Babies with Candy III

Commenting on Krugman's latest offering:

Comrade Greenspan wants us to seize the economy’s commanding heights.

O.K., not exactly. What Alan Greenspan, the former Federal Reserve chairman — and a staunch defender of free markets — actually said was, “It may be necessary to temporarily nationalize some banks in order to facilitate a swift and orderly restructuring.” I agree.

I don't. The idea is, on the surface, appealing. As was Paulson's asset buyout plan, and for the same reason. The promise of a single (more or less) bold, swift, unapologetic government action, which will stop the bleeding and right the ship. But, as I'll describe below, upon anything close to rigorous reflection, its hard to conclude that it stands much chance of working as advertised.

The case for nationalization rests on three observations.

First, some major banks are dangerously close to the edge — in fact, they would have failed already if investors didn’t expect the government to rescue them if necessary.

Second, banks must be rescued. The collapse of Lehman Brothers almost destroyed the world financial system, and we can’t risk letting much bigger institutions like Citigroup or Bank of America implode.

Third, while banks must be rescued, the U.S. government can’t afford, fiscally or politically, to bestow huge gifts on bank shareholders...

Let’s be concrete here. There’s a reasonable chance — not a certainty — that Citi and BofA, together, will lose hundreds of billions over the next few years. And their capital, the excess of their assets over their liabilities, isn’t remotely large enough to cover those potential losses.

Arguably, the only reason they haven’t already failed is that the government is acting as a backstop, implicitly guaranteeing their obligations. But they’re zombie banks, unable to supply the credit the economy needs.


While Krugman's observations are entirely correct, the conclusion he draws will not directly follow. He -- and this appears a tendency amongst liberal commentators -- conflates the two objectives of supplying banks with public capital.

As he noted, without the public backstop, these banks would likely implode. From the Lehman experience, we understand that the public has an strong interest in avoiding said imposion.

The second objective is that the banks need more capital "to supply the credit the economy needs."

The amount of capital required for the second is far in excess of the amount required for the first. Zombie banks may not juice the economy, but they don't threaten to destroy the global financial system the way a disorganized bankruptcy would.

Additionally, as I argued previously on this blog, claiming that credit is what the economy needs is akin claiming that alcohol is what Mickey Mantle needed. We are in economic pain due to systematic over-reliance on credit. For our long-term economic health, we need credit-withdrawal, not another fix.

To end their zombiehood the banks need more capital. But they can’t raise more capital from private investors. So the government has to supply the necessary funds.

But here’s the thing: the funds needed to bring these banks fully back to life would greatly exceed what they’re currently worth. Citi and BofA have a combined market value of less than $30 billion, and even that value is mainly if not entirely based on the hope that stockholders will get a piece of a government handout. And if it’s basically putting up all the money, the government should get ownership in return.

As discussed, Krugman pulled a logical fast one. He substituted the factual observation that we can't let Citibank go the way of Lehman, with the less solid assertion that we need to restore Citi fully back to life.

He is correct in noting that these banks cannot be recapitalized privately. He does not explain why this is the case. The reasons include: regulations that limit who can invest capital in banks and, more importantly, private capital is scared off by the spectre of nationalization.

Put differently: if the government formalized its role as backstop, which is required in any event, guaranteed that it would not punish new private capital, and reduced restrictions on where that capital could come from at least some banks would likely be able to recapitalize privately.

Tangentially, it seems like much of Obama's policy has reflected this pattern: He pursues some policy which restrains, or crowds out, private activity, and then uses that private inaction to justify dramatically increased government intervention. I am unsure if this is innocent or intentional, but its hard to imagine that Obama suddenly lost the rigor and discipline that characterized his campaign.

In any case, the "should" in "should get ownership in return" is, then, perhaps political necessity, but its not simply in the public's economic interest.

Still, isn’t nationalization un-American? No, it’s as American as apple pie.

Lately the Federal Deposit Insurance Corporation has been seizing banks it deems insolvent at the rate of about two a week. When the F.D.I.C. seizes a bank, it takes over the bank’s bad assets, pays off some of its debt, and resells the cleaned-up institution to private investors. And that’s exactly what advocates of temporary nationalization want to see happen, not just to the small banks the F.D.I.C. has been seizing, but to major banks that are similarly insolvent.

