Value Relevance of Other Comprehensive
Income
In a financial setting in the United States of America
Name: Rizwana Butt
Student number: 11424478
Thesis supervisor: G. Georgakopoulos
Date: August 20, 2018
Word count: 12,843
MSc Accountancy & Control, specialization Accountancy Faculty of Economics and Business, University of Amsterdam
Preface
For the last seven months, the author has been working on this paper. During this period, the author tried to contribute to prior study just to focus on one specific industry, namely the financial industry. Before the author has started with this paper, she did not had any knowledge about this subject. So, the last couple of months were very interesting for the author, in which she learned a lot. Finally, as stated on the next page, this paper is my work. I used the study of other researchers to motivate my choice.
During the period I had supervision of my supervisor Georgios Georgakopoulos. I want to thank him from the bottom of my heart for providing me important and relevant feedback. He was almost every time available, which I appreciated a lot.
Abstract
This thesis examines the value relevance of other comprehensive income (OCI) and its elements. The research is done for the financial industry in the United States of America (USA). The research period contains 6,909 observations, from year 2006 until year 2016. Only one database was used during this research, namely CRSP/CompStat merged database. This research expand on prior literature, but only put focus on the financial industry of USA. The outcome show that some elements of other comprehensive income are value relevant namely, minimum pension liabilities adjustment, net income, and unrealized gains/losses on marketable securities. Moreover, this research contains an additional test to explore if all other comprehensive income elements are value relevant. Which are stated in a financial statement. Unfortune the outcome presented that this was not the case. Because special items are only value relevant for the sample that was taken for this research. This research contributes to previous knowledge.
Statement of Originality
This document is written by student Rizwana Butt 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.
Content
1. Introduction ... 5
2. Literature review & hypothesis ... 8
2.1 Value relevance ... 8
2.2 Comprehensive income ... 9
2.3 OCI in financial statement ... 10
2.4 Literature review of prior research ... 11
2.5 Hypothesis ... 17
2.5.1 OCI elements’ value relevance ... 17
2.5.2 Special items’ value relevance of OCI and its elements ... 18
2.5.3 Capital ratios and value relevance of OCI and its elements ... 18
3. Research design ... 19
3.1 Sample collection ... 19
3.2 Value relevance ... 20
3.2.1 OCI elements ... 20
3.2.2 Additional element ... 21
3.2.3 New financial regulations ... 21
4. Results ... 24
4.1 Correlational and descriptive statistics ... 24
4.2 Value relevance ... 25
4.3 Additional element ... 27
4.4 New financial regulations ... 28
5. Conclusion and discussion ... 30
6. Bibliography ... 33
7. Appendix ... 40
7.1 Frequency distributions ... 40
1. Introduction
Within the financial industry, the subject of the value relevance of other comprehensive income (OCI) and its elements has not been examined (e.g. Biddle and Choi, 2006; Skinner, 1999), though prior research has found that OCI is more valuable for financial industries than for other industries (BEA, 2011; Biddle, 2004; Biddle and Choi, 2006; Cahan et al., 2000; Dhaliwal et al., 1999; Hirst et al., 2004). For this reason, this thesis will focus on the value relevance of OCI and its elements in the financial industry sector of the United States of America (U.S.).
The analysis of OCI and how its elements have affected shareholder value has increased significantly due to the Statements of Financial Accounting Standards 130 (SFAS 130). The Financial Accounting Standards Board (FASB) has developed the SFAS 130, which involves OCI reporting. Before the standard was created, OCI had to be reported somewhere in the financial statement, but there was no standard defining how to report OCI. This all changed in 1997 when a clear standard was released by FASB dictating how and where to report OCI.
The main impact of this change is that the elements of OCI must now be reported. Other comprehensive income is defined as an increase or decrease in net assets, and it consists of diverse elements including foreign currency translation adjustments, minimum pension liability adjustments, net income, unrealized gains/losses on derivatives, unrealized gains/losses on marketable securities, and other (additional) components (Financial Accounting Standards Board, 1997).
Subsequent the introduction of this standard, a large quantity of research has examined the explanatory value of these elements. Only a clear and detailed clarification is lacking due to the mixed results of the past research. Dhaliwal et al. (1999), Ohlson (1999), and Skinner (1999) have all come to the conclusion that not all elements are value relevant. On the other hand, Biddle and Choi (2006), Cahan (2000), and Goncharov and Hodgson (2011) have stated that those same elements are value relevant. The main purpose of the present research is that past research has not included the financial industry. In 2011, the financial industry represented 8% of the U.S.’s gross domestic product (GDP). This means that the financial industry is the third largest sector in the U.S. (BEA, 2011), and is therefore an effective part of the U.S. economy.
According to Dhaliwal et al. (1999), marketable securities are the main elements of OCI in the financial industry that need to be clarified for returns. The present research provides evidence concerning the nature of OCI elements, but Ohlson (1999) has stated that it should be transitory, so they extended the previous study to the financial sector. For this reason, the following research question is central to this thesis: are comprehensive income and its elements value relevant in the American
financial industry?
Prior studies have found mixed results regarding whether or not OCI is value relevant. Some studies have found that OCI is not value relevant (Barton et al., 2010; Goncharov & Hodgson, 2011), but others have found a relationship between OCI and stock returns (Biddle, 2004; Biddle and Choi, 2006; Cahan et al., 2000; Dhaliwal et al., 1999; Hirst et al., 2004). Since there is no clear agreement concerning the value relevance of OCI, this thesis provides extra evidence concerning whether or not OCI and its elements are value relevant in the financial industry. Moreover, since the financial crisis that took place in 2008, the financial industry has faced many significant organizational changes. According to Hirst et al. (2004) and Hodder et al. (2006), the financial sector is not entirely satisfied with the present ruling concerning OCI. The timeline of this study contains the financial crisis and its outcome, creating an opportunity to analyze whether or not the new financial regulations are affected by the value relevance of OCI. Other comprehensive income and its elements do not include all value relevant elements (Jones and Smith 2011). For example, special items are considered to be value relevant and serve are an element of OCI. An extra element is therefore included in OCI in order to prove whether or not OCI contains all elements of value relevance. Special items consist of few elements like: nonrecurring items, nonrecurring profit or loss (sale of asset), bad debt expenses, investments and securities, and adjustments apply for previous years (Fairfield, 2009).
The input data for this thesis was gathered using a quantitative research method. The goal was to collect data from an existing database. Thereafter, the information was analyzed using a data analysis program. Only one database, called CRSP/Compustat Merged, was used to gather the input data for this thesis. The data, namely the stock return and financial statements, was collected from this database.
Three hypotheses are tested in this thesis, the first being that all elements of OCI are value relevant. Not all value relevant elements, especially the special items, are individually stated in the financial statement, so the second hypothesis is that special items will decrease the value relevance of OCI and its elements. The value relevance of OCI is higher for organizations that follow the new financial regulations; therefore, the final hypothesis is that higher capital ratios will increase the value relevance of OCI and its elements.
