In the CAMELS system, the problem is measurement, not management

• Key insight: CAMELS does not need less management, it needs better measurement.
• What's at stake: A common evidentiary basis by which to assess management would empower firms as well as supervisors.
• Forward look: The FFIEC should preserve the management rating as an early warning signal and foster a public-private dialogue aimed at establishing a common evidentiary basis by which to determine when that signal is material.

Responding to the Federal Financial Institutions Examination Council's recently closed comment period on proposed CAMELS reform, industry participants argued broadly that management ratings — the "M" in the acronym — should be grounded in material financial risk. Many supervisors, however, fear that narrowing the scope of the management component may delay needed action on emerging problems. Both arguments are valid.

The FFIEC rightly demands materiality, but this cannot imply deterioration that is already evident in financial metrics, lest CAMELS ratings serve merely to describe observable distress while failing as a tool that works to anticipate such. The value of a management rating lies precisely in its ability to identify leading indicators of material risk not yet reflected in financial statements.

Reform of the management rating standard would start with a clearer definition of terms. "Material risks" are any that may credibly threaten market confidence in the continued viability of a firm.

Note that this definition is deliberately neutral about origin. Risk is "material" not where it begins, but where it leads. A threat to confidence in a firm's viability may begin in capital or liquidity. So too might it begin in governance, culture, technology or compliance. As such, at least in this context, we can set aside unhelpful distinctions between financial and nonfinancial risk. As an adjective, "financial" may describe the potential consequences of, rather than the source of, material risks.

And note that "credibly" does essential work in the definition of material risk offered here; it prevents prospective supervision from becoming speculative supervision.

Before identifying a risk governance or control failure as a prudential matter, examiners must demonstrate the causal pathways that may credibly lead to material harm: What behavior was observed? Was it patterned, persistent or worsening? How did it impair the institution's capacity to identify, escalate and remediate risk? What exposures could it create or amplify? Finally, how did that combination threaten the firm's viability or the market's continued confidence in such?

As I have argued in these pages previously, we must modernize how Management is measured. Supervisors can readily inspect formal risk governance "inputs" – e.g., policies, processes and control frameworks. And they can observe harmful performance "outputs" – e.g., losses, compliance breaches and consumer harms. But the greater informational opportunity lies between those inputs and outputs, amidst the organizational "throughputs" that determine whether governance arrangements work as intended, before risks are allowed to compound.

Culture is relevant here not as a judgment that it is "good" or "bad." Rather, culture is rightly viewed as a driver of organizational behaviors that underpin – or undermine – effective risk governance.

In this connection, our recent Supervisors on Supervision global stock-take found broad agreement among supervisors and industry participants worldwide that the bank failures and near-failures of 2023 were not principally balance-sheet problems in origin; instead, weaknesses in governance and culture became confidence problems. This highlights the importance of reliable management ratings.

Read more:

• Walmart-linked fintech's banking incursion reaches autos
• GAO flags disclosure gaps exposed by 2023 bank failures
• SEC moves to nullify pay-to-play ban for federal contracts

• Fed ends enforcement action against United Texas Bank

But assessments of management must draw on observable behavioral tendencies, calibrated over time and against relevant peers, and evidenced by transmission paths that credibly lead to material harm. When the financial components of a CAMELS rating remain sound, while the management component is seen to be deteriorating, such divergence may not suggest inconsistency. Evidenced properly, it provides precisely the early warning signal that a CAMELS rating is meant to offer.
Notably, while stock-take participants agreed that supervisory judgment remains necessary, they also insist that it must be more evidence-led, explainable, reviewable and contestable. Examiners must afford institutions an opportunity to rebut supervisory assessments of management. A risk metric is not a verdict; it is a diagnostic instrument that should make supervisory reasoning more consistent and challenges thereto more analytical. This demands a common evidentiary basis for rendering relevant judgements, on either side of the table.

A common evidentiary basis by which to assess management would empower firms as well as supervisors. Firms would gain clearer expectations and a common language for testing themselves, comparing business units, and reviewing operations across geographies and jurisdictions. Supervisors would be equipped with forward-looking information for horizontal reviews that facilitate earlier remedial intervention. Disagreement would not disappear, of course, but it would at least turn on a shared set of agreed reference terms rather than competing subjective impressions.

Establishing such a reliable evidentiary architecture is a collective-action problem. A framework designed by supervisors risks becoming an impracticable expectation set established by fiat, while one designed by industry only may be designed as exculpatory and lack sufficient public legitimacy.

The FFIEC should thus help to catalyze a collaborative, public-private effort involving supervisors, executives and board directors, academics, and other specialists, and tasked with establishing the evidence protocols, analytical frameworks, and illustrative metrics by which management can be assessed reliably, consistently, fairly, and proactively.

Supervision can cause harm in two directions; opaque judgment can impose significant financial consequences without adequate justification, and delayed judgment can leave institutions and their stakeholders suffering preventable losses. A reformed CAMELS framework must reduce both errors through instrumentation. Management ratings must be prospective but not speculative, behavioral but not impressionistic, quantitative but not mechanical, and consequential only when connected to a credible threat to confidence in the continued viability of a firm.

CAMELS does not need less management, it needs better measurement. The FFIEC should preserve the management rating as an early warning signal and foster a public-private dialogue aimed at establishing a common evidentiary basis by which to determine when that signal is material.