Ending the Accounting-for-Intangibles Status Quo

Abstract

The surge of corporate intangible investments is the hallmark of developed economies, radically transforming the business models, strategies, and performance of business enterprises. Accounting standard-setters, however, by and large, are oblivious to this world-wide development. I establish in this study that this accounting resistance to change seriously harms investors and the economy-at-large, and accordingly I propose feasible remedial changes to the accounting system to adapt it to economic reality. I discuss implementation issues of the proposed change, and the reasons for the three-decade resistance of accounting standard-setters to change the accounting of intangibles. Finally, in order to facilitate the accounting change, I outline a wide-ranging, policy-oriented research agenda on intangibles and related issues.

Keywords:

• Intangibles
• Accounting for
• Reporting of
• Accounting change
• Research agenda

Acknowledgments

Thanks to Feng Gu and Xi Wu for their assistance to this study.

Notes

1 An exception is the development costs of software for sale which should be capitalized (FASB, 1985), yet most major software developers ignore the standard and immediately expense this investment (Aboody and Lev (1998): ‘We find that software capitalization is value-relevant to investors … . [and] associated with subsequent reported earnings.’ Abstract).

2 The FASB’s, 1974 reasoning for the R&D expensing include odd statements like: ‘A direct relationship between research and development costs and specific future revenue has not been demonstrated … ’ (Section 41), which is contradicted by extensive research, such as Hall (2011). This alleged lack of R&D relation to future benefits will surely come as a surprise to, say, Pfizer executives who spend $8 billion annually on R&D, apparently unaware that there is no relationship between this expenditure and future revenues. Yet another bizarre FASB statement: ‘ … at the time most research and development costs are incurred the future benefits are at best uncertain.’ (Section 45). Are the future benefits of the capitalized acquired R&D in-process, or, in fact, the benefits of any other investment, certain? Such logic underlies the accounting expensing of intangibles.

3 Originally from Corrado and Hulten (2010), and updated by the authors to 2016 at my request.

4 Other developed economies underwent similar transformations, see Corrado, Haskel, Jona-Lasinio, and Iommi (2013).

5 Both the assets and earnings of intangibles-acquiring firms will generally be higher than those of firms that self-produce the intangibles, particularly when the intangibles growth rate is positive. However, the ROA and ROE of internally-generating (expensing) firms will generally be higher due to the missing intangible capital from the denominator of the ratios. A highly confusing situation for investors and analysts.

6 Some firms provide certain information on intangibles in conference calls or sustainability reports, but such disclosures are not uniform across firms and often inconsistent over time, thereby detracting from their usefulness to investors.

7 See, for example, Skinner (2008a) defense of the accounting status quo: ‘I argue that the case for [intangibles] reform is surprisingly weak, and does not support claims that large-scale reforms are necessary.’ (p. 191) (see ensuing discussion in Elwin, 2008; Lev, 2008; Skinner, 2008b).

8 Years ago, I was asked by a senior U.S. Senator about the issues surrounding intangibles. When I described the inadequate accounting treatment of intangibles, he reacted: ‘Where is the harm? Who is damaged by the outdated accounting? Investors? Pension funds? U.S. competitiveness? Unless you point at specific societal harms, I am not interested.’ concluded the Senator. This emphasis on specifically documenting societal consequences, in contrast to just discussing research findings (as common in, e.g., earnings management, or the accounting anomalies literatures), was an important lesson for both my own research, and that of my Ph.D. students.

9 The FASB clearly states that investors are the prime intended users of financial information: ‘The objective of general purpose financial reports is to provide financial information about the reporting entity that is useful to existing and potential investors, lenders, … ’ (FASB Statement of Financial Accounting Concepts No. 8, OBI, 2010).

10 Thanks to Feng Gu for performing this analysis for me.

11 Early discussion of these issues can be found in Lev and Zarowin (1999).

12 The returns were calculated from investing, in each quarter, in the firms that will subsequently meet or beat the consensus, two months prior to quarter-end, and liquidating the investment after the earnings announcement. These size-adjusted gains are averaged and presented in Figure 3. An early version of this figure was published in Gu and Lev (2017, p. 27).

13 See also Srivastava (2014) for the fast deterioration in the traditional returns-earnings relation (change regressions) due to increasing intangibles intensity. Barth, Li, and McClure (2017), using an unconventional statistical methodology, report that the relation between a set of financial report variables and market values didn’t decrease over time. (Interestingly, the variables contributing most to the constant relation are internally-generated and acquired intangibles). But even with their methodology, Barth et al. (2017) document a significant deterioration of the relation between market values and earnings, the focus of investors’ analysis. I am not familiar with a study that doesn’t show an earnings-market deteriorating relation.

14 One could argue that it’s not the expensing of intangibles that damages earnings, rather the fact that intangibles-intensive firms are characterized by high risk, which, in turn, adversely affects earnings measurement. To address this concern I regressed the gains from perfectly predicting consensus meets and beats (Figure 3) on the firms’ intangibles intensity and sales volatility (a firm risk measure). Results indicate that the coefficient on intangibles intensity is negative and highly significant, whereas the coefficient of sales volatility is positive but insignificant.

15 Notably, accounting standards-setters, both the FASB and IASB, abandoned in the 1980s the traditional income statement model, which focuses on income measurement and the revenue-cost matching, for the balance sheet model, emphasizing the measurement at fair values of assets and liabilities. For elaboration on this, misguided, in my opinion, regulatory shift, and the evidence on its adverse consequences on the usefulness of financial reports, see Lev (2018).

