U.S. bank regulators are reexamining their risk-based capital rules. Many believe the rules are not working well and should be modified significantly or even discarded.
Risk-based capital assigns different risk weights and capital to the categories of assets held by a bank. For example, a Treasury note has a risk weight of zero, a single-family mortgage loan 50%, and a commercial loan 100%.
Risk-based capital measures have been around for a long time. Various Federal Reserve banks experimented with them for decades. They became official when the Basel Committee of international bank supervisors adopted them a decade ago.
During my tenure as chairman, the Federal Deposit Insurance Corp. was opposed to risk-based capital standards. We believed the risk-based rules had a number of deficiencies, such as:
They look backward, not forward, by focusing on the past performance of portfolios. Some of banking's most serious problems have developed in portfolios that were once deemed relatively risk-free (e.g., sovereign risk loans in the 1970s).
They are enormously complex, and yet they fail to measure risk adequately. For example, they measure only credit risk, not market and operating risks. (Newly issued rules will attempt to cover market risks.) Moreover, they assign the same risk weight to all commercial loans, whether secured or unsecured.
They minimize examiner judgment and discretion in favor of rigid rules that can be circumvented. They give false comfort that a bank in technical compliance with these arbitrary rules is in great shape. It might or might not be.
They single out banks and ignore the competitors of banks, which are now much bigger players than banks in the financial services marketplace. Securities firms and insurance companies have lower and less-rigid capital requirements. Most other competitors have no regulatory requirements for capital.
They can lead to credit allocation. If we want to encourage certain types of loans, such as housing loans, we simply reduce the risk weight assigned to them.
Although the shortcomings of risk-based capital rules are now generally recognized, there's no consensus on what to do about them. Federal Reserve Chairman Alan Greenspan suggested recently that banks develop their own risk-based capital models.
The regulators would rely on those models if they appeared reasonable. Presumably the bank would be penalized severely should its model fail to perform.
Though this approach is probably superior to the current system, it, too, has serious flaws. Can a model based on past experience predict future events with sufficient accuracy? How will we know before it's too late? How can we enforce such a system in a way that's fair to all banks? Will we need different rules for large and small banks?
I believe the regulators should return to the system they used prior to the risk-based capital rules. They should establish a minimum ratio of tangible equity capital-to-assets (say 5%).
No bank, no matter how good it looks, could maintain less capital than this minimum. Bank examiners, officers and directors, auditors, and the marketplace will dictate how much additional capital any given bank should have based on their assessments of that bank's risk profile.
This system is simpler and easier to understand and enforce. It allows management, regulators, auditors, and marketplace participants to exercise discretion and judgment in determining the appropriate level of capital for each bank.
The system has checks and balances. If any one of these groups believes a bank's capital is too low, the bank will need to respond.
There's no foolproof system for determining capital adequacy. Indeed, that's the principal flaw in the current risk-based capital rules: They purport to be more sophisticated in measuring capital adequacy than they in fact can be.
A simpler system that purports to do less would be superior. The risk- based capital rules should be scrapped.