Arnold Kling  

Tyler Cowen on Bank Bailouts (Again)

Against Human Weakness... More on Monopoly...

He writes,

Two years ago, my view was this: if there is a banking crisis, the Fed should be loose with money and let the market sort out which banks deserve support and which do not. And stop there. When Bernanke let Lehman go, I remember being happy and thinking he finally had "real guts." I now view that attitude as mistaken, as I had not forecast how badly some credit markets would freeze up, most of all the collapse of repo as analyzed by Gary Gorton. Two years ago I also had not imagined that the U.S. economy could end up with so many insolvent banks at once. I had thought that the presence of "real money on the line," from bank CEOs, would stop such an outcome. That was wrong too.

Read the whole thing, which frames the issue of bailouts really well. Note that his reference in the addendum to "Arnold Kling comments" is not to what I am writing here, but to a previous blog post of mine. My response to Tyler's post is below.

Let me frame the question this way: When, if ever, should you allow the financial sector to shrink? Possible answers:

1. Never. Citigroup, Freddie, Fannie, and AIG are not dead. They're just restin'. Pinin' for the fjords. I characterize Gary Gorton this way.

2. Kill 'em all and let God sort 'em out. I characterize this as what Tyler is afraid the alternative to bailouts would have looked like.

3. Lord make me chaste, but not yet. That is, you want the financial sector to shrink at some point, but not in the middle of a crisis. I characterize Tyler this way.

4. Triage. My preferred approach, which gets some shrinkage but not the end of the world.

My preferred banking policy would have worked this way.

1. You look at the balance sheet of a bank. You allow some assets to have a range of values. The low value is the current market value of the asset. The high value is what you think that asset would be worth if the market were not in a state of panic.

2. If a bank is insolvent even assuming assets are valued at the high range, then shut it down. Go through the appropriate process (FDIC, bankruptcy, conservatorship), and deal with the consequences. My guess is the Citigroup, Freddie, Fannie, Bear, Lehman, and some others fall by the wayside under that scenario. However, my guess is that fewer than 5 of the top 20 commercial banks would have had to be closed. I assume that AIG winds up in this category, with probability .75.

3. If a bank is clearly solvent even if assets are valued at the current market value, then say so out loud. Be prepared to lend to the bank in case of panic. My guess is that between 3 and 8 of the top 20 commercial banks were in this good shape.

4. If a bank is solvent assuming assets are in the high range but not if they are in the low range, then put it under close supervision. Lend to keep it operating, but freeze its assets--no new risk taking. The bank can roll over existing loans (continue to finance businesses that it already is financing) and honor its letters of credit, but it cannot make new commercial loans. My guess is that about half of the top 20 commercial banks would have been put into this mode. Maybe AIG winds up in this category, with probability .25.

The banks under (4) might have needed a lot of loans from the central bank to deal with liquidity issues, in part because they will have assets tied up in the bankruptcies of the firms that fall under (2). That is ok. I'd rather be lending to banks that are close to solvent than bailing out banks that are clearly insolvent.

The biggest problem with my scenario is with foreign banks. Other countries do not have as many banks as we do, and my approach does not shovel as much money at them (as we did in the AIG bailout, for example). I can't promise you that my approach keeps the German financial system from collapsing, and that would not be a good thing.

I think we should not have put off shrinking our financial sector. The result of the bailouts is that we are maintaining credit markets based on false information and artificial prices. You may have pulled the airplane out of the dive, but you are flying with faulty instruments, and I don't think you are going to be happy about where you wind up.

Comments and Sharing

TRACKBACKS (1 to date)
TrackBack URL:
The author at in a related article titled Samizdata quote of the day writes:
    "I think we should not have put off shrinking our financial sector. The result of the bailouts is that we are maintaining credit markets based on false information and artificial prices. You may have pulled the airplane out of the dive, but you are fly... [Tracked on September 2, 2009 10:47 AM]
COMMENTS (12 to date)
David writes:


Was there time to apply this approach?

