September 29, 2008 Financial Accounting Standards Board (FASB) Attn: Technical Director, File Reference No.: 1570-100 401 Merritt 7 P. O. Box 5116 Norwalk, CT 06856-5116 Re: Comments on Conceptual Framework of Financial Reporting: The Objective of Financial Reporting and Qualitative Characteristics and Constraints of Decision-Useful Financial Reporting Information, Exposure Draft On behalf of the Financial Reporting Committee of the American Academy of Actuaries 1, we offer the following comments on the FASB/IASB (International Accounting Standards Board) Exposure Draft. The exposure draft is evidence of a high level of effort and skill devoted to its preparation. The overall clarity is particularly appreciated during reading and study of the document. We do have a few comments about the Exposure Draft which we believe should be considered. The members of the American Academy of Actuaries are highly trained practitioners in mathematical and statistical approaches to quantifying risk, and many practice in the insurance industry as well as broader financial areas in the industry. Actuaries work in roles which include preparers, auditors, and users of financial statements. As such, we won t comment on all questions in the discussion paper, but instead on selected areas that we believe most directly impact the insurance industry and in which actuaries are recognized experts. We have provided our responses to your specific questions further on in this material. Many of the questions touch upon more than one area we wish to address, and most of the areas we address touch upon more than one question. Therefore, for emphasis and ease of reference, we have utilized the following presentation. Immediately following is summary list of our broad areas of concerns and comments concerning those areas. Attachment 1 presents a discussion by subject of those areas of concern and comment. Attachment 2 specifically addresses each of your questions. 1 The American Academy of Actuaries is a 16,000-member professional association whose mission is to assist public policymakers by providing objective expertise and actuarial advice on risk and financial security issues. The Academy also sets qualification, practice, and professionalism standards for actuaries in the United States.
A. General Matters 1. Objective We are not convinced that financial report encompasses all of the decisions of equity investors (Page i, Chapter 1, Item 3, OB 13, BC 1.29). 2. Cost and Materiality We are pleased to see and heartily endorse the recognition of these as limiting factors (S 7, QC 29, QC 30). 3. Estimates, Judgments, and Models (OB 14) We are pleased to see the importance of these factors recognized. We do offer some suggestions that we hope will help you enhance the presentation (QB 14). 4. Provision for Uncertainty (QC 10, QC 11, BC 20, BC 21) We are concerned that this matter, as near as we can tell, is not discussed in the material. B Other Matters These concern matters of fairly narrow scope. They are listed here for convenience. 1. Use of the phrase freedom from error (QC 11) 2. Reference to a level of accuracy (QC 11) 3. Possible misinterpretation of material on estimates and probabilities (QC 20, BC 2.24) 4. An apparently mechanical approach to verifiability and an apparent tension between the use of verifiability in different portions of the document (OC 20, QC 21, BC 2.21, BC 2.27, BC 2.28) 5. Limited scope of types of other information (BC 2.24) Each of these is addressed in some detail in Attachment 1. Responses to the specific questions presented in the Exposure Draft are contained in Attachment 2. Attached are: 1. Discussion of items mentioned in above summary, and 2. Responses to questions presented in the Exposure Draft We appreciate this opportunity to comment. We will be pleased to respond to any questions you may have and provide our thoughts to the FASB as you continue to work on this important project. Sincerely, Henry Siegel Chairperson, Financial Reporting Committee American Academy of Actuaries 2
DISCUSSION OF AREAS OF COMMENT This discussion presents first the comments on general matters, followed by comments on other matters. Within each group, the comments are ordered by their first reference. A. General Matters 1. Objective: Page i, Chapter 1, Item 3, OB 13 BC 1.29 We doubt that financial statements can, encompass all of the decisions that equity investors make as stated in Page i, Item 3 and BC 1.29. This limitation seems to be recognized in OB 13 which states that: Financial reporting by a particular entity is but one source of information needed by capital providers. Those users of financial reports also need to consider pertinent information from other sources. A minor modification such as using the phrase, are of primary importance to most of the decisions in Item 3 and BC 1.29 would clarify the meaning. 2. Cost and Materiality: S 7, QC 29, QC 30 We believe these concepts are particularly applicable to financial reporting for the insurance industry. The reason for this is that various proposals have been presented to introduce accounting techniques, the most notable of which is bifurcation which could impose significant costs and complexity for doubtful value. Bifurcation of the insurance contracts, for example, into service and financial components, could involve a great deal of increased cost and complexity, a reduction in relevance, a result that is not consistent with the intentions of parties to the contract, and would provide questionable, if any benefit. 3. Estimates, Judgments, Models: OB 14 These forms of estimation are extremely important in the context of insurance industry financial reporting and are areas of actuarial expertise. We recognize that they are important in other industries involving uncertainty. We believe the reference to, transactions and other events and circumstances that have happened or that exist is appropriate; often reference to estimates is thought of as applying to future contingent events. Should not future contingent events also be referenced? This might be done by expanding the wording in the middle of the 1 st sentence to read: happened or that exist or may happen. The penultimate sentence in OB 14 refers to, models based more on accounting conventions than on the concepts and the framework. We believe it would be desirable to base standards 3
