Banking Soundness and Monetary Policy


Charles Enoch, and J. Green
Published Date:
September 1997
  • ShareShare
Show Summary Details

Julio Escolano’s paper, “Tax Treatment of Loan Losses of Banks,” provides a comprehensive discussion of the issues by adopting an international comparative perspective. In the following comments, I have attempted to summarize and discuss the basic themes and conclusions of Escolano’s paper. Since his emphasis is on the fiscal aspects of taxing loan losses, my focus will be directed to the relevance of the tax treatment for prudential standards and banking system soundness, which should be of particular interest to central banks and bank supervisory agencies.

Escolano offers two reasons why the tax treatment of loan losses should be of interest to fiscal authorities. First is the concern that a poorly designed tax treatment of loan losses may lead to a decline in tax revenue. Second is the fact that international convergence in defining credit risk and capital adequacy establishes a common basis for an internationally harmonized tax treatment of loan losses. A third reason might be that introduction of an adequate regime of tax deductibility of loan losses sometimes entails major revenue shortfalls in the short run. Even though such revenue shortfalls occur only once (and are not a permanent reduction of revenue), such a one-time expense may be sizable relative to the country’s budget. This problem has occurred in many transition countries. One policy response would be to cushion the budgetary impact through a phase-in period.

Although not treated directly in the paper, the issue of the tax treatment of loan losses is of great interest to banking authorities because a well-designed tax system constitutes a strong positive incentive for prudent bank behavior. When banks have a financial interest in adequately recognizing losses, they will be more motivated to do so than without such an incentive. On the other hand, poorly designed tax treatment of loan losses weakens prudential incentives.1

This is a pragmatic observation, which recognizes that bank managers have a tendency to postpone recognition of losses because in the short run it lowers their profits. Experience shows that management almost always prefers to postpone the bad news for some later point. Of course, it could be argued that regardless of the tax system, prudential rules (supported by accounting and disclosure requirements) can also force banks to recognize loan losses and are equally effective. This is true, but enforcing such rules (whether by the authorities, by independent auditors, or by bank shareholders) is labor intensive, time consuming, and, therefore, expensive. Enforcement also adds to the cost of capital. When bank supervisors can operate with a tailwind from the tax system, they can economize on such expenses, and the banking system can provide services more efficiently.

The banking authorities have a strong self-interest in a well-designed tax system. From this observation, an important policy implication follows: where tax rules fail to permit adequate tax deductions for loan losses, the banking authorities should be actively engaged in a dialogue with the relevant authorities in order to introduce a satisfactory tax regime. A case in point was the international debt crisis in the 1980s. Banking authorities in many countries became involved in coordinating the prudential and tax aspects of writing down problem cross-border loans held by commercial banks. This was done to make it easier to reschedule debts. The case of “country loans,” as these were called, has remained a special and exceptional case. However, as globalization of banking markets proceeds, the coordination of regulatory and tax issues becomes relevant for all aspects of bank lending.

The basic thesis presented in Escolano’s paper is that loan losses constitute a necessary business expense of engaging in lending activity and should, therefore, be deducted from the tax base; loan losses are equivalent to the depreciation of fixed assets. It would be useful to explore this parallel more fully. As in the case of bank loans, the monetary value of fixed-asset depreciation is not always readily observable and must be estimated. While depreciation allowances are widely recognized as a tax-deductible expense, the tax treatment of loan losses is controversial in many countries. It can be a useful exercise to illustrate (as is the case in many countries) that the estimates for loan losses are based on much more objective indicators and much more closely mirror the loan’s current market value than other depreciation schemes. This is a strong argument in favor of granting tax deductibility of loan losses.

If one accepts the tax deductibility of loan losses, when should these losses be recognized? Correctly, Escolano states that a loan loss should be recognized as soon as an economic loss has occurred. But what is the definition of an economic loss? Although there is no unambiguous operational definition, economic losses must be defined within a given institutional, legal, and accounting environment. Because there are significant differences among countries, making a globally valid definition of economic loss nearly impossible. Therefore, Escolano tries to determine the most suitable method to approximate economic losses.

What is the best method to reflect the economic loss? Escolano discusses three different approaches to recognize an economic loss and, justifiably, concludes that the “specific provisioning” method appears to be best. This approach suggests that an economic loss occurs when a bank recognizes a reduction in the current value of a loan by establishing a specific provision. Of course, as an accounting principle, a specific provision is a cost item; it is (or should be) booked on the asset side of the balance sheet as a “contra-account,” thus lowering the value of total assets. The specific provisioning approach for the tax treatment of loan losses builds on loan valuation methods developed and applied for prudential and risk management purposes.

The specific provisioning method has several elements. It is worth expanding on Escolano’s conclusion that the specific provisioning method is mostly successful in capturing losses soon after they have occurred. Banking authorities in many countries apply loan classification systems whereby each loan is categorized in classes such as “substandard,” “doubtful,” and “loss.” Such systems use specified indicators of declining loan value, including criteria such as the number of days debt payments are overdue, bankruptcy of the borrower, and so on. For each class of loans, banks may be required to set aside a required provision, for example, 25 percent for “substandard,” 50 percent for “doubtful,” and 100 percent for “loss.” Such rules and provisioning requirements may be set up either by bank regulators or by the banks themselves (and should be subject to scrutiny by external auditors). There is no single best system to account for depreciating values of loans. A good model, of course, should fit the national institutional setting, as well as the type of loans offered in the banking market. From a prudential point of view, what counts is that such a system has proven to reflect loan performance accurately, that it is internally consistent and stringently applied. Under such circumstances, it should be broadly acceptable to the fiscal authorities as a guideline for tax deductibility as well.

An important point to emphasize is that loan classification systems reflect an estimate of the actual (current) market value. They do not reflect anticipated (future) losses, as is often falsely assumed. The interpretation of loan loss provisions as anticipated, as opposed to actual losses has led the tax authorities in some countries to refuse tax deductions for specific provisions.

In which countries are there controversies between fiscal and monetary authorities over the tax treatment of loan losses? Escolano approaches this issue by looking at the conformity between regulatory and tax treatment. The sample is limited to industrial countries and within the sample four different groups are identified: the “high” conformity countries, such as Denmark, France, and Germany, the “low” conformity countries, including Australia and the United States, and two groups of countries in between. Grouping countries by “conformity,” however, poses some problems. First, conformity is only a positive sign if the prudential system is adequate and well enforced. Second, conformity of rules does not necessarily permit any conclusions on the actual tax burden. As illustrated in the case of the United States, low conformity of tax and regulatory approaches does not necessarily mean that on any given asset portfolio, U.S. banks are subject to a higher tax burden than are their counterparts in Germany, a country with high conformity. As pointed out earlier in the paper, even though U.S. tax rules do not permit the tax deductibility of specific loan loss provisions, the method used in the U.S. (the charge-off method) in practice turns out to be very similar to the specific provisioning method, because of the possibility of partial loan write-offs.

This example shows that a “ranking” of countries’ tax deduction rules, according to the degree to which they reflect economic loss, remains a difficult task. Conformity of rules, as such, does not appear to provide a satisfactory answer. Defining best practice standards of appropriate tax and regulatory treatment of loan loss provisions is, perhaps, an area for future research.

For a discussion on the relationship between regulatory and tax treatment see Claudia Dziobek, “Regulatory and Tax Treatment of Loan Loss Provisions,” Paper on Policy Analysis and Assessment 96/6 (Washington: International Monetary Fund, 1996).

    Other Resources Citing This Publication