5.2 Previous studies on compliance

This section describes the previous studies performed on compliance and more in particular the comply or explain principle as far as this is of interest to the underlying study. Studies regarding specific firm or country corporate governance characteristics or corporate governance ratings in relation to firm performance (e.g. La Porta, Lopez-de-Silanes et al. 1998) (Gompers, Ishii et al. 2003) are disregarded, hence the more specific scope of this study on code compliance and the comply or explain principle. Studies on code compliance - unlike studies on corporate governance characteristics - are quite rare, maybe even in their infancy and (with one exception (RiskMetrics Group 2009)) focus on one country, hence the added value of the underlying study. Most studies conducted - as discussed below - regard the relation between compliance and performance. Their diffuse outcomes are mentioned briefly below. Subsequently the focus is on the reasons for achieving or not achieving high compliance, the often reoccurring deviations and the reasons provided for non-compliance; all for the purpose of understanding and clarifying the specific choices made in the underlying study (see section 5.4).

Several empirical studies have tried to reveal whether higher rates of code compliance are related to superior financial performance or lower rates with inferior financial performance (Talaulicar and Von Werder 2008) (MacNeil and Li 2006) (Bassen, Prigge et al. 2008) (Weir and Laing 2001) (Arcot and Bruno 2007). The findings are, however, premature, diffuse, and difficult to compare due to the differences in the empirical studies conducted. Although MacNeil en Li are rather negative regarding the comply or explain principle and the self-regulatory status of the Combined Code, they conclude, based on their rather limited study of FTSE100 serial non-compliers, that a prima facie link between share price performance and investors' tolerance of non-compliance with the Combined Code exists (MacNeil and Li 2006, p. 487). Bassen, Prigge and Zöllner find that four of the German code provisions are significantly positively connected with the Tobin's Q ratio1 of the companies under review. When they review the compliance overall, a rather strong and negative connection is visible (Bassen, Prigge et al. 2008, p. 25). Weir and Laing conclude the same for the UK code; complete compliance does not appear to result in superior performance compared to partial or non-compliance, only the provisions on the remuneration committee have a positive effect (Weir and Laing 2001, p. 279). Goncharov, Werner and Zimmerman find that the degree of code compliance is value-relevant, which shows that the capital markets find the rules in the code meaningful and that there is capital market pressure to adopt the code. They also conclude that although there is capital market pressure suggesting a broad adoption of the German corporate governance code, no consistent answer can be provided to the question whether the compliance declaration provides value-relevant information (Goncharov, Werner et al. 2006). Moreover, Nowak, Rott and Mahr analyse that neither for the short nor long term do higher levels of code compliance have a (positive) impact on stock price performance compared to lower levels of compliance (Nowak, Rott et al. 2006, p. 33).

Possibly code compliance is more complicated and a differentiation in kinds of levels of compliance must be distinguished. Arcot and Bruno's results show that there is no relation between fully compliant versus non-compliant companies and performance. Nor does a relation exist when simply counting the number of deviations. Nevertheless, their own calculated corporate governance score (a sum of compliance and/or good explanations) shows a positive (0.0015) and significant (at 5% level) relation with performance. They therefore conclude that one size does not fit all; companies that depart from provisions for genuine reasons outperform all others (Arcot and Bruno 2007, p. 26). Arcot and Bruno state that their analysis: "provides support for the principle that in corporate governance regulation one-size-does-not-fit-all. We find that companies that depart from best practice because of genuine circumstances outperform all others. On the contrary, mere adherence to general accepted principles ofgood corporate governance is not necessarily associated with superior performance" (Arcot and Bruno 2007, p. 26).

Besides 'one size does not fit all', opportunistic behaviour of firms is also detected. Shabbir shows that firms become more compliant when their stock market performance declines, that mergers and acquisitions tend to decrease compliance, and that reorganisations and restructurings are followed by an increase in compliance as well. Shabbir states that "companies tend to 'ride the waves' becoming more compliant when the going gets tough, and less so as performance improves" andinhis opinionthisclarifies whycompliance possibly does not relate to performance (Shabbir 2008, p. 1). Talaulicar and Von Werder argue that it makes a difference how seriously the code is implemented and whether the discretion left by the code norms is utilised appropriately. Moreover, like Bassen, Prigge and Zöllner and Weir and Laing, they argue that the (non-)compliance with specific code provisions is probably more influential than the issue of total code compliance (Talaulicar and Von Werder 2008, p. 256).

