The Quality of
Information Disclosed
in XBRL
Bachelor Thesis Accountancy and Control
Tanja Langendijk 10331506
Universiteit van Amsterdam Amsterdam Business School
dhr. J.H. ten Berg MSc June 28, 2015
2
Statement of Originality
This document is written by Student Tanja Langendijk who declares to take full responsibility for the contents of this document.
I declare that the text and the work presented in this document is original and that no sources other than those mentioned in the text and its references have been used in creating it.
The Faculty of Economics and Business is responsible solely for the supervision of completion of the work, not for the contents.
3
Abstract
This thesis examines the influence disclosing information in XBRL might have on the quality of information. The quality of information is evaluated through three factors: comparability, information asymmetry, and reliability. XBRL, eXtensible Business Reporting Language, is a relatively new way of reporting financial information over the Internet. The significant growth of the Internet, as well as its use for financial reporting, led to the development of this reporting language. This development has started since the end of the last century and adoption has started in the U.S. since 2008. XBRL is found to have significant benefits to users, but no research has yet been conducted to determine how the quality of information in XBRL is being affected. Therefore, this aspect is the focus of this literature review.
Based on what was found in prior literature, it is found that comparability and information asymmetry, are both increased and decreased as a result of reporting in XBRL. Evaluation of the reasons behind both findings suggests that the increase is more sustainable for both comparability and information asymmetry than is the decrease. The decrease in comparability seems to be the result of the early stage XBRL is in, this is found to diminish overtime thus comparability is overall found to increase. Information asymmetry is overall found to increase, but this effect might also diminish as investor learning may take place. Reliability seems to remain constant as assurance can also be provided on the reports in XBRL, however, audit procedures must still be altered to enable more efficient evaluation of the filings.
Research suggests the quality is and will be affected positively by the use of XBRL. These findings might contribute to prior knowledge by indicating that
disclosing information in XBRL does actually improve the quality of information, this might be helpful in end-user acceptance as it facilitates more efficient
decision-making. More efficient decision-making found to lead to an increase in firm value, it is thereby also beneficial to filers.
4
Summary (Dutch)
Deze scriptie onderzoekt de invloed die het openbaar maken van informatie in XBRL mogelijk heeft op de kwaliteit van informatie. De kwaliteit van informatie wordt
geëvalueerd door middel van drie factoren: vergelijkbaarheid, informatie asymmetrie, en betrouwbaarheid. XBRL, eXtensible Business Reporting Language, is een relatief nieuwe manier van rapporteren via het Internet. De enorme groei van zowel het Internet als het gebruik hiervan voor financiële rapportage heeft geleid tot de ontwikkeling van deze rapportage taal. Deze ontwikkeling is het eind van de vorige eeuw op gang gekomen en de invoering ervan is in de V.S. in 2008 begonnen. XBRL blijkt vele voordelen te hebben, maar er is nauwelijks onderzoek uitgevoerd om te bepalen hoe de kwaliteit van de informatie in XBRL beïnvloed wordt. Daarom is dit aspect de focus van deze literatuurstudie.
Op basis van wat gevonden is in eerdere studies, kan worden gesteld dat vergelijkbaarheid en informatie asymmetrie zowel kunnen verbeteren als
verslechteren als gevolg van rapporteren in XBRL. Evaluatie van de redenen voor beide bevindingen leidt tot de conclusie dat de toename van zowel vergelijkbaarheid als van informatie asymmetrie duurzamer is dan de afname. De afname in
vergelijkbaarheid lijkt een gevolg te zijn van het vroege stadium waarin XBRL zich bevindt, onderzoek wijst uit dat dit effect na verloop van tijd vermindert waardoor de vergelijkbaarheid over het algemeen lijkt te stijgen. Informatie asymmetrie neemt over het algemeen toe, maar ook dit effect kan na verloop van tijd afnemen doordat investeerders mogelijk “leren”. Betrouwbaarheid lijkt constant te blijven aangezien zekerheid ook kan worden gegeven op de rapportage in XBRL, hoewel audit procedures moeten worden aangepast om efficiënte evaluatie van de rapportage mogelijk te maken.
Onderzoek veronderstelt dat de kwaliteit positief beïnvloed wordt, en zal worden, door het gebruik van XBRL. Deze resultaten kunnen bijdragen aan de huidige kennis doordat deze uitwijzen dat de kwaliteit van financiële rapportage daadwerkelijk wordt verbeterd, dit kan helpen bij de acceptatie van XBRL door eindgebruikers aangezien het efficiëntere besluitvorming mogelijk maakt. Onderzoek wijst uit dat efficiëntere besluitvorming kan leiden tot een toename van de
bedrijfswaarde en is daarmee ook voordelig voor de verschaffers van financiële rapportage.
5
Content
Statement of Originality ... 2 Abstract ... 3 Summary (Dutch) ... 4 List of abbreviations ... 6 1. Introduction ... 7 2. Background ... 9 2.1 Financial Reporting ... 92.2 Internet Financial Reporting ... 10
2.3 XBRL ... 12
2.3.1 Development of XBRL ... 12
2.3.2 XBRL Taxonomy ... 12
2.3.3 Implementation of XBRL ... 14
3. Factors Affecting the Quality of Information ... 15
3.1 Comparability ... 16 3.1.1 Disclosures... 17 3.1.2 Taxonomies... 18 3.1.3 Extensions... 20 3.2 Information Asymmetry ... 21 3.2.1 U.S. Studies ... 22 3.2.2 Korean Study ... 24 3.3 Reliability ... 25 3.3.1 Auditability of information ... 26 3.3.2 Security on XBRL-filings ... 29 4. Discussion ... 30 5. Conclusion ... 33 6. Limitations ... 35 References ... 35
6
List of abbreviations
GAAP Generally Accepted Accounting Principles
HTML HyperText Markup Language
IFR Internet Financial Reporting
IFRS International Financial Reporting Standards PDF Portable Document Format
SEC Securities and Exchange Commission XBRL eXtensible Business Reporting Language
XFRML eXtensible Financial Reporting Markup Language XML eXtensible Markup Language
7
1. Introduction
Live stats of the number of Internet users show clearly that the Internet has increased in its use over the last two decades, and it still does. Users are defined as individuals that have access to the Internet at home, these stats do not say anything about the frequency of use. In 1995, only 0.8% of the world population had access to the Internet at home. Nowadays this number has increased to over 40%of the world, whereas most developed countries are even more penetrated. Bermuda is relatively the most penetrated country, 97.75% of its residents was an Internet user in 2014, whereas big countries such as the US and China had a penetration of respectively 86.75% and 46.03%. In absolute terms, this means that there are over three billion Internet users around the globe (Internet Live Stats, 2015).
