July 19, 2002 Mr. Jonathan G. Katz, Secretary
Securities and Exchange Commission
450 Fifth Street, N.W.
Washington, DC 20549-0609

File No. S7-16-02
Proposed Rule: Disclosure in Management's Discussion and Analysis about the Application of Critical Accounting Policies
Release Nos. 33-8098; 34-45907
International Series Release No. 1258

Dear Mr. Katz:

KPMG LLP appreciates the opportunity to submit this letter in response to the Securities and Exchange Commission's request for comments on its proposed rule for disclosure in Management's Discussion and Analysis (MD&A) about the application of critical accounting policies (the Proposed Rule or proposing release).

KPMG strongly supports the Commission's efforts to restore investor confidence in the U.S. financial reporting system and the capital markets. In particular, KPMG supports the Commission's goal of increasing investors' understanding of how earnings are subject to variability, and efforts to improve and strengthen the quality of information available to the U.S. capital markets. We believe that it is important that investors understand that reported net income is not a precise scientific measure, but rather the outcome of a collection of management's best estimates of an enterprise's assets and liabilities. Initiatives that assist investors in understanding these matters can help the capital markets function more efficiently and better serve investors and registrants.

The difficulty lies in how best to communicate this sensitivity of reported results to management's good-faith judgments. We agree that better disclosure about the critical estimates inherent in the development of the financial statements can help users better understand the risks, sensitivities and limitations underlying the "bottom line." However, to be truly effective, disclosure requirements ultimately adopted must find the appropriate balance between qualitative and quantitative information about the financial statements that will enhance the reader's understanding and predictive power, but not overburden the reader with hypothetical outcomes that may be confusing and undermine the relevance and importance of financial statements prepared in accordance with generally accepted accounting principles (GAAP).

The Proposed Rule would, in essence, require registrants to provide a matrix of reasonably possible results based on quantitative sensitivity disclosures for a selected number of critical accounting estimates. This approach presents financial statement users with a number of different outcomes to analyze and evaluate against reported amounts in accordance with GAAP, and could lead to "pro forma like" analyses being generated by such users. This type of analysis may actually undermine the current accounting model, which relies on careful development by management of its best estimates to be used in preparing the financial statements, using all information available at the time the estimates are made. Investors may be misled to believe that it is just as acceptable to use other estimates in the financial statements as it is to use management's best estimate.

We encourage the Commission to consider a simplified approach to address investors' need for information about the variability of earnings that builds more directly on the existing requirements under GAAP. Specifically, we believe that existing disclosures required by GAAP relative to accounting policies and estimates, including Accounting Principles Board (APB) Opinion No. 22, Disclosure of Accounting Policies, and AICPA Statement of Position 94-6, Disclosure of Certain Significant Risks and Uncertainties, should be enhanced to address certain of the disclosure requirements included in the Proposed Rule. In this regard, we believe the Commission should request a private sector standard setter to undertake a project to enhance or interpret existing GAAP to consider certain disclosures included in the Proposed Rule. We believe that the framework under existing GAAP is adequate, with interpretation and enhancement, to provide the qualitative disclosures addressed in the Proposed Rule and achieve the objectives so stated in the proposing release. As an interim step, pending completion of the project by the private sector standard setter(s), the Commission staff could adopt certain of the proposed disclosures through the issuance of a Staff Accounting Bulletin.

If the Commission proceeds to final rulemaking on Management's Discussion and Analysis and amendments to Regulation S-K, Regulation S-B and Item 5 of Form 20-F, we offer the following recommendations. Most of our recommendations, except where already addressed in GAAP, would apply to the aforementioned proposed project to be requested of private sector standard setter(s), and to any Staff Accounting Bulletin issued as an interim step.

Proposed Disclosure about Critical Accounting Estimates

Definition of a Critical Accounting Estimate

In the proposing release, critical accounting estimates must meet two criteria to qualify for discussion. An estimate is critical if it (a) requires assumptions about matters that are highly uncertain at the time of estimation and (b) different estimates that could reasonably have been used, or reasonably likely changes in those estimates, would have a material impact on the financial statements.

