A lot of people are grading resolution authority on a Pass/Fail binary spectrum. It is there, or it is not there?
I want to help develop a language to understand, explain and critique resolution authority, not as a present/not present but as a practice. Specifically a practice of detecting problems in the system and as a way of deterring trouble ahead of time.
Here is Rob Johnson’s paper from MMBM, and here is his roadmap for how to set up credible Resolution Authority:
That might be a little overwhelming, and since I want to target this post at a general audience I’m going to simplify it:
Think of resolution authority as a relationship between deterrence, detection and resolution.
Ideally we’d like to be able to detect firms that are going to fail beforehand, and use the financial sector’s regulatory powers to push them back on a stable path. So resolution isn’t just about failing a firm, it’s about taking steps to tell a firm that they must take action to become safer before they are resolved. This is what regulation at commercial banks do all the time – they create limits and caps and explicit capital ratios. The fact that if they fail, the government will detect it and force changes acts as a deterrence – if they are going to get caught, why even bother?
This was the argument between bankruptcy and resolution authority – Republicans thought that if you just made the resolution above painful enough, through an ugly bankruptcy, that would be enough to force deterrence, and we wouldn’t even really have to bother with detection. The problem is that financial bankruptcies have externalities for banks that have no problems, as well as crushing the lending channels our real economy needs to grow and survive. And the parachutes are so golden, reputation incentives don’t see to do the trick. And the payoffs are so asymmetric and the books are so easily cooked, shareholders can’t bring disclipine. And so on. This is why we need regulation, specifically regulation to expand this pattern from commercial banks to the largest financial firms.
So here is what we need to do for each step:
Resolution
Resolution isn’t just a legal process at bankruptcy. It’s about the time period right before the firm starts to spin out of control, where regulators can see that the books are starting to look weak and can demand that financial firms raise more capital, divest of certain business lines, make themselves less interconnected. This is essential for financial firms because as a financial firm gets closer to bankruptcy, it has more incentives to gamble big and hope that it pays off, leaving more costs to the real economy. This is, post 1991, done for commercial banks with strict rules for guideposts, and with a fair amount of discretion in-between those triggers that heighten scrutiny and triggers that mandate resolving the firm, through a process called prompt corrective action. EoC had a recommended writeup on how prompt corrective action is essential for making this work, and I had more here.
The Dodd Bill has “early remediation requirements.” It functions exactly like its title suggests: earlier to resolution, regulators will help remediate a large failing financial business to try and bring them back to solvency. Will this work as part of resolution authority? It depends on when the Fed will invoke: “the initial stages of financial decline”, which is how the language is written. Ideally we want to intervene when there is still a chance that the business is salvageable, and I’m not sure the language suggests that.
Detection
Now in order to invoke early remediation, you need to be able to detect problems. This is fairly easy for the George Bailey bank. The Federal Reserve and FDIC kick ass at it, in fact. Will it be easy for the Federal Reserve to detect problems before they happen on the books of a 21st century financial firm, a Lehman or a Goldman Sachs?
One way would be for derivatives reform. I’ve written a ton about derivatives here, and you can see the MMBM chapter on it by Michael Greenberger. If you read Partnoy’s post on the valuation of derivatives and non-exchange traded items at Lehman by the Fed, it gives a great narrative of current practices as a disaster. So getting derivatives to clear would help manage the counterparty risk, and getting them and securitized debts onto exchanges would help with managing the books. Indeed one of the reasons the biggest firms are so big is because they have a business model of warehousing a ton of derivatives – this makes detection by regulators incredibly difficult.
The other obvious problem for detection is off balance sheet reform. Right now the balance sheet is a meaningless document at the top 20 firms. Repo 105 is just the latest in many ways to make the balance sheet say whatever traders want it to say for regulators. Since this will be part of the issue for detection that leads to resolution, it’s a major problem.
There’s talk about a living will, but any such will would have to be global and virtually real-time in order to be credible. This is exactly the kind of stuff private agents at exchanges can do for a nice profit, tying incentives and transparency is a powerful way, but we’ll see if any type of derivatives exchanges make it through.
Deterrence
The last step is blunt deterrent items that prevent there from being problems that need to be detected. Things like capital ratios, or interconnected limits. This leads to another type of deterrence, the very first line of defense, self-deterrence by agents at the financial firm itself, to not try and game the system.
