Embedded Derivatives Summary of Comments and Responses

Transcription

1 October 18, 2006 Embedded Derivatives Summary of Comments and Responses Comment Letter Abbreviations AAA American Academy of Actuaries AG Het Actuarieel Genootschap (Netherlands) ASSA Actuarial Society of South Africa CIA Canadian Institute of Actuaries DAV Deutsche Aktuarvereinigung e.v. FEE Fédération des Experts Comptables Européens IAIS International Association of Insurance Supervisors IAJ Institute of Actuaries of Japan NAIC National Association of Insurance Commissioners PWC PricewaterhouseCoopers LLP RB Ralph Blanchard SA Svenska Aktuarieföreningen UK Faculty / Institute of Actuaries (United Kingdom) Comment Letter Assessment G grammatical change T technical change C conceptual change that may require re-exposure S showstopper that may stop the standard being issued N new comment to incorporate the standard RR recommended response

2 Embedded Derivatives and Derivatives 1. This preliminary ED is relatively complex, due in part to the complicated nature of its subject. If you believe it too complex for educational guidance, do you have any specific suggestions regarding where it might be simplified? Comments (ASSA) Much of our sentiment has been expressed in general Q1 above we believe it is too complex. The flowchart in section 4.1 is very useful. Section would benefit from a more pictorial demonstration, whether it is a venn diagram of three circles (viz IAS 39, IAS 32 and IFRS 4) or a flowchart, with additional text underneath. We have also suggested certain text changes below that may assist in removing some of the complexity. Also, including examples (especially those common to many jurisdictions) may help to provide additional clarity. (AG) The PG embedded derivatives is not too complex, we think it is suitable for the purpose. It should not contain extra information, an example per country would be very relevant, but this would not be in the scope of the IAA. (CIA) We found this preliminary ED too complex for educational guidance, to the extent that it did more harm than good in our efforts to interpret the related accounting guidance. We believe this document needs extensive revision to make it understandable, and it should not be released as a PG until such revisions have been completed and exposed for comment. We include a number of examples of where the document is difficult to understand in the detailed comments below. (IAIS) We believe that providing an overview of relevant IFRSs is useful for such a complicated issue. The provision of practical guidance on how to apply IFRSs would help further and could be added as experiences are accumulated in the future. RR There is a mixed response, reaching from very complex / not suitable for release to not too complex / suitable for the purpose. Since the positive comments came from both actuarial and accounting associations, it is recommended to continue with the due process, but to make some improvements in wording and organization. The IASP is apparently helpful to those who are seeking more theoretical guidance than available in IAS 39, especially more insurance related issues relevant for identifying embedded derivatives in more advanced products. IAS 39 provides mainly examples from a non-insurance-world, which are rarely helpful in applying the guidance to special insurance issues. Considering the variety of issues, it is necessary to provide theoretical views of the rules in IAS 39 rather than examples. Actuaries can be expected to work with theoretical concepts rather than to compare specific situations with examples. Regarding the request for more examples, we have to consider two points. First, we are at the beginning of the process and it is currently not clear which examples might have international relevance. The Insurance business has very local characteristics and currently there is no known embedded derivative requiring separation, that has international relevance. IAS 39 refers to examples of international capital markets, which do not need long explanations, since they are well-known world-wide. IFRS 4 Implementation Guidance refers to examples which either, do not require separation (and do not need to be repeated in the IASP) or, are only of local relevance. The features mentioned in the Implementation Guidance requiring separation are unknown in a large number of countries. Referring to examples of only local relevance from the insurance business we would first need 2

