Print
E-MailBernard Black is the Nicholas D. Chabraja Professor of Finance at Northwestern University Kellogg School of Management, and Antonio Gledson de Carvalho is Assistant Professor at Fundação Getúlio Vargas School of Business at Sao Paulo. This post is based on his recent article forthcoming in the Brazilian Review of Finance. Related research from the Program on Corporate Governance includes The Elusive Quest for Global Governance Standards (discussed on the Forum here) by Lucian A. Bebchuk and Assaf Hamdani; What Matters in Corporate Governance? (discussed on the Forum here) by Lucian A. Bebchuk, Alma Cohen, and Allen Ferrell; and Learning and the Disappearing Association between Governance and Returns (discussed on the Forum here) by Lucian A. Bebchuk, Alma Cohen, and Charles C.Y. Wang.
This article provides an overview of the evolution of corporate governance (CG) in Brazil, over the decade from 2010-2019. Since the turn of the century, the government and many private institutions have adopted a number of measures to promote improved CG. In 2009, the Brazilian Securities Commission, Comissão de Valores Mobiliários (CVM), created a mandatory CG reporting system for publicly traded firms (formulário de referência), with reporting beginning in 2010. Using this rich data, we describe how CG practices in Brazil evolved over the period from 2010-2019.
Brazil constitutes a good ground to study the evolution of CG because strong and weak legal regimes for CG coexist, against the background of other legal rules and cultural influences that apply to both CG regimes: In 2000, the São Paulo Stock Exchange (then called Bovespa, succeeded by B3) created three listing levels with increased firm level corporate governance (FLCG) requirements relative to the regular listing level (for which minimum governance rules are set by law): Novo Mercado (NM, highest FLCG), Level II (L2), and Level I (L1). NM and L2 are identical, except that NM forbids the use of non-voting shares, while L2 allows it. L1 is similar to the regular listing, but somewhat stricter, mostly reflecting improved disclosure rules. NM and L2 require strong FLCG and foresee resolution of conflicts by arbitration which, according to Brazilian law, has to provide a resolution within 180 days. Thus, the two types of listings, NM and L2 (which we abbreviate as NML2), versus L1 and regular listing (which we abbreviate as L1R) can be seen as providing two distinct CG and dispute resolution regimes for listed firms. The requirements for each listing level have not meaningfully changed since 2000. In 2010, L1R had almost twice as many firms as NML2 (232 vs. 134). However, over time the number of firms in L1R has been decreasing while the number of firms in NML2 have been increasing (most initial public offerings since 2000 have been listed on NML2). By 2019, both types of listing had similar numbers of firms (164 vs. 163)
We rely on objective CG elements reported in Formulários de referência to build both an overall Brazil Corporate Governance Index (BCGI) and subindices for Board Structure (7 elements), Board Procedures (4 elements), Minority Shareholder Rights (6 elements), and Disclosure (8 elements). We then measure BCGI as the average of the four subindices.
We show that CG improved significantly between 2010 and 2019. The average BCGI increased by 16 points (from 49 to 65 points). This improvement has two components: firms adjusting their CG practices and an increasing proportion of high-standard listings (NML2) versus low-standard listings (L1R).
The adjustment in CG was not uniform across firms. The improvement was stronger for firms in NML2, but was concentrated in the first half of the sample period. For NML2 firms, BCGI went from 68 in 2010 to 81 points in 2015, but then largely leveled off, reaching 82 points over the period from 2017-2019. For L1R firms there was rapid improvement in 2010-2013, from 38 points in 2010 to 44 points in 2013, and continued gradual improvement after that, to 48 points in 2019. The average BCGI score in 2019 for L1R firms (48 points) was well below NML2 throughout the sample period. As the improvement was stronger for NML2 than for L1R (14 vs. 10 points), the overall gap in CG between the two groups of firms was higher in 2019 than in 2010.
The second component of the overall increase in average scores is the rise of NML2 versus L1R. Over the decade, the number of firms in NML2 increased from 134 to 163, while the number of L1R firms decreased from 232 to 164. Consequently, over the decade, the proportion of listings in NML2 increased from 37% (134/366) to 50% (163/327). This contributed to the increase in BCGI for the overall market.
In the first half of the sample period, both sources of CG improvement were important. However, firms do not adjust their CG continuously. In the second half, overall improvement reflects an increasing proportion of NML2 firms, plus gradual continued improvement in L1R CG levels; with nearly constant NML2 levels. One can decompose the 16 points increase in BCGI from 2010 to 2019 into two components: 13 points for the average increase in CG for the balanced panel of firms and 3 points for the increase in the proportion of NML2 listings.
The number of L1R firms shrank considerably over the sample period. However, averages scores of BCGI for a balanced panel L1R firms is very similar to what we obtained in the unbalanced panel (full sample) of L1R firms. This suggests that survivorship bias does not drive the improvement of BCGI in L1R. The firms that delisted had CG similar to those that remained.
CG improved across all subindices, for both NML2 and L1R firms. Overall improvement was stronger in NML2 than in L1R. Improvements were stronger for Board Procedures and Disclosure. The time pattern of changes is similar for the subindices as for BCGI: stronger improvement earlier in our sample period and continued but slower change after that, driven by L1R firms.
The full paper can be downloaded here.
