Global indicator of adherence to the best practices of corporate governance: Mexican stock market corporations Indicador global de adhesión a las mejores prácticas de gobierno corporativo: sociedades anónimas bursátiles mexicanas

This article offers indicators of adherence to Corporate Governance (CG) in México useful to monitor compliance with common regulations at the Mexican Stock Market Corporations (MSMC). We use Exploratory Factor Analysis to obtain indicators that reflect different dimensions of CG. The sample comes from the answers provided by the MSMC to the questionnaire of adherence to the Code of Best Corporate Practices in the period 2010-2016. Thirteen composite orthogonal factors that measure the quality and degree of adherence of companies and sectors to CG best practices are extracted. These factors show the relative position of each company, as well as its temporal dynamics. The proposed indicators allow the analyst to determine the degree of compliance of CG in different dimensions: strategic management, risk management, audits, operations with related parties, board of directors, among other corporate practices and functions, among others. There are no performance indicators of compliance with the CG practices that are systematically used in the literature or in the practice of CG in Mexico. The global indicator of adherence to the best Corporate Governance practices (ADHECORP) and the thirteen composite indicators proposed in this article cover this practical gap and allow to enhance research on various CG issues in Mexico. JEL code: G34, G35, G38


Introduction
Corporate Governance (CG) is a fundamental element in improving economic efficiency, boosting growth, and fostering investor confidence. CG encompasses an entire set of relationships between the governing body of a company, its board of directors (Board), its shareholders, and other stakeholders, providing a structure for the setting of objectives by the company, as well as the means to achieve them and to monitor compliance (OECD, 2004).
The institutionalization of how the management structure should relate to shareholders and stakeholders drove the analysis of CG in companies (Richart, Martínez, & García, 2011), a line of research that, until then, had been considered part of agency theory, thus giving it its own identity. According to Farinha (2003), the debate and growing interest in CG stems from 1) financial scandals, e.g., high executive salaries and benefits, as well as executive-driven anti-blocking measures; (2) the effectiveness of control mechanisms; (3) the effectiveness of laws to limit the actions of managers in the steps they take; and (4) the open debate on the effectiveness and efficiency of the leading CG models and structures: American, British, and Commonwealth; the two-tier continental European; Japanese; the family-based Asian model (Tricker, 2015, pp. 147-155); and that of emerging economies.
Authors such as Gompers, Ishii, and Metrick (2003) and Brown and Caylor (2006) have sought to monitor shareholder rights (Cremers & Ferrel, 2014). Others, such as Sarkar et al. (2012) examine elements of CG such as board composition and performance, information disclosure, auditing, and even dividend policy. Some others consider the monitoring of specific CG characteristics, such as attendance at Board meetings and whether the guidelines are publicly stated (Brown & Caylor, 2006).
For emerging countries, Briano and Rodriguez (2016) assessed whether institutional factors determine the degree of CG compliance in major companies listed in emerging markets in Latin America, including Mexico. The Corpora#te Governance index used by the authors derives from country codes of corporate practice from 43 reagents grouped into four subindices that are, in turn, integrated with variables to which they assign the same weight. This study allows a comparison of the degree of compliance with CG practices in the four countries included.
Effective Corporate Governance in enterprises strengthens the growth of the economy by fostering the confidence necessary for the proper functioning of a free market economy (OECD, 2004). Monitoring and measuring such effectiveness are essential and should recognize the fact that there is no single CG model since each country has its institutional structure, which determines the checks and balances for compliance.
The interest in the construction of CG Indices (CGI) to quantify the efficiency of companies, reducing multiple characteristics in a few dimensions or composite indicators, has been the object of more studies. However, the diversity of CG models prevailing in the world, as well as the legal, regulatory, and normative conditions specific to each political and economic jurisdiction, make it challenging to propose standardized global indicators from a single model applicable to all contexts.
