Ownership, board, and enterprise risk management

Otero, Luis; Rodriguez Gil, Luis-Ignacio; Durán Santomil, Pablo; Tamayo Herrera, Araceli Article Ownership, board, and enterprise risk management European Journal of Family Business (EJFB) Provided in Cooperation with: UMA Editorial, Universidad de Málaga Suggested Citation: Otero, Luis; Rodriguez Gil, Luis-Ignacio; Durán Santomil, Pablo; Tamayo Herrera, Araceli (2020) : Ownership, board, and enterprise risk management, European Journal of Family Business (EJFB), ISSN 2444-877X, UMA Editorial, Universidad de Málaga, Málaga, Vol. 10, Iss. 1, pp. 42-53, https://doi.org/10.24310/ejfbejfb.v10i1.6690 This Version is available at: https://hdl.handle.net/10419/231730 Standard-Nutzungsbedingungen: Die Dokumente auf EconStor dürfen zu eigenen wissenschaftlichen Zwecken und zum Privatgebrauch gespeichert und kopiert werden. Sie dürfen die Dokumente nicht für öffentliche oder kommerzielle Zwecke vervielfältigen, öffentlich ausstellen, öffentlich zugänglich machen, vertreiben oder anderweitig nutzen. Sofern die Verfasser die Dokumente unter Open-Content-Lizenzen (insbesondere CC-Lizenzen) zur Verfügung gestellt haben sollten, gelten abweichend von diesen Nutzungsbedingungen die in der dort genannten Lizenz gewährten Nutzungsrechte. Terms of use: Documents in EconStor may be saved and copied for your personal and scholarly purposes. You are not to copy documents for public or commercial purposes, to exhibit the documents publicly, to make them publicly available on the internet, or to distribute or otherwise use the documents in public. If the documents have been made available under an Open Content Licence (especially Creative Commons Licences), you may exercise further usage rights as specified in the indicated licence. https://creativecommons.org/licenses/by-nc-sa/4.0/ European Journal of Family Business (2019) 9, 5-20 http://dx.doi.org/10.24310/ejfbejfb.v9i1.5468 2444-877X/ © 2018 European Journal of Family Business. Published by Servicio de publicaciones de la Universidad de Malaga. This is an open Access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/). *Corresponding author E-mail: [email protected] EUROPEAN JOURNAL OF FAMILY BUSINESS http://www.revistas.uma.es/index.php/ejfb Revisiting Internal Market Orientation in family firms Sergey Kazakovab* aPrograma de Doctorado en Economia y Empresa Universidad de Málaga (Economics and Business Administration Phd. Programme at University of Málaga, Spain) bNRU HSE –Higher School of Economics, Moscow, Russian Federation Received 08 January 2019; accepted 29 November 2018 JEL CLASSIFICATION M310 KEYWORDS Internal Market Orientation; Job Satisfaction; Employee Commitment; Business Performance; Family Business Abstract The present conceptual paper depicts Internal Market Orientation (IMO) theory development conceptualization with a contemplation of new conditions, realities and technologies available to modern businesses in service industries. Based on the results of a conceptual study, this study proposes a novel IMO framework which reflects the noted global changes that affects family businesses. The denoted model introduces novelty variables including Information Communication Technologies (ICTs) and Outsourced Personnel structural constructs. They avail to measure the effect of IMO implementation on job satisfaction and employee commitment that, in their turn, exhibit a positive impact on business performance in service industries. CÓDIGOS JEL M310 PALABRAS CLAVE Orientación del Mercado interno; Satisfacción en el trabajo; Compromiso de los trabajadores; Desempeño del negocio; Empresa Familiar Revisión de la orientación del mercado interno en empresas familiares Resumen El presente estudio conceptual presenta el desarrollo de la teoría de la orientación del mercado interno (OMI) mediante la discusión de las nuevas condiciones, realidades y tecnologías disponibles para negocios modernos en empresas de servicio. Basado en los resultados de un estudio conceptual, esta investigación propone un nuevo marco OMI que refleje los cambios globales que afectan a las empresas familiares. El modelo indicado introduce variables novedosas tales Tecnologías de la Información y Comunicación (TIC) y las subcontrataciones de personal. Se valora la medición del efecto de la implementación de la OMI en la satisfacción laboral y el compromiso de los empleados que, a su vez, muestran un impacto positivo en el desempeño del negocio en empresas de servicio. INSTITUTO DE LA EMPRESA FAMILIAR EUROPEAN JOURNAL OF FAMILY BUSINESS 1 Cátedra Santander de Empresa Familiar Universidad de Málaga www.revistas.uma.es/index.php/ejfb ISSN 2444-877X 10 VOLUMEN 1 ISSUE EMPRESA FAMILIAR INSTITUTO DE LA Cátedra Santander de Empresa Familiar Universidad de Málaga ENERO - JUNIO 2020 Special issue. Strategic and Organizational Change. Research papers: Guest editor: Katiuska Cabrera Suarez, Santiago Lago-Peñas, Elena Rivo-López Entrepreneurial orientation, learning orientation, market orientation, and organizational performance: Family firms versus non-family firms Remedios Hernández-Linares, María Concepción López-Fernandez Dimension as a business strategy Sara Terrón-Ibáñez, María Elena Gómez-Miranda, Lázaro Rodríguez-Ariza