Amar Bhidé is a smart man with a very stupid idea:
We need to take away the reason for any depositor to fear losing money through an explicit, comprehensive government guarantee. The government stands behind all paper currency regardless of whose wallet, till or safe it sits in. Why not also make all short-term deposits, which function much like currency, the explicit liability of the government?
Why not? I can answer that question in one word: Ireland. That country’s blanket government guarantee of all bank deposits was the single dumbest decision of the global financial crisis, and I can’t quite believe that anybody thinks it would be a good idea here. Yes, I understand that the US can print money while Ireland can’t. But that doesn’t mean it makes sense for the US government to take on untold trillions of dollars in extra contingent liabilities.
But let’s rewind to the beginning of Bhidé’s op-ed, and try to work out how on earth he arrived at this rather crazy notion. He begins by saying that “central bankers barely averted a financial panic before Christmas by replacing hundreds of billions of dollars of deposits fleeing European banks” — a statement which comes with no footnote or hyperlink, which makes it hard for me to know exactly what he’s referring to. I suspect it might be the coordinated liquidity operation which was announced on November 30, but I don’t recall anybody at the time talking about “hundreds of billions of dollars of deposits fleeing European banks”.
There was certainly a liquidity crisis in the European banking sector at the time, but there’s a world of difference between a liquidity crisis and a bank run. A liquidity crisis is when banks don’t lend to each other; a bank run is when depositors withdraw the money they have on deposit and move it elsewhere. And while deposits have certainly been flowing out of Greek banks in particular and the European periphery in general, I haven’t seen reports that European banks in toto are seeing massive deposit flight, or that deposit flight was in any way the reason for the November 30 move.
But Bhidé is convinced that there was a bank-run panic in Europe and that there might be one in the US as well:
The Federal Deposit Insurance Corporation now covers balances up to a $250,000 limit, but this does nothing to reassure large depositors, whose withdrawals could cause the system to collapse.
Again, I’d need a lot of argument to be persuaded that a bank run by large depositors in the US is a real danger. According to the FDIC’s Statistics on Depository Institutions, total deposits in the US, as of September 30, were pretty much exactly $10 trillion. Of that, $8.5 trillion was held domestically, and of that, $5.4 trillion is insured. Which means that there’s about $3.1 trillion of uninsured deposits in the US, and $4.6 trillion of uninsured deposits at US financial institutions. (Although some of those international deposits will be insured by the countries where they’re held.)
Certainly a $3.1 trillion bank run would cause the US banking system to collapse — there’s no doubt about that. But where does Bhidé think the money would go? And what makes him think that such a bank run is a real possibility? I can certainly see a run on some individual bank, if it looks like it might be in trouble — we saw that in 2008, at WaMu, although in the end there all depositors ended up with 100 cents on the dollar. But what I can’t envisage is a run on the whole system, where depositors move their money somewhere else entirely — partly because I have no idea where they might move it.
It’s worth noting here that the US differs from Europe in two crucial ways. The first is that companies can’t deposit their money directly with the central bank, in the way that Siemens (SI) for instance does. (The only exceptions are the handful of companies which own banks, like GE (NYSE:GE) and Target (NYSE:TGT).)
More importantly, deposits in the US are senior to banks’ bonds: depositors get their money back before bondholders get anything. In Europe, by contrast, depositors are often pari passu with other unsecured creditors. Which means that deposits are riskier in Europe than they are in the US.
All of which helps explain why the total amount of uninsured deposits in the US has continued to rise since the crisis hit: when I ran the numbers in September 2008, using July 2008 data, total domestic uninsured deposits were $2.1 trillion. Which seemed like a lot of money at the time, but it’s gone up by a full trillion dollars since then.
All evidence, then, points to the fact that US bank depositors aren’t worried about the safety of their deposits at all, and that they believe — correctly — that US banks, in general, are a perfectly safe place to park their funds.
But Bhidé’s not happy with that: he wants short-term deposits to be guaranteed just in the way that paper currency is guaranteed. But here’s the thing: the guarantee of paper currency is meaningless, because paper currency isn’t a government liability in the way that a deposit at a bank is a bank liability.
You can walk up to a bank and ask for the money you have on deposit to be converted into paper currency, and the bank has to give it to you in cash. But if you have cash and you walk up to Treasury, there’s nothing the government is obliged to give you in return. That’s the whole point of fiat currency: it is what it is. You can buy stuff with it — you can even buy a Treasury bill, and turn your cash into a US government obligation. But when that Treasury bill matures, it just becomes cash again.
And the fact is that we really don’t want a blanket guarantee of bank deposits in any event. Bank deposits are dangerous things — they’re informationally-insenstive assets which do a really good job of housing tail risk in an invisible and impossible-to-measure manner. If there were a blanket deposit guarantee, you can be quite sure that total domestic deposits would rise substantially from their current $8.5 trillion, and I’m not at all sure that I want to give trillions more dollars to the US banking sector, which has proved itself time and time again a very bad custodian of such funds. Bhidé seems to think that you could regulate away any risk: that’s naive in the extreme. If banks don’t take risk, they’re not banks any more.
Sometimes, banks will fail. That’s a feature, not a bug: it’s necessary to impose discipline on them. In Bhidé’s utopia, banks are so stringently regulated that they never fail. That places an impossible burden on regulators, and infantilizes the important capital-allocation function that banks provide in the economy.
“The next time a panic starts,” writes Bhidé, “markets may just not believe that the Treasury and Fed have the resources to stop it.” He’s right about that. So what makes him think that markets will believe a blanket deposit guarantee? If you guarantee everything, you guarantee nothing. Let’s keep private banks private. Because as Peter Thal Larsen says, the logical conclusion of what Bhidé wants is the nationalization of every bank in America. And even François Mitterrand never went that far.
Update: This page says there’s a total of $6.8 trillion in insured deposits, although it doesn’t say how many uninsured deposits there are. And it’s worth answering the point that Ireland guaranteed all bank debt, not just deposits. Which is true. But if there’s an unlimited government guarantee on bank deposits, then banks will simply fund themselves through deposits and not through bank debt at all.