For much of the past four years, the Trump Administration’s financial policymakers have focused on rolling back rules enacted in the wake of the 2008 financial crisis. The federal banking agencies have relaxed capital requirements, liquidity rules, stress tests, and other financial safeguards, insisting that adjustments are necessary to simplify the regulatory framework and reduce regulatory burden.

With these deregulatory initiatives now largely complete, Trump-appointed officials recently turned to a separate—but equally worrisome—objective: weakening the supervisory process by which the agencies monitor risk in the banking system. These efforts, encapsulated in a speech given last week by Federal Reserve’s Vice Chair for Supervision Randal Quarles, rest on flawed assumptions that could shield banks from effective oversight even after the Trump Administration leaves office.

Supervision is an essential counterpart to the federal banking agencies’ better-known function, regulation. As supervisors, the Federal Reserve, Office of the Comptroller of the Currency, and Federal Deposit Insurance Corporation assess banks’ compliance with laws and regulations, detect emerging risks, and ensure that banks address any shortcomings. As former Federal Reserve Governor Daniel Tarullo put it, supervision “complement[s] the framework of statutes [and] regulatory rules” on which U.S. bank regulation relies.

For the past two years, however, the federal banking agencies have waged a war on bank supervision—or, in the words of Professors Peter Conti-Brown and Sean Vanatta, they have “de-supervised.” Echoing the concerns of bank lobbying groups, Trump-appointed officials are working to overhaul the modern supervisory process that they claim is too discretionary, too opaque, and too unpredictable.

The banking agencies have fought the war on supervision on multiple fronts. For example, the Office of the Comptroller of the Currency slashed its supervision budget—funded by assessments on national banks—by 10 percent in 2019 and another 10 percent in 2020. The Federal Reserve moved systemically-important foreign banks, including Credit Suisse and notoriously-troubled Deutsche Bank, out of the group of firms subject to the most intense supervisory review. In addition, the agencies implemented a congressional mandate reducing the frequency with which they examine smaller banks. The Federal Reserve also eliminated the “qualitative” objection from its annual stress tests, limiting supervisors’ ability to sanction firms for unsatisfactory capital planning. Most recently, the banking agencies and the Consumer Financial Protection Bureau proposed to codify a rule preventing examiners from criticizing a bank for failing to comply with supervisory guidance.

These “de-supervisory” actions all fit a familiar pattern. They will not make the financial system safer, detect misconduct, or protect consumers. Rather, the war on bank supervision will insulate Wall Street from oversight and erect unwarranted obstacles for examiners trying to mitigate excessive risks in the financial system.

The latest salvo in the war on bank supervision came last week, when Vice Chair Quarles delivered a speech attacking the ratings framework on which modern bank supervision is based. As part of the annual examination process, a bank’s supervisor assigns numerical ratings measuring the bank’s financial position, management, and overall condition using a standardized rating scale. A bank must receive strong supervisory ratings to enjoy certain regulatory privileges. For example, a firm generally must be deemed “well managed” to engage in interstate mergers or participate in certain financial activities, such as securities underwriting and dealing.

In his speech, Quarles asserted that, despite more than a century of experience with supervisory ratings, the banking agencies do not “have a particularly well-developed theory of ratings—principles regarding the internal processes and standards that promote consistency and predictability.” In Quarles’ view, bank ratings are too discretionary and lack analytical rigor. Quarles contends that the banking agencies lack “support for [their] view about where to draw the line between a satisfactory rating and an unsatisfactory rating.” In other words, Quarles doubts supervisors’ ability to reliably distinguish between banks that are “well managed” and banks that are not. Perhaps for this reason, Quarles and other Trump-appointed financial regulators have chosen not to limit the activities of the more than 40 percent of large banks that are not currently considered “well managed.”

To be sure, Quarles is correct that certain aspects of the supervisory ratings system could be improved. Balancing supervisory discretion with procedural regularity and accountability is a difficult task that benefits from continuous refinement. The agencies have made some progress in enhancing the transparency and predictability of supervisory ratings in recent years—for instance, by publishing for public comment revisions to the ratings frameworks in 2017 and 2019. The agencies should continue efforts to refine the ratings process, for example, by involving a broader range of agency employees in assigning ratings to ensure consistency and fairness, as Quarles suggested in his speech.

But the intensity with which Quarles attacked the supervisory ratings system threatens to undermine its very effectiveness. By calling into question the agencies’ processes for assigning ratings, Quarles invites a legal challenge to the rating system itself. While the Trump Administration has opted not to revoke privileges from banks that fail to maintain satisfactory ratings, the incoming Biden Administration could choose to do so. But when a Biden-appointed official blocks a bank from merging or engaging in financial activities because it is not “well managed,” that bank may well cite Quarles’ speech in a lawsuit attacking the supervisory ratings system as arbitrary and capricious.

This result would be unwarranted. Empirical data on bank supervision are limited—in large part because the agencies zealously guard the confidentiality of supervisory information. But the available evidence demonstrates that supervision is effective at reducing bank risk-taking and enhancing bank performance. For example, one recent study showed that supervisory enforcement actions reduce banks’ contributions to systemic risk. Earlier studies demonstrated that supervisory ratings are useful predictors of future bank performance and insolvencies. Moreover, Congress expressed faith in the banking agencies’ supervisory rating system as recently as 2010, when Dodd-Frank established a new requirement that a bank holding company itself (rather than just its subsidiary bank) must be “well managed” to engage in expanded financial activities. These considerations justify confidence in the agencies’ supervisory practices and undercut Quarles’ unsubstantiated claim that the supervisory ratings system is unreliable.

The biggest problem with Quarles’ speech, though, is that he erroneously places the burden on supervisors to demonstrate that a bank is not “well managed.” A better way to think about the supervisory process is that a bank must show that it is well managed. Under the statutory framework established by Congress, “well managed” status is a special designation that a bank must earn to enjoy privileges like interstate mergers and expanded financial activities. “Well managed” status is not a default assumption about a bank’s condition, as Quarles apparently assumes.

Recasting ratings in this light has important consequences for the agencies’ supervisory practices. Assuming, as Quarles insists, that the banking agencies are unable to reliably distinguish between banks that are “well managed” and those that are not, the appropriate policy response is not to continue allowing banks to enjoy special privileges even if they have unsatisfactory supervisory ratings. Rather, if supervisory ratings are as unreliable as Quarles claims, the banking agencies should place a mortarium on interstate bank mergers and expanded financial activities until the agencies determine how to more accurately assess which banks are “well managed” and which are not.

I do not expect the banking agencies to take such a dramatic step. But if the agencies are sufficiently confident in their assessment that a bank is “well managed” to permit it to engage in mergers and expanded financial activities, the agencies must also be confident enough to revoke those privileges when a bank no longer satisfies the relevant supervisory criteria. Regrettably, Quarles’ attack on the ratings system and the broader war on supervision threaten to undermine the banking agencies’ ability to mitigate risks in the financial system—and to hold banks to account when they fall short.

Jeremy Kress is the NBD Bancorp Assistant Professor of Business Administration and an Assistant Professor of Business Law at the University of Michigan’s Ross School of Business. He was previously an attorney at the Federal Reserve Board of Governors.