We do not understand why the level of uncertainty associated with critical estimates must be "high." Financial statements require a number of estimates, many of which are not "highly uncertain" at the time of estimation, that subsequently change with dramatic effect on the financial statements. For example, residual values for an auto leasing company often are initially estimable with a relatively high degree of precision (an active market for second hand vehicles exists with an observable market price, and market information is readily available on the internet and through dealerships). Yet, minor fluctuations in the estimates can have a significant impact on the results of operations and the statement of financial position. In fact, this estimate is probably the single most important estimate for a leasing company, yet it would appear not to be a critical accounting estimate under the definition outlined in the proposing release. We believe investors would be better-served if this type of estimate was considered "critical" as well.

We are unclear as to what the Commission intends in the definition of critical accounting estimates by "different estimates that could reasonably have been used" in the financial statements. GAAP requires that the financial statements reflect management's best estimates based on the information available at the time the financial statements are issued. It is unclear whether the Commision is focusing on the availability of different methods for making a particular estimate or if the Commission believes that estimates other than management's best estimate that are based on different assumptions but using the same methodology are acceptable in preparing the financial statements, even though they are not the "best" estimate. Either the Commission should clarify this point in its final rule, using an illustrative example, or delete this point altogether.

We believe that the definition of critical accounting estimates could be simplified and still allow investors and other users of the financial statements to identify those estimates that are likely to be the cause of the largest increases or decreases in reported earnings, financial position or cash flows, primarily because there is uncertainty associated with the assumptions underlying the estimates themselves.

One alternative for the definition would be to link the MD&A discussion more directly to the requirements of AICPA Statement of Position 94-6, Disclosure of Certain Significant Risks and Uncertainties. Paragraph 13 of SOP 94-6 requires disclosures about an estimate when, (a) it is at least reasonably possible that the estimate of the effect on the financial statements of a condition, situation, or set of circumstances that existed at the date of the financial statements will change in the near term due to one or more future confirming events and (b) the effect of the change would be material to the financial statements. For qualifying estimates, SOP 94-6 requires disclosure of the nature of the uncertainty and an indication that it is at least reasonably possible that a change in the estimate will occur in the near term. We believe that the definition of a qualifying estimate under SOP 94-6 would (1) better capture all estimates having a reasonably possible likelihood of resulting in a material impact on a registrant's financial presentation and (2) be consistent with the Commission's intent to identify critical accounting estimates. This definition also would provide a clear link between what is required in the financial statements under existing accounting standards and management's discussion of known trends and uncertainties as required under Item 303 of Regulation S-K. We therefore recommend that the Commission adopt the SOP 94-6 definition in its final rule.

Qualitative Disclosures

We generally support the disclosures proposed by the Commission in the proposing release that identify and discuss matters relevant to critical accounting estimates, including methodologies employed and underlying assumptions, and how the estimates affect the financial statements. However, we suggest that the disclosures relative to methodologies employed and underlying assumptions be included in the final rulemaking as items to consider for disclosure, as opposed to required disclosure for each critical accounting estimate. Registrants should be allowed to exercise judgment in how best to convey the uncertainties surrounding the estimate and the reasonably possible effects on the financial statements if future experience differs from that assumed. The Commission also may want to consider encouraging companies to disclose what steps management has taken, or plans to take, to mitigate the uncertainties underlying certain estimates. For example, in the case of inventory, an inventory shrinkage estimate may be less sensitive if a company conducts frequent physical inventories. In this way, investors may better understand why companies within the same industry may not view seemingly similar estimates as critical.

Quantitative Disclosures

We agree that disclosures about past changes in critical accounting estimates during all periods presented is useful information and should be carried forward to the Commission's final rule. However, with respect to transition, we strongly encourage the Commission to make the disclosure requirement prospective, in recognition of the fact that the information necessary to provide full disclosure may not have been prepared or retained. For example, if the final rulemaking is effective for a registrant's 2002 Form 10-K, the disclosure regarding past changes in critical accounting estimates in MD&A would commence with the 2003 Form 10-K covering changes occurring subsequent to the balance sheet included in the 2002 Form 10-K. We also recommend that the Commission clarify in the final rule what disclosure would be expected in situations where, due to changes in facts and circumstances, an estimate that was considered critical in an earlier year is no longer considered critical or where an estimate that previously was not considered critical is identified as critical in the current year. The Commission could specify that disclosure is required for all years presented for any estimate that is identified as critical in any period presented.

