Why bank deposits shouldn’t be an asset class

By Felix Salmon
January 5, 2012
responded to my piece about his proposal to guarantee all bank deposits.

" data-share-img="" data-share="twitter,facebook,linkedin,reddit,google" data-share-count="true">

Amar Bhidé has responded to my piece about his proposal to guarantee all bank deposits.

First, he says that the real stupidity in Ireland wasn’t the blanket guarantee of bank deposits; after all, he says, “virtually all Western governments implicitly or explicitly guaranteed all bank deposits (and other forms of short term cash, such as money market funds) in 2008″. Instead, it was Ireland’s guarantee of bank bonds. “this was dumb”, says Bhidé, “but it has nothing to do with whether or not deposits should be guaranteed”.

I think the two are in fact intimately connnected. Deposits are a funding source for banks, as are bonds. If you take in more of the former, you’ll issue less of the latter; cheap deposits will quickly chase out expensive bonds. If investors start flocking into 1-year and 3-year certificates of deposit on the grounds that they’re federally guaranteed, it’s hard to see how investors would find 5-year or 10-year uninsured bonds particularly attractive, unless they yielded a lot more than the CDs. And it’s equally hard to see why the bank would feel the need to pay through the nose to issue expensive 10-year debt, when it had unlimited access to much cheaper short-term funding instead.

Bhidé says that capital requirements would force banks to issue bonds — but I don’t see it. If there’s one thing we learned during the financial crisis, it’s that no one really considers subordinated debt to be useful capital: no bank has ever defaulted on its bonds and survived. In other words the kind of capital you get by issuing bonds is the kind of capital no one particularly wants. Instead, investors and regulators are increasingly looking at pure equity, in the form of metrics like TCE. I would hope that we’ve moved away from the paradigm that when a bank wants to improve its capital ratios, it issues a bunch of new liabilities which mean certain death if they’re ever defaulted on. That doesn’t strengthen a bank at all, and everybody knows it.

Bhidé’s proposal wouldn’t just destroy price discovery in the bank-debt market. It would also have the effect of raising interest rates more broadly, and increase the US government’s cost of borrowing. After all, bank CDs are always going to yield more than Treasury notes. If they carry exactly the same government guarantee, then there will be a significant move out of Treasuries and into bank deposits. I can’t see why the government would want that.

And yes, a lot of governments did implicitly or explicitly guarantee all bank deposits during the crisis. But here’s the thing: implicit guarantees are better than explicit guarantees, and temporary guarantees are better than permanent guarantees. Bhidé wants a permanent explicit guarantee, which is the worst of all.

An explicit and permanent guarantee on bank deposits would, overnight, create a whole new risk-free asset class — in a world where we’ve learned time and time again that risk-free asset classes are a Really Bad Thing. Investors tend to put inordinate value on safety, when what we really want to encourage is risk-taking. And highly-regulated banks with massive government-guaranteed bank deposits are not the best mechanism for allocating risk capital. Markets might not be perfect, but they’re surely better at capital allocation than that.

As companies like Apple build up their cash reserves into the tens of billions of dollars, we’re already beginning to see a world where bank deposits are becoming something of an asset class. It’s not a world we should encourage — let alone one we should institutionalize with a blanket government guarantee. Money should be invested; the government has no business encouraging corporations and institutional investors to simply park it in a bank instead.

13 comments

Comments are closed.