Democracy, as Winston Churchill said, is the worst form of government except for all the others that have been tried. Regulators may discover the same is true of Libor as a benchmark for the cost of money.
Andrew Bailey, CEO of the Financial Conduct Authority, announced plans last month to phase out the London Interbank Offered Rates by the end of 2021. Because there aren't enough underlying transactions to base Libor on, banks have to use their judgment -- or, to be frank, guesswork -- to decide what the cost of borrowing should be.
That long lead time is supposed to give markets time to shift to a replacement. But what if Libor could be refurbished to make it fit for purpose, without the hullabaloo of changing the reference rate for $350 trillion of securities that reference it?
Each day, Libor attempts answer a fundamental question: What is the true cost of money? It does this for a range of currencies (dollars, euros, pounds, etc.) and for a range of maturities. Scrapping it could cause chaos for derivatives, loans and other financial agreements tied to the benchmark.
Instead of ditching Libor and making lawyers rich amid a scramble to rewrite the gazillion contracts that are tied to the benchmark, maybe just changing the composition and calculation of Libor makes more sense.
While interbank lending has collapsed, there's no shortage of observable market rates to suggest what a dollar, a euro or a pound should cost to borrow. Corralling those rates, which are based on actual transactions, to generate a benchmark, wouldn't be beyond the wit of man.
Companies have sold more than 4,000 bonds so far this year worth almost $1.5 trillion. On every single new issue, there's a clearing price at which the borrower gets the money it wants and for which investors are willing to lend. So the market for new corporate bonds supplies a protean landscape of interest rates that could inform a new version of Libor.
The derivatives market generates indexes based on credit-default swaps that add yet another flavor of price discovery that could be useful inputs to Libor. And there's also the commercial paper market, where companies issue unsecured promissory notes repayable in no more than 270 days. That generates interest rates that are pretty close to Libor.
It shouldn't be too difficult to derive Libor from this matrix of borrowing costs. Sure, there's a debate to be had about whether bank credit risk should be included or excluded. There's room for argument on how much weight each element should have. And there's a tricky legal issue about whether and how often those weightings should be revisited, especially during times of stress when some markets get out of whack as was the case with European banks during the euro crisis earlier this decade.
Using existing market rates to generate Libor would help address the FCA's twofold discomfort with continuing to oversee the provision of the benchmark. Firstly, there's its tarnished nature, as evidenced by the $9 billion banks have paid in fines for rigging the rates. Secondly, while the FCA has the power to oblige banks to participate in the process, compulsion is clearly a less than ideal way to run the system.
Harnessing actual borrowing costs in the bond, commercial paper and derivatives markets would obviate the risk of rigging and manipulation. It would also automate the process, replacing judgment with mathematics. And it would improve Libor's ability to reflect the true cost of money by extrapolating from corporate borrowing costs, rather than just the price of interbank finance.
Failing that, the regulator could revisit an idea I first proposed back in May 2008, when it became clear that Libor didn't reflect economic reality: send the body that calculates the benchmark down to its local sports store, buy 16 sets of dartboards and darts, set up an interest-rate wall in the chimpanzee enclosure at London Zoo, and let the apes set the levels. Almost a decade later, their time may have come.
This column does not necessarily reflect the opinion of Bloomberg LP and its owners.