Last Friday, a top British top financial regulator issued a report suggesting numerous reforms to the London Interbank Offered Rate (Libor). Libor, a benchmark interest rate that affects more than $300 trillion in financial transactions, is constructed from the cost of funds estimates submitted by a handful of large banks. This summer, e-mails surfaced showing that a number of these submissions were manipulated to benefit submitting banks and their clients.
If Libor were to be reformed, the proposals outlined on Friday would constitute a good first step. The problem is that Libor shouldn’t be reformed; it should be abolished and replaced with a market-determined indicator.
The report issued by Martin Wheatley, managing director of Britain’s Financial Services Authority calls for the administration of Libor to be taken away from the British Bankers’ Association and assigned to independent, well-regulated authority. Banks that falsify their submissions would be subject to criminal penalties.
The report also calls for the reduction in the number of Libor rates issued, from the current 150 different currency-maturity combinations (e.g., one week euro, six month dollars) to 20 or so of the most commonly used. It also recommends an increase in the number of banks submitting bids, so that Libor is calculated by averaging the submissions of larger sample of banks.
Finally, the Wheatley report recommends a number of technical fixes, including basing the submissions on transactions data and delaying the release of individual Libor submissions until three months after the fact, rather than publishing them daily.
Certainly, if Libor is to be maintained, as the Wheatley report recommends, these would be useful reforms. They strengthen governance, improve the quality of Libor, and make fraud less likely. The suggested reforms have already received favorable comments from regulators in the United States, Europe, and Japan.
The problem is that Libor is not fixable.
Although better administration and criminal penalties for fraudulent submissions could deter the type of fraud we have seen, given the huge amount of money riding on Libor, the incentive to cheat will still be overwhelming. And since Libor submissions are estimates, it will be hard to prove that a high or a low submission was fraudulent. The recent fines that were doled out over Libor submission were only possible because the perpetrators were stupid enough to describe their actions in e-mails.
Reducing the number of currency-maturity combinations is also sensible. After all, a relatively small portion of the $300 trillion in financial transactions consists of Danish kroner loaned for two weeks or New Zealand dollars for six months. So eliminating the minor currencies from Libor makes sense. There are a number of market-determined alternatives for the major currencies—why not use them?
Neither delaying the release of individual submissions nor requiring banks to submit large volumes of transactions data will fundamentally fix the process. Even if banks no longer manipulate their submissions at the request of fellow bankers or clients, they may be tempted to do so to appease the regulators.
It has been alleged that during the subprime crisis some banks lowered their submissions at the behest of regulators, who wanted banks’ cost of funds to appear lower than it really was in order to increase confidence in the financial system during a turbulent time. Going forward, banks may try to generate submissions that are in line with other banks in order to avoid seeming “out of step” with other banks and drawing increased scrutiny from the regulators.
The only way to fix Libor is to scrap it and replace it with a market-determined indicator. The new indicator could be the GCF Repo index published by the Depository Trust & Clearing Corp, or some newly constructed index based on default swaps transactions, corporate bonds, and commercial paper, as suggested by Bloomberg. This is the only way to insure that the rate will be appropriate and fairly determined.
Recent financial crises and scandals have weakened public confidence in the financial system. Propping up a flawed financial benchmark will erode that confidence even further.
It’s really not as simple as replacement with GCF, or Wheatley would have simply recommended that. There are many reasons, but LIBOR supports trillions in notional of securities and derivatives so any new index would have to slot into this existing population. LIBOR is supposed to represent prime bank unsecured credit: GCF is secured/government credit. This is not an academic distinction, as it makes GCF behave very differently both in terms of level and ability to hedge non-government credit. Whilst I hope GCF does take on an increasing role, it’s apples and oranges to LIBOR and not the solution.
The real LIBOR problem is lack of volume in the underlying loan markets. GCF has volume to back it, but is the wrong type of rate. An index like the Bloomberg proposal get you nowhere because the CDS and corporate bonds referenced have no more volume than the loan markets: this is nothing more than a complex Chimera.
Wheatley is right in its conclusions: what is proposed isn’t great, but it is the least bad solution to a not very tractable problem.