The recent demise of Silicon Valley Bank and a few other regional lenders is forcing policymakers to revisit a difficult question: What should the government do to prevent such damaging bank runs? Should it yet again expand deposit insurance?
There’s a better way.
The SVB debacle illustrated three key weaknesses of modern-day banking. First, in an era of online transactions and social media, runs can happen with unprecedented speed. Second, uninsured depositors aren’t a reliable source of market discipline: They have just two modes, complete inattention or total panic. Third, authorities lack adequate tools to prevent panic. The “systemic risk exception” they used to rescue SVB’s uninsured depositors can be invoked only after a bank has failed and only on a case-by-case basis so as to offer limited comfort to depositors in other institutions.
To address these shortcomings, some politicians and policymakers want to raise the federal deposit insurance limit from the current $250,000 per account — or even remove it entirely. But this would create its own problems. For one, it would cover all institutions no matter how they were managed and thus would encourage lenders to take even greater risks — as the savings-and-loan crisis of the 1980s demonstrated. Second, it would require a large increase in banks’ contributions to the deposit insurance fund, effectively forcing conservatively run institutions to subsidize their more aggressive counterparts.
What’s needed is a backstop that creates the proper incentives for banks to manage risk. With this in mind, I would propose an expansion of the Federal Reserve’s lender-of-last-resort function (along the lines of a mechanism proposed by Mervyn King, who headed the Bank of England during the 2008 financial crisis). Instead of offering a blanket guarantee, the Fed would promise to lend banks the money needed to pay all their uninsured depositors — but, in exchange, banks would have to pledge sufficient collateral to cover those deposits. Higher-quality collateral would get more credit: The central bank might agree to lend $99 against $100 in short-term Treasury bills, but only $50 against $100 in risky, longer-term corporate loans. With such a backstop in place, uninsured depositors would no longer have an incentive to run, because they would know that their bank would always be able to obtain sufficient cash to honor their withdrawal requests.
Well-capitalized banks with ample high-quality assets would have no problem meeting the Fed’s collateral requirements. By contrast, risker banks with a lot of uninsured depositors would have more difficulty and would face powerful incentives to change how they operated. Silicon Valley Bank, for example, would have run out of collateral to cover its uninsured deposits long before the panic started. To get back into compliance, it would have had to raise capital, increase its share of insured deposits or reduce the riskiness of its assets — all steps that could have prevented the bank run from happening in the first place.
