August 8, 2002

Mr. Jonathan G. Katz, Secretary
Securities and Exchange Commission
450 Fifth Street, N.W.
Washington, D.C. 20549

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

Dear Mr. Katz:

The Securities Industry Association ("SIA")1 is pleased to submit this response to the proposed rulemaking of the Securities and Exchange Commission (the "Commission") in Release Nos. 33-8098; 34-45907 (the "Release"). The proposed rulemaking would require disclosure about critical accounting policies in the Management's Discussion and Analysis of Financial Condition and Results of Operations ("MD&A") section of annual reports, quarterly reports, registration statements and proxy and information statements.

I. Overview

SIA commends the Commission's efforts to improve disclosure regarding the subjective nature of material accounting judgments. We agree that investors would benefit from a better understanding of accounting estimates in the context of a company's financial condition and results.

We believe that the Commission's efforts will be most effective if undertaken in a manner that is consistent with other important goals, such as providing clear and concise disclosure through the eyes of management and disclosure that is well integrated with a company's other financial disclosures. We agree that quantitative insights into critical accounting estimates could be an important part of this disclosure framework. We are concerned, however, that a sensitivity analysis may not always be the only, or the best, means of doing so. For financial institutions operating in a mark-to-market environment, the proposed sensitivity analysis presents issues of both interpretation and complexity due to the number of factors involved and the interaction of those factors. The potential for data that can easily be misinterpreted by investors or that overstates the true risks involved would be significant in our view. As described below, we believe that the Commission should encourage companies to provide quantitative disclosure, while leaving to management the judgment as to the best and most meaningful way to do so.

The member institutions of SIA have a long history of constructive dialogue with the Commission and other regulators to improve their public disclosures and internal procedures so as to foster transparency and reduce risk. Member institutions have participated in the Shipley and Fisher reports2 and the work of the Derivatives Policy Group (the "DPG"),3 the Group of Thirty4 and the Counterparty Risk Management Policy Group (the "CRMPG").5 SIA is also currently contributing to the deliberations of the Joint Forum Working Group on Disclosure (the "Joint Forum").6

Given the complexity of some critical accounting estimates, including fair value, we believe that the Commission's goals would be best served by a similarly collaborative approach. The use of fair value is a critical risk discipline and consistent with how firms approach risk management and accounting disclosures. In the area of complex financial instruments, a major contributor to uncertainty is market illiquidity. For this reason, we believe there should be enhanced transparency of disclosure about the illiquid portion of firms' portfolios of financial instruments, as well as about their valuation frameworks and internal control procedures. These procedures, particularly the valuation control procedures, help reduce the risk of a systematic valuation bias across diversified portfolios. We believe that by working together with the Commission we can achieve the goal of furthering accounting and risk disclosures to address these issues and provide more meaningful disclosure to investors and regulators alike.

In addition to modifying the quantitative requirement of the proposal, SIA believes that the Commission should implement the following recommendations to improve the quality and workability of the proposal:

Our recommendations are discussed in the sections that follow.

II. Sensitivity Analysis

A. The Commission should encourage, rather than require, quantitative disclosure, and sensitivity analysis should not be mandated as the exclusive model for this disclosure.

The Commission's proposal would introduce a mandatory quantitative sensitivity analysis for each critical accounting estimate identified in a company's MD&A. SIA believes that the Commission should adopt a more flexible approach. We believe the Commission should explicitly encourage companies to provide quantitative disclosures to the extent appropriate and consistent with management's views about the relevant risks, and we urge the Commission to afford management the flexibility to select and develop different approaches to disclosure for different estimates. We discuss the Commission's proposal in the specific context of fair value accounting in Section II.B below.

Many accounting estimates are complex and are based on multiple interrelated assumptions. In most cases, the potential impacts of changes in these estimates and assumptions are not additive. An investor may well be better informed by a clear, concise qualitative discussion that includes management's views regarding the uncertainty of an estimate and the scope of its potential impact (e.g., material, severe, catastrophic),7 than by a lengthy, complex sensitivity analysis. This is particularly true in the case of estimates or assumptions for which it is difficult to define a unique range of alternative values. The sensitivity analysis would in that case be of limited utility, as the investor would have no meaningful information about the probability of the indicated changes.

