Institutional Investors on Boards and Audit Committees and Their Effects on Financial Reporting Quality

Manuscript Type. Empirical. Research Question/Issue. This study examines how the presence of representatives of institutional investors as directors on boards or on audit committees enhances financial reporting quality, reducing the probability that the firm receives qualified audit reports. We focus on directors who maintain business relations with the firm on whose board or committee they sit (pressure‐sensitive directors). We also investigate the specific role of bank directors and examine their effects on financial reporting quality when they act as shareholders and directors. Research Findings/Insights. Our results suggest that institutional directors are effective monitors, which leads to higher quality financial reporting and, therefore, a lower likelihood that the firm receives a qualified audit report. Consistent with the relevant role of business relations with the firm, we find that directors appointed to both boards and audit committees by pressure‐sensitive investors have a larger effect on financial reporting quality as it is more likely that the auditor issues an unqualified audit opinion. Nevertheless, when analyzed separately, only savings bank representatives on the board increase financial reporting quality. Theoretical/Academic Implications. The results confirm that board characteristics have an important influence on financial reporting quality, in line with the views that have been expressed by several international bodies. The findings also suggest that both researchers and policy makers should no longer consider institutional investors as a whole, given than directors appointed by different types of institutional investors have various implications on financial reporting quality, measured by the type of audit opinion. Practitioner/Policy Implications. This study makes its core contribution by empirically showing that directors appointed by different types of institutional investors have diverse implications on financial reporting quality. This evidence can be potentially helpful in providing a basis for regulatory actions, namely those aiming to influence the structure of the board of directors. The results have important implications for supervisors and regulators, who will benefit from an understanding of how the presence of directors on boards of savings and commercial banks in nonfinancial firms affects financial reporting quality in a bank‐based system.


INTRODUCTION
T he recent persistence of accounting scandals has led to serious reconsideration of the workings of boards and audit committees. Research has shown that board characteristics may affect the quality of the board's supervision of the financial reporting process (e.g., Beasley, 1996;Xie, Davidson, & DaDalt, 2003), and extant research on this issue has focused on board composition, specifically on the presence of independent directors (Klein, 2002;Peasnell, Pope, & Young, 2005). However, other board members have received scant attention in the literature, including directors appointed by institutional investors.
Institutional investors are among the most important controlling shareholders in continental Europe, where the principal agency conflict focuses on the expropriation of minority shareholders' wealth by controlling shareholders. In civil law countries the importance of institutional investors as supervisors compensates for the weaknesses of investor protection laws (Faccio & Lang, 2002). The specific agency problems in European continental countries have led to large blockholders, especially institutional investors, becoming directors. Thus, directors appointed by institutional investors (hereafter: institutional directors) have a significant influence on European continental boards, accounting for 40 percent of directorships in Spanish firms, compared to 2 percent in British firms (Heidrick & Struggles, 2011). In this context, unlike in the Anglo-Saxon environment, institutional investors who are also board members can exercise control over the firm as part of the internal decision-making process (e.g., Weinstein & Yafeh, 1998). These directors act as representatives of core shareholders in monitoring management and ensuring that the firm operates in their best interests.
Whereas recent studies have shown the prevalence of large institutional shareholdings around the world, research on the influence of institutional investors as directors is still scarce. In addition, whether the role of non-independent non-executive directors (also known as grey directors) is more like that of inside directors or outside directors remains ambiguous in the corporate governance literature (Hsu & Wu, 2010). Research shows that institutional directors have an important influence on leverage (Booth & Deli, 1999;García-Meca, López-Iturriaga, & Tejerina, 2013), firm value (Kumar & Singh, 2012), and earnings management (García Osma & Gill-de-Albornoz Noguer, 2007).
Given the importance of institutional investors in allocating capital to corporations and their role in firm governance, an understanding of how their presence on boards affects financial reporting quality is undoubtedly needed. Dechow, Ge, and Schrand (2010) state, "Higher quality earnings provide more information about the features of a firm's financial performance that are relevant to a specific decision made by a specific decision-maker." However, accounting quality is a noisy theoretical construct that is operationalized using a range of measures, the validity of which is unknown relative to the theoretical constructs of interest (Suberi, Hsu, & Wyatt, 2012). 1 In line with Bartov, Gul, and Tsui (2000), Butler, Leone, and Willenborg (2004), Chen, Chen, and Su (2001), Farihna and Viana (2009), and Pucheta-Martínez and de Fuentes (2007), among others, we use the audit opinion as a proxy for accounting quality. According to Farihna and Viana (2009), whereas the audit opinion is observable, discretionary accruals (i.e., the proxy for earnings management generally used in the prior literature) are unobservable and must be estimated using models that can lead to wrong conclusions due to specification problems (Dechow, Richardson, & Tuna, 2003;Dechow, Sloan, & Sweeney, 1995;McNichols, 2000).
To the extent that directors appointed by institutional investors monitor management more effectively than inside directors and have financial training, we hypothesize that companies with a greater proportion of institutional directors will be more likely to enhance financial reporting quality and, therefore, the likelihood that the auditor issues a qualified audit report will be lower. This study fills a gap in the literature as, to the best of our knowledge, we are the first to examine the influence of directors appointed by institutional investors on financial reporting quality.