and accounting guidance on the concepts in the framework whenever possible and believe that is intended. We suggest that replacement wording be developed to reflect this. Perhaps: although guidance is to be based on the concepts and the framework wherever feasible, there may be some instances where there is a need to accept estimates, judgments, and models based more on accounting conventions than on the concepts in the framework, although these conventions should be as consistent with the framework as possible. would help clarify. 4. Provision for Uncertainty: QC 10, QC 11, BC 20, BC 21 We are concerned does not appear to reference a provision for uncertainty. The provision for uncertainty is of utmost importance in the insurance industry (where it is now often referred to as risk margin ) and is also applicable to the activities of other industries. In fact, in Preliminary Views on Insurance Contracts (IASB May 2007), risk margin is given in Chapter 3 as one of the three building blocks for the measurement of insurance liabilities. Also, an entire subsection of Chapter 3 of that document is devoted to the subject of margins. FAS 157 (Fair Value Measurement) also recognizes, in paragraphs B4 through B10, the importance of a risk margin when an estimate is made under conditions of uncertainty. Both the IASB and FASB have recognized the need for a risk margin in insurance financial reporting. With regard to the conceptual framework, we believe this document will be read by many who do not think of the provision for uncertainty in connection with liabilities and assets in a financial statement and will not read the conceptual framework document to apply to and embrace the provision for uncertainty element. In fact, a reasonable interpretation of this material is that it contradicts the Preliminary Views on Insurance Contracts which explicitly and at length requires a risk margin and also FAS 157; we are sure this was not intended. Again, we are concerned that the document, as written, could be read to preclude a provision for uncertainty due to the omission of any mention of it; but a provision for uncertainty is needed in many areas of financial reporting. Hence we suggest explicit mention of the need for a provision for uncertainty. We believe explicit discussion of the point is needed so that QC 10, QC 11, and related matter is not read or interpreted to preclude provisions for uncertainty (whether labeled risk margin or by other terms). This can be addressed by explicit mention of the provision for uncertainty in the conceptual framework document. This could be done in a manner consistent with FAS 157 by stating: Certain measurements are made under conditions of uncertainty because the cash flows used are estimates rather than known amounts. It is appropriate that such measurements incorporate an implicit or explicit risk premium so that the estimate reflects a faithful representation of the value of such uncertainty. Another possibility is to add a new third sentence to QC 11, for example: 4
Some elements, (such as mineral reserves, provisions for uncertainty and provisions for legal settlements) may be based almost entirely on estimation when there is no available directly verifiable data. The matter of judgment is mentioned. We think adding a sentence in QC 11 after the present 3 rd sentence would make the paragraph more meaningful and robust. We suggest stating the simple fact that: As directly verifiable data for the estimate become more and more sparse, the relative importance of judgment in the estimation process increases. **************************************************************************** B. Other Matters 1. Error: QC 11 We were left unclear as to what type(s) of error this paragraph addressed, and suggest clarifying whether intentional errors, careless errors, estimation errors, or some sort of other errors are included or excluded. 2. Accuracy: QC 11 First, the present 4 th sentence conveys the impression that there is some specific level of accuracy applicable to all occasions which is required and needs to be equaled or exceeded. Also, the structure of these sentences can lead to the possible interpretation that whether or not there needs to be disclosure (see last sentence of section) depends on whether or not the level of accuracy is exceeded. We do not believe this is intended. Perhaps the issue could be addressed and the possibility of misinterpretation lessened by wording such as the following:...however, some minimum level of accuracy which would depend on the circumstances is also necessary for an estimate to be a faithful representation of an economic phenomenon. A particular situation where this discussion is applicable is the presentation of loss reserves of a property and casualty insurance company. 3. Estimates and Probabilities: QC 20, BC 2.24 In some cases, we support the concept of disclosure by such means as use of ranges may be appropriate. Disclosures of this nature alert the reader to the presence of considerable uncertainty and give at least some information about the extent of that uncertainty. However, there is a point of diminishing returns. More specifically, if QC 20 and other portions of the conceptual framework are read to require that the quotation given in BC 2.24 from Paragraph 72 of Concepts Statement 2 requiring, an indication of the probabilities attaching to different values, we believe disproportionate cost and effort will be required of small to medium size companies and perhaps of some larger ones. We suggest that BC 2.24 be strengthened to indicate that the 5