The important outcomes are that besides the element unrealized gains/losses on marketable securities and net income, that the OCI element of minimum pension liability adjustment is also value relevant. Moreover, not every value relevant element is stated individually in OCI financial statements. The outcomes depict that the variable, special items, is also value relevant. Finally, the last hypothesis turned out to be impossible to test due to the small sample of enterprises that do not conform to financial regulations. So the test for hypothesis three is done on the basis of the
remaining sample that was almost the same as hypothesis one. Which says that there is essentially no discrepancy in the value relevance of OCI elements. Because the sample size was not equal, it made comparing difficult.
This thesis consist of the following paragraph. Paragraph 2 explains the literature review and the development of the hypothesis. After this, paragraph 3 consist of research design. Subsequently, paragraph 4 gives the results. And the last paragraph 5, will be closed with conclusion, implications, and discussion.
2. Literature review & hypothesis
2.1 Value relevanceThis subsection discusses the topic of value relevance, which is important because the word is an essential part of the research question.
Kanagaretna et al. (2009) have used a short definition for accounting information, and accounting information is value relevance. They have explained accounting information as any information that is gathered from a firm’s financial statements. In their study, accounting information is value relevance when there is a statistical relation between market returns and accounting information. Even with a small, significant statistical link, accounting information impacts market returns and market prices and is therefore value relevant. Wang et al. (2003) came to the same conclusion, defining value relevance as a statistically significant link between stock returns and accounting information. According to Wang et al. (2003), an OCI element is only considered for stock return if it is also relevant. Stock returns must therefore be correctly stated when information concerning an element is included. This is also in accordance with the research of Amir et al. (2003), who have studied whether the link between market price and accounting measures or market return is higher, since accounting information is included in financial statements. Amir et al. (2003) have found this information to be value relevant.
The previously discussed studies have found that accounting information is value relevant because it strengthens the link between market price and accounting information or market returns. Both Mozes (2002) and Ohlson (1999) have drawn similar conclusions to those of the researchers named above. For this reason, (abnormal) earnings and equity value determine the market price of an organization (Feltham and Ohlson, 1997; Ohlson, 1995; Ohlson, 1999). Mozes (2002) has demonstrated that information is value relevant when it is linked to equity values. Ohlson (1999) has also found that information is value relevant when it is added to the predicted equity value.
Dehning and Ratliff (2004) have defined value relevance more broadly. They have discussed the fact that accounting information decreases investors’ costs for forecasting cash flows and earnings or assists in differently valuing a value relevant organization. Valuing a firm is subjective and can be done in many ways. To ensure that value relevance is treated equally, the following subparagraph defines comprehensive income (CI).
2.2 Comprehensive income
As a later subparagraph will explain, in 1997 the FASB issued the standard Financial Accounting number 130, called Reporting CI. Ten years later, the International Accounting Standard Board (IASB) pursued and issued the corrected International Accounting Standard (IAS) number 1, called Presentation of Financial Statement, in which CI reporting was made mandatory for all listed firms.
Other comprehensive income is a part of CI, so understanding the concept of CI is essential to the remainder of the present research. This subparagraph explains the concept and goal of CI reporting. The FASB defines CI as follows: “The change in equity [net assets] of a business enterprise during
a period from transactions and other events and circumstances from non-owner sources. It includes all changes in equity during a period except those resulting from investments by owners and distributions to owners” (FASB, 1997, p. 6).
The FASB’s definition does not fully explain what CI particularly indicates, yet many researchers have explained, based on this definition, that CI is the sum of OCI and net income (Brimble and Hodgson, 2005; Cahan et al., 2000; Chambers et al., 2007; Wang, 2003). Comprehensive income can be defined as the change in equity of non-owners supplies in a period that includes OCI and net income. The next part of the thesis explains the different elements of OCI.
Compared to net income, OCI is not easy to calculate. It is essential to understand the elements of OCI before calculating the total amount of OCI. “Although total comprehensive income is a useful measure,
information about the components that make up comprehensive income also is needed. A single focus on total comprehensive income is likely to result in a limited understanding of an enterprise’s activities. Information about the components of comprehensive income often may be more important than the total amount of comprehensive income” (
FASB, 1997, p. 7). The FASB says that CI is an effective measure for appreciating an organization’s activities. To provide a better understanding of OCI elements, two meanings of OCI are discussed. Cahan et al. (2000, p. 1273) defines OCI as “all other non-income statement, non-owner related transactions.” Chambers et al. (2007, p. 560), define OCI more broadly: “OCI consists of revenues, expenses, gains and losses that are excluded from net income and are consistent with one of four classifications: (1) foreign currency translation adjustments; (2) available-for-sale marketable securities adjustments; (3) minimum required pension liability adjustments; and (4) adjustments on derivative securities that qualify for cash flow or foreign currency hedge accounting treatment”.
In contrast with Cahan et al. (2007), Thompson et al. (2002), describe eight elements that are reported as OCI. These elements are in line with the four categories identified by Chambers et al. (2007). To complete the explaining of CI, the eight elements of OCI are gains and losses on intercompany foreign currency transactions that are of a long-term investment nature (consolidated, combined, or accounted for by equity method), gains and losses on foreign currency transactions that are designated as economic hedges of net investments in foreign entities, foreign currency translation adjustments, subsequent decreases (if not another-than-temporary impairment) or increases in the fair value of available-for-sale securities previously written down as impaired, unrealized holding gains
and losses on available-for-sale securities, changes in market value of futures contracts under FASB Statement No. 115, unrealized holding gains or losses that result from debt security being transferred, and net loss recognized pursuant to FASB Statement No. 87 as additional pension liability (Thompson et al., 2002, p. 2-3). These components are the same as those defined by Chambers et al. (2007); they are simply explained in more detail by Thompson et al. (2002).
The next subparagraph states what OCI actually is in a financial statement. 2.3 OCI in financial statement
The goal of a financial statement is “to provide information on the financial condition achievements and cash
flows in an enterprise – such information is useful for a wide circle of users who make decisions, and to show the results of the trust function performed by the entity’s management over the provided resources” (Strojek-Filus, 2013, p.
14). The first of January 2009 a financial statement signed on likewise with International Accounting Standards (IAS) and International Financial Reporting Standards (IFRS) is an assertion of an entity’s CI for a period. In that assertion, an economic unit is obligated to add the OCI and the total CI collected in a reporting period, aside from the net profit or loss. Both standard, IFRS and IAS do not give a motive of the following assertion; defining its minimal scope, it contains the following components: “revenues, financial costs, participation in profits and losses of affiliated units and joint ventures evaluated according to the ownership settlement method, tax liabilities, total sum including the amounts of profit or loss after tax on discontinued operations and at the time of fair value measurement, net profit or loss, each element of other comprehensive income classified by type, and total comprehensive income” (Strojek-Filus, 2013, p. 14). Two forms of this assertion are acknowledged if the point is on the framework. Entire incomes and costs are represented either in 1) statements of CI or in 2) statements. Which is divided into profit and loss account and as a statement on CI which starts with a profit or loss and shows the components of OCI (paragraph 81 of IAS1).
The following are examples of OCI: “profits or losses due to revaluation of fixed assets, intangible assets (IAS 16, Tangible assets, IAS 38, Intangible assets), profits and losses resulting from revaluation of the items in financial statements of an entity operating abroad (IAS 21, The results of changes in foreign currency exchange rates), profits and losses due to revaluation of the elements of financial assets available for sale (IAS 39, Financial instruments: recognition and evaluation), the effective part of profits and losses related to hedging instruments within a cash flow hedge (IAS 39, Financial instruments – recognition and evaluation, actuarial profits and losses due to a programme for, specific employee benefits, (§ 93 A IAS 19„Employee benefits)” (Strojek-Filus, 2013, p. 14;15).