Due to this increase in accessibility, the Internet can now be called a
widespread medium. This medium has the benefit over traditional mediums, such as newspapers, that it has international reach, whereas newspapers may only reach a local audience. This medium may, therefore, be used as a communication device of financial information to a widespread and international audience. The act of
communicating financial information over the Internet is referred to with terms as Corporate Internet Reporting and Internet Financial Reporting. Financial reporting over the Internet may have a lot of advantages relative to traditional paper-filings such as its widespread reach and its more efficient way of communicating
(Ashbaugh, Johnstone, & Warfield, 1999; Lymer, 1999). However, it is also found to raise some problems as mentioned by Debreceny and Gray (2001). These problems involve resource discovery, attribute recognition, and standards. Respectively these refer to the difficulty of finding information, classifying information and consistent reporting.
XBRL, eXtensible Business Reporting Language, is a rather new way of reporting financial information to stakeholders over the Internet. It is thought of as a way to solve some of the challenges mentioned before. Benefits of XBRL are
discussed in many descriptive articles such as those of Hodge, Kennedy, and Maines (2004) and Jensen and Xiao (2001). XBRL is among others expected to increase comparability an efficiency, and to diminish the information gap between user groups. Those are widespread notions but Debreceny and Gray (2001) even state in their conclusion that reporting mechanisms such as XBRL may improve the value of financial reporting for stakeholders, whereas Yoon and Ciganek (2011) state that
8
stock prices may increase due to XBRL-filings. Such findings suggest XBRL may have a significant impact on how reporting is done and the effect it might have.
The SEC has mandated the phase-in of XBRL during the 2009-2011 period. These XBRL-filings are supplements to the regulatory paper-filings. The phase-in of XBRL means that, in the first filing, only financial statement line items have to be tagged and the notes should only be tagged as blocks. The year after that, the notes should be extensively tagged too. XBRL is not only being developed and
implemented in the US, but also in other developed countries in the world such as Australia, Canada, France, Germany, and Taiwan (Debreceny & Gray, 2001). Even in the Netherlands, the process of implementation has started since 2013 (Pasmooij & Hilvoorde, van, 2012). The system is thus quite widespread already, however, it is only in its beginning stage.
The purpose of this study is to evaluate how the quality of information is being influenced by disclosing it in XBRL. Mainly the following three factors expected to influence quality are being used for this purpose: comparability, information
asymmetry, and reliability. Findings may help XBRL in being accepted in two ways. First, investors may be motivated to use XBRL-filings as high-quality information may affect the efficiency of decision-making. Studies such as those of Debreceny and Gray (2001) and Yoon and Ciganek (2011) suggest that there is a relation between the way reporting is done and firm value and thus stock prices. If a relation between those is in place, this must mean the investment decision is being influenced by the information provided through XBRL-filings. End-user acceptance is necessary for the successful implementation of a system. Second, if information quality tends to
increase due to the use of XBRL for financial reporting purposes, this might cause more efficient decision making which may cause firm value to increase.
This study, therefore, focuses on the question: How is the quality of
information being influenced by disclosing it in XBRL? This question is being
answered by the use of a literature review of scientific articles. The remainder of this paper is organized as follows. Section 2 contains the theoretical framework
necessary to understand the research, it discusses financial reporting as well as how the Internet has changed this, developments such as Internet Financial Reporting and XBRL are being introduced in this part. Section 3 discusses how comparability, information asymmetry, and reliability are being influenced by the use of XBRL. Section 4 contains a discussion about how quality is influenced based on that was
9
found in previous sections. Section 5 provides a conclusion and finally, Section 6 contains some limitations to this study.
2. Background
To be able to research the question provided in the introduction, it is necessary to understand some definitions and mechanisms. This chapter provides a brief overview of what financial reporting is and how the Internet has changed how financial
reporting is being done. 2.1 Financial Reporting
IAS 1 (FASB, 2007) states: The objective of general purpose financial statements is
to provide information about the financial position, financial performance, and cash flows of an entity that is useful to a wide range of users in making economic
decisions. The content of financial reports is standardized through financial reporting
standards such as GAAP – Generally Accepted Accounting Principles – and IFRS – International Financial Reporting Standards. The financial report includes financial statements, notes to the financial statements and possibly other regulatory additions. The financial statements that should be included in the report are the statement of the financial position, the statement of comprehensive income, the statement of changes in equity, and the statement of cash flows. Financial statements are the key part of the financial report since these tend to address the objective from IAS 1, they provide information about the firms’ assets, liabilities, equity, income and expenses as well as gains and losses, contributions by and distributions to owners, and cash flows. Along with the notes, these should provide a reliable view of the firm’s financial position, financial performance, and cash flows.
The need for financial reporting arises from the separation of ownership and control. Ownership is with the capital providers, from the agency theory point of view these are seen as principals. Control is with the management of the corporation, management is responsible for the value creation by means of the capital provided. Managers are typically seen as agents by agency theorists. Owners should be able to control management through investment choices, which demand information. This indicates the need for financial reporting, it enables efficient decision-making.
Efficient assumes no information asymmetry exists, the same information is available to both the principal and the agent. The agency problem encountered at this point is
10
that information asymmetry presumably does exist, this problem explains the value of auditing. Auditing provides assurance on whether the financial report fairly presents the financial position of the corporation in accordance with a financial reporting framework such as GAAP or IFRS. Reliable information is necessary for efficient decision making by capital providers such as investors.
2.2 Internet Financial Reporting
The increase in the accessibility and use of the Internet over the last two decades has caused it to be an effective medium for communication. Mainly this growth in use has been studied by many researchers such as Deller, Stubenrath, and Weber
(1999) and Debreceny, Gray, and Rahman (2002). Deller et al. focused on the widespread use of the Internet in the U.S., UK, and Germany whereas Debreceny et al. focused on the variations in the use of the Internet for financial reporting. Deller et al. found that by January 1998, the Internet was already considered a useful medium in distributing financial information since respectively 91%, 72% and 71% of U.S., UK, and German corporations used their websites for investor relation activities. The objective of those activities is comparable to that of financial reporting as its purpose is to provide information to investors for capital allocation decisions.
Internet Financial Reporting (IFR) is the reporting of financial information through the Internet and this has been done since the mid-1990s. The accessibility and use of the Internet has grown and so has the number of web pages available via the World Wide Web, this raised a number of problems for the way financial reporting via the Internet was done. First, the location of the information was difficult as no standards existed for this, this is the resource discovery problem. Second, there was no standardized schema for the classification of data either, referred to as the
attribute recognition problem. Third, financial reporting over the Internet is done
inconsistently, the standards problem (Debreceny & Gray, 2001).
IFR can be divided into three stages as indicated in a study of the IASC (1999). In the first stage, the Internet is only used to provide reports identical to the traditional paper filings through formats such as PDF. The second stage is
characterized by the same information, only through the use of a format that allows for Web interaction such as HTML. Finally in the third stage, additional information is provided that could not be provided efficiently through the use of paper filings. These
11
stages are used in several studies, among which the study of Debreceny et al. (2002).
