«Leif Atle Beisland University of Agder Dissertation submitted to the Department of Accounting, Auditing and Law at the Norwegian School of Economics ...»
OSE. We have access to 741 firm-year observations from firms reporting either according to IFRS or NGAAP. After removing the 1% upper and lower tails, we have 381 IFRS observations and 341 NGAAP observations. After controlling for other known drivers of response coefficients unrelated to the choice of reporting regime, which reduces the sample from 722 to 635 or 590 firm-year observations due to missing control variables, we find that the investors are more responsive to IFRS than to NGAAP book values; there is evidence of a significant structural break between the two samples. We therefore reject the null hypothesis in favour of our alternative. The sources of these results are identified to be increased emphasis by investors on net operating assets as well as on net financial debt, consistent with the expected effects of increased recognition of intangibles and increased measurement at fair value. This result is found to be quite robust to changes in the empirical specification and the statistical test procedures.
Our second hypothesis is that there is a structural break in the earnings response coefficient when firms alter reporting regime from NGAAP to IFRS. We believe that the earnings effect caused by the shift in accounting regime is more ambiguous than the book value effect. First, increased recognition of intangible assets may lead to improved ‘matching’ of investment expenditures with future revenues. This is believed to increase the correlation between stock returns and price-deflated earnings. There is also evidence that increased recognition of intangible assets increases the value relevance of earnings; see e.g. Lev and Sougiannis (1996), Aboody and Lev (1998) and Lev and Zarowin (1999). Second, more measurement at fair values induces more non-recurring or transitory revaluations into earnings, which tends to decrease earnings response coefficients; see Beaver (1998, pp. 74-76). Previous empirical studies find support for transitory earnings due to fair value revaluations having a smaller re
lin and Subramanayam (2007).
Our empirical tests of the second hypothesis provide evidence that the earnings response coefficient under NGAAP is larger than under IFRS, after controlling for value relevance drivers not directly related to financial reporting. This result holds both when tested by price and return regressions, and when the sample is constant, consisting of the same firms both in the IFRS and NGAAP sample. The main source of this finding is that the earnings response coefficient of firms reporting according to IFRS is reduced due to transitory items, relative to the ones reporting according to NGAAP.
This paper is outlined as follows. Section 2 gives a short summary of the differences between IFRS and NGAAP and a short review of related literature, develops the hypotheses and outlines the test design. The data, the selected sample, descriptive statistics and analyses of simple correlations are given in Section 3. Section 4 performs the statistical tests and discusses the results. The sources of the findings in Section 4 are further examined in Section 5. Finally, Section 6 concludes.
2. Accounting Differences, Previous Research, Hypotheses Development and Test Methodology This section starts with a brief overview of the accounting differences between IFRS and NGAAP. Then the previous research on the value relevance differences between IFRS and NGAAP are reviewed, as well as the most important studies focusing on the value relevance effects of increased recognition of intangible assets and increased measurement at fair value.
Based on the accounting differences and previous research, we form our hypotheses regarding
shifts from NGAAP to IFRS. Finally, we outline the methodology for testing these hypotheses.
2.1 Major Differences between IFRS and NGAAP International Financial Reporting Standards, IFRS, are the accounting standards issued by the International Accounting Standards Board, IASB, in London. These standards aim at being the accounting regime utilized internationally. In 2002, the European Union decided that all exchange listed firms within the European Economic Area, EEA, must adopt IFRS in their consolidated financial statements from 2005 – and many countries outside the EEA have chosen to follow. IFRS are based on a balance sheet oriented conceptual framework, which defines assets and liabilities2. Equity is the residual, i.e. assets minus liabilities. The IFRS has increasingly pointed at fair values as the principle of measurement after the initial recognition of assets and liabilities, but transactional historical cost is widely accepted when there is no reliable measurement of fair value.3 In principle, revenue is an increase in assets or decrease in debt; costs are increases in debt or decreases in assets. Thus, comprehensive income is the change in equity not related to capital expansions or withdrawals. Although some fair value revaluations are booked directly in equity, other revaluation gains and losses are reported in earnings, making earnings more non-recurring than if earnings are governed by the matching of transactional costs with earned revenue (see Penman, 2007).
Norwegian Generally Accepted Accounting Principles, NGAAP, are the accounting regulation in Norway, a member of the EEA. The most important regulations under NGAAP are the For a discussion of the balance sheet orientation of IASB and FASB, see Dichev (2007). Dichev calls for a reassessment of the balance sheet approach and argues that the income statement approach to accounting is the natural foundation for financial reporting for most firms.
Whittington (2008) claims that a fair value view is implicit in IASB’s public pronouncements. On the other hand, Cairns (2006) maintains that the use of fair values in IFRS is not as extensive as many imply.
dards Board, NASB. The NGAAP is based on an earnings oriented conceptual framework in which expenditures are matched with earned revenue to calculate the period’s earnings, based on unbiased estimates of for example economic lives. The matching principle is combined with prudence in which the book value is written down to fair value if there is an impairment loss, and reversed maximum to historical cost if fair value increases again. In principle, there is no other revaluation, i.e. write-ups to fair value when it is above transactional cost. But liquid financial instruments are to be measured at fair value. From 2005, all Norwegian firms have the option to report financial accounts according to IFRS – not only the exchange listed firms which are required to do so.4 The difference between IFRS and NGAAP may appear considerable as their main principles of valuation of assets and debt in the balance sheet are fair value and past transactional cost, respectively. Still, in practical terms, the two accounting regimes are not dramatically different for important classes of assets such as most inventories, fixed and intangible assets. For example; the cost model is usually chosen for fixed and intangible assets by firms reporting according to IFRS, even though the revaluation model is an equally emphasized option.5 According to Gjerde, Knivsflå and Sættem (2008), the most important differences between IFRS
and NGAAP are:
Regulation No. 1852, 17 December 2004; compare IFRS 1.
