Not So Fast, SOFR: Is It a Viable LIBOR Replacement?
Paul Noring
Here’s why SOFR could prove deeply problematic for commercial and consumer lending—and alternatives the Fed should consider
LIBOR, the interest-rate benchmark that underpins approximately $200 trillion in financial contracts in the US alone, is set to sunset at the end of 2021, and some financial institutions are raising significant doubts about the initially selected replacement rate.
In June 2017, the Alternative Reference Rate Committee (ARRC), made up of financial- and public-sector officials, with the support of the Federal Reserve, chose the Fed’s Secured Overnight Financing Rate (SOFR) as LIBOR’s replacement.
But some major US banks are sounding an alarm on SOFR’s presumptive coronation. Their objection to using SOFR as a benchmark for commercial and consumer lending is serious and credible—and represents only one of several major issues. The transition to SOFR could have systemic and deeply problematic ramifications for commercial lenders and borrowers. And the rationale for declaring SOFR the new universal reference rate is even shakier considering that other viable options for the loan market are beginning to emerge.
To assess banks’ objections, and the alternatives, it is important first to understand who was behind the decision to move forward with SOFR. It’s also crucial to understand the challenges and potential problems SOFR could create for main street financial institutions—and potentially for nearly every American business.
SOFR vs. Libor: Objections and challenges
Contracting spreads, decreased lending in times of stress
In a September letter to top US banking officials, leaders of 10 large regional banks voiced their primary concern about SOFR: because it is a secured rate (backed by collateral for the full value of the loan) derived from the Treasury repurchase market (LIBOR, in contrast, is a dynamic unsecured rate, derived from banks’ self-reported lending rates), in times of significant market stress it could push borrowers’ rates lower while banks’ cost of funds are rising.
The consequence would either be “a reduction in the willingness of lenders to provide credit in a SOFR-only environment…or an increase in credit pricing through the cycle,” the letter says.
The bankers who signed the letter represent institutions ranging from the sixth-largest in the US (with $400 billion assets) to the 38th (with $70 billion). And yet they united to raise their objections, because none of them was a member of ARRC when it made the initial decision to proceed with SOFR as LIBOR’s replacement. In fact, ARRC’s nine bank members are either top five US institutions, major foreign banks or investment banks.
Operational challenges
ARRC’s decision to make SOFR the official US benchmark rate leaves the vast majority of main street banks not only facing possible contracting credit spreads in a distressed economy, but also confronting serious operational challenges and a significant increase in volatility, which only recently has been addressed via the Federal Reserve’s emergency repurchase operations.
Operationally, bank executives in charge of converting from LIBOR to SOFR face a drastic change when it comes to how their institutions calculate rates for borrowers. That’s because LIBOR is calculated for specific terms, up to one year, at the beginning of those terms; SOFR currently has no associated terms and is calculated at the end of each trading day. That would force banks to shift to calculating interest rates in arrears—requiring massive changes to lending and borrowing processes and systems.
Shifting to an arrears calculation after decades of knowing rates in advance means developing new system requirements, operational processes, training and business process reengineering. That would not only take several quarters but could wreak havoc on the servicing of products like student loans and adjustable rate mortgages.
We all remember the mortgage servicing issues, like improper foreclosures, that emerged when banks tried to quickly stand up loan modification programs in response to the financial crisis; it is not unreasonable to predict similar processing breakdowns in the conversion to SOFR. Will banks and non-bank consumer finance entities be ready for a similar regulatory scrutiny—from the Consumer Finance Protection Bureau, state attorneys general and the class action plaintiffs’ bar?
Volatility and intervention
Because SOFR is tied to repurchase agreement lending, it can also experience more volatility than LIBOR, particularly around quarter-end and large Treasury auctions. For instance, in September 2019 certain transactions exceeded a 9 percent overnight rate, forcing the Federal Reserve to step in with highly public market-stabilizing transactions to bring the rate into a more reasonable range.
That intervention calls into question whether SOFR is even an appropriate replacement rate, as it behaved so differently than an interbank rate like LIBOR would, absent Fed support. Also, if the Fed is supporting SOFR, is it truly a reflective market rate, or is it being managed by a third party? Ironically, that was the very problem that led to LIBOR’s demise: a handful of London banks could (and did) easily manipulate it.
Why not a different benchmark for loans?
Nobody is objecting to SOFR as a benchmark for derivatives. Based on the notional value of contracts, approximately 95 percent of the $200 trillion plus in financial instruments referenced to US-dollar LIBOR are derivatives, with the remaining 5 percent in the cash markets. However, on a net balance-sheet exposure basis, the percentage of cash contracts is easily 20 to 1—more than the inverse of the above ratio.
For that reason, at least 95 percent of the conversion effort will be directed toward lending. That’s a good example of how ARRC did not sufficiently vet the downstream consequences of SOFR, and it’s why the regional banks are appropriately voicing a clear warning to the heads of the three primary bank regulators.
It could be that one benchmark rate should not fit all financial instruments. Institutions could conceivably use one reference rate for derivative contracts and one for commercial and consumer lending. The Fed, and ARRC, should at least consider that possibility—and take a thorough look at the potential alternatives.
Three alternatives to SOFR
If they are willing to reopen their process, ARRC would do well to start with three alternatives, each with a dynamic credit spread that would be better suited than SOFR to lending products. As they exist today, none is an ideal replacement, but each holds significant promise.
AMERIBOR—An interest rate benchmark by the American Financial Exchange that commenced trading in 2015 and reflects the actual unsecured borrowing costs of over 1,000 American banks and financial institutions. Transaction volume has steadily increased, and the rate has proved to be much more stable than SOFR, even during periods of quarterly volatility. In addition, it has achieved compliance with the key regulatory principals required for a benchmark rate.
Bank Yield Index—To be published by the Intercontinental Exchange, this is a forward-looking, credit-sensitive benchmark designed specifically as a potential LIBOR replacement for US dollar lending activity. The index seeks to incorporate key LIBOR properties that cash market participants would like to retain in a US-dollar lending benchmark.
Commercial Paper Rates—The Board of Governors of the Federal Reserve System publishes daily commercial paper rates derived from data supplied by The Depository Trust & Clearing Corporation (DTCC), a national clearinghouse for the settlement of securities trades and a custodian for securities. DTCC performs these functions for almost all activity in the domestic commercial paper market.
With LIBOR’s eventual phaseout a near certainty, two things are crystal clear about a replacement rate for commercial and consumer loans: it must have a dynamic credit spread to ensure the safety and soundness of all but the largest US banks; and it must be a forward-looking rate, as there is no way all commercial and consumer loan systems and processes can be changed in time to avoid potentially major operational issues.
With fewer than two years before LIBOR’s retirement, regulators and private-sector leaders have no time to waste; they should move quickly to assess the full range of challenges and alternatives. Meanwhile, financial institutions should start or expand their testing of alternative benchmarks—preferably with actual transactions tied to those rates.