3 Embedded Derivatives and Derivatives (PWC) The identification of embedded derivatives under IFRS is very complex and guidance on this area will be helpful. In particular, we welcome the use of a flow chart to guide actuaries in the process of identifying and accounting for embedded derivatives. (UK) We did not find anything we felt was not technically correct and overall our impression of the practice guideline (PG) is that this is very useful, particularly for actuaries who have not yet had to deal with the embedded derivative concepts under IFRS, who we would have thought were the target audience of the PG. However, we felt that it would be made more accessible and hence more useful it were shorter and made more readable without losing essential guidance, for example by including more illustrative examples. The flow chart on page 3 is an example of something that is very easy to follow. (UK) The guidance in this area we found useful was that which relates the rules to specific examples. A good example of this is IG2 of IFRS 4 which gives 29 examples of embedded derivatives within insurance contracts and their treatment. This is very clear and covers the types of EDs that most insurance companies have. (UK) The prevailing standards in this case are IAS 39 and IFRS 4 which set out the technical accounting requirements. Each standard already has its own implementation guidance and basis for conclusions to explain the accounting technicalities. The purpose of this PG should then be to provide actuaries with the necessary bridge between their actuarial understanding of insurance contracts to an accounting view of the business and concepts. We think the PG is useful in this sense: The decision tree provided right up front is a clear and concise description of the identification, measurement and disclosure processes involved. It makes clear that embedded derivatives needed to have an explicit to explain the product design, since it is known only locally. Second, the number of embedded derivatives in insurance contracts requiring separation is very small. There are large international insurance groups reporting under IFRS not reporting any separate embedded derivative in their annual report. The main purpose of the draft IASP was to provide guidance clarifying how restrictive the rules in IAS 39 are and that in most cases it is possible to demonstrate that embedded derivatives need not be separated. The issues raised in the draft IASP reflect the experience in many countries and large insurance groups of how to assess that many of those features, which on first sight require separation, do not actually qualified for that treatment. Clearly, to overcome a first view requires in depth analysis and the IASP is intended to provide motivation to go to that level of analysis. It should be noted that for any issue referred to a practical case existed, where that in depth view to the rules of IAS 39 was needed. But considering the very special and unique nature of those examples, it was not seen as appropriate to refer to those, but to provide the general principle found. Hence, different commentators found different issues to be important, depending on which cases they were already confronted with in practice. The Subcommittee is considering to provide in a separate document outside the IASP to prepare an appendix with examples. However, these examples will be intended to illustrate the way through the guidance. We will attempt during the next month to collect suitable examples and to produce as Subcommittee information on the IAA website. A first review of the draft generally supported the flow chart but commented that it could be improved. Some modifications where included as requested by the comments. Further, the placement of some paragraphs turned out to be suboptimal. This may have caused to some extent the opinion that the draft IASP is too complex, since the placement disturbed the logical structure and sequence of the paragraphs. Especially in the first part of the PG, the sequence of 3

4 Embedded Derivatives and Derivatives clause in the underlying insurance contract. This is one of those points that sometimes seem so obvious that people forget to make them but that can be quite important when considering some unusual practices. It was helpful to see a description of how to value unit-linked business and whether or not they contain an embedded derivative (with some reference to how to treat them for classification purposes as well) in Section paragraphs has been re-organized. Further, significant changes in wording and sentence structure were made to make the PG more understandable. Overall, the changes are not in substance but in wording and style and in our opinion do not require re-exposure and appropriately reflect the concerns of commentators. The re-drafting was used to eliminate any noted paraphrasing of IFRSs and to use exact wording. (CIA) 1. Scope We found that much of the guidance in this preliminary ED is nonactuarial in nature, expanding on the guidance in the related accounting guidance. Though the actuary has a role in identifying and measuring derivatives and embedded derivatives, such identification and measurement is the responsibility of management and not the actuary. If the guidance in this PG only applies to actuaries and not those responsible for identifying and measuring derivatives, then divergent practices might arise. We strongly recommend that this guidance appear within the IASB material (perhaps as "implementation guidance") rather than as IAA guidance. RR CIA: Most IASPs published refer to accounting issues and the interpretation of accounting guidance. Considering the special role of actuaries in assessing classification issues for insurance products (since actuaries have a thorough understanding of the economic substance of such products) it is necessary that actuaries deal with those issues in depth. They need a deep understanding of the accounting rational underlying the decision regarding separation. That is the main purpose of most IASPs. At the moment, there is, in the absence of new measurement requirements for insurance contracts, little need for education in actuarial issues. There are other IASPs issued which deal with even less technical issues, e.g. Changes in Accounting Policy. The purpose of Class IV IASPs (PGs) is also to indicate to non-actuaries, how actuaries as experts in complex insurance contract design understand the application of IAS 39. In this regard, such a PG is intended to be useful to all. The scope of this draft IASP is not specific to that IASP, but refers to the scope of all IASPs, which is to IFRS 4. Hence, no widening of the view in this draft is suitable in reference to the broader scope. See later for detailed comment. 4

5 Embedded Derivatives and Derivatives (PWC) In the appendix we also identify a number of areas where we believe the guidance in the ED is inconsistent with the requirements of IAS 39 and IFRS 4. The guidance note should be redrafted to avoid paraphrasing IFRS, and to focus on providing illustrative examples of the application of the accounting standards to derivatives embedded in contracts typically issued by insurance entities. Examples would be useful to illustrate the various steps of the flow chart. RR PWC Special care was taken in reviewing the Standard to address these comments. (SA) The preliminary exposure draft of the IASP is too complex, which is partly due to the complicated nature of the subject, but even more so due to the illogic structure of the document and the difficult language used. Even for readers with reasonable knowledge of the subject, it increases rather than reduces confusion. 2. Is there a need for further application guidance, particularly with respect to the measurement of embedded derivatives? If you believe there is a need, can you identify or provide specific examples for which guidance should be provided? Nothing to do. Comments (ASSA) No, we do not believe there is a need for further guidance on measurement. Fair value is adequately covered in IAS 39. (PWC) There is already sufficient guidance elsewhere on the practical measurement of derivatives that is applicable to the valuation of embedded derivatives. The IASB is also about to issue an exposure draft on measurement at fair value. Therefore, we do not believe that further guidance is necessary at this stage. 3. Do you agree with the decision taken to exclude from the Scope the topic of hedge accounting? Nothing to do. 5