The Commission also proposes to require sensitivity analysis and permits a registrant to discuss the financial statement impact of changes in an estimate using either (a) reasonably possible, near-term changes in the most material assumptions, or (b) the ends of a range of reasonably possible amounts determined when the estimate was formulated.

We believe that the Commission's objectives to enhance investors' understanding of how estimates are used in the financial statements and the sensitivity of those estimates are met by limiting the required disclosures to the qualitative disclosures as currently proposed by the Commission. Specific identification of key assumptions affecting each critical accounting estimate, and tailored explanations of how changes in those assumptions and other relevant facts and circumstances could impact the amounts included in the financial statements, would provide information useful in analyzing the financial statements presented. This information would also empower investors with the ability to predict how significant events that may occur in the future may affect a company's financial statements. The Commission has proposed the timely reporting on Form 8-K of many significant events under a separate rulemaking project. In combination, this information will allow investors to continuously revise their own assessments of the likely impacts on a company's financial statements. However, for the reasons discussed below, we do not believe that quantifying the effects of using different estimates would produce information that an investor could put to practical use, since any amounts disclosed would be subject to the same or greater risks and uncertainties underlying the amounts included in the audited financial statements.

We believe the quantitative disclosures proposed by the Commission to demonstrate the degree of sensitivity of critical accounting estimates will be substantially inoperable in practice and will result in investors formulating a matrix of results that may be confusing and potentially misleading. A matrix of reasonably possible results that focuses on the effects of a selected number of critical accounting estimates, is somewhat similar to pro forma presentations that focus the reader on changing certain lines of the income statement based on a few assumptions. It does not address the multitude of other smaller assumptions, estimates and accounting policies that make up the financial statements, and, in the aggregate, materially impact the presentation of financial position and results of operations. The amounts that fall at the low and high end of the matrix are ultimately less accurate than the amounts chosen as optimal and recorded in the financial statements by management.

It is unclear to us how the effect on the financial statements of a change in estimate should be calculated. Are both the direct and indirect impacts of changes included? Direct impacts, such as a change in a key assumption used in the valuation of inventory, derivatives or pension obligations are relatively easily identified. However, indirect impacts could result from direct ones. For example, increases in a retailer's inventory obsolescence reserve may indicate decreasing future revenues, which could in turn result in impairment of the long-lived assets that generate lower than originally expected cash flows. Would the resulting impairment of the long-lived assets also be included in the result of a change in estimate? How many iterations of "knock-on" effects would a company be required to consider?

We agree that investors need more information about the potential variability of earnings and an understanding that, had management made different assumptions, net earnings might be different than what is reported in the audited financial statements. However, quantifying those impacts on the financial statements may be excessively burdensome, unwieldy, and produce amounts that themselves will rely upon assumptions and estimates. To illustrate, consider Example 3 in the proposing release. In describing the financial statement impact of a 10% reduction in the company's estimate of future cash flows on the recoverability of equipment balances, the disclosure indicates that "...an impairment loss of approximately $30 million, equal to the difference between the fair value of the equipment (which we would have determined by calculating the discounted value of the estimated future cash flows) and the reported amount of the hard drive-related PP&E." In this example, the company must perform a hypothetical detailed calculation of impairment. To calculate the possible impairment loss, the company must select the method of determining the fair value of the assets and make certain assumptions, e.g., discount rate, in determining the fair value. The quantitative disclosure would require management to make those selections without the benefit of the presumably more thorough research and analysis of available methodologies that would be appropriate for a "real" impairment calculation. The impairment loss could be materially different than the hypothetical $30 million disclosed in the example if assumptions used in the calculation were altered. This estimation and disclosure process could be inordinately time-consuming, as it would require management to, in effect, account for the same event or transaction twice - once for the financial statements and again for the sensitivity disclosure. In a limited survey that we conducted of senior management, board of director members and audit committee members of our client base (a total of 40 responses), 70% (28 out of 40) of our survey respondents indicated that the proposed disclosure would significantly increase the cost of financial reporting.

