Taking the L-I-E Out of Libor

Illustration by Ethan Buller

By Kim Schoenholtz and Lawrence J. White Jul 26, 2012 6:45 PM ET

The recent revelations by Barclays Plc (BARC) probably spell doom for the London interbank offered rate, at least in its present form.

Many banks facing huge legal risks could decide to end their participation in the rate-setting process. And even if most institutions remain involved, Libor needs fundamental reform if it is to restore its credibility as a benchmark for hundreds of trillions of dollars of financial contracts.

How to ensure that a damaging scandal won’t happen again? The answer seems straightforward: Wherever feasible, benchmarks for financial contracts should derive from actual transactions, not surveys, as is the case with Libor.

The current “survey Libor” is an old-school model that can’t endure. Each day, a group of banks -- 18 for the U.S. dollar panel -- submit to the British Bankers’ Association the hypothetical interest rate at which they believe they could borrow from other banks. To determine the rate, the BBA disregards the top four and bottom four responses, and averages the remaining 10.

Hypothetical Benchmark

In the old days, the BBA maybe could have counted on an honest answer to this hypothetical question. But the incentives for banks to lie increased as Libor became widely used as a benchmark, making it a potential means of manipulating trillions of dollars of contracts.

The global financial crisis that began in 2007 added a powerful reason for banks to lowball Libor submissions because any institution that revealed that its funding costs were higher than those of its counterparts could be putting itself at risk of a run by its fearful depositors. The now-public record of Barclay’s activities shows how such considerations can overwhelm a bank’s compliance safeguards and reputational considerations, not to mention its ethics.

Transactions (at least in large volumes) don’t lie. The deeper the market, the more difficult it is to manipulate. That is why investors confidently trade trillions of dollars in futures and options contracts based on transaction prices in liquid markets such as U.S. Treasury bonds or theStandard & Poor’s 500.

In the case of the interbank loan market, the effective rate actually paid -- weighted by the volume of transactions at each rate -- would be a natural benchmark for other financial contracts.

How could the BBA’s “survey Libor” be replaced with a transactions measure? One approach would be a clearinghouse to broker interbank loans. It would make the interbank market more efficient, and need only exist electronically, allowing banks to post fund bids and offers to counterparty banks of their choosing. It also would make it possible for willing banks to transact and settle in real time.

Much as the Federal Reserve does with the federal funds market, the Bank of England could use the anonymous transactions data from the clearinghouse to post an effective Libor rate at each active maturity. The central bank could also disclose transaction volume and information on the dispersion of interbank lending rates.

Naturally, not all maturities (or currencies) provided for by the existing Libor process would be active on a daily basis. Especially in a crisis, the volume of transactions at longer maturities can be expected to plunge, as it did in 2007-2008.

Pricing Rules

Yet making volume information available would allow the parties to Libor-linked financial contracts to agree in advance on robust pricing rules -- which would be less vulnerable to egregious manipulation -- to be implemented for illiquid maturities or episodes. The prices of exchange-traded interest- rate swaps might provide useful guides in such instances.

Continuing with the Libor status quo appears untenable. The costs paid by Barclays alone -- including a fine of 290 million pounds ($450 million) and a 24 percent plunge in its stock price since the settlement -- should be sufficient to make banks reconsider their involvement in the process. If they can no longer justify to shareholders the risks of participation, “survey Libor” will die.

Now that Barclays has settled, regulators must address the possible misbehavior of more than a dozen other Libor-panel banks. And prosecutors may still bring criminal or civil charges against banks or their employees for fraud or collusion.

The potential liabilities and risks are astounding. Aggregate penalties and fines could easily amount to billions of dollars. Lawsuits by injured parties could add billions more. In a worst-case scenario, a criminal conviction for U.S. antitrust violations as a result of Libor collusion could cost a bank its charter. If several institutions were threatened with this fate, a bank run or a broader financial crisis would be possible.

To ensure a safe and speedy Libor transition, U.K. authorities could pay the new system’s setup and initial operating costs, which are likely to be modest. Over time, the electronic clearinghouse could become a publicly owned and operated utility, or it could be sold and chartered to a strictly regulated private owner.

Like Fedwire, the platform provided by the Federal Reserve for large interbank transfers in the U.S., a new low- or no-cost service for Libor probably would attract the overwhelming share of transactions in the market. The resulting benefits of a transparent and credible Libor benchmark would far exceed the minimal cost of operating the new system.

(Kim Schoenholtz and Lawrence White are professors of economics at New York University’s Stern School of Business. The opinions expressed are their own.)