AMERICAN ACCOUNTING ASSOCIATION 1997 Annual Meeting Dallas, Texas August 18, 1997 DANGEROUS IDEAS: A Sequel Remarks by Michael H. Sutton Chief Accountant Office of the Chief Accountant United States Securities and Exchange Commission Washington, DC __________________________________ The Securities and Exchange Commission, as a matter of policy, disclaims responsibility for any private publication or statement by any of its employees. The views expressed herein are those of Mr. Sutton and do not necessarily reflect the views of the Commission or the other members of the staff of the Commission. DANGEROUS IDEAS: A Sequel Introduction Good morning. It is a pleasure to be here with you again and to share some thoughts on accounting and financial reporting. Last December at the AICPA's annual SEC conference in Washington, I spoke about some of the ideas -- or proposals -- for accounting and financial reporting that we encounter frequently in practice and often find troublesome. These are ideas that, on the surface, seem reasonable in some contexts, but that, on closer inspection, can pose perils to the credibility of financial reporting. Some are advanced so frequently that they tend to take on an aura of plausibility, and only when they are examined more closely are flaws revealed. I described those ideas as "dangerous ideas." It was suggested to me that this general subject would be of interest to this audience, so what I plan to do this morning is to explore again and elaborate on a few of these notions. The ideas I would like to comment on today are as follows: * The matching principle should shape the financial statements. * The balance sheet fails to recognize important intangible assets. * Some intangible assets have infinite lives. * Accounting should reflect the intent of business decisions. * Disclosure can be an effective substitute for recognition in the financial statements. The Matching Principle One of the most common ideas that we encounter is that the matching principle should take precedence in shaping the financial statements. A corollary idea, and one also frequently advanced, is that current period losses can create assets. At first blush, those ideas can be understood. After all, why would a company incur a cost if a future benefit is not expected? A closer look, however, reveals that both of those notions, when and if implemented, have the potential to undermine the integrity of the balance sheet and, therefore, the credibility of a company's financial reporting. It is hard for investors to understand why things like deferred advertising costs or customer acquisition costs or start-up costs represent assets. When a registrant adopts an accounting policy -- based on the matching theory -- that results in the deferral of costs normally associated with ongoing operations, concerns about the quality of the asset created and, therefore, the quality of earnings follow. To appreciate some of the pitfalls, let's look for a minute at an industry that uses specialized accounting practices that rely heavily on matching theory -- the public utility industry. Under FASB Statement 71, Accounting for the Effects of Certain Types of Regulation, which addresses the accounting for rate-regulated industries, many costs that would be expensed by non-regulated enterprises are capitalized and carried as "regulatory assets". This practice -- described in Statement 71 as reflecting "the economic effects of the rate-making process" -- is an elevation of the matching principle to recognize a specific circumstance in which future revenues can be attributed to current period costs and existing legislation provides a strong expectation of recovery of those costs. Again, it is a practice that permits rate-regulated utilities to defer and capitalize costs that other companies, in other industries, are required to expense as incurred. Thus, if the assurance of recovery that leads to a different accounting model for utilities -- and on which investors rely -- is not realized, and unexpected losses accrue, the credibility of financial reporting and investor confidence will be eroded. Statement 71 made more sense in the era in which it was drafted than it does today. Fifteen years ago, it could be argued that there was little likelihood that the industries affected would be deregulated. Today, the regulatory promise of specific future revenues to recover current costs seems much less assured. Outside of rate-regulated industries, usually there is no direct and traceable link between current period expenses and a specific future period revenue -- even though it may be possible to associate, at least in part, current period activities with future revenues. But, the arguments for cost deferrals and the recognition of questionable assets continue. Experience shows, however -- time and time again -- that companies that select accounting policies that postpone recognizing expenses lose credibility. They frequently come under pressure to change their accounting policies, and they find that explaining why the new policy is preferable -- but that the old policy was acceptable too -- is a difficult task that strains investor confidence. As basic as the deferred cost issue is, it is one that comes up often, and I encourage all of us to look with skepticism at proposals that elevate the matching idea to the point that it overrides the definition of an asset. Intangible Assets Let me turn to a question that is receiving a lot of attention these days -- how should critical information about intangible assets be captured in the financial statements? Underlying that question is an implicit suggestion that, if the balance sheet fails to include the cost of developing important intangible assets, assets that in today's economic environment sometimes are far more important than tangible assets, financial reporting may lose its relevance and usefulness. It is true that, historically, accounting has been strongly influenced by the reporting needs of a manufacturing-based economy. It also is true that, as the US economy has become more service oriented, some high technology companies now have market capitalization equal to or greater than some industrial giants. That reality has led to questions about how well our accounting model is working. For example, if Microsoft's common stock has a market value of 11 times its net book value, while Exxon's common stock has a market value to book value ratio of three to one, doesn't that suggest that something is missing from Microsoft's balance sheet? Academic research should help answer questions like that, and I am aware that substantial work has been done in the past and substantial