As a matter of policy, the Securities and Exchange Commission disclaims responsibility for any private publication or statement of any SEC employee or Commissioner. This speech expresses the author's views and does not necessarily reflect those of the Commission, the Commissioners, or other members of the staff. |
It's a pleasure to be here. Today I am going to share some thoughts on Consolidation of Variable Interest Entities under FIN 46R,1 the Accounting for Income Taxes, and the Accounting for Benefit Plans.
First I would like to share with you our views on a few FIN 46R issues.
We have seen a number of questions about whether certain aspects of a relationship that a variable interest holder has with a variable interest entity (VIE) need to be considered when analyzing the application of FIN 46R. These aspects of a relationship are sometimes referred to as "activities around the entity". It might be helpful to consider a simple example. Say a company (Investor A) made an equity investment in a potential VIE and Investor A separately made a loan with full recourse to another variable interest holder (Investor B). We have been asked whether the loan in this situation can be ignored when analyzing the application of FIN 46R. The short answer is no. First, FIN 46R specifically requires you to consider loans between investors as well as those between the entity and the enterprise in determining whether equity investments are at risk,2 and whether the at risk holders possess the characteristics of a controlling financial interest as defined in paragraph 5(b) of FIN 46R.3 It is often difficult to determine the substance of a lending relationship and its impact on a VIE analysis on its face. You need to evaluate the substance of the facts and circumstances. The presence of a loan between investors will bring into question, in this example, whether Investor B's investment is at risk and depending on B's ownership percentage and voting rights, will influence whether the at risk equity holders possess the characteristics of a controlling financial interest.
Other "activities around the entity" that should be considered when applying FIN 46R include equity investments between investors, puts and calls between the enterprise and other investors and non-investors, service arrangements with investors and non-investors, and derivatives such as total return swaps. There may be other activities around the entity that need to be considered which I have not specifically mentioned. These activities can impact the entire analysis under FIN 46R including the assessment of whether an entity is a VIE as well as who is the primary beneficiary.
In another situation involving activities around the entity, investors became involved with an entity because of the availability of tax credits generated from the entity's business. Through an arrangement around the entity, the majority of the tax credits were likely to be available to one specific investor. Accordingly, the staff objected to an analysis by this investor that 1) did not include the tax credits as a component of the investor's variable interest in the entity and 2) did not consider the impact of the tax credits and other activities around the entity on the expected loss and expected residual return analysis.
We also see a number of questions about FIN 46R in the area of related parties. FIN 46R looks to FAS 57, Related Party Disclosures, for the definition of related parties and also includes de facto agents as defined in paragraph 16 of FIN 46R. If the parties satisfy these definitions, there is a related party group. Once a related party group has been identified, paragraph 17 specifies how to determine which of the related parties is the primary beneficiary. It is important to read the words in paragraph 17 plainly. Paragraph 17 requires an overall assessment of which party is the most closely associated with the entity. When considering questions under paragraph 17, the staff considers all the factors in paragraph 17 and any other factors that may be relevant in making this overall assessment. We do not view paragraph 17 to be a matter of checking the boxes for the four factors listed and adding up who has the most boxes checked. Instead we look at all relevant factors in their entirety considering the facts and circumstances involved. We have also been asked whether any of the factors in paragraph 17 carry more weight than any others or whether any of the factors in paragraph 17 are determinative. There is no general answer to this question. Instead, the facts and circumstances of the situation should be considered to determine whether one factor or another is more important.
I would also like to address the information scope out in FIN 46R. While the staff recognizes that FIN 46R is a challenging area, it is a company's responsibility to prepare financial statements in accordance with GAAP. The staff believes that an investor has the same responsibility for analyzing whether to consolidate a variable interest entity, and for preparing financial statements in which a variable interest entity is consolidated, as they do in the case of a voting interest entity. FIN 46R only includes an information out for enterprises involved in entities created prior to December 31, 2003.4 We, therefore, expect that all the information necessary to make a FIN 46R assessment and, if required, to consolidate a variable interest entity is available for entities created after December 31, 2003. Additionally, in those cases where a company believes they can avail themselves of the information out for entities created before December 31, 2003, companies should be prepared to support how you have satisfied the exhaustive efforts criterion.
Now let me move on to accounting for income taxes.
Some have argued that recognition of contingent tax benefits should be assessed like any other asset. Others argue that a higher threshold should be applied in the recognition of tax benefits and that a FAS 5, Accounting for Contingencies, probable notion is appropriate. As you may know, the FASB is currently working on an interpretation related to tax contingencies that should clarify the model to be used. Until that interpretation is finalized, we expect registrants will use a consistent method and have a reasoned basis for their accounting. We would also encourage disclosures where this policy could have a material effect on the financial statements.
