Ryan Chittum applauds Gretchen Morgenson and Dean Baker for trying to calculate what they all refer to as the “costs” of having a too-big-to-fail “policy”. But I don’t like this lens at all.
For one thing, what Baker is totting up here aren’t costs at all, they’re benefits to the banks concerned. At no point is the key logical step spelled out: if a big bank benefits by being too big to fail to the tune of say $1 billion a year, then there must be an equal and opposite $1 billion a year cost to taxpayers. What’s more, Morgenson implies, that cost can be compared directly to other government spending:
If Mr. Baker is correct about the estimated size of the subsidy, the costs of too-big-to-fail are substantial when compared with other government programs. At $34.1 billion a year, the subsidy is more than twice the grant given under Temporary Assistance to Needy Families, a $16.5 billion program that helps recipients move from welfare to work. A $6.3 billion subsidy would be roughly what the government spent in 2008 on the Global Health and Child Survival program, an initiative aimed at preventing malaria, AIDS and tuberculosis.
What Morgenson’s talking about here aren’t contingent liabilities, they’re cash outlays. The government isn’t giving that level of cash to the banks; there’s not even an explicit government guarantee on the banks’ unsecured debt. Instead, there’s an implicit government guarantee, which, being implicit, can’t really be charged for. Instead, the government is seeking to increase large banks’ minimum capital ratios so as to reduce their profits by at least as much money as the extra profits that Baker is calculating.
Conceptually, then, I don’t like the idea of painting this benefit to the banks as a cost to the government, because it really isn’t. After all, if the government were forced to bail out any of these banks’ bondholders, the sums involved would be vastly greater than the numbers that Baker is throwing about. In that sense, the cost to the government is either zero, or it’s in the hundreds of billions of dollars. It’s not somewhere in between.
I also don’t like the idea of painting too-big-to-fail as a “policy”, as though it was a decision implemented by some technocrat somewhere down the line. It’s not. Pretty much every major economy in the history of capitalism has had too-big-to-fail banks. It’s a natural state, which governments can try to manage, but doing so carries its own costs. Morgenson quotes Baker:
“There is a subsidy here, and we either have to say we are going to break up the banks and get rid of the subsidy, or if we don’t do that, then we have to be confident that we have put in enough regulation to offset the subsidy.”
Breaking up too-big-to-fail banks is, in the best-case scenario, an extremely costly and disruptive thing to do. That doesn’t mean it shouldn’t be done, of course. But there is absolutely a cost-benefit tradeoff here. What’s more, it might not even be possible: if you break a trillion-dollar bank into two $300 billion parts and one $400 billion part, all three parts are still too big to fail, especially in a time of crisis. How big is small-enough-to-fail? And is it possible to grind America’s largest banks into chunks that small? These are non-trivial questions, especially when you’re dealing with broker-dealers which need large balance sheets just to be able to make liquid markets.
Baker’s alternative involves offsetting subsidy with regulation — how is that supposed to work? Can you calculate regulatory costs and debit them from too-big-to-fail benefits? Goldman Sachs currently has a tier-1 capital ratio around 16%; if the government increases its minimum capital ratio from 8% to 12%, what is the associated cost to Goldman? Yes, it loses a certain amount of option value — the ability, if times turn tough, to reduce its capital ratio sharply while still remaining in regulatory compliance. But how would you even begin to calculate that option value?
I think that Baker’s exercise is a useful one, especially when he comes to the conclusion that in the case of a bank like Capital One, all of its profits and then some might be a function of the implicit government guarantee. (Although, is Capital One really too big to fail? When WaMu and Lehman were not?) But let’s not pretend that we could, at a stroke, take the benefits that accrue to banks when they get too big to fail, and spend that money on something entirely different. We can’t.