PDF, Portable Document Format, is an Adobe Acrobat format used by many corporations for reporting purposes. This format enables firms to define how
information is presented to stakeholders without it being altered. PDF is especially used during the first phase of IFR since its presentation is exactly the same as that of traditional printed filings. The identical reports offered through the use of PDF only beat the traditional filings in being available on the Internet. As previously discussed, the Internet is now a widespread medium useful for communications. However, mass audiences can only be reached through Adobe’s own search engine (Debreceny & Gray, 2001). The benefit of the Internet of having a widespread reach is thus being limited by using the PDF-format.
HTML, HyperText Markup Language, uses metatags. Metatags contain data about the data, information that can be used by search engines. It probably does not solve the attribute recognition problem mentioned before for a variety of reasons such as the tags used for HTML are not standardized, not all search engines use this kind of data although it might be available, financial statement terminology is not standardized and the format has limited semantic power. Semantic power refers to the power to have a meaning, HTML does not give a clear relationship between the tags and what is represented by it, it provides mostly layout information to the browser (Debreceny & Gray, 2001).
The use of the Internet for financial reporting has several benefits when compared to traditional financial reporting. First, the audience to whom reports can be distributed is much greater since Internet use is now widespread among both corporations and with people. Second, since reports do not have to be printed
anymore, this must ultimately lead to cost savings. Third, analysts now often have to re-enter the financial information provided by printed reports for them to be able to perform any analyzes on the data. If this information would be offered through a readily analyzable and comparable format, this would reduce time spent and errors made during this process, and thus increase efficiency.
12
2.3 XBRL
2.3.1 Development of XBRL
XBRL is an XML-based open standard for communicating financial information in a standardized manner. The AICPA first started with some investigations regarding the development of an XML-based financial reporting language in 1998. XML is short for eXtensible Markup Language, it is a mechanism that may help in the attribute
recognition on the Internet through the use of tags. This lead to a prototype of the eXtensible Financial Reporting Markup Language (XFRML), as it was then called, by the end of 1998. The language was then only focused at external financial reporting, based on U.S. GAAP and for commercial and industrial enterprises. Nowadays, this focus has spread, namely to other regulatory purposes such as tax returns, to other standards such as IFRS and to nearly all industries (Debreceny et al., 2002).
By the year 2000, the name was changed to XBRL, abbreviate for eXtensible Business Reporting Language. XBRL is thus an XML-based reporting language that uses electronic tags, it is different from XML as this particular language focuses on the communication of financial information instead of providing a mechanism to describe the structure of documents. This tagging of data has the ability to improve the accuracy and efficiency of data extraction and aggregation by providing
information about the data through the tags (Cong, Hao, & Zou, 2014). It is also expected to increase comparability and to decrease the information gap between user groups, thereby improving analyzability and decision-making. The use of XBRL makes information interchangeable between different information systems and thus enhances the exchange of business information (XBRL International, 2015a).
2.3.2 XBRL Taxonomy
The tags used for XBRL-filings contain information about what information is communicated by that data, these are gathered in a hierarchical taxonomy. A
taxonomy is thus like a dictionary for data. Taxonomies contain both concepts, along with their attributes, and relationships between concepts, these relationships make a taxonomy hierarchically structured (XBRL International, 2015c). Attributes of a concept may include elements such as its meaning, its datatype, the period it refers to, whether it is a debit or a credit element, and the class of the financial report it is part of such as Assets on the Balance Sheet. The meaning of a concept is just a
13
short description of what type of values is to be included. The datatype may be values, units, or even shares (Debreceny et al., 2002).
The taxonomy does also include relationships such as calculations,
presentation, and dimensions, these are called linkbases. Calculations refer to how concepts are connected to each other, for example: Cash and cash equivalents +
Current Assets + Non-current Assets = Assets. In this simple example, the
subordination of the line items is given by a parent-child hierarchy, Assets can be seen as the parent whereas the other line items are children. A presentation linkbase defines how the instance document, the XBRL-document that can be read and
understood manually, should look. Such a presentation linkbase can, for example, provide information on how the different line items in the Balance Sheet should be ordered. Finally, the dimensions refer to which structures should be used to support different segments such as the industry it is part of, these dimensions of the
taxonomy may indicate industry-specific values such as those of the Financial
Services segment. Revenues may be indicated in the HTML-filing of the firm as FinancialServiceRevenues, the appropriate tag found in the taxonomy is in that case Revenues in the FinancialService dimension. (Debreceny et al., 2002) Every line item
gets its own tag that refers to the taxonomy and that can be understood, analyzed, used, and so on by computers.
Taxonomies are developed by entities such as regulators, accounting
standard setters and government agencies. They are based on accounting standards such as U.S. GAAP and IFRS. Accounting standards are quite flexible, it leaves room for deviation which causes financial statements to differ between corporations.
Because of that flexibility, not all possible concepts can be identified and contained within the foundation taxonomy. Some concepts contained by the regulatory filings in a format such as PDF or HTML may not be covered by the applicable taxonomy (XBRL International, 2015b). When no appropriate tag is available in the applicable foundation taxonomy, extensions should be developed by the corporation itself. The foundation taxonomy cannot be altered, but an extension taxonomy can be
developed to account for provider-level differences. Extensions are only possible if the taxonomy is open, it encourages firm-level and industry-level differences, as is the case for U.S. GAAP and IFRS taxonomies (Debreceny, Farewell, Piechocki, Felden, Gräning, & Eri, d’, 2011).
14
XBRL taxonomies are standardized although extensions and other taxonomies are allowed. Multiple foundation taxonomies exist since almost every government develops one based on its applicable accounting standards framework. If the
foundation taxonomy is not sufficient for a given corporation, it is expected to develop extensions to give firm-specific information. As indicated before, HTML does not involve standardized tags, which limits the benefit of those tags. At this point, XBRL appears to be a clear improvement as compared to HTML (Debreceny & Gray, 2001).
By combining all tags in an XBRL-document, an instance document can be created that can be exchanged over the Internet and read by computers. An instance document can be used by stakeholders such as investors and analysts to analyze the financial information contained by the tags. By the use of a style sheet,
information can be put in each format by every user individually to make the tags visually readable. This enhances the accuracy and efficiency of analyzing since it is not necessary anymore to re-enter every element of the financial report for it to be analyzable. Accuracy is improved since re-entering may come with typos, and so is efficiency as less time is wasted to the mechanics aspect of analyses which involves the retrieval and gathering of information.
2.3.3 Implementation of XBRL
Although research had already taken off since 1998, it was not until 2005 that the first filings in XBRL-format had been done in the U.S.. These were voluntary filings and were done through the Voluntary Filings Program (VFP), this program was meant to help stakeholders gain an insight in and improve the process of filing in XBRL. The filings done during this VFP were seen as not filed, this created a limitation of the liability (Bartley, Chen, & Taylor, 2011).