2.4 Literature review of prior research Beginning of the topic OCI
Since the topic of OCI and its relevance began to be discussed in 1930, studies have revealed the need for extra OCI elements in order to create a full picture of OCI’s operating performance. According to Paton (1934) and Littleton and Paton (1940), elements should be reported if they impact the income. The assertions of Littleton and Paton (1940) are supported due to a report, namely Reporting the Results of Operations (number 9), which was issued by the Accounting Principles Board (APB) in 1966. The important point in that report is that it is the committees’ opinion that there should be a common assumption that all the components of profit and loss (P&L) are recognized for a period, which assumption is then used in deciding the amount disclosed as net income. An opportunity to this assumption would be with regard to components which, when combined, are materially important regarding a firm’s net income and are apparently not detectable with if it do not include from the current or typical firm procedures of the period (APB, 1966). The FASB has developed this idea further in Concept Statement 5. This statement suggests that organizations have to issue a complete set of financial statements. This includes the cash flows, CI, financial position, investments, and net income (FASB, 1984).
Changing accounting approaches
Beginning in 1990, some changes in accounting approaches (e.g., joint inclusion of hedges and derivates in the financial statements) raised the demand for an independent CI statement to ensure that an independent statement could not bypass the dirty surplus elements of the income statement. At that time, the financial analysts were not pleased with the reporting guidelines for income statements, which was a significant issue (Beresford et al., 1996; Johnson et al., 1995; Johnson and Swieringa, 1996; Foster & Hall, 1996). These issues were recognized in paragraph 40 by FASB as: those two logic were to discontinue the method of taking some elements of CI straightly to equity and to give a vehicle for setting future accounting problems, such as the presentation of unrealized losses and gains linked with financial instruments. The Association for Investment Management and Research (AIMR) noted in that report that “it has a long supported the all-inclusive income idea” (FASB, 1997, p.17). These alterations to the accounting guidelines are aligned with the assumptions of the analysts.
Alternatives choices for reporting OCI
Additionally, there is some resistance to the reporting of OCI. This resistance is primary based on the alternatives the FASB has granted concerning accounting for OCI. The FASB permits three different alternatives for reporting OCI. First of all, it can be stated as a separate assertion, of which
net income is an element. Secondly, it can be stated as two separate assertions. Or lastly, it can be stated in the statement of changing in equity (FASB, 1997). A significant amount of research has been published in the area of which way the OCI is reported. Lipe (1998) and Fitzpatrick et al. (2010) have found that reporting OCI in income statements increases the attention it is given. Moreover, Chambers et al. (2007) have found that when OCI and its elements are reported in the best area of the financial statement, there is a greater chance that investors will take these elements into consideration in their evaluations.
Different research method applied
Chambers et al. (2007), are the first with a different research approach – archival study – to investigate whether or not reporting OCI influences the value relevance. Maines and McDaniel (2000) further explored these outcomes and found that non-professionals were likely to value OCI when it was stated in an income statement. Thereafter, Rees and Shane (2012) have discovered that the findings of Chambers et al. (2007) contradict those of prior researchers who used an experimental approach.
Current research is still attempting to determine whether or not the elements of OCI are value relevant. According to Ohlson (1999), the elements of OCI should be naturally transitory. Unrealized gains/losses on marketable securities and foreign currency translation adjustments are in fact transitory by nature (Barton et al., 2010). Ohlson has also stated that performance measurements are more relevant, especially when the core expenses (operations) are added and the transitory components are excluded. Moreover, Skinner (1999) has stated that OCI could not have an additional significant relation with firm performance.
On the other hand, Mitra and Hossain (2009), have studied the value relevance of different OCI elements, such as foreign currency translation adjustments, pension related adjustments, gains and losses on hedging activities, and unrealized gains and losses from available-for-sale securities. They have used a several cross-sectional regressions to analyze their outcomes. These authors have discovered a statistically important positive link between share returns and foreign currency translation adjustments and pension related adjustments.
Studies conducted in U.S.
According to Black (1993), accounting information should not include transitory earnings, because the earnings should be based on value. Hirst and Hopkins (1998) have also stated that a comprehensible display of CI and its elements guarantees transparent earnings management. Pinto (2005) has researched the value relevance of an OCI element, namely foreign currency translation adjustments. He found that particular element to be value relevant, because it is a useful tool for
providing information to investors. At the same time, while OCI is transitory, researchers have also suggested that it is priced by American capital markets (Cheng, Cheung, & Gopalakrishnan, 1993; Dhaliwal, Subramanyam, & Trezevant, 1999; Choi, Das, & Zang, 2007; Barton, Hansen, & Pownall, 2010; Zülch & Pronobis, 2010). Furthermore, the value relevance of OCI concerning net income is most acceptable when it is investigated where the enterprise is established (Schaberl & Viktoravich, 2015). Although Biddle and Choi (2006) have pointed out that OCI is defined in SFAS 130 as the most meaningful predictive ability related to alternative income measures. Another discrepancy exists between two studies conducted by Isidro et al. (2004, 2006). In 2004, these authors discovered little evidence that eliminating dirty surplus elements would be good for valuation process failures. In 2006, however, Isidro et al. found weak proof that there is a connection between valuation flaws and eliminating dirty surplus elements, though only for American firms. Isidro et al. (2006, p. 341) found that “dirty surplus flows are a consistent source of error in applications of accounting-based valuation models.” After all, none of these components will exist in the organization’s future cash flow. There are, however, researchers who have argued in favor of OCI positively influencing firm performance and demonstrated different outcomes. For example, Dhaliwal et al. (1999) have found that all the elements, besides for the element of unrealized gains/losses on marketable securities, hardly include noise to the idea of OCI.
Studies conducted in European Union (EU)
Lin (2006) has stated that OCI elements are more value relevant than net income when listed in combined statements of OCI and net income of U.K. organizations. Lin, Ramond, and Casta (2007) have also discovered that OCI provides more incremental value relevance than net income for many European Union (EU) organizations. Devalle, Onali and Magarini (2012) have researched the value relevance of CI and OCI elements in the EU, but their research did not find any proof that OCI is a better performance measure than net income, so their conclusion is that OCI elements are not value relevant. Ernstberger (2008) has discovered the value relevance of OCI elements including gains and losses on cash flow hedges, changes in foreign currency translation adjustments, and unrealized gains and losses on available-for-sale financial assets. The research was based on firms that practice American Generally Accepted Accounting Principles (GAAP) or IFRS. Ernstberger (2008), has demonstrated that OCI does not provide incremental value relevant information that net income does not. Still, OCI as IFRS provides more incremental value relevant information compared to OCI as American GAAP. Additionally, only unrealized gains and losses on available-for sale-financial assets such as IFRS provide value relevant inavailable-formation. Moreover, Goncharov and Hodgson (2011) have discovered that some elements of OCI are value relevant, although only when
the elements are measured individually. They have also stated that using unrealized elements in the valuation procedures might create noise through the doubt that value will be created.
