Benchmarking Greek Corporate Governance against Different Standards

• Date: 01 September 2007
• Author: Argiris Galarniotis,Annita Florou
• DOI: http://doi.org/10.1111/j.1467-8683.2007.00614.x
• Published date: 01 September 2007

Benchmarking Greek Corporate

Governance against Different

Standards

Annita Florou* and Argiris Galarniotis

The rapidly increasing worldwide focus on corporate governance has resulted in a prolifera-

tion of rating systems that proxy for governance quality. This study develops a governance

rating for Greek listed companies by benchmarking their governance structures against three

levels: (a) the minimum requirements under Greek regulation (lower level); (b) the incremen-

tal recommendations of the Greek code (middle level); and (c) the additional international best

practices, prescribed by the UK Combined Code (higher level). Using available data on 274 out

of 340 Greek listed companies in 2003 and based on information collected primarily from

annual reports, we f‌ind that the average governance rating at the lower level is 65.5 per cent;

this scoring reduces signif‌icantly as we move to the middle and higher level. The average

governance rating is 44 per cent. Second, governance scores increase with f‌irm size. Although

there may be good reasons explaining these patterns, we f‌ind that Greek companies do not

provide explanations, i.e. do not practice the “comply or explain” recommendation. Third, our

middle level aggregate governance rating is much lower than that reported by prior research

that uses a different data gathering and weighting approach. This divergence has important

methodological implications. Finally, we document a relatively high lack of transparency in

relation to Greek governance practices. This is a sign of “bad governance”.

Keywords: Corporate governance quality, governance ratings, governance disclosures

Introduction

The rapidly increasing worldwide focus on

corporate governance,particularly follow-

ing the recent spate of corporate scandals, has

resulted in a proliferation of rating systems

that provide a comprehensive evaluation of

the governance practices of listed companies.

Governance quality scores are developed by

rating agents and other professional bodies

(e.g. Institutional Shareholder Services, Gover-

nance Metrics International, etc.) as well as by

academics (e.g. Drobetz et al., 2004; Black et al.,

2006).

In response to this growing interest in gov-

ernance ratings, the current study aims to

assess the quality of the governance structures

and disclosures of Greek listed companies by

constructing a multi-level and as comprehen-

sive as possible corporate governance index.

Following the decline of the Athens Stock

Market in mid-1999, the Greek government

attempted to re-establish investor conf‌idence.

The Capital Markets Commission set up a

Committee on Corporate Governance, which

initiated discussions with market participants

and experts from the relevant f‌ields of corpo-

rations, auditors, legal practitioners and in-

vestors. These discussions eventually led to

the publication of a voluntary Code of Con-

duct, entitled Principles of Corporate Governance

(Committee on Corporate Governance, 1999).

The code includes some of the recommenda-

tions of the OECD Principles of Corporate Gov-

ernance (OECD, 1999). Three years later, the

Ministry of the Economy issued a law (No.

3016/2002), which, for the f‌irst time, obliges

the Greek listed companies to enforce a set of

governance guidelines. This law effectively

mandates compliance with a subset of the

*Address for Correspondence:

Department ofAccounting and

Finance, University of Mace-

donia, 156 Egnatia Street,Thes-

saloniki 540 06, Greece. E-mail

anf‌lorou@uom.gr; Tel: +30

(2)310 891666; Fax: +30 (2)310

891278.

BENCHMARKING GREEK CORPORATE GOVERNANCE AGAINST DIFFERENT STANDARDS 979

Volume 15 Number 5 September 2007

© 2007 TheAuthors

Journal compilation © 2007 BlackwellPublishing Ltd, 9600 Garsington Road,

Oxford, OX4 2DQ, UK and 350 Main St,Malden, MA, 02148, USA

principles contained in the code. It is worth

mentioning, however, that the Federation of

Greek industries has a very f‌irm position sug-

gesting that governance codes should be vol-

untary and the regulator should refrain from

legal enforcement.

