The influence of CEO’s financial literacy on SMEs technological innovation: the mediating effects of MCS and risk-taking

Abstract

Previous literature showed mixed results about the impact of CEOs’ financial literacy (CFL) on small and medium-sized enterprises’ (SMEs) innovation. This relationship can be motivated by relevant variables, which are missing in the previous literature and make a difference as mediators. In this sense, based on the theoretical framework related to upper echelon theory and resource-based view, this study focuses on the mediating effect of risk-taking attitude and management control systems (MCS) variables. Empirical data from 310 SMEs gathered using a qualitative research questionnaire are analyzed using structural equation modeling methodology. Specifically, estimations are carried out considering the partial least square method. Findings show that MCS and managers’ risk attitudes fully mediate the relationship between financial literacy (FL) and innovation. Between these two mediating variables, the implementation of MCS stands out because it also enables the mediating effect of CEOs’ risk-taking in the CFL–technological innovation relationship. As the results do not support the significant direct relationship between FL and risk attitude, they confirm an indirect effect through MCS. Furthermore, based on the study findings, SMEs’ directors and owners, business associations, and public authorities can improve SMEs’ technological innovation by implementing training programs and policies to foster CFL. They can also acknowledge the interdependency between organizational factors and individual characteristics to enhance SMEs’ technological innovation.

Introduction

Research focusing on small and medium-sized enterprises (SMEs) is an increasing trend worldwide because of their unquestionable importance to emerging and developed economies in terms of employment and gross product. This behavior remains in different cross-cultural contexts, in occidental (European Union, Ireland, Italy, and Sweden) (Barrett et al., 2021; Cosenz and Bivona, 2021; Hilmersson and Hilmersson, 2021; Kraus et al., 2020) and oriental economies (China, Indonesia, Japan, and Korea) (Arsawan et al., 2022; Colovic, 2021; Lee, 2021). In Spain, SMEs also have an essential role in the economy, considering that the official figures show they represent 99% of global firms, 62% of gross value added, and 66% of employment. SMEs are also very relevant in job creation, with more than eight million people (DIRCE, 2020).

Given the relevant role of SMEs in the economy, the competitiveness of such kinds of firms needs to be maintained or even improved. In that sense, research on the factors that promote competitiveness in SMEs is essential. In particular, we consider technological innovations as they have been linked to firm competitiveness, sustainable and firm growth, or business success, among other critical aspects (Chatzoglou and Chatzoudes, 2018; Martínez-Alonso et al., 2022; Razavi et al., 2016; Santana et al., 2015). Over the last six decades, considerable technological innovation research has existed. In recent years, studies in the field of technological innovation mainly focused on research and development (R&D), technological innovation management, models of technological innovation, and impacts of technological innovation (Akbari et al., 2020). Previous literature recognized the need for research on the technological innovation behavior of SMEs as the innovative capacity, speed, and frequency of innovations are definitively essential for SMEs in a global context subject to a complex and changing environment (Hilmersson and Hilmersson, 2021). Existing research highlighted the importance of technological innovation to SMEs’ competitiveness and permanence in the market (Sharif and Huang, 2012). Most importantly, for SMEs that are often resource-restrained (Musteen et al., 2010), technological innovation is considered a remarkable driver of environmental sustainability practices (Ramírez-Orellana et al., 2022), expansion into other markets and, finally, SMEs’ performance (García-Lopera et al., 2022) and growth (Donbesuur et al., 2020). However, most previous literature paid scarce attention to technological innovation in SMEs (Akbari et al., 2020; Arsawan et al., 2022).

Additionally, knowing the relevance of considering technological innovations when researching on competitive factors of SMEs, we also consider other variables related to technological innovations. An example is the individual characteristics of CEOs, who are the leaders and responsible for the decision-making process in SMEs. In that sense, an emerging stream of research focused on the influence of the features of individuals managing the firm on the technological innovation outcomes of SMEs instead of considering firms’ characteristics only (Geyer, 2016). There has been a recent spike in interest regarding financial literacy (FL) in SMEs (Molina-García et al., 2020, 2022). This case is motivated by the observation that FL diminishes financial restrictions (Nkundabanyanga et al., 2014), increases firm performance (Eniola and Entebang, 2017), and could bring benefits in terms of the recognition of opportunities (Anwar et al., 2020), the attitude and management of corporate risk (Kulathunga et al., 2020; Ye and Kulathunga, 2019), or entrepreneurship (Riepe et al., 2022). Thus, in the framework of the innovation capacity of SMEs, the present study focuses on CEOs’ FL (CFL), which is becoming a relevant factor to be considered. CFL comprises the financial knowledge related to resources, markets, risks, management, legal, or tax issues related to financial matters (Koropp et al., 2013). CFL also includes making suitable financial decisions (Tian et al., 2020). In this sense, literature exploring the impact of managers’ FL on innovation capacity in SMEs is limited (Hutahayan, 2021; Liu et al., 2021; Tian et al., 2020). The scarce previous literature showed that CFL has a double effect (direct and indirect, by alleviating financial constraints) on the technological innovation of SMEs (García-Pérez-de-Lema et al., 2021). Specifically, based on upper echelon theory (UET), this study emphasizes how the specific characteristics of CEOs, namely, their FL, positively impact firm innovation, bettering expenses and income management, and/or long-term planning. Furthermore, the study shows that financially literate CEOs can alleviate financial constraints by gaining more access to existing financing, which may notably enhance the firm’s likelihood of undertaking technological innovation.

