FinServ Chairman Hensarling has lobbed a firebomb at the banking agencies, telling them to eschew “subjective” reputational-risk determinations in favor of sanctions based solely on CAMELS ratings. The problem with this is that CAMELS are just as subjective as rep-risk calls. And, history has sadly proven them also to be wholly unreliable. In 2001, I told Congress after the then-costliest failure ever that CAMELS should be disclosed so markets can judge them and supervisors can be held accountable for gaffes. But, for all the talk of “Pillar 3” disclosures and market discipline, the agencies still shroud CAMELS in secrecy. Time to out them, I say.

Why are CAMELS just as subjective as rep-risk calls? Think about the letters and let me count the ways. The C stands for capital. Seemingly a nice, objective number, capital is only a useful criterion if subjectively judged in the context of a bank’s risk profile. That is, meeting the nominal capital requirements isn’t safe at all if a bank has a very concentrated, high-risk book or a lot of structured off-balance sheet assets. Been there, done that.

A is for assets – i.e., how sound they are. This judgment rightly corrects for capital by pairing the C assessment with a look into the books. But, again, judgments here are subjective – valuations themselves don’t tell the whole tale and historical-cost accounting has its pitfalls. Again, been there, done that.

M is for management – yes, very subjective. How do supervisors judge the diligence, competence, and honor of top brass and the board? Not well a lot of the time, and surely also subjectively.

Earnings is the E criterion. Like C, it’s a number; also like C, it’s a number that only matters in context. How much earnings is enough earnings depends on a lot of factors that must be balanced with judgment, not just models. L is for liquidity, another number that only makes sense in context, and sensitivity – the S word – is again quantitative in character, but qualitative in impact.

We know what goes into CAMELS, but not what comes out when examiners assign it. That’s because, as I said, the magic number is a deep, dark secret. One rationale for this is that knowing whether an insured depository is a 1 or 2 might advantage it versus other institutions, making it hard for the weaker ones to recover. True, but isn’t that the “market discipline” long vaunted as a critical prudential-policy pillar? The flip-side of this argument – also used against my sunlight proposal – is that, when the poor dears know that their bank is a 3, 4, or 5, they’ll queue up, start runs, and destabilize financial markets. Maybe, but word has it we’ve got very generous deposit insurance these days. Better, I think, for sophisticated counterparties to get wind of looming problems and force a 3-rated bank to clean itself up or shut itself down before the FDIC is forced to do so at far greater cost.

Chairman Hensarling’s concern about rep-risk calls goes deeper than his preference for what he hoped might be a more objective way to boss banks around. He also argues that telling banks with whom they can do business uses private enterprise for public purpose.

Here, he’s got a point — as I’ve discussed before, public policy is increasingly recrafting insured depositories as public utilities. We may ultimately need to do this, at least for the very biggest banks. In fact, this might even have some upside for them, as well as for public policy. But, that’s only if so profound a paradigm shift is done on purpose and with forethought, not as an accident of one-off supervisory and regulatory actions.

So, CAMELS are just as subjective as other regulatory judgments, but all of them combine since the crisis to reconstruct commercial banking. More disclosure by making CAMELS public will correct for regulatory secrecy and market torpor, but not for the far larger question Mr. Hensarling has rightly posed: what are banks for?