TELEPHONE: 404-885-3000
FACSIMILE: 404-885-3900

December 6, 2002


Jonathon G. Katz, Secretary
Securities and Exchange Commission
450 Fifth Street, N.W.
Washington, D.C. 20549-0609

    RE: File No. S7-42-02, Disclosure in MD&A

Dear Mr. Katz:

We appreciate the opportunity to comment with respect to Release Nos. 33-8144; 34-46767; File No. S7-42-02. Our comments are limited to proposed Items 303(a)(5) and 303(c) of Regulation S-K.

We agree with the SEC that disclosure by public companies in their MD&A's needs to improve, especially with respect to contractual commitments and contingent liabilities. Accounting standards often do not require disclosure of these items in financial statements or the accompanying notes, in part because the accounting profession has not developed a methodology to produce a "standardized" or "normalized" value for them that accurately or consistently measures their likelihood of occurrence or financial impact. Because contractual commitments and contingent liabilities have the potential to quickly destroy the liquidity of even very large companies, better disclosure is needed. However, we are concerned that proposed changes would not provide the needed information in a usable format yet would significantly increase the cost to reporting companies of providing disclosure.

Contractual Obligations Table

Proposed Item 303(a)(5) would require reporting companies to disclose contractual obligations in a tabular format with prescribed payment periods as column headings and reporting company-determined types of obligation as row headings. This approach presumes that obligations can be (1) pigeon-holed into prescribed time frames, (2) readily quantified and (3) grouped into categories in a meaningful fashion. This simply is not the case.

A helpful way to demonstrate the problems reporting companies would face applying this approach is by example. If the initial term of a contract is for five years (and there are two years remaining on the five year term), but

  • each party has the right to cancel the contract, with or without cause, on 30-days notice, and

  • either party has the right to renew the contract for two years at the end of the initial contract period and the first renewal period,

into which time categories do payments under this contract fall? Should payments be assumed to stop under the contract at the end of the initial term so that payments are reported only under the "less than one year" column and the "one to three year" column? Should the reporting company interpret the renewal provisions as requiring disclosure of payments in all of the columns? Should this conclusion depend upon the reporting company's assessment of the likelihood (or necessity) of renewal, even though it may be premature from a business perspective to explore renewal? Should the reporting company interpret the 30 day cancellation provisions as permitting disclosure of payments only under the "less than one year" column? Do the answers change if: (1) the cancellation rights are held only by the reporting company, (2) the cancellation rights must be triggered by a breach of contract which is not timely cured or (3) the automatic renewal provision only applies after the expiration of the initial term but not after a renewal term?

The possible permutations of the facts presented in the example above are as endless as the reasonable conclusions reporting companies could draw. If reporting companies footnote the table in the level of detail required to make disclosure not misleading, the resulting table would become difficult to produce, prone to error and confusing to investors. The example given above also does not account for such real world situations as companies conducting business: (1) on no written contract, (2) on an expired contract that parties continue to honor, (3) on a contract which is not fully executed or (4) on an umbrella agreement with amendable schedules which may or may not have been amended to conform to current practice.

Even after a reporting company determines the type and time categories of its contracts, it still faces challenges associated with determining the obligation amounts which should be presented in the table. For example, a contract may be a "requirements contract" in which the amount of product or service purchased varies. A contract also may contain provisions which periodically adjust the price per unit according to some market indicator and/or total amount purchased. The problems presented with variable production and price provisions in contracts are illustrative of an overall problem with the contractual obligations table: for most reporting companies only a relatively few contracts -- usually finance or capital structure related -- are readily quantifiable.

Just as the time categories required in proposed Item 303(a)(5) do not reflect how reporting companies and investors synthesize, plan or analyze expenses, the types of categories suggested in proposed Item 303(a)(5) do not reflect actual practice. Although the types of categories are suggested and not required, the implication is that commitments can be divided into a few simple categories. But the categories suggested by the SEC show just how hard categorization actually is. The first three categories -- "long term debt," "capitalized lease" and "operating leases" -- are reasonably straight forward, provide information that already is tracked by most companies and correlate with contracts that typically (at least in the first two cases) are filed as exhibits.

