Unreliable Accoutning of Intangible Assets in a Digital Era: A study on the association between reliability and value relevance of intangible assets

Unreliable Accounting of

Intangible Assets in a Digital Era

A study on the association between reliability and value relevance of intangible assets

Authors: Andreas Danielsson & Fredrik Lindblad Supervisor: Elias Bengtsson

Co-assessor: Håkan Locking Examiner: Andreas Jansson Semester: VT21

Course Code: 4FE17E

Master Thesis in Business Administration

Acknowledgements

Elias, you have encouraged us to develop a critical mindset by questioning our thoughts and way to express ourselves. You have also inspired us to learn how to write clear and concise. Håkan, you have helped us to improve our statistical knowledge. Thank you both!

To our opponents, thank you for your feedback.

All the best!

Växjö - May 26

, 2021

Andreas Danielsson and Fredrik Lindblad

Abstract

Master Thesis in Business Administration

School of Business and Economics at Linnaeus University, VT 2021 Authors: Andreas Danielsson and Fredrik Lindblad

Supervisor: Elias Bengtsson Co-Assessor: Håkan Locking Examiner: Andreas Jansson

Title: Unreliable Accounting of Intangible Assets in a Digital Era – A study on the association between reliability and value relevance of intangible assets Background: The purpose of accounting and financial reporting is to provide useful information for its users in their decision-making. The importance of intangible assets for companies in the modern digital economy has led to debates concerning how it should be presented in accounting. As of today, the relevance and reliability of intangible assets can be questioned since large values of intangibles are left out of accounting and financial reports.

Purpose: The aim of the study is to investigate the association between reliability and value relevance of intangible assets.

Method: Using the Feltham and Ohlson (1995) model, we test the association between intangible assets and market valuation of firms. We divide the sample into reliable and unreliable and test whether there is any difference in value relevance of intangible assets between the groups.

Conclusion: The result indicates, without statistical significance, that reliability seems to matter more for goodwill than for total intangibles and other identifiable intangibles. Moreover, we suggest that investors seem to focus more on accounting standards and uncertainty than management discretion when assessing reliability of intangible assets. However, we are not satisfied with how reliability has been operationalized earlier and this study reinforces our doubts. Thus, among other suggestions we propose further research directed towards investors to find out whether reliability is important and what they consider as reliable.

Keywords: Intangible Assets, Value Relevance, Reliability, Accounting,

Financial Reporting

Table of Contents

1. Introduction 6

1.1 Background 6

1.2 Problem 9

1.3 Aim and Research Questions 12

1.4 Contributions 13

1.5 Outline 14

2. Scientific Method 15

2.1 Research Approach and Theoretical Foundation 15

2.2 Research Design 17

2.3 Source Evaluation 18

3. Regulatory Framework 20

3.1 Intangible Assets 21

3.1.1 Goodwill 21

3.1.2 Identifiable Intangible Assets 22

4. Theoretical Framework 24

4.1 Accounting Information Framework 25

4.1.1 Economic Relevance 26

4.1.2 Reliability 27

4.1.3 Value Relevance 28

4.2 Positive Accounting Theory 30

4.2.1 Debt/Equity Hypothesis 31

4.3 The Trade-Off With Management Discretion 32

5. Hypothesis Development 35

5.1 High Reliability → High Value Relevance 35

5.2 Reliability Matters More for Goodwill Than for Identifiable Intangibles 36

6. Empirical Method and Data 38

6.1 Sample Selection 38

6.2 Data Collection 40

6.3 Models to Measure Value Relevance 41

6.4 Operationalization of Reliability 45

6.4.1 Tangibility 47

6.4.2 Leverage 48

6.4.3 Big Four Audit Firms 48

7. Results and Analysis 50

7.1 Main results 50

7.2 Additional Tests 54

7.3 Analysis of the Results 56

8. Concluding Discussion 59

8.1 Conclusion 59

8.2 Are Uncertainty and IFRS More Important Than Management

Discretion? 59

8.3 The Trade-Off With Management Discretion 61

8.4 Contributions, Limitations and Further Research 62

Bibliography 65

Appendix 68

Table of Figures

Figure 1: Accounting Information Framework. 26

Figure 2: Accounting Standards effect on reliability and relevance.. 33

Table of Tables

Table 1. Sample Selection. 39

Table 2. Industry distribution of sample firms.. 39 Table 3. The distribution of the total amount of intangible assets reported by

the firms. 40

Table 4. Descriptive statistics for dependent and independent variables. 43 Table 5. Pearson Correlation for variables used in Model 1 and 2. 45

Table 6. Operationalization of reliability. 47

Table 7. Results for the regression Model 1 and 2. 50 Table 8. Results for the regression Model 3 and 4. 51 Table 9. Results for the additional test using Model 1 and 2. 55 Table 10. Results for the regression Model 3 and 4 for the additional test. 55

Table 11. Summary of the main findings. 57

6 (70)

1. Introduction

The purpose of accounting is to provide useful information for its users. In the modern economy, the importance of intangible assets leads to discussions about how it should be presented in accounting. There are only few studies on how reliability of accounting numbers affects firm valuation. The aim of this study is to empirically investigate that relationship.

1.1 Background

The purpose of accounting and financial reporting is, according to the conceptual framework of the International Accounting Standards Board (IASB), to provide useful information for decisions of its users, such as investors, lenders, and other creditors. For the information to be useful it has to achieve two fundamental qualitative characteristics: relevance and representational faithfulness (IASB, 2018).

Accounting information is relevant when it can make a difference in the decision-making by its users and faithfully represented when it is representing what it is supposed to represent, which also means that it is reliable

(IASB, 2018). Hence, if the accounting information is relevant and reliable it can be a useful tool for its users to make better investment, lending, and other business decisions.

However, current standards on intangible assets seems to be rather restrained regarding the possibilities of including them in the balance sheet. Under IFRS,

1 In this study, we use reliability because it is the established term in the extent literature. In 2010 reliability was replaced by representational faithfulness as fundamental qualitative characteristics in the conceptual framework of accounting, besides relevance, but the content of the concept was similar after the change (Erb & Pelger, 2015).

