• No se han encontrado resultados

Diferencias fundamentales entre mandato/comisión,

5. Contratos de colaboración: intermediación y distribución

5.2. Los contratos de intermediación

5.2.1. Diferencias fundamentales entre mandato/comisión,

Financial accounting information and disclosure are very important tools for investors (Healy & Palepu, 200 1 ; Lambert, Leuz, & Verrecchia, 2007) as financial accounting information and disclosure supply a key quantitative representation of individual corporations (Bushman & Smith, 2003). A high level of disclosure quality can reduce the cost of capital of a company (Ashbaugh, Collins, & LaFond 2006; Krishnamurti, Sevic, & Sevic, 2005b). Moreover, as a result of the increased global isation of financial and product markets, i nterest of both market participants and regulators i n financial reporting quality is developing worldwide (Kothari, 200 1 ).

While much attention is given to the quality of financial reporting and indeed the phrase "financial reporting quality" is widely used, the concept of financial reporting quality is elusive and has been interpreted in a variety of ways (Ball et al., 2003 ). There has been no agreement on the definition of or the framework for financial reporting quality among researchers and accounting professionals (Jonas & Blanchet, 2000). As stated by McDaniel, M artin, and Maines, (2002, p. 1 44) "the SEC, auditing profession and national exchanges (in the US) have not specified an explicit

defin ition of or a framework for financial reporting quality". As a result, there are various interpretations of or proxies for financial reporting quality.

Most pnor studies use either disclosure quality (for example, Wright, 1 996) or earnings quality ( Bushman, Piotroski et al., 2004) as a proxy for financial reporting quality (refer Appendix A for a summary of the studies). Very few studies use multiple proxies for financial reporting quality (see for example Barton & Waymire, 2004; Han, 2005; Rajgopal & Venkatachalam, 2008). This has motivated the current study to provide an understanding of the concept of financial reporting quality through multiple proxies.

The current study assumes incorporation of both disclosure quality and earnmgs quality as being important because it has been shown that companies with high quality disclosure substitute enhanced disclosure for low quality of earnings. that is, earnings are managed and delayed earnings recognition of value-relevant events is overcome by providing high quality disclosure (Shaw, 2003). In other words, even if a company' s disclosure qual ity is high, this does not necessarily mean its earnings quality is also high. Therefore, taking only disclosure quality as a proxy for financial reporting quality misleads users of financial reports.

Only taking earnings quality as a proxy for financial reporting quality is seen as inadequate as earnings information in investment decision-making is often i nsufficient (Schadewitz & Kanto, 2002). It is clai med to be insufficient because it is based primarily on historical figures (Coll ins, Maydew, & Weiss, 1 977), and therefore limits a prediction of a company' s future prospects. On the other hand, according to Schadewitz and Kanto (2002), disclosure allows management to communicate detailed information about not only historical information but also the future prospects of a company' s business activities.

General ly, a review of prior studies in the area of financial reporting quality can group them into two mai n categories; those that use disclosure quality and those that use earnings quality as a proxy for financial reporting quality. Other than these major categories, there are studies that refer to financial reporting quality in relation to certain characteristics or attributes. The following sub-sections discuss these two

maJ or proxies of financial reporting quality and also look at other attributes of financial reporting quality. The review of disclosure and earnings quality studies reported in the following sections focuses on understanding the concepts and measurements instead of the findings of those studies. However, Appendix A provides the details of the studi es, including their findings. Figure 3 . 1 shows the proxies for financial reporting quality established from the literature review and the discussions of the proxies in the fol lowing sections.

• • • • • • •

Figure 3. 1 : Proxies for Financial Reporting Quality

• Extent of disclosure • Disclosure quantity • Level of disclosure • lnfonnativeness • Timeliness. details, clarity • Comprehensiveness • Potential usefulness • Amount of d isclosure • Information contents Develop I n dex ( Items: 1 0-296\ Financial experts Investors Literatures Loan officers Authors Financial analysts Security analyst Professional RlltinPsll nclex • Cl FAR AIMR S&P FAF • • • • • • • • • • •

:

FRQ - Sec. 3.3

Sec. 3.3.2 Earnings persistence. precision

Earning sustainabil ity

Earning response coefficient Predictability of cash flows Discretionary accruals. abnormal accruals

Feedback value. neutral ity, timeliness, representational faithfulness

Earnings smoothing. timely loss recognition

Reflection of earni ngs on current economic activities

lnformativeness of accounting earnings Conservatism

Earnings composed primarily of operati ng cash flows

Information content F:arnin12 usefulness

FRQ : Financial Reporting Quality DQ : Disclosure Qual ity

EQ : F:arnings Qual ity

• • • • • • • • • • • • • Other Attributes Sec.3.3.3

Jones model ( 1 99 1 )/Modi fied Jones model ( 1 993 )