Studies conducted others
Brimble and Hogson (2005) have researched the value relevance of CI and OCI. Their research was based on the Australian market from 1988 to 1997. The outcome was that CI had a lower value relevance. Compared to the outcomes of OCI, the elements were incrementally value relevant. Exposure draft
With respect to the financial industry, Yen et al. (2007) have counted the letters commenting on the exposure draft from SFAS 130. The financial industry wrote 93 comment letters, of which nine letters reported OCI, suggesting that the financial industry was not comfortable with the rules suggested by the FASB. An action arose because OCI is a low performance measure and is approximately volatile in relation to net income and the concern that two income characters had to be reported. Another comment made was that the banking industry had to adjust its operational strategy in order to reduce the volatility of OCI. Specialists from the Independent Bankers Association of America (IBAA) stated that to prevent eventually volatile adaptions in market value that being reflect in OCI. Other way is to necessarily shorten the ability of, available for sale, protection and abandon yield and the capability to enlarging the financial performances of the organization. The accounting guidelines are driving the strategic and financial decisions of the community banks (Yen et al., 2007).
After the exposure draft
After the SFAS 130 was implemented, however, Dhaliwal et al. (1999) stated that the financial industry’s OCI was the better firm performance measure with respect to the another income measure. Cahan et al. (2000) have also supported this argument, having found that investors have appreciated OCI since the reporting guidelines changed. They have discovered that reporting the separate elements, the focus was on foreign currency translation adjustment and fixed assets revaluation, of OCI has no extra advantage for the investors. Conversely, Hirst et al. (2004) have found that “the full fair value income have a low possibility of misinterpreting interest percentage risk due to OCI.” Moreover, this thesis highlights the difference between hedged and exposed banks. The results suggest that exposed banks value OCI more than hedged banks do. The first argue supports the prior matter of financial industry that volatility of OCI is a concern for the valuation process. This is in agreement with later research done by Hodder et al. (2006), who found that the volatility of the full fair value income is much greater than that of OCI and net income. Otherwise, the study identifies
the concern that OCI is negatively related to unrecognized changes in the fair value of financial tools. The outcomes demonstrate, however, that the elasticity of the full fair value income is underestimated by SFAS numbers 115 and 130.
Study on special items
The other concern is the elements that are present in OCI. Jones and Smith (2011), have distinguished the value relevance of OCI with special items. These authors have discovered that special items and OCI are value relevant, while special items are more important than CI. Moreover, Jones and Smith (2011) have stated that certain elements of special items are temporary. The treatment should therefore be under the same accounting procedure as the treatment for OCI.
Persistence
The author Lipe (1986) discovered that different elements of income have different persistence, and that the size of market reaction to each elements of income is linked to own persistence of each element. According to several authors special items tend to be transitory, which is to have persistence (Fairfield et al. 1996; Bradshaw and Sloan 2002; Burgstahler et al. 2002).
Table 1 outlines the prior literature overview, listing the name of the author(s), the topic, the period in which the research was conducted and the important results.
In general, there is no clear agreement concerning the use of OCI. Today, the major discussion centers on whether or not OCI contains all relevant elements. Either or the accounting method should be maintained to full fair value income analysis. These concerns led to the three hypotheses, which are illustrated in the next subparagraph
Authors Topic Period Findings 1 Biddle and Choi (2002) Value relevance of different
elements 1994 till 1998 All elements are incremental value relevance (+) 2 Biddle and Choi (2006) Value relevance of different
elements
1994 till 1999
All elements are incremental value relevance (+)
3 Brimble & Hodgson
(2005) Value relevance of OCI and CI 1988 to 1997 OCI elements are incremental value relevant (+) 4 Brimble & Hodgson
(2004) Value relevance of OCI and CI 1999 – 2003 OCI elements are value relevance (+) 5 Chambers et al. (2007) Value relevance of elements of
OCI before and after SFAS 130 1994 till 1997 Before SFAS 130 only SEC is value relevance and after SFAS 130, SEC and FCT are value relevant (market returns) (+/-)
6 Dhaliwal et al. (1999) Value relevance of elements of
OCI 1994 and 1995 Only SEC is value relevant (+/-) 7 Cahan et al. (2000) Report individual elements of
OCI is incremental value relevance
1993 till
1997 There is no incremental value relevance for OCI elements (-) 8 Kanagaretnam et al.
(2005) Usefulness of reporting of OCI 1994 - 1997 Only SEC is value relevant for financial industry (+/-) 9 Kanagaretnam et al.
(2009) Usefulness of reporting of OCI 1998 till 2003 Only SEC and Hedge are value relevant (+/-) 10 Ernstberger (2008) Value relevance and OCI
elements 2001 to 2004 OCI under IFRS are more value relevant information than for U.S. GAAP. (+/-)
11 Mitra & Hossain
(2009) Value relevance of different OCI elements 2005 to 2006 Only two elements are value relevance (+/-) 12 Goncharov &
Hodgson (2011) Value relevance and OCI 1991 to 2005 Some OCI elements are value relevance (+/-) 13 Devalle & Magarini
(2012) Value relevance of CI and OCI 2005 to 2007 OCI elements are not value relevance (-) 14 Wang (2003) Value relevance of dirty surplus 1988 till
1997
Reporting under SFAS 130 not value relevant for OCI elements (-)
15 Pinto (2005) Value relevance of foreign currency translation adjustments element
1991 - 1996 The OCI component is value relevance (+)
16 Rees and Shane (2012) Value relevance of 1 OCI element: available-for-sale marketable securities
Not
applicable The OCI component is value relevance (+) 17 Skinner (1999) Critique / review on Dhaliwal et
al. (1999) research Not applicable Did not find the results surprising or interesting 18 Lin (2006) OCI elements and value
relevance 1993 - 1998 The element is significantly, only negatively linked with firm value (-) 19 Jones & Smith (2011) OCI and special items 1986 - 2005 OCI is value relevant (+)
20 Choi and Zang (2006) OCI and analyst forecast 5,200
observations OCI elements are value relevance (+) Table 1: overview prior studies
2.5 Hypothesis
2.5.1 OCI elements’ value relevance
Before the SFAS 130 was issued in 1997, Black (1993) stated that the earnings should not be based on changes in value, but rather on value itself, since it excludes transitory earnings from accounting information. The FASB took the advice by establishing the SFAS 130, which ensures that transitory earnings will be reported separately. In order to attract the investors’ attention, it is important to report OCI individually (Chambers et al., 2007; Fitzpatrick et al., 2010, Lipe, 1998; Maines & McDaniel, 2000). Following Black’s approach, Ohlson (1999) has argued that all the elements of OCI should be naturally transitory. There are only two elements that are transitory by nature: unrealized gains and losses on marketable securities, and foreign currency translation adjustment (Graham & Lin, 2017). For this reason, an additional study has been conducted to provide a clearer understanding of the impact of OCI and its elements on firm performance. As of yet, there is no clear consensus on whether or not the elements of OCI are value relevant. According to Barton et al. (2010), performance measures have a higher value relevance as transitory components like OCI are excluded. Arguments given by Dhaliwal et al. (1999) and Skinner (1999) have suggested that OCI cannot be a significant explanation of firm performance because none of these transitory components exist in expected cash flows, and that the OCI elements hardly add noise to valuation processes.
