The Good Bank

I liked the idea of this proposal when I first read it, and as Buiter points out, it has been floating around in various forms both before and since his original comments. I thinkt he idea of starting from scratch with a good bank funded by the government makes a lot more sense than trying to sort out how to prop up the system we have by creating a bad bank to take over the toxic waste. I mean, from the perspective of the public, all we need is a functioning banking system. This is why the existing banks are too-big-to-fail -- because it is impossible to replace the Ponzi scheme intermediation of the banking system overnight, not because they are somehow magical.

When you look at it from this larger perspective, in which you see banking as a relatively low-risk, public-good, utility-like business that shuttles capital from mom and pop's pocketbook to the guy who wants to build a factory, you wonder why this isn't a perfect opportunity to get into the banking business. The competition has its hands tied dealing with its earlier excess. No one is making loans despite the cost of money being essentially zero. Sounds like easy money, no? And you already do see this, only the government's clumsy lumbering around the sector changing the rules around every five minutes doesn't help. It's already a big undertaking to start a bank, and if you don't have any idea what the landscape will look like in 6 months because you know who the government will bail out or how, you will be reluctant to go into this business. The government could solve all of this in one swell foop by itself going into the banking business in partnership with private capital, which is what Buiter proposes.

The logic is simple. Many (probably most, possibly all but a handful) high-profile, large border-crossing universal banks in the north Atlantic region are dead banks walking - zombie banks kept from formal insolvency only through past, present and anticipated future injections of public money. They have indeterminate but possibly large remaining stocks of toxic - hard or impossible to value - assets on their balance sheets which they cannot or will not come clean on.

This overhang of toxic assets acts like a tax on new lending. Banks are required, by regulators or by market pressures, to hoard capital and liquidity rather than engaging in new lending to the real economy. The public financial support offered in the form of capital injections (in the US mainly through preference shares and other non-voting equity), guarantees for assets and for liabilities (old and new), insurance of toxic assets (as provided to Citigroup by the US sovereign) and possibly in the future through direct purchases by the state of toxic assets (using TARP money in the US) and the creation of one or more publicly owned 'bad banks' has been a complete failure.

The bad bank proposals the Obama administration and other governments are considering are non-starters, for the simple reason that they require the valuation of assets whose true value (even on a hold-to-maturity basis) can only be guessed at. The good bank proposal only requires the valuation of those assets on the balance sheets of the existing banks that are easy to value: transparently valued assets. The toxic stuff is left on the balance sheet of the existing banks, which become the legacy bad banks.

Offering to pay enough to the existing owners of the toxic assets to induce these owners to sell them would require paying over the odds. That might not leave enough fiscal resources to support the new lending activities that are so urgently needed. It would also be an unfair and moral-hazard-maximising bail out of the existing owners and creditors of the banks. Nationalising the dodgy banks (or even the entire cross-border universal banking sector) would only solve the valuation problem of the new owner (the state) after nationalisation. The toxic assets could be transferred into a bad bank at any valuation, including zero. The owner of the bad bank and of the cleansed bank are the same. Nationalising the dodgy banks would not solve the problem of how much to pay for the banks, however, because that would depend in part on the valuation of the toxic assets. The good bank proposal creates new, publicly owned banks which only purchase good assets from the legacy bank. There is no valuation issue involving toxic assets for the tax payer.

...

Two things are systemically important. The first is to restore the operation of key financial markets that have become illiquid. The Fed is doing a reasonable job in that regard. The second is to restore bank lending to the real economy. Neither objective requires that the existing banks be saved, let alone that their existing shareholders and creditors receive any financial support from the state. We can save banking without saving the banks or the bankers. The 'good bank' proposal demonstrates how to do this.

Feliz Salmon objects to this idea, but I don't find what he says terribly convincing.

For one thing, it's far from clear that there's a shortage of "good banks" in the US right now. One of the big differences between the US and Europe is that the US has many more banks; even the huge ones don't have anything like the market share that big European banks do, and there are thousands of small and tiny banks -- not to mention credit unions -- which have really no European counterparts at all. Many of them -- and even many pretty big banks -- are doing fine.

Putting government capital into new "good banks" would create unhelpful competition for the existing good banks, while providing no help whatsoever for the existing bad banks. Remember that the single biggest problem with too-big-to-fail banks is that they can't be allowed to fail. So even if setting up good banks was a good idea, it would be an insufficient idea, since the big bad banks would still need to be rescued: there's simply no way we can afford a big bond default by the likes of Citigroup or Bank of America, let alone even hint that their uninsured depositors might be in danger.

While it's true that there are good banks, they are finding it tremendously difficult to take advantage of the opportunity afforded them because no one trusts and banks at all any longer. The government good bank could also be structured as essentially just a big pool of capital that lets these banks expand dramatically in a short period of time. That is, instead of competing with the existing good banks, the government could be partnering with them in a transparent fashion. This would be precisely the opposite of "unhelpful competition".

As to his second point, we have to really define what we mean by too-big-to-fail. Why can't they fail? What would happen? Yes, their equity would be worthless. And their bondholders would have to try a recoup their money from the bankruptcy, but so what? I see three objections to letting these banks fail:

1. Depositors get screwed -- There hasn't been a run on these banks by depositors even though their insolvency is pretty common knowledge at this point, and anyway the problem with a run on the banking system is that the money comes out completely and goes into a sock. If the run simply takes the form of moving your deposits to another account, how is this a systemic problem?
2. They default on their obligations to other institutions -- There is no question that Citibank announcing that they cannot pay a bond would have an impact on the market. But why does this have to have such an impact? Clearly the equity can got to essentially zero without total armegeddon, given that they equity is already at zero, give or take. Whoever holds the bonds would need to write them down in the case of a bankruptcy. But by how much? From a systemic perspective, the problem with a bank going bankrupt is that it has to liquidate its loan portfolio (including the good loans) at fire sale prices, which, if everyone is doing it at the same time, can result in a general deflationary environment that spirals out of control. But what if they didn't have to liquidate in a hurry. What if the government manages the bankruptcy by taking all of the easy to value assets off the books at good prices? No good loans are called in quickly to meet obligations of the bankrupt estate because the new good bank is happy to make these loans. The bondholders can be paid off from the proceeds of selling these good loans. Any losses to the higher unsecured tranches or the quity will be realized over time as in any bankruptcy. What's the big deal?
3. New loans don't get made -- We still need somone to go out and make loans and roll over existing good loans. That's the whole point of having a new good bank though. It's not like Buiter, Romer or Soros are liquidationists that propose doing nothing. They simply propose to focus entirely on what is systemically important about the banking system so that its insolvency doesn't force us needlessly stop making as much stuff as we are.