July 19, 2002

Mr. Jonathan G. Katz
Secretary
U.S. Securities and Exchange Commission
450 Fifth Street, N.W.
Washington, DC 20549-6009

Dear Mr. Katz:

Deloitte & Touche LLP is pleased to respond to the Securities and Exchange Commission's (the "Commission") request for comments on its proposed rule regarding Disclosure in Management's Discussion and Analysis ("MD&A") About the Application of Critical Accounting Policies.

I. Introduction

We strongly support the Commission's goal to improve the quality and transparency of companies' financial disclosure. However, while certain of the disclosures discussed in the proposal may be useful to investors, much of the disclosure that would be required would result in excessive, detailed disclosure that is not useful and is potentially confusing.

Certain of the proposed disclosures about critical accounting estimates and the initial adoption of accounting policies are duplicative of current disclosure required in the footnotes to the financial statements under generally accepted accounting principles ("GAAP"). We do not believe it is necessary to repeat these footnote disclosures in MD&A. We believe that certain of the other proposed disclosures that are currently not required under GAAP may be useful. But, we believe that many of these disclosures would more appropriately be included in the footnotes to the financial statements instead of MD&A. An analysis of this information, including any sensitivity analysis that may be required, could then be provided in MD&A. In our view, the remaining proposed disclosures identified and discussed below would not be useful and should not be required. In addition, we believe that a "principle-based" approach to MD&A disclosure about critical accounting policies would be more effective than the prescriptive "rule-based" approach used in the proposal.

II. Usefulness of Information about Critical Accounting Policies

For many companies implementation of the proposed rule would result in a vast amount of additional, detailed information. This would result in numerous additional pages of disclosure in MD&A for many companies. This "disclosure overload" could make it more difficult for investors to focus on the arguably more important disclosures about material events and changes, trends, demands, commitments, events, and uncertainties. Many investors already confess that annual reports and MD&A are currently so long and complex that they do not read or understand the information, so a significant expansion of MD&A with excessive detail would not seem consistent with these investors' concerns. MD&A has become so lengthy that the Commission is even considering a proposal for a summary of MD&A.

We believe that certain of the disclosures proposed, such as an identification of the most critical accounting estimates and the most significant assumptions used in that estimate and why management changed an estimate during the current period from assumptions used in prior periods, may be useful to investors. However, as proposed, we do not believe it is useful to provide much of the other detailed disclosure, such as the quantified impact of available alternative accounting treatments, the methodology used by management to develop the estimate, and why materially different estimates might have been used in the current period. Investors may not understand how to interpret and use much of this detailed information. Investor confidence in the financial statements may actually be diminished as investors may question the appropriateness of legitimate estimates and accounting policies used because of the detailed disclosure about all the other estimates and alternative policies that could have been used. Further, an investor may not know how to compare the information to competitors of the company because of the different assumptions and estimates used and the different facts and circumstances of each company, which may result in speculative and inaccurate conclusions.

Consistent with the modernization and simplification of accounting standards, we believe the Commission should avoid the "rules-based" approach used in the proposal that could led to boilerplate, generic disclosures. Instead, general disclosure principles and guidance should be provided to allow companies to identify critical accounting estimates and policies and determine the appropriate level of disclosure to explain the impact of the accounting estimates and policies on the financial statements.

As acknowledged by the Commission in the proposal, current GAAP (including APB No. 20, Accounting Changes, APB No. 22, Disclosure of Accounting Policies, SOP No. 94-6, Disclosure of Certain Significant Risks and Uncertainties, and SFAS No. 5, Accounting for Contingencies) already require certain disclosures in the footnotes about accounting estimates and accounting policies that would be required in MD&A under the proposal. We do not believe it is necessary to repeat this information in MD&A. However, we believe that other proposed disclosures not currently required under GAAP may be useful, such as basic disclosures needed to understand critical accounting estimates, including a description of the estimate, an identification of how the estimate affects the financial statements and an identification of the affected financial statement line item.

