It wasn’t supposed to be this way. When, in 2018, then Federal Reserve Board vice chair for financial supervision Randal Quarles proposed “bespoke” tailoring of US financial regulations to better fit them to bank size, he argued that regulatory burdens on banks should be proportional to the risks posed. In theory, the larger the bank, the greater the systemic risk. It is therefore wholly appropriate that the biggest banks should be subject to the full suite of regulatory requirements mandated by the Dodd-Frank Act, which was enacted following the global financial crisis. Smaller banks pose less risk, Quarles argued, and thus should not be required to meet the same onerous standards.
All that is reasonable enough — except that recent failures of mid-sized US banks have shown the practice isn’t as straightforward as the theory: to minimize the fallout from the failure of Silicon Valley Bank (SVB) and Signature Bank in March, the authorities invoked the systemic risk exception, which allows the Federal Deposit Insurance Corporation (FDIC) to protect uninsured deposits to safeguard financial stability. So why was the practice so different from the theory behind bespoke tailoring of financial regulations?
How We Got Here
Start with the cause of the bank failures. There is no one single factor, and the relative importance of various causes will likely be debated for some time. Nevertheless, there is broad agreement on three key contributors to the failure.
First, governance failures. As pointed out by a recent Federal Reserve review of the March turmoil, a large share of the blame rightly goes to inadequate bank management. This is especially so with respect to SVB, whose board and management simply did not have adequate basic risk controls in place. Rapid growth in bank deposits led to an expanding, unhedged portfolio of higher-yielding government bonds; in effect, SVB was reaping the spread between bond yields and deposit rates and betting that interest rates would remain low indefinitely. Had interest rates remained low, their bet would have paid off. But somebody within the bank should have been warning that higher inflation would force the Fed to tighten monetary policy, driving up interest rates that would result in capital losses on the bond portfolio as bond prices fell.
Second, supervisory failures. The Fed’s review also attributes blame to itself, as SVB’s primary supervisor. Even if governance failures within the bank led to inadequate risk management, the risks posed by an unhedged bond portfolio in a rising interest rate environment should have triggered measures for prompt corrective action. It is clear, however, that the Fed did not take sufficient action. In part, this is because SVB was growing so rapidly that it was transferred from one supervisory team to another at a critical juncture; a bank that climbs the regulatory ladder, meanwhile, gets a reset on its supervisory review. But this is far from exculpatory: a bank that triples in size in two years should set alarm bells ringing. If those alarms were ringing, they may have been muted by the deregulation ethos driving the amendments to the Dodd-Frank Act that animated the Fed’s bespoke tailoring. The goal, in effect, was to shift the focus of supervision from confrontation to collaboration. Again, while good in theory, that now seems somewhat less felicitous (if not slightly naive) in practice.
Third, the digital herd. Recent bank failures have featured massive deposit runs that have played out over remarkably short periods. These runs reflect the digital herd, which in periods of calm is content with the returns it is earning on deposits. The herd can quickly stampede, however, driven by bad news or possibly merely by rumours of bad news. And what makes this behaviour so worrying and potentially fatal to a bank is the ease with which depositors can drain their deposits with the press of a button. SVB lost billions in deposits in this way.
In a sense, the loss of deposits unleashed a self-fulfilling prophecy. If SVB had been able to hold its portfolio of bonds “to maturity,” it could have avoided writing down the value of the bonds. But news of the decline in the underlying market value of the bonds, and of the bank’s efforts to raise capital to close the gap in its balance sheet, triggered deposit outflows, forcing the bank to sell bonds to fund deposits being withdrawn.
One prominent banker, meanwhile, has speculated that post-financial crisis regulations — specifically, Basel III liquidity requirements, which require banks to hold increased liquidity in the form of high-quality liquid assets (HQLA), essentially government securities — stoked banking turmoil. Without these requirements, the argument goes, banks could have held floating-rate business loans that would have performed better over the Fed’s tightening cycle. This perspective conveniently ignores the fact that enhanced liquidity requirements were put in place because of the financial dysfunction that erupted in the fall of 2008, as liquidity drained from markets in the frenzied days following the bankruptcy of Lehman Brothers. Deflecting blame from the absence of effective risk management, where it should rest, to regulation could be considered somewhat short-sighted. The fundamental purpose of banks, after all, is risk management.
There is little question that a single large bank that accounts for a sizable share of the financial system poses a threat to financial stability. But what about the case of several smaller banks that together account for the same share?
