Learning to live without LIBOR

A RESTAURANT CHAIN in Huntsville, Alabama, draws an extra few thousand dollars from its working-capital facility with a local bank. Meanwhile, its employees’ pension scheme needs to convert the variable interest rate on $10bn of its assets into a fixed revenue stream. The scheme agrees to an interest-rate swap with a hedge fund, which wants to bet on the Federal Reserve raising rates. It places the wager using a margin loan from its prime broker, one of Wall Street’s larger banks.

Each transaction needs a benchmark to decide what interest rate should be charged. The bizarre thing is that they all use the same one: the London interbank offered rate (LIBOR), an estimate of the rate at which big banks in London lend to each other in an obscure corner of the money markets. Every day, it gives borrowing costs for each of five currencies, for periods ranging from overnight to a year. Those for the dollar alone are used to determine the interest rates on $223trn of debt and derivatives—more than two-and-a-half times annual global GDP. But the benchmark is not long for this world. Fixings for the euro, sterling, Swiss franc and yen will be discontinued at the end of 2021, and those for the dollar in June 2023. What will replace them?

LIBOR is best known for a scandal that broke in 2012, when it emerged that banks and traders had been illicitly manipulating it for years. That had become even easier after the market it was meant to measure all but evaporated. For most of the five decades since its invention, banks used unsecured loans from their peers as a source of day-to-day funding. But these dried up during the financial crisis of 2007-09 and never returned to anything like their previous volume. An interbank lending market with daily transactions of around $500m now underpins contracts pegged to dollar LIBOR that are worth about 450,000 times that. This disparity led regulators to call time on LIBOR in 2017.

The most striking feature of the transition is the diversity of the candidates to replace it. Start with the leading alternatives in each of LIBOR’s five currencies. Those are the secured overnight financing rate (SOFR) for the dollar, the sterling overnight index average (SONIA) for the pound, the Tokyo overnight average rate (TONAR) for the yen, the Swiss average rate overnight (SARON) for the Swiss franc, and the euro short-term rate (€STR) for the euro. All are so-called “near-risk-free” rates, which measure the cost of overnight lending. But they do this in different ways, which reflect local funding conditions. British, European and Japanese banks lean on unsecured deposits for their day-to-day liquidity. So SONIA, €STR and TONAR are daily averages of the rates banks pay on such deposits. But big American and Swiss banks tend to rely more on repos (secured loans collateralised by government bonds). So SOFR and SARON measure transactions in these markets instead.

Further fragmentation is emerging within the dollar market. SOFR is less useful for the regional bank in Alabama that lends to local businesses, since it is unlikely to be a heavy user of repos. Such lenders tend to be more reliant on short-term, unsecured debt not dissimilar to that underpinning LIBOR, albeit at rates set for small banks in the local market rather than big ones in London. Hence the creation of AMERIBOR, an index that measures borrowing costs of small, medium and regional banks in America. Two indices produced by Bloomberg, a financial-data firm, and by the Intercontinental Exchange (ICE), aim to perform a similar function for larger banks that derive some of their funding from long-dated bonds rather than short-term repos, and that wish to lend using a benchmark that reflects their normal borrowing costs.

These alternatives to SOFR may be better tailored to certain types of lenders, but they come with two drawbacks. The first is that they might be too LIBOR-like. Thomas Wipf of the Alternative Reference Rates Committee (ARRC), a body convened by the Federal Reserve to steer the dollar market’s transition, warns against relying on reference rates that may share LIBOR’s shortcomings. A bespoke index representing banks’ funding costs may seem like a fine idea, until a liquidity crunch hits and the transactions underpinning it dry up, making it unstable in times of stress. SOFR, which is based on over $1trn of daily repo transactions, was chosen by the ARRC to avoid precisely this eventuality. By contrast, AMERIBOR and ICE’s Bank Yield Index are underpinned by transactions worth $2.5bn and $15bn respectively.

The second concern is liquidity, and hence trading costs, in the markets indexed to the more tailored benchmarks. Under LIBOR—and, in the future, SOFR—the restaurant chain in Huntsville has access to the same derivative market as the hedge fund. That vast trading pool means that small businesses can buy contracts to protect themselves from rising interest rates relatively cheaply. AMERIBOR might hew more closely to the regional lender’s funding costs, but trading debt and derivatives linked to this niche yardstick is likely to be slower and more expensive.

These flaws could mean that some of the new benchmarks fall into disuse. Yet, on balance, the constellation of successors to LIBOR seems like a good thing. It is fundamental to the concept of debt that borrowing costs depend on who you are, who your lender is and which currency you borrow in. The new measures should do a better job than a relic with little relevance for today’s financial system. ■