6 Embedded Derivatives and Derivatives Comments (ASSA) We agree that hedge accounting should be excluded. (AG) A separate PG on hedge accounting as well as shadow accounting would be relevant. In our opinion it should not be part of the Embedded Derivatives PG, but a (short) separate paper. (PWC) Yes, we agree that hedge accounting should be excluded from the scope of the ED. 4. Are there any other embedded derivative or derivate related issues that should be addressed here? Comments (ASSA) No further issues need to be addressed. (PWC) No. (UK) We believe the areas where the PG could be developed are: At times the discussion can be quite difficult to follow. This may be because the PG is trying to be generic rather than say US or UK specific. However, this can lead to a very technical description of the standards where one has to read very carefully to understand exactly what is going on. For example, Section on identifying embedded derivatives is quite hard to follow, which is a pity since the flowchart in the preceding section shows that the identification process is relatively straightforward. As it currently stands, we cannot see people referring to the PG on a regular basis. Once people are familiar with the concepts involved they would probably find IAS 39 or IFRS 4 more useful. However, if it contained a few more flowcharts, decision trees or checklists then I could see it providing a more useful reference - perhaps even for a wider audience than actuaries. There seem to be a few small words like "is" or "not" missing in a few places. RR We attempted to improve the language of the standard. However, it is agreed that the Standard should not be US or UK specific and should address issues generally in a principle based manner. It is intended to improve the applicability of the Standard by adding in Appendix C (outside the IASP) examples applying the guidance of the Standard. Section is now moved to a more logical place (to embedded derivatives rather than to derivatives) and is significantly reworded. A significant effort was made to improve the language. As far as possible, the original wording of IFRSs is included and replaces the rephrasing. 6

7 Embedded Derivatives and Derivatives In summary, we consider the PG on Embedded Derivatives and Derivatives to be useful where it supplements the IAS guidance rather than just rewords it 5. Other comments. Comments (ASSA) Where text is copied from the standards, it would be better if it were in quotation marks. RR Quotation marks were included. It was indicated in the Background paragraph, that there are differences in definition. It was not seen as appropriate to define terms at that stage before the terms are actually required. (DAV) General remarks: In the beginning there should be a brief introduction to and explanation of the three different key terms derivative, embedded derivative and embedded derivative cash flow. 7

8 Detail Comments (DAV) Page 3 The last box of the flow chart should be completed by the option to not separate a surrender right at fixed values. (ASSA) Section 3 : Background First sentence. Refer to IAS as well, since embedded derivatives are not covered in para 9 but in fact in para 10. (CIA) 4.1 Overview The chart and accompanying notes begin with Contracts subject to the scope of IFRS 4, and yet the scope of the PG is not limited in this way. Further, the chart and accompanying notes ignore the two exceptions in IFRS 4 as to when an embedded derivative must be separated from its host contract (i.e., if the embedded derivative is itself an insurance contract and a policyholder s option to surrender an insurance contract). The chart and accompanying notes should be completed. Added. Added. CIA: The chart was modified to exclude any embedded derivative within the scope of IFRS 4 from separation. IFRS 4 grants an accounting option to separate or not to separate surrender rights. That option is now described in the last box of the chart. (PWC) In our cover letter we commented on the draft flow chart within the ED and the fact that it does not fully reflect the current text in IAS 39 and IFRS 4: Our review of the proposed flow chart has identified a few instances where this does not reflect the requirements of IAS 39 and IFRS 4. In particular the diagram omits any consideration of whether the embedded derivative/component meets the definition of an insurance contract. This has been the single most common reason for not separating embedded derivatives in practice and therefore merits separate consideration. We have attached in the appendix to this letter an alternative flow chart and provided our detailed comments on the IAA draft flow chart. PWC: The recommended change in flow chart is welcome. However, there is the need to distinguish between IFRS 4.7, which prohibits separation (if not allowed under the existing accounting policy or based on a permissive change in accounting policy), and IFRS 4.8/9 which provides an accounting choice to separate or not. This has been addressed 8