In addition, sensitivity analysis is not as easy to apply in practice as is suggested by the examples included in the Release. We describe below interpretive questions that arise in the context of fair value, which we believe illustrate this problem very clearly. Even if the sensitivity analysis could be readily applied to all estimates and assumptions, the Commission's approach may overwhelm investors with numerous potentially confusing quantitative disclosures.8 At a minimum, the sensitivity analysis could give investors a false sense of precision in areas that are highly subjective, as well as a false sense of the comparability of disclosures of different companies, whose application of the sensitivity analysis will necessarily be idiosyncratic. By focusing investors on what may be an arbitrary quantitative analysis, the Commission's approach will distract attention from what in our view is the most important disclosure - that is, a qualitative discussion of the scope, nature and integrity of the procedures that companies have put in place to assure that their estimates and assumptions are reasonable.

The importance of the Commission's proposal is such that the potential burden to issuers should not drive the final rule. Nevertheless, we believe that it is appropriate to consider the relative burdens and benefits of the proposal. In this regard, we believe that the burden of preparing the proposed sensitivity analysis could be significant, particularly where companies would otherwise use alternative risk measures. This burden will be aggravated if the Commission proceeds with its proposal to accelerate the filing deadlines for annual and quarterly reports.9 We are skeptical that the resulting costs will be matched by benefits to investors of comparable magnitude.

B. Fair value accounting in the context of a diversified portfolio is an illustration of the problems presented by the proposed sensitivity analysis.

The Commission requests comment on whether there are critical accounting estimates for which a sensitivity analysis would be inappropriate, not meaningful or otherwise unhelpful to investors. The valuation of diversified portfolios of non-exchange traded financial instruments accounted for at fair value is one good example.

The Commission signaled the need for enhanced disclosure in this area in its Statement about Management's Discussion and Analysis of Financial Condition and Results of Operations ("FR-61"),10 and we believe that many of the qualitative disclosure requirements proposed in the Release would build upon and strengthen the initiative of FR-61. For the reasons described below, however, we believe that it would be a mistake to require a sensitivity analysis in this context.

1. The sensitivity analysis would be difficult to apply, and may not be meaningful, in the context of fair value.

The difficulties in applying a sensitivity analysis in the area of fair value begin with the definition of "critical accounting estimate." We are concerned that the proposed definition does not provide sufficient guidance as to whether, in the case of a diversified portfolio, the "estimate" is the aggregate portfolio, each business segment, each asset category or each instrument. Alternatively, the proposed definition could be interpreted to require each individual assumption to be measured across all affected instruments. Each of these approaches presents significant challenges for any sensitivity analysis, especially where the fair value of a broad range of instruments is derived from models or otherwise derived based on uncertain inputs.11

For these financial instruments, subjective estimates and assumptions relating to the valuation of the portfolio fall into three principal categories: (i) model structure, (ii) market data that is not directly observable or derived from reliable third-party quotation sources due to market illiquidity, and (iii) assumptions regarding the correlations of these inputs.

As a threshold matter, it is unclear how to apply the sensitivity analysis to assumptions in the form of model structure. This element of uncertainty is not, of course, susceptible to variances that could be reflected in a meaningful quantitative way. One could simply vary the numeric result of the operation of a model (i.e., the "estimate" resulting from the assumption). However, nothing in the model provides a basis on which to inform the range within which it would be appropriate to vary the result of the computation for the purpose of generating an alternative range of outcomes. Disclosure in this area would be further complicated by the fact that valuation models, the assumptions underlying those models and the composition of a company's portfolio all often involve highly sensitive and proprietary information.

We noted above that applying a sensitivity analysis could involve the identification and variability of many individual assumptions. This problem would be compounded in the area of fair value. For example, if an input or correlation is varied for purposes of quantifying its impact as a critical accounting estimate for a particular product, it must also be varied for purposes of quantifying the impact of the variability on the balance of the portfolio that is sensitive to the particular market risk factor. The scope and number of analyses that may be appropriate for disclosure are potentially very large, and determining the materiality of the results in light of relevant Commission pronouncements (notably Staff Accounting Bulletin No. 9912) makes the exercise more difficult.