Family Business in the health care sector: Past and future Francisco Reyes-Santías, Elena Rivo-López, Mónica Villanueva-Villar Ownership, board, and enterprise risk management Luís Otero González, Luis-Ignacio Rodriguez Gil, Pablo Durán Santomil, Araceli Tamayo Herrera Barriers to change in family businesses Jose Daniel Lorenzo Gomez How Do Family Responsible Ownership Practices Enhance Social Responsibility in Small and Medium Family Firms? Cristina Aragon, Cristina Iturrioz-Landart 71-84 71-84 71-84 71-84 71-84 71-84 EUROPEAN JOURNAL OF FAMILY BUSINESS 1 Cátedra Santander de Empresa Familiar Universidad de Málaga www.revistas.uma.es/index.php/ejfb ISSN 2444-877X 10 VOLUMEN 1 ISSUE EMPRESA FAMILIAR INSTITUTO DE LA Cátedra Santander de Empresa Familiar Universidad de Málaga ENERO - JUNIO 2020 Special issue. Strategic and Organizational Change. Research papers: Guest editor: Katiuska Cabrera Suarez, Santiago Lago-Peñas, Elena Rivo-López Entrepreneurial orientation, learning orientation, market orientation, and organizational performance: Family firms versus non-family firms Remedios Hernández-Linares, María Concepción López-Fernandez Dimension as a business strategy Sara Terrón-Ibáñez, María Elena Gómez-Miranda, Lázaro Rodríguez-Ariza Family Business in the health care sector: Past and future Francisco Reyes-Santías, Elena Rivo-López, Mónica Villanueva-Villar Ownership, board, and enterprise risk management Luís Otero González, Luis-Ignacio Rodriguez Gil, Pablo Durán Santomil, Araceli Tamayo Herrera Barriers to change in family businesses Jose Daniel Lorenzo Gomez How Do Family Responsible Ownership Practices Enhance Social Responsibility in Small and Medium Family Firms? Cristina Aragon, Cristina Iturrioz-Landart 71-84 71-84 71-84 71-84 71-84 71-84 6-19 20-32 36-43 44-55 56-65 66-79 European Journal of Family Business (2020) 10, 42-53 Ownership, board, and enterprise risk management Luis Otero Gonzáleza, Luis-Ignacio Rodríguez Gila, Pablo Durán Santomila, Araceli Tamayo Herreraa aDepartamento de Economía Financiera y Contabilidad, Universidade de Santiago de Compostela, España Received 2019-07-25; accepted 2020-01-21 JEL CLASSIFICATION G34, G32, D23 KEYWORDS Family ownership, Corporate governance, Enterprise Risk Management (ERM) CÓDIGOS JEL G34, G32, D23 PALABRAS CLAVE Propiedad familiar, Gobierno corporativo, Gestión integral de riesgos (ERM) Abstract This paper analyses the effect of family ownership and the characteristics of the board of directors on the implementation level of enterprise risk management (ERM) in Spanish non-financial companies. The sample consists of 162 Spanish non-financial companies listed on Spanish stock exchanges and markets during 2012–2015. The results obtained show that the relationship between the level of family ownership concentration and the implementation level of an ERM system has a non-linear structure. Therefore, a reduction in implementation for moderate ownership levels is observed, although this increases with high ownership values. Regarding corporate governance, our study confirms the importance of certain characteristics of the board of directors, such as the size and the figure of the shareholder director in the implementation of formal ERM systems. Propiedad, consejo y gestión del riesgo empresarial Resumen Este trabajo analiza el efecto de la propiedad familiar y de las características del consejo de administración sobre el nivel de implementación de la gestión integral de riesgos (ERM) en las empresas españolas no financieras. La muestra consta de 162 empresas españolas no financieras que cotizan en Bolsas y Mercados Españoles durante el período 2012-2015. Los resultados obtenidos muestran que la relación entre el nivel de concentración de la propiedad familiar y el grado de implementación del sistema de gestión integral de riesgos presenta una estructura no lineal, de modo que se observa una reducción de los niveles de implementación para niveles medios, pero que se incrementa en valores elevados de propiedad. Respecto al gobierno corporativo, nuestro trabajo confirma la importancia de ciertas características del consejo de administración como el tamaño y la figura del accionista-consejero en la implementación de sistemas formales de gestión del riesgo. https://doi.org/10.24310/ejfbejfb.v10i1.6690 2444-877X/ © 2020 European Journal of Family Business. Published by Servicio de publicaciones de la Universidad de Mlaga. This is an open access article under the CC BY-NC-SA license (http://creativecommons.org/licenses/by-nc-sa/4.0/). *Corresponding author E-mail: [email protected] L. Otero-Gonzlez, L. I. Rodríguez-Gil, P. Durn-Santomil, A. Tamayo-Herrera43 Otero-González, L., Rodríguez-Gil, L. I., Durán-Santomil, P., Tamayo-Herrera A. (2020). Ownership, board, and enterprise risk management. European Journal of Family Business, 10(1), 42-53. Introduction Ownership structure and the characteristics of the board can play an important role in the level of risk assumed by the company (Tufano, 1996, Boubakri et al., 2013). In the case of family businesses, previous literature has explained this relationship based on socio-emotional aspects (Gómez-Mejía et al., 2007 and Su and Lee, 2013) and in Agency Theory (Amihud & Lev, 1981; John