The calling -- and secrecy -- of bank supervision
We launched our book on the history of bank supervision. In today's post, I engage two key questions about the calling and secrecy associated with this vital system.
On Tuesday, June 10, 2025, my colleague and friend Aaron Klein hosted a symposium to launch my book, with Sean Vanatta, Private Finance, Public Power: A History of Bank Supervision in America.
Part of it, I will admit, was a party. It was so wonderful to be with some close friends and colleagues to celebrate this book, so many years in production. (Read this piece that I wrote in 2019 on confidential supervisory information and groan with me, tragicomically, on my thought that we were close to finishing the book; we were close to a draft manuscript, but one that we substantially rewrote twice over before publication day.)
The symposium was fun, yes, but it was primarily a substantive occasion to discuss and debate what bank supervision is, where it comes from, and what it means for the present and future.
I was blown away by the quality of the discussions. Readers should watch the whole thing and learn from some of the giants of the field. At one point, as I was struck by the sheer eloquence and forcefulness of the debate between Jeremy Kress and Christina Parajon Skinner, I realized that I might be watching the future Vice Chairs of Supervision at the Fed in a Democratic and Republican Administration, respectively, squaring off. (For the record, I would enthusiastically support the candidacy of either one, but watch their full panel to get a sense of where the debates about regulatory independence and bank supervision are drawn.)
For today’s post, I want to pick up on two questions that we debated: the “calling” of bank supervision and the apparent addiction to secrecy that lives deep inside the traditions of bank supervision.
Is bank supervision a calling?
The keynote fireside chat between Aaron, our host, and former FDIC Chair Sheila Bair, was a fascinating conversation from the jump. These two seasoned veterans of Capitol Hill, the US Treasury, and (in Bair’s case) a bank regulator shared some insights into how supervision occurs behind the scenes. Aaron’s story about the tribal bank with a national charter whose CAMELS score was lower than it would otherwise be because of the nature of its tribal customer base was especially fascinating.
At one point, Aaron posed the question whether we might think of the task of bank examination differently from other jobs, more as a calling than simply a job.
Bair liked the point and took it further: she thinks examiners should enjoy the prestige, independence, and quality of life similar to the Foreign Service.
I find these ideas very compelling, with caveats.
What compels me here is that their ideas could contribute to reversing the disease of hatred for government workers. I say “disease” because, even if we stipulate that government can and should be more efficient, or that there are places where government action should be completely absent in favor of private forces, this does not say anything about the quality of government services that we do in fact need. No serious person I have ever encountered thinks we should abolish bank supervision, full stop. Only dyed-in-the-wool Randian libertarians entertain such notions. (But I repeat myself…)
The question, then, isn’t whether supervision should exist, but what form it should take. More sensible supervision that is more sensitive to priorities set by politicians and their appointees must be of the highest quality we can find, subject to the constraints that we face. If that’s true, then we must ask: how do we convince people to sacrifice more lucrative careers in private-sector risk management in favor of government service?
Perhaps viewing their roles as “callings” is part of that answer. My caveat here is that most times when I hear people tell workers or consumers that they should dedicate themselves to a “common sacrifice,” I want to reach for my wallet. In cycling, there can be cults of devotion to local bike shops, some of which will sell new cyclists high-margin items of dubious utility simply because they have sales incentives to do so. Mentioning that you saw the same item for half the price elsewhere is the cyclist’s version of farting in church. (To be clear, my local bike shop, Cadence Cycling, is fantastic; but it’s also not the closest to me, precisely because of this dynamic at other, closer places.)
My point is that we can and should dramatically enhance the prestige and quality of bank supervision by enhancing the attractiveness of that job to experts. That might be best handled not by motivating bank supervisors for a spiritual connection to their work, but by making it among the best options for their skill set.
My friendly amendment, then, is that we should (1) make bank examiner jobs harder to get by (2) paying substantially more than they are paid now with (3) more rewards for success and (4) clearer paths for demotion/attrition/layoff in the face of failure. This is a bit pie-in-the-sky perhaps, but as Sean and I document in the book, there have been times in our history when bank examiners’ jobs were very desirable indeed. Are they still? Can we improve here? I think we can do much better, starting with some of Aaron and Sheila’s ideas.
The cult of secrecy
A few years ago, a colleague at a law school was reviewing the draft of the manuscript as it existed at the time. Legal scholars are always asking the (unfortunate, in my view) question of how the research they are consuming or producing can change the world. My colleague asked: what is your normative claim from the book? What statute would you pass to make what is bad better?