additional work is underway. But, even if evidence shows that, in fact, the market is trying to derive information that is missing from the financial reporting model today -- about the values of intangibles such as master copies of computer software, or the values of new products under development, or even the value of a talented work force -- does it follow that the way to address that information need is to capitalize some -- or even all -- of the costs incurred in developing those intangibles? I am troubled by the suggestion that the way to improve financial reporting for intangibles is to capitalize development costs. Recently, the International Accounting Standards Committee, IASC, proposed an approach to accounting for intangible assets under which internal development costs would be expensed up to some point in the process and then capitalized after certain criteria are met. The objective of the proposal is to require recognition in the balance sheet of the cost of developing an intangible asset if, and only if, it is probable that future economic benefits will flow to the enterprise and the cost of the asset can be measured reliably. The SEC staff has challenged this proposal on two grounds. First, the reliability of accounting produced by such a model is questionable -- it is based on inherently judgmental capitalization criteria that are not likely to be applied comparably by different entities or consistently across time. History has shown us that in capital markets that are -- or are believed to be -- earnings driven, cost capitalization accounting models are very difficult to apply credibly, and that the risk of earnings management and accounting abuse is high. For example, experience with the application of similar accounting models in the US -- FASB Statement 86, Accounting for the Costs of Computer Software to Be Sold, Leased or Otherwise Marketed, and AICPA Statement of Position 93-7, Reporting on Advertising Costs, has demonstrated that different enterprises make very similar "line calls" with very different results. As a result, under the IASC's proposed model, it should be expected that some enterprises would seek to capitalize internal development costs -- some in large amounts -- and other enterprises would seek to avoid capitalizing those costs by failing to meet the specified criteria. Second, the staff's concern does not relate solely to questions about the reliability of the accounting produced by the proposed model. We also have fundamental reservations about the relevance and usefulness of recording -- in the balance sheet -- an amount that can only be described as an accumulation of some, but not all, of the costs of certain internally developed intangibles. It would seem to be unlikely that the reported amounts would convey to investors any meaningful information about the value of the intangibles, the efficiency or productivity of the company developing the intangibles, or about how the intangibles might contribute to or otherwise impact future cash flows. I do believe that investors need better information about how companies invest in and benefit from activities like research and development, the efficiency and productivity of those investments, and how those investments can be expected to impact future cash flows. As I said earlier, we must question whether a cost capitalization accounting model can convey relevant or reliable information that will help answer those questions. Infinite Useful Lives A related idea about intangible assets that often is put forward is that some of those assets, such as brand names and goodwill, can have infinite lives and, therefore, need not be depreciated, so long as they are subjected to regular and rigorous tests for impairment. I have two thoughts about that suggestion. First, under our historical cost accounting model, amortization is a necessary consequence of the decision to recognize intangibles as long-lived assets. Under our model, the initial cost of long-lived assets is allocated (amortized) over their estimated useful lives. That principle applies to intangible as well as tangible assets. Second, it would be very difficult to devise and apply an impairment test that could be applied with objectivity and consistency in the absence of a systematic allocation process. Under such an "indefinite useful life with an impairment test" approach, we could expect inconsistent, unexpected patterns of write-downs. As you may recall, this approach essentially was the accounting model that preceded APB Opinion 16, Business Combinations, and APB Opinion 17, Intangible Assets, and it didn't work very well. Likewise, it seems to me that there should be some reasonable limit on the allocation period. Even assets that remain in use for many years -- even the most valuable brand names -- require significant ongoing support to maintain their value, and the distinction between value acquired and value created by subsequent ongoing maintenance blurs quickly. For example, I doubt that, when trying to decide whether the original cost of a brand name has been impaired, it would be feasible to separate the value -- and the additional life -- created by current expenditures for advertising and other marketing activities from the value and life that existed at the time the brand name was originally acquired. Certainly, these are not new problems, and finding answers will not be easy. Still, I think we have to be careful to avoid repeating old mistakes. Accounting and Business Decisions We often hear that business decisions are driven by accounting and that accounting should reflect the intent of business decisions. This idea takes a variety of forms, including suggestions that, if a particular accounting treatment is not allowed, it will not be possible to consummate a particular transaction. Trying to develop accounting rules based on a prediction about what intent is, or how business decisions might be made, may be the financial reporting equivalent of "the tail wagging the dog." While, as accountants, we should take pride in the important role that financial reporting plays in business and in our capital markets, we should remember that accounting is -- and should be -- first and foremost, a vehicle for effective communication. Accounting should report the facts in the most relevant and reliable way possible, and it should be unbiased. As an example, when the staff is engaged in discussions about the applicability of pooling of interests accounting, or other forms of carry-over basis accounting, we often hear that the preferred accounting is critical to the marketability of the stock being sold and, thus, accepting that accounting would encourage