- Key insight: The FDIC should release all the underlying data in its "Dissecting Depositor Flight: An Analysis of the Spring 2023 Bank Failures."
- What's at stake: Regulators continue to push banks to rely on funding that is now among the least stable and riskiest.
- Supporting data: Many types of brokered deposits discouraged by the regulators may have been beacons of stability and many that regulators favor likely triggered the liquidity crises.
Last month,
The study is fascinating for both what it includes and what it notably leaves out. As a next step, the FDIC should release the underlying
The study includes important new data further questioning the volatility of deposit types long seen by regulators as stable. Business operating (or checking) accounts are singled out as most concerning: "business accounts at all three banks declined rapidly and steeply, falling by 36 to 60 percent between March 7 and March 17." And yet, business operating accounts are typically placed in the most favorable regulatory liquidity buckets.
Historically, business accounts were, in fact, sticky. Despite large uninsured balances, operating accounts were hard to move (going to a physical branch, long settlement times for money movement, etc.), even when a bank was under stress. Mobile banking, real-time payments, etc. have changed all that. The times have changed, but regulatory presumptions remain stuck in the past. Regulators continue to push banks to rely on funding that is now among the least stable and riskiest.
The Federal Reserve governor warned in a speech Saturday that lower capital requirements and lighter supervision could create a credit 'sugar high' that could spur excessive risk-taking, with potentially significant long-term consequences.
Surprisingly, the study does not focus much at all on brokered deposits. The FDIC has, for decades, been intensely focused on the perceived risks from brokered deposits (funds collected through intermediaries). Untold hours and thousands of pages of regulation have been devoted to the topic to prevent bank runs.
If any subset of deposits classified as brokered had contributed to the bank runs, one imagines that would have been highlighted in the analysis. Oddly, the only reference is to brokered CDs. The study, quietly and without any sense of irony, reports that the much-denigrated brokered CDs actually did not contribute in any way to the deposit flight that brought these banks down. In fact, brokered CDs increased during the crisis, offering the banks a liquidity lifeline that was, unfortunately, insufficient.
As with business accounts, the shift in depositor behavior is perfectly logical. In the market of the 1980s, when the relevant rules were birthed, brokered deposits were generally uninsured with big incentives to move at the first sign of trouble. As the study itself acknowledges, a "primary attraction of brokered CDs" now is full deposit insurance. Fully insured, these depositors had little incentive to pull their money, and in fact did not.
Based on the limited data that is available to the public about these bank runs, it appears that this ironic reversal of behavior was widespread: Many types of brokered deposits discouraged by the regulators may have been beacons of stability and many that regulators favor likely triggered the liquidity crises. The data on which this FDIC study is based must be far more detailed and insightful than what is publicly available. It is reasonable to suspect that other conflicts between regulatory presumptions about funding stability and the on-the-ground truth may lie within that data, waiting to be unearthed.
Building on this important study, the FDIC should go further and release the underlying data, appropriately anonymized, to inform solid policymaking based on data rather than presumption. Since the funding structure of banks, large and small, are significantly influenced by the preferences of regulators, the soundness of banks in future periods of stress depends on getting this right, based on that data.