On the other hand, Cahan et al. (2000) and Dhaliwal et al. (1999) have stated that since the guidelines of the financial industry have changed, the elements are now value relevant. According to Biddle and Choi (2006) and Goncharov and Hodgson (2011), OCI has the highest prognostic ability relative to other income measures, and the elements of OCI are value relevant, but only if they are measured individually. Finally, Isidro (2004) has argued that there is a little evidence that dirty surplus elements account for valuation processes. Since the present thesis’ focus is only on firm performance, the dependent variable will be consistent during the different tests. According to the papers of Cahan et al. (2000) and Dhaliwal et al. (1999) the first hypothesis is:
2.5.2 Special items’ value relevance of OCI and its elements
Special items are reported infrequently; they are one-time events that occur unexpectedly and are not included in the analyst’s forecast. Organizations report these items separately in income statements to accentuate that they are different from OCI, since they are transitory items (Burgstahler et al., 2002) and are less persistent than the remaining elements of income (Bradshaw and Sloan, 2002). Special items include nonrecurring items, nonrecurring profit or loss (sale of asset), bad debt expenses, investments and securities, and adjustments applied for previous years (Fairfield, 2009).
The second hypothesis is based on the research of Jones and Smith (2011), Elliot and Hanna (1996), and Cready et al. (2010). These researchers have stated that additional elements of special components/items should be included in OCI. Adding special items will temper the impact of the value relevance of the OCI.
Meanwhile, there is proof that special items and OCI gains and losses cannot automatically be tempered, because Elliot and Hanna (1996) and Cready et al. (2010) have discovered that some special items persist over time. Cready et al. (2010, p. 1579) have discovered that “consistent with the idea that managers find it attractive to classify ordinary recurring operating expenses as ‘special items.” So the second hypothesis is presented here below:
“Hypothesis 2: Special items will decrease the value relevance of OCI and its elements” 2.5.3 Capital ratios and value relevance of OCI and its elements
One of Hirst et al.’s (2004) findings is that disclosed financial institutional can be valued by OCI, if that financial academy will take OCI into their account. This allows for the development of a clear understanding of the impact capital demand has on firm performance. The expectation is that the analytical power of OCI will increase, since the hedged banks will have decreased their capital ratios (Elliott, 2010; Hirst et al., 2004). With respect to capital ratios, new regulations for financial institutions have come out following the financial crisis (Elliott, 2010). Because of the regulations, organizations are less likely to apply hedge accounting and to take the regulation into account (Elliott, 2010). The final hypothesis is therefore that adding capital ratios will impact the value relevance of the OCI elements. So this will lead to the last hypothesis:
“Hypothesis 3: Higher capital ratios will increase the value relevant of OCI and its elements”. Described above hypothesis mean that all the elements of OCI are value relevant, however adding an extra element, special items, or capital ratios may limit this impact of value relevance. The next paragraph will contain the research design which is divided into the sample collection,
3. Research design
For each formulated hypothesis there will be a model described. With that model the OCI elements can be tested.
3.1 Sample collection
The sample of this research is a firm-year observation. This information was obtained from the CRSP/Compustat database. The collection only includes the financial industry based in the U.S. as observed from 2006 until 2016. If no data was available for one or more variable, it was not included in the sample collection. The study period is broad, so the financial crisis and its impact and restoration are all included. With respect to capital ratios, new regulations for the financial institutions have come out following the financial crisis. Because of these regulations, organizations are less likely to apply hedge accounting (Elliott, 2010). The financial crisis lasted from 2006 until 2009, its aftermath occurred during 2010 and 2011, and the restoration period lasted from 2012 until 2016. Next, the focus of financial institutions is on investing portfolios, because they have more significant amounts of OCI. Unrealized gains and losses on marketable securities, unrealized gains and losses on derivatives, and the value relevance of the foreign currency translation adjustment are therefore expected to be more significant (Dhaliwal et al., 1999). This selection has a shortcoming, however, because it is only useful for the financial industry in the U.S., not for other industries and countries.
The data on OCI elements, special items, and capital ratios was consulted using the CRSP/Compustat database. Stock return data is also obtainable from the CRSP/Compustat database in section of security monthly. Since stock return numbers are monthly based and the financial statement numbers are yearly based, the stock return numbers were converted into yearly data. Following Biddle’s (2004) approach, the stock return will be from eight months prior until three months after the fiscal year. All the variables will be winsorized at the first and 99th percentiles to outliers into account.
There were in total 13,381 data collected. 3,356 observations did not meet the criteria, so I had to delete them from this research. In total 6,909 datapoints were left to test. As this research contains the financial industry, so the, Standard Industrial Classification, (SIC) code starting with 60-67 are included in this thesis. To be more specified the following codes are used:
6020 6099 6162 6282 6351 6512 6798
6035 6111 6172 6311 6361 6513 6799
6036 6141 6199 6321 6411 6531
3.2 Value relevance
To test all variables, a linear regression model will be used, which all the variables will be incorporated.
3.2.1 OCI elements
First of all, the regression tests examined whether or not the OCI elements are value relevant. The OCI elements are value relevant if the elements are related to the dependent variable; in this case, the dependent variable is the stock return. The models of Dhaliwal et al. (1999) and Goncharov and Hodgson (2011) were followed in this thesis in order to prove that the elements are value relevant. Dhaliwal et al. (1999), however, did not take the other adjustment elements into account, while Biddle and Choi (2006) added adjustments in market securities to the other adjustments element. An individual element for the other adjustments and unrealized gains and losses of derivatives was therefore added. The model for OCI elements is as follows:
Sr = α0 + β1 * INt + β2 * MARKSt + β3 * PLIAt + β4 * FORVTAt + β5 * UnrDert + β6 * Addt + εt
In this model, Sr stands for stock return, which is measured by the accumulative return that begins at three months after the fiscal year has come to an end. This reveals the impact of prior findings and decisions of the return on equity (ROE) and follows the studies previously completed by Sloan (1996) and Biddle & Choi (2006). All the other elements are based on fiscal year, because those values are relevant for analyzing the stock return. Stock return was accessible through the CRSP/Compustat database. INt is the net income from an income statement. MARKSt stands for unrealized gains and losses on marketable securities. PLIAt is the adaption for pension liability. FORVTAt stands for foreign currency translation adjustment. UnrDert stands for unrealized gains/losses on derivates. Addt is another element that contains dirty surplus. That element is not covered in any other elements. If the net income element is not included in CI or in unusual component, is called by clean surplus (Hodgson and Mark, 2014). All the elements that are positive correlated with stock returns should be value relevant.
