Four Ways of Looking at a TBTF Subsidy: A Reply to Dean Baker

October 17, 2015

The discussion over a Too Big To Fail (TBTF) subsidy, where the largest banks are able to borrow more cheaply as the result of potential future bailouts, is back in the discussion. Paul Krugman referenced it with a link to my review of two studies arguing the subsidy has largely declined since the crisis. Dean Baker has responded with critical thoughts on the studies.

My point isn’t to say that the subsidy is completely over. Nor, as I’ll explain in a bit, is it to say that TBTF is over. Instead, understanding this decline lets us know we should push forward with what we are doing. It debunks conservative narratives about Dodd-Frank being fundamentally a protective permanent bailout for the largest firms that we should scrap, and provides evidence against repealing it. And ideally it gets us to understand this subsidy as just one part of the more general TBTF problem that needs to be solved.

I’ll first respond to Dean Baker. Then I’ll map out four different ways of understanding what we mean by a TBTF subsidy, and what is and isn’t fixed, because that might clarify other responses I’ve been getting.

A Response to Dean Baker

Dean’s core argument that the subsidy wasn’t measured in 2006, so something is off. “Obviously, the big banks did enjoy too big to fail protection in 2006, since only Lehman was allowed to fail in the crisis, yet the GAO analysis implies that this held very little value.” Why is this? According to Dean, the subsidy collapses to zero in good times, but explodes in bad. So you can only measure it in the crisis.

Three responses to this. First, we do know that TBTF subsidies are real, relevant and measurable with CBO technology in good times. We know this because they could easily see them for the GSEs. Here’s CBO giving 2001 testimony on the subsidies it found and measured in the capital markets for the GSEs, where GSE debt is “more valuable to investors than similar private securities because of the perception of a government guarantee.”

If a bailout guarantee falls to zero value for TBTF firms in good times, it should have also fallen for the GSEs here, where similar methodology is being deployed. But it didn’t.

Second, if it only exists in moments of crisis but doesn’t change any actions in the capital markets, it’s a different kind of subsidy than the normal “permanent bailout” language, and has an order of magnitude less of an effect in terms of market distortion in good times. I actually think we should think of it as a different type of problem, which I’ll explain in the next section.

Third, I see no reason to think that in 2006 the largest investment banks and subprime players were getting credit cheaper or taking actions because they thought the government would bail them out. So the lack of a TBTF subsidy in 2006 isn’t sufficient proof, or really any proof, to me.

I think those actors were thinking things like “housing will only go up” or “the suckers I’m selling this to have to deal with the consequences, and lol I’ll be cashed out by then” or “the ratings agencies said it was fine.” I don’t think “the government will bail this all out” played a role, certainly not enough to account for the size and scope of the actions taken.

Dean also brings up the idea that the CBO’s methodology can’t tell the difference between bailing out the biggest players versus bailing out all players in finance, which is true. What if all banks are TBTF?

Two responses. First, the IMF study, by only looking at large firms and their CDS prices, does account for that, and it finds similar conclusions. That’s why I find both together more convincing than the sum of their parts. Second, if no bank is allowed to fail, period, nobody told the FDIC. I count 490 since 2009 on the FDIC’s failed bank list, including 140 in 2009 alone. This is a sizable amount that will easily be understood by the financial markets as meaning that smaller banks have losses that need to be priced.

Four Ways of Having a Too Big To Fail Subsidy

Let’s map out what forms a Too Big To Fail subsidy could take. As Ganesh Sitaraman has argued there are a lot of ways to understand TBTF. I want to discuss four. I’m going to phrase this in a little extra economics language to tease out the specifics:

1: A TBTF bank can exercise economic and political power, capture regulators, collect monopoly rents, and engage in fraud without consequences. Not combating that is a type of subsidy. This would result inhigh profitswages that look like huge rents, and, perhaps, weak productivity. All of which we’ve seen.

2: A TBTF bank could impose large externalities on the economy as a whole in the case of failure. This is the result of contagion, panics, financial instability and so forth, just like we saw in the 2008 crisis. Not being forced to absorb these external costs is a form of a subsidy. In this case, the bank would have more debt and be larger than it would otherwise.

3: A TBTF could impose serious macroeconomic risk through its creation and allocation of credit. This is independent of its effects in a sudden failure. To think it through, imagine if Lehman Brothers failed but there was no financial crisis. But there’s still be a Great Recession from a deleveraging cycle, millions of foreclosures, plunder of black wealth in housing, and so forth. This is another negative externality but from activities, not failure.

4: The financial markets could assume that the government would absorb some of the losses of a TBTF bank through some sort of bailout process. This implied bailout reduces the credit risks of investing in TBTF banks, meaning TBTF banks can access funding cheaper. This would show up in measures of credit risks, such as bond spreads and credit default swaps.

I argue the leading studies say that the fourth subsidy here has declined.

Those studies can’t tell us much about the first three, since they just look at the credit risk measures. That includes the second subsidy, which means there could be no cheaper credit for big banks but they could still cause a lot of damage in failure. I think a lot of the discussion around a “near-zero value” subsidy that blows up in the crisis is more alongside this risk, which is different. We should, of course, be taking actions to fight the first three on all fronts.

This may seem esoteric but there was a fourth subsidy during the crisis in the aftermath of TARP. If that subsidy stayed the same or got worse, that would be a serious problem. It would mean we are in permanent bailout mode. That it is in decline is an important first measure of success. And it shows that higher capital requirements, liquidity requirements, living wills, restructuring, derivatives clearing, etc. is a smart path for combating it further.