Blockholder Heterogeneity and Audit Fees: Does Private Information Matter?

ABSTRACT

We analyze the impact of Small Blockholders’ (SBH – with 5% to 10% of voting rights) heterogeneity and their access to private information on the demand for audit services. By promoting enhanced audit services, we expect SBH to have a moderating effect on the relation between Large Blockholders (LBH – more than 10% of voting rights) and audit fees in a context of low shareholder protection. Drawn on Swiss public firm data over the 2002–2019 period, our results show that the presence of SBH flattens the concave relation between ownership concentration and audit fees found in prior studies. This moderating effect is mainly driven by the uninformed SBH, who given their lower – compared with informed SBH – access to private channels of information, are more likely to rely on audited public financial statements. Our findings contribute to the literature on the role of SBH in the company’s corporate governance.

Keywords:

• Audit Fees
• Blockholder Heterogeneity
• Small Blockholders SBH
• Principal-Principal Conflict
• Agency Theory

Disclosure Statement

No potential conflict of interest was reported by the author(s).

Notes

1 Along with a growing literature regarding the role of multiple large shareholders (MLS) in the governance of firms, we define LBH as shareholders owning 10% or more of the voting rights (Attig et al., 2013; Boubaker et al., 2017; Cai et al., 2016; Fattoum-Guedri et al., 2018), and we restrict the definition of SBH as those owning between 5% and 10% of the voting rights of a company. Alternative thresholds are tested in the Robustness section.

2 The relevance of our research feeds from the specificity of stakeholder countries (for instance Switzerland) characterized by ownership concentration and low investor protection. Prior research has evidenced that these two patterns are closely linked, suggesting that low legal protection increases the likelihood of ownership concentration (Aminadav & Papaioannou, 2020; La Porta et al., 1999). Switzerland shows one of the lowest minority investor protection in developed countries according to the minority investor index – Doing Business – published by the World Bank (2020). According to the doing business report published in May 2020, Switzerland obtained 50 in the minority investor protection score, compared to 84 in the UK, 71.6 in the US, 62 in Germany and 68 in France - giving to LBH the possibility to expropriate minority shareholders. Also, high ownership concentration, pyramid structure and the prevalence of multiple large blockholders in Swiss firms as evidenced by our descriptive statistical data make relevant to examine the impact of SBH presence, power, diversity, and nature on the relation between LBH and audit fees.

3 Aminadav and Papaioannou (2020) confirm the results found by La Porta et al. (1999): there are large differences in corporate control across legal families (i.e. the ownership of controlled shareholders is highest among French civil-law countries, followed by German and then Scandinavian civil-law countries).

4 For instance, Aminadav and Papaioannou (2020, p. 1192) argue that ‘while many researchers work with databases on cash flow, looking at voting rights is conceptually more appealing’. Also, the authors point out the difficulty to identify control from a myriad of complex corporate ownership structures.

5 Typical examples of ‘Shareholder’ Corporate Governance countries are common-law countries, such as the US, the UK, etc.

6 Typical examples of ‘Stakeholder’ Corporate Governance countries are code-law countries, such as Germany, France, etc.

7 In this paper, we focus on stakeholder countries given our focus on the impact of SBH in presence of LBH. LBH are expected to be more present in stakeholder countries than in shareholder countries. We invite the reader to refer to Barroso et al. (2018) for the developments about the shareholder countries. Indeed, there are few studies examining SBH and their contestability of LBH in shareholder countries, particularly the US and UK. First, insider trading regulation in these countries is very stringent compared to other countries, which discourages multiple large shareholders to invest in monitoring costs and incentivise them to sell their shares and vote with their feet (Admati & Pfleiderer, 2009). Second, ownership is more dispersed and, consequently, the conflict of interest between LBH and SBH is not prevalent compared to stakeholder countries. For instance, the UK minority investor protection regulation (The Takeover Code and the UK Company Law) creates obstacles to building controlling stakes by raising acquisition cost. ‘For example, an outside blockholder who owns 15% or more of the equity of a firm must make public their intentions of launching a takeover. Where a stake of 30% or more has been acquired, there is a compulsory tender provision for all remaining shares. Purchases of share blocks in excess 3% together with the identity of the buyer must be disclosed to the market’ (Franks et al., 2001, p. 240).

8 In Barroso et al. (2018), ownership structure is measured by the simple addition of cash flow rights of blockholders above 5%, with no consideration of the diversity of the shareholdings nor ultimate ownership.

9 https://www.reuters.com/article/sika-investors-idINL8N1CK4F4?edition-redirect=in

10 According to Ruigrok et al. (2006), Swiss banks still hold a strong influence over Swiss companies through three different mechanisms beyond direct ownership: first, as creditors; second, in some cases, they can send their representatives to the board of companies; and third, they can use the voting rights of depository shares in annual shareholders’ meetings.