Our analysis follows three steps. First, we study the impact that directors who represent institutional investors, both on boards and audit committees, have on financial reporting quality. In a second step, and according to recent literature, we assume that institutional investors cannot be considered as a homogeneous group due to their different incentives and ability to engage in the corporate governance (Almazán, Hartzell, & Starks, 2005;Chen, Elder, & Hsieh, 2007;Cornett, Marcus, Saunders, & Tehranian, 2007). We propose that the type of business relations between firms and institutional investors is a key issue in describing the role of institutional directors and, thus, their effects on financial reporting quality. Accordingly, we make a distinction between those who maintain business relations with the firm on whose board they sit and institutional investors whose business activity is not related to the company in which they hold a directorship. To our knowledge, no prior works that focus on the continental Europe context analyze the relation between institutional directors and financial reporting quality measured by the audit opinion. Finally, in a third step, we focus on the specific role of bank directors on boards and audit committees and analyze their effects on financial reporting quality when they act as shareholders and directors.
We use a sample of Spanish listed firms between 2004 and 2010. Spain is a good paradigm to study the effectiveness of institutional directors because it has the highest presence of institutional investors on the boards of large firms among European countries (Heidrick & Struggles, 2011). Corporate governance and auditing systems are different between Spanish and Anglo-Saxon markets (Fernández & Arrondo, 2007). Unlike the Anglo-Saxon capital markets, Spain is characterized by a high ownership concentration and the lack of liquid capital markets. This explains why the board of directors -which is marked by the presence of the large blockholders, especially institutional investors -is the prevalent mechanism of control in Spain. Finally, Spain offers a unique opportunity to analyze the conflicts of interest that arise when banks have simultaneous roles as shareholders, creditors, and directors.
Our results suggest that institutional directors are effective monitors, which results in higher quality financial reporting and, therefore, a lower probability of a firm receiving qualified audit reports. Consistent with the relevant role of business relations with the firm, we find that directors appointed by pressure-sensitive investors, that is, those who maintain business relations with the firm on whose board or audit committee they sit, have a greater effect on financial reporting quality since the likelihood that the auditor issues a qualified audit opinion is lower. Nevertheless, when analyzed separately, commercial banks and savings bank representative directors show different attitudes. Specifically, only savings banks representatives on the board help to enhance financial reporting quality. This finding may be justified by the specific composition of these entities, in which the regional and local governing bodies exercise a decisive power in firm strategy. 2 Even though the Spanish Unified Good Governance Code (2006) highly recommends forming audit committees of entirely independent and institutional directors, we fail to find a significant impact of independent directors. Instead, we find that institutional (specifically pressure-sensitive) board and institutional audit committee members influence financial reporting quality, as it is less likely that the auditor issues a qualified audit opinion. One explanation for this finding may be the substitution effect hypothesis between institutional and independent directors. That is, different levels of control provided by a single mechanism may be equally efficient, depending on the intensity of the control performed by other mechanisms available (Fernández & Arrondo, 2007). To a certain extent, this lack of consistency may also lie in the lack of investor confidence in the role and true independence of "independent" directors in Spain (Fundación de Estudios Financieros, 2007;Lorca, Sánchez-Ballesta, & García-Meca, 2011).
Overall, our results confirm that board characteristics have an important influence on financial reporting quality, in line with the views that have been expressed by several international bodies. Because the principal agency conflict in continental Europe and many other countries focuses on the expropriation of minority shareholders' wealth by controlling shareholders, the analysis of the institutional directors' influence on financial reporting quality highlights a priority research question. The findings in this research partly support the importance of the monitoring function of nonexecutive directors on the board and audit committee. This study makes its core contribution by empirically showing that the impact made by an institutional director on financial reporting quality varies depending on the type of institutional investor that appoints the director. This evidence could be potentially helpful in providing a basis for regulatory actions aimed at influencing the structure of the board of directors. An understanding of the factors associated with audit qualification can also act as an aid to the auditor's assessment of the engagement risk, including the planning process.

INSTITUTIONAL BACKGROUND
The Spanish corporate governance system is characterized by the presence of a few large dominant shareholders who may exert a strong influence on management, low independence on boards, low developed capital markets, and no active market for control. In this context, Spain has experienced significant legal and institutional changes with the basic objective of increasing the transparency of the stock markets and the level of protection of minority shareholders. One of the consequences has been the issue of several codes of corporate governance: the Olivencia Report was released in 1998, followed by the Aldama Report in 2003, and finally the Conthe Code or Unified Good Governance Code in 2006, which was characterized by a "comply or explain" principle in the enforcement of corporate governance regulations.
In Spain, important institutions such as the state, large banks and recently privatized companies have become controlling shareholders, exerting important influence in solving relevant issues in corporate governance (Crespí, García-Cestona, & Salas, 2004). Most of the controlling shareholders take important positions on boards, represent-ing the interests of large shareholders and institutional investors. As a result, the board of directors has become the prevalent mechanism of control with an increased presence of directors appointed by these large blockholders.
The Spanish Unified Good Governance Code (2006) distinguishes three types of directors: executive, independent, and proprietary or institutional directors. Executives are those directly involved in the management of the firm and both proprietary and independent directors are classified as non-executives, although they clearly have different agendas and their incentives to monitor managers also differ. Institutional directors represent institutional investors who are considered reference shareholders, often banking and insurance companies or investment funds.
The board of directors performs two primary functions: monitoring and advising top management (Pugliese, Bezemer, Zattoni, Huse, van den Bosch, & Volberda, 2009). While the advising role involves assisting management in strategy formulation and execution, as well as providing counsel in other areas of top-level decision making, the monitoring role involves overseeing management with a view to minimizing potential agency problems and improving the quality of financial statements. This implies becoming more involved with all the professionals responsible for financial reporting and internal control and engaging in meaningful discussion with auditors and management about financial reporting quality. The Spanish Unified Good Governance Code (2006) states that the most basic and inalienable duty of the board is "the general oversight function", divided into three key responsibilities: to guide and promote the company's policy (strategic responsibility), control its management echelons (stewardship), and liaise with its shareholders (disclosure).