et al., 2008). In general, they explained increased risk aversion and the incorporation of non-strictly economic incentives (not necessarily monetary), where capital preservation and business transfer determine risk taking. However, there has been little research on the relationships between ownership, the board of directors and risk management. Increased risk aversion could result in greater involvement in risk management, both through the adoption of Enterprise Risk Management (ERM) and coverage. Among the definitions of Enterprise Risk Management (ERM), COSO II1 defines corporate risk management as “a process carried out by the board of directors of an entity, its management and remaining personnel, applicable to the definition of strategies throughout the company and designed to identify potential events that may affect the organization, to manage its risks within the accepted level and to provide reasonable assurance regarding the achievement of objectives”. As is clear, all the people who are part of the entity must be involved, although we should highlight the role that the board of directors ought to play as the main driver of these strategies. Due to the link between the board of directors and ownership, the latter will also play a decisive role in the implementation level of this process. However, concentration of capital in the hands of family businesses can have a negative effect on the adoption of an Enterprise Risk Management (ERM) system and on risk coverage. This approach is proposed by Beasley et al. (2005) and Brustbauer (2016), who believe that the implementation of a risk management system (ERM) requires full support from the owners and awareness of the value it provides. Therefore, they consider that when the person who controls the company is a manager-owner and not a professional manager, it is more likely that there will be less involvement in the implementation of ERM. On the other hand, the existence of other strong investors, in particular, institutional, ones, could make the interest in incorporating ERM systems vary. Regarding the influence of the board of directors, authors such as Kleffner et al. (2003) consider that it is the most determinant factor of the company for implementing Enterprise Risk Management systems. However, this aspect has hardly been studied in the economic literature, particularly, factors that may be relevant such as types of directors, gender diversity and the size of the board. The aim of this paper is to evaluate how ownership and company governance affect the adoption of Enterprise Risk Management (ERM) models, as well as risk coverage programmes for Spanish non-financial listed companies. This paper makes several contributions to the literature that relate the level of assumed risk with ownership and corporate governance. Specifically, the aim is to evaluate the effect of family ownership concentration, as well as the influence of other shareholders with significant interests, on the implementation level of formal risk-management processes. The characteristics of boards of directors have also been included, meaning that it can also be a significant factor. Thus, it is one of the few papers that addresses this issue, while it also considers a large number of variables which are representative of risk management. The study is limited to the Spanish case, given that it is a market with a significant presence of family businesses and with heterogeneous characteristics that allow us to test the hypotheses considered. The results obtained show that the relationship between the level of family ownership concentration and the implementation level of the risk management system (ERM) has a non-linear structure, so a reduction in the implementation levels for moderate ownership levels is observed, although there is an increase for high ownership values. The presence of institutional investors is very decisive, affecting all the variables related to risk management very positively. Regarding corporate governance, our work confirms the importance of certain characteristics of the board of directors in implementing formal risk management systems. This paper is structured as follows: the theoretical reference framework is presented in the second section; next, the third section describes the sample and the variables and hypotheses used; the fourth section discusses the methodology and the results obtained; finally, the fifth section summarises the main conclusions. Previous literature and hypotheses Regarding the importance of the ERM system, different academic researchers have justified risk management based on the costs of non- 1. Committee of Sponsoring Organizations of the Treadway Commission (2004): Enterprise Risk Management - Integrated Framework. Otero-González, L., Rodríguez-Gil, L. I., Durán-Santomil, P., Tamayo-Herrera A. (2020). Ownership, board, and enterprise risk management. European Journal of Family Business, 10(1), 42-53. L. Otero-Gonzlez, L. I. Rodríguez-Gil, P. Durn-Santomil, A. Tamayo-Herrera 44 systematic risks involved for the company. Stulz (1996) states that risk management adds value by reducing the probability that the value is destroyed during financial crises and by reducing or eliminating the so-called “costly lower-tail outcomes” (Beasley, Pagach, and Warr, 2008; Baxter, Bedard, Hoitash, and Yezegel, 2008). For Nocco and Stulz (2006), ERM can create competitive advantages by allowing access to capital markets and other resources, while also helping managers and employees at all company levels to manage risk. Therefore, ERM helps to reduce the probability of there being /the risk of financial problems. In addition, ERM can also lower other types of costs, in particular, risk coverage costs and the so-called “costs of contracts”. For Hoyt and Liebenberg (2011), by including decision-making in all types of risks handled by the company, risk management expenses that may occur from their individualized treatment are cut and this allows for natural coverage of risks with different business activities. One of the first studies to investigate the implementation of Enterprise Risk Management, which was carried out by Colquitt, Hoyt and Lee (1999), showed via surveys that the role of risk managers was evolving in such a way that they faced an increasing number of risks. In addition, Kleffner, Lee and McGannon (2003) concluded that 31% of the members surveyed of the Canadian Risk and Insurance Management Society had found the current organizational structure and resistance to change as the main obstacles to implementating a risk management system (ERM). These authors showed that Canadian companies that had adopted ERM had done so by being encouraged by the board of directors. This implies that the factors that can be decisive in overcoming these obstacles and thus, favour implementing Enterprise Risk Management systems are related to the ownership structure and the characteristics of boards of directors, which are aspects that have seldom been studied in the literature on risk management (ERM). For this reason, in this section we review the literature and propose hypotheses regarding the effect that the concentration of capital in family businesses, the presence of institutional investors and the characteristics of the Board can have on a greater involvement in risk management. Family ownership Concentration of capital in the hands of family businesses is considered to have a negative effect on the adoption of a risk management system (ERM). As Brustbauer (2016) points out, implementing ERM requires a great deal of support from the owners and for them to be aware of the value it brings (Beasley et al., 2005; Brustbauer and Peters, 2013). That is why they consider that when the individual who runs the company is an owner-manager and not a professional one, it is more likely for there to be less involvement in the implementation of a risk management system (ERM). Brustbauer (2016) found in his study that family businesses have fewer incentives to implement a risk management system. In turn, Paape and Speklé (2012) point out that when the owners also manage the company and there are no agency problems between owners and managers, the value of implementing ERM systems is lower and, therefore, less likely to be supported. At the empirical level, he also shows that it is less probable for companies managed by their owners to invest in ERM. H1: Family businesses have less incentive to implement ERM systems. Institutional investors. The presence of institutional investors could lead to better risk management practices being applied in the company (Mafrolla, Matozza and D´Amico, 2016). One theory/hypothesis is that many of them have a small stake and who expect high quality information (Kane & Velury, 2004). On the other hand, Mafrolla, Matozza and D´Amico (2016) claim that when institutional investors have a higher stake, they perform professionally raising management standards and, consequently those of their risk system. In addition, Paape and Speklé (2012) state that as institutional investors are more powerful than individual ones, their presence will lead to a higher level of ERM implementation. At the empirical level, Brustbauer (2016) finds a positive relationship between institutional participation and the implementation of risk management systems (ERM), while Paape and Speklé (2012) find no evidence. H2: The presence of institutional investors encourages ERM systems to be adopted. Board of Directors and ERM According to Kleffner et al. (2003), the boost giv- en by the board of directors is the most important factor that influences the implementation of ERM in companies. The importance of the Board is also shared by other authors such as Beasley, Clune and Hermanson (2005), Desender (2007), Altuntas, Berry-Stölzle and Hoyt (2011) and Baxter, Bedard, Hoitash and Yezegel (2013), who maintain that Management teams and boards of directors have a significant influence on the implementation of ERM. Beasley, Branson, and Hancock (2009) defend this based on an increased demand for greater risk transparency with the aim of reducing the probability of possible fraudulent or opportunistic behaviour. Desender (2007) measures the risk management system (ERM) by using public information and finds that L. Otero-Gonzlez, L. I. Rodríguez-Gil, P. Durn-Santomil, A. Tamayo-Herrera45 Otero-González, L., Rodríguez-Gil, L. I., Durán-Santomil, P., Tamayo-Herrera A. (2020). Ownership, board, and enterprise