Ending the use of dollar Libor, the scandal-tinged benchmark bank funding rate, was always going to be problematic. Some Libor traders went to jail for collusion and self-enrichment. The Fed and its fellow regulators put together a public-private committee on Libor replacement big enough to swamp a ferry boat.
That hasn’t entirely worked. The use of Libor as a base rate for funding costs is bigger than ever — around $225tn of derivatives, consumer loans, corporate loans and cash investments. Nevertheless, the use of Libor is supposed to end, mostly, on December 31 for some Libor rates and by mid-2023 for those remaining.
The process of finding practical ways to replace it have led to increasingly audible shouting and blame trading between the major dealing banks and the Fed, along with the central bank’s entourage of agencies, academics, policy wonks and whisperers.
The major point of argument is the Fed’s ambition for the banks to adopt Sofr, a “secured overnight funding rate” intended to represent the cost of very short term borrowing using US government debt as collateral. The mechanics of using Sofr to replace Libor have been the object of fiddling by the officially blessed Alternative Reference Rates Committee since 2014.
As with so many public private partnerships, the ARRC began with grim co-operation and is now an object of revulsion for some key members.
For some, Sofr may have pristine collateral and governance, but does not reflect the real world of trying to price future interest rates and possible credit losses or systemic risk.
Unlike for Libor, there is not currently a Sofr rate for different time periods that would provide an explicit benchmark for borrowing for say, three or six months. In addition, there is not a Sofr rate with collateral based on private sector loans. This is a risk for banks.
For example, as the Bank Policy Institute has pointed out, in times of stress, rates based on Sofr would fall as the traditional haven asset of US Treasuries rallies. At the same time, banks’ costs of funding would rise as short-term credit rates spike. Thus bank lending margins would be squeezed at a difficult moment.
Thomas Pluta, JPMorgan’s global head of linear rates trading, agrees that “yes, [Sofr] creates a lot of credit basis risk. A lot of banks, particularly regional banks, have been vocal about the lack of credit sensitivity. If you are funding yourself in unsecured markets, that is a pretty obvious concern.”
A Sofr futures contract proposed this week by the futures exchange CME addresses some of the problem of the lack of different time terms. But even then it is based on derivatives. Fed economists apparently believe that a set of derivatives contracts would conjure up a liquid cash market for the underlying funding instruments. The bankers do not share this faith. And the CME contract does not address the lack of sensitivity to credit.
Fed-world may think that there can be simple tweaks leading to a Sofr-plus-a-standard-credit-spread formula, but the bankers strongly disagree. What happens during another pandemic or war if the “actual cost” of funds shoots up even as the secured rate declines?
One New York banker says the result is that banks will seek to hedge this risk by imposing a higher spread than it would with a more credit sensitive rate such as Libor. “The effect is to contract credit,” he says. Political poison.
The Fed crowd still just does not like Libor. In testimony to Congress last month, Mark Van dear Weide, the general counsel of the Fed board, criticised Libor as “vulnerable to collusion and manipulation”. All true. But it still worked.
Despite its shortcomings, Sofr is not going away. The CME’s Sofr contract will trade. Government-sponsored entities are force majeure adopters of Sofr deals. Even the banks have issued Sofr-indexed paper, perhaps funding positions in Treasuries.
That still leaves a lot of credit-risk-sensitive bank assets that will need to be funded. How can the banks determine the price of those risks?
Fortunately there are some prospective private market solutions. Bloomberg has developed a proprietary short term bank yield index, BSBY, (pronounced “bis-bee”). It aims to represent a series of credit sensitive reference rates and define a structure for borrowings of different terms.
Other prospective alternative offerings include indices such as the ICE Bank Yield Index, and Ameribor, an index from the American Financial Exchange oriented towards smaller banks’ overnight funding costs. Unlike Libor and BSBY, these are all proprietary data series. Fees for their use would inevitably be passed on to borrowers.
I have been surprised at the banks’ bitterness about the Sofr “solution”. This is not over.