Most of the institutional investors on Spanish boards belong to investment funds and banks. Banks have traditionally maintained a significant presence in economic and business development in Spain not only as creditors but also as controlling shareholders and board directors in firms. This relationship between banks and firms differs from that of the Anglo-American financial system, in which the financial markets play a much more important role, and institutional investors are not significant members of the board. The Spanish banking system is an industry with two main institutions, commercial banks and savings banks, which compete with each other for loans and deposits. Spanish savings banks have a special governance structure because they are controlled by politicians and public entities (Crespí et al., 2004;Sapienza, 2004). Regional and local governments exercise decisive power in the renewal of the governing bodies and the development of strategy for savings banks and some non-financial firms (Fonseca Díaz & González Rodríguez, 2005).
Empirical findings support the increasing role of directors appointed by institutional investors in relation to independent directors. Some studies set in Spain (e.g., García-Meca & Sánchez-Ballesta, 2009;García Osma & Gill-de-Albornoz Noguer, 2007;Lorca et al., 2011) show that independent directors are not effective at monitoring management, noting that the effect of board independence depends on investor protection rights. The majority of the studies on the monitoring role of independent directors in continental Europe show inconclusive results on the role of independent directors compared to those appointed by institutional investors (García-Meca & Sánchez-Ballesta, 2009). Some research suggests that the supervising role is not played by independent directors, as US-and UK-based research suggests, but by institutional directors, that is, those who represent controlling shareholders (García Osma & Gill-de-Albornoz Noguer, 2007).
We therefore argue that the Spanish setting provides a unique opportunity to study the effect of institutional directors, especially directors appointed by financial institutions, on financial reporting quality. Such a relation is more difficult to capture in a US or UK setting, where it is less common for directors to be appointed by institutional investors.

PREVIOUS LITERATURE
Accounting quality is defined by Palepu and Healy (2013) as the extent accounting measurement processes, and their implementation by the firm, captures the firm's underlying economic reality. Researchers use various measures as indications of "earnings quality," including persistence, accruals, smoothness, timeliness, loss avoidance, investor responsiveness, or restatements and SEC enforcement releases (Dechow et al., 2010). Earnings quality proxies can be organized into three broad categories: properties of earnings (e.g., earnings persistence and accruals), investor responsiveness to earnings (e.g., earnings response coefficient), and external indicators of earnings misstatements (e.g., accounting and auditing enforcement releases).
Although most of the accounting literature about earnings quality is based on discretionary accruals, this proxy is unobservable and must be estimated using models that can lead to wrong conclusions due to specification problems (Dechow et al., 1995;McNichols, 2000). In this sense, García-Meca and Sánchez-Ballesta (2009), in a meta-analysis on corporate governance and accounting quality, found that accruals models that do not consider long-term discretionary accruals choices may prompt us to make misleading inferences about the role of these corporate governance mechanisms in earnings management behavior. They find that when different discretionary accruals models are used, results can change considerably.
An external indicator of earnings quality, such as audit opinion, is observable and identified by an outside source (Dechow et al., 2010). Financial statements are representations of management, and individual investors who use them to make decisions rely on the auditor to verify their credibility (Chen, Su, & Zhao, 2000;Chow & Rice, 1982;Dopuch, Holthausen, & Leftwich, 1986). The external auditor is the key factor that guarantees the quality of financial statements, and thus it plays a crucial role in the corporate governance scheme. The audit opinion, as the observable output from an otherwise unobservable process, represents a vital piece of information for financial statement users. 3 The presence of a qualified audit opinion has been shown to be informative with respect to stock returns (Choi & Jeter, 1992) and bankruptcy events (Kennedy & Shaw, 1991). Bartov et al. (2000), Butler et al. (2004), Chen et al. (2001), Farihna and Viana (2009), among others, use the audit opinion as a proxy for accounting quality.
Prior research concerning corporate governance and financial reporting quality has tended to focus on board characteristics, especially board independence. In that respect, Beasley (1996) finds that companies committing financial statement fraud have a lower number of independent directors than firms without financial statement fraud. Klein (2002) and Peasnell, Pope, and Young (2000) showed that the independent board mitigates earnings management. In the same context, Bushman, Chen, Engel, and Smith (2004), Karamanou and Vafeas (2005) and Vafeas (2005) advanced that the information quality increases with the percentage of outside directors. Similarly, Beekes, Pope, and Young (2004) noticed that board independence allows the disclosure of information of good quality by firms in the United Kingdom, and Dimitropoulos and Asteriou (2010) and Firth, Fung, and Rui (2007) confirmed that the presence of independent directors improves the earnings quality of firms.
On the other hand, Vafeas (2000) shows that the informativeness of earnings, as proxied by the earnings-return relationship, is unrelated to the proportion of independent directors. Other studies suggest that the independent directors are not competent enough to control the managers and their presence on the board has no effect on reporting quality (Ahmed, Hossain, & Adams, 2006;Bradbury, Mak, & Tan, 2006;Petra, 2007). The assumed benefit of improved board independence stems from the belief that independent directors are better monitors of management than are inside directors (DeFond & Francis, 2006). However, insiders and others close to the company might have firm-or industryspecific knowledge that would aid director performance (Donaldson & Davis, 1991;Kiel & Nicholson, 2003). Thus, some prior studies at the board level suggest that the market places value on inside directors on the board (Klein, 1998;Rosenstein & Wyatt, 1997). Hence, although outside directors serve as monitors and help minimize agency conflicts between shareholders and management, inside and affiliated board directors provide firm-specific expertise that is valuable for planning the firm's operations and development (Klein, 2002).
