May 19, 2000
Mr. Jonathan G. Katz
Securities and Exchange Commission
450 Fifth Street, N.W.
Washington, D.C. 20549-0609
Dear Mr. Katz:
William M. Mercer thanks the SEC for the opportunity to comment on its February 2000 concept release concerning the standards published by the International Accounting Standards Committee (IASC) which is attached hereto. As requested, three copies of this letter and its attachments will be couriered for Monday, May 23rd delivery.
Ethan E. Kra, FSA, Ph.D.
SEC Concept Release on International Accounting Standards (File No. S7-04-00)
Submission by William M. Mercer (May 2000)
I. Introduction and Overview
William M. Mercer ("Mercer") thanks the SEC for the opportunity to comment on its February 2000 concept release concerning the standards published by the International Accounting Standards Committee (IASC). Mercer is the world's largest consulting firm that advises international organizations and multinational companies on all aspects of the employer/employee relationship and employs more actuaries than any other employer. Mercer has particular expertise in the application and interpretation of standards for employers' accounting for employee benefits around the world. For many years, we have been assisting clients in:
Our submission, therefore, is confined to our experience and views on the detailed application and interpretation of IAS 19 in comparison with FAS 87 and other relevant standards. Our answers to the 26 questions posed in the concept release as they apply to accounting for employee benefits are set out in Part II. We summarize below our views and the thrust of our submission.
II. Answers to the 26 Specific Questions in the Concept Release
A. Criteria for Assessment of the IASC Standards
1. Are the Core Standards Sufficiently Comprehensive?
Do the core standards provide a sufficiently comprehensive accounting framework to provide a basis to address the fundamental accounting issues that are encountered in a broad range of industries and a variety of transactions without the need to look to other accounting regimes? Why or why not?
As mentioned previously, we are confining our response to this and all other questions to employers' accounting for employee benefits. IAS 19 addresses all the various types of benefits under five categories in a single standard. (These five categories are listed in paragraph 4 of IAS 19.) By contrast, under U.S. GAAP, employee benefits are addressed under several standards (e.g., FAS 87 and FAS 88 for pensions and FAS 106 for other postretirement benefits). Looking at the five categories under IAS 19, there are only broad disclosure requirements for the last category, equity compensation benefits, instead of the more detailed requirements under FAS 123 in U.S. GAAP. However, IAS 19 is more prescriptive for certain other benefits, e.g., jubilee benefits under the third category, other long-term employee benefits.
Should we require use of U.S. GAAP for specialized industry issues in the primary financial statements or permit use of home country standards with reconciliation to U.S. GAAP? Which approach would produce the most meaningful primary financial statements? Is the approach of having the host country specify treatment for topics not addressed by the core standards a workable approach? Is there a better approach?
Under U.S. GAAP, employers with rate-regulated operations can expense the pension cost specified for rate-making purposes and recognize the difference between that amount and the FAS net periodic pension cost as an asset or liability. If the accounting standard is changed to IAS, most such companies would have to account separately for the difference between IAS 19 and U.S. GAAP (until the rate regulators could be convinced to accept the IAS expense).
Are there any additional topics that need to be addressed in order to provide a comprehensive set of standards?
We believe that a prescriptive standard for expensing equity compensation benefits is necessary for proper comparability. This is essential for certain industries, e.g., high tech.
2. Are the IASC Standards of Sufficiently High Quality? Why or Why Not?
Are the IASC standards of sufficiently high quality to be used without reconciliation to U.S. GAAP in cross-border filings in the United States? Why or why not? Please provide us with your experience in using, auditing or analyzing the application of such standards. In addressing this issue, please analyze the quality of the standard(s) in terms of the criteria we established in the 1996 press release. If you considered additional criteria, please identify them.
Generally, we believe that IAS 19 and the corresponding U.S. GAAP standards are of comparable quality, as there are no major differences in terms of actuarial methodology and principles. We comment on the differences and implications below.