Furthermore, we consider the role of some mediating variables that connect a CFL with technological innovations, which is the case for management control systems (MCS) and CEOs’ risk-taking. Accordingly, the implementation of MCS plays an essential role in stimulating technological innovation (Henri, 2006), and CFL may be vital to develop and implement MCS in SMEs (Rostamkalaei et al., 2022). Similarly, CFL is one of the attributes for unraveling CEOs’ risk-taking (e.g., Buratti and Allwood, 2018). CEO risk-taking propensity has been demonstrated to be a driver of technological innovation (García-Granero et al., 2015; Kraiczy et al., 2014). However, no research has measured the mediating effect of MCS and the risk-taking attitude of CEO on the relationship between FL and SME technological innovation (Molina-García et al., 2022). In this vein, the current study responds to the following research questions: Does MCS have a mediating effect on the relationship between CFL and the technological innovation of SMEs? Does MCS have a mediating effect on the relationship between CFL and CEOs’ risk-taking? Does CEOs’ risk-taking have a mediating effect on the relationship between CFL and technological innovation of SMEs? Is the relationship between CFL and technological innovation sequentially mediated by MCS and CEOs’ risk-taking?

We draw in UET and resource-based view (RBV) to answer the former questions. According to UET, agents’ features and behaviors are associated to organizational outcomes (Hambrick and Mason 1984). In this vein, CFL and CEOs’ risk-taking, as two of SME CEOs’ upper echelon characteristics, can impact the implementation of MCS and technological innovation performance, respectively. Based on the RBV, the implementation of MCS can be considered a source of competitive advantage (Barney, 1991; Songling et al., 2018). Then, the implementation of MCS can be a source or capability that might be a precursor of CEOs’ risk-taking and, ultimately, SME technological innovation.

To fill this research gap, we developed an empirical study addressed to 310 managers of SMEs. Research analysis follows the structural equation modeling (SEM) method based on partial least squares (PLS) path modeling. Our findings indicate that MCS and CEOs’ risk attitude have a full mediating effect on the relationship between CFL and technological innovation of SMEs, conceding an essential role to both variables. The results also show that CFL affects risk-taking behavior through MCS. Therefore, the use of MCS becomes fundamental because it not only mediates the effect of the CFL on SMEs’ technological innovation but also allows the mediating influence of CEOs’ risk-taking on the former relationship.

This study has several contributions to the literature and practice. First, this study is particularly intriguing because the previous research is almost silent regarding the indirect effects on the relationship between CFL and SME technological innovation. This research thereby extends and challenges the current literature by analyzing two particular factors that mediate the CFL–technological innovation relationship, namely MCS and CEOs’ risk-taking. Second, by relying on UET (Hambrick and Mason, 1984), this study is one of the first to use CEOs’ novel individual-level characteristic, CEOs’ FL, as a determinant of technological innovation in SMEs, contributing to developing this theoretical view further. Third, the findings of this study also yield relevant practical implications for innovation policies that can foster the competitiveness and performance of SMEs. SME CEOs can generate more technological innovation by increasing their FL, which enables organizations to build and enhance MCS and affect the risk-taking behavior of CEOs. Thus, governments can develop policies that encourage SMEs to augment CFL. Eventually, this case will lead to a higher implementation of MCS and better CEO risk-taking behavior, which in turn will increase the technological innovation of SMEs. Therefore, improving technological innovations in SMEs would lead to gaining competitiveness and business growth, maintaining the essential role of SMEs to sustain the welfare state of economies worldwide.

Empirical and theoretical background

Our study attempts to create novelty by analyzing two factors that mediate CFL–technological innovation, namely MCS and CEOs’ risk-taking. The current section is devoted to examining briefly the scarce studies focusing on the effects of CFL on technological innovation to back our research with previous empirical research results. This section also explains the theory that supports the mediating role of MCS and CEOs’ risk-taking.