Dechow and Dichev model ( 2002), Penman and Zhang model ( 2002 ), Leuz, Nand a & Wvsocki model ( 2003 )

F ASB's conceptual framework

Report by CFRA

Change in investors' assessments, Survey method, ratios Behaviour of security prices

Earnings response coefficient and future earni ngs growth Magnitude of abnormal accruals. the timeliness and relevance of earnings

Deviation of net income from operating cash flows Earnings return relation

Volatility of accruals and volatility of earnings Conservation index (C Scores) & earnings quality indicator (Q Scores)

Investment strategy based on the rank of the unexpected earnings and stock returns on the contemporaneous level and change in earnings Measurement • Relevance. reliability. comparability • •

Employed analysis judgement SEC's assessment criteria

• Transparency. full d isclosure.

comparabi l ity

• Relevance. reliability. clarity and management to assess qual ity

• •

Used audit committee members. Auditors and management to access quality

3.3.1 Disclosure Quality and Its Measurement

Various i nterpretations of disclosure quality have been put forward by prior studies. It has been referred to as adequacy of disclosure (Buzby, 1 974); comprehensiveness of information disclosure - the fact that no important aspect has been left undisclosed ( Imhoff, 1 992; Wallace and Naser, 1 995); the extent of disclosure (Bushee, 2004; Cooke, 1 989, 1 992), as wel l as the degree of compliance with standards requirements (Naser and Nuseibeh, 2003) . Unlike the studies that carried out annual report content analysis, M itton (2002) considers companies to have indicators of high quality disclosure if the companies have a listed American Depository Receipt (ADR) and if their auditor is one of the Big Four1 1 international companies.

In determining disclosure quality, pnor studies have used either their own self­ developed disclosure index (for example Buzby, 1 974; Cooke, 1 989, 1 992; Naser & N useibeh, 2003; Robbins & Austin, 1 986; Singhvi & Desai, 1 97 1 ; Wallace & Naser, 1 995); indices of professional bodies (such as Chartered F inancial Analysts I nstitute - CF A 1 2; Financial Analysts Federation - F AF; the Center for Financial Analysis and Research - C I F AR; or Standard and Poors - S&P) or the professional bodies' disclosure ratings. The disclosure index procedure involves an evaluation of the i nformation items disclosed in a report (such as an annual report), based on a pre­ defined list of the possible index items. The disclosure index used i s either weighted or un-weighted. A weighted index takes into account the importance of information items whereas an un-weighted i ndex assumes all items are of equal importance.

The studies that developed weighted disclosure indices include those of Singhvi and Desai ( 1 97 1 ), Buzby ( 1 974), Firth ( 1 979), Hooks, Coy and Davey (2002) and Naser and Nuseibeh (2003). Singhvi and Desai ( 1 97 1 ) developed an index of thirty-four items to assess the adequacy of disclosure of listed and non-listed companies' annual reports. Buzby ( 1 974) developed a weighted index of thirty-nine items to measure the extent of disclosure of financial and non-financial items in annual reports of small and medium size companies. The index was based on the importance of each

1 1

At the t i m e of M itton's (2002) study, i t was the Big Six.

1 2

Formerly known a s t h e Association of Investment Management and Research (AIM R).

of the items for disclosure in annual reports as perceived by financial analysts. Similar to B uzby's ( 1 974) study that measured the extent of disclosure, Firth ( 1 979) also developed a disclosure index made up of forty-eight voluntary items. The index developed in this study was a weighted index where the voluntary items were weighted based on their importance to financial analysts working for stockbrokers and investment institutions. By also developing and applying a weighted index, Hooks, Coy and Davey (2002) measured the extent and quality of disclosure based on seventy-six information items, where the weighting of disclosure importance was based on l iterature and a panel of expert opinions.

In contrast to the above weighted indices to measure disclosure quality, there are studies that have used an un-weighted i ndex . These include Cooke ( 1 989), who used a dichotomous procedure in developing and applying a disclosure index in order to measure the disclosure qual ity of annual reports of Swedish companies. The procedure identified whether an item was present in the companies' annual reports or not. A score of 1 was allocated to each item disclosed and 0 for non-disclosure. The ratio of actual scores awarded to the total expected (maximum possible) scores indicated the quality of disclosure. In Cooke' s ( 1 989) study, the un-weighted index was made up of 229 items. Also using the un-weighted index procedure, Wallace and Naser ( 1 995) constructed a disclosure index of thirty items to assess the comprehensiveness of disclosure.

