TEL: (212) 612-9200
FAX: (212) 612-9253

July 29, 2002

Mr. Jonathan G. Katz,
U.S. Securities and Exchange Commission,
450 Fifth Street, N.W.,
Washington, D.C. 20549-0609

Re: Disclosure in Management's Discussion and Analysis
about the Application of Critical Accounting Policies;
File No. S7-16-02 ___

Dear Mr. Katz:

The New York Clearing House Association L.L.C.1 (the "Clearing House") appreciates the opportunity to comment on the Securities and Exchange Commission's proposal (the "Proposal") regarding disclosure of critical accounting policies in Management's Discussion and Analysis (the "MD&A").

Our member banks strongly support the basic goal set forth in the Proposal of improved disclosure. A number of the recommendations in the Proposal advance this goal. We are concerned, however, that certain other recommendations in the Proposal will create confusion rather than clarity and mislead rather than inform.

We agree that investors should be informed when material aspects of an issuer's financial statements involve estimates. We believe, however, that it is often not merely unhelpful, but actually counterproductive, to attempt to quantify the consequences of different estimates. We also believe that it is essential to recognize that the Commission's understandable interest in expanded disclosure, and its advocacy of current as opposed to historical accounting, require increased judgments by registrants and that these judgments are often subjective and complex. We urge the Commission to make clear that estimates made in good faith do not violate the federal securities laws.

1. Definition of Critical Accounting Policies: Need for Comparability (Part C,1).

Although we understand the Commission's need to be flexible in determining what constitutes "critical accounting policies", an equally important goal should be feasibility. If registrants are unable to implement the proposed disclosure in a concise and understandable manner because the definition of "critical accounting policies" is too broad or too vague, the investor is likely to be confused rather than enlightened.

The proposed definition begins with the criterion that the matter must be "highly uncertain" at the time the estimation is made. How is this criterion applied in practice? For example, does it refer to the degree of likelihood that the estimate proves incorrect to some extent, the degree of likelihood that, if the estimate proves incorrect, the deviation is substantial, or both?

The second criterion compounds the uncertainty. Depending upon a registrant's reading of this criterion, it is likely to be either too inclusive or too exclusive. Once a company has made its best judgment as to an estimate, it can be argued that there is frequently not any different estimate that qualifies as reasonable. On the other side of the issue, it can be argued that estimates by their nature are so lacking in certainty that there must always be reasonable alternatives. Likewise, the argument could be made that estimates are almost always reasonably likely to change from period to period -- particularly if this refers not to each successive quarter but over a series of quarters.

The materiality test does little to add precision to this definition. At a time when the market often reacts sharply to an earnings report that differs from analysts' estimates by as little as one cent, almost any change in an estimate could have a material impact as defined in SAB 99.

We believe that the appropriate approach is that which the Commission adopted only a few months ago. Registrants should be asked to discuss the small number of accounting principles that the registrant considers most important to the presentation of its financial results.

2. Qualitative Discussion (Part C,2).

We generally agree with the qualitative discussion provided for in the Proposal. The breadth of the qualitative discussion proposed by the Commission obviates the need for a quantitative discussion, which, as discussed in the next part of this letter, has serious flaws. An investor has the information that he or she requires and that is clarifying rather than confusing, when: (i) the policy is identified; (ii) the methodology used by the registrant in developing the policy is described; (iii) relevant assumptions are identified; and (iv) known trends or uncertainties that are likely to affect the assumption or methodology are identified.

3. Quantitative Disclosure (Parts C,3 and C,4).

At the heart of our concerns about the Proposal is our strong disagreement with the quantitative disclosure approach. At least in the case of banking organizations, quantitative disclosure of other possible estimates will prove confusing rather than illuminating.2 Indeed, we are concerned that, at a time when financial statement integrity is a paramount consideration, such disclosure will cast doubt on the credibility of all financial statements.

We can all agree that transparency of financial statements is a laudatory goal. It is important, however, not to confuse numbers and transparency. Unless numbers are derived with accuracy and are truly meaningful and comparable, they create obfuscation rather than transparency.

We believe that the validity of our concerns can be illustrated by a discussion of what is probably the single most important critical accounting policy for most banking organizations -- the determination of the loan loss reserve. The determination of this reserve is based on a wide variety of both macro and micro factors. They include the state of one or more regional economies, the state of the national economy, the state (in many cases) of the global economy, interest rates, conditions in specific industries, a wide variety of government economic and other policies, geo-political developments, legislation and regulation, and the specific financial condition of what may be as many as tens of thousands of individual, corporate and other borrowers.