FL and technological innovation

Research focusing on the FL-technological innovation relationship is at a very early stage. Thus, Győri et al. (2019) studied specific aspects of the innovation activities of SMEs by considering entrepreneurial financial culture and social–economic environment. They found that, among other considerations, employing a full-time financial expert positively contributes to carrying out implemented and planned innovation in SMEs. Hutahayan (2021) analyzed how SMEs adjust and further develop business competencies, innovations, and performance using market orientation, learning orientation behaviors, and FL. Their findings confirmed the positive influence of CFL on firm innovation. Hasan et al. (2021) demonstrated that CFL promotes the usage of most innovative financial instruments as financing alternatives (FinTech) based on the use of technologies.

Studies focused on the relationship between CFL and technological innovation (García-Pérez-de-Lema et al., 2021; Liu et al., 2021; Tian et al., 2020). Liu et al. (2021) analyzed the impact of CFL on SMEs’ innovation using a large survey database of Chinese SMEs in 2015 and 2017. They found that an entrepreneur’s FL is positively associated with firm innovation engagement and that risk tolerance is a transmission mechanism for the influence of FL on innovation (Liu et al., 2021). From another aspect, Tian et al. (2020) examined the relationship between CFL and firm innovation based on a dataset obtained from surveys of Chinese SMEs in 2019. They confirmed that CFL significantly enhances R&D investments by alleviating financial constraints and improving risk management. Finally, García-Pérez-de-Lema et al. (2021) analyzed how CFL influences a firm’s technological innovation and the mediating role of alleviating financial constraints of SMEs using a sample of Spanish SMEs in 2016–2017. Their results showed that by alleviating financial constraints, CFL exerts a direct and an indirect impact on a firm’s technological innovation.

UET and RBV

This study adopts UET and RBV to explain the mediating role of MCS and CEOs’ risk-taking on the CFL–technological innovation relationship. Upper echelon theorists highlight the relevance of examining the executives’ roles and attitudes because organizational outcomes are considered consequences of the features of powerful actors in the organization (Hambrick and Mason, 1984). Thus, according to UET, CEOs’ idiosyncrasies can shape firm strategic decisions. The impact of CEOs’ disposition, expertise, and resulting behavior on the organization outcomes is even more intense in SMEs, where decision-making is concentrated in the CEO’s hands, and the number of TMT members is often reduced (Kraiczy et al., 2014). Accordingly, CEOs’ decisions and actions are led by their own analysis of the strategic choices they face, which in turn is conditioned by their experiences, values, and personalities (Calabrò et al. 2019). Hence, CEOs’ risk-taking and CFL may be identified as drivers of CEOs’ actions and organizational outcomes. The former refers to the willingness of a CEO to devote important resources to exploit opportunities or to get involved in behaviors with undetermined results (Gilley et al., 2002). The latter comprises both financial knowledge and making financial decisions (Koropp et al., 2013; Tian et al., 2020). In this vein, CFL and CEOs’ risk-taking, as two of the SME CEOs’ upper echelon attributes, can eventually affect SMEs’ organizational choices (e.g., the design of MCS) and results (e.g., technological innovation performance), respectively.

From another aspect, research based on RBV emphasizes that SMEs have distinctive capabilities and resources that may be sources of competitive advantage (Barney, 1991) and, in particular, may contribute to innovation success (Sirmon and Hitt, 2003). In this sense, the implementation of MCS usually involves applying formalized routines and procedures using information to maintain or change patterns in organizational activity and adopt suitable decision-making (Simon 1987). MCS can be considered a particular organizational resource and capability that can help improve risk management, changing CEOs’ risk-taking. MCS can also be utilized to generate technological innovation (Simons, 1995).

Hypothesis development

The mediating effect of MCS on the CFL–technological innovation relationship

Knowledge is crucial for firm performance (Serra and Kunc, 2015) in large organizations and SMEs (Ganesh and Mehta, 2010). Similarly, the use of MCS plays a vital role in the firm performance, making possible managers’ appropriate decisions through proper planning, budgeting, analyzing, measuring, and evaluating organizational information (Cosenz and Noto, 2015). However, the use of MCS is particularly challenging in SMEs, and much remains to be learned about it in these organizations (Goffee, 1996; Chepngetich, 2016). Additionally, proper knowledge can contribute to the better utilization of MCS, improving the performance of SMEs (Kulathunga et al., 2020).