16 The revenue-cost association is measured by the coefficient on current costs from annual, cross-sectional regressions of current annual revenues on current, previous, and next year’s annual costs.

17 There is a wide-spread but wrong belief that the immediate expensing of intangibles is conservative. In accounting, what’s conservative now, will become aggressive later. Earnings are understated (conservative) by intangibles expensing as long as the firm’s rate of growth of investment in intangibles is positive. When this rate turns negative, which is bound to happen sooner or later to every firm, reported earnings will be overstated: current revenues are recorded without the cost of the resources used to generate the revenues. For comprehensive analysis, see Lev et al. (2005).

18 The proliferation of one-time items in income statements (from assets and liability adjustments to fair values, goodwill impairment, restructuring costs, etc.) also adversely affects the revenue-cost matching, but as shown by Dichev and Tang (2008), this effect is minor.

19 We exclude from the figure the two hard-to-predict financial crisis years, 2008–2009.

20 Specifically, firm j’s earnings in year t+1, are regressed on its market value and earnings in year t. The regression is run across all firms in a given year, classified by high and low (above and below industry median) intangibles-intensity. The estimated coefficient on year t market value reflects share price informativeness, and is portrayed in Figure 6.

21 Relatedly, Peters and Taylor (2017) show that by including capitalized intangibles in the denominator of Tobin’s q ratio, the performance of this measure in terms of explaining and predicting firms’ investment improves significantly. The authors make the intangibles-adjusted values available on (https://wrds-web.wharton.penn.edu/wrds/queryforms/navigation.cfm?navld=421).

22 It is sometimes argued that R&D capitalization doesn’t really matter because investors can capitalize on their own the R&D expensed in the income statement. This is a misconception. Only managers have access to information about technological feasibility test results and target market conditions underlying firms’ decisions to capitalize development costs.

23 In the 1980s, for example, several drug and biotech firms tried to issue shares on their R&D and patent portfolios. These attempts were largely unsuccessful, mainly due to ‘adverse selection’ concerns (firms will sell shares on worthless patents).

24 See Haskel and Westlake (2018) for elaboration on the unique attributes of intangibles and their consequences. Legislation can reduce investors’ risk, like the one which directed the U.S. Patent Office in 2001 to disclose the entire patent document 18 months after patent application. See Lev and Zhu (2018) for investors’ risk reduction due to these early disclosures.

25 Storey and Storey (1998, p. 71).

26 On standard-setters’ unfortunate, in my opinion, switch from the income statement to the balance sheet model, and the serious adverse consequences of this regulatory shift on the quality of reported earnings, see Dichev (2017) and Lev (2018).

27 Certain intangibles, such as in-house employee training, or the development of organizational capital (internal systems and business processes) will often be excluded from capitalization by this condition because firms generally don’t track the costs involved in their creation

28 An example of amortization rates of acquired intangibles, from Cisco Systems’ 2016 report: Technology: 5 years, customer lists: 6-7 years, In-process R&D: 10–11 years.

29 For example, General Electric’s 2016 balance sheet included $25 billion of ‘Contract Assets,’ which are based on the costs and projected revenues of its long term projects, capitalized accordingly to the end-of-year percent completion of the projects.

30 The total annual investment in intangibles in the U.S. was estimated by Corrado and Hulten to have exceeded $2 trillion in 2016. Of this, total R&D amounts to about $350 million. The development part of R&D (distinct from the initial research) is perhaps half this sum. So, if IFRS were to apply in the U.S., and all firms followed the capitalization rule, less than 10% of the total intangible investment would have been capitalized. In most countries following IFRS, R&D is substantially lower than in the U.S., and not all firms follow IFRS capitalization rule.

31 This likely is the major reason for the dearth of substantive accounting change proposals made by academics, in sharp contrast to economic and legal scholars who often advance proposals to change legal and institutional settings.

32 See also, De Waegenaere, Sansing, and Wielhouwer (2017) on the merits of development cost capitalization. However, Cazavan-Jeny and Jeanjean (2006), using a sample of French firms over 1993–2002, reported that capitalized R&D was negatively associated with stock prices and returns.

33 There are quite frequent trades in patents and other intellectual property (brands, copyrights), but details of such deals are generally not disclosed.

34 The only exception are enterprises in a steady-state (no growth) of investment in intangibles. For these enterprises, the periodic investment in intangibles equals the intangibles’ amortization. So, immediate expensing of intangibles will be similar to their capitalization and amortization. But very few firms, if any, are in such a steady state for a protracted period.

35 Economists Corrado, Hulten, and Sichel (2009, p. 680) wrote: ‘What is surprising is that intangibles have been ignored for so long, and they continue to be ignored in financial accounting practice at the firm level.’

36 Khan et al. (2018) report that the R&D expensing standard was among the top FASB standards decreasing shareholder value.

37 Years ago, I served on the FASB’s advisory committee on the corporate acquisition and goodwill project (SFAS No. 142, 2001). I well remember that all the executives on this committee vigorously pushed for the immediate expensing of goodwill.

38 The PCAOB, the U.S. auditor regulator, cited in recent years several major accounting firms for deficiencies in the auditing of acquired intangibles, particularly goodwill (see Kim, 2018).

39 Recall Oswald et al. (2017) finding that U.K. firms which switched from expensing to capitalization enhanced their R&D spending.

40 Strategic assets are benefit-generating resources whose supply is limited (scarce), and cannot be easily imitated by competitors (see Lev & Gu, 2016, Ch. 11).