In Mexico, the Business Coordinating Council (Spanish: Consejo Coordinador Empresarial) issued the Code of Best Corporate Practices (Spanish: Código de Mejores Prácticas Corporativas, CMPC) in 1999, with its most recent version released in 2018. This code is the basis for the questionnaire to evaluate the degree of adherence and compliance by companies to generally accepted corporate practices. From this questionnaire, individual variables such as i) management discipline, ii) transparency, iii) independence, iv) the effectiveness of the board, v) the responsibility of the board and the directors, vi) rights and equitable treatment of shareholders, and vii) social responsibility, among others, are derived.
However, apart from the contribution of Briano and Rodriguez, there is no composite global indicator to monitor the degree of CG compliance of each company to the CG code in Mexico. There are also no composite indicators that reveal the various dimensions of CG, mainly derived from multivariate analysis allowing for systematic and consistent monitoring of CG dimensions, e.g., management, risk management, transparency, oversight, activities of the board of directors, among others.
The proposal herein contrasts with that of Briano and Rodríguez (2016), who assume equal weights for only four dimensions. Conversely, this proposal eliminates subjectivity in the selection of dimensions, as well as that of their loads.
This study includes all variables derived from the questionnaire, allowing the Factorial Analysis to group the indicators according to the correlation between the variables. This procedure also eliminates the subjective elements from the selection of variables. The resulting grouping in each factor comes from objective statistical criteria; therefore, each factor represents composite characteristics of corporate governance with analytical support. This approach also recognizes that there is no single CG model beyond the one that gives rise to the adherence questionnaire itself.
This article contributes with the generation of a Global Indicator of Adherence to Best Corporate Governance Practices (ADHECORP) that quantifies the degree of adherence of Mexican Stock Market Corporations (SAB) to the recommendations contained in the CMPC. The above is an indicator of the degree of compliance with the Mexican Best Corporate Practices constructed from 13 dimensions (compound factors) of the variables extracted from the SAB. A characteristic of the ADHECORP global indicator proposed in this article is its ability to monitor the main dimensions of adherence and compliance contemplated in the CMPC. Additionally, it suggests new dimensions that allow for the followup of stylized facts of corporate governance identified by international bodies such as Management, Transparency and Monitoring, and the Board of Directors (OECD, 2017).
Four sections comprise this article. The first section presents a review of the literature on the development of Corporate Governance indicators at the international and national levels. The second section presents the methodology used to obtain the CG indicator, in particular, the multivariate analysis used. The third section presents a descriptive analysis of the data collected from the CG Composite Indicators taken from the CMPC and the extraction of the Composite Indicators.
The study also shows a case analysis to illustrate the use of the indicators and, finally, the fourth section of this article is the conclusion.

Corporate Governance
There are different theories and models to explain the types of CG in the world. The traditional approach rests on the principle of maximizing wealth for shareholders, whose origin comes from the postulates of economic theory based on Adam Smith. Later, Adolf Berle and Gardiner Means, proposed the concept of separation of ownership and control (The Modern Corporation and Private Property), asserting that, if the ownership of large corporations disintegrates, the actions that their owners can take concerning the management of the enterprise are practically nil. In his article, The Nature of the Firm, Ronald Coase (1939) identifies the business as an instrument that is mostly at the service of economic efficiency. Agency theory sees the company as a node of contracts between the principal (shareholder) and the agent (management team), where it is necessary to align the conflicting interests of both parties to minimize management costs.
In emerging economies, such as Mexico, the institutional context makes the application of agency contracts more costly and problematic due to a high concentration of ownership and the absence of an efficient institutional CG, generating asymmetries between majority and minority shareholders-the Primary/central conflict (Watkins, 2013). Depending on the context, agency theory gives rise to the study of CG under two other approaches. One is the institutional theory approach, promoted by Thorstein Veblen (1904), founder of the institutionalist current of the social sciences, who explains that institutions do not exist as a function of social benefit, but rather because of the inertia of the system; institutions are the rules of the game. The stakeholder theory also arises as a contextual complement to agency theory, broadening the recognition of contracts or relations, implicit and explicit, applied to different groups in the creation and distribution of economic value, among which are identified the government, employees, and creditors, as well as networks of customers and suppliers.