As of June 2009, the SEC started mandatory filing in XBRL through a phase-in process. This process meant that the largest filers were mandated to file their
financial reports in XBRL for the reporting period ending on or after June 15 of that year. The first filing was characterized by tagging all of the line items of the financial statements and only blocks of the disclosures in the notes. The second filing was characterized by the same tagging-requirements for the financial statements, but the disclosures had to be tagged in detail now too. This meant that every element
15
started from June 2011 for the remaining filers and followed the same pattern as it did for the first filers. Some corporations already started filing as off April 2009, this was voluntary. The phase-in process enjoyed a limitation of the liability as was the case during the VFP, they were not to be held liable for failures to meet tagging-requirements during that period. Yet, they were expected to correct such an error immediately as it was detected (Debreceny et al., 2011).
An important notion made by the SEC is that XBRL-filings should be similar to the traditional filings of the corporation, possibly in a format such as HTML or PDF. Both numerical and non-numerical facts contained by the official HTML-filing should also be contained by the XBRL-filing, this covers the full financial report including both financial statements, notes, and other regulatory additions. Those facts should also reflect exactly the same information (SEC, 2015). This creates the need for extensions as XBRL-filings should be aligned with the traditional filings and some line items found in the HTML-filings may not be covered by the foundation taxonomy.
XBRL may eventually even be used to tag the transactions recorded in an information system within a corporation. This would allow the system to
independently aggregate those transactions into a financial report by the use of the information provided through those tags. It would be able to exchange and analyze the financial report without the involvement of any people. This kind of XBRL is called XBRL Global Ledger, which also has the benefit of not losing any information due to aggregation of the data. Individual transactions can still be “seen”, along with their details (XBRL International, 2015a).
3. Factors Affecting the Quality of Information
In this section, three factors that influence the quality of information are evaluated. Those factors are comparability, information asymmetry, and reliability. The quality of information is related to the objective of financial reporting, which is to provide users with useful information for economic decision-making, as was also discussed in Section 2.1. Users of financial information include internal, users within the
corporation, as well as external parties, but the implications for external parties are mainly discussed in this thesis. External users may include parties such as investors and analysts, comparability, information asymmetry, and reliability are expected to be important to them for the following reasons.
16
Decision-making consists of several stages, one of which is the evaluation of alternatives (Schwenk, 1984). Comparability is a factor necessary for efficient evaluation of alternatives. Information asymmetry refers to the case in which information available differs between parties, an information gap exists. This gap causes the less-informed party unwilling to trade at a low cost, the market is
inefficient. Users benefit from trading at low cost, thus information asymmetry should be as low as possible for the sake of external parties. Finally, reliability refers to the note that information should be useful, useful information should be relevant as well as reliable. Users would not base their decisions on unreliable information, unreliable information may, therefore, lead to inefficient markets. This does also explain how auditing may have value in providing users with information. Both comparability and information asymmetry are explicitly mentioned as to be benefits to XBRL, XBRL is expected to respectively increase and decrease both factors, this assumes that both are influential to economic decision-making.
The three factors mentioned above are evaluated in many articles individually and are discussed in the subsequent subsections. Other factors that may affect quality are timeliness, accessibility, how information can be used, and relevance. These factors are found to be improved through XBRL. Timeliness, accessibility, and how information can be used are discussed by Ashbaugh et al. (1999). Timeliness and accessibility are related to the use of the Internet for communicating purposes, timeliness is increased as only a limited time gap should have to exist between the formation of information and the uploading on the Internet. Accessibility of the information is increased as the accessibility of the Internet in general has
aggressively increased over the past two decades. How information can be used is related to the customization that is described by Jensen and Xiao (2001). As
information can also be used to analyze data, through the use of style sheets, customization is reached as all users can individually determine what information should be presented. Therefore, no extensive research on these factors has been done.
3.1 Comparability
The research of Debreceny et al. (2011) refers to a note made by the CFA Institute in a report from 2009, this report states that timeliness, transparency, comparability, and consistency are requirements investors set for information. Bovee, Ettredge,
17
Srivastava, and Vasarhelyi (2002) also note that SFAC No. 2 states that usefulness of information is affected by, among others, comparability. This indicates that there exists a need for comparable information among stakeholders. Decision-making under uncertainty, as is the case in the real world, involves the activities of goal formulation, problem identification, alternatives generation and evaluation and selection (Schwenk, 1984). Evaluation of alternatives is only possible if alternatives are comparable, this explains the need for the comparability of data provided by XBRL-filings. Comparability is discussed in articles such as those of Bovee et al. (2002), Debreceny, Chandra, Cheh, and Janvrin (2005), Vasarhelyi, Chan, and Krahel (2012), and Bonsón, Cortijo, and Escobar (2009), and is evaluated in this subsection.
3.1.1 Disclosures
Hodge et al. (2004) did a research on how the transparency of information is being influenced by the use of facilitating technologies. An example of a search-facilitating technology is XBRL, the tags assigned to every line item of the financial report can be used by such a technology to retrieve the information within the report. It can also be used to retrieve related information, such as information in disclosures related to information in the financial statements itself. Related information can be used to direct users’ attention as well as to clarify managers’ financial reporting choices by simultaneously presenting it. The degree to which these benefits do actually improve with the use of such a technology is discussed in this article.
Accounting policies vary between corporations, this makes it harder to
compare since the same information may be disclosed in another way. The research (Hodge et al., 2004) notes that previous research points out that financial statement users’ reactions towards recognized information, thus in the financial statements itself, are stronger than towards disclosed information, disclosures in the notes to the statements. If the research on search-facilitating technology concludes that no
significant difference exists between the value given by participants to respectively disclosed and recognized information, this technology thus may be helpful in comparing information. Cross-firm comparability is then increased as a result of an increase in the transparency firms’ disclosure policies. Higher transparency means that firms benefit less from different disclosure policies as these are now highlighted, different policies can be compared to each other since related information can be
18
presented simultaneously. Conclusions from the research do point out that search-facilitating technology may help users to evaluate and combine the information from disclosures with the financial statement information. Cross-firm comparability is therefore expected to be increased by the use of a technology such as XBRL.
3.1.2 Taxonomies
The power of XBRL over formats such as HTML is stated to be the existence of standardized tags, contained in taxonomies (Debreceny & Gray, 2001). Taxonomies are developed for different reporting purposes such as external reporting and tax return, so more than one taxonomy exists. Mostly used are the U.S. GAAP and the IFRS taxonomy since these are based on widespread accounting standards. It is important to know if what is contained within the taxonomies is also what firms try to disclose, the fit between the taxonomy and firms’ reporting practices.