9 We suggest that the following chart be adopted instead. (Secretariat note.- See Attachment A hereto.) The ED flow chart presents boxes on disclosure requirements. We recommend including them as a footnote (see our example above) rather than an outcome of the decisions in the flow chart. In addition we note that the generic reference to IAS 32 is not particularly useful. The ED correctly acknowledges in section 4.6 that IAS 32 does not provide any specific guidance on disclosures for embedded derivatives. Because the ED is focused on embedded derivatives in host insurance contracts we would recommend referring only to the relevant requirements in IFRS 4 paragraph 39. As we are concerned that the disclosure requirements might be overlooked too easily as a result of focusing on separation, we decided to keep it in this prominent place. Other recommendations were considered. (ASSA) 4.1 : Overview The flowchart ignores IFRS 4.7 which states that an embedded derivative which meets the definition of an insurance contract does not need to be separated and measured at fair value. One probably needs a decision block before the last rectangular block asking the question whether the embedded derivative meets the definition of an insurance contract. (An example could be an endowment with a base benefit of 100 units, together with a rider where the benefit is 1000 units * the difference between the growth in an index and 5% pa. The rider is an embedded derivative, is not already measured at fair value, is stand-alone and is not closely related, but yet does not have to be separated out since it meets the definition of an insurance contract). This needs to be built into note 6 as well. The second sentence in note 6 could become note 7, since it is an independent point. (CIA) 4.2 Identification of derivatives according to IAS 39 Much of this section was difficult to understand. Some examples: (4.2.1 pp.1) A component of a contract that would on a stand-alone basis be subject to IFRS 4 or that would be a service contract is considered to be an embedded derivative. ASSA Considered now in note 3. Now covered in note 4, since related to that step. CIA The sentences were rephrased. 9

10 (4.2.1 pp.5) If a component of an insurance contract satisfies the definition of an embedded derivative, even though it falls within the scope of IFRS 4, and would not be a derivative as a stand-alone contract, the disclosure requirements in IFRS 4.39 (e) apply if the component is not measured at fair value. (4.2.1 pp.7) Nevertheless, for contracts within the scope of IAS 39, the stand-alone requirement (IAS 39.11(b)) is not the only requirement for the separate measurement of a component to apply IAS 32 and IAS 39. (ASSA) Section 4.2 : Identification of derivatives according to IAS 39 First para: sub-bullet (c) is indented too far. Last para: replace disposition with features. (DAV) Section 4.2 Sub-items (a) and (b) should be on the same level as (c). ASSA Corrected. Considered. DAV Corrected. (PWC) There are instances where the ED has paraphrased IAS 39 without producing a clearer text than the standard. In other instances the guidance provided has used language that appears to be interpretative. Both risk misunderstandings leading to errors or omissions. We have noted below in page order our most significant comments in these two areas. (PWC) Section 4.2 On page 5 there is a statement that the criterion of future settlement in the definition of a derivative is not usually relevant in the case of contracts considered in this PG. This statement is incorrect. We suggest replacing the reference to relevance with a statement that for almost all cases of embedded derivatives in insurance contracts this criterion is met. However, there may be circumstances where this criterion has to be considered. (PWC) Section The two opening paragraphs in this section are not entirely clear. If an embedded derivative meets the definition of an insurance contract it PWC Considered. The entire section was reviewed and rewritten. It is clarified now that the disclosure requirements result from IFRS 4. 10

11 would not be under the scope of IAS 32 as suggested in the second paragraph. It would be under the scope of IFRS 4. The first sentence of the third paragraph could simply say that Components with significant insurance risk transfer or DPF can be embedded derivatives under IAS 39.10, even though IAS 39.11(b) does not require separation. In the sixth paragraph the statement that a contract that contains a derivative is also within the scope of IAS 39 should be clarified by explaining that the existence of an embedded derivative brings the contract as a whole into the scope of IAS 39. However the host contract may not be in the scope of IAS 39, such as a leasing contract. The last paragraph of this section (page 7) is incorrect for the same reason noted in our comment above for leasing contracts. The purpose of the sentence was different. It was reworded to clarify the intention better. Simplified now in the reviewed scope section. (DAV) st para This para is too strict and contradicts a similar statement in the last para on p. 6. It should be made clear that it is meant as a summary of assumptions which are dealt with in (DAV) Section (Page 6), 2 nd para This para describes more or less the same as para 2 on p. 7. (DAV) Section 4.2.1, 3 rd para In order to make the unconditional prerequisite clearer, this para should start with If instead of To the extent. (CIA) Section 4.2.1, pp. 9 says: IFRS 4 states that components of contracts subject to the scope of IFRS 4 need not be separated according to IAS 39 if, as a stand-alone contract, they would be within the scope of IFRS 4. This statement is incorrect. IFRS 4 only makes this exception for embedded derivatives that are insurance contracts (as opposed to all contracts subject to IFRS 4). (ASSA) Section : Consideration of scope of IAS 32 and IAS 39 We don t believe that this is true. We suggest replacing this sentence DAV Mistake was solved by rewording. Solved by rewording. Solved by restructuring and rewording. CIA: Regarding pp. 9: The scope of IAS 39.2 (e) excludes any derivative embedded in a contract in the scope of IFRS 4, if the derivative itself would be in the scope of IFRS 4 and is more far reaching than IFRS 4.7 which excludes only insurance components from the application of IAS 39. The wording was corrected and it not longer refers to a rule in IFRS 4. Done. 11