risk management. European Journal of Family Business, 10(1), 42-53. the independence of the board of directors is not enough on its own to lead to higher levels of ERM, but only when the position of the general manager or chief executive officer (CEO) and the chairman of the company are held by two different individuals. Beasley (1996) shows a positive relationship between independent directors and ERM. Altuntas, Berry-Stölzle and Hoyt (2011) find via a survey that companies that report using ERM generally have better corporate governance and a more appropriate organizational structure for risk management. Baxter, Bedard, Hoitash and Yezegel (2013) state that companies with the highest quality of risk management (ERM) are those with better corporate governance, with the presence of risk committees and senior management boards. The size of the board is also another factor that can play a significant role due to its ability to control managers’ actions (Daud, Haron & Ibrahim, 2011). Finally, regarding gender diversity in Boards of Directors, it is considered that the presence of women provides differing varied points of view (Joecks, Pull and Vetter, 2013) so much so that females more often than not tend to be seen as being more averse to risk than their male counterparts when it comes to investing (Eckel and Grossman, 2002; Fehr-Duda, de Gennaro and Schubert, 2006; Eckel and Grossman 2008; and Borghans, Golsteyn, Heckman and Meijers, 2009). We believe that a greater presence of female directors can positively influence the implementation of a risk management system (ERM). H3: The size of the board and the presence of women positively affect the implementation of ERM and risk coverage. H4: The presence of shareholder-directors negatively affects the implementation of ERM. Empirical analysis The empirical analysis was carried out based on information obtained from the SABI Database and Morningstar Direct. The sample is formed of the 162 Spanish companies that are listed on the stock exchange, excluding financial and real estate companies. The data related to ownership and other economic-financial data were obtained from the SABI database. This information was complemented with the risk indicators available in the Morningstar Direct database. Several dummy variables obtained from the information in the listed companies’ reports were used as ERM indicators. The independent variables are mainly made up of the percentage of ownership in the hands of family or individual investors, the presence of strong investors and other indicators related to the characteristics of the board of directors. Thus, the aim is to analyse the impact that ownership and the characteristics of Boards of Directors have on the implementation of risk management systems. Variables used Next, the variables used in the work will be discussed. Dependent variables The dependent variables determine the implementation of a risk management system in each company, as well as the quality of the implemented system, based on whether or not they have particular characteristics, which are indicators of good practices in risk management. Table 1 shows the variables used, keywords used in the search and their description. It is simply considered if the company has a risk committee (Risk_committee) and a chief risk officer (CRO) in its organizational structure. It is also borne in mind if the company measures its risks with a risk map and has established risk tolerance levels. We use the variables ISO 31,000 and COSO as an indicator of having ERM being used in the company, which entails that it has an enterprise risk system. Finally, we have included three indicator variables of coverage for the main financial risks, exchange and credit risk being most worthy of mention. Table 1. Definition of Variables Related With Presence Of Risk Management System Name Key Word Specification Risk_com- mittee Risk Committee Existing risk committee in the company CRO Chief Risk Officer, CRO Presence of a manager in charge of the company's risk management Risk_map Risk Map Existing risk map in the company Risk_toler- ance Risk Tolerance Existing risk tolerance level ISO 31000 ISO, 31000 Monitoring of the ISO 31000 standard COSO COSO Monitoring of the COSO framework Cov_int_ rate Derived Financial Instruments, Coverage Existing financial instruments dedicated to risk coverage of interest rate variation Cov_exch_ rate Derived Financial Instruments, Coverage Existing financial instruments dedicated to risk coverage of currency exchange rate variation Cov_credit Derived Financial Instruments, Coverage Existing financial instruments dedicated to risk coverage of credit rate Source: own elaboration. Otero-González, L., Rodríguez-Gil, L. I., Durán-Santomil, P., Tamayo-Herrera A. (2020). Ownership, board, and enterprise risk management. European Journal of Family Business, 10(1), 42-53. L. Otero-Gonzlez, L. I. Rodríguez-Gil, P. Durn-Santomil, A. Tamayo-Herrera 46 As observed in Table 2, only 20% of companies report the presence of a risk committee, and to a lesser extent (in 9% of cases) of a risk manager. However, it is quite common for Spanish listed companies to measure their risks (57%) and use the risk map in decision making while about 