El Presidente writes:

I mostly agree with Tyler's sentiment.

I think your caveat in your second to last paragraph is an illustration of the type of objection that is bound to be raised and cannot be answered definitively. Kudos for incorporating it in your remarks. It's balanced and it prompts thoughtful deliberation.

Acknowledging that these are more matters of judgement than calculation, I think it's important to evaluate the costs of alternatives in terms of both 'how much' and 'for whom'. Sometimes, like Bernanke has indicated, you redeem the irresponsible so that you protect the responsible from suffering the negative externalities you failed to prevent in the first place. It's a bitter pill, but it resonates with a valuable nurturing impulse. If we are all discipline and no nurture, we lose balance and things get out of whack. Your suggestion strikes a balance. I simply wonder whether it is the best one. It's certainly a reasonable one.

Arthur_500 writes:

With all the calculations we use for economics we never get away from the reality that people act in irrational ways. Confidence in our markets makes the capitalist system work and will break it.
When we learn that a financial insistution is on rocky ground we will avoid doing business with them and that will drive them under - to a point. Just as we shop at the local grocery store we will continue to keep our money in the local bank and pay our loans to that local bank. Many of us will never know our loan is now owned by someone thousands of miles away. Of course we have confidence in the system because it is guaranteed by the FDIC/ Fed Reserve.
What destroys confidence is knowing that the bank is a loser and seeing our taxpayer dollars prop it up. We feel helpless. We don't want to do any more business with them than is absolutely necessary. We get higher interest from the pillowcase than from those taxpayer funded institutions and that makes us angry. We want justice and we want to see them fail - NOW.
If the financial system operated as it was designed, rather than special interest bailouts, those institutions would have to re-organize. There are levels of reorganization available to them and they could be broken up and sold off. There would be time to do this because there has to be time to do this. The courts would assure there was time to properly deal with the assets and oversee the liquidation of the liabilities.
But then, you would have to believe that our courts and businesses and common people like me are as capable or even more so than an elected politician.

Robert Simmons writes:

I posted this comment at MR, but want to get your take on it. I see it as very similar to your #4.
Why can't lending at penalty rates combined with regulatory leniency work for insolvent institutions? They'll eventually go bankrupt, but this would put off the day of reckoning a while and allow for orderly renegotiation and trading of claims and such.
If the placement in the capital structure is done right, depositors wouldn't be hurt, and we wouldn't lose money on the deal.

ao writes:

By announcing which banks are in group #3, and taking action on banks that are in group #2, you also implicitly announce which banks are in group #4 (perhaps you would even advocate explicitly stating this). But by doing this, won't you push banks currently in group #4 into group #2? In the old model of a bank run, the widespread belief (either true or false) that a bank is insolvent can lead to insolvency, thus the belief is a self fulfilling prophecy. What happens when you announce that a bank is almost insolvent? Sure you allow them to roll over their existing loans, but what rate of return do lenders require on almost insolvent banks? Probably a rate high enough to push them into insolvency just like an old fashioned bank run (this is the Gary Gorton story, I think). My guess is that group #4 is not a sustainable group and that you will de facto create only two groups, solvent and insolvent. By your accounting this would leave 15/20 banks insolvent.

Shayne Cook writes:

Tyler poses the question: "If you disagree with me on bailouts, a simple starting question is this: without the bailouts, and with loose monetary policy only, how many of the twenty largest banks do you think would have ended up in bankruptcy court within a fairly narrow time span? Until you've addressed that question, we're not getting to the bottom of the substantive issue."

I would answer his question thus: "Every bloody one of them that had earned their right to be in bankruptcy court, through defective business practices!"

Further, I would argue Tyler's question is not the question to answer in order to "get to the bottom of the substantive issue." Like the bailout itself, it focuses far too much attention on the failed institutions, instead of focusing attention on the banks that were in no real danger of failing.

