Carrick Mollenkamp has a worrying piece in the WSJ today about Libor in general, and the much-benchmarked three-month Libor fixing in particular.
Jitters have made many banks unwilling to extend loans to each other for more than one week. As a result, the rates they quote for loans of three months or more are often speculative, because there’s little to no actual lending for that time period, brokers say. "It amounts to an average best guess," says Don Smith, an economist at ICAP, the London broker of interbank loans and derivatives.
This is a genuine and big problem, and one can see how Yves Smith could slip gently into hyperbole in response:
I saw this when working briefly in Mexico in 1984. The local McKinsey office confirmed that there was no reliable data in the entire economy.
It’s not that bad, at all. For one thing, as the article points out, it’s not in banks’ best interest to underreport the rates they’re paying on interbank loans, since their own assets are largely tied to Libor. If you’re earning understated Libor plus 20bp, while in reality you’re paying understated Libor plus 30bp, then you’re in deep trouble.
But it is peculiar, to say the least, that unsecured debt is seemingly being marked at lower rates than secured debt:
The Federal Reserve recently auctioned off $50 billion in one-month loans to banks for an average annualized interest rate of 2.82% — 0.1 percentage point higher than the comparable Libor rate. Because banks put up securities as collateral for the Fed loans, they should get them for a lower rate than Libor, which is riskier because it involves no collateral.
There’s no easy solution to this problem, beyond looking only at the more reliable Libor fixings, such as the overnight and one-week rates. But with many loans tied to one-month and three-month Libor, that won’t help a great deal.
Update: The British Bankers’ Association now says it will ban anybody deliberately misquoting interbank rates.