Even where individual estimates or assumptions are material by any standard, the disclosure of sensitivity analyses for each of them could create a misleading impression of the aggregate uncertainty affecting the portfolio's valuation. Precisely because the aggregate impact is not the algebraic sum of the underlying variances, the disclosure of those figures could grossly distort the aggregate amount of variability.

In addition, many financial instruments are hedged or are themselves hedges. A sensitivity analysis that demonstrates independently the impact of changes in assumptions on the hedged and hedging instruments will also generally overstate the true impact of changes in the relevant assumptions on the company's results.13 Meaningful disclosure about valuation uncertainty at the portfolio level would imply a framework that accounts for these considerations, among other relevant factors. We do not believe that satisfactory quantitative tools now exist that could be adapted readily to provide this framework.

2. The Commission should permit companies to develop meaningful fair value disclosures that build on internal valuation and control procedures and supplement existing risk disclosures.

SIA believes that the Commission should permit companies to develop disclosures about fair value that build on the foundation of internal management and valuation control procedures and supplement existing risk disclosures consistent with management's own view of the relevant risks.14

In a mark-to-market environment, a company's income statement incorporates an ex-post measure of its risk profile as period-to-period changes in the mark directly affect earnings. The quality of this profile, and whether it represents in fact a true reflection of risk, is a function both of proper starting valuations and systematic application of independent internal and external checks and balances to assure process integrity. SIA believes that the existing internal control procedures developed in the context of market risk could be a useful starting point for considering the scope of additional disclosures. We recognize that market risk disclosures are not themselves responsive to the Commission's concern in proposing disclosure about critical accounting estimates. Market risk disclosures address future valuation uncertainties, whereas (at least in the context of fair value) the Commission is now seeking information about current valuation uncertainty. Nevertheless, these procedures are instructive in understanding what financial institutions now do to assure the integrity of their risk profile as reflected in their financial statements. Supplementing existing risk disclosures would also promote a more coherent presentation of information about current and future valuation uncertainties, which are inextricably related.15

Current valuation control procedures and related disclosures go well beyond those mandated under the Commission's existing rules. They reflect the work of the Group of Thirty in 1993, the DPG in 1995 and the CRMPG in 1999, among others. Together, these initiatives called for a more uniform framework for risk measurement and management, including a valuation framework for fair value. The framework included:

We believe that these procedures and the related disclosures provide important protections to investors and limit the possibility of a systematic bias in estimates or assumptions underlying valuations across a portfolio. We believe that supplementing these disclosures and procedures would do more to achieve the Commission's stated goal in the Release than the application of a sensitivity analysis. A key focus should be the portion of the portfolio that is most subject to the uncertainty addressed by the Commission's proposal, which we believe to be centered on illiquid instruments. Subject to developing a framework for identifying this portion of the portfolio, companies could provide, for example, disclosures that would result in greater transparency for on-balance sheet cash and derivative positions. Other quantitative tools, such as liquidity value-at-risk analysis, could also be developed or improved over time to address future valuation uncertainty associated with the illiquidity of these asset classes. Above all, more detailed information should be provided about risk policies and procedures, which are, in our view, central to an investor's appreciation of risk in a fair value context.

SIA stands ready to collaborate with the Commission in developing a workable approach to enhanced disclosure. The deliberations of past and current working groups, such as the DPG and the Joint Forum, make it clear that there is no "one size fits all" solution to the complex issues associated with valuation and risk of loss. "Best practices" are constantly evolving - and rightly so - to account for new products and new or improved techniques for analyzing them. A consistent, committed dialogue with regulators and accounting standard setters would allow this important process to continue uninhibited.

If the Commission nonetheless decides to require quantitative disclosures at this time, SIA believes that the Commission should do so through an appropriate phase-in program that would permit a range of permissible quantitative analyses. This would allow companies the time to develop appropriate systems to capture the types of required data and generate the required presentation. An alternative approach might be a pilot program. This approach would permit the Commission a better opportunity to observe and consider, together with registrants, the wide range of issues that will doubtless be raised and thereby develop a more refined and nuanced approach. In any event, if the Commission elects to retain the requirement of a sensitivity analysis, without providing for alternative methodologies, we believe that further discussions with financial institutions should be undertaken to develop the best means for implementing the requirement.