7 (70)

the current standard by IASB, acquired intangibles are being capitalized

, while most internally generated intangibles such as brands, customer lists, etcetera are being expensed. Due to the current accounting standards of today, there is an absence of important intangible assets in financial reports. Hence, it can be questioned if accounting and financial reporting do provide usefulness to its users, as is the main purpose.

Jenkins and Upton (2001, p. 4) state that traditional financial reporting is unable to capture the ‘value drivers that dominate the new economy’, referring to intangible assets. Lev and Gu (2016, p. 84) states that it is surprising that even though companies spend a lot of money on intangibles and the fact that these assets generate future benefits, the accounting framework does not treat them as such. The increased importance of intangibles can be explained by a modern digital economy.

For many modern businesses, intangible assets have become one of the most common sources of value creation. Companies like Facebook, Amazon, Apple, Alphabet & Microsoft constitutes the top five corporations in the U.S.

based on market-capitalization (Nasdaq, 2021) and they have one commonality, their digital business models depend on intangible assets. The importance of intangible assets as one of the major value creating sources to modern corporations has affected the business community in many aspects, such as what employees work with and how they do it, but accounting standards has only changed slightly (Holthausen & Watts, 2001). It appears that even though there has been a significant change in the economy,

2 Capitalization is the recording of an expense on the balance sheet instead of fully including it in the income statement instantaneously.

8 (70)

traditional financial reporting has not changed to currently recognize

the value creation of intangible assets for modern companies (Lev & Gu, 2016, p.

88).

The current accounting framework and traditional financial reports are arguably most relevant and useful for businesses with tangible assets. Standard setters could consider adaptation to a more digital economy with relevant information for firms in the launch- and growth stage, and with value in intangible assets, for the accounting information to provide relevant information for its users and be useful. Though, it seems like accounting standards are still conservative regarding the rules concerning the intangible assets.

Then, why are accounting standards so conservative and without leeway for management to adapt it to the specific business? Wouldn’t it provide more useful information with more flexible standards? By giving the company and managers more financial reporting discretion they are given the chance to report a better guess on the reality of the intangible asset. However, it could possibly lead to financial reporting where it is misleading because of opportunistic behavior. The trade-off between opportunity to present the best adapted accounting information, and standardized rules that prevent opportunistic behavior by management, has led to the precautionary principle of accounting and a conservative standard (Ji & Lu, 2014).

Altogether, it can arguably be questioned whether today’s accounting and financial reports do provide useful information for its users about intangible

3 Recognition means including an item that meets the definition for being booked in the balance sheet or income statement (IASB, 2018).

9 (70)

assets in the modern digital economy, in which intangible assets are the value- drivers for many businesses. Therefore, relevance and reliability of accounting and financial reporting can also be questioned. Hence, the relevance and reliability of intangible assets are both interesting and important to investigate further for accounting to keep its value as one of the major information sources for the firms’ stakeholders, and fulfill its purpose that is to provide useful information.

1.2 Problem

Considered the increased importance of intangible assets, the question about the value of accounting and financial statements emerges as a hot field of research. Value relevance is an established term in this field, and an accounting amount is defined as value relevant if “it has predicted association with equity market values”

(Barth, Beaver & Landsman, 2001, p. 79). Researchers find value relevance interesting since that knowledge will provide information about how investors assess various accounting information (Holthausen &

Watts, 2001).

Relevant information has the ability of influencing economic decisions of users (Wyatt, 2008). For instance, investors might value a firm higher if it controls a computer software, that can be used to either produce value that can lead to cash flow, or to sell, than if it does not. That is, the information about the firm’s possession of the computer software is relevant for investors. By studying how relevant information is to users, conclusions can be drawn about

4 Notice the difference between relevance in the Conceptual Framework of Accounting by IASB, and value relevance in accounting research. Relevance in the Conceptual Framework describes a characteristic accounting information shall have, while value relevance is a measure used in research to describe the association between accounting information and firm value.

10 (70)

investors considerations of the presented information when taking decisions, and to what extent information captured by accounting can explain firm value.

There is extensive research on value relevance of intangible assets in the existing literature and results are unanimous that recognized intangible assets are value relevant for investors, which means that capitalized intangibles increases stock valuation (Dahmash, Durand & Watson, 2009; Wyatt, 2008;

Holthausen & Watts, 2001; Barth, Beaver & Landsman, 2001; Oliveira, Rodrigues & Craig, 2010; Godfrey & Koh, 2001).

However, studies have found that intellectual property, R&D expenses, amortization, and impairment expenses were not value relevant for investors (Oliveira, Rodrigues & Craig, 2010; Kimouche & Rouabhi, 2016). In other words, recorded expenses seem to be interpreted as consumed capital, while capitalized intangibles are regarded as investments which are value-enhancing for the firm. This is consistent with how the difference between expenses and assets are described in accounting theory (Deegan, 2013).

Research on value relevance have found that further research on underlying economics and reliability are necessary to impact standard setters (Holthausen

& Watts, 2001). Standard setters have kept its conservative handling of intangibles even though there are consensus among academics that more information about intangibles should be recognized or disclosed to increase value relevance (Barth, Beaver & Landsman, 2001).

The negligence of these results depends partly on that it is regarded as a too

bold assumption that all wanted attributes of accounting are reflected in stock

prices (Holthausen & Watts, 2001). These studies prove an association, but not

causality, between recognized intangible assets and stock price which means

that we cannot infer that investors used the specific information in firm

valuation (Wyatt, 2008). The conceptual framework of accounting includes

several characteristics of accounting and measuring relevance only by the

11 (70)

relation to stock price is too primitive, which is why new approaches and measurements are needed (Holthausen & Watts, 2001).

To make future findings more useful, researchers have decomposed value relevance of accounting information to enable make more narrow studies (Wyatt, 2008; Maines & Wahlen, 2006). Value relevance depend partly on how well accounting can capture firms’ intangible assets and how they are allowed to be valued, and partly on whether the assessed value is trustworthy, referred to as reliability. Reliability has been found to be an underlying factor that affects value relevance, which implies a conditional relationship.