Many agree that before recognizing a tax deduction in the financial statements, the deduction should be probable of being sustained under IRS audit without consideration for audit detection risk. We recognize that some companies take positions on their tax return that do not meet the probable threshold. Some call this the "as filed" basis. The SEC staff believe that a contingent tax liability should be recorded to account for this difference between the book and as filed tax return positions. In addition, we believe disclosures for both recorded and unrecorded exposures to contingent tax liabilities are required by FAS 5, Accounting for Contingencies. I'm sure many of you are thinking "if I follow the FAS 5 model for accrual and disclosure of these differences, I will provide a road map of my tax positions to the IRS." We have empathy for this but remind registrants that concern over confidentiality of information and disclosure requirements has been with us for a long time and is not reserved to income taxes. This concern is present in other areas as well. However, confidentiality concerns are not valid reasons for non-compliance with GAAP or the SEC's disclosure rules.
Before leaving the topic of income taxes, I would like to remind companies and their auditors of their documentation responsibilities. Tax contingencies is an area where some people may have been inclined to document less rather than more due to confidentiality concerns and a fear that tax workpapers will end up in the hands of the IRS. We appreciate these concerns but remind you of your responsibilities under Sections 404 and 302 of the Sarbanes-Oxley Act related to CFO and CEO certifications. It is equally important for auditors to recognize their responsibilities under auditing standards. Audit documentation for taxes, including tax contingencies, should follow the same high standards as in other audit areas. The staff believes that the auditor's obligation to report to shareholders on the company's financial statements takes precedence over any request from the company for confidentiality. Inherent in this obligation is the need to make sure the auditor's workpapers support the auditor's findings in each area, in particular, the accounting for income taxes.
Now I would like to share some thoughts with you about accounting for benefit plans.
FAS 87, Employers' Accounting for Pensions, FAS 106, Employers' Accounting for Postretirement Benefits Other Than Pensions, and FAS 112, Employers' Accounting for Postemployment Benefits, have been with us for some time. In recent years there has been increased focus on certain aspects of the accounting and disclosure in these areas. I'd like to take this opportunity to bring these standards to the forefront of your considerations for financial reporting.
It is important to be cognizant of the fact that FAS 87, FAS 106, and FAS 112 are all driven from the concept of the "substantive plan".5 In most cases the substantive plan is the written plan, but it's not uncommon for the written plan to be modified by provisions documented in other places and by a company's past practice in providing benefits. Sometimes the substantive plan provisions will be documented outside the written plan in other contracts with employees or their representatives, such as union contracts. A company's substantive plan is the plan the employees have come to know as the plan and the plan the employees expect from the company's past practice. For example, a written plan might include employer/employee cost sharing provisions but these cost sharing provisions could be substantively modified, or eliminated, via the company's pattern of past practice or provisions documented in other places. We have seen a situation where a company's pattern of past practice indicated that the company was absorbing more of the plan costs than provided for in the written plan. As such, the SEC staff objected to the company's accounting since it did not incorporate the substance of the company's past practice of cost sharing.
Another area that impacts the accounting for benefit plans under FAS 87, FAS 106, and FAS 112, but that is not uncommon in other areas of accounting, is assumptions. Benefit plan accounting relies on the use of a number of assumptions that will vary by the nature of the plan and the benefits provided. Common assumptions include assumptions about employee demographics, retirement age, compensation levels, future costs, pay rates, turnover, mortality, expected long-term rate of return on plan assets, discount rate, amount and timing of claims, among other things.
Since the accounting under FAS 87, FAS 106 and FAS 112, relies heavily on the assumptions used in the benefit computations, the relevance, reasonableness and reliability of a company's accounting and financial reporting is therefore dependent upon the use of relevant, reasonable and reliable assumptions. Each plan assumption should reflect the best estimate of the plan's future experience6 and should be reflective of the employee base covered by the plan.
One key assumption that has received attention lately is the assumed discount rate. Companies sometimes benchmark their discount rates off the rate used by other companies. It's important to note that both FAS 87 and FAS 106 indicate that, "conceptually, the selection of an assumed discount rate should be based on the single sum that, if invested at the measurement date, would generate the necessary cash flows to pay the benefits when due".7 Each company should have support for the rate selected based on an evaluation, not just a comparison to other companies. For example, in some cases, a company may construct a hypothetical portfolio of high quality instruments with maturities that mirror the benefit obligation in order to accurately estimate the discount rate relevant to their particular plan.
Another key assumption in benefit plan accounting is mortality. We suspect that some plan valuations may be based on 20 year old mortality tables, even though more recent tables are available. In addition to using the most recent tables, it's important that the mortality tables used in benefit plan computations be reflective of the employee base covered under the plan. Since we are much more a world of service providers than we used to be when some of the common mortality tables were developed, selection of a mortality table should take this fact into consideration. If your company operates in a service industry and you use a mortality table developed from employee data in manufacturing industries, some might argue that the mortality table may not be reflective of your employee base.
Although I have only specifically addressed two of the key assumptions used in benefit plan accounting, we believe thoughtful consideration should be given to all the assumptions used in your benefit plan accounting. It is important to comprehensively assess your accounting in these areas on an ongoing basis to determine if the assumptions used in the computations are, and therefore your accounting is, relevant, reasonable, and reliable.
I thank you for your time today.