But what about "other long term obligations?" For many reporting companies this one category reflects many of the variables that are the most important to investors, yet the SEC in its example has bundled them together. What might fall into this category? Union contracts? Employment agreements? Franchise payments? Capital expenditures? Futures obligations? Warranty obligations? Severance obligations? Phone and other utility contracts? Computer maintenance contracts? Health insurance obligations? Pension plan and other retirement obligations? There are hundreds of other broad categories of expenditures that companies commit to in advance. It is not at all clear how a reporting company would or could meaningfully divide these up for presentation in the required table. But the diversity among them would require their separate identification in order for the disclosure to be meaningful or comparable.

Once the commitments are divided by type, do they need to be divided by segment or line of business? Where contracts apply to more than one segment should they be allocated among segments? Should contracts that reflect variable costs rather than fixed cost be delineated? What about contracts that a company expects to enter into -- in fact must enter into in order to continue operating -- but has not done so?

Proposed Item 303(a)(5) also requires that reporting companies footnote the contractual obligations table to describe provisions that create, increase or accelerate obligations or other pertinent provisions. Because categories -- particularly "other long-term obligations" -- could encompass hundreds, if not thousands, of contracts, many if not most of which will have substantially different terms, footnote disclosure is likely to be extremely detailed and confusing. Further, in the case of indebtedness, this provision, if read literally, would require the footnote to contain an elaborate disclosure of the financial and other covenants. Lastly, provisions which create or increase obligations are, by definition, contingent liabilities which are covered by proposed Item 303(a)(5)(ii). On its face, Item 303(a)(5) actually asks for contingent liability information three times, which is more than slightly confusing.

Contingent Liability Disclosure

Proposed Item 303(a)(5) would require reporting companies to disclose, either in tabular format or in text, their contingent liabilities that expire in prescribed time frames. These contingent liabilities also would have to be categorized by type in a "manner that is suitable for the registrant's business." Proposed Item 303(a)(5) would permit the "amount" of contingent liability to be presented as one of three values: (1) expected amount of contingent liability, (2) range of liability amount or (3) maximum liability amount. This disclosure suffers from each of the issues detailed above. In addition, disclosures regarding what probably are the most important factors with respect to contingent liabilities -- the likelihood of occurrence and the ability to cover -- are not even required. Further, on its face, proposed Item 303(a)(5) would require disclosure of claims and litigation.1 This would be inconsistent with long-standing claims and litigation disclosure practice and could require disclosure of confidential information such as litigation cost estimates or reserves which could undermine the reporting company's defense posture or result in waiver of the attorney-client privilege.

Disclosure would be more useful if it produced standardized values which would provide a basis for meaningful comparison among companies such as the type of disclosure required of market risk pursuant to Item 305 of Regulation S-K. The SEC has unnecessarily linked proposed Item 303(a)(5) with the congressionally mandated disclosure of off-balance sheet obligations. Although we agree that disclosure of contingent liabilities could potentially be of significant value to investors, the currently proposed rule would not produce information helpful to investors but would significantly increase the cost of disclosure to reporting companies.


Under almost any manner of categorization, it is the aggregate impact of a collection of individually immaterial contracts that yields the disclosures that the SEC appears to be seeking. With respect to material contracts, they already are required to be filed as exhibits and frequently are summarized. We are aware of almost no unregulated reporting companies (i.e., other than banks, utilities, insurance companies and the like) that comprehensively track immaterial contracts. In fact, we know of only a few reporting companies that have in place systems that would enable them to even start to credibly comply with the disclosure requirements of proposed Item 303(a)(5). Current internal controls systems are not required to track immaterial contracts. Hence, compliance with the disclosure requirements will require the implementation of tracking systems and will take time. In addition, this system will not only have to track contracts, it also will have to provide a mechanism for allocating payments to future periods and identifying material changes from period to period. (In this regard it is not at all clear how materiality would be measured.)

We believe that the costs estimates contained in the proposal are ridiculously low. Our best estimate is that for a diversified manufacturing company with sales of approximately $2 billion (i.e., at the lower end of the S&P 1000), tracking the necessary information would require initial expenditures for the development of software of between $75,000 and $125,000 and annual expenditures of between $90,000 and $120,000. This reflects the cost of one paralegal and one financial analyst. This does not include the cost of translating foreign contracts into English or reviewing the analysis, either internally or externally, in order to prepare the MD&A. On average, we would expect these functions to add an additional $25,000 to the cost of compliance.