capital formation. One submission pleaded that we should "rise above the technical merits" of the accounting analysis and not object to pooling of interests accounting so that the company's shareholders could realize greater share values. Those arguments, however, have another side. What may be good for selling shareholders in a particular circumstance may not be good for the buying shareholders -- the new investors and shareholders on the other side of the transaction. In other words, a good price for a seller may be a bad price for a buyer. Almost daily, we encounter other requests to accept or encourage an accounting treatment that will not deter some form of business or capital market activity that is believed to be highly desirable. We also hear these suggestions in the context of the accounting standard-setting process, and we are hearing today that the FASB's proposed accounting for derivatives will cause companies to forego available risk management tools and, therefore, expose themselves to greater market risk. When I hear this argument, I am reminded of the Board's project on accounting for health care benefits a few years ago that led to Statement 112, Employers' Accounting for Postemployment Benefits. In that project, some affected interests predicted that the FASB's new accounting would cause companies to abandon their health care plans and cause countless citizens to be without much needed benefits. That didn't happen, and I think with hindsight, most now agree that -- to the contrary -- improvements in the accounting brought needed attention to the health care cost issue. With every important standard-setting project, some will object strongly to change and predict dire consequences if change is made. I see nothing in the Board's derivatives accounting proposal that supports the argument that the new accounting would prevent companies from engaging in the risk management activities they deem appropriate. The new model would, however, provide investors with better information about those strategies and their impacts on the financial statements. We all should see "red flags" flying when we are told that the accounting will make or break a deal -- or create or destroy a particular market. While accounting can be an important factor in some decisions, we all should recognize that accounting that masks, or fails to capture, meaningful information for the benefit of all investors is not sound. We all should share the goal of enhancing the ability of accounting to express the story of an enterprise's financial performance in a meaningful, accurate and consistent way. Disclosure vs. Accounting Another idea that can be hazardous to sound financial reporting is the suggestion that supplemental disclosure can cure bad accounting. Without question, good disclosure is an essential component of effective financial reporting. It's the flesh on the bones of the balance sheet and income statement -- it's the words that make the numbers in the financial statements become three dimensional. But disclosures are, in point of fact, the modifiers of the accounting. Good disclosure doesn't cure bad accounting any more than an adjective or adverb can be used without -- or in place of -- a noun or a verb. In Accounting Series Release 4, the Commission recognized the importance of this issue when it concluded that no amount of supplemental disclosure can justify the use of unacceptable accounting principles. Thus, for example, cash basis accounting for cost of goods sold would be viewed as misleading, even if accrual basis amounts were disclosed fully in the footnotes. While using cash basis accounting for cost of goods sold may seem to be an extreme example, there have been a number of cases in which footnote disclosure has been urged as a remedy for something close to cash basis accounting. Remember, for example, that when the FASB proposed using accrual accounting for postretirement benefits, some urged that the pay-as-you-go approach be maintained and that footnote disclosures be improved. In Statement 106, Employer's Accounting for Postretirement Benefits Other Than Pensions, the Board concluded that disclosure is no substitute for recognition and that "the usefulness and integrity of financial statements are impaired by each omission of an element that qualifies for recognition." Again, this same issue has been raised in the context of the Board's derivatives and hedging project. Some have suggested that good disclosure, such as the information in the Commission's new market risks disclosure rules, overcomes the need to record and measure those instruments in the basic financial statements. The SEC's market risk disclosure rules and the FASB's proposed derivatives accounting standard, however, have different goals and address different investor needs. The SEC's rules are designed to provide investors with qualitative and quantitative information about the market risks a company faces and about how the company goes about managing those risks. The information is forward-looking in nature and seeks to provide investors with insights about what might happen in future market risk scenarios. The FASB's proposed standard addresses how derivatives should be accounted for -- how they should be recognized and measured in the financial statements. The proposal addresses how past transactions should be portrayed to investors in the body of the financial statements. Investors need and value an accurate accounting about past activity in making judgments about how a company will perform in the future. In one sense, the financial statements are a report card on what an enterprise has done. To say that forward-looking information is a substitute for credible accounting is like saying that a projection of next year's performance is a substitute for an accurate report on last year's performance. Accounting and financial reporting must keep up with the realities of the market place. The use of derivatives in a variety of roles, including risk management, has grown exponentially in the last 10 years, and we can expect that it will continue to grow. Investors need and deserve better accounting for those activities. The reality is that the current accounting model for derivatives -- which in many instances means no accounting at all -- just won't do, and again, supplemental disclosure is no substitute for unacceptable accounting. Conclusion I will conclude with just one thought. Sometimes, in financial reporting as in life, the biggest challenge posed by "dangerous ideas" is recognizing them, especially when they are presented in a familiar environment and have seemingly worthwhile goals. Thank you for your attention.