May 31, Ms. Leslie Seidman, Chairman Financial Accounting Standards Board 401 Merritt 7 P.O. Box 5116 Norwalk, CT

Transcription

1 May 31, 2013 Ms. Leslie Seidman, Chairman Financial Accounting Standards Board 401 Merritt 7 P.O. Box 5116 Norwalk, CT Reference: Accounting for Financial Instruments Dear Ms. Seidman: The Committee on Corporate Reporting ( CCR ) of Financial Executives International ( FEI ) wishes to share their views on the two Proposed Accounting Standards Updates related to Financial Instruments: Credit Losses and Recognition and Measurement of Financial Assets and Financial Liabilities. CCR agrees with the Board that there is a need for improvements in the areas being addressed by these proposals, but we also believe that these can be accomplished through targeted amendments to existing U.S. GAAP rather than through fundamental changes in accounting concepts and principles, which these documents would introduce. We believe that following a more targeted approach will improve the quality and speed of implementation, will substantially lower the costs of implementation and will enable a more timely adoption of the improvements. For ease of reference, our response is divided into three parts: our cover letter, which explains the basis for our recommendation, and Attachments A and B which provide further detailed comments on the proposed ASUs on credit losses and classification and measurement. FEI is a leading international organization of 15,000 members, including Chief Financial Officers, Controllers, Treasurers, Tax Executives and other senior-level financial executives. CCR is a technical committee of FEI, which reviews and responds to research studies, statements, pronouncements, pending legislation, proposals and other documents issued by domestic and international agencies and organizations. This document represents the views of CCR and not necessarily those of FEI individual members.

2 Page 2 Credit Loss ASU CCR s overall view is that recording lifetime expected losses at the date of recognition of financial instruments is not an improvement in financial reporting for investors. Moreover, we believe the significant differences between the FASB s proposal and the IASB s proposal are neither desirable nor sustainable over the long term. That is not to say that CCR is recommending adoption of the IASB proposal our concerns with the operability of that model remain unchanged, notwithstanding subsequent improvements that have been made. CCR believes that the Board can be responsive to the concerns that have been expressed with current GAAP by focusing on an event-oriented recognition approach that is more sensitive to environmental changes that trigger the need to record additional loss provisions. We do not believe that this requires an entirely new model, but rather could be accomplished through a combination of a lower confidence threshold for recognition of a loss and enhanced implementation guidance, including examples, that helps address past problems in practice. We think there are many questions and areas of concern that are still unresolved under the proposed approach. These include but are not limited to: How to combine historical experience, current estimates, and reliable/supportable forecasts in a robust and repeatable fashion throughout the credit cycle. How the concept of expected loss fits into the conceptual framework and the implications of this change for other contingencies, liabilities and asset impairments. How to adapt existing incurred loss estimation models to determine expected losses (e.g., moving beyond one-year probabilities of default). How the model applies to trade receivables. If there is no change in practice, why not? If there is a change, why does it produce better information? How the model applies to debt securities. Why is it necessary to change a wellunderstood and effective model? Why do non-credit related factors impact the measurement of impairment? How auditable is the calculation of expected losses, particularly the determination of how to combine the three disparate elements of the calculation? Given the above, there is concern over the potential for significant diversity in practice and how audit, banking and securities regulators will react to such diversity. We continue to believe that the relative success of any revised approach to be incorporated into U.S. GAAP will be determined by what the regulatory environment interprets, accepts and supports. The process for determining when to recognize a credit loss is inherently judgmental because such events are rarely directly observable until the very end. Accordingly, the problem with the incurred loss model is not with the concept itself, but rather with the extremely rigorous and demanding level of proof that has been required historically to substantiate the amount recorded, which inherently forces