Mainly, FL is a multifaceted construct that usually comprises knowledge about funding, financial markets, management systems, financial risks management, legal or tax issues related to financial matters, and so on (Koropp et al., 2013). Drawing on UET arguments, we propose that the CEO’s behavior regarding the implementation of MCS is a function of his/her possession of substantial FL, as one of his/her key possible characteristics. Thus, financially literate CEOs having an appropriate level of understanding of MCS can improve the use of MCS in SMEs. CFL may be vital to develop and implement MCS in SMEs, such as budgeting (Rostamkalaei et al., 2022), contributing to better decision-making and operations (Li et al., 2021), and finally, increasing SME performance (Iramani et al., 2018). In this vein, recent research supported the significant impact of higher financially literate CEOs adopting MCS, particularly enterprise risk management practices and systems (Kulathunga et al., 2020; Mabula and Dong, 2018). However, SMEs tend to reject investing in and devoting time to MCS, such as accounting systems (Kou et al., 2021) as they are not sure of their benefits for the organization (McMahon, 2001). In short, CFL is an efficient means to use MCS in SMEs.

From another aspect, according to Henri and Wouters (2020), research studies paying attention to the relationship between MCS and technological innovation are still insufficient. Nevertheless, previous literature identified that one of the core challenges when implementing an influential MCS is to deal with the trade-off between innovation and planned goal achievement by finding the balance between control and flexibility (Simons, 1995). In that sense, Ahrens and Chapman (2004) proposed enabling control when MCS (planning and administrative controls) allow the company to achieve a balance between efficiency and flexibility. Then, this balance of MCS fosters technological innovation. Furthermore, according to Henri (2006), the correct implementation of MCS becomes a condition to stimulate technological innovation as it is a valuable distinctive and imperfectly imitable capability, a source of competitive advantage following the RBV.

The present study clearly uses RBV to illuminate the MCS–technological innovation relationship in SMEs. We argue that the implementation of MCS is a starting point for catalyzing superior technological innovation and provides SMEs with the ability to manage their resources and risks more effectively than their competitors, leading to sustained competitive advantages (Barney, 1991; Colbert, 2004). In this vein, recent studies revealed the essential role of MCS in innovation processes by exploring how CEOs use MCS in different innovation contexts (Bedford, 2015; Chenhall and Moers, 2016). Thus, proper management controls increase the potential benefits from technological innovation (Bedford, 2015; Bisbe and Otley, 2004; Feranita et al., 2021). In particular, Bisbe and Otley (2004) found that an interactive use of MCS improves the performance effect of technological innovations in mature and manufacturing SMEs, mainly in an exploratory innovation context that requires a more straightforward diagnostic system and fewer resources than exploitative innovations. In particular, for SMEs, the managing director or CEO plays a critical position in the relationship between MCS and technological innovations. In that sense, the interactive use of MCS, particularly concerning customer relationships, promotes technological innovations in the case of SMEs (Pešalj et al., 2018). This previous study also highlighted the critical role of the CEO in balancing the diagnostic and interactive MCS continuously. Similarly, MCS based on cost information and non-financial performance indicators strongly influences technological (product) innovations in SMEs (Henri and Wouters, 2020). Furthermore, these authors showed that this influence is particularly relevant in contexts of environmental unpredictability, helping to promote technological innovations for SMEs.

In short, financially literate CEOs will likely lead to an increase in technological innovation. Their FL allows them to implement MCS, which may notably enhance the organization’s capacity to obtain technological innovation. Therefore, this study proposes that more excellent CFL will lead to higher technological innovation outcomes through a higher implementation of MCS. Based on the above arguments, we propose the following hypothesis:

H1: MCS positively mediates the relationship between CFL and technological innovation.

The mediating effect of MCS on the CFL–CEO risk-taking relationship

In the preceding sub-section, this study showed that prior research confirmed the positive effect of CFL on implementing MCS in SMEs. Based on the arguments of UET, prior research suggested that CFL can be reckoned as a factor that characterizes CEOs and has beneficial implications on the organization (Hambrick and Mason 1984), namely, through the higher use of MCS.