Similarly, other studies focus on audit committee attributes and explore whether audit committee characteristics affect accounting quality (Abbott & Parker, 2000;DeZoort & Salterio, 2001;Felo, Krishnamurthy, & Solieri, 2003). Most research finds that directors with financial expertise in audit committees help to ensure credible financial statements. For example, Carcello and Neal (2000) find a negative association between the percentage of affiliated directors on the audit committee and the likelihood of a financially distressed firm receiving a going-concern audit report. Although this evidence is suggestive, these studies do not directly investigate the relation between directors appointed by institutional investors and financial reporting quality.
The main interest of directors appointed by institutional investors is to create the maximum level of return for their beneficiaries. Thus, directors appointed by institutional investors seek to extend their influence to the decisionmaking board committees, given that increased share value resulting from direct supervision can compensate for any supervisory costs that may be directly incurred. This puts pressure on corporate managers to make the company look attractive to institutional investors and to create more shareholder value. Thus, monitoring by institutional investors is likely to result in improved firm performance because, as large and sophisticated shareholders, institutional investors have the incentive and expertise to monitor management and can do so at a lower cost than atomistic shareholders (Shleifer & Vishny, 1986). They are also able to exert enough influence to alter the governance structure and the firm's course of actions.
Institutional investors are known to influence various important corporate decisions. Almazán et al. (2005), Borokhovich, Brunarski, Harman, and Parrino (2006), Brickley, Lease, and Smith (1988), Bushee (1998), Ferreira and Matos (2008), and Hartzell and Starks (2003) show that institutional investors influence antitakeover amendments, research and development (R&D) investment decisions, and CEO compensation and profitability. Ramalingegowda and Yu (2012) note that higher ownership by institutions that are likely to monitor managers is associated with more conservative financial reporting. Ljungqvist, Marston, Starks, Wei, and Yan (2007) support the hypothesis that the presence of institutional investors provides incentives for analysts to publish unbiased or less biased research.
Despite the wide research on institutional investors, as far as we know, no works that focus on the context of continental Europe analyze the relation between institutional investors on boards and accounting quality as proxied by audit opinion. The ownership structure of firms, the weakness of the corporate control market, and the relevance of banks as core shareholders mean that the conclusions obtained in this study can be extended to other countries with similar corporate governance paradigms.

HYPOTHESES DEVELOPMENT
Given institutional shareholder incentives to supervise managerial actions, we expect a positive influence of institutional directors on financial reporting quality. Because earnings information is important for business valuation purposes, institutional directors demand high quality information and exert more influence than other board members. Institutional owners, as a group, command large amounts of capital that are professionally managed and employed in the equity markets. Using this capital, institutional owners can exert influence by buying and selling large blocks of a firm's securities and by holding voting rights that can be directly employed to influence the decisions of management (Kane & Velury, 2004). The existence of sophisticated institutional investors can also induce managers to provide high-quality audits (Felo et al., 2003). According to prior research, by doing so, institutions can delegate the actual task of monitoring to auditors, and the cost of that monitoring is borne by all shareholders within the firm (i.e., the delegation hypothesis). Chung, Firth, and Kim (2002), Jiraporn and Gleason (2007) and Rajgopal, Venkatachalam, and Kotha (2002) suggest that institutional investors serve as monitors, mitigating earnings management behavior. In this line, some authors find that the higher the proportion of non-executive board members, the lower the probability of accounting fraud (Beasley, 1996;Peasnell et al., 2005;Xie et al., 2003).
Regarding the audit committee, in Spain the Conthe code (2006) recommends forming audit committees entirely of external directors (i.e., independent and institutional directors) in a proportion similar to that of the board of directors. According to previous research, when audit committees are made up by a high proportion of institutional directors, they are more likely to be effective in protecting the credibility of the firm's financial reporting because they are external directors and independent of management (Pucheta- Martínez & de Fuentes, 2007). As such, it will also be more difficult for management to resist the adjustments proposed by auditors (McMullen & Raghunandan, 1996;Song & Windram, 2004).
In this line, Carcello and Neal (2000) and Klein (2002) find a significantly negative association between abnormal accruals and the proportion of outside directors on the audit committee. Similarly, García Osma & Gill-de-Albornoz Noguer (2007) find that the main role in constraining earnings management in Spain is played by institutional directors rather than independent directors. Hsu and Wu (2010) note that as the number of grey directors 4 on the board and audit committee of UK firms increases, the probability of corporate failure decreases. However, they find that more independent outside directors on the board and audit committee may not effectively decrease the likelihood of corporate failure.
Given the previous discussion, we hypothesize that a higher number of institutional directors will increase financial reporting quality.

Hypothesis 1a. Financial reporting quality is positively affected by the presence of institutional directors on boards.
Hypothesis 1b. The quality of financial information is positively affected by the presence of institutional directors on audit committees.
Theoretical work by Kahn and Winton (1998), Maug (1998), and Shleifer and Vishny (1986) highlights the choice that institutions face between exerting monitoring effort for shared gain versus simply trading for private gain. Institutional investors vary in a number of dimensions, including the skill of their employees, their resources or incentives to gather information, and the implicit or explicit pressure from firms in which they invest due to potential business relations (Brickley et al., 1988). Different authors note that the presence or absence of business relationships can condition the institutional investor's levels of influence. Researchers such as Almazán et al. (2005), Borokhovich et al. (2006), Brickley et al. (1988), Bushee (1998), Ferreira & Matos (2008), Hartzell and Starks (2003), and Ramalingegowda and Yu (2012) have shown that certain types of, but not all, institutional investors exert influence on antitakeover amendments, R&D investment decisions, CEO compensation, profitability, and earnings conservatism. García-Meca et al. (2013) also show that institutional directors have diverse incentives to engage in corporate governance, noting different effects on cost of debt depending on the type of institutional director.