3 Page 3 conservatism in the application of the principle. We believe that if revised guidance were issued that provides additional support to allow companies to record and substantiate the loss amounts in a timely manner the goals of the project can be achieved with less disruption and cost than would otherwise be the case. If the Board decided to pursue that approach, it could develop the guidance with the assistance of experts that have significant practical experience in dealing with these issues. Classification and Measurement ASU It is our understanding that the proposed ASU was not intended to introduce wholesale changes in how traditional loans and plain-vanilla debt securities would be categorized. However, we are aware of many examples, such as traditional loans that fail to meet the amortized cost requirements of the proposal, that expand the use of fair value measurement (FV-OCI or FV-NI). This is one of several unintended results that indicate the proposal exchanges one form of complex guidance, which is well-understood in practice, with different complex guidance for which new interpretive guidance will need to be developed. CCR supports more targeted amendments to current practice to better achieve the project s goals, such as recording the change in value of financial instruments due to changes in a company s own credit standing through OCI (rather than net income) and retaining the fair value option for both assets and liabilities. Cost-Benefit Considerations CCR believes that targeted enhancements to the existing model can satisfactorily address the core issues in these two areas while imposing less complexity and a much lower cost burden on the financial reporting community. If the goals of the projects can otherwise be met, we believe that it is far better to look for ways to augment existing guidance through targeted amendments rather than through the introduction of ground-breaking new principles. Fundamental changes, particularly those as broad as the concept of expected losses or the historical classification and measurement of financial instruments, will create enormous uncertainty and interpretive churn in the financial reporting and user communities. Even though the FASB Staff indicates that many of our existing models and methodologies can be relied upon to comply, we know that we will essentially be starting over in developing frameworks, policies, interpretive guidance, procedures and controls at a cost that will far exceed the benefits, if any, of moving to a completely new approach. As we move through that process, we will inevitably face questions that will be difficult to answer about whether our processes and policies have changed sufficiently and appropriately to comply with the new principles. We expect significant diversity in practice and are particularly concerned about how an auditor will be able to opine on these judgments in the present environment and how the PCAOB will reach a conclusion that, in each case, the auditor had sufficient basis upon which to opine. CCR is of the view that the existing loss estimation methods used in practice today are well-understood and are capable of being adapted to a broader, event-oriented model if appropriately supplemented with more specific guidance. If this approach was adopted, we believe the revised principles could be implemented sooner and with substantially less

4 Page 4 interpretive effort than a complete change in the model s principles. Based on the fact that this approach would augment existing concepts and principles, we believe that it is reasonable to assume that there will likely be significantly less diversity in practice with lower potential for restatements and audit issues, as well as less confusion within the user community. For these reasons, among others, we believe that an incremental approach is worthy of the Board s full consideration as an alternative to the wholesale changes inherent in each of the proposed ASUs. ***** In summary, CCR supports an incremental and targeted approach to changes in the accounting framework for the reasons explained above. We believe that this approach will yield demonstrable benefits compared with models that introduce fundamental new concepts, including those based on variations on an expected loss approach. These benefits include: development of an operational and cost-effective model, capability to require an earlier effective date, streamlined transition provisions, and higher quality implementation with the potential for significantly less diversity in practice. Our detailed comments on specific issues with each of the ASUs can be found in the attachments to this letter. Please feel free to contact Lorraine Malonza at (973) if you would like additional information on any of the issues or recommendations in this letter. Sincerely, Loretta Cangialosi Chair, Committee on Corporate Reporting Financial Executives International cc: Paul Beswick, Chief Accountant, U.S. Securities and Exchange Commission Hans Hoogervorst, Chairman, International Accounting Standards Board