The fit referred to in the previous paragraph is investigated by Bovee et al. (2002) for the XBRL-taxonomy from the year 2000 based on U.S. GAAP. Only financial statement line items were used in the sample, these were manually
reviewed on whether they existed within the given taxonomy. “Not-existing” line items were qualified as special attention items, SAIs. These were then divided into three categories: potential new tag items, grouped items, and firm-specific items. New tag items may be necessary to be included in the taxonomy, this creates taxonomy migration which causes taxonomies to grow. Grouped items refer to extensions that aggregate concepts included in the given taxonomy, and firm-specific items refer to filer-specific disclosures (Debreceny et al., 2011)
Descriptive statistics from the research (Bovee et al., 2002) gives reason for some significant conclusions. The existence of SAIs suggests that improvement can be made on the taxonomy, taxonomy migration causes taxonomies to grow by adopting new concepts that have not yet been addressed by the taxonomy. Most SAIs, or extensions, are found within the statement of cash flows, which indicates that this statement has the least fit with the 2000 taxonomy. Although this statement has the least fit, the type of extensions between the statements differ. For example, most grouped items are found within the balance sheet, which indicates that these are most aggregated compared to the taxonomy. Extensions are also evaluated by industry, results confirm that some industries’ reporting practices have a better fit with the taxonomy than others do. That result indicates the need for industry-specific
19
extensions, those may decrease cross-industry comparability but they may increase comparability within an industry. Although extensions to the taxonomy might be necessary, the overall fit is good since only 15% of the line items could not be tagged with concepts from the given taxonomy.
The taxonomy frequently used in Europe is the IFRS-GP Taxonomy based on the IFRS accounting standards (Bonsón et al., 2009). The same procedures as with the previously discussed research were used, the sample used was obtained May 2006. The results also indicate that SAIs exist, and thus improvement of the
taxonomy is possible. Most extensions were found within the statement of changes in equity, this has the least fit. This study also acknowledges the need for industry-specific extensions, especially for the banking and insurance sector which is already contained within the September 15, 2004 updated version but obviously still needs to be extended. It is to be noted that the study also states that many of the deficiencies detected are resolved in the newer versions of the taxonomy.
The existence of several taxonomies may cause comparability to decrease since these rest on different bases (Bonsón et al., 2009). They are based on standards that differ per country, financial statements are for that reason difficult to compare. No mechanism yet exists to compare financial information
cross-taxonomies although this is necessary for comparing firms’ performance. Wenger, Thomas, and Babb (2013) tried to address this problem with an ontological approach. Differences between taxonomies should be mapped by describing how differences relate to each other. These concepts and relationships are gathered in the ontology, which can be used to ‘translate’ financial reports from one taxonomy to another. The growth and change of taxonomies require constant reviewing and mapping of the taxonomies.
The mapping of how concepts included in different taxonomies are related to each other should increase the comparability since this creates the possibility of presenting all statements in the same taxonomy. Comparability may not be enhanced completely as cultural attitudes may make it harder to compare ratios although they reflect the same numbers: some countries may focus on debt financing whereas in other countries equity is preferred for example (Wenger et al., 2013).
20 3.1.3 Extensions
As mentioned in Section 2.3.2, two kinds of taxonomies exist. The first is the foundation taxonomy, which includes most of the concepts of given accounting standards such as GAAP or IFRS. Standardization enhances information comparability, but a need for customized information does also exist, different information for different users (Jensen & Xiao, 2001). Therefore, the second
taxonomy is the extension taxonomy that includes all extensions used by a firm to tag all line items in their financial report. The need for the extension taxonomy rises because the SEC mandated that the XBRL-filings should accurately align with the traditional financial report in a format such as PDF or HTML. Since not all line items included in a corporation’s report may be contained by the applicable foundation taxonomy, an extension taxonomy should be developed. Extensions are both discussed as to enhance predictive and feedback value as well as to decrease comparability.
This decrease of comparability is the main focus of this section. Extensions are the consequences of aggregation or disaggregation of the firms’ disclosures compared to the foundation taxonomy. Complete comparability may be achieved by only using tags found within the foundation taxonomy. However, this level of
conformity is not possible as this would cause XBRL-filings to differentiate from traditional filings. Different kinds of extensions are acknowledged by Debreceny et al. (2011), namely: unnecessary extensions, extensions to reflect filer-specific
disclosures, and extension that aggregate or disaggregate taxonomy elements.
Debreceny et al. (2011) focused on the first filing of firms between April 15, 2009, and June 30, 2010. This sample contained 1,565 filings and 12% of the tags used, or elements tagged, were extensions. Most extensions were categorized as unnecessary, this qualification refers to the case where an extension covers the same information as an already existing concept in the U.S. GAAP taxonomy.
Instead of the extension, the identical concept from the U.S. GAAP taxonomy should have been used, this would have enhanced comparability. One cause for these errors might be that such concepts only exist in the disclosure section of the taxonomy while it is a financial statement concept in the firms’ report. Filers are, however, required by the SEC to use these concepts and only change some of the linkbases. Another cause may be that filers do not use the dimensions that were mentioned in Section 2.3.2 and that allow for segment disclosures with the available
21
concepts. These errors may diminish over time as they may be related to the fact that first-filers are inexperienced with XBRL-policies, -procedures, and -taxonomies.
Extensions to reflect filer-specific disclosures are based on concepts that are not addressed by the U.S. GAAP taxonomy at all. These are necessary to reflect the information provided by the firms’ traditional filings and are stated to be useful
(Bartley et al., 2011). Most of these actually do not offer any new information as it is often only a small variation from what is addressed by existing concepts, or it only gives different subtotals. This might decrease the comparability of disclosures since different subtotals address different issues. Two other important extensions arise from aggregation or disaggregation. Aggregation arises if more elements in the taxonomy are aggregated into one element within the filer’s report. If one element is split into different elements, disaggregation is in place. Comparability is decreased between the “children”, but parent-level comparability is maintained (Bartley et al., 2011).
A note should be made to the influence extensions have on the comparability of information. Although extensions do actually make it more difficult to compare information cross-firm or cross-industry, dependent on the kind of extensions, this effect is expected to be diminished as filers become more experienced with filing in XBRL and taxonomies become more developed (Vasarhelyi et al., 2012).
3.2 Information Asymmetry
Information asymmetry exists when relevant information is known to some parties, but unknown to other. Some parties thus have an informational advantage compared to other, an information gap is existent. This factor is also discussed in the
background-section of this article as the cause of the agency problem and as the need for auditing. Information asymmetry causes markets to become inefficient since not all market participants have access to the information they need for decision-making. The value relevance of information asymmetry is clarified by the fact that performing in a high-quality and standardized format, such as XBRL, tends to
decrease perceived investor risk. Less risk may decrease cost of capital and thereby increase stock price (Yoon et al., 2011). Information is clearly being affected by XBRL as several articles are on this topic, these articles do not always agree on the way information asymmetry is being influenced by XBRL. This subsection, therefore, describes different researches on how information asymmetry is being influenced.
