Saying Goodbye to Libor Won’t Be Easy, but It’s Necessary

Amid the investigations into manipulation of Libor, policy makers are coming around to the idea that the world needs a new benchmark to price hundreds of trillions of dollars in loans, securities and derivatives. The challenge will be getting banks, investors and borrowers to agree on what that benchmark should be.

Even when it was created some three decades ago, the London interbank offered rate was far from ideal. Its aim was good: to provide an objective measure of banks’ borrowing costs, so that the parties in financial contracts would see that they were getting a fair deal. Lenders, for example, could peg payments to an interest rate that more or less tracked how much they paid to borrow money, while borrowers could consult the measure to be sure they weren’t being overcharged.

We’re now abundantly aware of Libor’s flaws. Instead of gleaning information from actual transactions, Libor relies on banks to report their borrowing costs honestly, something they spectacularly failed to do. The market it supposedly measures -- interest rates on short-term loans among banks, unsecured by collateral -- tends to disappear in times of crisis, making the rates no more than estimates. The group that polices the system, the British Bankers Association, is in large part run by the same banks it purports to oversee. (Disclosure: Bloomberg LP is a competitor of Thomson Reuters Corp., which conducts Libor surveys on behalf of the BBA.)

Observable Benchmark

Given the drawbacks, it’s hard to see how Libor can be fixed. What the market needs is a benchmark based on a large volume of actual, observable loans, one that is highly resistant to manipulation.

It’s encouraging that the world’s central bankers, who will meet in September in Basel, Switzerland, to discuss reforming the system, are beginning to zero in on some possible replacements -- at least for U.S. dollar Libor, which is the most widely used of the Libor benchmarks. In Congress last week, Federal Reserve Chairman Ben S. Bernanke mentioned two leading candidates: overnight index swaps and the general collateral repo index.

Yes, they’re both mouthfuls. Let’s take them one at a time.

Overnight index swaps are contracts based on the so-called federal funds effective rate, which is the interest U.S. banks charge one another on overnight loans. The underlying loans are observable: The Fed records them and publishes a weighted average interest rate every day. Problem is, banks tend to pull out of the market during times of stress, leaving it too small and too easily skewed to provide a true picture of borrowing costs. Currently, for example, the market is dominated by government-backed behemoths such as Fannie Mae and Freddie Mac, which are lending at abnormally low rates.

The general collateral repo index looks like a better option. It tracks the very large market for repurchase agreements, known as repos, typically overnight loans made against good collateral such as U.S. Treasuries. The Depository Trust & Clearing Corp. publishes a daily weighted average of the actual interest rates paid on these loans. Aside from being secured by collateral, a large portion of the loans are processed through a central counterparty that protects the system against default by any one participant. These features make the repo market, and especially the part that uses Treasuries as collateral, relatively resilient in times of crisis.

Keep in mind that most banks these days tend to package their loans into securities. They then pledge the bundled loans as collateral when they borrow in the repo market. Given the way modern finance works, an index that reflects the cost of secured lending might make more sense as a benchmark.

Critical Mass

Establishing a new benchmark won’t be easy. For a repo index to gain widespread acceptance, a market must grow up around it. And investors and dealers won’t want to hold securities tied to the index until there is a robust derivatives market providing them opportunities to hedge their holdings. Last week, NYSE Euronext contracts on the repo index, and banks are signing up to trade interest-rate swaps. Even so, the derivatives will have a hard time gaining critical mass until a lot of securities have been issued. It’s the classic chicken-and-egg problem.

Here, the U.S. government’s debt load can be an advantage. The U.S. Treasury could (as it it might) issue floating-rate notes -- that is, new government bonds whose interest rates would fluctuate with the repo index. As banks’ and investors’ holdings of the notes grew, so would their demand for derivatives. The notes’ prices and yields would also provide reference points for the issuance of mortgages, corporate loans and all kinds of other securities tied to the index.

Until a replacement emerges, regulators will have to do what they can to keep Libor honest. Prosecuting criminal cases against traders and bank executives would send a powerful signal. Requiring banks on the Libor panel to improve their internal controls and submit to external audits, as the authorities have done with Barclays Plc, would help, too. The New York Fed made some other useful suggestions in 2008. For example, Libor panel banks currently report borrowing rates for 15 different time periods, at many of which they rarely or never borrow money. Reducing the number of maturities would better align the benchmark with reality and make the reporting less of a guessing game.

Ultimately, the best way to deal with Libor’s flaws is to move to a better benchmark. Repo looks like a big part of the answer. With some help from the U.S. Treasury, the market might agree.

To contact the senior editor responsible for Bloomberg View’s editorials: David Shipley at