To understand institutional monitoring better and the sometimes conflicting evidence, we study institutional investors on boards of directors and audit committees and focus on pressure-sensitive investor directors, namely, those that maintain business with the firm in which they invest (i.e., basically directors who represent banking and insurance companies). Pressure-sensitive investor directors focus mainly on the firm's long-term viability as they have more incentives to collect and process information. In this line, Brickley et al. (1988) find evidence that firms with greater holdings by pressure-sensitive shareholders (banks and insurance companies) have more proxy votes cast in favor of management's recommendations. Porter (1992) argues that long-term or dedicated owners alleviate pressures for myopic investment behavior because their holdings provide incentives to monitor managers. Similarly, Dobrzynski (1993) and Monks and Minow (1995) argue that institutions that invest in firms with the intention of holding substantial ownership blocks over a long horizon have stronger incentives to monitor the firm. Han, Kang, and Rees (2013) show that firms are more likely to hire a Big Four auditor when long-term institutional ownership is high, suggesting that long-term institutional investors view high-quality audits as a viable means of improving corporate governance, while reducing their direct monitoring costs. Their results suggest that dedicated long-term institutional investors demand higher quality audits to enhance corporate monitoring and that short-term institutional ownership is positively associated with higher audit risk. Prior research (e.g., Chen et al., 2007;Gaspar, Massa, & Matos, 2005;Shleifer & Vishny, 1986) also suggests that institutional investors' demand for conservatism is more likely to emanate from monitoring institutions with a long-term investment horizon. In contrast, pressure-resistant investors put pressure on management to meet short-term earnings targets, which can increase the likelihood of financial misreporting. Coffee (1991) notes that short-term institutional investors may have incentives to sell their stock due to poor performance rather than initiate corrective action.
While this evidence is suggestive, these studies do not investigate directly the relation between directors appointed by pressure-sensitive investors and financial reporting quality. Thus, we pose the following hypothesis: Hypothesis 2a. Financial reporting quality is positively affected by the presence of pressure-sensitive institutional directors on boards.
Hypothesis 2b. Financial reporting quality is positively affected by the presence of pressure-sensitive institutional directors on audit committees.
Nevertheless, even within pressure-sensitive investors (insurance companies and banks) some differences exist. Banks are the most prevalent and identifiable representative of institutional investors, especially in continental countries. However, in the United States, earlier regulation created a corporate governance system that differs historically from that in other countries such as Spain, Germany, and Japan where, by design, institutions (particularly banks) play a large role in the ownership and monitoring of corporations (Gillan & Starks, 2003).
In Spain, banks are not only a major source of funding and financing for the country's business fabric but they also hold strong positions as company stockholders and members of boards. Bankers can play a certification role on the board because a banker joining the board of a firm can signal to the market that the firm is unlikely to experience financial distress. Hadlock and James (2002), Johnson (1997), and Lummer and McConnell (1989) report that long-term relations between banks and nonfinancial firms reduce the asymmetric information and allow banks to control a firm's decisions. Thus, they diminish the adverse selection and moral hazard problems. Ljungqvist et al. (2007) also provide evidence that analysts issue more optimistic recommendations when they are affiliated with banks that have an existing relationship with the firm and when they work for banks with larger businesses.
In addition, after regulation changes and press coverage following recent accounting scandals, the need has been stressed for financial expertise on corporate boards. Thus, if a bank develops specialized knowledge through lending to many firms in a particular industry, bankers can provide valuable industry-specific financial expertise as board members of firms in that industry. In addition, a qualified report is costly for a bank director because free-rider problems and information asymmetries make it difficult for firms to renegotiate with creditors.
Due to the differences between Spanish commercial and savings banks, it is interesting to analyze separately how the governance of these banks affects the financial reporting quality, measured by the audit opinion, when they are members of other firms' boards and audit committees. This comparison is relevant because both commercial and savings banks operate under the same regulatory framework and market conditions, but Spanish savings banks have a special governance and ownership structure (which is a lack of ownership) because they are controlled by politicians and public entities (e.g., Sapienza, 2004).
During recent years the Spanish regional regulation has increased the presence of public administration of savings banks at the expense of depositors' representation, leading the regional and local governments to exercise a decisive power in the renewal of the governing bodies and the establishment of the savings banks' strategy (Fonseca Díaz & González Rodríguez, 2005). These directors have also become an increasing part of nonfinancial firm boards. Consequently, the reform of the savings bank law in 2010 addressed this issue by reducing the political power of public authorities and strengthening the privatization and the professionalization of governing bodies with the aim of depoliticizing the government of savings banks. Given these arguments, we acknowledge the value of research issues such as whether the influence of the directors appointed by savings and commercial banks influences financial reporting quality. We therefore pose the following hypothesis:

Hypothesis 3a. Financial reporting quality is positively affected by the presence of commercial and savings bank directors on boards.
Hypothesis 3b. Financial reporting quality is positively affected by the presence of commercial and savings bank directors on audit committees.

EMPIRICAL DESIGN Sample
The sample is drawn from the population of Spanish nonfinancial firms listed on the Spanish Stock Exchange during 2004-2010. We exclude financial companies both because they are under special scrutiny by financial authorities that constrain the role of their board of directors and because of their special accounting practices. We obtain our data from two databases. Audit opinion, financial information, and firms' market value come from the "Sistema de Análisis de Balances Ibéricos" (SABI) database, and corporate governance information is collected from the annual corporate governance reports that all listed companies have been required to publish since 2003.
We build an unbalanced panel of 627 firm-year observations from 162 firms. Roughly, our sample accounts for more than 95 percent of the capitalization of Spanish nonfinancial firms. The panel is unbalanced because during this period some firms became public, and other firms delisted as a consequence of mergers and acquisitions. Nevertheless, the estimations based on unbalanced panels are as reliable as those based on balanced panels (Arellano, 2003).