35% adopt formal risk management policies implemented in accordance with the COSO or ISO standard instead. Finally, interest risk is the most common form of coverage, followed by exchange and then credit risk. Table 2. Descriptive data of independent variables Variable Obs Mean Std. Dev. Min Max risk_committee 577 .2062392 .4049553 0 1 cro_ 577 .0918544 .2890709 0 1 risk_map 577 .5719237 .4952293 0 1 iso_31000_ 577 .0433276 .2037701 0 1 coso_ 577 .3015598 .4593334 0 1 risk tolerance 577 .5459272 .4983182 0 1 cov_interest_rate 577 .4592721 .4987709 0 1 cov_exch_rate 577 .3379549 .4734235 0 1 cov_credit 577 .1975737 .3985142 0 1 Source: own elaboration. Table 3 includes the correlations between the dependent variables, where it can be observed that in general, the values are positive, in line with, expectations, since they are representative variables of risk management. Ownership variables Firstly, we considered a continuous variable that represents the percentage of capital concentrated in individual investors or families (Famcont). As shown in Table 4, stakes in Spanish family-run businesses capital are very common, with an average value of 40%. In 10% of cases, controlled capital exceeds 85%, while in another 10% there is no presence of families or other individual investors in the shareholding. Table 4. Distribution of representative family ownership variables Probability Values Obs 1% 0716 5% 0Mean 10% 00.4096369 25% 0.2 DT 50% 0.35 0.2865664 75% 0.65 Skewness 90% 0.85 0.4016694 95% 0.95 Kurtosis 99% 0.95 2.116003 Source: own elaboration. The presence of other owners, in particular, investment funds (FIcont), has also been considered. Thus, the aim is to evaluate to what extent the presence of other relevant partners can influence the implementation of ERM. As previously Table 3. Correlations between the dependent variables Risk_ committee Cro_ Map_ risk~_ Iso_ 31000_ Coso_ Tolerance~_ Cov_ interest Cov_ exch cov_ credit Risk_committee 1 Cro_ 0.2828 1 Risk_map~_ 0.3111 0.2145 1 Iso_31000_ 0.2492 0.2271 0.1325 1 Coso_ 0.4024 0.1833 0.3471 0.2311 1 Tolerance~_ 0.1896 0.2298 0.4210 0.1770 0.2731 1 Cov_interest_rate 0.1663 0.2608 0.2140 0.1626 0.2431 0.2188 1 Cov_exch_rate 0.2516 0.2929 0.2109 0.2259 0.2491 0.2984 0.3709 1 Cov_credit 0.0913 0.0532 0.1390 0.0868 0.0913 0.1990 0.2152 0.2252 1 Source: own elaboration. Independent variables Taking the above into account regarding the determinant factors for adopting an Enterprise Risk Management model or ERM, we have considered the variables listed below to specify our explanatory model for the determinants for implementing an Enterprise Risk Management system in the company. stated, the existence of multiple relevant shareholders can positively influence risk taking and management (Mishra, 2011). Variables related to the characteristics of company governance Variables related to Corporate Governance of companies have also been considered specifical- L. Otero-Gonzlez, L. I. Rodríguez-Gil, P. Durn-Santomil, A. Tamayo-Herrera47 Otero-González, L., Rodríguez-Gil, L. I., Durán-Santomil, P., Tamayo-Herrera A. (2020). Ownership, board, and enterprise risk management. European Journal of Family Business, 10(1), 42-53. ly, information regarding the number of members that make up Boards of Directors (Totalmembers). In addition, the number of women that make up Boards of Directors (Boardwom), and of shareholders who are members of Boards of Directors (Sharboard) have been calculated. In general, companies opt for boards with an average of 14 members, although in some cases they may have 40 representatives. Women participate in virtually all boards, which are outnumbered by men so that out of the 14 members mentioned above, women only account for 1.5 on average. In more than 85% of cases, managers are shareholders, an element that can contribute to aligning interests. Finally, we should point out that a high percentage of the members of Boards of Directors (31.2%) are also company shareholders. Table 5. Characteristics of Corporate Governance Variable Obs Mean Std. Dev. Min Max totalmembers 716 14.41899 8.238941 3 40 boardwom 716 1.586592 1.701349 0 8 sharboard 716 4.530726 4.313959 0 23 Table 6. Summary of ownership and corporate governance variables Name Specification Relating to company ownership Famcont % of capital in the hands of family FIcont % Investment fund held by companies that are listed on the stock exchange Relating to company governance Totalmembers Total members that make up the board of directors of the companies listed on the stock exchange Boardwom Number of female members on the Board of Directors Sharboard Number of shareholders that make up the board of directors Source: own elaboration. Control variables The level of ERM adoption is also related to the size of the company, since there are economies of scale and minimum sizes required to implement risk management programmes and these can incur very high costs. The size of the company is usually related to the diversification level. Therefore, larger companies can use their market power to obtain greater benefits (Ang et al., 1985) and have a greater capacity to face the effect of economic changes (Sullivan, 1978; Hardwick, 1997). On the other hand, smaller companies