3.2.2 Additional element
The same approach was used for Hypothesis 2, but with the addition of special items, which influence the value relevance of additional elements. According to Jones and Smith (2011), special items go straight into the net income, which is also value relevant. Special items are reported infrequently, the items are an one-time event. As those elements occurred unexpected, and they do not include in the analyst’s forecast. Organizations report the items separately in income statement to accentuate that it is different from OCI, as they are transitory items (Burgstahler et al., 2002). And are less persistence than remaining elements of income (Bradshaw and Sloan, 2002). Special items consist of few elements like: nonrecurring items, nonrecurring profit or loss (sale of asset), bad debt expenses, investments and securities, and adjustments apply for previous years (Fairfield, 2009). The model for Hypothesis two is as follows:
Sr = α0 + β1 * (INt- ISt) + β2 * MARKSt + β3 * PLIAt + β4 * FORVTAt + β5 * UnrDert + β6 * Addt + β7 * ISt + εt
The variables in Model 2 were measured the same as in Hypothesis 1, only not for the net income. In the first model, the net income already consisted of special items, so this was corrected. For this reason, the net income stayed the same as in Hypothesis 1, only the special items were deducted from the net income (INt - ISt). ISt stands for special item elements derived from financial statements. The hypothesis was that adding a special item as an element would not moderate the value relevance of OCI elements.
3.2.3 New financial regulations
The value relevance of OCI elements might be dissimilar for different kinds of financial institutions. Hirst et al. (2004) have argued that OCI elements are value relevant for exposed firms rather than for hedged firms. According to Hirst et al. (2004) and Elliot (2010), exposed bank have greater capital ratios, which leads to a higher value relevance of OCI elements. As bank has a greater capital which leads to a higher net income. And since net income is an element of OCI, this means that the value relevance of OCI element will be automatically higher. Capital ratios will be added as a dummy variable, the following model is:
St= α0 + β1*INt + β2 * MARKSt + β3 * PLIAt + β4 * FORVTAt + β5 * UnrDert + β6* Addt + β7 * D1(T1) + β8 * D2(T2) + εt
The variables in Model 3 were measured the same as those in Hypothesis 1, only not for the dummy variables. If D1(T1), has a value of one, this means that the ratio Tier 1 will have a value higher than 0.06 for capital and/or risk-weighted assets. The variable D2(T12) will have a value of one if ratio Tier 1 plus Tier 2, capital and/or risk-weighted assets have a value higher than 0.10.
(Elliott, 2010; Ingves, 2014).
Tier 1. Capital / risk weighted assets, tier 1 variable is determined as retained earnings plus common stock (Malcolm & Wurgler, 2015). The Basel account standard want that the banks should meet a minimum ratio of regulatory capital to risk-weighted assets (Francis & Osbornce, 2012). Regulatory capital compromises that this tier has a higher quality will leads to disadvantage absorbent capital components like an common equity (Francis & Osbornce, 2012). And Tier 1 + 2. Capital / risk weighted asset this includes other elements with capital like equity and subordinated debt (Francis & Osbornce, 2012).
The author believed that the outcome would be that the capital ratios would increase the value relevance of OCI elements.
The variables that will be used in this research are given below in table 2 along with the Compustat descriptions. All these variables are derived from the Comspustat Database.
Variable: Compustat description:
____________________________________________________________________________
Sr Yearly stock return
Monthly stock return adjusted in yearly data. Period: 8 months prior and three months after fiscal year
INt Net income
Derived from CompStat, named ‘CIBEGNI’
Addt Other adjustments
Derived from CompStat, named ‘CIOTHER’
ISt Special items
Derived from CompStat, named ‘SPI’
UnrDert Unrealized gains and losses on derivates Derived from CompStat, named ‘CIDERGL’
PLIAt Pension liability adjustment
Derived from CompStat, named ‘CIPEN’
INt – Ist Net income – special items
Net income adjust to show the impact of special items > stock return. As the Normal special items are included in net income
MARKSt Unrealized holding gains and losses on marketable securities
Derived from CompStat, named ‘CISECGL’
FORVTAt Foreign currency translation adjustment Derived from CompStat, named ‘CICURR’ Table 2: variables
The variable that is used in this thesis are explained below in table 3: Variable explanation:
____________________________________________________________________________ Net income is nothing else than to define profit or loss in a profit and loss statement (IAS, 2011) Special items are reported infrequently, the items are an one-time event. As those elements occurred unexpected, and they do not include in the analyst’s forecast. Organizations report the items separately in income statement to accentuate that it is different from OCI, as they are transitory items (Burgstahler et al., 2002). And are less persistence than remaining elements of income (Bradshaw and Sloan, 2002).
Unrealized gains and losses on derivatives is explained in ASC Topic Derivates and Hedging report as all derivatives are reported as an asset or a liability in balance sheet at fair value. If there is a gain or a loss from any changes in FV of derivatives is acknowledged directly in current gain or deferred as a component of OCI (Manchiraju et. al, 2016)
Pension liability adjustment occurs when the accumulate profit obligation for the pension plan is higher than fair value (FV) of plan’s asset (Carpenter & Mahoney, 2007)
Unrealized holding gains and losses on marketable securities, equity and debt securities that are not classified as trading security or held to maturity securities are reported at FV and classified as available for sale securities. The unrealized gains and losses are excluded from net income and are reported in a separate element of equity (FASB NO. 115, 1993)
Foreign currency translation adjustment is according to PwC (2016) a process of showing a foreign firm’s practical currency financial statements under the currency reporting section.
The cumulative translation adjustment account includes the translation adjustments and is an element of OCI.
_____________________________________________________________________________ Table 3: explanation variables
4. Results
4.1 Correlational and descriptive statistics
Table 4 depicts descriptive statistics used in regression models. Yearly stock return (RET) was determined as return rate estimated by lowering the year return by 1. The number 0.34 is the mean and indicates that the return rate was almost 34%. The variables UnrDer, FORVTA, Add, and PLIA have approximately low means, because most enterprises do not have sole values for these variables in their financial statements. Nearly every variable is not symmetrical in accordance with the skewness value. Only the following variables are symmetrical: unrealized losses/gains on marketable securities (MARKS), stock return (RET), and unrealized losses/gains on derivates (UNRDERT). Moreover, almost every variable that is not symmetrical is considered a negative value, so only for those variables where the values are low. As a result, they are asymmetrically distributed to the left.
Variable Obs Mean Std.Dev Min Max Skewness Kurtosis
RET 6909 0.3429807 33.78431 -85.21634 121.1577 0.2319897 4.34645 IN 6909 207.9743 979.8372 -658.229 8296.907 6.836608 47.97639 FORVTA 6909 -2.342922 50.29257 -411 276.55 -4.879575 55.69791 UnrDer 6909 0.7553994 43.53548 -293 206 0.9449409 30.75441 Add 6909 -0.4420601 5.754774 -53.322 15.511 -8.512456 80.40276 PLIA 6909 -2.392050 27.39881 -277.439 206 -6.333808 53.3442 MARKS 6909 17.02904 609.5956 -24439.17 10943.77 0.2940319 199.3555 SI 6909 -10.20354 80.42187 -449 198 -7.303434 46.07209 Table 4: descriptive statistics
Table 5 depicts the correlation of Spearman matrix concerning the independent and dependent variables. Regarding the dependent variable, RET, a few independent variables are correlated positively, namely special items (IS), net income (IN), and pension liabilities adjustment (PLIA). The previously mentioned variables have a correlation of 1% that is statistically significant. Approximately every independent variable has a correlation that goes in the expected direction. Because the assumption is that OCI elements impact the stock return, a growth in stock return will increase the independent variable. After all, the variable unrealized gains/losses on marketable securities is negatively correlated with stock return, so the growth in stock return will decrease the value of that variable.
