Reminder: Technically, Too Big To Fail Is Not Actually A Thing
May 31, 2012 10 Comments
Despite its almost comically hypermelodramatic title, there’s little to argue with in Simon Johnson’s post Jamie Dimon And The Fall Of Nations at Baseline Scenario. Little, that is, if you let this slide:
This bank [JP Morgan] is “too big to fail” – meaning that if it were to get into difficulties, substantial financial support would be provided by the Federal Reserve System (and perhaps other parts of government) to prevent it from collapsing.
This is stated so matter-of-factly it’s possible to gloss over it as you’re reading without quite stopping to ask…well, I mean…Huh?
Here’s the thing. I looked in my pocket copy of the U.S. Constitution for the section on “too big to fail”. Didn’t find it. Curious, I went over to the Stanford campus and borrowed Finance Professor Anat Admati’s textbook on introduction to banking. Turning to the index, I looked for Too big to fail and Fail, Too big to. Not there. The plot thickened.
When did we all agree that “too big to fail” was an objective thing that we weren’t going to argue with any longer? I’m still here. I’m still arguing.
JP Morgan is a company. There is a thing that can happen to companies when they can’t pay their bills for whatever reason, which is that they can declare bankruptcy. I am sorry, but that is the established process. If you are arguing for some other process, there is no legal basis for it, so you’re the one with the burden of proof, not me. Fuck this too big to fail bullshit already. I’ll see your too big to fail and raise you one chapter 11.
In fact this is something both left and right should (supposedly) agree on right? We’re all mad about the bailouts aren’t we? Isn’t the consequence of being mad about bailouts supposed to be, fuck bailouts? How did things get so convoluted that when everyone got mad about bailouts we proceeded directly to assuming bailouts were business as usual unchangeable facts of life?
And the thing is, if you take away this “too big to fail” nonsense, the rest of Simon Johnson’s article basically becomes moot. Try it at home. No too big to fail, no perverse asymmetric banker incentives to lead to Our Nation’s Fall. Done. Problem solved.
Now I know what you’re going to say, you’re going to say that I’m being dumb/naive because ‘realistically’, under the current political-economy context we have in place, Congress is very very very likely to treat certain banks as “too big to fail”, JP Morgan among them. Fair enough. So, we can both agree we have a problem on our hands, which is, and let me try to state the problem fully not partially, When (a) [certain stuff e.g. a bank blowup] happens, (b) Congress will [bailout/other stuff that treats a bank as "too big to fail"], (c) costing us money and probably also leading to the ill effects/incentives Simon Johnson talks about.
But let’s not forget that solutions to this problem (and avoiding (c)) can involve addressing/preventing (a), as most of conventional wisdom seems to be trying to do (futilely, IMHO), or they can involve addressing (b). All I am saying here is, give (b) a chance. It’s not like anyone’s making me super confident they have a good solution to (a) after all.
The “problem” with letting banks fail is that (a) banks are GSEs in every significant way and (b) banks are so heavily regulated in such a bad manner that there is effectively only one bank (with multiple management teams, of course).
What do I mean by (a)? Banks can’t operate without government guarantees – they’re all insolvent in the sense that their assets don’t cover their liabilities and they need to constantly roll over debt to continue to operate. They can operate in this manner only because of the implicit government guarantees of their liabilities.
What do I mean by (b)? All banks hold the same assets. If one goes down and holds a giant asset sale they all go down when they mark to market. Government regulation, of course, encourages all banks to hold the same certified assets.
You raise good points worth thinking about, as always.
I have some reservations about b and whether that’s how it really would play out (your decent points about procyclical regulation and the effects of MTM notwithstanding). Let’s continue the metaphor of there being only one bank. Now, imagine a bank in which one part has a massive writedown/scandal. More often than not, a year or two later I’ll show you some other part of that bank that ( we later learn) won an obscure internal political battle over the scandalized part, took it over, feasted on its carcass, and then went on to riches and promotions and glory for being such geniuses.
For example, do you ever wonder who now gets to take over JP Morgan’s CIO trading book with hundreds of billions of (for all we know) quite-decent-to-maturity risk in it that has just been marked down by 2? 5? billion under someone else’s name? And also has those AFS bonds that supposedly have +8 billion of mark to market baked in them, just waiting to be released? Because, like, I do. Whoever gets that book could be printing 2-4 billion of easy PNL under their name for the next 5 years.