III. Qualitative Disclosures about Critical Accounting Estimates

While SIA does not support at this time a mandatory quantitative disclosure requirement, we do support improved qualitative disclosure. To be meaningful, however, additional disclosures should build on the existing disclosure and oversight regime with respect to a company's financial disclosures. A crucial predicate is that the concepts and definitions applicable to any new disclosures be harmonized with existing pronouncements of the Financial Accounting Standards Board ("FASB") and the AICPA. SIA also believes that some of the Commission's proposed disclosures are also appropriately made in the notes to a company's financial statements. Doing so would not only result in the enhanced disclosure sought by the Commission, but would subject those disclosures to review and certification by the most appropriate gatekeeper, namely the company's independent auditor.

A. The definition of "critical accounting estimate" should be harmonized with the related accounting literature.

The Commission's proposal defines an accounting estimate as "critical" if both of the following are true:

"(A) The accounting estimate requires the registrant to make assumptions about matters that are highly uncertain at the time the accounting estimate is made; and

(B) Different estimates that the registrant reasonably could have used in the current period, or changes in the accounting estimate that are reasonably likely to occur from period to period, would have a material impact on the presentation of the registrant's financial condition, changes in financial condition or results of operations."16

The related accounting literature uses terminology that is similar but not the same as the Commission's proposed language. SOP 94-6 requires disclosure about an accounting estimate in the notes to the financial statements when known information indicates that both of the following criteria have been met:

"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.

b. The effect of the change would be material to the financial statements."17

SIA believes that the existence of differing standards for the disclosures regarding critical accounting estimates will be burdensome to issuers and confusing to investors, and we urge the Commission to take steps to harmonize the standards and terminology in this area. We would urge the Commission to work with the FASB and the AICPA to develop a coordinated and harmonized approach to accounting estimate disclosures.

B. The Commission should require increased disclosure about critical accounting estimates in a company's financial statements.

SIA recommends that the Commission amend Rule 4-08 of Regulation S-X to require that the notes to a company's financial statements include additional qualitative disclosure that identifies critical accounting estimates and material underlying assumptions. While existing accounting literature requires certain disclosures about accounting estimates in the financial statements,18 we believe the Commission should take this opportunity to strengthen those disclosures and harmonize them with new MD&A disclosures.

C. The Commission should not require disclosure regarding accounting estimates that do not involve significant management judgment.

The Commission requests comment on whether it should expand the definition of "critical accounting estimate" to include volatile accounting estimates that require the use of complex methodologies but do not involve significant management judgment.

We believe this would be a mistake. The stated goal of the Commission's proposed rules is to improve investor understanding of material accounting estimates that involve subjective judgments of management. To expand the rules to include complex accounting estimates that do not involve the exercise of management judgment would be to broaden significantly the scope of the proposed disclosures. It would run the risk of overwhelming investors with inherently complex disclosures about accounting principles while substantially increasing the disclosure burden of many reporting companies. This type of disclosure would be particularly burdensome in the context of valuations of financial instruments, where such instruments are routinely valued on the basis of complex models that are generally accepted in the marketplace and do not involve any significant management judgment.

IV. Audit Committee Disclosure

SIA agrees that the oversight role of the audit committee in discussing critical accounting estimates with management and a company's independent auditor is an important financial control. SAS 61 requires that the auditor discuss these estimates with the audit committee and that the audit committee be informed as to the basis for the auditor's conclusions about the reasonableness of sensitive estimates.19 We believe that management of most companies also participates in discussions about accounting estimates with both the audit committee and the independent auditor. The benefits to investors of a recitation of the occurrence of these discussions is at best arguable. Nor would such a requirement, taken by itself, provide meaningful encouragement to an audit committee or management to engage in more focused discussions in this area, or provide guidance as to how to improve the quality of that discussion. For these reasons, we believe that this element of the proposed rules should be deleted.

If the Commission nevertheless determines to require enhanced audit committee disclosure, we believe the disclosure is more appropriately part of the audit committee report included in a company's proxy statement pursuant to Item 306 of Regulation S-K.

