Why our Fundamental Approach to Banking Regulation is Inherently Unsound

By William K. Black

(cross-posted from Benzinga.com)

Greetings from the North American Securities Administrators Association (NASAA) annual enforcement conference in Charleston, S.C.  I’m giving the keynote address Monday.  I’ll discuss the NASAA members’ exceptionally important and often effective role against securities fraud in future columns.  This column, however, deals with “safety and soundness” banking regulation.  

Our current approach to banking regulation exposes us to recurrent, intensifying financial crises.  The good news is that because we reached an all time low in Basel II, Basel III almost has to be an improvement. The bad news is that Basel III has not re-examined the fundamental assumptions underlying the Basel process.  As a result, Basel III will be a variant on the common ineffective theme of banking regulation designed by economists and the industry. 

The Basel process is built upon three flawed assumptions.

1. Capital requirements are the ideal form of banking regulation.
2. Capital requirements can be set without establishing sound accounting.
3. Accounting control fraud is not a serious concern.

Capital requirements are the ideal form of banking regulation under conventional economic wisdom.  The attraction of capital requirements to neoclassical economists is elegance.  Their theory is that while private market discipline ensures that normal corporations are inherently safe, private market discipline poses an inherent dilemma for banks.  A bank run is a form of form of private market discipline.  Banks have very short-term liabilities and longer-term assets.  This exposes them to interest rate risk and liquidity risk.  A run is the ultimate liquidity nightmare for a bank.  The conventional economic wisdom is that runs are not a desirable form of market discipline.  Economists tend to use the word “panic” when they describe runs.  Economists fear that depositors are likely to be financially unsophisticated and to start runs on banks on the basis of false rumors that the banks are unsound.

Deposit insurance is designed to prevent depositors from engaging in private market discipline.  The insurance limit is often set at a sufficiently high amount that the overwhelming bulk of depositors’ accounts are fully insured – minimizing private market discipline.  Central banks often provide a “lender of last resort” facility to allow the central bank to trump any run.  Many nations with advanced economies are so opposed to runs that they provide both deposit insurance and a lender of last resort facility through the central bank.  

The conventional economic wisdom is that deposit insurance renders private market discipline ineffective because banks’ principal creditors are fully insured depositors.  It is expensive for creditors to undertake the monitoring and analyses required to impose effective private market discipline, so fully insured depositors should not discipline banks.  The conventional economic wisdom has a further prediction:  the absence of private market discipline will increase the risk of moral hazard.  The conventional theory gets quite fuzzy at this point about how moral hazard works, a point I return to below, but it predicts that moral hazard can lead banks to take excessive risks.  The conventional economic wisdom further predicts that imposing adequate capital requirements will successfully constrain moral hazard.  As long as the shareholders’ have material capital at risk of loss should the bank fail they will not cause the bank to take excessive risks.  The shareholders’ incentives will be aligned with that of the public and the banks’ creditors as long as the bank meets its capital requirement.  The conventional wisdom, therefore, requires that the regulators force the bank to be promptly recapitalized or closed if it fails to meets its minimum capital requirement.    

The above analysis begins to explain why the conventional economic wisdom is that capital regulation is the optimal form of bank regulation.  The key is the alignment of shareholder’s interests with the public interest, but capital also provides a buffer against loss to the insurance fund and the taxpayers.  When the incentives are right there is little or no need for additional regulation.  Any rules that constrained bank decision-making (when the incentives were correct) would constitute the regulators substituting their business judgments for those of the banks’ officers.  The conventional economic wisdom asserts that private sector business judgments are vastly superior to regulatory decision (Easterbrook & Fischel 1991).  It follows that the conventional economic wisdom was that the banking regulators that regulated the least produced the best banking results.  Increased regulation did not simply increase cost; it increased the risk of banking failures and crises.  Less banking regulation allowed financial intermediaries to be more efficient and increased economic growth.

James R. Barth, Gerard Caprio Jr., and Ross Levine.  Bank regulation and supervision: what works best?  Journal of Financial Intermediation 13 (2004) 205–248; Barth, J.R., Caprio Jr., G., Levine, R., 2001a. Banking systems around the globe: Do regulations and ownership affect performance and stability? In: Mishkin, F.S. (Ed.), Prudential Supervision: What Works and What Doesn’t. Univ. of Chicago Press, pp. 31–88.

(Barth and his colleagues eventually differed from the conventional economic wisdom in being skeptical even of capital regulation of banks and urging greater reliance on private market discipline of banks.)

The conventional economic wisdom also claimed that small levels of reported capital were sufficient to create the desired incentives among shareholders.  In a bubble, bank loan losses are normally greatly reduced.  Economists began to argue that the lower the banks’ capital requirement the greater the amount of productive loans that would be made and the faster the economy would grow.  Basel II substantially reduced capital requirements. 

