How Do You Replace the Most Important Number in the World?
(Bloomberg Opinion) -- In deciding to replace key interbank rates, financial markets have created their own version of a Y2K problem: how to ensure existing structures don’t collapse during the transition.
Hyperbolically termed the most important number in the world, Libor (London interbank offered rate) is a key market interest rate. It evolved in conjunction with euro-currency markets in the mid-1960s. The British Bankers Association formalized the arrangements in 1986, facilitating the use of Libor to price loans and act as a reference rate for interest-rate derivatives.
Today, Libor and its equivalents in other major currencies are used for a wide variety of transactions, including deposits, loans, bonds and derivatives. While comprehensive statistics are unavailable, it’s estimated that somewhere around $370 trillion of transactions are linked to the rate. While the bulk of transactions are short-term, one year or less, some have longer maturities that can extend to 30 or more years. Changing such ubiquitous and important benchmarks is a fraught challenge.
Several factors are driving the changes. The basic concept of Libor, which is based on a survey where participating banks advise the rate at which institutions can borrow from each other in the London interbank market, is flawed. Grounded in a different ethical era, the self-regulated system is vulnerable to abuse. Traders have colluded to influence rate sets to benefit their positions. In 2008-09, banks, with alleged implicit support from regulators, submitted artificially low rates to calm nervous markets.
Banks eventually paid around $10 billion in penalties, a few bankers were jailed and several senior staff resigned. In the post-crisis period, trading volumes in the interbank markets have declined, making the market rate less reliable. Potential legal liability has made banks reluctant to supply quotes used to determine the benchmark.
Following extensive reviews, several replacement rates have emerged. In the U.S., there’s SOFR (Secured Overnight Financing Rate), which is based on the cost of overnight loans, using repurchase agreements secured by U.S. government debt. The European Central Bank is developing an unsecured overnight rate based on transaction data available to member central banks. Japan favors TONA (Tokyo Overnight Average), based on unsecured money market rates administered by the Bank of Japan. The U.K. has selected Sonia (Sterling Overnight Index Average), which reflects the unsecured overnight funding rates of banks and building societies. Switzerland proposes Saron (Swiss Average Rate Overnight), based on repo rates administered by the SIX Swiss Exchange.
There are problems with all these potential replacements, however.
Competing proposals in some currencies and different approaches between rates create inconsistencies and complexity. ICE Benchmark Administration Ltd., currently responsible for overseeing Libor, has proposed switching to the U.S. Dollar ICE Bank Yield Index, which competes with SOFR. As Libor isn’t being eliminated -- regulators will simply no longer force banks to continue supporting the benchmark -- it’s possible that different rate mechanisms could co-exist in the future, creating confusion. The proposed benchmarks are untested. It’s unclear whether they will prove robust and less susceptible to manipulation.
Libor was designed to enable banks to lend at a spread over their marginal cost of funds. Some new rates don’t reflect the credit risk of banks. For example, SOFR and Saron reflect essentially risk-free funding costs. The margin between risk-free and bank rates can be volatile, especially in times of stress. Since 2000, the spread between three-month Libor and overnight repo rates (a useful proxy for SOFR) averaged around 30-40 basis points but rose to 460 basis points in 2008. Where government rates become the market benchmark, banks may increase lending rates to compensate for the uncertainty of their actual funding costs.
Unlike Libor, which fixes the rate for periods such as one, three or six months, some proposed benchmarks are overnight rates. The lack of term structure creates uncertainty in the cost of funding for borrowers. There are difficulties in agreeing on standard methodologies for pricing the credit and term risk to be added to the risk-free rate.
A further issue relates to hedging interest rate risk. The nascent derivative markets in new benchmarks don’t currently approach the quality, depth and liquidity of established Libor-based instruments. The new benchmarks, especially where the reference rate used in a loan or investment differs from that underlying the derivative, create mismatches that will be exacerbated in volatile conditions. This may create hedge accounting difficulties, forcing mark-to-market gains to be recognized as current-year gains or losses, creating earnings unpredictability that may discourage risk management.
Finally, with Libor being phased out by 2021, the transition from existing to new rate structures presents logistical challenges. It requires identification of affected transactions, selection between migration to the new rates or using fallback arrangements and then documentation of the changes. Procedural and authority issues, as well as legal and tax implications -- for example, from the termination of certain hedging arrangements -- complicate the transition. Asymmetric effects on parties, creating winners and losers from rate changes, increase the risk of litigation. Bank of England Governor Mark Carney has warned of the complexity of transition.
More than 200 years ago, the philosopher Edmund Burke cautioned against destroying an edifice or a system which has stood for any length of time without ensuring that there was something better to take its place. Financial markets and regulators would do well to heed that advice.
This column does not necessarily reflect the opinion of the editorial board or Bloomberg LP and its owners.
Satyajit Das is a former banker and the author, most recently, of "A Banquet of Consequences."
©2019 Bloomberg L.P.