The real question is why the Obama administration keeps coming up with proposals that sound like possible alternatives to nationalization, but turn out to involve huge handouts to bank stockholders.

As Krugman notes later, the phrase "nationalization" is inapt if what is being advocated in a government ordered liquidation akin to what the FDIC does. Nationalization implies something different.

The idea that the government can, with the insolvent big banks do what the FDIC does with small banks -- in effect government-ordered bankruptcies-- bears little srutiny.

For one, if I understand correctly, the FDIC seizing is generally less akin to chapter 11 then chapter 7, which is to say, it more generally liquidates bank assets and pays back debt-holders in an orderly fashion, then restructures failed banks into viable ones. The former, obviously, requires far less business savvy -- and therefore is far more safely entrusted to the government -- then the latter. The nationalization Krugman appears to have in mind is very much the latter, and therefore the FDIC analogue-justification inapt.

Furthermore, consider that there are some 8000 FDIC insured banks in the US. Two banks seized a week adds up to a little over 1% a year. Even with the increase in bank failures the FDIC has a robust private market into which the it sells the institution or its assets.

Imagine, by way of comparison, of what would happen if the FDIC was seizing 50 banks a week. There would be at least two easy to foresee effects:

One, it would have far less ability to move seized institutions back into private hands. More likely it would be "forced" to operate those institutions for an indefinite amount of time.

Secondly, seizures are more likely to snowball. So long as seizures number 1% per year, they serve -- by avoiding fire-sales -- to strengthen healthy banks. Seizing 20% a year may well -- by undermining the sector -- serve to weaken healthy banks.

The same is true of Citibank and BofA. Any nationalization is unlikely to be particularly temporary and nationalizing Citibank and BofA may predictably weaken Goldman, Morgan and JPM to the point where they too "require" nationalization.

For example, the administration initially floated the idea of offering banks guarantees against losses on troubled assets. This would have been a great deal for bank stockholders, not so much for the rest of us: heads they win, tails taxpayers lose.

Now the administration is talking about a “public-private partnership” to buy troubled assets from the banks, with the government lending money to private investors for that purpose. This would offer investors a one-way bet: if the assets rise in price, investors win; if they fall substantially, investors walk away and leave the government holding the bag. Again, heads they win, tails we lose.

Krugman's arguments here are blatantly dishonest. Earlier he observed the motivation for government involvement as avoiding a Lehman style implosion and Japanese style zombie banks. As that is the case, taxpayers wins and losses are measured by whether implosion is avoided and credit starts flowing again.

Also the public-private partnerships being discussed are not the simplistic heads they win, tails we lose scenario Krugman dishonestly describes.

To take a concrete example, lets say there is an asset that right now a bank will sell for $100, but outside investors would only buy for a far lower price, say $25. Which leads to it being stuck on the bank balance sheet inhibited lending.

The government is offering to finance ~90% of the purchase price of such assets which dramatically raises the price outside investors are willing to pay (the cost to the investor is now 10% * $100 + interest on $90). If an investor is now willing to buy the asset, the bank has cash instead of a toxic asset on its books and the amount of money it can lend out is increased (which is, for Krugman, a win).

The investor has put down $10 of their own money and borrowed $90. So long as the asset retains most of its value the investor will repay the loan and the government profits. If the asset value drops more substantially, the investor may walk away from the loan and leave the government holding the asset, which it, in effect bought for 9/10ths the asking price. The asset may well, at that point be worthless, but it also may have the capacity to recover with the government now owning the upside.

