There’s been a recurring theme in many recent regulatory actions – the slogan on the front of the box sounds great: “Now with 100% more efficiency, flexibility, clarity, AND transparency!” The trouble, as usual, is when someone actually opens the box, and sees what’s inside.
At first glance, the Federal Reserve’s stress test transparency proposal has an entirely respectable pitch – make one of the Fed’s most important supervisory tools more understandable, more accountable, and less vulnerable to legal attack. After all, stress tests help determine how much capital the largest banks must hold, and when a supervisory model effectively drives a binding capital requirement, banks understandably want to know how the machinery works.
But a recent letter from senior members of the Senate Banking and House Financial Services Committees argues that this otherwise benign framing misses the real danger. Their concern is not that transparency is inherently bad. It is that this particular kind of transparency may convert the stress test from a supervisory shock exercise into something closer to an “open-book exam.”
And it isn’t as though this is the first time this kind of warning has been raised. The letter cites the Fed’s own prior position that “full disclosure” of supervisory models could make the financial system more vulnerable by allowing firms to change their stress test results “without materially changing their risk profile.”
In his dissent, Barr warned that disclosing the models and scenarios would make the stress test “weaker and less credible,” creating the risk of “illusory comfort” in the banking system’s resilience. He also warned that the proposal could invite gaming, produce overly optimistic projections, and result in weaker capital requirements.
To be fair, the industry’s position is internally coherent: if stress tests set capital requirements, banks want clarity, stability, and the ability to challenge errors. But from a supervisory perspective, that same clarity can quickly become optimization. The industry letter says firms should better understand “how their activities and exposures are modeled” so they can evaluate the “regulatory consequences of their business decisions.” And while you could certainly argue that is prudent capital planning, you could just as safely argue that may be exactly the kind of model-management critics are worried about.
Not to put too fine a point on it, but if large banks all learn the same model, optimize to the same assumptions, and reduce capital based on the same blind spots, the entire system may become more correlated, not less risky. In turn, that model disclosure could create a “model monoculture,” where banks rely on similar models and miss bank-specific vulnerabilities. Barr made the same warning, emphasizing that model disclosure could cause banks to manage to the Fed’s framework rather than develop independent risk-management capacity.
Furthermore, some argue that the Fed’s proposed model changes would make banks look stronger under stress while reducing capital cushions. The senior members’ letter estimates that stress capital buffer requirements would fall by 2.2 percent, reducing capital cushions at the riskiest banks by roughly $35 billion.
Ultimately, the issue isn’t whether transparency is good. The issue is transparency for whom, for what purpose, and at what cost.
Brett Goodnack, JD, CAMS
Compliance Advisor