While the above studies developed and applied either a weighted or un-weighted index in order to assess disclosure qual ity, there are studies that have used both weighted and un-weighted indices (for example, Barrett, 1 976; Robbins & Austin, 1 986; Naser & Nuseibeh, 2003 ; Chow & Wong-Boren, 1 987). Barrett ( 1 976) constructed a disclosure index using seventeen categories of information. The quality of disclosure was i ndicated by the extent of financial disclosure that was determined from the application of the index and the degree of comprehensiveness of the companies' financial statements as determined by quality criteria identified by the researcher. In Robbins and Austin ( 1 986), the index was made up of twenty-seven items and used to measure the extent and importance of disclosure of sample companies' annual reports. Naser and Nuseibeh (2003), in assessing the quality of information disclosed by a sample of non-financial Saudi companies listed on the

S audi Stock Exchange, constructed a disclosure index which was weighted by the mean and median responses of several user groups of annual reports in Saudi Arabia. I n the study, the un-weighted procedure was also applied. Naser and Nuseibeh considered the extent, the importance of disclosure and the degree of compl iance to the statutory requirements as a measure for disclosure quality.

Similarly, by using an index and compared scores when weighting was added and not added, Chow and Wong-Boren ( 1 987) examined the extent of voluntary financial disclosure. The index consisted of eighty-nine items that were weighted for various degrees of importance by loan officers. Their comparison between weighted and un­ weighted scores revealed almost identical results and the finding has been used in a lot of subsequent research to defend the use of un-weighted indices (for example, Marston and Shrives, 1 99 1 ; Wallace and Naser, 1 995; Naser and Nuseibeh, 2003). These studies found that the use of weighted and un-weighted indices gave no material difference in results.

From the late 1 990s, researchers in the disclosure quality used disclosure ratings issued by professional bodies as a measure of disclosure quality. For example, Lang and Lundholm ( 1 996), Sengupta ( 1 998) and Shaw (2003 ) used companies' disclosure ratings as outlined in the report of the F AF.

In addition to the F AF ratings or scores, disclosure quality ratings issued by the CF AI AIMR were also used (for example, Bens & Monahan, 2004; Brown & H i l legeist, 2007; Bushee & Noe, 2000; Healy, Hutton, & Palepu, 1 999; Lee, Petroni, Shen, & Hirst, 2006). The AIMR ratings were based on the financial analysts' perceptions of the importance and quality of disclosure items selected. The disclosure quality scores issued by CIF AR have also been used in prior studies, for examples Bushman and Smith (2003); DeFond, Hung and Trezevant (2007) and Hope (2003 ). The CIFAR index largely covers the same items as S&P's Transparency and Disclosure i ndex and focuses on the quantity or extent of disc losure (Bushee, 2004 ).

While the above reviewed studies used the ratings/scores issued by the professional bodies as the construct of disclosure quality, there are studies that applied the index

used by professional bodies - such as the S&P's index - to the annual reports of their sample companies (for example, Dargenidou, McLeay, & Raonic, 2006; Patel, Balic, & Bwakira, 2002). This index was used by Patel et al. (2002) to assess the level of disclosure of ninety-eight possible information items which were divided into three sub-categories: ownership structure and investor relations; financial transparency and i nformation disclosure; and board and management structure and process.

In summary, the review of disclosure quality studies finds that there is no common understanding of the concept of disclosure qual ity. In terms of its measurement, prior studies have recognised the use of disclosure indices to measure disclosure quality. The index can be either weighted or un-weighted. In addition, there is no agreement on the number of items used in the index developed. Appendix A provides a summary of prior studies related to disclosure qual ity and disclosure indices ( including the above reviewed studies). The results of each study are also reported in the summary.

The disclosure quality assessed in the current study is that of annual reports. Annual reports are not the only source of corporate reporting; however, focusing on this source only will not reduce the quality of information, as it is general ly bel ieved that the annual report is one of the most important sources of corporate reporting (Botosan, 1 997). The definition of disclosure quality that is employed in the current study is in line with Cooke ( 1 989, 1 992), who considers the extent of disclosure as a construct of quality. Extensiveness ensures a sufficient amount of disclosure is provided to the users of financial reports to make economic decisions. It is an adequate measure of the quality of disclosure (Botosan & Plumlee, 2002). As this current study is concerned with the extent of disclosure as a proxy for the quality of disclosure, the use of a disclosure index is seen as appropriate. Chapter Five provides details of the development of the disclosure index.

3.3.2 Earn ings Quality and Its Measurement

Earnings quality has al so been defined and measured differently in previous studies. Earnings quality has been referred to as earnings informativeness ( Beaver, 1 968; Fan

& Wong, 2002; Vafeas, 2000), and the usefulness of earnings operationalised by the behaviour of security prices (Ball & Brown, 1 968).