If all banks are part of the One Bank, like you say, then when one (sub) bank fails I wouldn’t be worried about them all failing so much as I’d be wondering who will get to feast on the carcass of the one that fails. (it will always be Goldman, most likely.
Mmmm, it’ll be more persuasive if you comment on the down-side of letting them fail and explain why it doesn’t matter or is bad but less so than letting them go bankrupt.
First of all, huh? ‘fail’ = ‘go bankrupt’ in this context
Second of all, that comment seems completely serious/sincere, which is disorienting from you
Third, yes, I didn’t prove a full thesis. But as the ones advocating (or at least treating as inevitable facts of life) endless ad hoc blank check bailouts are pushing what is essentially an extralegal policy, I find it odd the onus isn’t on them to prove it’s good/less bad? No one has done so (I will not count hand waving about how letting a bank fail will end the world…says who! The bank CEO?)
am sleepy so unsure what you mean.. of course fail=go bankrupt.. but are you saying there’s no downside to that?
To be honest, I’m not clear on what the downside and its fallout would or could be (particularly in worst case scenarios) but you surely have thought this through (how many people would be severely out of pocket, see most their savings go down the drain – directly from this and how many other big businesses would/could collapse leading to how much further damage to other people, what chain of collapses and joblessness this would lead to, etc) and explaining this while explaining how it’s preferable to bailing them out would make your case more persuasive.
Anon., you used to be a lot more fun. What happened to you?
Anyway, I’m just going to go ahead and agree with you: if I were to work through, in all relevant detail, the ramifications of a hypothetical situation in which a sufficiently large bank is allowed to go bankrupt, calculating all the knock-on repercussions and ripple effects that would follow, and estimate the effect of that scenario on the average person, versus (by contrast) the cost to the average person of the government bailing that bank out, keeping all other things equal, and demonstrate that the latter is worse than the former for some wide and plausible class of peoples’ utility functions, then that would indeed make for a more persuasive blog post in support of what I’m saying.
Alas, I am not capable of doing that analysis. In fact, here is a complete list of humans who are capable of doing that analysis:
1. No One.
2. Not Him Either.
3. Nope, Sorry.
You seem to have forgotten your RWCG Catechism: All large calculations are wrong.
Fortunately, in most areas of life, we don’t actually attempt to solve this sort of optimization problem centrally before making policy decisions of this sort. We have (asdf’s comment notwithstanding) things such as ‘laws’ that we follow, and they tend to be grounded in stuff like principles and rights, rather than relying on the correctness of a centralized utilitarian impossibly-gigantic mega-calculation and analysis. But, I am clearly in the minority, as people now seem to want and demand that I attempt the mega-calculation before opining on policy. Even though (again): all large calculations are wrong. Oh well,
I generally agree with your commentary on this sort of thing, but I think you’re wrong here.
Counterparties and debt-holders should behave differently toward the largest banks, since they know that even the bank fails, there’s a very good chance that the bank (and especially it’s counterparties and debt-holders) will get bailed out. You have a point that there’s no guarantee that a large bank would be bailed out. However, the fact that there’s a non-zero (and very significant) probability that such a bank will be bailed out should effect its borrowing costs and interactions with counterparties.
Actually I think there IS now a (de facto) guarantee that they’ll be bailed out, and this is what I’m complaining about, because I don’t want there to be. Was I not clear on that?
You are right that, given the implicit bailout promise, it should (and presumably does) affect their borrowing spreads. Would that it all were not so.
Then again, I’m not really even in favor of FDIC. So I won’t hold my breath on any of this.
There is no “the law.” That includes bankruptcy law. That includes property rights. That includes everything. There is only, and there has only ever been, politics. Politics determines the law, because politics is power. The law is only the law if men with guns make it the law. And politics dictate that banks will get bailed out. The reasons for this are obvious and covered many times over in logical detail that is really hard to dispute.
You know this, you just don’t want to accept it. Because accepting it turns it from a really easy solution (just don’t bail them out) to a really complicated one (full of stuff you don’t like). However, that’s the real world, a bunch of complicated stuff you don’t like because we are just a bunch of tool using apes.
Pingback: Google Reader Purge « Rhymes With Cars & Girls