Corporate Governance is a means by which various stakeholders exercise control over a corporation by employing certain rights as outlined in existing legal and regulatory frameworks as well as bylaws (Kose & Senbet, 1998). By definition, according to the OECD (2016), CG includes a series of relationships between the management of a company, its board, shareholders, and other stakeholders. This series of relationships provides a structure for the setting of objectives by the entity, determines the means to be used to achieve them, and monitors compliance. The existence of an effective CG system contributes to increasing the degree of confidence necessary for the proper functioning of a market economy. This degree of confidence results in the reduction of the cost of capital, which encourages companies to use their resources efficiently, thus driving growth.
We distinguish two dimensions of CG. First, the institutional (Paz-Ares, 2004), external or country-level CG (Allayannis, Lel, & Miller, 2012), which is imposed externally and generalized by laws and other regulations as well as by the regulatory institutions of each country, relating to a series of external mechanisms including legislation, regulation, labor markets, and corporate control markets (Watkins, 2013). The second dimension is called contractual, internal, or companylevel CG, which is the one assumed internally by each business organization and which adapts to its own needs and strategic objectives.

Codes of Good Governance
In a global economy, there are several standards set by the countries that make up the global bodies. One of them, the Organization for Economic Cooperation and Development (OECD), of which Mexico is a member, issued the "OECD Principles for Corporate Governance" in 1999, revised in 2004. These principles are a reference for each country to publish its own, adapting them to its regulatory framework and business culture (Consejo Coordinador Empresarial, 2010).
According to López and Pereira (2006), each country issues its code of good practices or code of good corporate governance, constituting general recommendations on the appropriate structure of the governing bodies and the proper behavior of their members. It is possible to say that, in a global scenario in which markets are increasingly interrelated, codes of good practice seek to improve CG when external market discipline and applicable laws are insufficient to guarantee transparency and the creation of the necessary value for investors and stakeholders. The CMPC includes an annex entitled "Questionnaire to assess the degree of adherence to the CMPC." Companies with securities registered in the National Securities Registry are required to disclose, no later than May 31st, the responses to the questionnaire corresponding to the immediately preceding fiscal year, following the general provisions and with provision 4.033.00 section XI of the Internal Regulations of the BMV (Valores, 2017). In this way, the information can be used as metrics by interested persons and organizations such as authorities, the stock exchange, researchers, analysts, and investors, among others.

Corporate Governance Indices, conceptual origin
A CG index has the potential to reveal unobservable latent dimensions impossible to measure in an analysis of individual variables. A CG index is useful for investors to rate the quality of corporate governance (Pillai and Al-Malkawi, 2016); to examine the relationship between adherence to good CG practices and financial performance (Erem, 2017); and as a basis for empirical verification of various CG relationships, among other uses. However, the generation of a CGI is a complex task given the set of dimensions proposed by agency theory and stakeholder theory discussed in the previous section, in addition to the possible subjectivity introduced by the researcher in defining a priori CG measures (Lagos & Vecino, 2011).
The construction of a CGI has involved considering different dimensions not directly observable, focusing on the protection of shareholder rights (Gompers et al. 2003;Byun, 2007;Brown & Caylor, 2006). Subsequently, other fundamental dimensions of CG were incorporated, such as the composition and performance of the board, the disclosure of information, the existence and efficiency of audit committees, as well as, eventually, the dividend policy (Byun, 2007). Our GDI also considers other several variables: the annual election of directors; the use or authorization of the poison pill 2 shareholders; the non-revaluation in the last three years; attendance of directors to board meetings; publicity and compliance of board guidelines by shareholders (Brown & Caylor, 2006).
The methodological variations of the studies (i) consider that CG dimensions contain elements with different loads; (ii) emphasize defense against hostile takeovers; (iii) consider industry sectors, market or geographic region; and (iv) regularly update the elements to be considered according to trends in global CG (Bhagat, Bolton, & Romano, 2008).