We also employ several independent variables. We define INST as the proportion of institutional directors on the board. These are mainly directors appointed by institutional investors, and they often represent banking and insurance companies or investment funds. We define INDEP as the proportion of independent members on the board. In line with García-Meca et al. (2013), we define SENSIT as the proportion of the board members who are representative of pressure-sensitive institutional investors (i.e., banks and insurance companies). Given our special attention to the roles played by the different institutional investors, we define COM_BANK as the proportion of directors who represent commercial banks and SAV_BANK as the proportion of directors who represent savings banks.
We define analogous variables concerning the presence of these directors on the audit committee. Specifically, INSTAC and INDEPAC represent the existence of institutional and independent directors on the audit committee, respectively. SENSITAC is a dummy variable that equals 1 if pressuresensitive representatives sit on the audit committee. COM_BANKAC and SAV_BANKAC are dummy variables for directors appointed by commercial banks and savings banks, respectively, on the audit committee.
Following prior literature, we control for a number of factors (see Pucheta-Martínez & de Fuentes, 2007) that can potentially affect the audit opinion. SIZE is the log of total assets and is a measure of firm size. Carcello, Hermanson, and Huss (1995) and Mutchler, Hopwood, and McKeown (1997) report a negative relation between company size and the receipt of a qualified audit report. DeAngelo (1981) suggests that the issuance of qualified audit reports may cause companies to switch audit firms and thus cause the initial auditor to lose the quasi-rents associated with future audits of the client.
Previous literature shows that big auditors provide higher quality services (Teoh & Wong, 1993) and that they are also better able to express a qualified opinion (Farihna & Viana, 2009;Lennox, 1999). Thus, we propose BIGFOUR as a dummy control variable that equals 1 if the opinion is issued by a Big Four audit firm, and zero otherwise. Regarding the ownership structure, Sánchez-Ballesta & García-Meca (2005) report that director ownership is an effective monitoring device that leads to higher quality of financial reporting and, therefore, less likelihood of receiving qualified audit reports. Thus, we define DIREC_OWN as the percentage of stock owned by directors. In addition, we expect that larger audit committees reduce managers' ability to put pressure on a significant number of members, thus making it more difficult to resist the adjustments proposed by auditors. Therefore, we include as control variable AC_SIZE, defined as the size of the audit committee and measured as the number of audit committee members.
Given that prior literature identifies financial distress as a factor that may increase the likelihood that an auditor will issue a qualified audit report (Carcello et al., 1995;Mutchler et al., 1997), we include two variables to control for financial distress. These are LEV as a proxy for the agency cost of debt and measured by debt over total assets, and losses in the previous year (LOSS). On the other hand, previous evidence shows that return on assets (ROA) is a significant factor in explaining corporate failure (Al-Khatib & Al-Horani, 2012;Altman, 1968;Altman, Halderman, & Narayana, 1977;Beaver, McNichols, & Rhie, 2005;Ginoglou, Agorastos, & Hatzigagios, 2002;Izan, 1984;Laitinen & Laitinen, 2000;Lakshan & Wijekoon, 2013;McGurr & DeVaney, 1998;Zapranis & Ginoglou, 2000). Therefore, we also control for ROA, defined as operating income before interests and taxes over total assets. According to previous literature (e.g., Bradshaw, Richardson, & Sloan, 2001;Sloan, 1996), we expect a negative relation between audit qualifications and ROA 5 because, from the auditor's perspective, lower ratios mean a higher probability of corporate failure. Dummy variables for each year were used to control for time effects. Table 1 provides a summary of all the variables. Table 2 presents the mean value, the median, the standard error, and the 10th and 90th percentiles of the main variables. The results show that the representatives of large shareholders account for 44.39 percent of directorships on the board and 78 percent of directorships on the audit committee. Pressure-sensitive institutional investors represent 7 percent of directorships on the board and 20.60 percent of directorships on the audit committee. In accordance with the international trend to increase the importance of institutional investors (Cuatrecasas, 2012;Li, Moshirian, Pham, & Zein, 2006), the proportion of directors appointed by institutional investors in our sample increases from 42.97 percent in 2004 to 45.45 percent in 2010 on the board and from 77.78 percent in 2004 to 79.57 percent in 2010 on the audit committee. The mean presence of independent directors is 30.03 percent on the board and 84.0 percent on the audit committee. These data provide evidence that the percentage of institutional investors, pressure-sensitive directors, and independent directors is higher on the audit committee than on the board.