My answer: we must reform confidential supervisory information so that the relevant stakeholders can do the careful work of holding banks and bank supervisors alike accountable. The current infrastructure protects too much and provides insight into too little.
That was the legal scholar’s version of a fart in church.
My colleague dismissed this as too obscure. He wanted to know whether I thought supervision was “good” or “bad,” captured by industry or abusive of it, the finest example of independent technocratic government or a violation of banks’ due process.
Our book makes clear that such dramatic assertions about so dynamic a set of institutions like supervision are easy to make, but nearly always wrong in any kind of enduring way. Rather than “good” or “bad,” bank supervision is about a process of shared risk management across a public, private divide, where the frontline risk lies with banks and the residual risk ends up with government. (The pathology of requiring every legal scholar to be an activist-reformer is a jeremiad for another day.)
The fact is, though, that any given instance of supervision can indeed be evaluated with the right information. It is also the case that we, on the outside, almost never have access to the kind of information needed to make that evaluation.
Was Silicon Valley Bank’s failure a supervisory failure? Or a stunning success?
Consider the most important instance of bank supervisory news in the last decade or so—the so-called “regional bank crisis of 2023,” with the failure of Silicon Valley Bank at the center.
The consensus narrative that has arisen about that crisis is simple enough. Some bad bankers made some big bets on the direction of interest rates in 2020 and 2021. When interest rates went the other way as the Federal Reserve sought to combat growing inflation, those bankers couldn’t get away from the trade. Their uninsured depositors got spooked by bad things they read on Twitter and a massive run ensued, threatening the entire US financial system and reconvening the federal financial crisis brigade in the process. The supervisors, this consensus narrative continued, were too corrupt, or too focused on DEI, or too timid, or too captured by Trump Republicans at the Fed, or too incoherently structured across too many organizations, to be much use. It was, that narrative goes, a supervisory failure as much as it was a banking failure.
Candidly, there is very little about this consensus narrative I find convincing. (For the best single account of the 2023 crisis, see this paper by Steven Kelly and Jonathan Rose; it casts an awful lot of doubt on the entire consensus view.)
The condemnation of supervisors, though, bothers me especially. The fact is that the 2023 crisis may well have been among the most impressive supervisory successes in US history. In 2020, every single one of the 9,000 some odd depository institutions in America faced the same rate compression. This meant that their assets slowly converged toward lower values, their net interest income also correspondingly dropped. And then, by 2023, at best estimate fewer than a couple hundred of those 9,000 institutions were stubbornly stuck in that duration risk trade.
What happened to the other almost 9,000? Did their supervisors track these risks, flag them for banker review, and then watch as the banks—most of whom take supervisory suggestions very seriously—retreat from those trades and prepare their balance sheet for the long-telegraphed rise in interest rates?
If that account is true, then the bipartisan blame on bank supervisors is completely off base. Supervisors were slandered and libeled in 2023. They didn’t deserve the hostility, including from the Federal Reserve itself. They deserved medals.
Unfortunately, though, we have no idea whether the alternative account is true. Because all communications of every kind are wrapped in the steely bars of “confidential supervisory information,” we will never know.
That’s why the biggest normative take-home point from our book is that information wants to be free. We need not make all information free immediately if we want to preserve candor in the supervisory process.
But the current default presumption is that all information is confidential supervisory information. This default presumption should flip.
Where do we go from here?
You’ll hear a lot more from me on bank supervision in the weeks and months ahead, including some of the conversations that Sean and I will have on the book in public. There will be more of these conversations than I expected. As Aaron said at the conference, when we first put the program together he hoped it would be a rainy day so that people couldn’t leave when they realized that we were going to spend the day talking about bank supervision. The event was completely sold out, though. There is a huge enthusiasm for understanding better this shared space of risk management in banking.
I’m glad. It came just in time for us to sell a few books.
Your observation regarding the possible interpretation of the 2023 crisis as a success instead of a failure made me think of the penultimate chapter of Dan Heath's book Upstream (2020) where he makes a compelling case for perceiving the response to and aftermath of Hurricane Katrina as a relative success instead of an abject failure given the Hurricane Pam simulation the previous year that projected over 60,000 deaths given the response by emergency personnel (instead of the fewer than 1,500 which actually occurred with Katrina). If you haven't read the book (yet), I recommend it.
I wonder if a regular policy of releasing confidential supervisory information (meaning, examination reports or memos) at regular intervals wouldn't help improve examination practices.
A (perhaps not exactly common) complaint is that examiners don't understand the business, or are making up requirements, or are focusing on minor risks at the expense of important ones just to get an MRA. Regular exposure of examination work (with names of examiners redacted but the name of the bank not) might help address this complaint.