What are the important differences between U.S. GAAP and the IASC standards? We are particularly interested in investors' and analysts' experience with the IASC standards. Will any of these differences affect the usefulness of a foreign issuer's financial information reporting package? If so, which ones?
The main differences between the relevant U.S. GAAP standards (FAS 87, etc.) and IAS 19, and some brief comments on them, are as follows:
1. Transition from previous accounting method
Note: transition dates are much earlier for U.S. GAAP than IAS 19
Net obligation or asset at transition generally recognized over average remaining service of employees expected to receive benefits under FAS 87; immediate recognition possible under FAS 106
Net obligation at transition generally recognized immediately (or over 5 years at most); net asset recognized immediately
Transition from U.S. GAAP to IAS 19 could result in a significant change in earnings, which could affect comparability with prior periods.
2. Prior or past service costs from plan amendments
Generally recognized over average remaining service of employees expected to receive benefits under FAS 87; over average service to full eligibility date under FAS 106
Generally recognized immediately (or over average remaining vesting period)
This is an area of controversy; the relative merits are discussed below. IAS 19 may discourage plan improvements for past service.
3. Additional minimum liability (AML) test
Under FAS 87 only, with possible intangible asset and other accumulated comprehensive income
Not considered necessary on top of normal recognition and disclosure requirements
4. Limit on balance sheet prepaid or net asset
Yes, based on measure of maximum economic benefits available by way of future contribution reductions or refunds from plan
Limit has theoretical attraction, but stricter guidance should be given for consistent application.
5. Valuation of plan assets
Generally fair value (market-related value may be used for expense)
Always fair value
IAS 19 permits no flexibility to smooth equity fluctuations.
6. Measurement date
May be up to three months before balance sheet date
Always balance sheet date although estimates are permissible
An earlier measurement date permits adequate time to collect complex information albeit at a loss of immediacy of information.
7. Curtailments and settlements
FAS 88 differentiates accounting for curtailments and settlements
IAS 19 treats curtailments and settlements similarly
8. Equity compensation benefits
Detailed rules under APB No. 25 and FAS 123
No calculation method prescribed; only disclosure requirement
Prescription improves comparability and consistency.
9. Multi-employer plans
Recognizes the required contribution for the period as the expense, and the due and unpaid amounts as a liability. Discloses cost, liability, and plan description.
IAS 19 presumes that most multi-employer plans allocate expense components to each contributing employer and these amounts are reported on its financial statements. If not available, U.S. GAAP treatment effectively applies, with supplementary disclosures.
IAS 19's presumed allocation is typically not done in the U.S. Such a change would result in a significant cost to the multi-employer plan, which the Department of Labor could prohibit as an inappropriate plan expenditure.
10. Business combinations
Currently, business combinations can be accounted for under purchase accounting rules or pooling-of-interest rules. FASB issued an exposure draft in September 1999 to modify Opinion 16 to eliminate pooling-of-interest.
Requires purchase accounting
Proposed FASB modification to Opinion 16 to eliminate use of pooling-of-interest would align both accounting standards.
It should be noted that the difference in items 1, 2, 4, 5 and 6 could significantly affect the results for U.S. companies.
None of the above differences, in our view, is likely to affect the usefulness of a foreign issuer's reporting package under IAS, except perhaps for equity compensation benefits. It would also be helpful to have greater consistency of terminology between the two sets of standards. For example, IAS 19 includes pensions and other postretirement benefits in the definition of postemployment benefits, whereas FAS 112 (Employers' Accounting for Postemployment Benefits) excludes them.
Would acceptance of some or all of the IASC standards without a requirement to reconcile to U.S. GAAP put U.S. companies required to apply U.S. GAAP at a competitive disadvantage to foreign companies with respect to recognition, measurement or disclosure requirements?