11 In many cases, the annual report provides ownership information through the layers of ownership until reaching the ultimate shareholder. If this is not the case, we use the annual reports of each large shareholder through the layers of ownership.

12 This threshold is also used by Jentsch (2019) when examining the Swiss sample. ‘In Swiss corporate law, minority shareholders holding more than 10% of outstanding shares have certain important minority rights’ (Jentsch, 2019, p. 217).

13 We also use a 15% and 20% threshold. Results are similar using the 15% and 20% thresholds. However, beyond the 20% threshold, the statistical significance disappears. This is expected since we truncate the distribution at a point that is too close to the control threshold of around 30% documented by Barroso et al. (2018); see the Robustness Analysis section.

14 We classify Government shareholders as uninformed. These shareholders are usually characterized as not following strictly the market logic since they are often in strategic sectors in a (quasi-)monopolistic situation. Government large shareholders have less incentives to monitor managerial opportunism or controlling shareholders and to obtain trading information because of lack of financial motivation (see Isakov and Weisskopf (2014) and Lin et al. (2020)).

15 While a direct identification of Board members would be relevant to distinguish between Informed and Uninformed, we are unfortunately unable to identify those shareholders that are represented on the Board, unless the individual owns the shares. To overcome this difficulty, we have followed two tracks. First, for the individual or family shareholders, we can only directly identify the individuals or the family members that are themselves present at the Board, as they must disclose the shares they own. But individual or family shareholders may (and indeed often do) place trusted individuals (professional managers/directors) in the governing bodies of the company, including the Board, to represent them, with no mention of the link with the individual or family provided in the publicly available documents. For this reason, we have classified the individual or family shareholders as being Informed, only when we were able to connect insiders with the Family or individual SBH. This means that, to classify a Family shareholder as informed, the management or the board members must own a percentage of shares in the range of our definition of SBH (5–10%). As for the professional investors, we make the distinction based on the active approach to their investment (e.g. Private Equity), or their private access to financial information (e.g. Banks): we then assume that these categories of shareholders are Informed shareholders. Of course, being able to identify those shareholders that purely rely on Proxy advisors would certainly provide a clearer identification for the study. Unfortunately, we would need a more fine-grained data that are not available currently. This limitation is acknowledged in the conclusion. To get a direct access to insiders, we searched for a regulation mirroring the US Reg FD. That regulation does not exist in Switzerland, but we find a specific regulation about insider trading. According to Article 74a of the Listing Rules of the Swiss Stock Exchange (SWX) and the Directive on the Disclosure of Management Transactions, Equity transactions larger than CHF 100,000, must be disclosed within two trading days (Zingg et al., 2007). Introduced in 2005, Article 161 of the Swiss Criminal Code (SCC) tightens those guidelines, forbidding management transactions when they are based on confidential information, but only if those transactions have a serious impact on prices. In particular, it characterizes so-called ‘Primary insiders’, defined as – ‘Members of a management or supervisory body; and/or person who has access to insider information by reason of their ownership interest or function; shareholders or company professionals with access to confidential information of the company’. This characterization would have fitted well our need for defining the informed/uninformed shareholder status. Unfortunately, this list is not disclosed (our understanding is that this characterization is used ex-ante by the authorities in case of suspected illegal inside trading); we therefore cannot use this distinction as a proxy for our Informed /Uninformed variable.

16 As for the linear term, β₄, its moderating effect on the relation between FEE and LBH results mostly in moving the optimum of curve to the right or the left: If β₄ is significantly negative (resp. positive), then the optimum of the inverted U-shape between FEE and LBH will move to the left (resp. right). As presented in the Result section, this linear coefficient is non-significant in most regressions.

17 Using a 20% ownership threshold as in Faccio and Lang (2002) and Isakov and Weisskopf (2014), we find that 45% (37% in Isakov and Weisskopf (2014), 27.57% in Faccio and Lang (2002)) of the sampled firm are widely held, 41.5% (39% in Isakov and Weisskopf (2014), 48.13% in Faccio and Lang (2002)) are family controlled, 12.92% are controlled by non-family firm (17.76% in Faccio and Lang (2002)) and 0.52% have two ultimate owners. Our results are broadly consistent to similar samples.

18 This is the necessary condition to show that the curvilinear relation between audit fees and ownership concentration is altered (Dawson, 2014).

19 This variable is not kept in the main model in order to maintain the largest possible sample. It may be indeed an important omitted variable. However, the results in all our models remain robust with and without this variable.

20 We thank an anonymous reviewer for this suggestion.

21 Adecco 2004 Annual report.

22 http://server.capgroup.com/capitalinternational/private_equity/about/philosophy.html