It may also be useful in certain circumstances to disclose the assumptions used about highly uncertain matters and other material assumptions and to provide certain disclosures about the initial adoption of an accounting policy. However, we believe this type of information is more appropriately included in the footnotes instead of MD&A. The information already required under GAAP in the footnotes could be enhanced to include these types of disclosures. The SEC could consider additional guidance (through, for example, the issuance of a Staff Accounting Bulletin or a Financial Reporting Release) to enhance or interpret existing GAAP to include these additional disclosures. Alternatively, the Financial Accounting Standards Board or another standard setter could adopt such a requirement.

MD&A should primarily focus on an analysis of a company's results of operations and financial condition. For example, MD&A should include any sensitivity analysis that may be required and discuss the impact on the financial statements of critical accounting estimates that are reasonably likely to have a material effect should reasonably likely future experience differ from that assumed. MD&A should discuss the risks and uncertainties surrounding the estimate and the implications to the financial statements.

III. Definition of Critical Accounting Estimate

The two-part definition used in the proposal is complicated and may be difficult to understand and interpret. The use of "highly uncertain" is vague and may not be consistently applied. Instead, general guidance could be provided to describe a critical accounting estimate (for example, a critical accounting estimate is one where the financial statements would have been materially different if a different estimate that reasonably could have been used would have been used.) Companies are best able to examine their own particular facts and circumstances to determine which estimates are critical, without trying to use a strict definition. A strict definition may also allow companies to avoid disclosure about a particular estimate that might otherwise be considered critical except it does not fit into the literal definition of "critical accounting estimate".

The Commission has asked how many estimates would a company typically identify as critical accounting estimates under the proposed definition. Respondents to our survey¹ indicated that 13% would have 1 or 2; 67% would have 3 to 5; 13% would have 6 to 8; and 7% would have more than 8. The Commission also asked whether a company with multiple segments would have a greater number of accounting estimates than a company without multiple segments. Thirty-eight percent indicated they would have a greater number of critical accounting estimates if the critical accounting estimate was identified by segment: with 22% having an additional 1 or 2; 44% with 3 to 5 more; and none with 6 to 8 more; and 34% with more than 8. In response to the Commission's question of whether a maximum number of accounting estimates should be established, we believe general, practical guidelines may be helpful to avoid overly detailed disclosure. Although we believe no precise limit should be given, we generally believe that the substantial number of companies should emphasize no more than 3 to 5 critical accounting estimates and as the number of critical accounting estimates exceed six, the company should consider whether a practical limit has been reached (similar to the methodology and guidance used in paragraph 24 of SFAS 131, Disclosures about Segments of an Enterprise and Related Information.)

IV. Quantitative Disclosures to Demonstrate Sensitivity

Sensitivity analysis disclosures in MD&A may be useful in certain circumstances. Less than 20% of the respondents to our survey indicated that disclosure about the impact of different estimates that management could have used on the financial statements would be useful to investors. About half of the respondents to our survey indicated that quantitative sensitivity disclosures would be somewhat useful and half indicated that they would not be useful. Respondents indicated that "too much analysis can be confusing," the analysis would be "too complicated to be meaningful" and "too complex," the "potential for misinterpretation" of the data exists, the detailed analysis "creates an aura of ambiguity that is unnecessary," the disclosures would not be "reliable or quantifiably exact" and would be "too difficult to measure" and the "disclosures may only be useful for very significant matters." Other respondents indicated that the sensitivity analysis is still based on assumptions and still requires management judgment. Another respondent indicated that if the possible range is too large, investors might question the credibility of the financial statements. The significant number of possible outcomes may also not be useful for comparative purposes and comparison to competitors may not be possible.

Based on the above, we do not believe quantitative sensitivity disclosures should be mandated in all cases for each critical accounting estimate. Rather, a Company should be allowed to determine whether, based on an analysis of the company's particular facts and circumstances, the significance of the estimate to the financial statements, and other factors, quantitative disclosures to demonstrate sensitivity are useful for all, some, or none of the critical accounting estimates. We do not believe that the Commission should only mandate the two choices proposed or any specific choices for assuming changes relating to the critical accounting estimates to analyze sensitivity. Rather, the registrant should have flexibility to select one of the methods suggested by the Commission or a different method that best portrays its particular facts and circumstances. Of course, the company should disclose an explanation of the method used in the sensitivity analysis.