Sensitivity analysis using the ends of a range permitted by option (b) will produce disclosures that by definition must be less accurate than the number recorded in the financial statements, since GAAP requires that the amount recorded in the financial statements be management's best estimate. The amounts that form the upper and lower limits of the "reasonably possible" range are not adjusted for their probabilities of occurrence. While additional narrative discussion could mitigate this shortcoming, it would also increase the already voluminous disclosure proposed by the Commission and still may not deter the reader from only focusing on the amounts themselves.

We also note that the option of disclosing the ends of a range is very similar to the reporting requirements under Statement of Financial Accounting Standards No. 5 (SFAS 5), Accounting for Contingencies, and it appears that the proposed MD&A disclosure would encompass loss contingencies such as environmental, legal and tax matters. In such cases, the proposed MD&A disclosure would conflict with the disclosure limitations explicitly set forth in GAAP. Many registrants will have significant (and well-founded) concerns related to the disclosure of sensitive information for these types of loss accrual accounts. We have previously indicated our concerns regarding disclosure of sensitive information in response to similar proposed disclosures included in Commission Release Nos. 33-7793/34-42354, Supplementary Financial Information. Detail disclosure of changes in loss accrual positions may have significant negative implications for registrants and their investors. Specific information related to loss accruals for pending litigation, contingent income and franchise taxes, loss contracts and environmental issues may in fact provide a "road map" for third parties to exploit information that would otherwise have remained confidential. Existing GAAP and MD&A disclosure requirements have provided an effective means to communicate relevant information regarding these estimates to investors for many years, and we encourage the Commission to exclude these types of estimates from the final rule.

Lastly, we believe that some estimates do not lend themselves to the sensitivity analysis disclosures described in the proposal. For example, for long-term contracts accounted for under the percentage-of-completion method, an increase in estimated costs to complete may cause a decrease in profit recognized on some long-term contracts, but may trigger accrual of a full loss on another long-term contract with seemingly similar characteristics. Sensitivity analysis for transactions arising from individually unique contractual arrangements would require detailed calculations of hypothetical impacts on a contract-by-contract basis that would be extremely time-consuming to prepare and would produce limited meaningful information.

Similarly, this type of redundancy would be a problem for many financial services companies. Life insurance companies, for example, use a large number of estimates in estimating required reserves, changes in any of which would have a material impact on the financial statements. Some estimates, such as mortality rates, change over time as average life spans increase. Others, such as persistency or renewal rates, expected investment yields, and the age of the policy holder pool, can change more rapidly due to external factors. Insurance companies would, therefore, be faced with a range of assumptions to change, all of them key assumptions. The resulting permutations and combinations would entail voluminous disclosure, require extensive preparation and produce myriad information that may be more confusing than enlightening. Property and casualty insurers would be faced with their own difficulties. Many such companies believe it is impossible to estimate a range of outcomes for certain casualty events. For example, environmental exposures could result in essentially limitless losses under existing environmental legislation. Consequently, no upper end to a range may exist. Financial institutions offering derivative products face similar problems, in that a portfolio of derivative contracts may have values that react differently to changes in the same assumptions. An increase in a key index may cause a decrease in profit recognized on some contracts, but may trigger losses on another contract with seemingly similar characteristics. For many such entities, the proposing release would essentially require registrants to maintain multiple sets of accounting records.

Application to Segments

Item 303(a) of Regulation S-K currently requires companies to include a discussion of segment information where, in management's view, the discussion would be appropriate to an understanding of the company. Financial Reporting Release (FRR) No. 36 goes on to say that segment discussion should be included "to the extent any segment contributes in a materially disproportionate way to [revenues, profitability and cash needs], or where discussion on a consolidated basis would present an incomplete and misleading picture of the enterprise...".

The Proposed Rule would require a company to determine if a separate discussion of critical accounting estimates at the segment level is required, in addition to a company-wide discussion. Analysis and presentation of individual segment-level estimates would potentially increase the number of critical accounting estimates and the volume of disclosure, and is inconsistent with the Commission's current guidance that demonstrates a view toward limiting segment disclosure to only that which is critical to an understanding of the company as a whole. Identification and disclosure of critical accounting estimates at the segment level also is inconsistent with the provisions of SOP 94-6, which focuses on estimates as they affect the financial statements as a whole. Estimates critical to an understanding of the company as a whole would likely be covered by company-wide disclosure.