As noted above, although the methodology and principles underlying the two sets of standards are similar, the results could be very different. Since U.S. companies are not allowed the choice of the two sets of standards, they could be at a competitive disadvantage. Further, currently U.S. GAAP financial statements are acceptable for listing purposes on stock exchanges around the world. We understand that the European Commission is considering the possibility of requiring companies listed on EU stock exchanges to comply with IAS, which presumably would mean that U.S. GAAP would no longer be acceptable for listing there and reconciliation to IAS would be required. This would put U.S. companies at a competitive disadvantage, as they would have to bear the cost of complying with the two sets of standards. This is comparable to the situation currently faced by non-U.S. companies having to reconcile to U.S. GAAP for U.S. listing purposes.
Based on your experience, are there specific aspects of any IASC standards that you believe result in better or poorer financial reporting (recognition, measurement or disclosure) than financial reporting prepared using U.S. GAAP? If so, what are the specific aspects and reason(s) for your conclusion?
The table in our answer to question 5 highlights the differences between U.S. GAAP and IAS 19. We would make the following additional comments:
Transition: In our experience, most companies adopting IAS 19 have chosen to recognize the transitional amount immediately in the balance sheet as a charge or credit to retained earnings. It seems inappropriate to require a company that has complied with U.S. GAAP and is transitioning to IAS to recognize unrecognized gains immediately and unrecognized losses either immediately or over a period of up to five years when IAS 19 adopts the same structure for recognizing gains or losses as FAS 87/106. This issue will be important to current foreign issuers who may wish to use IAS in the future should the SEC allow it.
Past service costs (in IAS 19 terminology) or prior service cost (in FAS 87/106 terminology): There is no agreement within our firm as to which is theoretically appropriate. Some actuaries advocate immediate recognition on the basis that the costs are in respect of prior service. Others advocate deferred recognition using the arguments set out in the basis for conclusions under FAS 87 and FAS 106 (paragraphs 162-3, 167, 170-1 of FAS 87 and 285-7 of FAS 106). As noted in the table in our answer to question 5, immediate recognition (which is effectively the treatment for increases to pension benefits in many countries under IAS 19) may discourage plan improvements for past service.
3. Can the IASC Standards Be Rigorously Interpreted and Applied?
(a) The experience to date
Is the level of guidance provided in IASC standards sufficient to result in a rigorous and consistent application? Do the IASC standards provide sufficient guidance to ensure consistent, comparable and transparent reporting of similar transactions by different enterprises? Why or why not?
In our view, the level of guidance is sufficient, even though accounting for employee benefits is a complex subject. This is largely the case because of the substantial experience accumulated with FAS 87 and other related corresponding U.S. GAAP standards. The basis for conclusions in IAS 19 (IAS 19 is one of the few IASC standards to include such a basis) is also helpful. Since the standard is relatively new, we would expect further guidance over time. A document such as the FAS 87/88/106 Implementation Guides ("Questions & Answers") for IAS 19 would be helpful (or, failing that, at least a process for the IASC to collect feedback and issue guidance periodically). A number of areas which, to us, are open to potential misunderstanding or inconsistent application are as follows:
Are there mechanisms or structures in place that will promote consistent interpretations of the IASC standards where those standards do not provide explicit implementation guidance? Please provide specific examples.
Some of the uncertainties in interpreting IAS 19 mentioned above have been informally discussed with the IASC. The Standing Interpretations Committee (SIC) is an official committee at the IASC which meets quarterly to provide formal interpretations. However, the perception is that the official route would take too much time and resources. In the longer run, we believe, as mentioned in the answer to question 8, it would be useful for the IASC to compile and issue a question-and-answer booklet such as those produced by the FASB following issuance of FAS 87/88/106, or at least establish a process for the IASC to collect feedback and issue guidance periodically.
In your experience with current IASC standards, what application and interpretation practice issues have you identified? Are these issues that have been addressed by new or revised standards issued in the core standards project?
Apart from the difficulties of interpretation mentioned above, we have identified a particular problem with the IAS 19 definition of pension plan assets for certain trust funds and insurance contracts in France, Germany and Sweden. The definition was, in our view, unnecessarily restrictive in these countries and this led to the IASC's decision to start reviewing the matter in July 1999. At their March 2000 Board meeting, it was decided that the definition would indeed be revised and an Exposure Draft is expected to be approved at their next Board meeting in June 2000. This, in our view, shows that the IASC is prepared to consider changes to their standards in light of particular difficulties that emerge after their application in particular countries. However, the process can take too much time (as discussed further below).