Fixed and intangible assets are recognized at transactional costs both according to IFRS and NGAAP, as the cost equals the fair value at the time of the transaction. After recognition the two accounting regimes might differ in measurement. According to IFRS, a revaluation model or a cost model is to be chosen. The cost model, in which assets measured at costs are amortized over the best estimate of their useful lives combined with impairment to fair value (conservatism), is the only alternative according to NGAAP. Since fair values of fixed and intangible assets could be difficult to obtain, the cost model is chosen by most firms reporting according to IFRS.
Then there is no material difference in reporting. If instead the revaluation model is chosen according to IFRS, the carrying amount of fixed and intangible assets is fair value at the times of revaluation. Between revaluations, the assets are amortized like in the cost model. Revaluation should happen so often that the carrying amount of the asset does not differ materially from its fair value. But since write-ups are reported directly in equity, earnings are not affected relative to the cost model; write-downs are reported in earnings as in the cost model. However, equity will differ.
nually for impairment according to IFRS. Impairment testing is also present in NGAAP, but NGAAP also requires goodwill to be amortized over the best estimate of its useful
IFRS requires research expenditures to be expensed, while NGAAP allows the rarely • used option to recognize such expenditures as intangible assets. In contrast, development expenditures should be recognized as an intangible asset under IFRS, as long as it satisfies the definition of an asset. NGAAP permits immediate expensing of such assets – an option that is commonly used.
Provisions for future expenditures are more rarely recognized as debt according to IFRS • than NGAAP; as such expenditures have to satisfy the definition of debt to be booked as liabilities. For instance, NGAAP allows expected expenditures of future periodic maintenance to be reported as an accrued expense. IFRS treats periodic maintenance as an investment when incurred, i.e. as a component of the maintained asset. This is also allowed according to NGAAP.
IFRS requires biological assets to be measured at fair value if they can be measured • reliably. Investment properties may be valued at fair value or cost. NGAAP, however, requires transactional cost combined with prudence in both cases.
NGAAP measures financial assets and debts at cost unless they are short term financial • instruments traded in a liquid market. IFRS measures more financial instruments at fair
To summarize the differences between IFRS and NGAAP, we may conclude that IFRS has more recognition and measurement at fair value than NGAAP; increased recognition is re
ing points for making hypotheses regarding differences in the response coefficients of earnings and book value in subsection 2.3. The other starting point is the following review of the literature.
2.2 Previous Research on IFRS versus NGAAP, Recognition of Intangibles and Measurement at Fair Value Gjerde, Knivsflå and Sættem (2008) analyze the value relevance effects of adopting IFRS in Norway, focusing on the restated financial statements from NGAAP to IFRS in 2004, the year prior to the mandatory adoption of IFRS on the Oslo Stock Exchange. In 2005 comparable figures for 2004 had to be disclosed. The advantage of their approach is that they have two sets of financial statements, one according to NGAAP and one according to IFRS, representing identical underlying economic activity. Differences in value relevance are thereby caused solely by reporting differences. The cost of their approach is that the number of observations is limited, only 145. Nevertheless, Gjerde et al. find little evidence of increased value relevance after adopting IFRS when evaluating the two accounting regimes independently. However, they find that the reconcilement adjustment is incrementally value relevant both for the balance sheet and the income statement, after adjusting for non-recurring items in earnings.
Gjerde et al. attribute the marginal improvement in the value relevance of the balance sheet to increased use of fair value, and in earnings to an improvement in the value relevance of the net operating income caused by increased recognition of intangible assets (formerly expensed goodwill and development expenditures).
Capkun, Cazavan-Jeny, Jeanjean and Weiss (2008) also focus on restatements from local GAAP to IFRS, but include pooled observations from nine European countries. They have 1,722 observations, of which 98 are Norwegian. However, no separate results are reported for
equity reconciliations are not. The problem with such a pooled approach is that it is not easy to pinpoint what are the accounting drivers behind the result, since local GAAP might be inconsistent. Including countries with a weak tradition for high quality reporting, e.g. in terms of conservative accounting estimates, will contribute to change the expected effects of adopting IFRS relative to a benchmark like Norway with unbiased accounting based on transactional costs. A major advantage of utilizing data from several countries is the ability of focusing on broad economic effects in large samples, such as the effect on the cost of capital caused by more homogenous financial reporting; see also Daske, Hail, Leuz and Verdi (2007).
Several studies of individual European countries’ transition to IFRS have been conducted.6 For example, Horton and Serafeim (2008) investigate the market reaction and value relevance of First Time Adoption of IFRS in the UK. The reconciliation adjustment from UKGAAP to IFRS is value relevant for earnings (and not for equity). They attribute this result to differences in the reporting of goodwill, share-based payments, employment benefits, financial instruments and deferred taxes. Specifically, they find significant negative abnormal returns and positive trading activity for firms reporting a negative reconciliation adjustment from UKGAAP to IFRS earnings. Horton and Serafeim conclude that their findings are consistent with IFRS altering investors’ beliefs about risk-adjusted cash flows and, hence, also about stock prices.
We concluded the previous subsection by summarizing the main differences between IFRS and NGAAP being more recognition of intangible assets and more measurement at fair value Some studies focus only on the implementation of specific standards within the IFRS. For instance, Hamberg, Novak and Paananen (2008) study the value relevance effect of implementing IFRS 3 Business Combinations in Sweden.
vance effects caused by increased recognition of investment expenditures as intangible assets, relative to expensing them as incurred, and increased use of fair value, relative to transactional historic cost.