12 with wording from IAS Delete the last sentence of the second paragraph it is repetition of the previous sentence. Fifth paragraph: delete even though it falls within the scope of IFRS 4 these words appear to be superfluous. Fifth paragraph: last line, requirement should be requirements. Seventh paragraph: Change the last sentence to read. Nevertheless, for contracts within the scope of IAS 39, the stand-alone requirement (IAS 39.11(b)) is not the only requirement to assess whether separate measurement of a component is required IFRS 4.7 must also be considered. (DAV) Section Market factor should be written in capitals. (CIA) Section , pp. 4 says: When, for example, the mix of investments is subject to management s discretion and the average duration of the obligation is longer than the average duration of the current assets, the asset portfolio is not normally considered to be directly related to the contractual cash flows. It is not clear why the relative duration of the asset and liability portfolios is relevant to the issue. (ASSA) Section : Identification of market factors In the top paragraph on page 9, reference is made to contract surrender. We suggest that either this sentence is deleted, or additional reference be made to the fact that surrenders are exempt from being classed as an embedded derivative (IFRS 4.8). (DAV) Section , 2 nd para As the term correlated might be misleading for readers with statistical background it should be replaced by related to. An additional chapter should be added stating that the following discusses several items without specifying whether an embedded derivative can be assumed or not. Solved by restructuring. The mistake was corrected. Done. Solved within restructuring. DAV Done at first appearance of that term. CIA Relative duration of assets and liabilities: If the duration of assets is shorter than those of liabilities, there is a need for reinvestment at the discretion of the insurer. Hence, policyholders did not expect that their maturity value (especially in case of participating contracts,)is determined by specific assets but based on the discretionary investment policy of the insurer, which is specific to a party. ASSA: reference to surrender: IFRS 4.8 does not mean that the component is not an embedded derivative (IFRS 4.39 (e) applies in those cases), but establishes an accounting choice whether to apply IAS or not. DAV Reworded to influenced by. Considered in rewording. However, the question is not whether it is an embedded derivative but whether it is a derivative. (DAV) , para 6 12

13 Examples include unilateral rights should read as Examples may include unilateral rights. (DAV) , para 7 We recommend to delete the 2 nd sentence as it might be misleading. The 3 rd sentence should be corrected as The fact that rights or obligations inherent in the contract have different values under different but unspecific market conditions without relation reference to a specific market factor (such as the market value of an alternative instrument that has similar characteristics at outset) is not sufficient to create a derivative. (DAV) , para 8 Because an insurer s mortality experience is not a market factor the example should be precised by the addition term life insurance contract in which because of market pressure a one-time bonus. (PWC) Section The second and third paragraphs from the bottom of page 8 appear potentially inconsistent. Paragraph 2 says "...the underlying has to be specified in the contract". Paragraph 3 says "...the variable does not have to be named explicitly in the written contract." The definition of derivative in IAS 39 refers to specific variable without it being necessary that the contract identifies it explicitly. Paragraph 2 should be amended. The first and third sentences in the second paragraph on page 9 (the paragraph begins with the words In any case ) are apparently inconsistent with the second sentence. We recommend that this be redrafted to include a clear statement of the circumstances in which a derivative may be present. The last paragraph in this section is incorrect as it refers to ex-gratia payments (one-time mortality bonuses) being a derivative. They cannot be because they are based on discretionary decisions by one party. Reworded. Reworded to be correct. Considered. PWC Was corrected in new version. Redrafted, a not was missing in the second sentence. Solved by redrafting. 13

14 (PWC) Section On page 10 there is an incorrect reference to the fact that a change in value resulting from a change in the underlying has to be material to qualify as a derivative. Whilst materiality is relevant generally in determining how to apply IFRS, it is not a specified criterion in determining whether or not there is a separable derivative. (DAV) , 1 st para Its adequacy in the 2 nd sentence should be read as the market factor s adequacy. (DAV) , 3 rd para We recommend to change this para as follows Typically the measurement of the fair value of a surrender option in a contract with a savings element would reflect the relationship between the interest provided under the contract and market interest rates scenarios, which together would might be considered a market factor, as well as the surrender value itself. (DAV) , 6 th para, last sentence The precondition applies only if the units of fund are not measured at fair value. PWC Corrected by replacing material by relevant. Irrelevant effects do not need to be considered under substance over form. DAV Solved by overall rephrasing of that paragraph. Reworded. DAV: th paragraph: If the unit fund contains a relevant amount of derivatives, the entire investment component is a derivative. It does not matter whether the fund is measured at fair value. However, measuring the fund and the investment component at fair value, would mean that t no separation or specific disclosure is required. The issue was clarified in the draft. (CIA) Some illustrative examples would be helpful in understanding sections Required effect of the market factor on the value of a derivative and Interpretation of criterion (b): alternative investments. CIA Will be added in separate material 14