SIA does not, moreover, believe that audit committee disclosure should cover procedures used by the committee in its oversight function with respect to critical accounting estimates. The role of the audit committee is not one of independent verification, and will rather be directed toward an understanding of the coherence and scope of the procedures employed by management and the company's auditor.20 These latter procedures are precisely those of greatest importance and that merit, as we have recommended above, the spotlight of disclosure.

SIA also believes that any additional audit committee disclosure should not call for information about unresolved concerns of the committee in the area of critical accounting estimates. The meaning of "unresolved" is unclear in this context. Given their uncertain nature, accounting estimates will a priori lack resolution. A recitation that the audit committee will continue to review a volatile estimate seems of little utility to investors. Similarly, if there is in fact material disagreement among management, the auditor and the audit committee as to one or more estimates, it seems unlikely that the financial statements would be certified by the auditor or recommended to the full board in the first place. And, whatever the meaning of "unresolved" may be, disclosure of the details of an audit committee's review would surely serve to chill discussion among the most crucial gatekeepers of a company's financial disclosures.

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

The Commission solicits comment on whether it should require that MD&A disclosure regarding critical accounting estimates undergo an auditor examination or an auditor review under standards comparable to Codification of Statements on Standards for Attestation Engagements § 701 (Management's Discussion and Analysis).

This question presents complexities of its own, but in terms of process rather than substance. Indeed, in some respects, such a proposal would provide important advantages: the company's key financial gatekeeper - its independent auditor - would assume a measure of responsibility for the appropriateness of the proposed disclosures, particularly the quantitative disclosures; underwriters would have, at least indirectly, the comfort that such a review had been completed. Nevertheless, these procedures are not now commonly undertaken. We believe that both companies and their auditors balk at the expense of an examination or review, particularly since it would not appear that these procedures would result in better disclosure. Auditors may also be unwilling to increase their potential for liability as a result of errors in these disclosures beyond that which they assume under Codification of Statements on Auditing Standards § 550 (Other Information in Documents Containing Audited Financial Statements), under which they must review and consider the accuracy of MD&A disclosure. These practical considerations lead us to conclude that an examination or review by the auditors should not be required at this time. Instead, these concerns illustrate the need for a robust and effective safe harbor such as we propose in Section VIII below.

VI. Quarterly Updates; Past Changes in Estimates

The proposed rules would require the following disclosures in a company's quarterly reports on Form 10-Q:

SIA agrees with the first part of the Commission's proposal with respect to quarterly updates. If a company uses a critical accounting estimate that has not previously been the subject of disclosure, then the company should provide disclosure with regard to that estimate in its Form 10-Q for the first quarter in which the estimate is used.

By contrast, we believe that the scope of the second part of the Commission's proposal is too broad. The rule should focus on material changes in management's judgment or approach with respect to the accounting estimate, but not necessarily all material changes in the accounting estimate itself. Where the methodology for the estimate remains unchanged, but the estimate itself changes as a result of external factors, then quarterly disclosure should not be required. SIA also believes that any quarterly updating requirement should exclude accounting estimates that are not required under generally accepted accounting principles ("GAAP") to be considered on a quarterly basis, such as actuarial assumptions underlying pension and benefits expense.

The Commission also requests comment on the comparative period appropriate for disclosure regarding changes in critical accounting estimates. SIA believes that no comparative discussion should be required. Particularly in the area of fair value, it will be difficult, if not impossible, to disaggregate meaningfully changes in an estimate (e.g., default rate) from changes in a market factor (e.g., interest rate). If the Commission retains a comparative approach, we believe it should limit the period to two years, with the initial year being the year of implementation.

VII. Alternative Accounting Policies

The Commission requests comment on whether, in connection with a company's initial adoption of an accounting policy, it should require in the MD&A a discussion of the impact that alternative accounting policies acceptable under GAAP would have on a company's financial statements, even when the company did not choose to apply the alternative accounting policies. We urge the Commission not to adopt such a requirement. To require quantitative disclosure regarding accounting policies that were not adopted would be to require companies in many cases to keep multiple sets of books. The burden and cost of such a requirement would far outweigh the potential benefits of the proposed disclosure.