We recommend that the final rule clarify that segment-level disclosure be limited to those circumstances contemplated by Item 303 of Regulation S-K, that is, when any segment contributes in a materially disproportionate way to the revenues, profitability and cash needs of a company.

Senior Management's Discussions with the Audit Committee

We strongly support proposals that engage the audit committee more rigorously in a registrant's financial reporting and compliance process, and support the proposal to require a registrant to disclose in MD&A whether or not senior management has discussed critical accounting estimates with the audit committee.

We also support the Commission's proposal not to require disclosure of the substance of the discussions for the reason outlined in the proposing release. Of our survey respondents, 80% (32 out of 40) agree that this disclosure should not be required.

Auditor Examination or Review of MD&A Disclosure Relating to Critical Accounting Estimates

The proposing release requests comments on whether the proposed critical accounting policies to be included in the MD&A section should be subject to independent auditor examination. We would not support examining and reporting on only a portion of MD&A. By attempting to carve out a separate section of MD&A and subjecting this separate section to examination, but not MD&A in its entirety, would imply that this separate section is more important than another section. We believe a "piecemeal" approach to providing assurance on only limited portions of the MD&A section would also inappropriately elevate one component of MD&A over other equally or more important components. It also would be difficult for readers of the MD&A section to understand different levels of auditor involvement with the MD&A information when some information is subjected to the auditor's procedures under SAS No. 8, Other Information in Documents Containing Audited Financial Statements, and if other selected information, such as critical accounting policies, is subjected to examination-level assurance. This may actually mislead a reader into thinking that an auditor had examined a certain section when, in fact, it had not.

As noted above, by including certain critical accounting policy disclosures in the footnotes, appropriate independent assurance on such disclosures could be achieved in the context of the overall financial statements because the disclosures would be within the scope of the financial statement audit.

If the final rule requires auditor attestation of MD&A, we strongly recommend that this provision be made effective for annual reports filed a year after initial adoption of the final rule. We believe it is not practical to make this portion of any final rulemaking effective for 2002, for a number of reasons.

  • AT 701, Management's Discussion and Analysis, as it currently exists, will need to be amended in order to address the requirements of the Proposed Rule.

  • Our internal research indicates that attestations under AT 701 have not been widely implemented to date, and the profession will need some time to develop suitable examination programs that are tailored to the specific requirements of the final rule.

  • Auditing firms will need time to train personnel on these types of engagements to ensure effective coverage and appropriate procedures.

  • Registrants will need to be educated on the documentation and support required by auditors to attest to the disclosures in MD&A, and will need time to plan for this additional engagement.

  • An unprecedented number of registrants recently have changed, or anticipate changing by the end of the year, their independent accountants. A change in independent accountant requires a significant time commitment from both the registrant and the successor auditor, and with the unparalleled number of changes, the national accounting firms that audit the preponderance of registrants are integrating a large number of new personnel, in addition to the new clients.

    In short, the implementation of the Proposed Rule and other recent rulemakings, such as accelerated filing deadlines, on top of the other changes occurring in the current financial reporting environment, will place additional significant demands on the existing resources devoted to the financial reporting process.

    Delayed adoption of attestation would allow management, legal advisors and independent accountants to become comfortable with the disclosure requirements without the added burden of attestation. This approach would facilitate an orderly transition to revised business models for both registrants and auditing firms, and help to control the increased implementation costs borne by investors.

    Quarterly Updates

    We support the proposal to require a quarterly MD&A update on critical accounting estimates in Forms 10-Q and 10-QSB. The Proposed Rule would not explicitly require qualitative and quantitative discussion in quarterly reports concerning past material changes in critical accounting estimates. Item 303(b) of Regulation S-K requires interim period discussion sufficient to enable the reader to assess material changes in financial condition and results of operations. To the extent past changes in estimates have had a material effect on the quarterly financial presentation, we believe that discussion of those changes should be provided in quarterly reports on Forms 10-Q and 10-QSB. Consistent with our views on annual discussion of past material changes, this provision should be applied prospectively, beginning with the first quarterly report due after the effective date of the final rule.