(Kallapur & Kwan, 2004; Ji & Lu, 2014; Ji, 2018). However, there are only few studies made on the association and they have some limitations.

The limited number of studies along with the new context and measurement issues makes our study worth-while and contributing. As for now, we cannot draw any definite conclusions about the relationship between reliability and value relevance for several reasons. First, there are still a limited number of studies on reliability and its association with value relevance (Ji & Lu, 2014).

Second, there are limited studies in a European context on the association (Ji, 2018; Ji & Lu, 2014; Harto & Paranti, 2017). Third, there are measurement issues on reliability that have been recurring in discussions of research limitations and suggestions for further research (Maines & Wahlen, 2006; Ji

& Lu, 2014). Therefore, we find it valuable to do further research on reliability of accounting information on intangible assets.

Our results can provide standard setters with knowledge about how reliability

affect the usefulness of financial reports. For instance, if reliability matters for

value relevance of intangible assets, it indicates that investors assess

trustworthiness as important for firm valuation. This is consistent with Ji

(2018, p. 98):

12 (70)

“This implies that the reliability of intangible assets is significant information to help evaluate the relationship between accounting information and stock prices.”

In summary, previous research has proven that capitalized intangible assets have a positive association with firm value. However, standard setters have not considered these results because of the lack of a more detailed framework and empirical research on such. Thus, theories that decomposes value relevance into underlying economics and reliability have been developed. The relation between reliability, which is one factor of value relevance in the frameworks, and value relevance has been empirically tested only by a few and with measurement deficiencies, which is why we believe this study is worth-while and contributing.

1.3 Aim and Research Questions

The following is the aim of this study:

The aim is to investigate the association between reliability and value relevance of intangible assets.

The main purpose are subdivided into two main research questions:

1. Is there any difference between value relevance of intangible assets for reliable and unreliable accounting information?

2. Is there any difference in how investors value reliability regarding goodwill and identifiable intangible assets?

5 Intangible assets consist of goodwill and identifiable intangible assets and will be further explained in Chapter 3.

13 (70)

1.4 Contributions

The theoretical contribution to the existing literature on accounting of intangible assets of this study are twofold. First, we did not, unlike prior researchers, find support for a positive association between reliability and value relevance of total intangibles, goodwill and other identifiable intangible assets respectively. This implies that reliability may not be as important as prior researchers found. Second, we did not find support for, but indications that reliability is more important for goodwill than for other identifiable intangibles. This comparison has not been tested and analyzed by prior researchers.

The methodological contribution of this study is the combination of methods from the extent literature. For robustness we use two main ways and one additional way to measure reliability. We found differences which implies that the way to operationalize reliability is important to consider. Thus, we find it valuable with a direct study on investors to find out how they are determining reliability of a firm.

Furthermore, we contribute with a new way to make a statistical test on the differences of coefficients from the regression models. Prior researchers compare analytically while we add variables to be able to do a significance test on the differences as well.

This study has practical contributions for standard setters, firms and investors.

Since management has larger leeway for goodwill and the results indicate that reliability matters more for valuation of goodwill, standard setters should consider reliability when determining the level of management discretion. We propose, with support from prior research, that it is a trade-off between flexibility and reliability.

Lastly, our results also contribute with insights for investors and firms. With

the present accounting standards, reliability to the firms do not seem to be of

14 (70)

large importance. It also indicates that goodwill seems to be more value relevant, why firms have incentives to increase its goodwill rather than other identifiable intangibles to increase its market capitalization. This is especially important in the modern digital economy with intangible assets of growing importance.

1.5 Outline

The remainder of this thesis paper is structured as follows. Chapter 2 presents the choice of a deductive research approach and a quantitative method to investigate the association between reliability and value relevance.

Subsequently, Chapter 3 gives a brief description of the current accounting standards for goodwill and identifiable intangible assets. Furthermore, the Accounting Information Framework (AIF) and Positive Accounting Theory (PAT) are explained in Chapter 4 and the hypotheses are generated and presented in Chapter 5.

Chapter 6 describes the empirical method of the study. We use the Feltham

and Ohlson (1995) model to measure value relevance. We operationalize

reliability in two main ways and one additional. In Chapter 7 the results of the

tests are presented. We did not find any strong evidence for that reliability

affects value relevance. Finally, we analyze and discuss the results in Chapter

8 and the study’s conclusions and limitations are presented in Chapter 9.

15 (70)

2. Scientific Method

A deductive approach is taken to empirically investigate the association between reliability and value relevance through a quantitative method. The Accounting Information Framework (AIF) and Positive Accounting Theory (PAT) will constitute the theoretical foundation and with public financial data, a cross-sectional research design is used in our attempt to fulfill the aim.

Moreover, the empirical method will be presented in Chapter 6.

2.1 Research Approach and Theoretical Foundation

Our aim is to investigate the association between reliability and value relevance of intangible assets and to fulfill that a deductive research approach is taken. With foundation in existing theory and research of a certain phenomenon we develop and test falsifiable hypotheses. The deductive research has elements of induction as well since the results and conclusions may comprise confirmation or rephrasing of existing theory (Bryman & Bell, 2015).

The deductive research approach is commonly taken through a quantitative research method and in this study, we use a quantitative method because of its ability to handle large amounts of data and to make the results generalizable.

The main purpose of science is to create knowledge that can be generalized through time and space and after the initial creation of a theory through induction, a deductive research approach through a quantitative research method is suitable to perform that (CRQ, 2015).

Most prior research with similar aim has used a quantitative method (Ji & Lu,

2014; Dahmash, Durand & Watson, 2009; Ji, 2018) and we accompany but

with adjustments in the empirical method. We contribute to the literature with

changes in operationalization of variables and with the comparison between

goodwill and other identifiable assets. However, a qualitative research method

could have been used as well.

16 (70)

A qualitative alternative to our study could be to interview investors to investigate how they assess and value reliability in investment decisions.

