LIBOR May Converge With CD Rates

The media outfits these days love to scrutinize bank submissions of LIBOR, looking for some sort of a pattern. The chart below from the WSJ shows how frequently banks change their rate. What are those sneaky bankers up to? Naughty SocGen, so slow in updating those rates. And why is CS constantly changing theirs? The answer of course depends on who is running that department (usually within the corporate treasury) and what "methodology" is used to cook the numbers.

It's a thankless job because the scrutiny is high, the risks of getting it "wrong" and ending up in court are even higher (since there is no "right" answer), and there is no money to be made by submitting. Now that the regulators are taking over this process, the incentives to submit LIBOR on a daily basis will be even lower. The FSA will soon require proof of actual borrowing costs to justify the submission values.

Reuters: An expansion of the panel of banks that submit rates to Libor, a three-month blackout on the disclosure of individual bank's lending rates and a requirement to back submissions with evidence of borrowing costs in other markets are among reforms designed to reduce attempts to manipulate the rate.

But banks generally don't borrow from other banks for 3 months - so where is this proof going to come from? Some think LIBOR submission will simply converge to certificate of deposit (CD) rates. That's right, retail CD rates. After all the full CD curve for each bank is generally publicly available and is effectively how a bank borrows money from the public for a period of time.

JPMorgan: We expect that requiring a formal process for Libor submissions, together with the explicit requirement to reference a broad set of short term markets, is likely to produce convergence towards observable rates such as bank CD rates.

Of course it would be the rates for "jumbo" CDs in order to avoid the FDIC protection effect for CDs under $250,000 (FDIC protected amounts tend to have lower rates). The BofA CD rates for example (below) don't seem to distinguish the account types, but the curve is available on their website daily.

That means that the 3-month LIBOR rate should start approaching the 3-month CD average as the LIBOR regulation nears. It appears that convergence has already been taking place recently (possibly for other reasons). Hopefully it should make LIBOR submission process much easier. Rather than calling their basis swap desk for "guidance" on LIBOR, treasurers will start calling their retail banker for CD rates.