    Proposed Disclosures about Initial Adoption of Accounting Policies

    General Comments

    The Proposed Rule would require a detailed discussion of initial adoption of accounting policies, including the events giving rise to adoption, the principle adopted and method of application, and the impact, qualitatively, of such adoption on the financial statements. Further, the Proposed Rule goes on to require that where alternatives exists, a company must identify and explain those alternatives, justify its policy selection and provide a qualitative discussion of the impact on the company's financial statement that the alternatives would have had. Similar requirements exist where there is no existing literature governing the transaction.

    While, as the Commission notes in the proposing release, APB No. 22, Disclosure of Accounting Policies, requires the disclosure of a newly adopted accounting policy if it is considered to be a "significant accounting policy", the current requirements of APB No. 22 do not entail all of the qualitative disclosures considered in the proposing release.

    We believe that robust disclosure of initial accounting policy selection is useful to users of financial statements. However, we believe that certain of the disclosures required by the proposing release relative to the adoption of accounting policies would more naturally be included in the footnotes to the financial statements. We suggest that the Commission staff adopt certain of the disclosures in the proposing release through the issuance of a Staff Accounting Bulletin, as an interim step. Ultimately, the private sector standard setter, the Financial Accounting Standards Board, should be requested to add an agenda item to enhance or interpret existing generally accepted accounting principles and consider the disclosures included in the Staff Accounting Bulletin.

    We also note that as currently drafted, the proposing release does not seem to require discussion of initial adoption of a new policy until a Form 10-K, registration statement or proxy statement is filed. As a result, the disclosure on initial adoption appears to contemplate an annual requirement only, with no disclosure for a policy initially adopted during an interim period. We believe that to maximize effectiveness and utility to the user, the disclosure requirements also should apply to quarterly filings.

    Specific Comments

    While we support the overall objective of disclosure relative to initial adoption of accounting policies, we do not support other features of the proposing release.

    First, we believe that detailed and quantified disclosure of the impact of alternative accounting treatments that a company could have adopted for a transaction could result in confusion for the reader. Discussion of the impact on the financial statements of the multiple alternative policies and methods of application available, (i.e., in methods of inventory valuation or depreciation of fixed assets), could potentially obfuscate the alternative that management believes to be the most appropriate under the circumstances. For example, a number of methods of inventory valuation and depreciation exist. Disclosure of every possible alternative, or even several alternatives, would result in a lengthy narrative, with a limited informational or predictive value to the reader. It should be sufficient to note that other alternatives exist and to indicate, if applicable, that results could be materially different under the alternative(s) not selected.

    A second concern we have about this feature is the workload required to quantify the alternatives not selected and the relative cost/benefit of doing so. Most companies devote significant accounting and reporting function resources to preparing their financial statements using the policies management believes most appropriate. Adding an additional requirement to calculate the effect of accounting policies and methods not considered appropriate could, in many cases, strain resources further and distract management's attention from its primary reporting duties, and is inconsistent with the Commission's proposal to accelerate reporting deadlines.

    Third, while comparative analysis of policies selected against those applied by companies in the same industry could, theoretically, be useful and illuminating to the reader, practical application of the requirement is fraught with difficulties. Many companies are aware of the policy choices made by their competitors, but they may not have sufficient detail regarding competitor transactions to determine that transactions are similar enough to warrant the same accounting policy and/or methods of application, thereby making any comparison speculative. Similarly, while companies may be aware of policy choices made by their competitors, there may be other situations in which a registrant is not aware of the method of application of that policy, making comparison difficult, speculative and potentially ineffective.

    Fourth, seeking non-public data on industry or competitor accounting choices and methods from auditors is inconsistent with the role of an auditor. Auditors are precluded by standards of confidentiality from conveying confidential information to clients about other clients' application of policies. Looking to auditors or other financial advisors to determine whether a company's accounting policies generally diverge from industry practice seems inconsistent with the premise that MD&A, accounting policy selection and the resulting financial statements are management's responsibility.

    Finally, the disclosure requirements may put earlier-stage companies at a disadvantage relative to more established ones. Developing companies will be required to make detailed disclosures on the adoption of a new policy; an event that will likely be a more frequent occurrence than for a more established company that has already made its policy selections.