Although it is possible to describe many of these assumptions, it is not possible to describe others, and it is totally impossible to predict with any meaningful accuracy the impact if they prove incorrect. Four examples may help illustrate this point. First, in establishing reserves it is generally assumed that there is not any material change in the current U.S. taxation system. If, however, a change were to occur, its consequences are unfathomable. A change 15 years ago in the tax code is widely believed to have contributed significantly to the collapse of the real estate industry, which in turn resulted in substantial loan losses in the banking industry. Second, it is assumed that the audited accounts of borrowers do not involve fraud, massive or otherwise. Is it truly possible to predict the full consequences of a massive fraud at one borrower, much less massive frauds at multiple borrowers? Third, it is generally assumed that there is no major change in global geo-political conditions. Is it possible to predict the consequences of a development such as Iraq's invasion of Kuwait, the resultant sharp spike in oil prices and the impact on U.S. borrowers? Fourth, is it possible to quantify the impact of a decline in an entire industry, whether it be telecommunications or automobiles?

Moreover, the number of assumptions involved creates a virtually infinite number of possibilities. Which assumptions will be held steady and which will move? Will they move in offsetting directions or the same direction?

We also find it inconsistent for the Commission to ask bank holding companies to describe the impact of future events on the allowance. The Commission has argued vigorously that banking organizations should not consider possible future events when determining the amount of the allowance. Is it not inherently confusing to investors to require a description of events that might affect the amount of the allowance in the future, but cannot be taken into account when establishing the allowance in the present?

Neither of the two possible choices suggested in the Proposal for presenting changes resolves this problem. The identification of the most material assumption or assumptions is not realistic when so many assumptions are involved. As mentioned, perhaps the assumption that proved to have the single greatest impact on banks' loan reserves during the past 50 years -- when it proved incorrect -- was the steady state of the Internal Revenue Code. Yet, was such a change truly more than remote? Even if one were to broaden the assumption and conclude that the category was U.S. legislation -- so that the change is "reasonably possible" -- how could any institution conceivably quantify the impact of possible changes.

The "range" option is no better, because our banks (and we believe most others) do not utilize ranges in developing their overall loan loss reserves. Although ranges of possible losses may be used in the case of some specific borrowers, correlative moves are not likely. There is no reason to assume that each borrower for which a range is used will be either at the top or bottom of its specific range.

We respectfully suggest that the approach of quantitative disclosure is so fatally flawed that the Commission's specific inquiries on this issue are basically beside the point. No quantitative approach for describing changes will work; whether one or both of the approaches suggested in the Proposal is used will not ameliorate the problem; and standardized, i.e. numerical, tests cannot be applied to largely qualitative factors.

We want to mention one other problem with a quantification requirement. In a number of cases, it is likely to force registrants to disclose proprietary or competitively sensitive information. For example, if a company were required to disclose specifics about its litigation reserves and possible variations, that information could be very helpful to the litigants on the other side of the controversy. Quantification of possible changes in tax reserves could reveal a registrant's tax strategy. Although accurate disclosure should "trump" competitive considerations, the reverse should be the case for information of marginal or debatable value to investors.

The requirement for a quantitative discussion of past changes in accounting estimates is equally flawed. We support the Commission's proposal for a qualitative discussion, but the quantitative discussion suggests a precision that is unrealistic.

As discussed, the establishment of the loan loss reserve is a function of numerous assumptions. When, over time, a loan loss reserve moves from x to x+ or x-, that change represents an amalgam of all those assumptions without specific identification. A banking organization could rarely, if ever, attribute the difference to a single change in an assumption, much less allocate the change to a series of different assumptions.

The Commission has repeatedly stated that it wants investors to be able to view a company's financial statements the same way as management views them. We agree that this is a worthwhile objective. If, however, the Commission is committed to that objective, it must eliminate the quantitative approach contained in the Proposal. Bank management does not utilize such an approach.

The requirement of quantification is also inconsistent with the Commission's efforts to reduce the use of pro forma financial statements. Various quantifications of the impact of changes in accounting policies create a wide range of constantly changing pro forma statements.

Finally on this subject, we believe that the three examples provided in the release all arise out of a relatively simple manufacturing situation. They do not readily apply to financial institutions, and also do not apply to more complex issues, such as pensions, that apply to all registrants. We recommend that the Commission consider the establishment of industry working groups (which could include users as well as registrants) before deciding to pursue a quantification approach.

4. Senior Management Discussions with the Audit Committee (Part C,5).

Our member banks' senior managements currently discuss their critical accounting estimates with their audit committees, and we believe that these discussions should occur at all companies. This subject, however, is only one of a number of significant accounting issues that are, and should be, discussed with the audit committee. Accordingly, we question why this one matter alone should be disclosed in the MD&A as a matter discussed with the audit committee, which potentially creates an inaccurate negative implication.

