially a one-off shift of value from one book to another, between the client and the bank? How is this classified? This could change all your risk and capital models. One fundamental input would change, which would alter all the outputs. This can have financial, but also strategic risks. And which side of the value transfer are you on, enterprise-wise? The easy-to-identify risks are derivatives and loans. Our experience is that when clients complete scoping exercises to identify LIBOR benchmark risk, they may only get 70 percent of the exposures the first time around. We have 87 product categories on our LIBOR list. It will be important to understand exactly where and how this value transfer impacts you—by division, by geography, and by product.
- Hedge Accounting. The Financial Accounting Standards Board released a new Accounting Standards Update on October 25 that adds the Overnight Indexed Swap (OIS) rate based on SOFR as a U.S. benchmark interest rate. This was done to facilitate the LIBOR-to-SOFR transition and provide sufficient lead time for entities to prepare for changes to interest rate risk hedging strategies for both risk management and hedge accounting purposes. But it is unclear when a deep and liquid hedging market will develop, and thus it will difficult to conclude when to utilize and elect the new hedge accounting treatment.
- Training and Tooling. We have all grown up on LIBOR, in many cases subconsciously—in trading and sales, in risk, and in finance. This shift to overnight reference rates is generational. In order for an organization to grasp and solve all the related issues of this change, teams may need to be retrained and retooled.
There has been a big drive to try to generate transactions in SOFR “term” rates and build out a term curve. The SOFR Future has been active since May 7, 2018, and shown good initial transaction volume and participation. However (at the risk of sounding pedantic), the Futures contract is a Future of the overnight rate—it tracks and values the direction of the average of overnight rates. The future does not represent term risk in the way LIBOR did.
Once banks have priced in the difference between SOFR and LIBOR and what they will charge to compensate, could an OIS market establish itself off the back of those calculations of risk and be the starting platform for a real term benchmark curve in the future? It’s possible, but again one must recall that the benchmark has to stand based on actual underlying transactions. The underlying transactions are overnight repo of government securities transactions, and I am doubtful you could build an industry-agreed and regulatory approved benchmark where your underlying qualifying transactions are based on transactions in swaps deals.
The LIBOR transition will be a journey, a bumpy and complex one, where a map will be quite helpful. The next key steps are:
- the U.S. ISDA consultation and resulting ISDA derivatives documented protocols for the fallback benchmark—likely due at mid-year;
- the response to the U.K. FCA’s “Dear CEO” letter, asking major banks and insurers for the preparations and actions they are taking to manage the LIBOR transition (maybe we’ll hear next steps in the first half of 2019); and
- announcements from ARRC workstreams established to discuss the LIBOR transition outside of derivatives and aside from the largest global banks. This last item could include feedback and next steps on mortgages and loans while providing commentary from the likes of Fannie Mae and Freddie Mac. The process already began with the December 7 announcement of a consultation for fallback language for bilateral business loans and securitizations.²
Core to all next steps will be to grasp that alternative benchmarks for LIBOR will have “term” rates that will only reflect averages of overnight rates. Don’t fight that conclusion, adapt to its new order.