5 Page 5 Accounting for Credit Losses Our detailed comments on the Credit Loss ASU are as follows: Trade receivables We believe the existing impairment framework for trade receivables is effective. The underlying business model, nature of credit risk and credit risk management processes for trade receivables are fundamentally different than other business activities, such as investing and lending. In addition, due to the short-term nature of trade receivables, compliance with the proposed guidance should not result in a significant difference from the current credit loss framework. Accordingly, we believe the costs to comply with the new model significantly outweigh the benefit of achieving a single model for all financial assets. We encourage the Board to exclude trade receivables that are short term in nature or with expected credit losses that are insignificant from the proposed model. Debt securities impairment We believe that the current model for measuring other-than-temporary-impairment ( OTTI ) of debt securities functions well, is conceptually sound, widely accepted by financial statement users, is well understood by both preparers and users, and should continue without major changes. The Board made several refinements to the model during the financial crisis to address transparency of credit risk and consistency in application. When coupled with existing disclosure requirements, we believe the OTTI framework provides users with sufficient risk information related to debt securities and recognizes losses at the appropriate time. Rather than replace the OTTI model, we would prefer modest revisions to the OTTI model and offer the Board the following suggestions: The OTTI model should apply to all financial instruments measured at fair value through OCI The assessment of OTTI should be permitted on a homogenous pool basis as well as on an individual asset basis; OTTI estimates should be recorded through a valuation allowance rather than a basis adjustment; Both favorable and adverse changes in OTTI estimates should be immediately recognized in earnings; and The non-accrual concept should not apply to debt securities (see separate comment below). These suggested modifications will reduce complexity of the model, yield substantially similar impairment results compared to the current model and provide consistency with respect to the recognition of changes in credit loss estimates with the impairment model for loans. Practical expedient for individual instruments measured at fair value through OCI We believe the proposed practical expedient is too narrow in scope and will ultimately require preparers to quantitatively estimate impairment for a significant number of financial assets, even though they may exhibit little or no sign of credit deterioration.

6 Page 6 Many CCR members have investment portfolios that contain a significant amount of securities that are measured at fair value through OCI. Our members are concerned the practical expedient will result in unintended adverse consequences, such as the asymmetrical treatment of similar instruments, increased cost and burden to produce insignificant expected loss estimates and financial results with limited use. For example, consider identical securities purchased on different dates at different prices. One security may be eligible for the practical expedient while the other security may not simply due to a difference in purchase price rather than fundamental credit factors. Likewise, during a rising interest rate environment, high quality financial assets may not be eligible for the practical expedient due to a decline in fair value below amortized cost for non-credit related factors. If the practical expedient is retained, we strongly encourage the Board to remove the requirement that the fair value of an individual financial asset exceed its amortized cost basis. This modification would appropriately exclude financial assets in an unrealized loss position due solely to non-credit related factors. Expected credit loss calculation We have several concerns with the expected credit loss calculation: Consideration of credit quality: We do not believe that a measurement of lifetime credit losses based on the origination or acquisition of a financial asset is appropriate. We believe the recognition of expected credit losses should consider the credit quality of financial assets at each reporting date. Credit quality can be evaluated with commonly used credit quality indicators and portfolio and product characteristics, combined with appropriate loss estimation periods that contemplate expectations regarding current and future economic conditions. As a result, performing assets would not require immediate recognition of a less reliable estimate of expected lifetime credit losses. Wide range of outcomes expected: While the proposal permits significant flexibility, varying degrees of interpretation of the proposed guidance will likely yield wide ranges of possible outcomes for similar assets or groups of assets as preparers attempt to satisfy the objective of a lifetime loss calculation. These varying degrees of interpretation may inhibit comparability among peers and usefulness of financial information. As a natural outcome, the lack of consistency, comparability and usefulness of financial information could potentially drive the establishment of arbitrary rules/guidance that have unintended consequences. Audit support for lifetime loss estimates: The ability to satisfactorily support an expected loss estimate that reflects current expectations of the collectability of contractual cash flows rests on the collective understanding across all constituents involved. This collective understanding has to extend to auditors, the PCAOB and other regulators. While the exposure draft appears to go to great lengths to establish the ability of an entity to utilize existing risk management models, entities may struggle