22
According to Cong et al. (2014) the SEC believed new information
technologies, such as XBRL, to be useful in upholding transparency, liquidity, and efficiency. Market efficiency tends to be high if information asymmetry is low as information asymmetry refers to an information gap between market participants. If the information gap between different parties within one market is great, parties may not be willing to trade, which causes the market to be inefficient. Thus, as the study of Blankespoor, Miller, and White (2014) states, the SEC expects information asymmetry to be reduced by search-facilitating technology. Several studies contain research on this topic, some of those provide reasons for an increase in information asymmetry, and others explain how it may be decreased through XBRL-filings.
Information asymmetry is being studied through several measures such as bid-ask spread, price impact, and trading volume and is often combined with measurements of liquidity. The bid-ask spread seems to be the most common measure since it is thought to explicitly capture information asymmetry, it measures traders’ willingness to trade at a low cost. Information asymmetry may cause adverse selection, the more informed party may use its information advantage in making deals which may cause the other party to feel disadvantaged. Adverse selection may eventually cause less informed parties to be unwilling to trade because of their
disadvantage, or they want to be compensated for it. This compensation directly leads to a higher bid-ask spread if the information gap between parties is greater (Blankespoor et al., 2014).
3.2.1 U.S. Studies
Ashbaugh, Johnstone, and Warfield (1999) acknowledged that the effect IFR had on information asymmetry was unknown. This way of reporting had, according to them, potential to decrease this, but only if the Internet was equally accessible to all
investor groups. If the degree of access differs between groups, information
asymmetry may be increased through the use of IFR. Similar expectations existed with respect to reporting through the use of XBRL as is stated by Blankespoor et al. (2014). Their study focused on the market impact of XBRL after it got mandated in the U.S..
Blankespoor et al. (2014) measured the effects that XBRL may have on trading behavior through abnormal bid-ask spread, price impact of trade and trading volume. The research focused on 10-K filings, which is an annual report to the SEC
23
that differs from the annual report to shareholders by providing more detailed information about firms’ financial performance. A comparison was made by using different periods: the five-day window around the filing as well as a five-day window 45 trading days prior to the filing.
In contrast to what was expected, the results of their study show greater bid-ask spread for XBRL-adopting firms compared to non-adopting firms in the initial year. An increase in bid-ask spreads means an increase in information asymmetry since investors are apparently less willing to trade at a low cost, this indicates a greater information gap between market participants. These effects are found to diminish over the period following the filing. Further study was performed over the following two years, the effect XBRL-filing has with respect to the bid-ask spread seems to significantly weaken and the duration seems to have shorten compared to the initial year. This finding suggests that some investor learning may cause
concerns with respect to adverse selection to diminish over time as investors become more familiar with XBRL. Although information asymmetry is not found to decrease, this may be explained by the fact that mandated adoption has only initiated since 2009. The study states that this effect may change as the technology becomes more accessible.
Cong et al. (2014) argue that market efficiency may be promoted by the use of XBRL since the study expects the creation of idiosyncratic information, “private information”, and the infusion of this information to be improved. This improvement should facilitate the formation of prices as more information is available, thereby promoting market efficiency. It is argued that information asymmetry increases when investors trade to take advantage of the idiosyncratic information, this should mean that information asymmetry should be higher immediately after the filing. Since the capabilities necessary for XBRL-filing may differ between firms, this may influence information asymmetry negatively. Those firms that possess superior information-processing capabilities, mainly the bigger firms, are expected to benefit more from the filings in XBRL since the remaining firms will have an informational disadvantage due to resources available. Those disadvantaged parties are expected to be less willing to trade at a low cost, thereby increasing information asymmetry. Again, information asymmetry is measured through the bid-ask spread.
This study contains both 10-K filings and 10-Q EDGAR filings, 10-Q EDGAR filings are quarterly filings of unaudited financial statements to the SEC. The effect of
24
XBRL-filings is only expected to influence the period after the filing and, therefore, does not include pre-filing days as is the case with the previously mentioned study of Blankespoor et al. (2014). The sample periods used for comparison are two days immediately after the filing and an equal period two weeks prior to that. The sample contains filings from June 14, 2008, till June 15, 2012, this particular period is based on the phase-in of mandated XBRL-filing that took place from 2009-2011. Results confirm expectations, information asymmetry is significantly increased after the XBRL-filing, and this finding indicates that XBRL does facilitate the creation and use of idiosyncratic information. The supplementary information now available and used for decision-making seems to improve market efficiency, although information asymmetry is increased with it (Cong et al., 2014).
3.2.2 Korean Study
Since XBRL-adoption is not only an American phenomenon, studies have also been carried out outside the U.S.. The SEC seems to be the most progressive supervisory authority as it is mentioned greatly in studies, but actually, mandated XBRL-filing was already adopted in other countries before 2009. One such country is Korea, which adopted a mandatory filing program in XBRL format as of October 1, 2007. Yoon et al. (2011), therefore, mention in their study that the Korean stock market seems to be appropriate to investigate the initial effect of XBRL-adoption. Most Korean stock market characteristics are comparable to those of the U.S. stock market, with exceptions such as how transactions are executed.
XBRL leads to an increase in disclosure, the authors, therefore, expect to find a decrease in information asymmetry since prior literature finds that higher level corporate disclosure leads to a decrease in information asymmetry. This expectation seems intuitive since more information available, in a way that is better retrievable and analyzable, should reduce the information gap between parties. Another prognosis prior to the research was that large firms would benefit more from XBRL-filings than do small firms. This is based on the fact that larger firms attract greater capital providers that are more concerned with the future and thus evaluate their performance more accurately. Large companies often operate in more than one industry and, as is discussed in Section 3.1.1, reporting procedures are clearer
through the use of intelligent-software technologies such as XBRL. This indicates that cross-firm comparability is increased with the use of XBRL and also between
25
industries, although exception taxonomies may infuse this. Assuming that tagging is done accurately and taxonomies are well-formed, larger firms are expected to benefit more.
Again, the bid-ask spread is used as a proxy for information asymmetry. For this Korean study, findings suggest that information asymmetry after XBRL-filings has indeed decreased when compared to the period before those filings. Findings also show that the decreasing effect XBRL-filings have on information asymmetry is only significant for large firms, not for medium or small firms, as is also consistent with what was expected. XBRL thus has the effect of decreasing information asymmetry for large-sized firms, an effect that is not market-wide (Yoon et al., 2011).
XBRL is both found to increase as to decrease information asymmetry. An increase was found in U.S. studies whereas the Korean study reports a decrease. These findings are not likely to be related to the country in which the research was performed as the Korean study indicates that most market characteristics are the same. The difference may be the result of the fact that the Korean study focuses more on the amount of disclosures, whereas the U.S. studies focus on the
information provided through disclosures and on the providers of information. The amount of disclosures is more related to general financial reporting over the Internet whereas the other two points of focus are related to XBRL itself.