The purpose of the study is to benchmark

the governance quality of Greek listed com-

panies against three levels: (a) the minimum

requirements of the Greek regulation (lower

level); (b) the incremental recommendations

of the Greek code (middle level); and (c) the

additional international best practices, pre-

scribed by the UK Combined Code 2000

(higher level). To construct our governance

index we manually collect data on 47 vari-

ables, which we then sort into seven main

dimensions. We equally weight variables

within each dimension to obtain total dimen-

sion score and similarly we equally weight

dimensions to obtain an aggregate corporate

governance rating at each level separately and

overall.

The contribution of the paper is two-fold.

First, it complements a recent governance

paper based on Greek data by Tsipouri and

Xanthakis (2004). They use a questionnaire

designed to quantify the self-reported compli-

ance of 120 Greek listed companies with the

recommendations of the Greek corporate gov-

ernance code (our middle level) and construct

a governance index based on their self-

reported data. The current study expands the

work of Tsipouri and Xanthakis (2004) in the

following ways:

• It evaluates the governance quality of com-

panies listed on the Athens Stock Exchange

(ASE) benchmarked not only against the

middle level (as in Tsipouri and Xanthakis)

but also against the lower level (i.e. current

Greek law) and the higher level (i.e. Com-

bined Code). In doing so, the current study

attempts to provide a multi-dimensional

and as comprehensive as possible gover-

nance rating for Greek listed companies.

• A common criticism of questionnaire-

constructed indices (as in Tsipouri and

Xanthakis) is that they are potentially sub-

ject to self-reporting biases (Drobetz et al.,

2004). The current study avoids this type of

bias by collecting data regarding corporate

governance structures from company

annual reports and corporate web sites.

Public domain documents such as these,

although not necessarily objective, are less

likely to contain unrepresentative data

because they are subject to scrutiny by ana-

lysts, shareholders, creditors, regulators

and the general public. On the other hand,

our methodology may not necessarily be the

most eff‌icient way of acquiring data on

actual governance arrangements of f‌irms.

• Assigning different weightings to different

governance dimensions (as in Tsipouri and

Xanthakis) also assumes subjective judge-

ment (Van den Berghe and Levrau, 2003).

However, the existing governance litera-

ture does not provide guidance on specif‌ic

variables and dimensions quantifying

governance quality (Balling et al., 2006).

Accordingly, this study weights all the

governance dimensions equally, thereby in-

creasing the transparency and relative ob-

jectivity of the rating system. Clearly, our

methodology assumes that all governance

dimensions are equally signif‌icant.

The second main contribution of the paper

relates to corporate governance disclosures. If

they are to be effective, corporate governance

mechanisms should not only be in place but

also, and more importantly, should be known

to be in place by investors. Therefore, disclo-

sure of governance arrangements is a matter

of interest to both investors and researchers

studying governance quality. This paper adds

to the international corporate governance lit-

erature by providing details on the transpar-

ency of governance practices of Greek public

companies. In doing so, we carefully docu-

ment non-disclosure of all measured gover-

nance criteria. We then calculate the ratings

in two ways: f‌irst, missing values (i.e. non-

disclosure) are treated as absence of the

variable under study and hence f‌irms are

penalised in assessing governance quality;

second, missing values (i.e. non-disclosure) are

excluded from the analysis. The discussion of

the results is based primarily on the method-

ology, which rates the absence of data as

negative.

Based on data available on 274 out of 340

Greek listed companies in 2003, we show how

the level of corporate governance quality dete-

riorates as we move from the lower standard

(average score is 65.5 per cent) to the middle

and higher standard (average scores are 42.3

and 24.3 per cent, respectively). The average

overall corporate governance rating is 44 per

cent. Evidence suggests that voluntary adop-

tion of more advanced and less binding gov-

ernance standards is a relatively uncommon

practice in Greece. Although there may be

good reasons justifying this (e.g. specif‌ic com-

pany attributes including f‌irm size and level

of regulation), we document that most Greek

f‌irms do not provide reasons for “non-

compliance”. Second, we f‌ind that governance

ratings improve with f‌irm size. For example,

the overall corporate governance rating of

FTSE/ASE 20, FTSE/ASE 40 and FTSE/ASE

980 CORPORATE GOVERNANCE

Volume 15 Number 5 September 2007 © 2007 TheAuthors

Journal compilation © BlackwellPublishing Ltd. 2007