7 Page 7 reconciling this possibility since many existing loss estimation methodologies are not compatible with an expected lifetime loss model. Implications to conceptual framework: The proposed guidance represents a significant change from existing guidance for contingent liabilities. We do not believe that the recognition of credit losses at period of inception is supported by the existing conceptual framework or that the Board has adequately justified the unique treatment for loss contingencies related to financial assets. Inaccurate picture of earnings performance: Under the proposed model, all expected losses will be recognized immediately while interest income will be recognized prospectively. As a consequence, the earnings performance related to the financing activities of our member institutions will not be meaningful to users. While we acknowledge that information about expected credit losses may be useful on a supplemental basis, we believe users will likely have to adjust historical earnings and other performance indicators to more accurately assess business trends and periodic performance. Time value of money We do not support the inclusion of a time value of money ( TVM ) principle in the proposed credit loss model. We are concerned that the inclusion of the TVM principle will result in interpretive risk to preparers without providing real practical benefit to financial statement users. Accordingly, we believe the FASB should characterize current credit loss estimation methods as acceptable alternatives to approximate the present value under a discounted cash flow model, rather than as methods that implicitly reflect the time value of money. Since the inclusion of TVM makes several assumptions that may not be present in current credit loss estimation processes, we are concerned that without this clarification, auditors will require preparers to substantiate the assertion of the Board that current methodologies implicitly reflect TVM. Presentation and Disclosure We agree that the financial statements should include information that enables users to understand an entity s exposure to credit risk and how that risk is managed. We believe the most efficient way to provide investors with relevant and reliable information is to base disclosures on information used by management to manage credit risk that is tailored to the nature of the portfolios of debt instruments, including distinguishing between loans and debt securities. In some cases, the proposed incremental disclosure information will not be readily available from existing systems and will require significant cost and effort to build systems to provide data that has limited benefit to the users of the financial statements. Roll forwards of debt instruments measured at amortized cost and fair value through OCI are examples of burdensome disclosures with little added value. These disclosures provide limited duplicative information as to volumes of new loans and prepayments. Most of this information is already available in the cash flow statement and currently required

8 Page 8 disclosures that present significant sales, purchases or reclassifications of financing receivables. A recent study 1 on the volume and complexity of disclosures indicated that the sheer quantity of financial disclosures has become so excessive that the overall value of disclosures has been diminished. We believe the Board should postpone adding creditrelated disclosures until it more fully considers the findings of the Disclosure Framework project and an overall cost-benefit analysis. This will enable the Board to recommend disclosures that balance the cost of the disclosures against their perceived benefits. We also note that the proposed guidance requires structuring the financial statements in alignment with the proposed classification categories which represents a fundamental change in historical presentation. We believe that the usefulness of the financial statements may not be enhanced by the proposed structure, particularly for non-financial institutions. Should the Board retain the proposed classification and measurement approach, we believe that the issuer should have the discretion to determine the financial statement presentation that best conveys decision useful information to the user. Other matters: Non-accrual, Troubled Debt Restructurings ( TDR s), Acquired loans It is our understanding the Board intended to codify existing non-accrual concepts currently contained in regulatory literature. We believe this will have unintended consequences as the Board is using different words in an attempt to obtain a similar result. Further, the non-accrual concept is not appropriate when the impairment measurement represents the cash flows expected to be collected discounted at the effective interest rate. Lastly, we do not believe the application of a non-accrual concept is warranted for instruments measured at fair value through OCI as the practical result is the reclassification of interest income from retained earnings to OCI. None of these outcomes results in an improvement to financial reporting. CCR recommends eliminating the TDR designation. We believe providing disclosures for significant loan modifications (rather than a specific loan group) will simplify processing of loan modifications, provide more useful information to investors and further align the disclosures with IASB disclosures. We support the proposed revisions to the model for Purchased Credit Impaired ( PCI ) loans. However, we believe the Board should retain the existing OTTI model, along with our suggested revisions, for debt securities. We also believe the significant improvements to the proposed PCI loan model, should be applied to all loan acquisitions, whether acquired individually or in a business combination. The historically different models for PCI loans and non-pci loans inhibit comparability and have been a source of confusion for analysts and investors. The application of the proposed PCI accounting model to all loan purchases would eliminate this confusion and provide a uniform accounting framework for all loans study published by KPMG, Disclosure Overload and Complexity, Hidden in Plain Sight