Is there significant variation in the way enterprises apply the current IASC standards? If so, in what areas does this occur?
We have not had adequate experience under IAS 19 to determine if there will develop any significant variation in the way enterprises apply IAS 19. However, because IAS 19 is a sufficiently prescriptive standard, we would not expect any significant variation in its application.
(b) The need for a financial reporting infrastructure
After considering the issues discussed in (i) through (iv) below, what do you believe are the essential elements of an effective financial reporting infrastructure? Do you believe that an effective infrastructure exists to ensure consistent application of the IASC standards? If so, why? If not, what key elements of that infrastructure are missing? Who should be responsible for development of those elements? What is your estimate of how long it may take to develop each element?
We believe the discussion in (i) through (iv) below provides the various elements and we have contributed to the extent that we consider our particular experience with IAS 19 to be relevant to the general discussion. Beyond that, however, we do not have more detailed answers as they fall outside our areas of expertise and experience.
(i) The interpretive role of the standard-setter
What has your experience been with the effectiveness of the SIC in reducing inconsistent interpretations and applications of IASC standards? Has the SIC been effective at identifying areas where interpretive guidance is necessary? Has the SIC provided useful interpretations in a timely fashion? Are there any additional steps the IASC should take in this respect? If so, what are they?
As mentioned above, our perception is that the SIC might take too much time in dealing with questions of interpretation in a formal and authoritative way. Informally, we have found IASC staff members helpful when available. We feel that the IASC may generally suffer from a lack of resources and detailed experience as a standard-setter, particularly when compared to, say, the FASB.
(ii) The restructuring of the IASC
Do you believe that we should condition acceptance of the IASC standards on the ability of the IASC to restructure itself successfully based on the following characteristics: an independent decision-making body; an active advisory function; a sound due process; an effective interpretive function; independent oversight representing the public interest; and adequate funding and staffing? Why or why not?
We agree that successful restructuring of the IASC is necessary in the longer term to improve the quality of the standards and international convergence to those standards. However, we believe that there is very significant support for the IASC standards in Europe. This should ensure a successful restructuring in the fullness of time.
(iii) The role of the auditor in the application of the standards
What are the specific practice guidelines and quality control standards accounting firms use to ensure full compliance with non-U.S. accounting standards? Will those practice guidelines and quality control standards ensure application of the IASC standards in a consistent fashion worldwide? Do they include (a) internal working paper inspection programs and (b) external peer reviews for audit work? If not, are there other ways we can ensure the rigorous implementation of IASC standards for cross-border filings in the United States? If so, what are they?
We cannot comment in detail on the accounting firms' audit and quality control procedures. As actuaries and consultants, Mercer, among other things, calculates employee benefit costs in line with IAS 19. We have global quality control standards and peer review processes in place. IAS 19 encourages (but does not require) an enterprise to involve a qualified actuary in the measurement of material post-employment benefit obligations.
Should acceptance of financial statements prepared using the IASC standards be conditioned on certification by the auditors that they are subject to quality control requirements comparable to those imposed on U.S. auditors by the AICPA SEC Practice Section, such as peer review and mandatory rotation of audit partners? Why or why not? If not, should there be disclosure that the audit firm is not subject to such standards?
Is there, at this time, enough expertise globally with IASC standards to support rigorous interpretation and application of those standards? What training have audit firms conducted with respect to the IASC standards on a worldwide basis? What training with respect to the IASC standards is required of, or available to, preparers of financial statements or auditors certifying financial statements using those standards?
Our view is that IAS 19, being very similar to FAS 87 and other related U.S. GAAP standards, does not cause any particular training problems for international actuarial/consulting firms such as Mercer.