The proposal also would require specific quantitative and qualitative disclosures in MD&A about material past changes in critical accounting estimates. APB No. 20, Accounting Changes, already requires disclosure in the notes of the effect on net income and earnings per share for a change in estimate that affects several future periods. To the extent that changes in critical estimates materially affect measurements in companies' financial statements and their overall financial performance, we believe that is appropriate for a company to discuss the quantified impact of the change in MD&A, in addition to the required note disclosure. The proposal would require a company to also describe the reasons for the change. We believe disclosure of the reasons for the change could either be provided in MD&A or the notes, but if it is disclosed in the notes, we do not believe it necessarily needs to be repeated in MD&A.

V. Initial Adoption of Accounting Policies

We believe that certain of the proposed disclosures about the initial adoption of accounting policies would be useful to investors. Almost 100% of respondents to our survey believed that the accounting principle that has been adopted and the method of applying that principle would be useful, about 70% believed it would be useful to disclose the events or transactions that give rise to the initial adoption of an accounting policy, and about 85% believed that the quantified impact resulting from the initial adoption of the accounting policy on the company's financial condition and results of operations would be useful. However, we believe these useful disclosures are more appropriately included in the notes to the financial statements instead of MD&A. In addition, we note that APB 22 already requires disclosure in the notes identifying and describing significant accounting policies and the methods of applying those principles, including newly adopted accounting policies if they are considered significant.

Other proposed disclosures, such as an identification of all the acceptable alternatives, a description of why a company made the choice that it did, and a qualitative discussion of the impact that the alternatives would have had on the financial statements, would result in disclosure overload and would not be useful. Only about 15% of respondents to our survey believed that the estimated effect of adopting alternative accounting policies that the company could have adopted but did not adopt would be useful. Discussing every alternative, some of which clearly may not be preferable in a company's particular circumstances, is unnecessarily confusing to investors and may not provide relevant information. Such detailed disclosure would also seem to unnecessarily question and undermine the appropriateness of the alternative that management did choose and could diminish investor confidence. In addition, respondents to our survey indicated that additional costs of up to $2 million could result if disclosure is required about the effect of alternative accounting policies that are currently in place that could have been chosen but were not.

The Commission has also asked whether it should require in MD&A a discussion of whether the accounting policies followed by a company upon initial adoption differ from the accounting policies applied, in similar circumstances, by other companies in its industry, and the reasons for those differences. We do not believe the Commission should require this disclosure. A company may not know the policies applied in similar circumstances by other companies in its industry. A company may not be able to ascertain whether the circumstances are truly similar to its own circumstances based on publicly available information. They may not possess all the facts to make an accurate conclusion. Slightly different facts and circumstances, which may not be known, can lead to a different conclusion. Speculation about other companies "similar" circumstances would be inappropriate.

The Commission has also asked whether auditing firms may be able to assist companies in determining whether their accounting policies generally diverge from industry practice. Although auditing firms may be able to speak generally about industry practice, because of client confidentiality, they could not provide specific details and facts about other clients that may be necessary for a company to determine whether their circumstances are the same as another company or whether slight variations exist which could justifiably led to a different conclusion.

VI. Audit Committee Discussions

Slightly over half of the respondents to our survey currently always discuss critical accounting estimates with the audit committee and the remainder sometimes had these discussions. However, over half of the respondents did not believe disclosure of whether or not senior management had engaged in discussions with the audit committee about critical accounting estimates would be useful to investors. Some respondents indicated that the disclosure would add "marginal value" or "no value" and that "the discussion does not increase shareholder confidence." One respondent believed that audit committee discussions reflected a corporate governance issue and not an accounting policy decision. Another respondent believed that there should be a presumption or a requirement that critical accounting estimates were discussed with the audit committee and disclosure should only be required if the discussion did not occur. This is similar to the Commission's rules for disclosure of an interim review (that is, registrant's are not required to disclose that an interim review has been performed since interim reviews are required, but they must disclose if a review has not been performed.)

