Corporate ownership and the information content of earnings in Poland

Abstract

In this article we test the influence of ownership structure on the information content of earnings in Polish-listed companies. Our investigation is based on the notion that in a weak corporate governance environment expropriation of private benefits of control is pervasive and manipulation of financial disclosure is a way to conceal those benefits to avoid disciplinary action. Concentrated ownership can act as a substitute for missing country-level corporate governance mechanisms to limit acquisition of private benefits of control, reducing incentives to mispresent financial situation and thus improving the quality of public accounting information. We find that weak country-level corporate governance mechanisms in the transition environment are best substituted by concentrated holdings of several investors rather than a single large shareholder. The information content of earnings increases when a few blockowners jointly hold between 25 and 50% of voting rights. We argue that the overall beneficial effects on corporate governance practices come from each blockholder's incentives to protect themselves from being expropriated by managers and other blockholders. We also find a positive impact of managerial holdings on the information content of earnings, and we argue that the holdings effectively align managers’ and investors’ interests.

Acknowledgements

We would like to thank two anonymous referees, Martin Bohl, Jana Fidrmuc, Meziane Lasfer, Mark Taylor (the editor) and participants at the Workshop on Corporate Governance, Corporate Restructuring and Corporate Finance in Transition Economies at Brunel University for their very helpful suggestions and generous comments on earlier drafts of the article. All remaining errors are our own.

Notes

¹ Ratings of Polish-listed companies demonstrate that companies either do not introduce policies to protect shareholder rights or create regulations that are clearly controversial and do not comply with international standards. Two editions of the corporate governance rankings were prepared by the Polish Forum for Corporate Governance in 2001 and 2003. Detailed information is available on www.pfcg.org.pl

² The ownership studies are complementary to other investigations that test the relation between the information content of earnings and various governance and institutional characteristics. For example, earlier literature analyses the association between the informativeness of earnings and board composition and board size (Vafeas, 2000), board leadership structure (Gul and Wah, 2002), levels of stock compensation (Behn et al., 2002), auditor quality (Teoh and Wong, 1993), cross-listing in the US (Lang et al., 2003), country's legal origin (Ball et al., 2000), and the type of country's financial system and accounting and tax regulations (Ali and Hwang, 2000).

³ Leuz et al. (2003) acknowledge that, holding private benefits constant, higher penalties for the expropriation of outsiders in strong corporate governance systems could potentially lead to higher earnings opaqueness as insiders may have stronger incentives to conceal their private benefits to avoid a penalty. However, the authors conclude that the empirical evidence suggests that the relation between the quality of corporate governance and the quality of accounting information is mainly driven by international differences in private benefits of control and hence earnings management decreases with the country's quality of corporate governance.

⁴ The Emitent system is a counterpart of the UK Regulatory News Service (RNS) and is aimed at ensuring efficient and safe transfer of information required by the Act on Public Trading in Securities of 21 August 1997. Established in 1999, the system enables anyone to access the information transmitted by a company to the public without the intermediation of a news agency.

⁵ For example, Jermakowicz and Gornik-Tomaszewski (1998) and Gornik-Tomaszewski and Jermakowicz (2001) investigate companies listed on the WSE between 1995 and 1997, and 1996 and 1998, respectively. Grosfeld and Tressel (2002) study large shareholders in Poland in the period 1994–1998, and Trojanowski (2008) tests block transfers at the WSE from 1996 to 2000. Jindrichovska and McLeay (2005) study companies listed on the Prague Stock Exchange in the period 1993–1999.

⁶ According to the Decree of the Council of Ministers of 16 October 2001 the definition of managing person is ‘a person who has major influence on the management of the issuer, including: a management board member, person acting as a management board member, proxy, administrator, commissioner or liquidator’.

⁷ 15 NIFs were introduced as a part of the Polish Mass Privatization Program to act as major shareholders in 512 large and medium companies selected for privatization. NIFs were allocated 60% of share ownership in these companies, with the remaining shares held by the state and employees. NIFs were owned by adult Polish citizens through universal share certificates, subsequently converted into shares in each of the NIFs (Puntillo et al., 1996). In 2002, two NIFs merged.

⁸ Notoria is a data vendor providing data on all companies listed on the WSE to many news agencies including Reuters, and its database includes the following information for each company: company name and address, name of the current CEO, and quarterly (annual) financial statements and main financial ratios for last five quarters (5 years). Financial statements consist of balance sheets, profit and loss accounts and cash-flow statements.

⁹ The average end-of-year exchange rate in the period under consideration was around 4.00 PLN/USD.

¹⁰ Jermakowicz and Gornik-Tomaszewski (1998) control for size of the company indirectly by introducing a dummy variable for companies that are greater than the sample mean value. Moreover they use total value of assets at the end of the fiscal year as a proxy for firm size.

¹¹ Variability of earnings and persistence of earnings are two more control variables commonly employed to investigate cross-sectional differences in the informativeness of earnings. To calculate these variables, Warfield et al. (1995) use data for 16 quarters, Gabrielsen et al. (2002) for at least 7 years and Yeo et al. (2002) for 8 years. The length of our earnings time series restricted by the short history of the WSE is inadequate to ensure reliable estimates of earnings persistence and variability in our study.

¹² At the end of 2002, there were 216 companies listed on the WSE, among them 16 financial firms and 14 NIFs.

¹³ Jermakowicz and Gornik-Tomaszewski (1998) have 139 observations for 53 companies listed on the WSE. In the subsequent study Gornik-Tomaszewski and Jermakowicz (2001) include 231 observations for 77 firms. Jindrichovska and McLeay (2005) cover 63 companies listed on the Prague Stock Exchange with 317 observations.

¹⁴ We believe that the small correlation between stock returns and index returns is not a result of nonsynchronous trading. As a robustness check we re-estimated betas correcting for possible nonsynchronous trading and the estimates were largely unaffected.

¹⁵ We are unaware of any earlier study that tests the relationship between the informativeness of earnings and ownership structure in a transition economy. We use the methodological approach from other studies based in emerging and developed markets (e.g. Warfield et al., 1995; Jung and Kwon, 2002; Yeo et al., 2002).