LIBOR Reform: What Will Work?

by: Kurt Dew

Intercontinental Exchange (ICE) has recently introduced a potential LIBOR replacement instrument to compete with the Secured Overnight Funding Rate (SOFR), the Fed’s proposed LIBOR replacement.

The dealer banks have endorsed the ICE proposal.

The dealer banks have ample reason to object to SOFR.

SOFR will tend to fall relative to the actual cost of dealer bank money at times of crisis.

We need a third choice.

What index will work as a LIBOR replacement? Consternation over the question of replacing LIBOR continues. For two reasons, there is no guarantee that any choice of LIBOR replacement currently advanced by bank regulators will “work.” Things have not gone well for these replacement candidates. Problems include:

  • The failure of the SOFR-based derivatives instruments to gain much traction [for example, the Eurodollar Futures contact at CME Group ((CME)), traded roughly 1.8 million contracts ($1.8 quadrillion) on Tuesday, January 29th; the Secured Overnight Financing Rate (SOFR) contract, under 1.2 thousand contracts ($1.2 trillion)].
  • The basis risk associated with an interest rate like SOFR (overnight tenor and Treasury-related). LIBOR is a forward-looking 3-month rate on private debt. SOFR is a backward-looking overnight rate. The potential to play havoc with bank profitability during times of crisis is substantial.

The consequences of this replacement choice are enormous. Hundreds of trillions of loan instruments depend on LIBOR for valuation. The most successful futures contract, Eurodollar futures, depends on LIBOR. LIBOR-based pricing also dominates the even bigger interest rate swaps market.

As Steve Wunch, of Wunch Auction Associates, LLC argues “Going from a “manipulated” quote-based system to a transaction-based one, as regulators are demanding, effectively requires changing from a non-continuous call market structure to its opposite: a continuous market structure. Since the call market structure was critical to Libor’s success, its elimination will make it forever non-viable.”

On the other hand, the regulators have made it plain that any replacement not based on a continuous market structure is unacceptable to them.

Regulatory point of view.

Regulators around the globe have made one thing completely clear. They demand to see a pricing index that is firmly rooted in transactions. Wunch’s belief notwithstanding, an index based on market transactions can work under the right conditions. Witness the success of the S&P 500 stock index futures. Whether such an index can work in the LIBOR market, however, is an open question.

It is reasonably certain that ultimately the world will settle upon some specific LIBOR replacement. To predict the likely LIBOR replacement, it may be fruitful to consider how LIBOR became so ubiquitous. Regulators did not designate LIBOR. It evolved. The interest rate uncertainty that plagued the decade of the 1970's begged for a market-driven short-term rate that would provide a compromise between the banks’ desire to control the spread between their volatile wholesale deposit costs and their not-so-volatile loan rates, on one hand; and borrower desire for predictable loan payments, on the other. Volatility-induced losses on fixed rate loans forced the banks to develop the floating rate, index-based loan. That the index became LIBOR was the banks’ decision, supported at the time by the Bank of England.

Why regulators can block a LIBOR replacement, but cannot promote one.

Regulators are fundamentally in a defensive position. It is within their power to determine what a LIBOR substitute cannot be, but not within their power to insist on what the replacement should be. Such a Fed designation would leave the Fed with culpability if something goes wrong. Thus, the choice ultimately falls to market participants.

A further historical development that will affect the ultimate LIBOR replacement is the failure of an earlier competitor of LIBOR. The wholesale deposit markets of the early 1980's were split between wholesale Certificates of Deposit (CDs) issued in the US and London wholesale time deposits. The two forms of bank deposit debt went head-to-head at the Chicago Mercantile Exchange [now CME Group (CME)] when the exchange listed their associated futures contracts. CDs had the Fed-desired characteristic of being transactions-based.

Eurodollars won this contest, largely because of problems created by the CD settlement process. Since CD settlement was by delivery of an issue of a specific bank liability, two negative effects resulted in market disappointment with CD contract settlement. First, the market does not price all bank liabilities equally. Seller chooses the deliverable instrument and benefits most by choosing the highest yield issue. This affects buyer adversely, since the highest yield issue is the riskiest name.

Furthermore a high yield relative to that of a bank’s counterparts embarrasses the bank whose CDs are delivered, since the world learns though delivery that the delivered bank’s credit quality is in question, perhaps even affecting that bank’s stock price adversely.

The process of displaying the identities of the riskiest banks to the financial world concerns both the banks as a group and their regulators. The LIBOR “fixing,” based on rates the issuers supply, permits issuers to disguise individual bank weakness by submitting their ideas of the average market yield. On first blush, this seemed preferable to both banks and regulators.

Things that will not work.

That LIBOR defeated CD futures gives me reason believe there is no satisfactory existing single debt-based rate that can adequately replace LIBOR. There must instead be a new form of short-term debt, issued on demand, which has no direct link to a single issuer.

A singular conclusion. The implication is that no debt chosen by seller in delivery to buyer can pass muster with bank regulators, since neither bank regulators nor banks themselves will accept the implications for safety and soundness of the banking system. The LIBOR substitute must be issuer-independent, but the marketplace trades no issuer-independent form of debt now.