RET IN FORVTA UnrDer Add PLIA MARKS SI T1 T12 RET 1 - - - - IN 0.2616** 1 - - - - FORVTA 0.0223 0.0378 1 - - - - UnrDer 0.0382 0.0493** -0.0430** 1 - - - - Add 0.0051 -0.0344 -0.0054 -0.0578* 1 - - - - - PLIA 0.1218** -0.0277* 0.0702** 0.0692** 0.0051 1 - - - - MARKS -0.0363 0.2852** 0.037 -0.0453* -0.2478** -0.2450** 1 - - - SI 0.0507** 0.0249 -0.0206 0.0089 -0.0113 0.0223 -0.0231* 1 - - T1 0.3723** 0.1233* -0.0089 -0.0043 -0.0050 -0.0503** 0.0232 0.0601* 1 - T2 0.4178** 0.0704** 0.0023 -0.0094 0.0003 -0.0374* -0.0033 0.623** 0.8773** 1
** displays that the correlation is significant at 0.01 level (2-tailed) * displays that the correlation is significant at 0.05 level (2-tailed)
Table 5: correlation matrix
Generally speaking, there is no any significant evidence of multicollinearity in the OCI elements. The correlation values vary between 0.004 and 0.06, which demonstrates that multicollinearity does not exist in the first two regression models. There is, however, a chance of multicollinearity in Tier 1 (T1) and Tier 2 (T2). The correlation of these two variables is 0.8773, which is a relatively high number. This is explained by the fact that the T1 also exists in T2, since T2 is a merger of the T1 and T2 variables.
4.2 Value relevance
Table 6 presents the p-value, t statistics, and coefficient for the value relevance of OCI elements. As a group, OCI elements significantly impacted the stock return (F(6, 850 = 8.43, p < 0.000). Still, the entire model is not so strong, since the R2 is 0.0181. One reason for this could be the long study period encompassing the financial crisis, since enterprises have a greater probability of facing financial distress during a period of this duration and being restored after the financial crisis. For this reason, the data points for enterprises could differ extremely.
Looking at the variables in Table 6, IN, PLIA, and MARKS are 5% statistically significant. The variables IN and PLIA are also 1% statistically significant. The results for two of these three variables conform to prior research conducted by Biddle and Choi (2006), Goncharov and Hodgson (2011), and Dhaliwal et al. (1999). Only the significance of the variable PLIA differs from these studies. One reason could be that at the time of financial crisis, the pension liabilities increased due
to financial anxiety concerning investment enterprises and pension funds. Another possible reason is that investors priced PLIA higher because it could be important for organizations to either balance or increase those funds, which would lower the total OCI (Impavido & Tower, 2009).
Independent variable Sr = α0 + β1*INt + β2*MARKSt + β3*PLIAt + β4*FORVTAt + β5*UnrDert + β6*Addt + εt Robust standard error Constant -0.199 0.4723 IN 0.004** 0.0006 FORVTA -0.030 0.0154 UnrDer -0.020 0.0320 PLIA 0.081** 0.0183 MARKS 0.005* 0.0017 Add -0.006 0.1233 N 6909 R² 0.0181 Adjusted R² 0.0176 F statistics 9.34**
Table 6: value relevance
Notes: table 6 depicts the adjusted R² statistics for Hypothesis 1. The dependent variable is stock return (RET). The outliers in the model were adapted based on winsorization. The values begin in the first and 99th percentiles. The second column depicts the coefficients for the independent variables, the observation total, and the R² adjusted and F statistics for the entire model. The last column depicts the robust standard error for the independent variables, where * symbolizes outcomes that are 5% significant and ** symbolizes outcomes that are 1% significant.
____________________________________________________________________
The outcomes are in line with the study interpreted above (Biddle and Choi (2006), Goncharov and Hodgson (2011), and Dhaliwal et al. (1999). Not all elements in the financial industry are value relevant. It is possible that the variables Add, UnrDer, and FORVTA are not included in financial statements, as opposed to IN, PLIA, and MARKS. This is also stated in the table of descriptive statistics, where the relative means of Add, UnrDer, and FORVTA are very low compared to those of IN, PLIA, and MARKS. In general, Hypothesis 1 is only partly supported by these outcomes, which is also in line with prior research (Biddle and Choi (2006), Goncharov and Hodgson
4.3 Additional element
In the second model, where Hypothesis 2 was tested, an extra variable was added. The variables IS and IN were deduced using reported IS values. The hypothesis was that the IS element would temper the explanatory value of OCI elements. Table 8 depicts the outcome for the value relevance of including an additional OCI element. Altogether, the second model is statistically significant for clarifying stock returns (F(7, 850) = 8.04, p < 0.000). Moreover, the explanatory power for this model is higher (R2 = 0.0312) than for the previous model that included IS in IN. One reason could be that the variables IN and IS are relevant at the 1% level, which is why it creates a high explanatory power.
Independent variable
Rt= α0 + β1*(INt- ISt) + β2*MARKSt + β3*PLIAt + β4*FORVTAt + β5*UnrDert +
β6*Addt+ β7*ISt + εt Robust standard error Constant 0.106 0.5234 IN – IS 0.005** 0.0007 FORVTA -0.031* 0.0150 UnrDer -0.017 0.0223 PLIA 0.059** 0.0191 MARKS 0.005* 0.0017 Add -0.033 0.1551 SI 0.049** 0.2234 N 6909 R² 0.0328 Adjusted R² 0.0326 F statistics 8.04**
Table 7: additional element
Notes: table 7 depicts a model for the adjusted R² statistics from Hypothesis 2. The dependent variable is stock return (RET). The outliers in the model were adapted based on winsorization. The values begin in the first and 99th percentile. The second column depicts the coefficients for the independent variables, the total of observations, and the R² adjusted and F statistics for the entire model. The last column depicts the robust standard error, which is only for the independent variables. Here * represents outcomes that are 5% significant, and ** represents outcomes that are 1% significant.
____________________________________________________________________
Special items are value relevant for clarifying stock returns (t = 4.03, p < 0.000). This confirms the research previously conducted by Jones and Smith (2011). In comparison to the first hypothesis, FORVTA is also value relevant in clarifying stock returns (t = -1.97, p < 0.044). This can be interpreted as growth in degrees of freedom due to including the variable IS. Moreover, adding IS does not temper the value relevance of the OCI elements IN, PLIA, and MARKS from model one. All three of these elements have higher t-values in this model than in the previous one. This does not conform to the research conducted by Jones and Smith (2011). Their study did, however, focus on the service and manufacturing industries in addition to the financial industry, which could be a reason for the difference.
In total, these outcomes support the second hypothesis where the variable IS is only value relevant when IN is not included. These outcomes are not in line with the assumption that an additional element would temper the value relevance. One reason could be that it could raise an argument for including extra elements in OCI because the focus is on full fair value income, which is accepted by the financial industry (Elliot, 2010).
4.4 New financial regulations
The third model included two capital ratio dummy variables. Table 8 depicts the outcomes for the discrepancy in value relevance of financial industries. After all, only 73 observations did not meet the regulation requirements, so it is difficult to analyze the different results because there is an important difference in the number of observations. Moreover, because the sample is approximately the same as the sample from Model 1, there is essentially no discrepancy in the value relevance of OCI elements (F = 7.71, p < 0.000). The explanation power of the third model is R2 = 0.0169. The singular change in the elements of OCI is for MARKS, which in this model is less than 1% significant.