quoted type of analysis is not adopted by the IASB and FASB as the only way to disclose uncertainty. 4. Verification: QC 21, BC 2.21, BC 2.27, BC 2.28 We agree that verifiability can enhance the decision usefulness of relevant and faithfully represented financial reporting information. However, we are concerned that faithful representations of financial reporting information frequently result in information that is not directly or indirectly verifiable. This concern is addressed in Basis for Conclusions 2.27 and 2.28, but we are not satisfied that this concern is addressed sufficiently enough in Chapter 2 to guarantee the freedom to use the most appropriate actuarial projections. Our main concern is the mechanical spirit by which verification, both direct and indirect, is defined in QC 21. For example, indirect verification by checking the inputs and recalculating the outputs greatly oversimplifies the process that actuaries go through to derive estimates of unpaid claims and related liabilities. In fact, QC 21 seems to be contrary to BC 2.28 which states that, many pieces of information included in financial reports are not verifiable Given the complexity of the phenomena that actuaries quantify, the wide variety of actuarial methods that can be employed to derive estimates, and the inherent uncertainty surrounding actuarial estimates, actuarial judgment must be used to achieve reasonable results. As such, actuarial estimates that appear in financial statements can rarely be replicated, or verified, by the means suggested by this definition. One suggestion is to modify QC 21 by adding a statement that recognizes that some important pieces of information cannot be verified in the manner described in QC 21. If the words in BC 2.28, many pieces of information included in financial reports are not verifiable are included in QC 21, this would be a step toward addressing this issue. We ask the board to carefully consider language that makes QC 21 consistent with the concept of QC 20 that verifiability implies that different knowledgeable and independent observes could reach general consensus, although not necessarily complete agreement. We believe it is currently not consistent. 5. Type of Information: BC 2.24 The paragraph mentions statistical and econometrics. There are other types of information, including but not limited to, actuarial information. Of course, there is a question here of how many special areas of expertise and information are mentioned. We suggest in the second line of BC 2.24 instead of the word statistical the word other be used; and, in the fourth line, the wording be expanded to: econometric, actuarial, statistics, and other specialized areas. 6
QUESTIONS FROM EXPOSURE DRAFT The responses to the questions are from the point of view of our areas of expertise and activity. They do not apply universally to all preparers and users. A. Chapter 1 Question 1 Do you agree with the Boards conclusion (that an entity s financial reporting should be prepared from the entity perspective rather than the perspective of its owners) and the basis for it? If not, why? We have no response to this question. Question 2 Do you agree with Boards conclusion (to identify present and potential capital providers as the primary user group) and the basis for it? If not, why? We have no response to this question. Question 3 Do you agree with the objective (broad enough to encompass all of the decisions that equity investors, lenders, and other creditors make) and the Boards basis for it? If not, why? We do have observations concerning clarity. See our discussion regarding objective as referenced on Page i, Chapter 1, Item 3 and OB 13 and BC 1.29. B. Chapter 2 Question 1-a Do you agree that relevance and faithful representation are fundamental qualitative characteristics? (See paragraphs QC 2 SC 14 and BC 2.3 BC 2.24.) If not, why? Question 1-b Do you agree that comparability, verifiability, timeliness, and understandability are enhancing qualitative characteristics? (See paragraphs QC 16 QC 26 and BC 2.25 BC 2.34.) If not, why? Question 1-c 7
Do you agree that materiality and cost are pervasive constraints? (See QC 28 QC 33 and BC 2.58 BC 2.64.) If not, why? Is the importance of the pervasive constraints relative to the qualitative characteristics appropriately represented in Chapter 2? Question 2-a Financial reporting information that has predictive value or confirmatory value is relevant. Agree. Question 2-b(1) Are the fundamental qualitative characteristics appropriately identified and sufficiently defined for them to be consistently understood and useful? If not, why? Question 2-b(2) Are the components of the fundamental qualitative characteristics appropriately identified and sufficiently defined for them to be consistently understood and useful? If not, why? We agree with the wording directionally, but we have some suggestions for improving clarity. See our comments regarding Estimates, Judgments, Models QB 14; Provision for Uncertainty QC10, QC 11, BC 20, BC 21; Verification QC 21; Estimates and Probabilities QC 20, BC 2.24; Type of Information BC 2.24. Question 3 Are the enhancing qualitative characteristics (comparability, verifiability, timeliness, and understandability) appropriately identified and sufficiently defined for them to be consistently understood and useful? If not, why? Generally, yes. Again, we believe the wording is going in a good direction. See our comment on Verification QC 21, BC 2.21. Question 4 Are the pervasive constraints (materiality and cost) appropriately identified and sufficiently defined for them to be consistently understood and useful? If not, why? See our very supportive comments on Cost and Materiality S 7, QC 29, QC 30. 8