We also believe it would be a mistake for the Commission to impose a general requirement to disclose the impact of a company's choice among accounting methods under GAAP that are used in the company's industry. While this approach may have theoretical appeal, as a practical matter it raises important issues. These issues include interpretive issues (e.g., the definition of "industry" for these disclosures), practical issues (e.g., whether companies can obtain reliable information about competitors' approaches and whether such a comparison would be appropriate in light of the relative concentration of risk across companies in a given industry) and liability implications (e.g., how and whether a company could meaningfully verify the approaches used by competitors for purposes of a due diligence investigation).

VIII. Robust Safe Harbor for Forward-Looking Statements

The Commission notes in the Release and in the instructions to the proposed rules that the disclosures on critical accounting policies would require companies to make forward-looking statements.21 The Commission has requested comment on whether there is a need for further guidance in the application of safe harbors for forward-looking information.

SIA believes that the Commission should provide explicit safe harbor protection for any new disclosures on critical accounting estimates and that such protection should extend to all types of issuers and transactions. SIA favors the approach taken by the Commission in the context of market risk disclosures under Item 305 of Regulation S-K whereby (i) required disclosures are explicitly deemed to be "forward-looking statements" for purposes of the safe harbors, (ii) the safe harbors are explicitly extended to all types of issuers and transactions, and (iii) the requirement to provide "meaningful cautionary statements" under the statutory safe harbors with respect to the quantitative disclosures is deemed satisfied if the registrant satisfies all of the disclosure requirements under the rule. A similar approach in the area of critical accounting estimates would help promote thoughtful disclosure.

We believe that this approach is appropriate for disclosures regarding critical accounting estimates because they involve uncertainties that are similar in nature to those underlying forward-looking statements, and in many cases may be forward-looking. Under the proposed rules, "critical accounting estimate" is defined as an estimate that requires "assumptions about matters that are highly uncertain," and the definition includes any such estimate with respect to which changes "reasonably likely to occur from period to period" could have a material impact on the company's financial condition or results. While the estimate may be itself determined as of a particular date, forward-looking considerations will be implicit both in the identification of the estimate and, to some extent, in the assumptions underlying the estimate. As noted in the Release, other provisions of the proposed rules would specifically require a forward-looking discussion of potential period-to-period changes in estimates and underlying assumptions.

We do not believe it is sufficient for the Commission simply to call attention to the terms, condition and scope of existing safe harbor protections. Given the uncertainties inherent in the proposed disclosures, and the sensitive and complex nature of the disclosures for many companies, it would place an undue burden on reporting companies to require them to attempt to identify and qualify with "meaningful cautionary statements" those particular disclosures that contain forward-looking statements. It is not in the interests of reporting companies, investors or the Commission to leave doubts about the extent of potential liability in the area of critical accounting estimate disclosures unaddressed.

* * * * *

In conclusion, SIA supports the Commission's objectives set forth in the Release and believes that the proposal represents an important first step in promoting dialogue in the area of critical accounting estimates. While we believe that mandating specific quantitative methodologies is not desirable at this time, we believe that quantitative information can play an important and necessary role in this area. We look forward to working with the Commission to develop a framework for providing this information, particularly as it concerns fair value accounting.

We appreciate the opportunity to comment on the Release. If you have any questions about our letter or would like additional input from us, please feel free to contact the undersigned, Edward Rosen ((212) 225-2820) or Janet Fisher ((212) 225-2472) of Cleary, Gottlieb, Steen & Hamilton, special counsel to SIA for this matter, or Jerry Quinn ((212) 618-0507) of our SIA staff.

Sincerely yours,

Marc E. Lackritz
President

cc. The Honorable Harvey L. Pitt
The Honorable Cynthia A. Glassman
The Honorable Paul S. Atkins
The Honorable Jerome Goldschmid
The Honorable Roel C. Campos

Alan L. Beller, Director, Division of Corporation Finance
Robert K. Herdman, Chief Accountant
Annette L. Nazareth, Director, Division of Market Regulation