(iv) The role of the regulator in the interpretation and enforcement of accounting standards
Is there significant variation in the interpretation and application of IASC standards permitted or required by different regulators? How can the risk of any conflicting practices and interpretations in the application of the IASC standards and the resulting need for preparers and users to adjust for those differences be mitigated without affecting the rigorous implementation of the standards?
We are not aware of any variation in the interpretation and application of IASC standards permitted or required by different regulators. However, the issuance of implementation guides or other guidance should reduce the scope for variation.
Would further recognition of the IASC standards impair or enhance our ability to take effective enforcement action against financial reporting violations and fraud involving foreign companies and their auditors? If so, how?
We request comment with respect to ways to assure access to foreign working papers and testimony of auditors who are located outside the United States. For example, should we amend Regulation S-X to require a representation by the auditor that, to the extent it relied on auditors, working papers, or information from outside the United States, the auditor will make the working papers and testimony available through an agent appointed for service of process? If not, should we require that the lack of access to auditors' workpapers be disclosed to investors? Is there another mechanism for enhancing our access to audit working papers?
B. Possible Approaches to Recognition of the IASC Standards for Cross-Border Offerings and Listings
What has been your experience with the quality and usefulness of the information included in U.S. GAAP reconciliations? Please explain, from your viewpoint as a preparer, user, or auditor of non-U.S. GAAP financial statements, whether the reconciliation process has enhanced the usefulness or reliability of the financial information and how you have used the information provided by the reconciliation. Please identify any consequences, including quantification of any decrease or increase in costs or benefits, that could result from reducing or eliminating the reconciliation requirement.
For employee benefits, there is no doubt that FAS 87 and related U.S. GAAP standards have set the tone for other standards (including IAS 19) to follow. The quality and usefulness of the information to analysts and other users of financial statements applies not only in the examination of periodic financial statements but also in due diligence exercises for acquisitions where a realistic and reliable assessment of employee benefit costs is required.
Should any requirements for reconciliation differ based on the type of transaction (e.g., listing, debt or equity financing, rights offering, or acquisition) or the type of security (e.g., ordinary shares, convertible securities, investment grade or high yield debt)? Are there any other appropriate bases for distinction?
If the current reconciliation requirements are reduced further, do you believe that reconciliation of a "bottom line" figure would still be relevant (e.g., presenting net income and total equity in accordance with U.S. GAAP)?
Reduced reconciliation can often engender more confusion than clarity.
Should any continuing need for reconciliation be assessed periodically, based on an assessment of the quality of the IASC standards?
The continuing need for reconciliation should not be based on an assessment of the quality of the IASC standards but rather on how these standards compare with U.S. GAAP. Hopefully, over time there will be one global standard, so that reconciliation will only be required temporarily.
The IASC standards finalized as part of the core standards project include prospective adoption dates. Most standards are not required to be applied until fiscal years beginning on or after January 1, 1998, at the earliest. Should we retain existing reconciliation requirements with respect to the reporting of any fiscal year results that were not prepared in accordance with the revised standards or simply require retroactive application of all revised standards regardless of their effective dates? If not, why not?
For employee benefits, it is usually difficult and impractical to restate or reconcile prior year results. This is because actuarial valuations of employee benefit plans are costly and time-consuming, and carrying them out at effective dates before the latest available valuations would not normally add much value. As a consequence, we would normally estimate previous valuation results by projecting the latest results backwards. We think the most appropriate approach is to apply the transition provisions of IAS 19 at the latest applicable date (i.e., the first accounting period beginning on or after January 1, 1999). We would then project backwards using the provisions of IAS 19 to reconstruct prior years' results.
Does the existence of a reconciliation requirement change the way in which auditors approach financial statements of foreign private issuers? Also, will other procedures develop to ensure that auditors fully versed in U.S. auditing requirements, as well as the IASC standards, are provided an opportunity to review the financial reporting practices for consistency with those standards? If so, please describe these procedures. Alternatively, will the quality of the audit and the consistency of the application of the IASC standards depend on the skill and expertise of the local office of the affiliate of the accounting firm that conducts the audit?