Safe Harbor

Proposed Item 303(c) appears to extend the coverage of the safe harbor provided by the Securities Reform Act of 1995 to the information disclosed pursuant to proposed Rule 303(a)(4). It is unclear to us why proposed Item 303(c) does not cover the disclosures required by proposed Item 303(a)(5) as well. By their very nature, the projections with respect to payments due under contractual commitments or contingent obligations are forward looking. While they should represent the company's good faith estimate of the likely payments, they are no less susceptible to impact by intervening events and warrant the benefits of the safe harbor. Moreover, the safe harbor is not available in all instances, and the nature of the disclosures contemplated by Item 303(a)(5) are such that they may warrant even broader protection than what the Reform Act provides. We appreciate that the inclusion of Item 303(c) may have been intended to counter any implication that the safe harbor was not available for disclosure of off balance sheet liabilities because, arguably, the off balance sheet entities are not part of the company making the disclosure, but an unintended consequence of proposed Item 303(c) is the negative implication that the safe harbor is not available for the remaining parts of Item 303. We believe that this needs to be corrected.


What may be the most troubling aspect of the proposed changes to Item 303 is that it is not at all clear how this additional disclosure will assist investors. No two companies are alike, and the presentation of dissimilar information regarding dissimilar companies is not going to provide investors (or even professional analysts) information that they can effectively analyze. Further, even analyzing a single company in isolation will be difficult. Contractual commitments and contingencies are but a subset of the financial obligations that a company has, and the enormous gap between the information proposed to be provided by Item 303 and the total picture of a company's financial circumstances is so broad that the information that is to be provided will be of minimum use. Further, it is noteworthy that the trade association that represents professional analysts, when surveyed, did not view a lack of disclosure of contractual commitments as a major disclosure shortcoming. Instead, it cited the lack of disclosure regarding special charges as the item that most needed additional disclosure.


A table which requires that reporting companies try to neatly pigeon-hole contractual obligations into arbitrary time periods and type categories would be difficult for reporting companies to produce and of little use to investors. Improved contingent liability disclosure is definitely needed. But contingent liability disclosure is of no value unless it provides a measure of the risk contingent liability presents. When useful measures of contingent liabilities are formulated, measures which quantify and provide bases for fair comparisons such as the measures of market risk required by Item 305 of Regulation S-K, they should be required to be disclosed. The SEC should allow the time necessary to develop appropriate measures of contingent liability risk.

We also think that it is important for the SEC to recall the context in which the revisions contained in the proposed rule initially were introduced as an interpretation of existing disclosure requirements. It is widely accepted that the major auditing firms, under intense scrutiny as a result of Enron and the other erupting scandals, proposed the additional MD&A disclosure in an attempt to suggest that the scandals were largely the result of poor disclosure rather than poor accounting. As a consequence, their request for rule making dumped a broad range of items into MD&A, many of which would be better placed in financial statements, schedules, or elsewhere in SEC filings. The additional disclosure that the auditing firms proposed was adopted, almost verbatim, by the SEC in less than thirty days with minimal consideration. Now it is being reconsidered, but at a time when the SEC is overburdened and understaffed and traditional commentators do not have the resources to thoughtfully respond to all of the SEC's different rule proposals. We believe that changes as dramatic as those required by the proposed rule should be more carefully considered than they will be under the prevailing fire drill mentality that seems to be driving the rule-making process.

The SEC also needs to decide what role it wants MD&A to fulfill. With the addition of SAB 100 disclosure, critical accounting policies and the many other items that the Staff has informally placed into MD&A, it has become a dumping ground for disclosure and, thereby, risks diminishing its ability to provide investors a meaningful view of a company's operations "through the eyes of management." Piling on the disclosure required by the proposed rule just exacerbates this problem. We believe that the SEC would be well advised to rethink the long-term role of MD&A and to consider bifurcating it into both an operationally focused section and a financially focused section. Otherwise, it will be of decreasing, not increasing, value.

Finally, in advance of adopting any rules we urge the SEC to consider implementing a pilot program with a small group of companies as they attempt to enhance their disclosure consistent with the proposed rules. This approach worked well with the SEC's Plain English initiative and would provide the SEC with reliable insight into the practicability of the proposed rules at minimal cost. We believe that the proposed rules are more demanding than the SEC appears to envision. This intermediate process would work to improve the SEC's understanding of the issues and the ultimate quality of disclosures.

We appreciate the opportunity to comment on the proposed changes to Item 303 of Regulations S-K. If you have any questions regarding our comments, please contact either Brink Dickerson or Sandy White.



1 We also note that indemnity clauses are very common contingent liability provisions.