Contrast this with the nationalization Krugman desperately prefers. If the government seizes the bank, it now owns this asset. To restore the bank to health -- Krugman's avowed goal -- entails either writing the value of the asset down to near zero or selling the asset into the market. (Either action is likely to create further downward pressure on still-surviving banks.) The market is only willing to buy the asset at $25. (In the case of the nationalized Swedish banks, the government actually offered guarantees and subsidies to the buyers of some assets.) To keep the bank well capitalized (its assets safely greater then its liabilities) the government would then have to make up for the $75 loss with public money, only some of which it would get back if it sold the bank back into market. If it writes the asset value down to near zero, and the perceived riskiness of the asset drops substantively while the bank is nationalized, the government does stand to make some return. Still, it will likely be a long time before the perceived riskiness of most of these asset drops substantively.

In sum, there is no strong reason to believe that nationalization stands to be a better deal for the taxpayer then alternatives that promise roughly the same outcome.

Why not just go ahead and nationalize? Remember, the longer we live with zombie banks, the harder it will be to end the economic crisis.

I am not sure what he wants us to remember. He never explained why this is the case, and I rather suspect it is not the case.

The longer we live with zombie banks, the more we learn to live in a world of constricted credit. The economy that will emerge from that world will be far more sustainable then the product of free-flowing credit.

How would nationalization take place? All the administration has to do is take its own planned “stress test” for major banks seriously, and not hide the results when a bank fails the test, making a takeover necessary. Yes, the whole thing would have a Claude Rains feel to it, as a government that has been propping up banks for months declares itself shocked, shocked at the miserable state of their balance sheets. But that’s O.K.

The stress test is, likely, more about politics then economics. In practice, the outcome of a stress test is controlled by embedded assumptions. It will be easier politically to seize a bank that fails the stress test.

And once again, long-term government ownership isn’t the goal: like the small banks seized by the F.D.I.C. every week, major banks would be returned to private control as soon as possible. The finance blog Calculated Risk suggests that instead of calling the process nationalization, we should call it “preprivatization.”

And once again, this is either foolish or dishonest. If the government seizes Citibank and BofA it will likely be "forced" in short order to seize the rest. And if the government seizes the sector its hard to imagine that long term government ownership isn't the most likely outcome.

To put my money where my mouth is, I'll offer Paul Krugman the following $1000 bet straight-up. If the government seizes Citibank and BofA, five years later, Morgan Stanley will be bankrupt or in government hands.

The Obama administration, says Robert Gibbs, the White House spokesman, believes “that a privately held banking system is the correct way to go.” So do we all.

Gibbs finished that phrase by calling for government regulation. The difference between regulation and ownership is a fuzzy one. Its not clear to me what control the government would have as owner that it does not have as regulator.

I suspect Obama understands this, which is why he isn't eager to nationalize the banks. He can control the banks entirely without bearing the political cost of nationalizing them.

But what we have now isn’t private enterprise, it’s lemon socialism: banks get the upside but taxpayers bear the risks. And it’s perpetuating zombie banks, blocking economic recovery.

What we want is a system in which banks own the downs as well as the ups. And the road to that system runs through nationalization.

We own the downs because we are overly dependent on the overflow of credit the banks provide. If the banks suffer, they tighten the credit we can't live without. The road to that system, then, leads to an economy that is less dependent on credit. Alternatively, perhaps, the road to that system leads to an economy in which the flow of credit is less dependent on the health of a few large banks. The reasons it does in our economy are less market or structural and more legislative/regulatory in nature.

Its hard to see how the road to either place runs through nationalization. It is hard to see any sensible argument explaining nationalization gets us to a place where banks own the downs.

Finally, to krugman, zombie banks block economic recovery because they don't extend the credit the economy needs. It is illuminating to review precisely how that works. Crudely: the large banks take cash from lenders and investors, they, generally, "lend" by creating securities backed by loans (which they either originate themselves or support a market for originators), which they mostly sell of to investors in exchange for cash, which they can put right back to work. Part of what is going on now is that there is a dramatically reduced market for these securities. There are a number of reasons for this including: the opacity of some of the most complex derivatives, dramatically increased sale of government-debt securities crowding out the market for private-debt securities, and, perhaps, above all, the market questioning the credit worthiness of borrowers.

This is, in other words, another question and egg problem. The return of credit requires the return of debt securities markets. It is unclear that it is zombie banks causing frozen debt security markets as opposed to frozen debt security markets causing zombie banks.