Talaulicar and Von Werder justly conclude that more in-depth insight in the concrete form of code conformity is essential to help predict why some firms or countries feature specific corporate governance arrangements, as well as to analyse the consequences of those for the firm performance (Talaulicar and Von Werder 2008, p. 256).2 Future research should indeed - as the underlying study does - review the concrete form of code conformity together with underlying causes and not solely relate compliance to performance.

In 2006 Arcot and Bruno conducted a study of 245 non-financial UK companies for the period 1998-2004 (Arcot and Bruno 2006) (Arcot, Bruno et al. 2010). They documented the degree of compliance and more importantly the quality of the explanations provided in the case of deviations from the code provisions. Although they conclude, based on their univariate analysis, that the compliance of the companies does increase during the years under review, some drawbacks in the system are highlighted. They find that for an average of 17% of the deviations no explanations are provided. Moreover, in 51% of the cases the explanations are standard and uninformative and this even worsens for a company in which agency problems are likely to be serious (Arcot and Bruno 2006, p. 35). Von Werder, Talaulicar and Kolat examine the corporate governance statements of 408 companies listed at the Frankfurt Stock exchange that have to comply with the German corporate governance code as adopted in 2002 (Von Werder, Talaulicar et al. 2005). They distinguish some neuralgic provisions of which it is expected that they will not be complied with within the near future by more than 10 per cent of the companies. These concern i.a. personal liability, board compensation and the structure of the supervisory board. Moreover, Von Werder, Talaulicar and Kolat conclude that the acceptance of the code provisions tends to increase with the size of the companies, as will also be researched in the underlying study. Hooghiemstra and Van Ees analysed the corporate governance statements of 126 Dutch companies in 2005 and as a consequence thereof doubt the effectiveness of the comply or explain principle, since the explanations for non-compliance are relatively standard and not built upon firm-specific circumstances. The underlying study uses more or less the same subdivision in reasons for non-compliance as in the study of Hooghiem-stra, Van Ees and Van der Laan (see section 5.3.3 and Annex II). Hooghiemstra, Van Ees and Van der Laan conclude that an emerging 'one size fits all' approach is visible which they consider not to be in line with the fundamental logic of the comply or explain principle (Hooghiemstra, Van Ees et al. 2008, p. 29).

In some of the countries under review the compliance with the applicable code is also researched by, for instance, monitoring groups or (assigned) academics. As far as relevant, those studies were described in the previous chapter 4 when the corporate governance systems of the countries under review were researched. Since those studies provide country-specific trends in compliance and are neither comparable nor useful for the sections to come, no repetition of the outcomes is provided in this chapter but only references to chapter 4. The single international study on code compliance up till now is the "Study on Monitoring and Enforcement Practices in Corporate Governance in the Member States" performed by the RiskMetrics Group in September 2009. They examined the compliance with the applicable codes for 270 listed companies from 18 Member States for the financial year 2008 (15 companies per country). Although no annual developments in compliance can be reviewed, the study of the RiskMetrics Group showed some interesting results (Risk-Metrics Group 2009, p. 13); 86 per cent of companies they reviewed provide some kind of comply or explain information regarding a corporate governance code and 23 per cent of those companies state that they comply with all the code provisions. The companies reviewed have an average of approximately three explanations per company. Code provisions on the board of directors and on remuneration are the provisions explained most often. Other code topics repeatedly explained are shareholder rights and duties, disclosure and audit. The explanations provided for these deviations mainly involve the presence of an important shareholder, the specificity of the companies' activities and contracts set up before the implementation of the code (RiskMetrics Group 2009, p. 14). The average number of deviations is higher for mid-cap companies than for large-cap companies. In the end the RiskMetrics Group concluded that the comply or explain principle enjoys wide acceptance, although the quality of explanations is mainly considered to be at an unsatisfactory level and should be remedied by strengthening the comply or explain principle itself and by strengthening the role of market-wide monitors and statutory auditors (RiskMetrics Group 2009, p. 18).

Based on the studies discussed above, it can be concluded that future research should - preferably from an international perspective and over a period of more than one year - review the different concrete forms of code conformity together with underlying causes (such as the manner in which the code is implemented, the 'one size does not fit all approach', opportunistic behaviour or company's size) for the purpose of studying code compliance and the comply or explain principle more in-depth and, ultimately, to provide sufficient recommendations.