I’m not confident at all by the slack language in the bill. There should be harder limits upfront that can be cranked up, or supplemented as appropriate. One can also force harsher penalties under resolution to try and deter behavior: mandatory management resignations, mandatory haircuts on QFCs, equity dilution or wiping out, mandatory restructuring of creditors before any of the fund is touched. I believe the first two aren’t included, though I believe they are under the Frank Bill.
Notice that each of these three things make each other stronger, or make each other weaker, in a powerful way. Better detection and better prompt corrective action will deter people; harder deterrence upfront will require less high-stakes on detection. And financial reform needs to get a better job of getting each of these to strengthen each other. And the more transparency is emphasized, the better we are. This is the way to judge resolution authority.
Because looking forward, at the path down detection and down resolution, if you are a giant firm it might be worthwhile to play a game of chicken with the government as it stands. You’ve already made enough money for life at the biggest firms, you are big enough where your quick collapse would destroy a huge amount of the economy, and you’ve already called the government’s bluff once. I personally don’t want to play this game anymore.
Rortybomb 
US lawmakers obviously think that if you make resolution painful, financial firms would be deterred from making reckless actions. They don’t understand (or don’t acknowledge) that financials can take the entire economy hostage if they fail.
Their failure can never be as clean as GM’s, because they are a pipeline of capital to the economy, and 2B2F bankers know this. So the mere fact that they have this effect on the rest of the economy is grounds for strong regulation, to prevent them from spinning out of control in the first place.
It’s like educational discipline, no? Maybe in the past, children were allowed to be physically punished by the administration and it was fine and scary and it worked for the most part. But our children and our cultural philosophies about punishment / discipline have changed over time, where paddling a misbehaving child just doesn’t happen anymore and the children aren’t afraid of detention, etc.
So what’s a school administrator to do? Well, you have to take preemptive action before the misbehavior becomes too great. A proactive administration including teacher/classroom feedback, parental involvement, and a change in the way the child is educated all play a part (along with other factors). But there needs to be a strong system, otherwise the child will get lost in the cracks.
Pingback: First take on Dodd’s financial regulations « Epiphanyblog
“US lawmakers obviously think that if you make resolution painful, financial firms would be deterred from making reckless actions.”
Yeah, that’s pretty much the Republican reasoning Mike refers to. It has at least two hidden assumptions:
1. The people who run the firms want what is best for the firms, so they won’t deliberately take risks that are too dangerous to the firm.
2. The people who run the firms know enough about their industry to not take unreasonable risks without realizing that they are doing so.
The current crisis has clearly proved that both of these are false in practice.
Managers can make a bundle gambling with shareholders’ money, and walk away rich when their luck runs out; and the shareholders have effectively no way to stop this before it happens.
And even people with considerable experience in an industry can still make huge mistakes.
P.S. Clawbacks and other “skin in the game” reforms might make some progress against problem 1, which is a classic principal-agent problem.
But you can’t prevent mistakes by changing the incentives of the mistake-maker — only intentional actions respond to incentives.
Ultimately, ISTM that preventing financial chain reactions without massive moral hazard is going to require some kind of *appropriately priced* insolvency insurance (to protect the counterparties, not the equityholders of the insolvent), and the government is the only party big enough to act as (re)insurer of last resort. (A private insurer would itself go insolvent in a crisis.) And it has to be mandatory, otherwise firms will cut back on it to cut costs.
And the premium has to depend on the level of risk (funding bailouts through taxes is a major economic distortion because taxes are paid based on profit, not risk), which means the government agency that issues it needs enough information to evaluate the risk *and* needs to somehow be shielded against regulatory arbitrage and/or capture. In non-crisis times this insurance program should run a substantial surplus — to offset the amount of borrowing it’s going to have to do when a financial crisis strikes. That will make it perpetually vulnerable to pressure to cut its “excessive” rates, or to loot its rainy day fund if it has one.
That’s not going to be easy, but — does anyone have a better idea?
Pingback: Matthew Yglesias » Two Dimensions of Too Big to Fail
Pingback: The New FDIC Paper on the Resolution of Lehman Brothers | Rortybomb