We also question whether disclosure of an audit committee's procedures and deliberations will add materially to investors' mix of knowledge. Accordingly, we do not support the additional disclosures described in the Proposal's questions.

5. Divergent Accounting Policies (Part C,5).

The last question on audit committees relates to whether the registrant's accounting policies diverge, "to its knowledge", from the policies "predominantly" applied by other companies in the same industry. At the outset, even if this subject is to be discussed, there is no reason to link it to the audit committee.

More substantively, we agree that it may be theoretically meaningful for investors to know whether a registrant's accounting policies are divergent from peer policies. As practical matter, however, in most cases, the registrant will not only be unaware of such a divergence, but have no means of making an informed and complete determination.

Accordingly, if the Commission were to provide for this type of disclosure, it should: (i) specifically state that registrants will normally not have this knowledge and that disclosure is required only if a registrant has actual knowledge and (ii) provide that any divergence should not be reported unless it is material.

6. Segment Disclosure (Part C,6).

We believe that it is appropriate to include a discussion of critical accounting policies in the segment disclosure. Once again, however, we believe that such discussion must be qualitative only.

7. Independent Auditors' Examination of the MD&A (Part E).

We do not believe that it is appropriate to require the independent auditors to examine MD&A disclosures regarding critical accounting estimates. Stated simply, this would not be an audit function, but a management function. The auditors are not in a position to second guess management as to whether the relevant factors in arriving at estimates are the correct ones or whether they are properly weighted.

The Proposal inquires about the relevant costs and benefits. We believe that the costs will considerably outweigh any benefits, particularly after considering the inevitable qualifications that auditors will place on the results of their examinations.

8. Initial Adoption of Critical Accounting Policies (Part G).

The Clearing House has several concerns about the Commission's proposed disclosure regarding initial adoption of critical accounting policies.

First, in our view, when a registrant adopts a critical accounting policy, there is only one acceptable reason for its adoption: The registrant believes that the policy is the most appropriate in the circumstances to present fairly the registrant's results of operations and financial position. The Commission should not expect that the disclosure of the reasons for adoption would be more expansive, although there could appropriately be a discussion of the background and impact of the adoption of the policy.

Second, based on the experience of our member banks, there are very few situations where there is truly a "choice" among acceptable accounting policies. Rather, there is normally one "choice" that is best supported by the facts and circumstances, and, therefore, is the "right" accounting. There may be alternatives, but a discussion of alternatives that were rejected, typically because they did not seem as appropriate and informative as the accounting that was selected, does not improve financial statement disclosures. It would confuse rather than clarify, particularly because the impact of an "alternative" is often speculative.

Third, as discussed above (Section 5 of this letter), we believe that registrants will often not know about the accounting policies adopted by others. We believe that disclosure should only be provided in those rare situations where the registrant has actual, concrete knowledge.

9. Disclosure Presentation (Part H).

We agree that all disclosure should be as understandable as feasible. Nonetheless, it is important to develop an appropriate balance between disclosure that is meaningful and disclosure that must be reduced to a "lowest possible denominator" of financial expertise, particularly when the subject is as complicated as estimates. Based upon our experience, the Commission has devoted substantial resources to its "plain English" project, and, as desirable as the objective might be, we questions the cost/benefit analysis in a number of specific cases.

10. Application of Safe Harbors (Parts H and K).

We fully agree with the Commission that a discussion of crucial accounting estimates would qualify for the statutory safe harbor for "forward looking statements". If the Commission adopts quantitative disclosure, we can also understand why quantitative disclosure of the changes made to accounting estimates during the past three years ("backward looking" statement(s)) would not qualify.

We are concerned, however, by the suggestion in Part H of the Proposal about disclaimers of legal liability. We strongly believe that registrants should be entitled to disclaim legal liability -- even if statutory safe harbors are not directly available -- for estimates that are inherently subjective and often complex.

The need for such disclaimers becomes compelling if the Commission proceeds to require quantification. Investors should understand -- and registrants are entitled to explain -- that any such numbers represent "guestimates" and not accurate facts.

* * *

The Clearing House appreciates the opportunity to comment on the Proposal. If you have any questions, please call Norman Nelson, General Counsel of the Clearing House, at 212-612-9205.

Very truly yours,

1 The member banks of the Clearing House are: Bank of America, National Association; The Bank of New York; Bank One, National Association; Citibank, N.A.; Deutsche Bank Trust Company Americas; Fleet National Bank; HSBC Bank USA; JPMorgan Chase Bank; LaSalle Bank National Association; Wachovia Bank, National Association; and Wells Fargo Bank, National Association.
2 We note that the Proposal's three examples of disclosure about critical accounting estimates all relate to manufacturers. This suggests that the Commission may not have considered fully the implications of the Proposal for banking organizations.