Most of the research related to the construction of CGIs comes from developed countries. However, there are rigorous efforts generated in developing countries and emerging markets: Briano and Rodríguez (2016), propose the measurement of CG quality from a sample of companies from four Latin American countries and consider 43 aspects contained in four dimensions: i) Composition and performance of the Board of Directors, ii) Shareholder rights, iii) Ethics and conflict of interest, and iv) Other information related to CG. Al-Malkawi, Pillai, and Bhatti (2014)  The list of CG elements is extensive, so deciding which features to include in each unobservable dimension based on a single theory or model can generate biases. The basis of this study is the recognition that there is no unique CG model since each country has its institutional structure, which determines the weights and counterweights for compliance. There is broad recognition by institutions such as the International Monetary Fund, the World Bank, and the OECD that there is no single model of corporate governance, and that it depends on the political, legal, social and structural context in each country (Pillai & Al-Malkawi, 2016). The number of variables or elements of Corporate Governance extracted from the membership questionnaires to the CMPC is vast, with a diverse subject matter. From these individual CG characteristics, it was possible to generate several dimensions of Adherence to Best CG Practices. These dimensions can, in turn, be used to create CG indicators proposed by bodies such as the OECD to measure corporate governance, transparency, oversight, and the activities of the Board of Directors (OECD, 2017). The following section describes the method used to extract these CG factors.

Method of Extraction of Corporate Governance Factors
While analysis of questionnaire variables may be useful in describing some individual or specific CG characteristics, a method is needed to compactly identify the sometimes-unobservable dimensions of CG identified by the literature. Many of the variables can measure or approximate similar attributes, so it is desirable to have a method that groups them naturally into more compact sets of characteristics. This study uses the Principal Component Analysis (PCA) and the Exploratory Factor Analysis (EFA) as methods for extracting indicators from the data obtained from the questionnaire to evaluate the degree of adherence to the CMPC. These indicators display the quality of the CG of the companies.
The PCA is a descriptive geometric method; its general objective is to discover the underlying structure in a set of n units of study, under a series of quantitative p variables. It is the statistical behavior of the variables, i.e., the correlation and distance to the means, that reveals the underlying structure of Corporate Governance in Mexico.
Once the underlying structure of CG data is discovered, EFA is used to substantially reduce the size of the set In summary, while PCA allows for the reduction of the size of the original multivariate information to few components, EFA provides for the creation of indicators, regardless of the variation of the data that explains them and the generation of composite indicators. 3

Definition of Factors
The questionnaires used to evaluate the degree of adherence to the CMPC provided information on the behavior of the companies regarding the best practices of CG in Mexico. The CMPC is available on the website of the Mexican Stock Exchange, www.bmv.com.mx. The issuers report the questionnaires indicated annually, and they reflect the particular perception of each company concerning the practices related to CG.
Each of the years analyzed recorded a total of 241 variables from 108 companies for selected companies listed during the years 2010 and 2016. However, there were cases listed without reporting data. The structure of the code recognizes various issues and recommendations of the bodies known as the Shareholders Meeting and Board of Directors, as well as the functions of Audit, Evaluation and Compensation, and Finance and Planning.
First, the study extracts the main components through the maximization process described in the previous section.
The type of matrix used is a correlation, which allows the variables to be normalized and made consistent in terms of their variances. However, under this criterion, there are more than 15 components, which, after a detailed analysis, contain redundant information. This study selected 13 factors that together explain 72.77% of the variance of the CG data in Mexico, maintaining as a general criterion the interpretation of the indicators and the proportion of variance explained.
3 For further methodological details and formalization of the PCA and EFA methods, see the works of Everitt (2005) and Mendoza (2010). The factorial method employs the components to estimate the individual loads for each of the thirteen factors that comprise the study. The Varimax rotation method identifies the maximum loads, and orders variables according to their relative importance. This ranking results in the thirteen factors in Table 1, and they comprise 91 variables from a total of 241 captured. These factors include all of the topics recommended by the CMPC and consider 32 of the 51 recommended practices.
