Thinking About the Bank Bailout

I have been asked the following question: Should those who supported the bailout at the time believe now that it was a mistake? I don’t think so, but I think that there are important lessons we should learn from the past few months that should inform changes going forward.

I think the place to start is to recap the logic for the bailout at the time, and then review what new information we have now that we did not have in 75 – 90 days ago.

1. What was the original logic for supporting the bank bailout?

My reason for supporting it was that I believed that we were in an emergency situation, and that absent intervention we would face a serious risk of financial contagion and a resulting economic catastrophe. I argued that a depression-level event was less-likely-than-not even without a bailout, but that the risk was real, and would likely be reduced by the bailout. I very crudely estimated that the odds of this happening were something like 1-in-4 with no bailout, and something like 1-in-10 to 1-in-20 with the bailout. I further argued that the ideological costs created by this program would be numerous, severe and long-lasting. Weighing these against the risks of contagion, I concluded that the bailout was “the worst possible course of action, other than any available alternative”.

2. What has been the structure of the bailout as it has actually taken place, and how does this compare to the original expectations?

Paulson has invested $335 billion of the first tranche of $350 billion approved by Congress. This is not spending, but is mostly the purchase of bank equity. As such, we would expect to get most or all of it back when the government eventually sells these holdings. The biggest surprise here is that no investments have been made by buying specific troubled assets from banks under reverse auctions, which was how the plan was originally described. Acting within the incredibly broad mandate available to them, the administrators have instead directly purchased equity stakes in banks.

The basic idea of the original asset purchase plan was to create something like a good bank / bad bank structure that would enable the healthy parts of banks to resume normal operations, while the government provided liquidity to the “bad bank” assets while they were unwound. Presumably, this structure was found not to be practical in the short-term, and the alternative approach of equity injections was pursued. Many, many very smart economists have argued all along that this is more efficient. Maybe it is. Of course, this approach creates a real possibility of ending up with “zombie banks” that limp along, but can’t really do their job of making loans very well because they have marginal balance sheets and worse credibility. This is very much like what led Japan to have an essentially stagnant economy for the entire decade of the 1990s.

3. What has it accomplished, and how does this compare to the original expectations?

The financial crisis has not (yet) kicked off financial contagion and turned into a massive economic crisis. I argued at the time that this was the most likely outcome whether or not we had the bailout, so it is very hard to know what causal impact the bailout has had to date. Even the point that we would be in the intellectual limbo of not knowing whether or not it worked in the likely event of no meltdown was foreseeable at the time.

Scholars will be arguing about this question for decades, but the information we have now versus our knowledge at the time of the decision raises my estimate of the odds that the bailout helped prevent contagion. The basic indicator that I’ve always used as a quick-and-dirty measure of contagion risk is the TED Spread. Here is what has happened with this interest rate indicator:

There clearly was an epic dislocation in the credit markets. We have passed through at least the current stage of the credit crisis, and are back to something like the elevated level of interbank interest rates that we experienced earlier in the year. This is exactly the kind of pattern we would have expected to see if the intervention was both needed and successful, at least in the short-term.

On the other hand, it is unclear that the intervention has actually gotten banks to lend much again. It is very hard to get good, comprehensive data on this, but there is at least informed anecdotal evidence that banks are not responding to these equity injections by lending more. Now, of course, some of this is healthy – we want credit to be harder to get than it was for, say, a marginal homebuyer in Florida in 2005 – but ultimately we need a sustainable credit stance that will enable economic growth.

4. What costs has it imposed, and how do these compare to the original expectations?:

The widely-discussed “bailout nation” problem was predictable. In the first thing I ever wrote about it, I described four huge problems that would likely proceed from the Paulson / Bernanke approach. One of which was the likelihood that we would see many future requests for bailouts, saying that voters:

…will likely demand greater socialization of consequences of reasonably-foreseeable bad behavior by people who don’t make a million dollars per year. The ideological consequences of the last few weeks will take many years to play out, and conservatives are unlikely to happy about them.

But what has surprised me is how quickly Detroit came after the money. It only took a few weeks. I thought it would take longer for this problem to manifest itself at this level.

5. Summary and Conclusions:

The bailout appears likely to have helped stave off the immediate emergency we faced less than three months ago. It is very unclear how relatively fragile the credit situation is today versus September and October. While interest rate spreads have declined, it may be that Treasury knows of or suspects specific hidden weaknesses like credit default swaps that might put us right back to four-alarm-fire mode any moment. The endless stream of huge bailout requests has come on more quickly than I expected, and appears to be contributing significantly to the momentum behind incredible spending plans, especially stimulus plans, by the new administration. The equity injection approach that has been used to date runs the risk of avoiding the necessary restructuring and financial pain that will be required to get the financial sector healthy again; and there is at least anecdotal evidence that we may be developing exactly such a zombie bank problem right now. In sum, the bailout so far has been painful, expensive and sloppy, but we’re still in one piece.

Congress has the right to approve or deny any request that Treasury might make for the second tranche of the additional $350 billion. This should be treated as a separate request. We have the luxury of time, as compared to a few months ago, to vet this request with far greater rigor, and in light of what we have learned. Specifically, before authorizing this money, Congress should perform its oversight function, and demand to know: (1) what underlying risks, not to shareholders or employees, but of systemic financial collapse now exist or are latent that would justify this much money, and (2) how we will avoid the zombie bank problem, including potential application of lending requirements in return for capital, as has been done in the UK.

I supported the original bank bailout, and continue to believe that it was a painful, but correct, decision. However, supporting the next tranche will require a lot of convincing.