In recent weeks, policymakers have begun to turn from ever-more detailed revisions of the Basel III capital rules to a dawning awareness that these standards are unduly complex.

I hope they also recognize soon that these dizzying rules are dangerously contradictory when taken in context with other pending regulatory initiatives.

A cure to this chaotic situation is urgently needed, but it isn’t the "back to basics" approach advocated on Sept. 14 by Federal Deposit Insurance Corp. board member Thomas Hoenig. 

To reject all of the risk-based rules in favor of a single tangible-equity to tangible-assets standard, as Hoenig advocated, is akin to abandoning surgery with complex anesthesia for the bite-a-bullet option.  It’s simpler, but a lot more painful.

In his address to the American Banker's Regulatory Symposium, Hoenig made a forceful case for junking Basel I, II, II.5 and III.  He would, instead, rely on a revised version of the U.S. leverage standard, simplifying it further to the aforementioned tangible numerator and denominator.  This, he argues, will ensure that bank capital is a robust form of risk absorption immune to gaming by "brazen"big banks that use their model-building skills to exacerbate the plight of simpler, better community banks.

Like a major paper presented at the Fed's Jackson Hole conference – "The Dog and the Frisbee," by Andrew Haldane and Vasileios Madouro of the Bank of England – Hoenig's framework is premised on the belief that capital is the crux of bank regulation. He describes it as the "cornerstone on which a bank's balance sheet is built.”  This is true as far as a balance sheet goes, as the point of it is to segment holdings into assets and liabilities so one can quickly determine which is larger and, thus, how sound a bank may be.  But a "fortress" balance sheet based on a capital armory isn't the only cornerstone of safe and sound banking. 

As we have learned over and over again, banks fail for lots of reasons and sometimes solvent ones evaporate under attack.  This attack can come from within – management self-dealing or outright fraud (think S&L crisis) or from external shocks like the liquidity risk caused by Lehman's failure, which created severe contagion risk throughout the financial system, pummeling well-capitalized banks.  Operational risk comes not just from within, but also and often most devastatingly from without – 9/11, of course, and more recently the Japanese earthquake and tsunami. 

So, even if we built up enough capital to prevent credit or market risks from scaling a fortress bank's balance sheet, risks still remain that require a far more nuanced and, yes,  complex regulatory framework.

But assume Hoenig concedes that prudential policy needs more than just one plank comprised of capital.  Can we then give him his plan for the capital plank in this wider floor?

Let's separate the Hoenig ratio into its numerator and denominator to assess what's wrong with each.  First, the numerator: total tangible equity. 

Hoenig is right that complex hybrid instruments proved a weak form of capital on which to rest risk determinations.  Such instruments were indeed discounted by markets when banks came under stress in favor of tangible common equity.  Thus, TCE should be the bulwark of the capital numerator, as Basel III indeed demands.

But what of the other instruments now counted as capital, especially in Basel's Tier 2?  Some of these instruments may be a bit dodgy, but one – loan loss reserves – is a critical component of regulatory capital. In my view, it should be reflected in regulatory-capital calculations to eliminate the perverse incentive baked into all of the Basel rules: Adding reserves doesn't improve capital, making it easier for banks to skimp on reserves or hope that equity holds up under stress.  Capital may be meant for unexpected loss and reserves for the expected, but the more regulatory standards join accounting rules in discouraging high reserves, the less cushioning a bank has for all of the risks it runs.

The tangible-asset denominator is as full of traps as the seemingly simple TCE.  Basel I recognized that off-balance sheet assets cause risk and incorporated this into its risk-weighting scheme.  However, the rules exempted from capital calculations any off-balance-sheet obligation (e.g., a letter of credit) with a maturity of less than one year.  Guess what?  In about a nanosecond, every off-balance sheet position every bank held turned into a 364-day "risk-free" one. 

By 2008, this loophole in Basel I turned toxic.  Any number of very large banks sat on giant holdings of structured-investment vehicles and other highly-engineered financial instruments that hid risk from the regulatory-capital account.  So, if we just go to on-balance sheet assets, we go back to ignoring off-balance sheet ones, and thus unlearn a major hard lesson of the financial crisis.

What if the tangible-asset part of Hoenig's simple equation is finessed a bit to add off balance sheet instruments?  Does this prevent "brazen" bankers from gaming leverage?

Again, history is instructive.  Under Basel I in the U.S., we had both a leverage rule and, to discipline it, risk weightings.  But both were very simple.  Under this standard, a no-doc mortgage with no down payment or record the borrower had ever paid a bill got the same leverage requirement and 100% risk weighting as those given to a solid-gold corporate loan to a proven hometown employer.  Guess where the money went.

Very simple solutions are truly tempting as one tries to plow through Basel's blizzard of ever more complex standards based on increasingly opaque models designed to capture every risk under each scenario in all countries for thousands of banks.  That's too complex and can't be done.

Too simple may not be as risky as too complex – at least one gets a clear number from Hoenig's equation in contrast with the Basel III that standards even sophisticated banks have trouble calculating.  But, just because one gets a simple number doesn't make it the right number.  If it misrepresents risk or – worse – creates strong incentives to take on still more of it, simple is the wrong solution. 

The right answer is to build a few clear, transparent risk indicators – capital one among them – and then to hold banks and regulators accountable for meeting them.

Karen Shaw Petrou is a managing partner at Federal Financial Analytics in Washington.