3.3 Reliability
In 2009, the AICPA published the Statement of Position Reliability 09-1, this publication acknowledged that data contained in XBRL-files must be accurate and reliable in order to be useful (Srivastava & Kogan, 2010). Reliability, as well as accuracy, can be ‘assured’ by means of third-party assurance, this may enhance investor trust (Bartley et al., 2011). Reliability may thus influence the perceived quality of the information disclosed.
Reliability of XBRL-filings depends on two things: the auditability of information provided as well as the security of that information. The first aspect reflects the
agency problem in which the management should justify their actions towards stakeholders as they are the owners of the firm. This problem has already been indicated in Section 2.1 and indicates the need for auditing. The second aspect refers to the fact that the Internet is an insecure medium, security mechanisms should be
26
used to protect against unauthorized access. Both aspects are discussed in the following subsection, with the main focus on auditability.
3.3.1 Auditability of information
Auditing is the act of providing financial statement users with an opinion about whether those financial statements are stated in accordance with specified criteria. Those criteria include standards such as GAAS. The opinion given is to provide reasonable assurance that the statements are presented fairly, in all material respects, and give a fair view of the firm’s financial position in accordance with an applicable reporting framework. Auditing generally increases the trustworthiness of the source and perceived quality is higher if the source is trustworthy (O'Reilly, 1982), this indicates the need for reliable information.
In 2005, the SEC started a Voluntary Filings Program for XBRL-filing, the objective of this program was to help stakeholders gain an insight in and improve the process of filing in the XBRL-format. Bartley et al. (2011) acknowledge that problems in filings may result in decreased credibility and lost confidence in the data, thereby discouraging filers as well as users to use XBRL-filings. This, again, indicates the need for reliable information, otherwise the information would not be accepted and used. The voluntary filings issued during the 2006 VFP were studied to determine their accuracy and to compare those to later filings in 2008. Errors are indicated as the differences between the official 10-K filings and the voluntary filings. Six different errors have been classified: missing elements, incorrect amounts, sign flips, duplicate elements, incorrect elements, and incorrect display. The latter does not influence data provided, but it does create differences between the two filings, thereby reducing user confidence and thus affecting reliability (Bartley et al., 2011).
Results show that the 2006 filings contained relatively many errors, among which most are display errors, those do not actually affecting amounts indicated. When compared to 2008 filings, the number of errors seems to have diminished significantly from 478 to 200. Again, most errors are classified as an incorrect display. Explanations to the improvements in 2008 filings include the following: filers may have used feedback from the SEC about previous filings, software may have been improved thereby automatically detecting some errors, improvements in technical elements of XBRL that reduce errors resulting from complexities in XBRL, increased familiarity with XBRL-filing, and improvements made to the taxonomy. Errors do,
27
however, still exist due to the complexity of XBRL, thereby damaging the reliability of filings. This indicates that third-party assurance may be necessary to gain full
benefits of XBRL.
Debreceny, Farewell, Piechocki, Felden, and Gräning (2010) focused their research on the data quality of the first round of mandated XBRL-filings. The main focus was on the correctness of the mathematical relationships implied by the
taxonomy and used in the instance document. This study does also acknowledge six different errors, but some classes differ from the ones found by Bartley et al. (2011). The six groups include the exclusion as well as the inclusion of values in a
calculation, wrong fact values, debit/credit reversal, rounding errors and errors in the HTML 10-Q filing. The latter indicates that this study focuses on 10-Q filings instead of 10-K filings since most filings found were 10-Q, only five 10-K filings had to be excluded. Although 20% of the filers had already filed during the VFP, 26% of the filings included errors. The number of errors for these voluntary filers were even greater than for first-filers. Most errors are debit/credit reversals, followed by the exclusion of values, incorrect values, the inclusion of values, rounding errors, and finally errors in the HTML filing. The nature of errors found in the sample may clarify how reliability can be improved and audit can be conducted.
During the VFP, provisions were given by the SEC to limit the liability filers had. These provisions included that filers had to make an effort in complying to XBRL-requirements, and they had to correct failures within 24 hours after discovery. If these requirements were met, the firm could not be held responsible for failures in those voluntary filings (Bartley et al., 2011). After that period, auditing was still not required as the phase-in period of mandating enjoyed the same limitation of liability as the VFP did (Debreceny et al., 2011). No assurance on XBRL-filings is required in order to encourage filers to comply with SEC requirements without additional costs of third-party assurance. However, third-party assurance may be necessary to achieve end-user acceptance.
28
The AICPA publication SOP 09-1 provides some guidance as to how XBRL-tagging procedures should be done to improve completeness, accuracy, and consistency. Such assurance can be given with the help of practitioners such as CPAs, but no assertions exist as to how assurance should actually be done (Srivastava & Kogan, 2010). The main action auditors should take is to give an opinion about whether the XBRL-filing is a true representation of the official HTML-filing (Plumlee & Plumlee, 2008). Srivastava and Kogan (2010) identified ten audit assertions that should assure the true representation, these are presented in Fig. 3.1.
Assertions are the claims made by management regarding the financial statements, these claims should be tested to be able to provide an opinion about whether the financial statements are presented fairly, in all material respects, and give a fair view of the firm’s financial position in accordance with an applicable reporting framework. When compared to the HTML-filing, the data in the XBRL-filing should be complete, existent and accurate. Accuracy involves both accuracy in how tagging is done and whether the HTML-filing is accurate. The XBRL-filing should be well-formed, valid, and it should be a proper representation of the HTML-filing. The tags used for the XBRL-filing should comply with all XML- and XBRL-rules.
Assertions should also be made on the taxonomy itself, the taxonomy, extensions, and linkbases used should be appropriate, and extensions should also be valid. Evidence should be collected to provide reasonable assurance that these assertions are true, so by identifying these assertions, a base is given for the evidence that should be collected (Srivastava & Kogan, 2010).
29
Numerous problems are being encountered when third-party assurance should be given on XBRL-filings, mainly since no applicable assurance standards or guidelines exist. XBRL has the advantage of enabling selection of relevant
information and analyzation in the appropriate format by using style sheets. Third-party assurance is traditionally given on the statements as a whole, thus it provides a fair representation of the financial position when viewed overall. This assurance does not indicate that fair representation is assured too when line items are presented individually. Generally accepted auditing notions and processes should be translated when applied to XBRL-filings. Such notions include errors, materiality, sampling, and risk (Plumlee & Plumlee, 2008).
Boritz and No (2009) did a mock assurance engagement on the 10-Q filing of the third quarter of 2005 of United Technologies Corporation. Four phases were identified: client acceptance, planning, testing and evidence gathering, and evaluation and reporting, all encountered some problems. Client acceptance is related to the terms of the agreement with the engaging party. Problems are mainly encountered on the terms, XBRL-specialists may have to be approached for help since necessary knowledge may be unavailable within the firm. One requirement made by the PCAOB was that the financial statements to which the XBRL-filing is related should have also been audited by the auditor. The three remaining phases involve problems with respect to notions such as errors, materiality, sampling, and risk, as are also discussed by Plumlee and Plumlee (2008). Those theoretical problems are thus consistent with the findings of this case study.