Annex A

Summary of Risk Disclosure Practices of Financial Institutions

The following summary includes disclosure requirements of the Commission and the FASB, as well as disclosure recommendations made in the following reports: Report of the Working Group on Public Disclosure (chaired by Walter V. Shipley) (January 11, 2001); Final Report to Basel Committee on Banking Supervision, Committee on the Global Financial System of the G-10 Central Banks, International Association of Insurance Supervisors and International Organization of Securities Commissions, Multidisciplinary Working Group on Enhanced Disclosure (chaired by Peter Fisher) (April 26, 2001); and Recommendations for Public Disclosure of Trading and Derivatives Activities of Banks and Securities Firms, Basel Committee on Banking Supervision and Technical Committee of the International Organization of Securities Commissions (IOSCO) (October 1999). While we believe this summary reflects a growing consensus on risk disclosures for financial institutions, not all of our member institutions are currently providing all of the disclosures set forth below.

I. Market Risk Disclosures

Item 305 Disclosures Qualitative discussion of primary market risk exposures and how the company manages those market risks.

Qualitative discussion of past and expected future changes in market risk exposures.

Quantitative disclosure, to the extent material, for each market risk exposure category (i.e., interest rate, foreign exchange, commodity price, etc.), using one of three methods of presentation:

  • Tabular presentation of fair value information and contract terms relevant to determining future cash flows, categorized by expected maturity dates;

  • Sensitivity analysis expressing potential loss in future earnings, fair values or cash flows from selected hypothetical changes in market rates and prices; or

  • Value-at-risk ("VAR") disclosures expressing such potential losses over a selected period of time with a selected likelihood of occurrence.

Description of modeling techniques, key assumptions used to measure market risk and limitations of market risk measures.

Summary of market risk information for the previous fiscal year.

Additional Recommended Disclosures Periodic disclosure of intra-period VAR by major risk category (high, low and average for the period).

Aggregate high, low and average trading VAR over the quarter.

Quantification of how well market risk models performed.

Disclose the number of times (days) actual portfolio loss exceeded VAR.

Qualitative discussion of stress testing relating to market risk.

Description of organizational structure central to the risk management and control process for trading and derivatives activities.

Qualitative and quantitative disclosure of information produced by internal risk measurement and management systems on risk exposures and actual performance in managing those exposures.

II. Fair Value/Credit Risk Disclosures

FAS 107 Disclosures Disclosure of (i) the fair value of financial instruments for which it is practicable to estimate fair value and (ii) the methods and significant assumptions used to estimate the fair value of financial instruments.
FAS 119 Disclosures For companies that hold derivatives for trading purposes, disclosure of average and end-of-period fair value of derivative instruments and of net trading gains or losses from derivative instruments, by category of financial instrument.
FAS 133 Disclosures Disclosure of objectives for holding or issuing derivatives and discussion of risk management policies, including a description of the items or transactions for which risks are hedged.

Description of the purpose of derivatives not used for hedging.

Disclosure of effects on earnings of holding derivatives instruments.

Rule 4-08(n) Disclosures Description of accounting policies used for derivative instruments and methods of applying those policies that materially affect financial position, cash flows or results of operations.
Additional Recommended Disclosures Qualitative description of policies for identifying, measuring and managing counterparty credit risk.

Qualitative discussion of methods used to mitigate counterparty credit risk, including netting and collateral agreements.

Quantitative disclosure of concentration of counterparty credit exposure, reflecting the effects of netting, collateral and other credit protection, by:

  • Maturity profile;

  • Credit rating of counterparty;

  • Type of counterparty; and

  • Geographic location of counterparty.

Quantitative disclosure of credit losses.

Qualitative discussion of potential future exposure to counterparty credit risk.

Qualitative discussion of stress testing relating to counterparty credit risk.

Qualitative discussion of credit risk control function, internal controls, risk limits and limit monitoring.

Qualitative discussion of how performance in managing credit risk is assessed.