VII. Auditor Examination of MD&A Disclosure Relating to Critical Accounting Estimates

If auditor examination of critical accounting policies is to be mandated, then all of MD&A prepared as part of an annual filing should be examined under AT 701 and the auditor's examination report should be filed with the Commission as part of the annual report on Forms 10-K and 10-K/SB. We do not support examining and reporting on only a portion of MD&A. Applying varying levels of assurance to different parts of MD&A in a piecemeal approach (AT 701 examination vs. SAS 8) would be confusing and readers may be led to believe that the auditor had examined a certain portion of MD&A when in fact they had not. Furthermore, requiring examination of only a portion of MD&A would imply that this portion was somehow more important or more relevant to users. All of MD&A is relevant and one portion should not be made to appear more important then any other.

The filing of separate reports on the audit of the financial statements and the examination of MD&A should be permitted; however, to best communicate those reports the Commission should encourage firms to combine these reports. It is also imperative that the auditors' reports not overshadow the responsibility of management for the accuracy of the financial statements and the assertions in MD&A in the eyes of the user. >We also believe that annual MD&A should be examined and not reviewed. It is in the best interest of the public that the highest level of assurance be expressed.

VIII. Other Issues and Questions Considered

Scope: Confidential Information

Some respondents to our survey were concerned about the competitively sensitive nature of proprietary information that may be required to be disclosed (for example, certain insurance rates, and pricing methodologies and assumptions.) Also, disclosure of assumptions about loss contingencies, such as disclosure about legal, environmental, and tax contingencies may be harmful to companies. Therefore, we believe that the scope of any required MD&A disclosure should consider these concerns. Particularly, we believe that specific loss contingences for litigation and other actual or possible claims and assessments accounted for and disclosed in accordance with the requirements of SFAS 5, Accounting for Contingencies, should be excluded from the scope.

Cost and time considerations

The disclosures required under the proposal may require significant time and cost to prepare and the usefulness of the additional disclosures to investors may not justify this additional time and cost. If the SEC accelerates the due dates for Forms 10-K and 10-Q, the additional disclosure required under the proposal could cause even further time constraints. About 55% of the respondents indicated that the estimated time to prepare the additional disclosures on an annual basis was 1-2 days; 25% indicated that the disclosures would take 3 to 4 days; 13% indicated they would take between 5 to 7 days; with the remaining 7% of respondents estimating more than 7 days. The additional cost estimates ranged from less than $100,000 to $1 million for disclosures about critical

accounting estimates and the initial adoption of accounting policies. As discussed above, respondents indicated that additional costs of up to $2 million could result if disclosure is required about the effect of alternative accounting policies that are currently in place that could have been chosen but were not.

Segments

As required under the proposal, we believe it may be useful information to identify the segment(s) affected if a company operates in more than one segment and a critical accounting estimates affects fewer than all of the segments. However, a separate discussion of the critical accounting estimates for each identified segment in addition to the disclosure on a company-wide basis will led to disclosure overload, and the added costs to prepare the information would not appear to justify the benefit or usefulness of the information. As discussed above, about 40% of respondents to our survey indicated that they would have a greater number of critical accounting estimates if the critical accounting estimate was identified by segment, with over 75% having an additional 3 or more.

Quarterly Update

Under the proposal, in Forms 10-Q and 10-QSB, companies would be required to provide an update to the MD&A information related to critical accounting estimates discussed in the company's last filed annual or quarterly report. The Commission has asked whether there are some accounting estimates or material assumptions or methodologies that would normally be considered by companies only on a less frequent basis than quarterly. About three-quarters of the respondents to our survey indicated that they do not have critical accounting estimates that are prepared less frequently than quarterly. However, the other one-quarter of respondents indicated that certain accounting estimates, such as estimates involving year end actuarial studies including certain pension and postretirement estimates, may in fact be prepared on a less frequent basis than quarterly. We believe the Commission should consider omitting these types of estimates from the quarterly updating requirement.

IX. Conclusion

Although certain of the proposed disclosures may be useful to investors, we believe that many of the other proposed disclosures would lead to disclosure overload, may be confusing, and are not useful. We also believe the Commission should consider using a "principles-based" approach in any final rule to avoid unnecessary complexity and confusion.

If you have any questions, please contact John Wolfson at (203) 761-3741 or Christine Davine at (202) 879-4905.

Very truly yours,

/s/ Deloitte & Touche Footnotes

¹ We surveyed 27 registrants (about half have revenues of more than $1 billion and about 75% have a market capitalization greater than $75 million) requesting comments on a series of questions about the proposed rule.