Independent variable Rt= α0 + β1*INt + β2*MARKSt + β3*PLIAt + β4*FORVTAt + β5*UnrDert + β6*Addt + β7*D1(T1) + β8*D2(T2) + εt Robust standard error Constant 1.426** 0.4328 IN 0.004** 0.0005 FORVTA -0.024 0.0131 UnrDer -0.019 0.0213 PLIA 0.073** 0.0169 MARKS 0.003* 0.0018 Add -0.003 0.1386 N 6764 R² 0.0169 Adjusted R² 0.0152 F statistics 7.71**
Table 8: new financial regulations
Notes: table 8 depicts the adjusted R2 statistics for Hypothesis 3. The dependent variable is stock return (RET). The outliers in the model were adapted based on winsorization. The values are in the first and 99th percentiles. Moreover, D1(T1), Tier 1. Capital / risk weighted assets, observations with higher than 6% are included. And for the observation of D2(T2), Tier 1 + 2. Capital / risk weighted asset that are higher than 10% are also included.
The second column depicts the coefficients, the observation total, and the R² adjusted and F statistics for the independent variables of the entire model. The last column depicts the robust standard error of the independent variables, where * represents outcomes that are 5% significant and ** represents outcomes that are 1% significant.
5. Conclusion and discussion
Since the SFAS 130 was issued, the concept of OCI and its elements has been widely studied; however, much of the research has failed to include the financial industry, which has separate regulations and rules. The author’s assumption was that the impact of OCI is essentially greater in the financial industry, because that industry focuses more on investments. Comprehensive income is significantly impacted by other comprehensive income elements.
Subsequent the introduction of this standard, a large quantity of research has examined the explanatory value of these elements. Only a clear and detailed clarification is lacking due to the mixed results of the past research. Dhaliwal et al. (1999), Ohlson (1999), and Skinner (1999) have all come to the conclusion that not all elements are value relevant. On the other hand, Biddle and Choi (2006), Cahan (2000), and Goncharov and Hodgson (2011) have stated that those same elements are value relevant.
According to Dhaliwal et al. (1999), marketable securities are the main elements of OCI in the financial industry that need to be clarified for returns. The present research provides evidence concerning the nature of OCI elements, but Ohlson (1999) has stated that it should be transitory, so they extended the previous study to the financial sector.
Prior studies have found mixed results regarding whether or not OCI is value relevant. Some studies have found that OCI is not value relevant (Barton et al., 2010; Goncharov & Hodgson, 2011), but others have found a relationship between OCI and stock returns (Biddle, 2004; Biddle and Choi, 2006; Cahan et al., 2000; Dhaliwal et al., 1999; Hirst et al., 2004). Since there is no clear agreement concerning the value relevance of OCI, this thesis provides extra evidence concerning whether or not OCI and its elements are value relevant in the financial industry. Moreover, since the financial crisis that took place in 2008, the financial industry has faced many significant organizational changes. According to Hirst et al. (2004) and Hodder et al. (2006), the financial sector is not entirely satisfied with the present ruling concerning OCI.
The present research examined the value relevance of OCI elements. The sample selected for this study was from 2006 until 2016. Moreover, it also explained whether or not there could be additional/extra value relevance elements that are not separately recorded in OCI statement. Finally, the impact of the new financial regulations on the value relevance of OCI was tested.
To test the aforementioned research questions, input data for this thesis was gathered using the database CRSP/Compustat. All information about financial statements was easily accessible; the monthly security data just had to be transformed into yearly data, since the financial statement data is also yearly. The monthly security data was gathered eight months before and three months after each fiscal year had come to an end.
There are three hypothesis tested in this thesis, namely “all elements of OCI are value relevant”, “not all value relevant elements are individually stated in the financial statement especially the special items, so the second hypothesis is special items will decrease the value relevance of OCI and its elements, and value relevance of OCI is higher for organizations that obligate to the new financial regulations, therefore the last hypothesis is “higher capital ratios will increase the value relevant of OCI and its elements”.
The present research contributes to the area of study in several ways. First of all, there has never been a clear general agreement on whether or not OCI elements are value relevant. The present research focused solely on the financial industry, whereas previous studies (e.g., Skinner, 1999; Biddle and Choi, 2006) did not include the financial industry in their samples. Moreover, financial industry would have the full-fair-value income measure than to have an OCI approach, that is nowadays used by that industry (e.g., Hodder et al., 2006; Hirst et al., 2004). The present research examined whether or not their claim is enough to analyze the SFAS 130. Finally, Jones and Smith (2011) have stated in their research that not all the value relevant measurements exist in OCI. This claim was examined by including special items in this research.
Overall, the outcomes suggest that some elements of OCI are value relevant. The following elements were found to be value relevant: minimum pension liabilities adjustment, net income, and unrealized gains/losses on marketable securities. In addition to these elements, some prior studies also found that unrealized gains/losses on marketable securities and net income are value relevant (e.g., Dhaliwal et al., 1999; Biddle, 2004; Biddle and Choi, 2006; Hirst et al., 2004; Cahan et al., 2000). In contrast with this research, prior studies did not find any value relevance for minimum pension liability adjustment. This could be because of the financial crisis and issues regarding pension funds, since investors have found that an enterprise’s pension adjustment determines financial anxiety (Impavido & Tower, 2009). Additionally, in line with research conducted by Jones and Smith (2011), the special item element is value relevant and could possibly be included as an additional element. Finally, there was no outcome regarding the impact of new financial regulations on the value relevance of OCI, because the sample that did not conform to the new regulations and rules was insignificantly small.; therefore, there was no basis for an equation with observations that complied with new rules.
This study has several implications. First of all, the period studied may have been too broad. Secondly, the outcomes of this study are only applicable for U.S.-based companies that performs under American GAAP standards, so there may be errors in the data on the stock return. At first, the idea was to examine the value relevance yearly and then subdivide within observations that comply with the regulations and rules designed for the financial crisis and the restoration that
followed it, even though this sample consists of organizations that almost comply with the regulations and rules determined by Elliot (2010). Because of the first proposal, the sample period covered 11 years of data and includes enterprises that went bankrupt or suffered financial anxiety. The outcomes of this research can only be applied to American organizations, so it cannot create an overall understanding of whether or not OCI and its elements are value relevant. Finally, there is a chance that there are errors in the calculations of yearly stock return information made when the stock return data was transformed into yearly data and the data was merged. It is possible that something went wrong in the two datasets before the merging procedure began.
Further research could focus on whether the element of minimum pension liability adjustment could also be value relevant in different kinds of industries and sample periods. Moreover, more financial regulations and rules could be added to measure whether or not the financial crisis, its impact, and the subsequent restoration yield dissimilar results regarding the value relevance of OCI. Hereafter, another examination could look at the three period that the author has described in the introduction. So what will be the effect of value relevance on OCI elements in pre-crisis period. What will be the effect of value relevance on OCI elements in the crisis period, and what is the effect of value relevance on OCI elements in the post credit crisis period. Finally, future research could try to divide the special item elements in order to discover which components might be value relevant.
