The collapse of Silicon Valley Bank (SVB), and several other banks in its wake, marked the third financial crisis since the passage of Dodd-Frank. Commentators have rightly pointed to the so-called “tailoring” of banking regulations during the Trump administration—which curtailed safeguards and preventative stress-testing for mid-sized banks like SVB—as a key cause of the crisis.
But the problem is deeper than that. SVB failed in part because of rules that have been on the books for decades: rules that lay out what activities are considered “safe” for banks to extend credit to, and what activities are not.
Banking’s social contract is warped, with cycles of private speculation and profit layered on top of a public commitment to absorb the cost. In the rubble of SVB, the public subsidy of the banking sector only intensified: the Federal Reserve functionally guaranteed all bank deposits, even above the statutory FDIC insurance cap. Now there is talk of codifying this implicit commitment into law. Yet to fail to act was to court disaster for regional banks across the country. How do we find out way out of this bind? Strange as it may seem, the rules that helped fell SVB offer the seeds of a better social contract for banking.
The Rules of the Game
Supervision sits at the core of the banking franchise. In exchange for the right to create money bearing the government’s full faith and credit, banks are subject to chartering requirements and face ongoing oversight from federal regulators. Supervisors pore over banks’ books to make sure they are operating in a “safe and sound” manner, raising and resolving concerns in a targeted fashion specific to individual banks. At least, they’re supposed to. From the 1980s onwards, regulators started to deemphasize individualized bank supervision in favor of reliance on bright-line rules delimiting acceptable conduct. The goal was to set rules that would function “like the discipline the market would impose,” with regulators imposing the same requirements on a bank that its shareholders would if they had full access to the content of its balance sheet.
But this market-imitating regime has failed at its own goals. For one thing, an investor in a bank has a very different risk appetite than the government that insures it; bank activities that bolster shareholder dividends can also increase risks to the financial system as a whole. The quest to replicate market discipline has authorized an era of increasingly speculative financial engineering. And time and again, the market’s purported talent for pinpointing exactly how risky a given bank activity is has proven demonstrably incorrect. In the absence of effective supervision, banks are incentivized to “reach for yield,” taking on as much risk as they can within the formal regulatory limits. Out of this heady brew: the SVB failure.
The regulations governing banks’ minimum capital embody this regime. Capital requirements are meant to make sure that banks aren’t over-leveraged: banks must fund enough of their lending activities through equity and other sources that can absorb losses before depositors’ claims are threatened. At mid-century, the bank regulators maintained that numerical capital ratios were “arbitrary” and a “first-approximation” of safety and soundness to be determined through discretionary supervision. But from the Greenspan era onwards, formulas ruled the day. Regulators apply “risk-weights” to every loan and asset on a bank’s balance sheet corresponding to its perceived level of credit risk. A bank’s overall capital requirements correspond to the risk-weights applied across its balance sheet. If a bank has lower risk-weighted assets, it has lower capital requirements, and can lever its deposits and other funding sources to take on greater risk.
Enter SVB. The California-based bank parked its cash in in long-term U.S. government bonds. When the Fed started to raise interest rates, the price of the bonds dropped, and SVB (having not hedged against this interest rate risk) stared down serious losses if it had to sell the bonds before maturity. Venture capitalists texted their friends, pulled their deposits, and sank the bank. So why was it so attractive for the bank to invest in Treasuries in the first place? Among other reasons, because U.S. government bonds have the lowest-possible risk-weight of zero. For purposes of the core risk-based capital requirements, banks need not fund any of these exposures through equity. An effective supervisory regime would have flagged this practice as plainly unsound. A regime that prioritizes formula adherence allowed it to go on until it was too late.
Tilting the Board
In a forthcoming Article, I trace the histories of why assets like U.S. Treasury bonds got the risk-weights that they did. To be sure, Treasuries are seen as the quintessentially safe financial assets. But as the SVB episode highlighted, lack of credit risk (for now!) does not mean that an asset will retain its par value on the secondary market. The weighting reflects something beyond a detached actuarial assessment of risk: a policy choice to encourage banks to hold government debt.
The exemption of Treasuries from risk-weighted capital requirements can be traced to the prescriptions of the Basel Committee for Bank Supervision — a transnational group of the major rich countries’ bank regulators. Spearheaded by the U.S. in the late 1980s, the Basel Committee seeks to develop uniform transnational policies so that banks cannot shop between countries for favorable regulatory treatment. To achieve uniformity, the Basel Committee members had to make difficult decisions about how to set risk-weights. One thing the members could agree upon was setting risk-weights that ensured deep markets for their own governments’ debt. Deliberative documents from the Basel Committee show complex political negotiations about which countries’ sovereign debt should (and should not) be given the favorable risk-weight treatment. These rules weren’t simply market-takers, seeking to replicate investor discipline. They were market-makers, nudging bank credit toward particular uses.
LPE scholars have shown that ostensibly neutral rules can have skewed distributional consequences. For risk-weights, the veneer of neutrality is a thin one, indeed. Time after time, they have deviated from assessments of actuarial risks to encourage (or discourage) lending to a particular sector. Favorable risk-weights encouraged residential mortgage lending in alignment with U.S. housing policy. And they discouraged banks from engaging in physical commodity trading after a series of exposes cast a shadow on the practice.
Take the example of commercial real estate lending. Concerned about excessive speculation in that sector, the federal bank regulators sought to impose a higher risk-weight of 150% on most commercial real estate loans. Community banks and realtors decried the proposal, arguing that it would choke off credit to an important part of the economy. But when the regulators mostly refused to budge, they successfully took their fight to Congress. The same “tailoring” legislation that reduced regulatory requirements for SVB lowered the risk-weight applied to commercial real estate lending. Members of Congress did not claim to disagree with the regulators’ assessment of risk—they instead argued that ensuring “access to capital” was of paramount importance. Risk-weights were their tool for allocating this capital.
Where Does the Money Flow?
Despite simmering debate about how high overall bank capital levels should be, there is scant attention to the risk-weights in the denominator of the formula. In the Article, I take on the specific question of whether risk-weights should be adjusted to reflect climate-related financial risks. The limited availability of reliable climate risk information can make quantifying such risks an empirical challenge. But the legal question is a different one: whether safeguarding the financial system from prospective, uncertain harms posed by climate change is consistent with Congress’ delegation to set capital requirements. On this point, I argue that regulatory action is clearly lawful.
But peeling the veil of neutrality off of risk-weighted capital requirements has a broader significance. It speaks to a lost feature of the centuries-old debate about the proper relationship between public and private in American banking. In the neoclassical account, the Federal Reserve decides how much money should flow through the economy, and private banks decide where the money goes by allocating credit. But the public has always had a stake in these latter decisions. The historic preference for unit banks serving local communities over branch banks, the Community Reinvestment Act’s incentives for lending to low-income communities, bank chartering requirements based on “economic need,” and affiliate transaction limits are among the many features of the bank regulatory scheme reflecting this long historical tradition. Risk-weighted capital requirements were hailed as an effort to recast regulators’ role in the mold of the unfettered market. But bank lending decisions have always been a matter of public concern. The regime we have, however imperfectly, reflects this reality.
As we move to recast banking regulation, we should pair discussion about how to shore up the liability side of banks’ balance sheets with greater oversight of what they do on the asset side. Risk-taking is not simply a means for banks to generate profit, but to fulfill their function as core parts of our financial infrastructure. We already have a regime that works to shape how banks allocate the government’s full faith and credit—but in a highly diluted way, that perversely increases financial sector risk. When banks lend money, they act as instrumentalities of the public. We should get them to act as such.