When banks fail, who really pays?
In spring 2023, as first Silicon Valley Bank (SVB) and then Signature Bank failed, Michael Ohlrogge was teaching his “Introduction to Banking and Modern Finance” course. His students had questions. Why had the banks failed, and why had the Federal Deposit Insurance Corporation (FDIC) guaranteed all deposits at both banks—even those with balances too high to qualify for deposit insurance?
“I remember spending pretty much my entire spring break trying to figure out what actually happened at SVB and how that actually related to what was in the laws and laws that had changed,” Ohlrogge says. “Because there was a lot of misreporting and oversimplification going on.”
The result is an article published in the NYU Law Review last year—“Why Have Uninsured Depositors Become De Facto Insured?”—that examines a striking shift in how the FDIC handles bank failures. Formally, deposit insurance is capped at $250,000 per depositor. In practice, however, uninsured depositors at failed banks are now far more likely to be protected than they once were.
The FDIC is generally required to resolve failed banks using the least costly method of protecting insured depositors, Ohlrogge explains. If officials want to protect uninsured depositors for broader financial-stability reasons, even when doing so is not the least costly option, the law allows them to do so under the systemic risk exception, which requires approval from top federal officials.
Of the FDIC ‘s two main ways to resolve a failed bank, one is a piecemeal approach: pay insured depositors, then sell the failed bank’s assets to different buyers, or transfer only selected assets and liabilities. The other resolution method is a whole-bank sale, in which one buyer takes over the bank more or less intact, including uninsured deposits. The systemic risk exception can be invoked and considered when choosing the resolution method, if the priority is to reduce risk instead of cost.
The distinction between the two resolution methods matters, Ohlrogge says, because a whole-bank sale is often the costlier option. A buyer taking an entire failed bank may have to accept a mixed bag of loans and other assets that it does not especially want or cannot easily value. “It’s kind of like those stories about how you can buy big crates of returned items from Amazon,” Ohlrogge says. “You have no idea what you’re going to get. Probably some of it you might like. A lot of it is going to be junk for you. So what are you going to pay? You’re not going to pay very much for that crate.” Selling assets separately can instead let them go to the parties that value them most, which can lower the FDIC’s losses.
Whole-bank sales also tend to protect uninsured depositors. If a buyer assumes all deposits, then depositors with balances above the insurance cap are made whole, too. By contrast, a piecemeal resolution is more likely to leave uninsured depositors exposed.
Ohlrogge argues that the FDIC has increasingly structured resolutions around whole-bank sales even when those resolutions are frequently not the cheapest lawful option. The data in the paper show how dramatic the change has been. Between 1992 and 2007, uninsured depositors took losses in 63 percent of bank failures. Since 2008, they have taken losses in only 6 percent of failures. At the same time that uninsured depositor losses have plummeted, the FDIC’s costs of resolving failed banks have nearly doubled.
“It’s only when you bring in a lot of data and you look at these really, really clear patterns,” Ohlrogge says, “that it starts to become pretty clear, in my opinion, that the FDIC has not been following the law.”
Why did this happen? Ohlrogge offers two main explanations. One is institutional capacity: during the 2008 financial crisis, the FDIC faced a wave of failures, and more complicated resolutions may have been harder to carry out quickly. But what may have begun as an emergency response appears to have hardened into a long-running practice even after the crisis passed, he says.
The second explanation is what he describes as mission creep. In the paper, Ohlrogge argues that the FDIC may have developed a preference for protecting uninsured depositors whenever possible, even though Congress has repeatedly tried to limit that kind of discretionary rescue. Ohlrogge’s research highlights a tug-of-war between the FDIC and Congress that has existed since the agency’s creation, in which the FDIC repeatedly veers towards rescuing uninsured depositors and Congress periodically reins it back in.
The FDIC’s choices appear to have been expensive for the American public. “I estimate that FDIC resolution costs since 2008 have gone up by about $45 billion, compared to what they would have been if we could have maintained the efficiency of resolutions that we had pre-2008,” he says. “And I estimate that only $5 billion of that has actually gone to benefit uninsured depositors.”
In fact, the paper does not argue that uninsured depositors should never be protected. Rather, Ohlrogge says, when officials believe broader protection is necessary, they should use the legal mechanism that Congress created for that purpose. Most of the increased cost, he says, reflects the inefficiency of relying on whole-bank resolutions more often than is economically justified. “If the FDIC just got the political authorization that the current law requires them to get, they could just directly compensate uninsured depositors when they deem it essential, rather than jerry rigging a convoluted and inefficient resolution process to try to claim that a whole-bank sale is the only viable way to resolve a bank, and that it's just a happy accident that uninsured depositors get protected,” he says.
In Ohlrogge’s view, the current system is the worst of both worlds: it often protects uninsured depositors without doing so transparently, and it does so through a process that appears far more expensive than it needs to be.“I will say it’s an easy decision that, if you are someone in Congress and you think we should do more for uninsured depositors, our current approach is a really bad one,” he says. “So you at least want to change the law.”