The fundamental disconnect with making capital requirements the pillar of banking regulation is that “capital”, “net worth”, and “equity” are accounting concepts.  They have no meaning outside of accounting.  Worse, they are all residual accounting concepts.  Accountants do not, and cannot, count a modern bank’s “capital.”  They determine assets and subtract liabilities to determine capital.  The implication of that is that the accuracy of reported “capital” depends on the accuracy of the valuation of every asset and liability.  That means that capital is not only an accounting concept, but the accounting concept most subject to error.  For a large bank, there are literally tens of thousands of ways to use accounting to distort reported capital by enormous amounts.  Beyond the obvious – understate liabilities and overstate asset values – banks are the perfect vehicles to self-fund “capital.”  Accountants do purport to count “capital” when there is a purchase of newly issued stock or a capital contribution.  Savings and loans and the Big Three Icelandic banks self-funded the purchase of newly issued stock by insiders, cronies, and shills.  Anglo-Irish Bank self-funded the purchase of shares from a distressed shareholder to prevent the sale of a large block of shares in the market.    

Banks can self-fund purported “capital contributions.”  The person controlling the bank, for example, can purport to contribute $10 million in capital to the bank by contributing real estate (improperly) valued at $25 million to the bank and receiving $15 in cash from the bank.  If the real estate actually has a market value of $10 million he will make a profit of $5 million.  The bank will suffer a real loss of $5 million but will falsely report that its capital has increased by $10 million.   Its capital will be overstated by $15 million.

Banks also self-fund reported “income,” which can flow through to capital.  I discuss this in more detail below, but the overall result that needs to be understood is that self-funding can be used to report guaranteed, record income and capital.

All of this means that accurate accounting is essential for banking regulation premised on capital requirements to succeed.  The Basel process relies primarily on capital regulation, but ignores the accounting games that allow banks to create their reported capital.  Bank examination and supervision, globally, puts only minimal emphasis on accounting in the era leading up to the crisis.    

The failure of Basel and the regulators to make accurate bank accounting their central priority would be dangerous even if accounting control fraud did not exist.  In the world of modern finance where accounting is the “weapon of choice” for control frauds, the failure to take accounting seriously was catastrophic.  The four-part recipe that bank control frauds use to produce guaranteed, record fictional short-term income turns regulatory regimes based on capital regulation profoundly perverse. 

1. Grow extremely rapidly
2. By making loans to the uncreditworthy at premium yields
3. While employing extreme leverage
4. While providing only trivial loss reserves (ALLL)

Akerlof & Romer (1993) emphasize that accounting fraud is a “sure thing.”  If a bank can produce guaranteed, record income then it can appear to be healthy.  Regulators are taught to worry about banks showing losses – not record gains.  A bank reporting record income can pay its controlling officers huge compensation and still have plenty of fictional net income to flow through to fictional capital.  Regulators are taught to believe that firms reporting adequate capital have the correct incentives and have a buffer that will protect the FDIC against losses.

The fictional increase in income and capital makes it easy for the bank to meet the first ingredient – extremely rapid growth.   It also makes the regulators feel comfortable about the bank employing extreme leverage.  The fourth ingredient is an essential ingredient of accounting control fraud.  The first three ingredients maximize real losses.  The expected value to the bank, for example, of making liar’s loans is sharply negative.  That means that the loss reserves (ALLL) that the bank should establish under GAAP should exceed the net income from the loan (i.e., the loss reserves should be large enough that the lender recognizes a loss on the liar’s loans when they are originated).  That would have meant ALLL provisions in the 20% range for liar’s loans.  Instead, ALLL fell each year in the peak of liar’s loan originations to roughly one percent.

Basel III is premised on the assumption that raising capital requirements will greatly reduce the risk of future failures and crises.  One can understand the logic.  Basel II reduced capital requirements and failed banks followed extreme leverage.  Special investment vehicles (SIVs) employed exceptional leverage and many SIVs failed.  The regulators are correct that leverage matters – it is the third ingredient in the lenders’ accounting fraud recipe.  What the regulators have not taken into account is a series of means of gaming reported capital that render capital requirements malleable.  Instead of correcting these accounting abuses they have stood by, or in the case of Ben Bernanke encouraged, the destruction of the remaining integrity of accounting standards.  Bernanke encouraged the Chamber of Commerce and the banking lobbyists to use their political allies to extort the Financial Accounting Standards Board (FASB) to junk the rules requiring banks to recognize their losses.  This massively overstates asset valuations, which massively overstates reported capital – evading the requirements of the Prompt Corrective Action law.  It also overstates income, allowing bank officers to enrich themselves through bonuses they had not earned.  Having just gimmicked the accounting rules to achieve their goals of covering up the scale of the crisis (and claiming to have “resolved” the crisis for a pittance), it is bizarre that the banking regulatory agencies treat capital requirements as if they had meaning independent of accounting.  A sound system of banking regulation cannot be based on capital regulation as it is conceived in the Basel process.