15 (PWC) Section The first paragraph paraphrases the wording in IAS 39.9 in a way that introduces ambiguity. IAS 39 does not use the terms sufficiently small as not to be material but it requires no initial net investment or an initial net investment that is smaller than would be required These do not have the same meaning. For example, under the IAS 39 definition, a deep-in-the-money option can meet the definition of a derivative, even if there is a significant up-front premium, provided it is smaller than the amount paid for the non-derivative instrument. The guidance should use the same terminology as IAS 39 to avoid confusion. (PWC) Section We do not agree with the statement under number 3 in this section that suggests that IAS 39 requires an economic activity to meet the definition of a derivative. (DAV) , 2 nd para The sentence should be changed to is not considered to constitute a derivative (DAV) , sub-item 5 Insurable interest is not an IFRSs term but only mentioned in BC25ff. It should be explained either here or in the glossary. (DAV) , 2 nd para This para should be worded more precisely Under-pricing might be explained by assuming that the contract was intentionally priced below cost. or if it indicates that there is no investment or less investment in the contract than in the alternative investment. If it can be demonstrated that a contract is an investment contract rather than just being inadequately priced, it would not have the character of a derivative. Distinguishing is needed to cases where there is no investment or less investment in the contract than in the alternative investment. (CIA) 4.3 Identification of embedded derivatives according to IAS 39 We found the distinction between sections 4.2 and 4.3 blurred. Much of PWC Corrected by using the original wording from IAS 39.9 The sentence is formulated as a sufficient condition, not a requirement. Generally reworded. Explanation added. Entirely reworded. CIA The difference between sections 4.2 and 4.3 is that section 4.2 refers 15

16 section 4.2 deals with identifying embedded derivatives. Again, much of this section was difficult to understand. Some examples: (4.3.1 pp.5) If at the outset the expected cash flows can be significantly affected by refunds of premiums reflecting the cost of other benefits, the refund conditions do not constitute conditions that modify cash flows in a way that would otherwise be contractually required. (4.3.3 pp.3) Derivatives can be based on market factors that affect only the value of the derivative without affecting the cash flows by affecting the inherent time value of money. Embedded derivatives are based on market factors directly affecting the cash flows. (PWC) Section This section uses foreign currency embedded derivatives an example. It concludes with if the cash flows are denominated in the functional currency, the contract does not contain an embedded foreign currency derivative. This is a correct statement. However, the section also lists the criteria of IAS 39 AG 33(d) as to whether foreign currency embedded derivatives are closely related or not. These are two different issues and we suggest that they should be analysed separately in the final version: If the cash flows are denominated in the functional currency of the entity, there is no foreign currency embedded derivative. If the cash flows are denominated in the functional currency of a party to the contract (but not the functional currency of the entity), there is an embedded derivative (cash flows are modified), but the derivative is closely related. The reference to AG 33 (d) is appropriate in this case only. (PWC) Section The first paragraph contains a circular argument in relation to the effect of market factors on conditions and cash flows that are not affected by market factors. We suggest deleting from To identify components to the end of the first paragraph. (ASSA) Section : Identification of embedded derivative cash flows The first paragraph is key can a way be found to give it extra only to market factors and alternative investments. Section 4.3 refers to specific identification issues of embedded derivatives, without providing further guidance for the definition of market factors or alternative investments. Considered by complete reference to AG 33 (d). Addressed by rewording that paragraph to clarify the rational behind the introduction of an embedded derivative cash flow. Yes, was reworded. 16

17 prominence? Second paragraph: move (i.e. increasing or decreasing) to after that financial risk, and move (positively or negatively) to after correlated financial risk. (CIA) Section (pp. 2,3) uses the word may in a manner that could be interpreted in two different ways. It could mean either (a) the indicated course of action is permitted but not required, or (b) there are certain conditions (which should be discussed) under which the practitioner would take the indicated course of action. This should be clarified. Was reworded to make understanding easier. CIA use of the word may : The IAA makes extensive use of open language, which is consistent with the fact that the IASP is a Class IV standard. (DAV) 4.3.4, sub-item 1 We recommend to add The behaviour of parties to a contract, e.g., policyholders, can be considered to be a non-financial variable specific to a party and hence not a market factor. (DAV) 4.3.5, 2 nd para The next step In that case it is to be determined whether the overall contract qualifies as a derivative. (ASSA) Section : Identification of the component Move the first sentence of the second paragraph up into the first paragraph. The second sentence will thus start The next step is to determine whether Third paragraph: delete the clause even though this determination is not reached as a result of the definition. (PWC) Section The first two paragraphs on identification of the component seem contradictory. We believe that the step to be taken after identifying the embedded derivative cash flows of a component is to classify (rather than identify) this component. The step in the second paragraph could also be clarified. We believe Done. Reworded. Reworded. Done. PWC 4.3.5: Embedded derivative cash flows are not a component of a contract. The component of the contract to be identified is the contract clause that causes the cash flows to respond to market factors. (the last sentence seemed a restatement of the first) (felt the comment was not required) 17