9 Page 9 Recognition and Measurement of Financial Assets and Liabilities As discussed in our cover letter, CCR understands that the proposed ASU was not intended to introduce wholesale changes in the classification and measurement of traditional loans and plain-vanilla debt securities. However we believe the guidance in the proposal could have significant unintended consequences, including instances where loans and debt securities with common features would be classified as fair value through net income. We believe that further refinement of this proposal is unlikely to resolve this phenomenon and will inevitably result in significant interpretive and potentially standard setting activity post issuance. CCR therefore recommends that the Board take a more pragmatic approach by targeting specific areas within existing GAAP that are in need of revision and retaining the guidance where concepts are already well-understood and tested by auditors, regulators and the PCAOB. We believe that the prospects for a successful standard that improves financial reporting are greater if the Board avoids introducing complex new concepts that require significant implementation guidance and could result in unintended consequences and additional, avoidable implementation costs. CCR s recommendations in response to the proposed ASU are as follows: Guidance on embedded derivatives CCR recommends retaining the current guidance requiring bifurcation of embedded derivatives that are not clearly and closely related to the host contract for financial assets. We are concerned the proposed solely payments of principal and interest ( SPPI ) framework for evaluating the cash flow characteristics of financial instruments is just as complex as the existing framework. While we appreciate that this is an effort to converge with international accounting standards, we think that it would be more effective to pursue improvements to the existing framework. While we acknowledge the existing framework is complex, it is well understood and implementation guidance exists that has been vetted over several years to drive consistency of application and comparability of results. We do not believe the improvement in financial reporting will be achieved by substituting existing complex practices with the equally complex and new SPPI concept. The SPPI test will likely require the same level of effort to interpret and develop implementation guidance as has already been invested in practice under the current guidance. If the desired end-result is intended to produce very limited changes in how traditional lending products and plain-vanilla debt securities are reported, it appears to provide little or no improvement to financial reporting and the potential for significant unintended consequences. Based on a preliminary review of the requirements, it appears several instruments accounted for at amortized cost and FV-OCI under today s requirements will fail the SPPI test because of common features (e.g. prepayments, interest rate resets and other than senior structured securities), resulting in fair value through net income measurement. We find this result to be troubling, particularly when it is our understanding the Board did not intend wholesale changes to the accounting for traditional loans and debt securities. While it may be possible to modify the principles to