Zéghal and Maaloul (2011) mean that inadequate accounting treatment of intangibles leads to misevaluation of firms and they therefore suggest that further research should explore the value of disclosures through interviews with management and investors. The benefit with a study like that would be more nuanced answers and an ability to find whether investors in fact takes reliability into consideration, or if the association found in other studies are due to omitted variables and bad research design. We suggest this as an opportunity for further research.

We are critical to the way prior researchers operationalize reliability but understands the difficulty in doing that. Therefore, a qualitative study on investors’ evaluations of reliability would give useful insights. Having said that, our purpose is to test the association between reliability and value relevance described by Wyatt (2008) and Maines and Wahlen (2006), and we value the ability to generalize results that quantitative and deductive research provides.

Our hypotheses are based on the Accounting Information Framework (AIF) by Wyatt (2008) and Maines and Wahlen (2006). Further, Positive Accounting Theory (PAT) describes and predicts management’s choice of accounting method, based on management incentives. The present accounting standards allow a certain leeway for management to use evaluations. Since PAT claims that management act in self-interest and accounting standards allow evaluations, we can test how investors value reliability by rating reliability based on management incentives.

Furthermore, research should aim to provide understanding rather than making

assertions of what accounting standards should be like. Holthausen and Watts

(2001) criticize prior studies on value relevance for being too normative. They

opine that researchers should hold the main objective of accounting (decision

17 (70)

usefulness) as given, and from that observe whether the stated objective is properly met through the present accounting standards and regulations. We believe it is a fair point, and therefore aim to contribute with knowledge about how reliability affects value relevance, which by extension can help to assess how important reliability is to achieve the main objective of accounting.

Lastly, scientific research should not be a political discussion but instead aim to understand how the world works. This means that theory on accounting standards is different from actual standards and regulation. However, in this study on the association between reliability and value relevance we must consider the present standards since the leeway in accounting standards are a fundamental part of our tests. We do our tests with the presumption that reliability to accounting can be improved within the frames of the current accounting standards. With no leeway for management, the study would not have been possible. In other words, the hypotheses are built upon the theory, but the standards are a premise for the tests.

2.2 Research Design

This study uses a cross-sectional research design by studying multiple firms’

accounting numbers and stock prices. We gather data from several firms to be able to quantify the information into variables and test relationships and dissimilarities. Prior researchers have used the cross-sectional design for similar studies (Ji & Lu, 2014; Dahmash, Durand & Watson, 2009; Ji, 2018).

Further, Ji and Lu (2014) and Dahmash, Durand & Watson (2009) have also

used the longitudinal design, but to investigate changes following the

implementation of IFRS by comparing pre- and post-implementation. This is

an alternative for us as well, but it has been 16 years since the implementation

of IFRS in Sweden, why we believe the cross-sectional research design is more

contributing. It makes it possible to investigate relationships between variables

at determined points in time.

18 (70)

Furthermore, prior researchers have used simulations and experiments to investigate value relevance (Kadous, Koonce & Thayer, 2012; Healy, Myers and Howe, 2002). We believe that an experiment would be possible to fulfill the aim. For instance, through designing a case study in which we let investors value a firm, which we manipulate into two groups: reliable and unreliable, we could test if reliability is associated with value relevance. In that case, half the sample gets the reliable firm and half the unreliable, and we could have compared their valuations of the firms. The advantage of that type of study would be that the reliable and unreliable firms would have been identical except for our change of reliability, which means we would not have to consider control variables like we must in the cross-sectional design.

However, that method is problematic for several reasons. The case would have to be realistic, comprehensive, and detailed because investors take a lot into account in their investment decisions. That would demand a lot from the respondents and without compensation or any other incentives it would have been hard to get enough responses. Consequently, this study is a cross- sectional study of financial data to enable multiple dimensions due to the massive amount of data available.

2.3 Source Evaluation

Scientific research must be based on credible sources to be useful and that is why we have been careful in our choice of sources. The sources used in this thesis is almost exclusively peer-reviewed scientific articles, which are considered the most credible sources. In addition, textbooks and accounting standards are used to explain for instance the current accounting standard in Chapter 3 (Regulatory Framework) and our choice of scientific method and empirical method. The theories and literature review that are the foundation of our empirical study are based on scientific research.

The risk of omitting important research findings is present for all researchers.

We have managed this risk in several ways that we believe is sufficient to

19 (70)

motivate research with this aim. Completeness is treated using multiple databases including Linnaeus University OneSearch and Google Scholar.

Further, a rigorous search through reference lists of key articles (Wyatt, 2008;

Ji & Lu, 2014; Ji, 2018; Dahmash, Durand & Watson, 2009; Gülec, 2021) enhances our understanding of the research field.

The last one is a very recent literature review, why we believe, if we can trust that Gulec (2021) made an adequate work, that we will cover most literature through that article and references. We did also search through citations of our key studies. However, the literature included is not exhaustive but representative of prior research and includes both the latest (Ji, 2018; Harto &

Paranti, 2017; Gülec, 2021) and some of the most cited (Wyatt, 2008; Ji & Lu,

2014) scientific articles.

20 (70)

3. Regulatory Framework

Current accounting standards are conservative regarding capitalization of intangible assets. The present accounting standards for goodwill and identifiable intangible assets are presented in this chapter.

According to Holthausen and Watts (2001), studies within the value relevance literature should consider the objective of standard setting to improve the usefulness of the findings and conclusions within value relevance studies.

They criticize prior studies for being too normative and that the studied associations have to aim to help standard setters achieve its main objective.

They suggest that:

“Simple assertions by authors that standard setters should consider that attribute desirable are not sufficient for scientific research.”

(Holthausen & Watts, 2001, p. 4).

Further, knowledge about present standards is necessary to critically evaluate the method and arguments for why management discretion is interesting to study. The present standards, give opportunities for evaluations that can be made by management to direct accounting numbers to its own advantage.

Hence, a brief review of the existing accounting standards is necessary to understand the current situation of manager’s discretion.

The accounting standards of IASB that manages intangibles are IFRS 3 – Business Combinations and IAS 38 – Intangible Assets. These standards explain how goodwill and identifiable intangible assets are to be identified, accounted for, and valued, as well as certain instructions regarding disclosures.