Descriptive Statistics
In addition, the mean size of the company is 13.56 (log of the total assets); 86 percent of the companies are audited by one of the Big Four auditing firms; 27.03 percent of the directors of the board held shares; and the size of the audit committee, on average, is 3.5 members. Mean board size (unreported) is 10.65 members. Finally, the level of leverage of the companies is, on average, 58.64 percent; 12 percent of the companies reported losses in the previous year; and the companies report a return on assets, on average, of 3.43 percent. Table 3 presents the Pearson correlation matrix to test for multicollinearity. The correlation between most of the pairs is low, generally below 0.3, and not significant. None of the correlation coefficients is high enough (>0.80) to cause multicollinearity problems (see Archambeault & DeZoort, 2001;Carcello & Neal, 2000), except the pair SENSITAC-SAV_BANKAC, which is correlated by construction. According to these results, we can, therefore, conclude that the models are free of multicollinearity problems. Table 4 shows the mean difference of INST, INDEP,  SENSIT, COM_BANK, SAV_BANK, INSTAC, INDEPAC, SENSITAC, COM_BANKAC, and SAV_BANKAC between firms with unqualified and qualified audit reports to test for the presence of differences in means between both groups of companies. The analysis of the results reveals that the presence of institutional investors, pressure-sensitive institutional investors, and savings banks on the board (INST, SENSIT, and SAV_BANK) and on the audit committee (INSTAC, SENSITAC, and SAV_BANKAC) is higher in companies that receive unqualified audit reports. This finding implies that institutional investors, pressure-sensitive institutional investor, and savings banks directors on the board and on the audit committee enhance the quality of the financial information. However, the mean difference for independent and commercial bank directors on the board and on the audit committee between unqualified and qualified audit reports is negative and positive, respectively, but neither is statistically significant. Therefore, institutional investors, pressure-sensitive institutional investors, and savings banks directors appear to exert much more control on the board and audit committee than independent and commercial banks directors to enhance the quality of the financial information. Table 5 shows the results of the logistic regression for the board. We build three models. Model 1 analyzes the proportion of institutional directors (INST) and independent directors (INDEP) on the board. Model 2 examines only the variable proportion of the board directors who are representative of pressure-sensitive institutional investors (SENSIT). Operational income before interest and taxes over total assets

Regression Results
Model 3 examines the proportion of the board that represents commercial banks (COM_BANK) and savings banks (SAV_BANKS). The Chi-squared test shows that the three models are statistically significant at 1 percent. According to our predictions, Model 1 of Table 5 shows that the variable institutional investors sitting on the board (INST) offers the expected sign and is statistically significant at the 5 percent level. Thus, we can accept Hypothesis 1a and conclude that the proportion of institutional investors sitting on the board enhances financial reporting quality as their presence reduces the likelihood that the auditor issues quali-fied audit reports. The variable proportion of independent directors sitting on the board presents the expected sign, but it is not statistically significant. This finding shows that institutional investors on the board exert much more influence than other board members regarding the demand for high quality of the financial information. This result is in line with prior studies (e.g., Almazán et al., 2005;Borokhovich et al., 2006;Brickley, Lease, & Smith, 1988;Bushee, 1998;Ferreira & Matos, 2008;Ljungqvist et al., 2007;Ramalingegowda & Yu, 2012), which also provide evidence of the positive impact of this class of directors on firms. Mean, standard deviation, and quartiles of the main variables. Panel A and B show the continuous and dummy variables, respectively. IA equals 1 if the company receives a qualified audit report; INST is the proportion of institutional investors on the board; INDEP is the proportion of independent directors on the board; SENSIT, COM_BANK, and SAV_BANK are the proportion of directors who represent pressure-sensitive institutional investors, commercial banks or savings banks on the board, respectively; INSTAC equals 1 if institutional directors sit on the audit committee; INDEPAC equals 1 if independent directors sit on the audit committee; SENSITAC, COM_BANKAC, and SAV_BANKAC equal 1 if, respectively, pressure-sensitive institutional investors, commercial bank directors, and savings bank directors sit on the audit committee; SIZE is the log of total assets; BIGFOUR equals 1 if the company is audited by one of the Big Four auditing firms; DIREC_OWN is the proportion of shares held by directors; AC_SIZE is the number of members on the audit committee; LEV is the book value of debt over total assets; LOSS equals 1 if the company reported losses the previous year; and ROA is operational income before interest and taxes over total assets.  the previous year; and ROA is operational income before interest and taxes over total assets. ***Significant at 1%. **Significant at 5%. *Significant at 10%. Table 5 analyzes the influence of the pressuresensitive institutional investors on the quality of financial reporting. The variable SENSIT, which represents this category of directors, presents the expected sign and is statistically significant at the 5 percent level. As a result, Hypothesis 2a can be accepted, and we conclude that pressure-sensitive institutional investors (banks and insurance companies) sitting on the board increases financial reporting quality. This evidence is in line with prior research that reports that not all institutional investors, but only some types, exert influence on corporate decisions (Almazán et al., 2005;Brickley et al., 1988;Ferreira & Matos, 2008;García-Meca et al., 2013;Hartzell & Starks, 2003;Ramalingegowda & Yu, 2012).

Model 2 in
In Spain, the presence of institutional investors representing banks is more prevalent than institutional investors representing insurance companies. In addition, bank representatives on the board exert more control in the company and demand a high quality of the financial information, because poor-quality financial information can make renegotiating with creditors more difficult. For this reason, in Model 3 in Table 5 we analyze the impact of this type of director (COM_BANK and SAV_BANK) on the quality of the financial information. The results show, contrary to our predictions, that COM_BANK is not statistically significant. However, SAV_BANK presents the expected sign and is significant at the 5 percent level. Consequently, Hypothesis 3a is partially accepted because only the presence of savings banks on the board increases financial reporting quality as the likelihood that the auditor issues qualified audit reports will be lower. Prior research reports the relevant role that bank directors on the board play in the companies (Hadlock & James, 2002;Johnson, 1997;Ljungqvist et al., 2007;Lummer & McConnell, 1989).
In the three models all control variables show the expected sign, but only SIZE and ROA are statistically significant at the 1 percent (Model 1) and 5 percent level (Models 2 and 3). Therefore, these results provide evidence that large companies with high levels of return on assets are likely to provide higher financial reporting quality.