10 Page 10 reduce the incidence of these unintended consequences, we recommend that the Board retain existing GAAP in this area and focus its efforts on improving those areas believed to be in need of different guidance or clarifying principles where there is diversity in practice. Loan and debt security accounting under the business model principle We recommend retaining the current practice of separate accounting models for loans and debt securities. As one considers the broad range of instruments affected by this principle, at one end of the spectrum the only difference between a security and a loan may be the legal form, while at the other end of the spectrum there are differences that are substantive. The business model framework must acknowledge these substantive differences that can exist in the way in which loans and debt securities are managed due to differences in information availability, existence of active markets, and the ability to actively manage credit risk or other unforeseen circumstances that did not exist at the origination/acquisition date. CCR believes these substantive differences are creating difficulties developing a business model framework that can be equally applied to loans and debt securities. For example, we are concerned the business model parameters for the hold to collect model are too restrictive and result in outcomes that are counterintuitive. In this regard, we note the proposal would suggest that selling assets after credit deterioration is consistent with a hold to collect model while selling those same assets to manage credit risk prior to deterioration is not. The business model should facilitate the reporting of an entity s activities, not dictate how an entity manages its activities. It appears that the restrictive nature of the parameters in the hold to collect model are similar to the restrictive criteria for the held to maturity securities under today s model for debt securities, which is rarely used in practice. It appears the difficulty in developing the framework is a direct result of the tension that exists between loans and debt securities. To address concerns of amortized cost classification for debt securities, various restrictions need to be contemplated. However, these restrictions inherently fail to accommodate normal and prudent risk or business management behavior thus resulting in the potential for classification that does not reflect economic reality. Therefore, we believe the best solution is to retain separate accounting for loans and debt securities allowing the Board to focus on developing a business model framework that properly addresses debt securities. In developing that framework we would encourage the Board to think beyond the current held to maturity guidance as a starting point for the hold to collect model. While we have pointed out the counterintuitive results of the credit example here, we would highlight there are other instances in which hold to collect may be appropriate (e.g. rebalancing the book to fund liabilities). Equity method investments held for sale criteria CCR recommends providing further clarification regarding the held for sale criteria for equity method investments. We are concerned the interpretation of the criteria could potentially be too broad, resulting in long-term strategic ventures being accounted for at fair value, simply because it is common practice to have a defined exit strategy. This could have a pervasive and significant impact on practice as the formation of joint

11 Page 11 ventures is the primary means for expanding business activities in some geographic regions. It appears this guidance was developed to address the removal of the fair value option for equity method investments. However, the consequence of this proposed change in the guidance is to add to the burden of proof for preparers to satisfy auditors and regulators that there is no exit strategy for the investment and, therefore, substantiate that normal equity method accounting is warranted. OTTI for equity method investments As we have indicated above for other types of instruments, we recommend retaining the current guidance for evaluation of OTTI in equity method investments. Equity method investments are accounted for based upon methodologies rooted in consolidation theory and the concept of significant influence. CCR believes that an OTTI assessment is the appropriate way to evaluate such an investment given that the investor, through its involvement and influence over the investee, generally can look through to and adequately assess the underlying operations of the business to determine when an OTTI is warranted. In this regard, the methodologies and thought processes for the impairment evaluation are much closer to those applied to long-lived assets. In addition, we note that the equity method produces changes in the carrying amount each reporting period for earnings, dividends and the investor s proportionate share of OCI items. As a result, we believe that applying a passive investor approach to impairment testing yields flawed results. Specifically, there are adjustments to the carrying value of an equity method investment (e.g. for the investor s share of OCI related to cash flow hedges) that have nothing to do with the change in the entity s fair value that could potentially increase the carrying value of the investment and trigger an impairment. Consequently, in some instances, the results of applying a one-step impairment test under the model will not reflect economic reality. Equity investments CCR recommends a classification option for equity securities similar to today s available for sale. Many of our members invest in strategic long-term equity investments in sectors of the economy to stimulate growth, create new business opportunities, and expand global markets for their products. We do not believe that recognizing unrealized gains and losses for these strategic investments through net income will lead to reporting results that are representative of ongoing central operations. There are several instances our members can point to whereby significant gains or losses have reversed over the time these strategic investments have been held. Keeping this volatility restricted to the Statement of Comprehensive Income preserves the value of the Statement of Income for our users by presenting the measurement of operating performance and not the periodic impact of the economic volatility from financial instruments. The fair value option We acknowledge that providing unlimited options in accounting standards does not always promote consistency or comparability. However, we believe that retention of the fair value option to alleviate accounting mismatches, appropriately apply risk management

12 Page 12 strategies and reduce complexity is warranted given our mixed measurement attribute model. Accounting for fair value changes attributable to an entity s own credit It is our understanding that when liabilities are measured at fair value, most users disregard the portion attributable to an entity s own credit as this is not viewed as a true economic event. We agree with this view and recommend that the Board amend existing guidance to record the change in fair value attributable to an entity s own credit to OCI. We believe this change will represent a meaningful improvement to financial reporting and is consistent with how users analyze financial results.