Schipper and Vincent (2003 ) came out with an extensive review of earnings quality constructs and measures which were classified into four sources - the time-series properties of earnings; the relationships between income, cash and accruals; selected qualitative characteristics m the F ASB's Conceptual Framework; and implementation decisions (Schipper & Vincent, 2003, p.99). Earnings constructs have been mostly derived from the first two sources. The time-series-based and accrual-based constructs have been then modified and/or combined in subsequent studies for the purpose of measuring earnings quality. For the time-series classification, three constructs have been identified (Schipper et al ., 2003, p.99) - "persistence" where earnings are viewed as "more permanent and less transitory"; "predictive ability" which is referred to as "the ability of past earnings to predict future earnings" (Lipe, 1 990 in Schipper et al., 2003, p. 99); and "variabi lity" which is identified from whether the earnings are naturally smoothed earnings or result from income smoothing activities.

Studies that used earnmgs quality constructs derived from the time-series classification, as reviewed by Schipper et al. (2003) include Kormendi and Lipe ( 1 987), Collins and Kothari ( 1 989), and Leuz Nanda and Wysocki (2003) . In more recent studies, DeFond et al. (2007) also measured earnings quality using the time­ series c lassification - a variation of the earnings management metric used by Leuz et al. (2003).

The second earnings construct classified by Schipper et al. (2003, p.99) was mostly related to accruals which include "changes in total accruals", "direct estimation of discretionary accruals" and the "relations of accruals-to-cash". According to the researchers, changes in total accruals i ndicate manipulations by managers, in that the greater the changes, the lower the quality of earnings. The direct estimation of discretionary accruals was initially i ntroduced by Jones ( 1 99 1 ) using accounting fundamentals - revenues adj usted for receivables or plant, property, and equipment. In this approach, total accruals are regressed on the accounting fundamentals and the residuals from the regression are the discretionary accruals which indicate earnings

management that reflects lower earnings quality. Jones' s ( 1 99 1 ) model was also used in the studies of Bedard, Chtourou and Courteau (2004), Cahan ( 1 996) and Myers, Myers and Omer (2003).

According to Schipper et al. (2003), Jones's ( 1 99 1 ) model was improved by Dechow and Dichev (2002) by capturing aspects of the relations of accruals-to-cash. Thi s approach i nvolves a regression of changes in working capital accruals on prior, current, and next period cash flows. The estimated residuals from the regression describe an estimation error in unintended and manipulative accruals and indicate an opposite measure of earnings quality. The extent to which working capital accruals map onto operating cash flow realisations reflects accruals quality (Francis, LaFond, Olsson, & Schipper, 2005). Dechow and Dichev' s (2002) model has been employed i n a number of research studies, for example Francis, Huang, Raj gopal and Zang (2008a); Francis, Nanda and Olsson (2008b ), Francis, LaFond et al. (2005) and Chen, Shevlin and Tong (2007). Francis, LaFond et al . (2005) and Francis et al. (2008a; 2008b) integrate Jones' s ( 1 99 1 ) model and Dechow and Dichev's (2002) models in measuring earnings quality. Dechow and Dichev's (2002) model is able to identify a direct link between cash flows and current accruals.

Whi le the above studies employed time-series properties-based and/or accrual based earni ngs quality constructs, Basu ( 1 997) operational ised earnings quality as timely recognition of economic losses. This operationalisation of earnings quality was then used in other studies (see for example, Ball et al ., 2003 ; Ball & Shivakumar, 2005). Ball et al . ' s (2003 ) highlighted the fact that financial reporting qual ity was ultimately determined by the underlying economic and political factors influencing managers' and auditors' incentives, and not by accounting standards per se. However, Bal l et al. (2003) did not empirically examine the relationship between political factors and financial reporting quality. This provides an opportunity for the current study to investigate the relationship.

If Ball et al. (2003) and Ball and Shivakumar (2005) focused on timely recognition of economic losses, Ashbaugh et al. (2006) used timeliness and value relevance (transparency) of accounting earnings as one of the proxies of earnings quality. According to Ashbaugh et al. (2006), more transparent and current earnings reflect a

company's current economic activity information and contribute to higher earnings quality. In Ashbaugh et al. (2006), the construct of timeliness and value relevance of earnings was combined with two other constructs - discretionary accruals and the i ndependence of the audit committee. The magnitude of the three constructs was used as the proxy of earnings quality.

Simi lar to Ashbaugh et al. (2006) in combining more than one construct in determining earnings quality, Franc is et al. (2008a; 2008b) took into account