The variables that dominate the first factor define the "strategic management" of the company (ADMEST).
The second factor comprises variables associated with the management, identification, control, analysis, and disclosure of risks, accompanied by concepts such as quality assurance or value creation, among others. This factor is called "risk management." (ADRISK).
The third factor is composed of variables related to the follow-up that the company gives to the audit function. It captures the audit process to the annual financial information. For this reason, this factor is called "audit follow-up." This negative load may reflect the fact that a more significant amount of independent directors exercises an enormous counterweight to the management of the activities of the related parties.
The seventh factor, efficiency of the basic functions of the board of directors (CONADM), includes variables related to the work of the board of directors such as compliance with evaluation and compensation functions and the finance and planning function; the corporate practices carried out. With a negative load, the adequate anticipation of the information for the board meeting. This factor measures efficiency in the fulfillment of functions and practices. From this bipolarity, it is possible to establish the hypothesis that there is (at least empirically) an inverse relationship between the fulfillment of the functions and the anticipation of timely information to the board. According to the literature review, there is no theoretical body of knowledge to explain this possibility. The result is probably due to how the function compliance variables are measured (binary variables) with regard to the information anticipation variable (discrete variable measured in days). Factorial analysis based on distances and variation detects these differences in the measurement and points out in these results the difference with a change of signs.
The eighth factor is made up of the following variables: category and professional activity of directors, the evaluation of confidential matters by directors, the specification of the type of directors in the annual report, and the presentation of the risk report dealt with by the directors. This factor is called "characteristics of the directors" (CONSEJ).
The ninth factor, called "transparency and ethics" (TRANSP), is composed of the following variables: verification of legal operations by the OIA; compliance with the code of ethics; disclosure of improper facts; protection of informants; confidentiality mechanism, among others; and, with a negative load, the integration of independent directors to intermediate bodies. The negative burden of this last factor suggests that greater transparency, ethics, and legality are associated with fewer independent directors of intermediate bodies.
The tenth factor quantifies the impact of independent directors, alternate directors, the number of board members, independent patrimonial directors, and the number of board meetings per year. This factor is known as "Board composition." The eleventh factor is the "relationship between directors and alternates" (COTISU), which identifies the fact that the company considers that having an independent proprietary director implies that the alternate director is also independent, in addition to the proprietary director proposing the person to occupy the position of alternate director.
The twelfth factor considers variables related to business ethics and legal compliance by the Board. We name this factor "Monitoring" (VIGILA). It integrates variables such as the verification of legal compliance by the Board; the performance of the auditing function; the information system and protection of informants; the existence of a business ethics code; the proposal for a declaration of a socially responsible company; and the policy for the use of company assets by directors and advisers.
The last factor, number thirteen, identifies the existence of a communication mechanism between proprietary and alternate directors; the effect of patrimonial directors; the establishment of a lower limit considering independent directors and proprietary directors; the link of the OIA with an internal audit; and, with a negative burden, recognizes the related directors. This factor is called "Director Independence" (INDCON). The load associated with the variable director is the highest weight and negative, indicating that the higher the number of related directors, the lower the communication between the proprietary and alternate directors or with the link between the OIA and the internal audit.

Application: Obtaining CMPC adherence factors of Mexican SABs
The relative score per company for each of the thirteen CG indicators results from the sum product of factorial loads and the value of each variable, reported by each company in the questionnaire. Table 2 displays, as an example, the individual score for each company that makes up the industrial sector, as well as the average for each factor. The score that reflects CG quality as measured by the Global Adherence to CG Best Practices index. 4

Global Indicator of Adherence to CG Best Practices
The previous year presented individual indicators of adherence to best corporate governance practices, that is, the average score obtained by each company and sector for the indicator of strategic management, risk management, and, in the same way, for the thirteen adherence factors. The Global Indicator of Adherence to Corporate Governance Best Practices in Mexico is a weighted average of each factor's variance, shown in the fourth column of Table 1. Thus, Table 2 displays in the last column, the average global adherence score obtained by each company and by sector in 2016. For instance, the overall score achieved by the Materials sector was 87.906. Table 3 presents the summary of the Global Adherence Indicator average scores for CG best practices achieved by each sector for the 2010-2016 period. The global indicator summarizes the adherence of each sector to corporate good governance recommendations. The sectors called Fast-moving products, as well as Materials, rank first and second throughout the seven years of the study, while the financial, health, and industrial sectors compete for the last places regularly.