Further, goodwill and intangible assets without definite useful life are not

systematically amortized. They are instead tested for impairment, which is

described in IAS 36 – Impairment of Assets.

21 (70)

3.1 Intangible Assets

An intangible asset is defined as an asset without physical substance which is controlled by the entity and expected to provide future economic benefits.

Intangible assets can be divided into goodwill and identifiable intangible assets. All intangibles that are not goodwill are identifiable intangibles and thus, per definition, have to be able to be identified.

3.1.1 Goodwill

Goodwill is an intangible asset that accounts for the difference between purchase prices and fair value of the identified assets in an acquired company.

Thus, internally generated goodwill cannot be capitalized. The items included in goodwill are often not easily quantifiable, for instance, brands, customer lists, location, human resources, staff, and etcetera, thus making them unidentifiable. Goodwill is in some cases motivated by the expected synergy effects of the acquisition.

Moreover, goodwill cannot be sold or bought separately and thus only occurs upon purchase of another company. IASB assumes that acquired goodwill are more reliable and relevant because it has been object in a market transaction and therefore more reliably measured. Since goodwill arises through acquisitions no internally generated goodwill is capitalized.

Systematic amortization, which refers to the distribution of expenses of an asset over its estimated lifetime, is not allowed for goodwill. This because it is assumed that goodwill has an indefinite useful lifetime and is therefore instead tested for impairment every year in compliance with IAS 36 – Impairment of Assets.

This means that goodwill is annually tested if it carries more value in the

balance sheet than its recoverable amount, which is the highest of fair value

less costs of sale and value in use. Goodwill cannot generate cash flow

independent of other assets, because it is unidentified. Therefore, it is treated

22 (70)

as a part of a cash-generating unit or group of cash-generating units, which is expected to benefit from the synergy effects of the combination.

The cash-generating unit to which goodwill is allocated is then tested for impairment. If the carrying value is larger than the recoverable amount, goodwill will be written down to the recoverable amount. Reversal of impairment loss for goodwill is not allowed since it is hard to evaluate if it is retained value of impaired goodwill or if it is internally generated goodwill.

To facilitate the annual test of impairment companies shall distribute the goodwill on cash-generating units directly at the acquisition time. This unit will be subject to the annual test of impairment, which makes it easier if the goodwill is already distributed. Allocation of cash-generating-units is hard, which means it is subject for evaluations made by management.

3.1.2 Identifiable Intangible Assets

Unlike goodwill, some intangibles can be identified and distinguished, and these are treated different in accounting. Common examples of identifiable intangible assets are computer software, customer lists, mastheads, patents, brands, and licenses. Identifiable intangible assets can either be acquired or internally developed. Acquired identifiable intangible assets are easier to recognize in the company’s balance sheet than internally generated assets as they are initially valued and recognized by its acquisition costs.

Expenses for internally generated identifiable intangible assets are only

capitalized if certain criterions are met. For R&D expenditures it depends on

which phase the project is currently at. In the research phase, all costs are

accounted as expenses, while the costs in the development phase can be

capitalized as intangible assets. The company has to intend and be capable of

completing the intangible asset and either use it or sell it and demonstrate how

the asset will produce future economic benefits, for development costs to be

capitalized.

23 (70)

Intangible assets with a definite useful life are amortized based on the asset’s

estimated lifetime, while intangibles with an indefinite useful life are treated

the same way as goodwill, which means annual impairment test and no

systematic amortization. However, the identifiable intangible assets are not

allocated to cash-generating units if the asset generate cash flows

independently. Lastly, reversal of impairment is allowed because the assets are

identifiable unlike goodwill.

24 (70)

4. Theoretical Framework

The Accounting Information Framework (AIF) decomposes value relevance into concepts that can be observed and analyzed isolated, which provide a deep understanding of the theory and logic behind accounting and accounting standards. Further, Positive Accounting Theory (PAT), which has its origin in opportunistic theory, can be used to explain management discretion that is one component of reliability. Lastly, another perspective suggests a trade-off between management’s opportunity to adapt accounting to the business, and increased reliability due to less management discretion.

In the beginning of the century, Holthausen & Watts (2001) criticized the extent value relevance literature for not being supported by descriptive underlying theory and thus not very useful for standard setters. Prior researchers investigated the association between intangible assets and stock prices, but without any established framework to explain the results. Since then, Wyatt (2008) and Maines and Wahlen (2006) have developed elaborated frameworks for accounting information.

The framework contributes with associations, problems, and opportunities to understand accounting. It is based on prior research and describes the relationships between underlying economic reality, accounting and stakeholder assessment which can be used as foundation for empirical research. Thus, a detailed explanation of the model is essential and will be done in Section 4.1.

In the Accounting Information Framework (AIF) three constituents of

reliability is presented: uncertainty, present accounting standards, and

management discretion. Management discretion is an assumption in the AIF

which are supported by the Positive Accounting Theory (PAT). Briefly

summarized, PAT aims to explain the choice of a specific accounting method

25 (70)

based on management incentives. PAT is presented in Section 4.2 with particular focus on the debt/equity hypothesis.

Lastly, another perspective suggests a trade-off between management’s opportunity to adapt accounting to the business and increased reliability due to less management discretion. In Section 4.3 we convey this in addition to AIF to lay a foundation for the discussion in Chapter 8.

4.1 Accounting Information Framework

The Accounting Information Framework (AIF) aims to explain the relationship between accounting information and economic values in a firm.

The frameworks of Wyatt (2008) and Maines and Wahlen (2006) are in many ways similar. Both frameworks are a decomposition of value relevance into several parts that aims to explain the relationship between accounting information and stock prices. We have integrated the two models into three parts.

First, the economic relevance describes what and how value is created through economic activities. Second, reliability focuses on how the economically relevant resources and undertakings are measured and how trustworthy the presentation and measures are. Third, accounting information are part of the basis for firm valuations by investors. The framework explains a logical chain from what and how value is created, through how it is presented in financial accounting, to how it is valued by stakeholders. The last part is referred to as value relevance.

