At the heart of Silicon Valley Bank’s failure are uninsured depositors — specifically startup companies who held far more than the insured limit of $250,000 and who couldn’t make payroll without access to their accounts. It’s tempting in light of SVB’s failure to assume that the insured deposit limit needs to be raised, but that solution creates new problems. A better approach would be for the U.S. to follow the example of other countries and create “payment banks” that take little-to-no risk, are highly regulated, and have access to the payment network. They would be a place where companies could park funds — like VC investment earmarked for payroll — without exposing themselves to the risks that normal banks create.
The failure of Silicon Valley Bank has highlighted underappreciated fragilities of the U.S. banking system. While banking crises have historically centered on credit risk, this recent crisis of confidence stemmed from unrealized losses on safe securities that made depositors anxiously seek liquidity. The liquidation of those securities crystallized mark-to-market losses and amplified the anxieties of these depositors, and a bank run ensued.
While insured depositors have no reason for anxiety, the recent crisis has highlighted the critical role of large uninsured depositors, who are understandably prone to anxiety. They constitute more than $8 trillion — or approximately 40% of all U.S. deposits.
And one particular anxiety stands out: The prospect of many companies being unable to make payroll was a critical aspect of this crisis, as it became clear that some uninsured depositors were business clients that couldn’t pay their employees without access to their accounts.
The Problem of Uninsured Deposits
As an emergency response, it became necessary for the FDIC to effectively uncap the limit on deposit insurance and to declare the troubled banks as systemically important to restore calm. That solution is problematic for many reasons. In the absence of many new regulations, uncapped deposit insurance gives banks horrible incentives. And the regulations required to mitigate those horrible incentives can stifle risk-taking across the broader economy.
The deeper solution to this problem lies in understanding the dilemma of the uninsured depositor and addressing their needs in a more direct way. It’s easy to caricature the uninsured depositor as a reckless riskseeker that flits between banks seeking yield. That caricature is not worthy of a bailout or much sympathy. But the reality is that many uninsured depositors are facing an enormous dilemma.
Consider the problem of private-sector payrolls, which constitutes more than $9 trillion in annual fund flows in the U.S. alone. Large amounts of money must be facilitated on a regular basis and that money must be housed within a bank in order to access the payment system. These deposits simply have no alternative but banks and, therefore, are exposed to the actions of banks who can lend or buy assets with those large deposits. In that process, all our salaries become exposed to the decisions of bankers who can accept these large, volatile deposits, take risks with them, and then socialize the losses when we are forced to uncap deposit insurance.
The Case for “Payment Banks”
The problem of uninsured depositors is really the problem of accessing the payment system — a system monopolized by central banks and then delegated to banks. The payroll problem is a notable example of this problem as payroll funds necessarily must get parked in banks, where they are exposed to the risks mentioned above.
Fortunately, other countries have begun to figure out solutions to this problem. The United Kingdom, Australia, and Singapore have all been innovating and we can usefully learn from their efforts. There are effectively two possible solutions: Allow nonbanks to access the payment system as the UK and others have allowed, or create banks that do nothing more than solve this “payroll problem.” We prefer the latter.
To solve the uninsured creditor problem without distorting incentives for risk-taking, the U.S. should create a special class of bank called a “payment bank” that does nothing more than process payments. Their deposit bases would be large and potentially volatile, they would be very tightly regulated (even more so than money market funds), and they would be unable to take any credit or maturity risk. In short, they would take payroll deposits and other similar large B2B transactions and facilitate access to the payments system.
What would the business model be for these payment banks? There are two possibilities: They could earn a safe return by investing these deposits with the Federal Reserve at the federal funds rate, or they could charge their clients a very small fee for facilitating these large payments. Investing large amounts of these deposits for very short periods in a riskless manner can yield sizable revenues, especially in the current environment, and it’s possible that some of this revenue could even be rebated back to the depositors.
While we have characterized this as a payroll problem, there are numerous other economic agents that have large, volatile deposits that only seek to access the payment system. Consider a $100 million revenue business that has $70 million of annual costs and prudently keeps cash equivalent to a month’s expenses in a bank to cover payments. Alternatively, consider a venture capital or private equity fund that is seeking to raise capital or deploy capital to acquire companies.
Currently, these funds must access traditional banks in order to access payment functionality. Indeed, that is precisely the business model for both Silicon Valley Bank and First Republic Bank. But every bank has these kinds of customers. Indeed, the broader terrain of card-based merchant payments — where $9 trillion of card payments have to make their way to merchant bank accounts via the merchant acquirers — has similar features.
By creating payment banks, the large, volatile deposits that far exceed any reasonable deposit insurance limit will find a suitable home in a tightly regulated bank that takes effectively no credit or maturity risk and can facilitate their transactions. More importantly, the entire banking system will no longer bear the burden of these uninsured deposits and can return to their core function of retail deposits and making prudent lending and asset-liability decisions. And we can avoid uncapping the deposit insurance limit and making all banks systemically important. In some sense, this solution is a less ambitious and much more realistic effort than to use stablecoins or a central bank digital currency to facilitate B2B payments on alternate payment rails. In many ways, this idea reflects the industrial strength principles of clearing and settlement employed in financial markets to a broader set of payments.
The reality is that the U.S. banking system has become much less dynamic since the global financial crisis. Entry is nearly non-existent. While the number of U.S. banks may be high relative to many other countries, the truth is that we don’t need more traditional banks — we need different kinds of banks. Crises are terrible things to waste, and this one could lead us to a much safer banking system by recognizing the problem of the uninsured depositor and creating a home for them.