A way out?

There are, however, nearly issuer-independent debt instruments, for example asset-backed instruments. Multiple issuers originate the assets that produce the value supporting the liabilities created in an asset-backed structure. However, once the supporting assets are bundled together, the asset-backed originator never replaces the original assets. Thus, a given asset backed security decays in quality over time. This decay is undesirable for a LIBOR substitute index.

Another close relative to the imagined issuer-independent form of debt is the exchange-traded fund (ETF). The ETF fund manager has discretion to swap investments in and out of the fund if consistent with the fund’s stated investment objectives. For example, an S&P index fund changes investments when a particular stock has value sufficient to replace another in the S&P 500 list. However, the S&P’s size criteria is not a good one for formation of debt portfolios. The size of debt outstanding is a negative indicator of debt quality.

A third potentially-issuer-independent form of debt is the money market fund. This kind of debt portfolio is also problematic for our purposes. As with the other two potential LIBOR substitutes listed above, the portfolio manager strikes a balance between risk and return, since the instrument itself must appeal directly to investors. The 2008 experience with the Reserve Primary Fund, the collapse of which sent shock-waves through the entire financial system when it “broke the buck,” shows that these short-term liability-backed funds are prone to unstable behavior as they aggressively seek return at the expense of higher risk. This will not do to provide support for a LIBOR substitute.

How should a LIBOR substitute look?

There are three desirable characteristics of asset building blocks used in formation of a name-independent debt instrument – diversifiability, name-irrelevance, and liquidity of the portfolio supporting assets.

First, the debt should be the highest quality in the private market. Moreover, this quality should not decline over time, unless the quality of all private short-term debt is declining. Second, the debt should be covenant-lite: free of issuer call or cancellation conditions. Backing the LIBOR substitute with high quality, diversified, uncollateralized, overnight, unsecured issues would meet these criteria. Overnight commercial paper has several advantages. A decline in quality permits the portfolio manager to drop the name from tomorrow’s portfolio without the adverse effects on issuer credit reputation of selling the issue into the market, further exacerbating the issuer’s problems.

If the fund manager does not seek to maximize returns, focusing solely on minimizing instrument credit risk and maximizing portfolio diversification, the LIBOR replacement would instantly become the lowest risk short-term debt available in the private credit market. Furthermore, the issuer might change the portfolio names day-to-day, further reducing that risk.

Consumers of LIBOR-based products should guide selection of a successful LIBOR replacement.

The two camps debating the choice of LIBOR replacement now are the bank regulators and the dealer banks themselves. They are, and will remain, hopelessly deadlocked. They will never agree.

Regulators demand a transaction-based instrument backed by a large liquid market. However, there is no financial instrument meeting the regulators’ conditions that will produce the forward-looking (maturity longer than overnight) instrument that the banks absolutely require to prevent the squeezes on bank net interest margins that SOFR and the others are doomed to create.

Bankers conceivably might develop a LIBOR replacement. There is nothing preventing the banks from building the kind of instrument I describe. That is not what the banks have decided to do. They irrationally attempt to seek control of the index rate the Fed will not permit them.

The International Exchange, Inc. ((ICE)) has recently proposed a LIBOR replacement endorsed by the large dealer banks. ICE’s benchmark, called the ICE Bank Yield index, considered here, is a failed attempt to meet the regulators halfway.

In the words of ICE:

“First, the index seeks to measure the average yields at which investors are willing to invest in the unsecured debt obligations of a broad set of large, internationally active banks for specified forward-looking tenors. These are key elements of interest rate benchmarks that users in the cash markets have historically sought.

“Secondly, the index is underpinned entirely by transaction data representing short-term, unsecured bank investment yields. This solid transactional foundation should make the index robust, avoiding any requirement for judgement in the methodology.

“Thirdly, the index utilizes data from both primary funding markets and the secondary bond market. This enables the index to better represent yields on short-term, unsecured bank debt, given the increased use of the bond market for bank funding since the financial crisis.”


The question is, will the market (or for that matter, the regulators) trust an index provided by the banks, based on vague selection criteria from multiple illiquid market sources.

I suggest the answer is no. As ICE currently describes the index, the banks have too much discretion in picking the rates that will determine the index value. There is no obvious way to determine whether the banks will build an index that makes index-based borrowing more costly than it was under LIBOR. Maybe later ICE will be more specific.

The world awaits a single market price-based rate that it can see is not bank-manipulated. That index will necessarily be transaction market-sourced and based on a single market price of a liquid term instrument, of the kind I propose here.

But only those institutions and interest groups that will suffer most from a large bank-controlled choice of LIBOR replacement will develop this instrument. Among them are borrowers of loans based on LIBOR, and various municipalities, the mortgage agencies, student loan providers, and others currently paying LIBOR plus a spread for credit. This is an enormous asset class (amounts are measured in the hundreds of trillions) with no cohesive leadership. Add to that the technicality of the issues associated with the decision. I conclude that someone will have to produce a LIBOR replacement for them.

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.