The separation and clarification of economic relevance, reliability and value relevance is a way to both provide understanding for the relationships, and a way to unify research around concepts and establish terms to prevent misunderstandings.

However, it can be argued that this is not a theory because it is affected by

standard setters’ interpretations and how they design the standards, but this

26 (70)

framework explains the theoretical concepts given the current standards. If standards are changed, new results will be found in studies on the associations, but the framework itself and the relations are still valid.

Figure 1: Accounting Information Framework. Source: (Wyatt, 2008; Maines & Wahlen, 2006). To see the original figures, see Appendix.

4.1.1 Economic Relevance

The first part of the AIF in Figure 1 describes what value is created through economic activities. This can be divided into value construct and value creation process. Value construct refers to a circumstance in which value is constructed or transferred, for instance through R&D expenditures, goodwill acquisition, or the making of a brand. These are actions and transactions that initiate something that will create value for the firm.

Value creation process refers to the process in which value is expected to be

captured, for example synergy effects from goodwill acquisition or increased

productivity due to R&D expenses into a new computer software for inventory

optimization. Value construct is the actions or transactions that initiates the

creation of value, while value creation process describes how value will be

captured from that. Accounting standards are based on economic relevance

27 (70)

which means the real economic values are independent of how accounting standards are designed (Wyatt, 2008).

The asset has a value when it is assumed to provide future cash flow or benefit.

Therefore, value construct and value creation process are dependent on expected future benefits (Maines & Wahlen, 2006). If no future benefit is expected, the resource has no real value for the stakeholders. To sum up, economic relevance determines what resources are relevant to stakeholders by predicting future benefits.

4.1.2 Reliability

The second part of the framework explains how reliably the resources and undertakings are measured and presented in the financial statements.

Reliability refers to what is recognized in accounting and how it is measured.

This is explained by (Maines & Wahlen, 2006, p. 403):

“Reliability is the degree to which a piece of accounting information (1) uses an accounting construct that objectively represents the underlying economic construct it purports to represent, and (2) measures that construct without bias or error using the measurement attribute it purports to use.”

Reliability is impacted by three factors: uncertainty, present accounting standards, and management discretion (see Figure 1). The uncertainty in how a resource will provide future benefits affects the reliability of the measure.

This is because it will be hard for stakeholders to trust a measure if it is difficult to estimate future value of an economic activity (Wyatt, 2008).

The value of an asset is a function of expected future benefit and the risk or

variation in the level of that benefit. For example, the present value of a

software application is determined by the expected future income from that

software with respect to when in time it is expected and the risk and variation

in the level of it. Furthermore, it is affected by the risk in present benefit, that

28 (70)

is if it in effect has utility for the firm today or it if based on expectations that it will be valuable in the future. Uncertainty cannot be affected by standard setters, but it is a crucial aspect to consider.

Reliability to financial statements also depends on how well accounting standards require and enable firms to represent the economically relevant information (Maines & Wahlen, 2006). For instance, what intangibles that are allowed to be capitalized and how items may be measured at acquisition and continuing influences the reliability. Properly designed standards improve the opportunity to adopt financial accounting to the particular business.

However, flexible standards also impair reliability due to management discretion since they have opportunity to engage in opportunistic behavior. If management have incentives to signal anything different from the truth and standards allows a leeway, accounting information may be assessed less reliable. For instance, management may want to capitalize expenses for intangible assets before they are probable to provide future benefits.

Capitalization of expenses increases earnings and inflates assets, why management discretion is suggested to be another factor that affects reliability.

Management discretion will be further presented and discussed in Section 4.2 where Positive Accounting Theory (PAT) is explained.

To conclude, uncertainty, accounting standards, and management discretion are the three elements of reliability, which in turn affects value relevance.

Reliability has a causal effect on value relevance (see Figure 1). Factors that affect reliability also affect value relevance but not vice versa, which proves that the relationship is unidirectional (Kadous, et al., 2012).

4.1.3 Value Relevance

The third part of the framework describes that accounting information are part

of the basis for firm valuations by stakeholders. Accounting numbers are

29 (70)

explained to affect firm valuation, which completes the chain between economic relevance and value relevance via reliability (see Figure 1).

The framework is applicable to all stakeholders, such as creditors and investors. The value relevance for creditors depends on how well accounting information provides useful information for their decisions to lend to the company. For investors, on the other hand, it is determined by how well it purports expected future cash flows which is the basis for stock valuation.

If the value construct and value creation process were easy to assess and the future were known, we would not have economic relevance and reliability issues. However, this is not the case and uncertainty affect the ability to clearly define which resources will provide future benefits. Especially for intangible assets, the uncertainty is vast (Wyatt, 2008), which implies that the economic relevance is hard to determine.

For instance, the purchase of a building is easier to predict future benefit from than an internally generated software, or goodwill. The building provides value through its capability to be sold in an active market and the use of it in business activities. The valuation of such building can be based on market value, value of use, or other relatively reliable measures.

On the other hand, the economic properties of intangible assets make financial information on intangibles less reliable. Intangibles are often characterized by a less active market, a more firm-specific product and often more uncertain in how it will provide value in the future. The precision in how accounting information capture future benefit is weaker for intangibles than for tangibles since the connection between the value creation process and future benefits are stronger for the tangibles. In other words, the nature of intangibles entails weaknesses in the reliability and value relevance step as well.

In summary, the value relevance is affected by economic relevance and

reliability. The value relevance describes how stakeholders value accounting

30 (70)

information. If the links are weak from start value relevance will be low. In the following section Positive Accounting Theory (PAT) explains this more in detail but summarily PAT expects managers to act in their own self-interest and choose accounting method based on their incentives (Deegan, 2013, pp.

272-273).

4.2 Positive Accounting Theory

Positive Accounting Theory (PAT) tries to describe and predict why a certain accounting method is chosen in preference to others based on managers’

incentives. This as an economic effect for various stakeholders (Watts &

Zimmerman, 1986, p. 359). For instance, investors may use the information to predict future cash-flows for stock valuations, while creditors use it to ensure that they will get their payment (Deegan, 2013, p. 302).