______________________________
1 SIA brings together the shared interests of nearly 700 securities firms to accomplish common goals. SIA member firms (including investment banks, broker-dealers and mutual fund companies) are active in U.S. and foreign markets, and in all phases of corporate and public finance. The U.S. securities industry manages the accounts of nearly 80 million investors directly and indirectly through corporate, thrift and pension plans. The industry generates $358 billion of revenue and employs approximately 760,000 individuals. More information about SIA is available on our homepage: http://www.sia.com.
2 Report of the Working Group on Public Disclosure (chaired by Walter V. Shipley) (Jan. 11, 2001); Final Report to Basel Committee on Banking Supervision, Committee on the Global Financial System of the G-10 Central Banks, International Association of Insurance Supervisors and International Organization of Securities Commissions, Multidisciplinary Working Group on Enhanced Disclosure (chaired by Peter Fisher) (Apr. 26, 2001).
3 Framework for Voluntary Oversight, Derivatives Policy Group (1995). The DPG included representatives of six leading investment banking firms.
4 Derivatives: Practices and Principles, Global Derivatives Study Group (1993). The Steering Committee of the Group of Thirty included representatives of U.S., non-U.S. and multi-lateral banking institutions, the accounting profession and academia.
5 Improving Counterparty Risk Management Practices, Counterparty Risk Management Policy Group (June 1999).
6 The Joint Forum, which is chaired by the Director of the Commission's Division of Market Regulation, includes a wide range of U.S. and non-U.S. regulatory participants, such as the Federal Reserve Bank of New York, the Comptroller of the Currency and the Bank for International Settlements.
7 See, e.g., Statement of Position No. 94-6 ("SOP 94-6") of the American Institute of Certified Public Accountants (the "AICPA"), which uses these terms to distinguish certain types of risk that are subject to enhanced financial statement disclosure thereunder.
8 In this regard, we are concerned that the Commission may be unaware of the number of critical accounting estimates that companies may have. The Commission suggests that most companies will have three to five critical accounting estimates. This may be true for a company that operates in a single country and in a single segment. However, the estimates for a company with cross-border operations spanning several segments or business lines will surely be a significant multiple of the Commission's suggested range.

Since the number and relative complexity of critical accounting estimates is likely to vary widely, we believe that it would not be in the interests of companies or investors for the Commission to impose a maximum number of critical accounting estimates. Imposing a limit could expose a company to significantly increased risk of litigation if an estimate not selected for enhanced disclosure, even in the exercise of management's good faith judgment, is nevertheless found to be at the heart of later financial difficulties experienced by the company. These problems are exacerbated for companies operating in multiple segments. If the Commission imposes a maximum number of estimates, the requirement should apply on segment-by-segment basis.

9 Commission Release Nos. 33-8089; 34-45741 (Apr. 23, 2002).
10 Commission Release Nos. 33-8056; 34-45321; FR-61 (Jan. 22, 2002).
11 It should be emphasized that the source of an estimate is not necessarily correlated with the estimate's reliability. For certain instruments, third-party quotations can involve significant uncertainty. For other instruments, valuation estimates may be computed by models using reliable market data inputs observed in active markets.
12 Commission Release No. SAB 99 (Aug. 12, 1999).
13 A product-by-product approach would require special measures to ensure that the uncertainty associated with the component legs of a hedged position is offset rather than doubled. The widespread use of cross-hedging and dynamic hedging could significantly complicate this analysis.
14 We note that the Commission adopted this flexible approach with respect to fair value disclosures in FR-61, stating that "registrants should consider the need to furnish information [relating to the fair value of non-exchange traded contracts], quantified to the extent practicable . . . ." (italics added). See also Recommendations for Public Disclosure of Trading and Derivatives Activities of Banks and Securities Firms, Basel Committee on Banking Supervision and Technical Committee of the International Organization of Securities Commissions (IOSCO) (Oct. 1999). This report recommends that companies provide quantitative disclosures, to the extent appropriate and practicable, that are consistent with management's views about the relevant risks and that are closely linked to the framework used by management to measure, monitor and manage risk.
15 We have summarized certain risk disclosure practices of financial institutions in Annex A.
16 Proposed Item 303(c)(2)(ii) of Regulation S-K.
17 SOP 94-6, paragraph 13.
18 See SOP 94-6.
19 Statement of Auditing Standards ("SAS") No. 61, Communication with Audit Committees (Apr. 1988).
20 The Commission requests comment as to whether there should be disclosure as to whether the audit committee has reviewed the MD&A. We believe that many audit committees already review MD&A as a matter of "best practice."
21 The Commission notes that those statements would be entitled to protection under existing statutory and regulatory safe harbors, provided that the relevant conditions are satisfied.