18 this paragraph should explain that there are situations where there is only one component in the contract and in that case the contract may be a derivative itself if it meets the conditions set out in IAS 39. This would be consistent with the second part of this paragraph where it is explained that in IAS 39 an embedded derivative can only exist if there is a non-derivative host contract. (PWC) Section 4.4 Other than in relation to the functional currency the intention of the entity is not relevant for the assessment of whether an embedded derivative is closely related or not. IAS 39 and IFRS 4 include a number of examples and rules to determine the application of the closely related criteria. The example of IG A.1 of IAS 39 (forward purchase of a commodity) relates to the own use exemption in IAS 39 paragraphs 5 and 6 and is out of context here. 4.4 There are examples in reinsurance contracts, where the contracts formally include derivatives, but there was never an intention to make use of them. The intention and the substance of the contract is to receive the reinsurance coverage.the linkage to market factorsis merely a formality and does not influence the value for both parties.the example in IAS 39.IG A1 is similar. (CIA) 4.4 Separation requirement of IAS The description in the first paragraph of this section implies that the exceptions in IFRS 4.6 are included in the list below, but they are not. (CIA) The second paragraph seems to say that if something is a derivative according to its contractual terms, it might, nevertheless, not be a derivative if that s not the intention of the reporting entity. This could be perceived to expand the application of the substance over form principle beyond what was intended in IAS 39. (CIA) In section (pp 1), the phrase exceeds a related insurance coverage should be clarified to, for example, extends beyond the expiry of a related insurance coverage. Wording reviewed to clarify. Further guidance from IAS 39 was added to clarify. Considered. 18

19 (CIA) Section and imply that accounting guidance for prepayment rights applies to the right of surrender of a contract. Prepayment rights and surrender rights are not the same thing. (CIA) Sections and would benefit from illustrative examples. (CIA) Both sentences in the third paragraph of section are incomplete, requiring (for example) and if not leveraged at the end in order to be correct. Prepayment rights: Prepayment rights are options allowing one party to create cash flows earlier than otherwise due under the contract. The decision to execute that option is based on market prices for cash (time value of money) and other market prices. In some jurisdictions, surrender rights are understood as a right to require the payment of today s value of the future maturity value prematurely, i.e. to require a prepayment. In this respect, they are comparable with surrender rights in the context of derivatives. The difference in comparison to surrender rights is that in other jurisdictions the amount of the surrender value is not seen as a reflection of the future rights given up by surrendering the contract. The difference is in the intention in determining the amount paid at surrender, not in the risk. IAS 39 provides guidance for prepayment rights and only very limited guidance for surrender rights. The main problem is that IAS 39 provides guidance for many examples and it is necessary to find the most relevant example. Here, the optionality of timing of payment is key. To be considered in separate material. Considered. (PWC) Section The concept of an insurance contract measured at fair value is not yet part of IFRS with the exception of business combinations and portfolio transfers involving insurance contracts. We do not agree with a statement that suggests that an entity could develop a fair value for insurance contracts based on IAS 39. If an insurer changes its accounting policies to introduce more current assumptions and mirror fair value measurement concepts the hybrid contract is still not at fair value through profit and loss and the It was clarified that the measurement has to comply with the requirements of IAS 39. However, if an approach is applied that would be acceptable in some situations (e.g., in the case of a business combinations), there is little rationale to object that this is a fair value. That was clarified. 19

20 embedded derivative would have to be separated in accordance with IAS 39 paragraph 11 (c) if it does not meet the two other conditions in paragraph 11. (CIA) The second paragraph of section seems to contradict itself. Even if the right to pay additional premiums might constitute an embedded derivative that could be separated implies that it could be. But the next paragraph says The right to pay additional premiums. is not an embedded derivative that has to be separated. (ASSA) Section 4.4.: Separation requirement of IAS In the first sentence, IFRS 4.6 is the incorrect reference we suspect it should be IFRS 4.8. Again, reference needs to be made after (a) (c) that if the embedded derivative is itself an insurance contract it does not need to be separated out. (ASSA) Section : The principle Change the last sentence of the third paragraph as follows: The embedded derivative is also closely related to the host contract if it can be demonstrated that the embedded derivative is closely related to another component of the contract. Fourth paragraph: it is worth showing some of the examples in the ED (even as an Appendix) since not everybody will have the standards at hand. (ASSA) Section : Index-linked benefits Is the second sentence of the first paragraph really necessary? (DAV) st para IFRS4.6 should be read as IFRS4.8 Such a right is clearly an embedded derivative, but not an embedded derivative to be separated. Corrected. Done. Done. It can be generally assumed that reading IFRS-related IASPs cannot be reasonably done without having the IFRS at hand. ASSA : The second sentence is considered necessary, since there is a common misconception that unit-linked contracts may be viewed as a normal fixed-interest instrument that is modified by a clause that adjusts the outcome in response to a change in a unit value. Corrected. 20

