“Who cares about director independence?” P. Santella, C. Drago, G. Paone EALE 2007: “Corporate Law and Economics and Corporate Governance”. Copenhagen Business School (CBS) 13 September 2007 Independent directors and company disclosure: the case of Italy in 2003 The paper is available free of charge on IDEAS: and SSRN:

P.Santella, C.Drago, G.Paone2 Independent directors and company disclosure in Italy Data and methodology Data are collected for 40 companies making the S&P MIB index in 2003: 284 independent directors (101 belong to not financial companies and 183 from financial companies) Two sets of independence criteria (EC and Preda) and 12 single criteria considered (7 EC and 5 Preda) Every rate for each criterion is assigned (“from the investor point of view”) using three different qualitative outcomes: “yes” compliance, “no” not compliance “ns” not disclosure. We derive from the initial data the adjacency matrices of cross-directorships and dual directorships. Free float data (to study the relationships between free float and global rates) source: Consob

P.Santella, C.Drago, G.Paone3 Independent directors and company disclosure in Italy Exploratory data analysis: general compliance Global rates by company Company profiles Single criteria by company RankingScores RankingScores Network analysis By set of criteria (Preda and EC) By single criteria Company level Network level Network structure

P.Santella, C.Drago, G.Paone4 Independent directors and company disclosure in Italy Highest compliance and highest disclosure: 80 Fiat, Enel and Finmeccanica EC criteria: Companies that show higher compliance are Fiat, Enel, Finmeccanica. Such companies also show higher free float These companies are characterised by an higher level of compliance and disclosure (lower ns rates) but not always lower levels of not compliance “Ns” rates (yellow) “Yes” rates (blue) “No” rates (red) EC Global rates by company

P.Santella, C.Drago, G.Paone5 Independent directors and company disclosure in Italy The situation does not change much for Preda criteria. Similar companies on top of the ranking (Enel, Fiat and ENI), for which also we observe high levels of not compliance but also high levels of disclosure by considering the median level (60%) as global compliance rates we can build two visual groups. These two groups of companies are characterised by different data structures (higher compliance and disclosure versus lower not compliance and lower disclosure)

P.Santella, C.Drago, G.Paone6 Independent directors and company disclosure in Italy Lower compliance for three key criteria: other commitments and business relationships Mosaic plots: counts by rate are proportional to length (of the split rectangle fin\not fin)

P.Santella, C.Drago, G.Paone7 Independent directors and company disclosure in Italy

P.Santella, C.Drago, G.Paone8 Independent directors and company disclosure in Italy

P.Santella, C.Drago, G.Paone9 Independent directors and company disclosure in Italy In general higher compliance and disclosure for not financial companies either for preda and ec

P.Santella, C.Drago, G.Paone10 Independent directors and company disclosure in Italy Star plot of global rankings: Each star represents a company. Every rank of global rate is proportional to the length of the ray.

P.Santella, C.Drago, G.Paone11 Independent directors and company disclosure in Italy Higher free float companies: strong “environmental pressure” on higher disclosure (and higher compliance) Cooperative banks: higher not compliance and lower compliance for EC criteria

P.Santella, C.Drago, G.Paone12 Independent directors and company disclosure in Italy Higher free float companies: higher compliance but known lower performances in relational criteria Cooperative banks: lower performance on EC criteria Star plot of ranking by criteria: Each star represents a company. Every rank of criteria is proportional to the length of the ray.

P.Santella, C.Drago, G.Paone13 Independent directors and company disclosure in Italy Cross- directorships: financial companies form the centre of the network; not financial companies are the periphery. In the box the EC compliance rate and in brackets the incidence of cross-directorships on total number of independent directors

P.Santella, C.Drago, G.Paone14 Independent directors and company disclosure in Italy Mediobanca Unicredito Eni Autostrade STML Assicurazioni Generali Banche Popolari Unite Tim Mondadori Enel Mediaset Mediolanum Fondiaria SAI Banco Popolare Verona e Novara Autogrill RCS RAS Telecom Italia Pirelli E-Biscom Benetton Italcementi Luxottica Finmeccanica Alleanza Assicurazioni 1 (0.14) 4 (0.33) 2 (0.33) 1 (0.50) 4 (0.20) 5 (0.33) 1 (0.17) 1 (0.33) 1 (0.20)1 (0.25) 3 (0.43) 3 (1.00) 2 (0.40) 2 (0.50) 4 (0.31) 1 (0.33) 1 (0.20) 1 (0.14) 1 (0.33) 1 (0.13) 1 (0.11) 1 (0.10) 1 (0.05) 1 (0.25)1 (0.33) Banca Intesa 1 (0.14) Dual directorships: Enlarged network size but similar structure In the box the number of directors involved and in brackets the incidence of dual directorships on the total number of company directors.

P.Santella, C.Drago, G.Paone15 Independent directors and company disclosure in Italy Dual and cross directorships networks: The position defines the quality of compliance. Higher compliance for peripherical (and higher free float) companies In the box the average company compliance for EC and Preda

P.Santella, C.Drago, G.Paone16 Independent directors and company disclosure in Italy Cross-directorship links tend to be associated also with other links and the number of independent directors in these cases is lower (higher incidence) Higher free float companies tend to have a very low incidence of cross- directorships Cross-directorships incidence (red) Dual- directorships incidence (blue) Cross and Dual incidence considered together (green)

P.Santella, C.Drago, G.Paone17 Independent directors and company disclosure in Italy Horizontal and vertical red lines indicate the median for compliance rates (y axis) and free float (x axis). The 1- labels represent the partecipation to a cross-directorship network. Similar data structure for the two sets of criteria (companies that perform better in EC tend to perform better also in Preda) Positive linear association between free float and compliance for not financial companies. The same data structure does not hold for financial companies

P.Santella, C.Drago, G.Paone18 Independent directors and company disclosure in Italy Higher disclosure for not financial companies (only those characterised by higher free float); lower disclosure for not financial companies part of the networks.

P.Santella, C.Drago, G.Paone19 Independent directors and company disclosure in Italy Conclusion A) General low level of compliance for both financial and not-financial companies. B) In particular, not-financial companies with more fragmented shareholder base show higher compliance rates (responsiveness to market pressure). C) Financial companies: - show an even lower level of compliance; - are connected through cross-directorship networks; - tend not to show higher compliance the more fragmented their shareholder base (poor responsiveness to market pressure) D) Financial and not-financial companies are connected through dual- directorship networks.