By Telis Demos
Last March, Silicon Valley Bank collapsed and kicked off the most significant U.S. banking crisis in years. A few months later, the Federal Reserve proposed a series of major changes to how big lenders are regulated. Those events are less related than you might think.
The Fed proposals are known as the "Basel III endgame," in reference to completing the implementation of global banking reforms first outlined well over a decade ago, following the 2008 financial meltdown. These reforms have substantially raised banks' capital levels and constrained many behaviors that were at the heart of that crisis, like bets on risky mortgages.
A larger group of banks would face some tougher standards under the proposal, including those the size of SVB. Yet this final batch of rules won't directly address some of the most salient causes of the collapses of SVB, Signature Bank or First Republic. After all, some of SVB's most troublesome assets were low-risk U.S. Treasurys. At First Republic, a big problem was mortgages for low-risk superprime borrowers -- not shaky subprime ones.
And in all three cases, a problematic funding source was their deposits via core customer relationships. This was the exact sort of thing a lot of people said banks needed to refocus on after the 2008 crashes of trading firms like Bear Stearns and Lehman Brothers.
Here is a short recap of what sparked last year's failures: During the pandemic, banks were flush with deposits and bought bonds, or made loans, at low, fixed interest rates. Then when the Fed started aggressively raising rates in 2022, the market value of those assets fell. That became a problem as banks faced higher deposit costs or increasing withdrawals as customers sought higher yields elsewhere.
But if that crisis was fundamentally about credit risk, then 2023 was about interest-rate risk. And addressing that may require some different ideas.
These pressures built slowly, then all at once last March. Many have pointed to social media and 'round-the-clock digital banking as responsible, arguing they facilitated depositors' panic. But instant connectivity is now a fact of life. Instead, history suggests that across time, there are other commonalities in bank runs that could still be better addressed.
One thread is the outsize role played by big accounts, rather than the popular image of individuals lining up at bank doors. These are also the types of customers that have likely had computer hookups to their banks for decades. The crisis at Continental Illinois in 1984, for example, has been described by a Federal Deposit Insurance Corp. history as a "high-speed electronic bank run."
"Available evidence indicates that major corporations, with hundreds of millions of dollars in deposits, were the main holders of large uninsured accounts and were the major drivers of runs in 1984, 2008, and 2022-2023," wrote Jonathan Rose, a senior economist at the Federal Reserve Bank of Chicago and a historian of the Federal Reserve System, in a note published last year. "In comparison, household and small business depositors appear to have played a limited role."
At SVB, the 10 largest accounts held $13.3 billion in deposits, according to the FDIC, or roughly 8% of its total on the eve of its run. At Signature, about 60 clients had deposit balances exceeding $250 million, or about 40% of the bank's total.
Yet past U.S. bank crises triggered in part by rises in interest rates -- like the bailout of Bank of the Commonwealth in 1972, which held long-term municipal bonds -- haven't translated into hard-and-fast quantitative capital rules explicitly addressing interest-rate risk.
One reason that this task remains incomplete, or left up to more guideline-based or holistic supervision assessments, is that deposit quality is notoriously difficult to measure. It must consider softer factors like a customer's relationship with the bank and the bank's business model. But it is clear that the status of some categories should be re-evaluated. That could include deposits tied to cryptocurrency, or coming from a common industry such as the tightly knit venture-capital ecosystem.
Another area ripe for reform is "held-to-maturity" accounting, widely used in recent years. This enables banks to avoid having to reflect changes in the value of those securities in their investment portfolios in capital levels.
Marking bonds this way protected banks on paper as the value of those securities fell. However, as investors and customers began to worry about banks' ability to meet deposit outflows, they began looking at the potential capital hits that lurked if banks did start to sell down securities like Treasurys or government-backed mortgage bonds held in cold storage.
The current Basel III endgame proposal would require a wider range of banks -- including those the size of SVB -- to reflect marks on other securities, those labeled "available-for-sale," in their regulatory capital levels. But while U.S. banks had $204 billion of unrealized losses on available-for-sale securities as of the fourth quarter of 2023, they had another $274 billion in held-to-maturity, according to FDIC data.
There are good reasons for HTM accounting. Forcing banks to reflect every quarter's price moves into capital could make them appear much more volatile than they really are. That said, maybe banks' ability to use this accounting method could be more tied to assessments of their deposit and funding stability by regulators.
"We learned a lesson about how sensitive a bank should be to HTM accounting," says Steven Kelly, associate director of research at Yale School of Management's Program on Financial Stability. "Auditors shouldn't be making those calls. It should be regulators and supervisors."
Other measures can also help banks raise liquidity when needed, instead of having to sell bonds or loans at a loss. One would be to encourage use of borrowing at a Fed facility known as the discount window, where banks could regularly pre-position collateral in case of emergencies. An issue at SVB was that it couldn't move quickly to use that facility. The Fed's vice chair for supervision, Michael Barr, said in February that he was focused on "how we can improve bank readiness to tap the Fed's discount window."
The next banking crisis surely won't be exactly the same as last year's. It might even look more like 2008's. But if history is any guide, something like 2023 will eventually come around again -- so the clock is already ticking.
Write to Telis Demos at Telis.Demos@wsj.com
(END) Dow Jones Newswires
March 09, 2024 05:30 ET (10:30 GMT)
Copyright (c) 2024 Dow Jones & Company, Inc.
Comments