Bridging Practice with Theory: Implications for Regulations
A more cogent criticism of regulation and supervisory practice concerns the fundamental inconsistency of having HQLA requirements and allowing banks to hold those assets at book value provided they are classified as “held to maturity.” To be blunt, an asset that is held to satisfy a liquidity constraint isn’t terribly helpful if huge capital losses are incurred precisely when it is needed for liquidity reasons. That paradoxical outcome points to necessary change: assets classified as held to maturity must either be hedged or (some fraction) excluded for the purposes of satisfying the HQLA requirements.
The SVB failure, meanwhile, underscores the utility of rigorous and meaningful supervisory stress testing. The failure of risk management, particularly with respect to SVB’s growing bond portfolio, could have been identified earlier with a plausible stress-testing scenario that involved Fed tightening. And rather than a costly and cumbersome process that is unnecessary for mid-sized banks, stress testing could possibly have saved the bank.
At the same time, the failure of SVB highlights a fundamental fallacy behind the Fed’s bespoke tailoring of regulations — the proposition that systemic risks are a function of size of institution. There is little question that a single large bank that accounts for a sizable share of the financial system poses a threat to financial stability. But what about the case of several smaller banks that together account for the same share? If these smaller banks are exposed to the same shocks, have identical asset and liability structure, and (more importantly) follow exactly the same hedging strategies, their risks of failure would be perfectly correlated. In effect, they are one institution; size of individual institutions is an imperfect indicator of systemic risk. In this respect, the likely reason US authorities triggered the systemic risk exemption and introduced a new liquidity facility at the Fed was the realization that SVB was not the only mid-sized regional bank with an impaired loan portfolio and exposure to highly runnable uninsured deposits.
To prevent further bank runs and the risk of financial dysfunction that outcome might entail, the FDIC paid out uninsured deposits, including those of wealthy individuals. And while that decision may be justified in terms of short-term financial stability considerations, it poses a long-term challenge in terms of moral hazard. Put simply, if large depositors know that they will be bailed out in the event of bankruptcy, they will have less incentive to exercise prudence in placing deposits in well-managed banks. A bank that seeks rapid growth, or that wants to engage in risky lending to generate higher returns, need only offer a slightly higher interest rate to attract depositors. The FDIC could try to assuage moral hazard risk by stressing existing deposit-insurance limits. But if depositors ignore those risks, and the threat of a systemic crisis once again looms large, the authorities would have a difficult choice to make between market discipline and financial crisis.
Unfortunately, there is no obvious policy option. One approach is to cover all deposits but introduce co-insurance, so that the depositor bears some (small) share of the loss. This proposal is unlikely to work in the age of the digital herd: even if, say, 90 percent of the deposit is insured, no investor will choose to incur a 10 percent loss when a click of a button can move the deposit from one bank to another. Rather than reduce the risk of deposit runs, the proposal could have the wholly unintended effect of making deposits more unstable. Similarly, options to introduce “gates” and “haircuts” (automatic penalties for withdrawal) triggered by predetermined thresholds on deposit outflows would likely exacerbate volatility, since depositors would be incentivized to move early. Even more problematic, such early-mover-advantage deposit runs could be triggered by an errant tweet either by chance or by design.
Another option is to explicitly insure all deposits, regardless of size. This approach would have the advantage of collecting insurance premiums and building up the size of the deposit insurance fund, thereby reducing potential costs to taxpayers. But with all deposits covered, the role of market discipline in containing risk taking would be suppressed, possibly increasing the potential costs of bank failures. To contain such risks, it would be necessary to enhance regulation and supervision: in the limit, allowing banks to hold only a limited set of assets. In effect, banks would become regulated utilities. Even this would likely prove ineffective in safeguarding financial stability, however, since risk taking would likely be driven from banking to other sectors of the financial market. Unless the regularity perimeter is extended to non-bank financial intermediation and effectively enforced, the only thing that would have been accomplished is creating safer banks in a riskier financial system.
This outcome highlights the fundamental reason why financial intermediaries — not just banks — are regulated: the threat to the economy posed by financial instability. It also points to a possible way to bridge the theory and the practice of regulation. In this respect, the most effective path to containing risk taking and minimizing the threat of deposit runs may begin in the institution, with the institution’s senior management and board of directors. Deferring a share of their compensation to align short-term decisions with the longer-term financial health of their institution, as was done in the wake of the global financial crisis, is a start. But consideration could also be given to introducing an explicit fiduciary responsibility to taxpayers as potential shareholders of last resort. And while there are a host of issues that would have to be worked through, this approach may offer the best balance between the need, on the one hand, for prudent risk taking that provides financing for new, innovative ideas and, on the other, for safeguarding financial stability.