The "substantive issue" is how best to support viable banks during a crisis - based on viability, not size (too big to fail) - AND both use their strength and elicit their help in minimizing ill effects. Emphasis on viability and viable banks. I sense that is Arnold's preferred conceptual framework as well.

q writes:

would you allow banks in your 'can't take on new business to':

-- hedge existing positions? (and if so on what dimensions -- you have to specify in some way or it's infinitely gameable)

-- allow them to write new contracts with counterparties who want to monitor their counterparty risk (ie if a counterparty is suddenly exposed to the institution due to a change in an existing contract, would you allow them to write a contract which hedges this)?

q writes:

also, can you clarify what you mean by 'shrinking the financial sector'? on what dimension? number of banks? profitability? square footage? what do regulators have control of here?

Drewfus writes:

The consensus that letting Lehman Brothers fail was a mistake, has no supporting theory or evidence to back it up. But the consensus in itself creates a terrible precedent - that next time, absolutely every company that fits within the subjective border of 'too big to fail' will be bailed out. The propensity for reckless behaviour, owing to moral hazard, has been increased enourmously.

This is very bad. We now have a society moving away from the principle of reward for success, and towards the principle of reward for failure. This inversion of the natural order can only lead to disaster.

People like Cowen are simply taking a narrow, short-sighted view of the situation. We desperately need better thinking.

axg writes:

Your approach depends on government being able to assess the relative health of banks ("You look at the balance sheet of the bank..."). I do applaud the caution and caveats you apply to your subsequent analysis.

But please read todays' report about SEC probes of Madoff. The SEC obviously isn't viewed as being that disfunctional, as witnessed by the fact that it is being allowed to continue basically unchanged into the future. So please tell me, do you there is a government organization that even remote approaches the intelligent discrimination your "preferred bank policy" depends on. If so, what, and upon what basis to you hold this belief?

Bill Woolsey writes:


Self-fulfilling expecations or even knowledge of insolvency won't push FDIC insured banks into insolvency.

This is all about a handful of giant, politically powerful wall street investment banks. Most of them were operating what amounted to giant thrift operations, borrowing by issuing quasi-deposits in the form of commercial paper, some of its as short as overnight, and then lending in home mortgages, though held indirectly in the form of mortgage backed securities. They had lent into a speculative bubble in housing. The proposal is to bail them out. Cowen is prposoing that the Fed bail them out by lending to them when they can no longer issue commercial paper. Of course, the Treasury bought stock in them, and there were various proposals that the Treasury buy the mortgage backed securities from them.

The commercial banks have a huge proportion of their portfolios in mortgages and also own a lot of mortgage backed securities. While many may be insolvent because they lent into a bubble, they will only close down if FDIC makes them close down. If FDIC were to just close down insovlent FDIC insured institutions, and leave them closed with all the depositors' money unreachable, and then liquidated them or reorganized them gradually, then there could be an extended period of time where many banks ceased to operate. That would be stupid.

The more reasonable scenario is that FDIC insured banks continue to operate even though they are insolvent, and FDIC does overnight reorganizations of them as fast as it can, creating healthy banks in the aftermath.

Now, having FDIC is a massive government intervention. If the alternatives are to have the Fed bail out investment and commercial banks, or else abolish FDIC on the spot, close all the insolvent banks immediately and have them reorganized through the bankrupcy courts, then then the first looks like a good idea.

However, FDIC is operating. The Fed bailouts of investment banks is an additional intervention piled onto the others.

Drewfus writes:

If the only way that depositors can be protected is by keeping insolvent banks operating, then then system is poorly designed.

Insurance of deposits needs to be completely decoupled from the continued functioning of banks. Insolvent banks should be closed, like any other failed business. Depositors should be paid out by the FDIC using fiat money. As the banks assets are gradually liquidated, an equivalent amount of base money should be withdrawn from the system.

That way, depositors are protected, there is no loss of money supply, but no moral hazard either. Banks that take poor risks suffer the consequences. The problem with the current system is that one function is dependent on another, with very adverse consequences.

Insure deposits, not depositors accounts!

Comments for this entry have been closed
Return to top