21 (DAV) , 3 rd para We recommend to replace the last sentence of this para by To be closely related to the host contract, it is sufficient to demonstrate that the embedded derivative is closely related to any component of the host contract considering the embedded derivative separated. (DAV) , 1 st para In the example the term Insurance component should be changed to insurance part. (DAV) , 1 st para PROVISION is not meant in the meaning of the glossary, so it should be replaced by amount. (DAV) 4.4.3, 4 th para This para should be worded more precisely Although IFRS 4.34(d) requires full application of IAS 39, the application of fair value IFRS to these embedded derivatives in insurance contracts and investment contracts with discretionary participation features should be given priority treatment also applies (IFRS 4.8 9). (DAV) 4.4.3, 4 th para A new subchapter Options should be added referring to other options than surrender options. (CIA) 4.5 Measurement issues and 4.6 Disclosure issues It appears that the term host contract is confused with hybrid contract in a few places. This should be clarified. (CIA) In section 4.6, the first sentence of the third paragraphs is not consistent with IFRS 4.39(e), in that the clause through profit or loss should not appear. Sentence was reworded to clarify that issue. Considered by changing the wording in general. Done. Wording was simplified. That should be done in examples, since no special further guidance is available in addition to the general guidance regarding counterparty s behavior. Done. Done, however it is impossible that an embedded derivative within an insurance contract is measured at fair value without being through profit or loss. IFRS 4 does not allow reporting anything without movements through profit or loss except in the case of shadow accounting. 21

22 (ASSA) Section 4.6 : Disclosure issues The ED needs to state that if there is an embedded derivative in a host insurance contract that is not closely related and if the insurance contract is measured at amortised cost, then the fair value of the embedded derivative is required to be disclosed. An example is IFRS 4 IG4 example 2.8. Also, if a host financial instrument is measured at amortised cost, the fair value of an investment embedded derivative must be disclosed if it is not fair valued in the underlying reserve held. (DAV) We refer to our remarks to PG Actuarial Practice with regard to an internationally clear and precise wording. Quotations should be marked precisely. Especially we would like to have a clear reference to all terms listed in the glossary whenever one of these terms appears in the PGs. It is not sufficient for the ongoing use if a glossary term is only marked in capitals when it appears for the first time in a PG. This might be confusing for readers who just want to look up a special item and need reference to the glossary in this special paragraph. (DAV) Appendix B List of terms defined in the Glossary Derivatives should be added to the list of terms. (FEE) We believe that the paper could be more clearly insurance oriented. The paper appears to focus mainly on financial instruments and IAS 39, and less on insurance contracts. We suggest improving it by adding examples illustrating issues related to derivatives in insurance contracts. Also, the IAA could gather difficulties experienced by its members and analyse them in the guidance. We consider that it would be beneficial to link the examples to the flow chart on page 3 using the flow chart to show how the examples can be treated. (FEE) Based upon our experience, many embedded derivatives once identified in insurance contracts are not required to be separated and measured at fair value because IFRS 4 paragraph 7 applies i.e. the Disclosure of Fair Value: We did not identify any requirement to disclose the fair value of an embedded derivative which is in the scope of IFRS 4 but not in the scope of IAS 32 or IAS 39. Presumably, the comment refers to guidance of ED 5 or it refers to a separation requirement and reporting the embedded derivative at fair value through profit or loss in the balance sheet and income statement, rather than as disclosure. Care was taken to achieve that. Done. That will be done in separate material and in an intended comment area on the website of the IAA. Done. 22

23 embedded derivative is itself an insurance contract. We therefore consider that the flow chart should be reordered and reworded to reflect this. This would improve the speed of assessment. Disclosures of the embedded derivative may still be necessary (IFRS 4 Para 39) but the avoidance of the need for separation and measurement are important relief. (FEE) We noted that the terminology used is sometimes different from IFRSs. Formal guidance like this should use the identical language as in the IFRS. We suggest that, whenever possible, terminology should not deviate from IFRS language and we recommend in particular that new language should not be introduced as it can cause misunderstanding. For example, in the first paragraph paraphrases the wording in IAS 39.9 in a way that introduces ambiguity. IAS 39 uses the words it requires no initial net investment or an initial net investment that is smaller than would be required rather than sufficiently small as not to be material in This may not be appropriate to deal with a deep-in-the-market option, as this would be a derivative even when there is a significant upfront premium. (SA) Our specific suggestions for improvements are: Full restructuring of the document alongside the following lines: Short and concise definition of embedded derivatives, omitting the confusing link to pure derivatives. Done. SA: For understanding embedded derivatives it is considered preferable to first explain derivatives. Further, the definition of derivatives is necessary to decide whether an embedded derivative needs to be separated. Therefore, didactically the first step was to introduce derivatives, then embedded derivatives, and then to explain when to separate the latter and what to do if separation is not required (disclosure requirements). Short introduction into the classification of contracts or parts of contracts insurance contracts, investment contracts and service contracts alongside the guidance set out in IASP 3. Identification of embedded derivatives, explained separately for each of these contracts types, and guidance on which contract part embedded derivatives would be assigned to on unbundling a contract. We do not believe it is necessary to repeat issues already covered in another IASP. To be done in separate material To be done in separate material 23