are affected by a number of financial disadvantages that result in economic restrictions, greater difficulties in acquiring medium and long-term financing (Hellmann and Stiglitz, 2000) and a higher financial cost (Melle, 2001). Thus, there are studies that identify size and sector as relevant factors (Colquitt, Hoyt and Lee, 1999; Beasley, Clune and Hermansom, 2005; Pagach and Warr, 2011; Baxter, Bedard, Hoitash and Yezegel, 2013), although the work by Liebenberg and Hoyt (2003) is inconclusive. While Liebenberg and Hoyt (2003) found no significant differences in the use of ERM from one company to another of a similar size or industry, Beasley, Clune and Hermanson (2005) found that the companies with a higher implementation of a risk management policies had risk managers, were larger and operated in the financial, insurance or education sectors. As for Baxter, Bedard, Hoitash and Yezegel (2013), they found that larger and more diversified companies had better ERM programmes. H5: Larger companies adopt ERM to a greater extent. There are also differing theories which have the aim of explaining the relationship between liquidity and risk. In this regard, Bonfim and Kim (2012) show that the relationship can be either positive or negative. Based on the agency theory of free cash flow (Jensen, 1986), a positive relationship is established between both variables, with the argument that there is a greater risk of inappropriate investment when there is a very high level of liquidity; this is because managers prefer to retain excess funds and have greater discretion, which at times, can be materialized by way of the implementation of investment projects having a negative net current value. On the contrary, Logue and Merviue (1972) and Moyer and Chartfield (1983), postulate a negative relationship between liquidity and risk, maintaining that high liquidity indicates a low level of short-term liability and therefore a lower risk, bearing in mind that a higher liquidity reduces risk because there are more resources available to meet the company’s obligations (Edge, 1998). In the initial investigation of the correlation between both variables, Beaver, Kettler, and Scholes (1970) found a negative relationship with risk. However, the empirical studies of Borde (1998), Rosenberg and McKibben (1973) and Pettit and Westerfield (1972) showed liquidity ratios to be positively associated with Otero-González, L., Rodríguez-Gil, L. I., Durán-Santomil, P., Tamayo-Herrera A. (2020). Ownership, board, and enterprise risk management. European Journal of Family Business, 10(1), 42-53. L. Otero-Gonzlez, L. I. Rodríguez-Gil, P. Durn-Santomil, A. Tamayo-Herrera 48 risk. Nonetheless, the studies by Gu and Kim (1998) and Logue and Merville (1972) found no significant relationship between liquidity coefficients and risk. Finally, profitability can also be decisive for the risk level, because following financial valuation models, a positive relationship between the profitability and risk of all investments has been confirmed both theoretically and empirically (Blume and Friend, 1973; Fame and MacBeth, 1973). If the company carries out aggressive strategies to increase profitability, which can increase risk (Edge, 1998), it seems logical that the riskiest investments are those that promise the highest rates of return. However, and as indicated by Bowman (1980), and Chen (2013), if the company is very profitable, there is a lower chance of incurring losses and bankruptcy. Bowman (1980) maintains that correlations between the accounting measures of profitability and risk are negative for most of the sectors analysed, that is, the most profitable companies have a lower risk. Consequently, the most at-risk companies obtain worse results on average. The same result is found by Fiegenbaum and Thomas (1988). This double relationship is justified due to the double attitude towards the risk that managers may take based on the prospect theory formulated by Kahneman and Tversky (1979). According to these authors, managers show a risk-averse attitude when the expected result is higher than desired, but they are prone to risk when the expected result is lower than desired. The values of the control variables considered in the study are shown in Table 7 Summary of control variables. As can be seen, companies are heterogeneous in terms of size, liquidity, solvency and profitability. Table 7. Summary of control variables logta 572 12,37317 2,190238 6,598531 16,96684 ratliq_ 572 1,246647 1,204686 0,006 6,739 rroa_ 572 0,3376346 14,02567 -73,205 62,517 Source: own elaboration. While, in general there is not a very high correlation between the variables considered. Table 8. Correlation coefficient between control variables logta ratliq_ rroa_ logta 1 ratliq_ -0.1003 1 rroa_ -0.0917 -0.0319 1 Source: own elaboration. Table 9 shows all the variables considered, as well as the expected sign for each case. Table 9. Variables and initial hypotheses Variable Expected relationship ERM Relating to ownership Famcont - FIcont + Relating to Corporate Governance Totalmembers + Boardwom + Sharboard + Control variables Size + Liquidity + Profitability + Source: own elaboration. Methodololgy Most empirical studies carried out, which this paper