That which is perceived as risky, that which scares off everyone – especially bankers – is a source of little danger to the banking system.

The really big danger for it has been, and always will be, what is perceived to be absolutely safe, based on assumption and prediction, but that with knowledge and hindsight turns out to have been very risky.

The capital of banks should therefore primarily cover the unexpected losses. And, since these by definition are unpredictable, and given the difficulties to determine which assets are more likely to generate unexpected losses, a sensible approach would be to require banks to hold the same level of capital against any asset.

But, no! Regulators decided, with Basel II, that the capital requirements for banks were to be based on the same perceptions of expected losses which banks already price for by means of interest rates, size of exposure and other terms…and so the regulators blithely ignored the possibilities of unexpected losses.

As a consequence, banks are now required to hold much more capital against "risky" assets than against "safe" assets. And as a consequence of that, banks now obtain a much higher risk-adjusted return on their equity when lending to "the infallible" than when lending to "the risky".

And so these risk-weighted regulations doom the banks to lend too much, on too easy terms, to "the infallible" borrowers, such as governments, the housing sector or the AAAristocracy, so that some of them as a result will, sooner or later, turn out to be very risky.

A prime example of the above, turning safe into risky, is what happened with securities backed with subprime mortgages. As a consequence of the approval of Basel II, in June 2004, financial institutions (especially investment banks overseen by the Securities and Exchange Commission and all European banks) could hold a meager 1.6% in capital against these securities, if they were rated AAA. And so the possibility of being able to leverage bank equity 62.5 times to 1 created an extraordinary demand for these bonds. If the market, being what it is, cannot supply enough real AAA rated products, it will find the way to satisfy the demand with Potemkin ratings.

(In the Financial Times in January 2003 I warned, "Everyone knows that, sooner or later, the ratings issued by the credit agencies are just a new breed of systemic error to be propagated at modern speeds. Friend, please consider that the world is tough enough as it is.")

Also just consider how European banks lent too much to Greece, only because they were required to hold minimal capital when doing so. 

The effect of these risk-weighted regulations is especially severe when banks, like now, suffer from a general lack of capital and, in order to adjust to more severe regulatory requirements, retrench into the "safe" assets which require less of it.

Simultaneously, this situation also dooms the real economy to having banks lend too little, and too expensively, to "the risky," those who in fact most need bank financing on competitive terms, like small and midsize businesses, entrepreneurs and start-ups.

Why did the regulators make this monumental mistake that so distorts the allocation of bank credit to the real economy? First and foremost, it happened because regulators really never defined the purpose of banks. And we also find some clues in Nobel Prize winner Daniel Kahneman’s extraordinary book "Thinking Fast and Slow."

In Chapter 32, "Keeping Score," Kahneman writes: "Enhanced loss aversion is embedded in a strong and widely shared moral intuition…The dilemma between intensely loss-averse moral attitudes and efficient risk management does not have a compelling solution."

This aversion would make it hard to digest the possibility of requiring banks to hold more capital against what seems safe than against what seems risky, even if the whole world knew that what is perceived as "absolutely safe" poses the biggest dangers to the banking system. The chances of marketing such a concept to an anxious general public seem very slim indeed.

But why is this horrendous mistake not even discussed and allowed to live on? In some ways Basel III makes it even worse by adding liquidity requirements which are also very much based on subjective perceptions already priced for by many in the markets.

In this respect, Kahneman writes: "The decision to invest additional resources in a losing account, when better investments are available, is known as the sunk-cost fallacy, a costly mistake that is observed in decisions large and small … Cancelling the project will leave a permanent stain on the executive’s record, and his personal interests are perhaps best served by gambling further with the organization’s resources in the hope of recouping the original investment-or at least in an attempt to postpone the day of reckoning"

And so… do we need to explain more? Basel III is the sunk cost error of Basel II, a doubling down on a soured bet.

Per Kurowski was an executive director at the World Bank from 2002 to 2004. He writes the Subprime Regulations blog from Rockville, Md.