Overall, the average global membership index exhibits very slight variations from 2010 to 2016 (see Table 3), except from 2011 to 2012, when the index dropped four points. If the dispersion of scores between sectors from the standard deviation (last line of the table) is measured, temporary stability is present, which can mean a generalized degree of adherence of companies to standards related to good CG or consistency in the application of good corporate practices within sectors.
The overall adherence rate can also be calculated within each sector to examine the relative position of each company. The tables of the authors, which the reader can obtain by correspondence, present the scores of the companies that make up all sectors. In the Materials sector, the case of the issuer Grupo Collado stands out, which obtained the lowest score in each of the years from 2010 to 2016 and which therefore reflects a very low adherence to the CMPC. Indeed, the Grupo Collado listing was suspended by the BMV in June 2010 for not presenting the report on the degree of adherence to the CMPC for 2009 (El Economista, 2010).
Finally, in addition to the examination by sector and company that can be carried out with the adherence indicators, subindices can be generated based on theoretical and normative CG in three areas emanating from the OECD Principles of Corporate Governance. In particular, the composite indicators herein can be grouped to examine particular attributes of adherence to CG, such as Management, Transparency and Oversight Functions, and the Board of Directors. Table 5 below displays the proposed grouping of adherence indicators.     The subindices obtained, orthogonal to each other, make it possible to monitor the underlying behavior of individual and joint corporate governance of companies concerning adherence and compliance in the CMPC. One of the advantages of this proposal of indicators is the elimination of subjective criteria in the selection of variables to form individual indicators and also the elimination of subjectivity in the choice of weighting factors, since PCA and EFA group the variables in an orderly manner according to their statistical behavior. The above contrasts with proposed indicators in the literature that rely on subjective criteria to select and weight the variables that make up the indicators, e.g., Briano and Rodríguez (2016). At the same time, this proposal recognizes that there is no single CG model but instead describes CG as a diverse set of dimensions.
according to the good practices recommended by the CMPC. The strongest cases and companies with weak CG practices are detected. This study identifies particular cases of companies with unfavorable events in their compliance with the regulations issued by the BMV, which receive low scores using the indicators in this study. The subindices also allow for the following of stylized facts of corporate governance, such as those identified by the OECD (2004) concerning Management, Transparency, and Board of Directors Monitoring.
The study demonstrates the application of subindices of CG by economic sectors and also by companies.
Additionally, it shows the usefulness of ADHECORP for monitoring the temporal variation of the essential characteristics of the degree of compliance with best practices.
The ADHECORP index is a country-specific indicator on a national scale. This prevents it from being implemented in other countries. Also, the annual frequency of the input data limits the timely reporting of the measurements. Some other challenges are the sensitivity of ADHECORP to changes in the questionnaires; and the possible omission of relevant information from instruments such as annual reports, individual corporate governance reports, press reports or other releases.
However, these restrictions do not invalidate the usefulness of the proposed indicators. In addition to monitoring and ranking the relative position of companies, composite indices make it possible to examine the degree of compliance, even of companies not listed on the Mexican stock market. This proposal allows for the evaluation of individual and joint characteristics of corporate governance, as well as to test hypotheses and theoretical relationships in this area in future studies. Future research using the ADHECORP index includes determining the degree of compliance with current CG precepts and the impact of financial (Gompers, Ishii, & Metrick, 2003), profit (Paz-Ares, 2004), financing (Durnev & Kim, 2003), and market value (Klapper & Love, 2002) variables, among others.