Much of the discussion of the appropriate LIBOR replacement revolves around two objectives. The regulators want a transaction-based substitute, to minimize the possibility of further LIBOR scandals. However, the users of the index, largely retail lenders and derivatives traders, want a forward-looking term rate for private debt, to prevent squeezing of net interest margins, endangering the profitability of lending activities during financial crises.
However, both regulators and regulated banks confront a real dilemma that demands a practicable solution. To wit: there is no liquid market that meets the combined needs of both banks and regulators.
This is not a statistically addressable problem. If large lenders cannot raise large amounts of cash at the rate used to replace LIBOR, financial stress will inevitably result. It is folly to search for ever better estimates of a rate at which borrowers cannot raise money. Thus, there must be a readily available liquid market for funds at the specific maturities used for a cost of funds index to make a loan/derivatives index viable.
The way forward for the CME (or whatever replaces it) lies in reversing the source and the use of liquidity. Futures are no longer derivatives of cash instruments. We are hung up on our vocabulary. Cash instruments in the short-term debt market now derive their demand from the more useful futures markets. They are the derivatives today, not futures.
We ought to reverse the direction of derivativity. Term cash instruments should derive their demand from futures markets. If the liquid futures markets were to create their own term debt instruments on settlement, the problem of term instrument liquidity would evaporate. Futures can liquefy the cash markets for short-term debt.
The regulators have advanced a rate, SOFR (Secured Overnight Financing Rate), based on Treasury-collateralized repurchase agreements. Regulators should think twice before securing bank cooperation with a decision that could lead to a financial crisis, which SOFR might do. If collateralized rates fall while bank costs of funds rise in a financial crisis, the result would be catastrophic. Lately, perhaps in recognition of this concern, regulators have become less aggressive in pressing banks to adopt SOFR, as described here.
To prevent a SOFR-induced disaster, the market must come to grips with the fundamental issue, liquidity of term money. There is no liquid market for 1-, 3-, and 6-month private debt. This is not a statistical or ethical problem; it is an existential problem. It demands an existential solution. The liquidity of markets with these maturities must be somehow jump-started.
The chart below, to be found in IBA's white paper, "U.S. Dollar ICE Bank Yield Index Update, April 2019" displays the problem. The regulator concern that the LIBOR replacement rate be transaction-based always had two purposes. First, the rate would be more difficult to manipulate if it is transaction-based. Second, the index must be in some sense, actionable. That is, the rate should be one at which money can be readily bought and sold.
The three daily yields ICE plans to give us to replace LIBOR are going to reflect neither market conditions nor the appetite for cash at the maturities in question. If the ICE Bank Yield index is used, the substitute will be an average over the course of a day, of short term rates generally, not an index of demand at each maturity. The chart below reveals planned source and use of transactions to produce a LIBOR replacement.
Even though there is a comforting cluster of data observations at the tenors (1-, 3-, and 6-months) that the method will generate, a couple observations are clear from the graph. The old polling method of generating LIBOR would probably have generated a 3-month rate about 10 basis points higher than the blue line rate. Furthermore, the slope of the index curve inevitably owes more to the illiquid longer-term rates than the near-term rates that it will generate for user purposes. If not for the use of transactions at longer terms, the one-month and three-month LIBOR rates would have been higher; the six-month LIBOR rate, lower.
I do not make this comment to criticize ICE' approach. Indeed, I have no ready way to improve it. Given the IBA remit - find a transactions-based, private funding term rate; I could not do better. One attractive characteristic of the ICE method is its transparency. In this regard, I doubt there is a better alternative among existing commercial methods.
Nevertheless, we must do better. Such data-based indexes suffer from the same shortcoming as LIBOR itself. Few ever borrow or lend at the proposed index rate itself. In other words, a transactions-based statistically estimated term rate is not necessarily a rate at which large volume transactions are possible. To get a reliable measure of a cost of term money, one must first produce a reliable way to borrow at the term in question.
In other words, the shortcoming is not procedural. It is existential. To get a rate representative of term borrowing costs, term borrowing itself must be available.
Jump-starting the term structure
To understand how to create a liquid short-term market, it is useful to remember how we destroyed it. The term market bloomed originally to provide a source of funding for term loans. However, methods that are more efficient have replaced bank funding at these terms. Banks can more cheaply borrow very short-term in the money market fund and deposit markets, hedging any residual risk using less credit risky Eurodollar futures or forward rate agreements (FRAs). Longer term, banks have replaced rolling LIBOR borrowing by interest rate swaps.
This phenomenon creates the existential problem LIBOR faces. On one hand, banks cannot run a LIBOR-based loan book without a hedge of the maturity mismatch created by overnight funding at a government-collateralized rate. On the other hand, the hedges banks use - futures and interest rate swaps - depend on the existence of the term deposit rates they have competed into extinction for their daily valuation.
Because derivatives are derived from cash prices. Or at least they have been until now.
Cash markets derived from futures
The nature of the problem suggests the design of the solution. If futures and swaps are the "real" market, then futures exchanges could use their liquidity to produce actual three-month money as a by-product derived from futures trading. The exchange could settle open positions in Eurodollar futures by delivery to buyer of an exchange-originated three-month term instrument at the closing futures yield on settlement day.
This way, futures exchanges could unify demand for term money futures contracts and demand for term money itself to create a single liquid pool using the essential reality of liquidity - liquidity begets itself. I describe the method here and explain in detail here.
Disclosure: I/we have no positions in any stocks mentioned, and no plans to initiate any positions within the next 72 hours. I wrote this article myself, and it expresses my own opinions. I am not receiving compensation for it. I have no business relationship with any company whose stock is mentioned in this article.