In sum, the analysis of the structure of the board shows that the proportion of institutional investors, pressuresensitive institutional investors, and savings banks directors enhances financial reporting quality. Similar results have been reported by Ramalingegowda and Yu (2012). Thus, these results reveal the important role that institutional investors on Spanish boards play as a mechanism of good corporate governance. Table 6 provides the results of the logistic regression for the audit committee. As with the board, we employ three models. According to the Chi-squared test, the three models are statistically significant at the 1 percent level. In Model 1, the variables that represent the presence of institutional investors and independent directors on the audit committee present the expected sign, but only the presence of institutional investors (INSTAC) is statistically significant at the 10 percent level. In line with Felo et al. (2003) andGarcía Osma &Gill-de-Albornoz Noguer (2007), independent directors on audit committees (INDEPAC) do not affect financial reporting quality. Thus, Hypothesis 1b is accepted. In Model 2, as predicted, the sign of SENSITAC, which represents the presence of pressure-sensitive institutional investors, is negative and statistically significant. Therefore, Hypothesis 2b is also accepted. In Model 3, neither COM_BANKAC nor SAV_BANKAC are statistically significant, although both have the expected sign. This last result can be explained because the audit committee, on average, has fewer members than the board. As a result, the presence of commercial and savings banks is likely to be smaller. Thus, Hypothesis 3b is not accepted. These conclusions reveal that the presence of institutional and pressure-sensitive institutional investors on the audit committee increases financial reporting quality, and then the companies where they are appointed are less likely to receive a qualified audit report. Thus, this evidence strengthens the role of institutional investors on the audit committee, and within this type of directors, pressuresensitive directors gain notable relevance. As with the board models, all control variables report the expected sign, but only the size of the company and the return on assets are statistically significant. In Models 2 and 3 the variable proportion of shares held by the directors is also significant. Thus, big and profitable companies whose directors held shares and where institutional investors and pressure-sensitive institutional investors sit on the audit committees are more likely to have higher financial reporting quality.

CONCLUDING REMARKS
The specific agency problems in European continental countries have led to an increasing presence of large blockholders as directors, especially directors appointed by institutional investors. Although considerable research has been conducted on institutional ownership, the literature to date has failed to reach a consensus on whether institutional investors perform a specific role in boardrooms. Thus, given the importance of institutional investors in allocating capital to corporations and their role in firm governance, an understanding of how their presence on boards affects financial reporting quality is undoubtedly needed. Our study contributes to the literature by providing evidence of the effect of directors appointed by institutional investors on financial reporting quality measured by the audit opinion. We study the effectiveness of institutional directors in Spain, the European country with the highest presence of institutional investors on boards of large publicly listed firms.
We propose that the type of business relations between firms and institutional investors is a key issue in describing the role of institutional directors and, thus, their effects on the quality of financial reporting. Accordingly, we focus on those directors who maintain business relations with the firm on whose board they sit (i.e., pressure-sensitive directors) and analyze their influence both on boards and audit committees. We also examine the specific role of bank directors on boards and audit committees and analyze their effects on financial reporting quality when they act as shareholders and directors. Our results suggest that institutional directors are effective as monitors of management, which leads to higher quality financial reporting. Thus, when boards are made up by a high proportion of institutional directors, they are more likely to be effective in protecting the credibility of the firm's financial reporting because they are also external directors and independent of management. The results suggest that, compared to independent outsiders, institutional directors may be more effective in overseeing management because they may have more firm-specific knowledge. In addition, institutional directors have a relatively close relationship with top management, and so they may also reduce the conflict between the board and top management. These results support the relevant role of institutional directors on boards and the lack of influence of independent directors in European countries, as suggested in the literature.
In addition, despite the fact that in the Spanish context the Spanish Unified Good Governance Code (2006) holds that the audit committee should be made up exclusively of independent and institutional directors, our results also show that only institutional, non-independent directors on audit committees influence financial reporting quality. This finding suggests that independent and institutional directors may play distinguishable governance roles on both boards and audit committees. The lack of significance of independent directors on both boards and audit committees may be related to the measure of independence, in communitarian studies in general, where there are many concerns that board members are not independent of those who nominate them. Another explanation could be the substitution effect between independent and institutional directors.
Consistent with the significant role of business relations with the firm, we find that directors appointed to boards and audit committees by pressure-sensitive investors have a higher impact on financial reporting quality. This finding confirms that institutions that invest in firms with the intention of holding substantial ownership blocks over a long horizon have stronger incentives to monitor the firm. Nevertheless, when analyzed separately, only savings banks representatives on the board increase financial reporting quality. This finding may be justified by the specific composition of these entities, where the regional and local governing bodies if institutional investors, independent, pressure-sensitive institutional investors, commercial bank, and savings bank directors sit on the audit committee; SIZE is the log of total assets, BIGFOUR is a dummy variable that equals 1 if the company is audited by one of the Big Four auditing firms; DIREC_OWN is the proportion of shares held by directors; AC_SIZE is the number of directors on the audit committee; LEV is the book value of debt over total assets; LOSS equals 1 if the company reported losses the previous year; and ROA is operational income before interest and taxes over total assets. ***Significant at 1%. **Significant at 5%. *Significant at 10%. exercise a decisive power in the firm strategy. The lack of influence of savings bank directors on audit committees may be related to their low representation on these committees. This study contributes to the literature by showing that one of the ways in which institutional investors play a monitoring role is through influencing financial reporting quality when they are both board and audit committee members. The findings also suggest that both researchers and policy makers should no longer consider institutional investors as a whole, because directors appointed by different types of institutional investors have varied impacts on financial reporting quality. The findings are pertinent given the concerns about the regulation and quality of auditing services. Finally, the results have significant implications for supervisors and regulators, whose role in safeguarding the financial system will benefit from an understanding of how the presence of savings banks and commercial banks in nonfinancial firms' boards affects financial reporting quality in a bankbased system. Thus, the results have implications for policy makers who are seeking a suitable model for board composition and as they define the role of independent directors. A greater discussion and analysis is required so that independent directors remain independent of the large shareholders and so that they are able to safeguard minority shareholder rights.