In the wake of Silicon Valley Bank’s collapse, which set off panic in the financial sector and concern across the global economy, a crucial question has been whether regulators could have intervened sooner. Recent reporting has indicated that, more than a year ago, the San Francisco Fed did notice problems—including how the bank managed its exposure to changes in interest rates and whether it would have enough cash in a crisis—and warned S.V.B. about them. (Between 2017 and the time of those warnings, the bank’s assets had quadrupled to more than two hundred billion dollars.) After the financial crisis of 2008, Congress passed the Dodd-Frank Act, which imposed stricter regulations on the banking sector; in 2018, Congress scaled back Dodd-Frank, raising the threshold for increased scrutiny of banks from fifty billion in assets to two hundred and fifty billion.
I recently spoke by phone with Peter Conti-Brown, an associate professor of financial regulation at the Wharton School and an expert on the Federal Reserve. During our conversation, which has been edited for length and clarity, we discussed whether the Trump Administration’s push to weaken Dodd-Frank helped cause the current crisis, whether the Federal Reserve should be in the business of regulating banks, and how much autonomy and power regulators really possess.
Late last week, Bloomberg reported that regulators had warned Silicon Valley Bank repeatedly about its risk-management practices, sending it warning letters that would precede more severe action. Can you talk about those letters, and how binding they are?
The supervisory processes include a number of different elements. Sometimes they include checking to make sure that specific procedures have been followed. Sometimes they are second-guessing risk-management practices. When something goes wrong, examiners and bankers will talk with each other to make sure that the questions are well understood. When they are, an examination report comes out, and that gets a response from the bank.
When an examination report highlights a failure to follow adequate procedure, or when the risk has been increased without adequate explanation, then the examiner—in this case, an employee of a Federal Reserve Bank—will work with a team to escalate the issue. That escalation goes through a process that culminates in a Matter Requiring Attention (M.R.A.), and the M.R.A. goes to the board of the examined entity. It’s not uncommon for just about every supervised bank to have M.R.A.s.
To escalate from an M.R.A. to a Matter Requiring Immediate Attention (M.R.I.A.) typically requires the approval of a supervisor of supervisors, sometimes called a chief examiner. That chief examiner covers a number of different banks and isn’t going to act unilaterally. Typically—again, this isn’t spelled out in any kind of public document—the chief examiner, in consultation with others at the Federal Reserve Bank and the Board of Governors in Washington, D.C., will reach the determination. Of the many M.R.A.s, there are a few that are especially significant and need to be resolved “immediately.”
But an M.R.I.A. is not a file that just gets parked on a desk—it is the basis for additional enforcement action. When there’s a failure to resolve M.R.A.s or M.R.I.A.s, that can easily turn into more aggressive action, such as a mandate that a bank not take on any more clients, or sell assets, or even shut down.
Regarding Silicon Valley Bank, how much could regulators have known about its problems?
Every single M.R.A. and M.R.I.A. was available to regulators. These are written by supervisors, and, while we don’t know the content, we can be almost certain that they referred to the deterioration of risk management on the asset side and the explosion of flighty deposits on the liability side. Sometimes M.R.I.A.s are especially egregious instances of, say, risk mismanagement or failures of processes. And sometimes they are escalated from M.R.A.s because the bank has done nothing about them. We don’t know what the M.R.I.A.s were, but I would imagine they’re both of those things.
I would imagine that the bank’s reliance on lower interest rates would be the type of thing that would generally set off alarms for regulators.
And I’ve read that there were other red flags. One of them was that Silicon Valley Bank was borrowing from the Federal Home Loan Banks system. Why was that a red flag?
Examiners sometimes will flag overreliance on the Fed’s own nonemergency lending facilities. And the reason is that the Federal Home Loan Banks, as with the Federal Reserve Banks, are creatures of Congress that are meant to be lenders of last resort. An aggressive use, as there was here, suggests that Silicon Valley Bank didn’t have first-resort creditors to which it could turn. And that’s usually a sign that markets are suspicious of the viability of the firm.
I want to take a step back and talk about Dodd-Frank. What did Dodd-Frank do to make the banking system safe? What are its most important legacies?
It is important to recall that Dodd-Frank is seventeen different laws. And the one that we are debating in this case is really Title I of Dodd-Frank. Title X is the Consumer Financial Protection Bureau. Title VII is about derivatives. And so, when we say Dodd-Frank, we’re talking about a lot of changes to the financial system. The biggest change of all was that all banks with assets above fifty billion dollars were subject to a much more aggressive, much more open-ended, and much more discretionary supervisory regime. This is known as enhanced prudential supervision. It includes stress tests. One stress test was created by Dodd-Frank itself, and the others are ones that the Fed had already been doing and which were now ratified by Congress.
The other part of this—this refers to Title II—is that the banks had to submit a so-called living will: in the event of a crisis, this is how we will die in an orderly way. And the orderly death did not include the Federal Deposit Insurance Corporation (F.D.I.C.) guaranteeing the uninsured liabilities for its depositors, as was done here. In that sense, a living will, had it been submitted—the change to the law in 2018 eliminated that requirement—would have identified other kinds of loss-absorbing capital, other kinds of liquidity.
Before the 2018 rollback, we would’ve been having stress tests in an annual cadence that would’ve picked up on the concentration of risks. The strangest part of this story is that Silicon Valley Bank mismanaged its risk using the safest, plain-vanilla assets you can imagine. The way that our capital regime denominates those securities means that they might have got by a stress-test system, had it been applied. That’s an indictment of the stress-testing regime; it’s not an excuse for Silicon Valley Bank. But that blunts the critique that the stress test would have saved Silicon Valley Bank had it still been applicable. At this point, I’m going to say that we don’t know yet whether a stress test, a living will, or enhanced prudential supervision would have done what we needed them to do.
Those are the innovations of Dodd-Frank. The biggest point of all, though, is that, even before Dodd-Frank, and even after 2018, the supervisors retained all of the tools necessary to pursue any question of risk concentrations or risk mismanagement up to and including forcing a bank into liquidation. And that’s what we did not see. We saw the red flags, and Bloomberg’s report says that the supervisor saw those red flags, too. What they didn’t do is act on them. And that’s the question we still don’t have an answer for.
How did the 2018 law change things—both very specifically in terms of what was mandated by law and also in the ways that regulators went about their jobs?
In this era of omnibus bills, often hundreds or thousands of pages long, the 2018 bill is interesting for a lot of reasons, not least because of how short it is. The entire thing is seventy-five pages. The part that is relevant to our discussion right now is only five pages long. In the original Dodd-Frank bill, banks with fifty billion dollars or more in assets were subject to enhanced stress testing and the rest. Those five pages moved that number to two hundred and fifty billion. So, for all banks above two hundred and fifty billion, nothing changed.