PAT derives from the criticism to the efficient market hypothesis (EMH). The EMH assumes that the capital markets react to all publicly available information in an unbiased and effective manner. Thereby, the EMH assumes that the share price of a firm is reflected by all available information and not only accounting information, which is seen as only one of many supplies of information to the capital market. If managers do not make honest accounting disclosures, the market will “see through” the choice of accounting method and it will be reflected in the share price (Deegan, 2013, pp. 276-277).

Ball and Brown (1968) investigated whether accounting information did provide useful information to the capital market or not, following their critique of EMH. Their results showed that companies’ unexpected accounting earnings produced abnormal returns in the capital market, supporting that the financial reports were useful for its users (Ball & Brown, 1968 cited in Deegan, 2013, pp.277-278).

However, they did not explain why managers of the firm adopted particular

accounting methods. (Deegan, 2013, p. 279). Therefore, Watts and

31 (70)

Zimmerman (1990) suggested that by looking at the contracts between the individuals who are involved in the accounting choices the choices of accounting methods could be explained. By introducing and using the assumptions of the agency theory, PAT was developed.

Within the agency theory, the company is seen as a nexus of contract and is constituted by several different individuals. Hence, there are various different contracts within the company and thus different relationship, for instance between the owner (principal) and the manager (agent) alternatively between the manager (agent) and the creditor (principal).

The contracts are put in place to reduce conflicts of interests between the individuals. The contracts are also supposed to ensure that the individuals within the company benefit the organization (Deegan, 2013, pp. 281-282).

However, based on the agency theory, PAT assumes that all individuals are motivated by their own self-interest which supports that they will act opportunistic and try to maximize their own wealth.

To summarize, PAT assumes that the agent and the principal are separated from each other and that the accounting choice is entirely given to the agent (Watts & Zimmerman, 1990). That in combination with the assumption of a self-interested agent implies that the agent is going to make accounting choices in an opportunistic manner on the expense of the principal.

Furthermore, within the PAT literature three different variables are presented that will affect the agent’s choice of accounting method. These variables have helped to predict what would happened within a specific scenario. These three variables are: bonus plans, debt agreements and political procedures and each one of them constitutes a hypothesis (Watts & Zimmerman, 1990). The one of particular interest in this paper is the debt/equity hypothesis.

4.2.1 Debt/Equity Hypothesis

Watts and Zimmerman (1990, p. 139) formulate the debt/equity hypothesis as:

32 (70)

“The debt/equity hypothesis predicts that the higher the firm’s debt/equity ratio, the more likely managers use accounting methods that increase income.”

The agency relationship between managers and external lenders causes lenders to undertake certain activities to minimize or eliminate the likelihood of not getting repaid. These are called debt covenants and can be described as restrictions that lenders insert on lending agreements to either limit the borrower or require the borrower to take action if clauses are violated (Deegan, 2013, pp. 285-286).

The debt covenants can for instance include restrictions about debt/asset ratio, total assets, or certain earnings requirements. PAT describes that debt contracts are made to reduce the conflicts of interests but that the opportunistic manager will violate these contracts if the firm is close to breaching the debt covenant. Hence, if the level of debt increases or if the earnings are too low, managers will act opportunistic and there are possibilities that the reported assets and income will be overstated (Deegan, 2013, pp. 289-290).

In summary, PAT describes that management will act opportunistic if they have leeway in the choice of accounting method. However, there are a trade- off for standard setters concerning management discretion. On the one hand, leeway in accounting standards provides opportunities to enhance accounting, but on the other hand it impairs reliability in the way described in this section about PAT.

4.3 The Trade-Off With Management Discretion

Reliability is affected by uncertainty, present accounting standards, and

management discretion. The extent to which judgement by management is

allowed affects both the reliability and value relevance (Ji & Lu, 2014). This

perspective proposes a trade-off for standard setters regarding the extent to

which judgement are allowed.

33 (70) Figure 2: Accounting Standards effect on reliability and relevance. Source: (Ji & Lu, 2014).

To see the original figure, see Appendix.

Figure 2 illustrates a model of management discretion and standard setting.

The input at top of the model are the present accounting standards. More management discretion allows firms to adopt accounting to the business and including more assets in the balance sheet. If firms can do estimates and judgment is involved, it can be used to enhance the quality of accounting and provide more information to improve decision-making.

On the contrary, the more judgement involved the more likely management engage in opportunistic behavior. If accounting standards allow more flexibility management have opportunities to exploit it to its own benefit. The leeway can be used to signal dressed up financials and therefore reliability is lower.

Further, if firms are not allowed to make professional judgment there are less possibility for management to manipulate the numbers to its own benefit, but on the other hand it does not enable them to adopt accounting to the specific business to enhance the quality.

Consequently, it seems to be a trade-off in the choice of level of management

discretion. For instance, larger opportunity to capitalize identifiable intangible

34 (70)

assets can enhance the financial reporting information because including more intangible assets in the balance sheet could better capture the value drivers in the business. However, the possibility that presented information would not be reliable would be bigger due to the opportunity of utilizing the accounting standards in an opportunistic manner.

Hence, standard setters face challenges and difficulties when determining if managers should have more leeway and opportunities to enhance accounting of intangible asset or restrain managers discretion to make financial reporting information more reliable.

To summarize Chapter 4, the AIF is the basis which describes the relationship between economic relevance, reliability, and value relevance. In the framework reliability is conditional on value relevance. Further, PAT supports that management will use the leeway in regulation to its own advantage.

Lastly, there are a contrary perspective which suggests a trade-off between

management’s opportunity to adopt accounting to the business and increased

reliability due to less management discretion.

35 (70)

5. Hypothesis Development

We hypothesize that intangible assets in more reliable firms will be valued higher than in less reliable firms (H1) and that it is true for both goodwill (H1a) and identifiable intangible assets (H1b). We also conjecture that reliability is more important for investors in valuation of goodwill than of identifiable intangible assets (H2).

5.1 High Reliability → High Value Relevance

The AIF explains the relationship between reliability and value relevance.

Value relevance is affected by the economic relevance and reliability. A reliable measure of intangible assets is hypothesized to increase investors valuation of the firm.