belongs to, test the hypotheses established in the theoretical framework by means of conditional probability models. Therefore, we have chosen to apply a logit model to analyse the implementation of the variables related to ERM. This method establishes a linear relationship between the set of independent variables and the dependent variable. The dependent variable, which varies in the [0; 1] interval, is the logarithm of the ratio of opportunities or probabilities (odds ratio), probability of a certain event (default) and probability of its complement (no default). We take P as the probability of the event occurring (value “1”) and 1-P, the probability of the complementary event occurring (value “0”). It is a Bernoulli or dichotomous variable whose mathematical expectancy represents the random result of default or no default. Taking the variable as Y, the probability that the company has some of the issues evaluated in the differing dependent variables is , where represents the explanatory or independent variables (Caballo, 2013). The logistic regression model approximates the probability of the event “1” with the value of the explanatory variable as follows: [1] thus, [2] If we now express the number of times that an event occurs versus how many times it does not occur (odds-ratio or probability ratio), we can deduce the following: [3] obtaining a linear relationship in both the L. Otero-Gonzlez, L. I. Rodríguez-Gil, P. Durn-Santomil, A. Tamayo-Herrera49 Otero-González, L., Rodríguez-Gil, L. I., Durán-Santomil, P., Tamayo-Herrera A. (2020). Ownership, board, and enterprise risk management. European Journal of Family Business, 10(1), 42-53. independent variables and parameters Caballo, 2013). The proposed model is as follows: Table 11. Difference of means considering the Risk Map variable Risk Map=1 Risk Map=0 Diff. Std. Error Obs. Famcont 0.3809 0.4421 0.0612*** 0.0235 577 Ficont 0.2048 0.1587 -0.0461*** 0.0075 577 Total members 18.3818 12.5709 -5.8110*** 0.6597 577 Boardwom 2.0303 1.251 -0.7793*** 0.1436 577 Sharboard 6.6879 3.17 -3.5178*** 0.3412 577 Source: own elaboration. As shown in Table 12, the differences are significant in all cases, as with the previous variable, demonstrating again that, on the whole, the results are the same as those with a risk map. [4] The regression signs are interpreted in the following way: a positive sign shows an increase in the probability that and a negative sign means the opposite. It is important to interpret the sign and not the magnitude since the latter must be done in terms of marginal effects that are calculated as: [5] Mean difference Before analysing the results of the logit models, we performed an analysis of mean differences for some of the dependent variables. First, we chose the implementation of COSO as a proxy for ERM, where we observed that in general, companies that had implemented the standard have significant and positive differences in terms of ownership and the boards of directors. Thus, in general they have a higher percentage of institutional participation in capital, larger boards of directors, more women on boards of directors and more shareholders who are members of the boards. Table 10. Mean difference considering the COSO variable as proxy of ERM COSO=1 COSO=0 Diff. Std. Error Obs. famcont 0.3871 0.4158 0.0287 0.0255 577 ficont 0.2224 0.1690 -0.0534*** 0.0080 577 total members 20,1609 14,0521 -6.1088*** 0.7136 577 boardwom 2,2931 1,4392 -0.8539*** 0.1547 577 sharboard 7,6207 4,1290 -3.4917*** 0.3730 577 Source: own elaboration. As we can see in Table 7 the differences are all significant, like before showing once again that in general, companies with a risk map have more institutional presence and larger, more diverse management boards, with more directors as shareholders. This does not happen with the family ownership concentration variable, which we have seen has a negative result, so a higher concentration implies a lower implementation level of the risk map. Table 12. Difference of means considering the Risk Committee Risk Committee=1 Risk Commitee= 0 Diff. Std. Error Obs. Famcont 0.3475 0.4226 0.0751*** 0.0288 577 Ficont 0.2122 0.1781 -0.0341*** 0.0093 577 Total members 20.3613 14.7336 -5.6277*** 0.8268 577 Boardwom 2.2857 1.5437 -0.7420*** 0.1773 577 Sharboard 7.3361 4.6223 -2.7139** * 0.4398 577 Source: own elaboration. The same applies to coverage (Table 13), which backs up how the differences are repeated in general with only the negative variable being the representative variable of family ownership once again. Table 13. Difference of means considering coverage Coverage int rates=1 Coverage int rates =0 Diff. Std. Error Obs. Famcont 0.3638 0.4439 0.0801*** 0.0232 577 Ficont 0.2011 0.1715 -0.0297*** 0.0075 577 Total members 19.1811 13.1026 -6.0786*** 0.6501 577 Boardwom 2.1962 1.2724 -0.9238*** 0.141 577 Sharboard 6.5245 4.0417 -2.4829*** 0.3539 577 Source: own elaboration. Results of the logistic regression In Table 14 we can see that the Famcont variable is significant in six of the estimated models, showing a U-shaped relationship. This means that in general, a greater concentration of capital in the hands of family businesses leads to a lower likelihood of the company adopting risk management