    Application to Foreign Private Issuers

    Our key concern in the application of the proposed rule to foreign private issuers is that of investor confusion if the proposed disclosure is provided for both primary financial statement GAAP and U.S. GAAP. The current requirement to reconcile the primary financial statements prepared under home country or international accounting standards (IAS) to U.S. GAAP is unique to reporting in the U.S. and is directed towards U.S. investors. However, the Commission's reporting requirements still center on the home country GAAP/IAS financial standards as the primary financial statements. Accordingly, Instruction 2 to Item 5 of Form 20-F requires the MD&A equivalent for foreign private issuers to focus on the primary financial statements, with a discussion of any aspects of the difference between foreign and U.S. GAAP not discussed in the reconciliation, that a registrant believes is necessary for an understanding of the financial statements as a whole. Staff Accounting Bulletin No. 88 (SAB 88) enumerates certain U.S. GAAP matters, including material changes in estimates and significant accounting policies and measurement assumptions not disclosed in the financial statements, that may warrant discussion in MD&A (if the registrant believes they are necessary for an understanding of the financial statements as a whole).

    We support applying any new MD&A disclosure requirements to foreign private issuers in a manner consistent with domestic registrants, and the existing instructions to Item 5 of Form 20-F and SAB 88. Consistent with that view, we believe that the disclosures should be required only for the primary financial statements. Accordingly, as currently required by Instruction 2 to Item 5 of Form 20-F and SAB 88, registrants should be required only to provide the additional disclosures relative to critical accounting estimates and the initial adoption of accounting policies specific to the application of U.S. GAAP if the registrant believes some or all of these disclosures are necessary for an understanding of the financial statements taken as a whole.

    For foreign private issuers that report under Item 18 of Form 20-F, the full disclosures required by U.S. GAAP would require SOP 94-6 disclosures as they relate to the primary financial statements. If the Commission were to adopt a definition of critical accounting estimate that coincides with SOP 94-6, the same link between MD&A and the financial statements that would exist for domestic registrants would result for foreign private issuers that prepare financial statements that comply with Item 18.

    With respect to the proposed disclosure of initial adoption of accounting policies, many European companies will be required to report under IAS and International Financial Reporting Standards beginning in 2005. If the Commission adopts a final rule requiring the MD&A disclosure requirements for foreign private issuers as it relates to the primary financial statements filed with the Commission, the rule should clarify how the requirements will apply for first-time adoption of international standards. We believe it would be an overwhelming amount of disclosure if each foreign private issuer was required to identify and explain every accounting policy adopted as of January 1, 2005.

    Application to Small Business Issuers

    We support the Commission's plan to exclude small business issuers disclosing their business plans instead of MD&A from the new requirements in the proposing release.

    Other Considerations

    In the proposing release, the Commission poses many questions to elicit comment on specific proposed disclosure requirements, many of which we have addressed elsewhere in this letter. We offer the following additional comments in response to certain of those questions:

  • While we do not disagree with the Commission's proposal to require the disclosures required under the Proposed Rule in a separate section of MD&A (subject to our comments included herein), we believe that a company should not be limited to that style of presentation. As stated in FRR No. 36, the Commission expects MD&A to provide a short and long-term analysis of the business of the company "through the eyes of management." In that spirit, we believe it appropriate to allow management to exercise its own judgment and present the required information in such a way that it believes is useful to investors and most indicative of how and why management views these matters as critical to an understanding the company's financial statements.

  • Throughout the proposing release, the Commission seeks comment on the need for quantitative thresholds, minimum percentage impacts, specified percentage changes, and standardized formats as a means to clarify the requirements, standardize disclosure and promote disclosure comparability among companies. We do not support mandating such measures as they are contrary to the presumption that management is best equipped to determine what is material to investors and what information is critical to an understanding of the financial statements. Use of specified benchmarks could produce impacts that, in management's view, are not reasonably likely to occur.

    ************************

    We appreciate the opportunity to comment on the Proposed Rule. If you have any questions about our comments please contact Sam Ranzilla at (212) 909-5837 or Melanie Dolan at (202) 533-4934.

    Very truly yours,

    /s/ KPMG LLP