Investors’ assessment of accounting information is divided into what are (and are not) recognized, and how reliable the presented numbers are. The reliability is assumed to have an impact on the perceived value because unreliable information is not a good basis for decisions. Depending on how trustworthy the measures are the less margin of safety is needed for the instrument, which entails a higher stock valuation.

Furthermore, PAT describes that management acts opportunistic in the choice of accounting method and that is hypothesized to be considered by investors.

We believe that more reliable firms will be valued higher than less reliable firms, consistent with the findings of Ji and Lu (2014) and Ji (2018). Thus, H1 is the following:

H1: In firms with higher reliability, the value relevance of intangible assets will be higher than in firms with low reliability.

Further, goodwill and identifiable intangible assets can be perceived

differently. Prior studies found differences in how reliability affects value

36 (70)

relevance for the two types (Dahmash, Durand & Watson, 2009; Ji & Lu, 2014). Therefore we test them separately and H1 is subdivided into:

H1a: In firms with higher reliability, the value relevance of goodwill will be higher than in firms with low reliability.

H1b: In firms with higher reliability, the value relevance of identifiable intangible assets will be higher than in firms with low reliability.

5.2 Reliability Matters More for Goodwill Than for Identifiable Intangibles

In H1 we hypothesize that reliability matters for value relevance of goodwill and identified intangible assets, respectively. To further deepen the understanding, we put these in relation to each other to find out whether reliability matter more for one than the other. These results in combination with the dimension of management discretion for the two types in the present accounting standards can contribute with interesting findings for standard setters.

Since goodwill and identifiable intangible assets are treated in different ways, it can be expected that the importance of reliability will differ between the two types of intangible assets. To create a surmise of whether reliability matters more for goodwill or identifiable intangibles, we recall the regulatory framework from Chapter 3.

Current standard setting allows managers some discretion in accounting for intangible assets. However, there is more management discretion for goodwill than identifiable intangibles. Goodwill is always tested for impairment annually whereas most identifiable intangible assets are systematically amortized. Further, goodwill cannot be tested for impairment isolated, instead it is allocated on cash-generating-units, which are tested (Marton, Lundqvist

& Pettersson, 2020). This allocation has been criticized for being subjective

since it is a process of evaluation and as (Davis, 1992 cited in Ibrahim, Said,

37 (70)

Latif & Shukur, 2003 p.107) says: “Goodwill is the most intangible of the intangibles”.

For goodwill, with more management discretion, reliability is expected to matter more for value relevance than it does for identifiable intangible assets.

Hence, the second hypothesis is as follows:

H2: Reliability has larger effect on value relevance for goodwill than for

identifiable intangible assets.

38 (70)

6. Empirical Method and Data

Using the Feltham and Ohlson (1995) model, we test the association between intangible assets and market valuation of firms. We divide the sample into reliable and unreliable and test whether there is any difference in value relevance of intangibles between the groups. Reliability is measured in two ways: Tangibility and Leverage. Furthermore, we discuss the validity of the ways to measure reliability and add a third measure for robustness: Big Four Audit Firm.

6.1 Sample Selection

The sampled firms are all listed firms on the Swedish Stock Exchange, including the following lists: Stockholmsbörsen (Large, Mid & Small Cap), Spotlight, Nordic Growth Market (Nordic MTF & Nordic SME) and First North. Our sample is limited to the listed firms since data on market values for non-listed firms are not available.

Further, access to comprehensive data on listed firms in databases simplifies the data gathering. We believe our results and conclusions can be generalized to non-listed firms as well because theoretically there are only a small difference in what price investors wants to pay for a listed firm compared to if it is not listed.

Table 1 provides information about the sample used in this study. We use a sample period of ten years between 2010 and 2019 to get the newest information while simultaneously getting enough observations after exclusions to get statistical significance.

Besides the falling off for firms in which data is missing, we exclude firms

from the financial sector since they differ from other firms. The nature of firms

in the financial sector makes their accounts significantly higher than other

firms, thus the results will be skewed from those firms. This is consistent with

the handling of prior researchers (Dahmash, Durand & Watson, 2009; Ji,

39 (70)

2018). Lastly, we also excluded firms with negative book value of equity. The final number of observations is 4,334.

Sample Selection Sample

Coverage (2010-2019) 10,720

Less

Firms with missing data (319)

Firms in the financial sector (GICS code = 40) (5,980)

Firms with negative book value of equity (87)

Final number of observations 4,334

Table 1. Sample Selection.

Table 2 presents the different types of intangible assets reported by the final number of firms between different industries. The table shows that the industrial sector is the most common among our sample of firms (23.12%).

The two subsequent industries are the health care (21.37%) and information technology (15.99%).

Industries GICS Firms in sector Goodwill Other intangibles

Energy 10 70 (1.62%) 7 (0.31%) 21 (0.59%)

Materials 15 248 (5.72%) 133 (5.97%) 220 (6.17%)

Industrials 20 1,002 (23.12%) 744 (33.38%) 908 (25.46%)

Consumer Discretionary 25 660 (15.23%) 443 (19.87%) 578 (16.21%)

Consumer Staples 30 171 (3.95%) 110 (4.93%) 154 (4.32%)

Health Care 35 926 (21.37%) 254 (11.40%) 800 (22.43%)

Information Technology 45 693 (15.99%) 396 (17.77%) 644 (18.04%) Telecommunication Services 50 75 (1.73%) 59 (2.65%) 62 (1.74%)

Utilities 55 115 (2.65%) 18 (0.81%) 78 (2.19%)

Real Estate 60 374 (8.63%) 65 (2.92%) 101 (2.83%)

All 4,334 (100%) 2,229 (100%) 3,566 (100%)

Table 2. Industry distribution of sample firms. Global Industry Classification Standard (GICS). The industry “Financials” – GICS 40, is excluded above.

Furthermore, Table 2 also presents that 2,229 firms (51.38%) out of the total

4,334 firms reported goodwill. Goodwill are most common by firms in the

industrial sector (33.38%), consumer discretionary sector (19.87%), and

information technology sector (17.77%).