3.3.2 Security on XBRL-filings
Ashbaugh et al. (1999) stated that IFR was able to provide two kinds of unreliable information, namely unaudited information and fraudulently transformed information. The latter category is the result of inadequate web security. The Internet is an
insecure medium which tends to be secured by message-level security controls such as encryption and authentication to assure confidentiality, integrity, and
authentication, as well as authorization and trust. This category is not the focus of this thesis, although worth studying (Boritz & No, 2005).
30
4. Discussion
This thesis focused on how disclosing information in XBRL influences the quality of information, quality is operationalized by three factors: comparability, information asymmetry, and reliability. The increase of comparability and the decrease of
information asymmetry are explicitly mentioned as benefits of XBRL. The objective of financial reporting involves providing users with useful information for economic decision-making. Useful information should be relevant as well as reliable, relevance is mentioned as a benefit of XBRL since this search-facilitating technology creates the possibility of customization of the information offered by the use of different style sheets. Reliability of information is doubtful as traditional audit procedures are not appropriate for assurance on XBRL and the insecurity of the Internet demands security controls.
Comparability of the information disclosed is an explicitly mentioned benefit of XBRL, this factor is found to be affected both positively and negatively. First, the way information being disclosed depends on the applicable accounting standards as well as on the accounting choices made by management as standards leave some room for deviation. Search-facilitating technologies, such as XBRL, are found to be helpful in recognizing these accounting choices. Previous studies pointed out that users reactions towards recognized information are stronger than towards disclosed information. However, the study of Hodge et al. (2004) indicates no significant difference between the value given by participants to disclosed and to recognized information. This implies that XBRL is more efficient, than other formats, in providing search-engines with data to integrate information from the disclosures with
information from the financial statements. Those choices seem more transparent to users when such a technology is in use, thereby increasing cross-firm comparability as no benefit can be attained through reporting choices. Second, taxonomies
represent the standardization of tags, which should enhance comparability but lacks with formats such as HTML. The existence of multiple taxonomies, based on different accounting standards or developed for other purposes, may actually decrease
comparability as bases are not the same. Therefore, it may be necessary to develop an ontology in which taxonomies are mapped and thereby allows for “translation” of financial reports into one comparable taxonomy (Wenger et al., 2013). The U.S. GAAP and the IFRS taxonomy are the most developed taxonomies, and these seem to represent what is actually disclosed by firms quite well (Bovee et al., 2002).
31
Industry-specific extensions seem necessary to reflect concepts only used within some industries, such as the banking and financial sector. Such extensions should enhance cross-firm comparability within a given industry (Bonsón et al., 2009). Lastly, extensions hurt comparability since those differ between firms or industries,
dependent on which kind of extensions is referred to. The creation and use of extensions may actually diminish as XBRL becomes more developed for two reasons. First, filers are still inexperienced in XBRL-tagging, thereby creating
numerous unnecessary extensions (Bartley et al., 2011). Second, taxonomies are still not completed, extensions may be adopted in the foundation taxonomy which would result in fewer extensions to be created (Debreceny et al., 2011).
Information asymmetry is clearly expected to be influenced by the use of XBRL as its decrease is mentioned as one of its benefits. Although this suggests that information asymmetry should decrease as a result of XBRL, studies show that this might not be the case. Information asymmetry refers to an information gap between market participants that might hurt trade, thereby creating inefficient markets. XBRL-filings should result in a higher level of disclosures, which is a result of using the Internet for reporting purposes, more information available diminishes the information gap. This effect actually only shows to be significant for large firms as these benefit more from transparency (Yoon et al., 2011). The creation and use of idiosyncratic information, “private” information that becomes available immediately after the filing, is expected to increase, thereby disclosing additional information. Investors will try to take advantage of this, thereby increasing information asymmetry immediately, however, this effect will diminish (Cong et al., 2014). Another reason for an increase in information asymmetry may be that the access to the Internet (Ashbaugh et al., 1999) as well as the information-processing capabilities between firms differ. The consequences of such provider-level differences may be greater for XBRL than for formats such as HTML or PDF as XBRL-filing is more standardized through policies and procedures which makes filing more strict and complex. Firms with more
resources available benefit more from XBRL, thereby enlarging the information gap (Cong et al., 2014). It should be noted that this increase may result from the early stage XBRL adoption is currently in. Although development has already started since the end of last century, the existence of such a search-facilitating technology does not automatically lead to user adoption. Wide publicity should be given to XBRL for it
32
to be accepted and used intensively (Hodge et al., 2004), as is now been done by mandating use.
Reliability of information is necessary for it to be useful and thus to achieve end-user acceptance. Third-party assurance can be useful for enhancing reliability, assurance may be necessary as research finds that errors are still existent in most filings. No auditing standards yet exist for XBRL-filings for various reasons. Several kinds of errors exist that should either be evaluated automatically or manually, Bartley et al. (2011) as well as Debreceny et al. (2010) separately identified six
distinct error groups within XBRL-filings. Although errors were found to decrease over time, assurance still seems to be useful since errors remained present in later
studies. Auditors should give an opinion about whether the XBRL-filing is a true representation of the official HTML-filing. The content of the official HTML-filing should have already been audited by the same auditor, the assurance on the XBRL-filing only provides an opinion about its alignment with the HTML-XBRL-filing. Additional assurance is thus necessary when filing in XBRL, compared to HTML. One problem related to assurance on XBRL-filings is the degree to which an error may influence user decision-making, materiality. In traditional audits, materiality is used for the financial statements as a whole. Since XBRL is data-centric, data can be used outside off the context of the instance document, overall materiality has little value. Another problem with this notion is that it is a quantitative term, tagging is qualitative which makes translation even harder. Another problem involves the audit risk model, this model includes audit risk and the risk of material misstatement which is the result of control risk and internal risk. It is not clear how such notions should be used on the XBRL-filings, control risk may relate to how tagging is been done and inherent risk may relate to the type of tags used. The risk of material misstatement may decrease the amount of evidence that should be gathered in order to provide an opinion about whether the XBRL-filing is a true representation of the HTML-filing, thereby affecting the sample. Sampling is done to reduce the amount of evidence that should be investigated in order to provide reasonable assurance that these assertions are true, thus that tagging is done appropriately. Sampling should be based on risk and on amounts, since the tags and amounts are not related, no reasonable sampling can be done. If no base can be found to sample upon, a full investigation of all tags
should be done, which would be expensive (Plumlee & Plumlee, 2008). Auditability of information may not be sufficient for it to be reliable as the Internet seems to be an