From another aspect, based on RBV, we argue that a positive association exists between the application of MCS—an organizational resource or capability—and CEOs’ risk-taking—an executive characteristic. CEOs’ risk-taking (propensity), defined as CEOs’ present predisposition to take or avoid risks, is a function of CEOs’ psychological, social, and cognitive abilities (Hambrick and Mason, 1984). CEOs’ risk-taking is an attitude that is simultaneously persistent and changeable over time because the CEO faces a continuous decision-making process (Sitkin and Weingart, 1995). Organizational control systems, such as MCS, can affect CEOs’ risk-taking because they focus on different aspects of the decision-making process (Giaccone and Magnusson, 2022; Kou et al., 2019; Sitkin and Pablo 1992). MCS makes more likely proper CEOs’ diagnoses and choices, affecting CEOs’ risk propensity. In this vein, the literature studied the relationship between the importance of the design and adequate implementation of MCS to drive businesses effectively and CEOs’ risk-taking propensity (Liem and Hien, 2020). The authors explored the connection between the use of MCS and CEO managerial risk propensity during the decision-making process. In particular, the authors considered that within MCS, management accounting systems have a relevant function in measuring top managers’ risks, which can be divided into risk aggregation, risk reporting, and risk monitoring. In that sense, MCS lessens uncertainty in the decision-making process and support risk management (Collier et al., 2004). A direct link exists between MCS and CEOs’ risk-taking as one of the purposes of MCS is to control the risks in the organization’s activities. In that sense, Sandino (2007) suggested a classification in terms of the objectives of MCS (basic, costs, revenue, and risk) in which “Risk MCS” are established to reduce risks and protect asset integrity. As part of MCS, internal audit quality is also related to CEOs’ personality in risk management. In this sense, implementing MCS, both the interactive and diagnostic use of budgets in conjunction with performance indicators, has favorable effects on risk awareness within the organization (Braumann et al., 2020). Organizational control systems can also affect the risk behavior of CEO and top managers through the mediating effect of risk perceptions of top managers (Sitkin and Pablo, 1992).

In sum, financially literate CEOs will likely augment CEOs’ risk-taking, given that their FL allows them to easily use MCS—this was explained in the previous sub-section, which may significantly improve CEOs’ risk-taking by diminishing uncertainty in decision-making and backing risk management. Therefore, this study proposes that more excellent CFL will lead to higher CEOs’ risk-taking through increased use of MCS. For the previous reasoning, the following hypothesis is considered:

H2: MCS positively mediates the relationship between CFL and CEOs’ risk-taking.

The mediating effect of CEOs’ risk-taking on the CFL–technological innovation relationship

On the one hand, CFL has been proven to have a positive effect on risk diversification and risk management (Bannier and Neubert, 2016). Financially literate CEOs are more likely to understand risks, have better risk perception and assessment (Buratti and Allwood, 2018), and be more prone to take risks (Hsiao and Tsai, 2018). Therefore, CFL is expected to increase CEOs’ motivation to take risks and affects CEOs’ risk-taking (Liu et al., 2021).

On the other hand, technological innovation usually requires considerable time, effort, and resources, is often linked to sharp learning curves, and is essentially risky as the chances of success are uncertain and small. However, if successful, then the high potential profitability associated with technological innovation makes many managers prone to pursue it (Wu et al., 2005; Zhou, 2006). Firms’ managers, mainly CEOs in SMEs, have to deal with the uncertainty inherent in technological innovation. In this vein, the attitude of CEOs when facing risks influences technological innovation results. Prior entrepreneurship and leadership research has been considerably interested in analyzing the risk-taking–innovation association (Latham and Braun, 2009). Similarly, studies on creativity have addressed this relationship (Gilson and Shalley, 2004). Managers’ risk-taking is often recognized as an important determinant of innovation outcomes (Craig et al., 2014; Gilley et al., 2002), as risk management implies to handle with innovation activities that usually involve potentially high-rewarding investments but are also very risky. Based on UET, some studies analyzed how managers’ and TMT’s risk-taking behaviors, depending on unique features, such as tenure or age, affect innovation performance (e.g., Liu et al., 2012). CEOs who are more prone to take risks value the potential benefits of technological innovation generation more favorably than more risk-averse CEOs (Pérez-Luño et al., 2011). Thus, by and large, scholars showed that CEOs oriented toward risk-taking generate more and better technological innovation outcomes (Gilley et al., 2002; Ling et al., 2008). Similarly, the literature focusing on creativity highlighted risk-taking as a relevant aspect to stimulate creativity and experimentation (Martins and Terblanche, 2003) and, in turn, improve innovation performance (Yuan and Woodman, 2010). More recently, Giaccone and Magnusson (2022) re-emphasized the significant positive effect of risk-taking on innovation performance, underscoring results from previous research (Cabrales et al., 2008; Craig et al., 2014; Gilley et al., 2002; Guimaraes and Paranjape, 2017; O’Connor et al., 2008).

In short, the above arguments suggest that CFL is a crucial antecedent for explaining CEOs’ risk-taking (Buratti and Allwood, 2018). Similarly, the positive influence of CEO risk-taking propensity on technological innovation is well established in previous literature (García-Granero et al., 2015; Kraiczy et al., 2014). Following the last literature review reasoning, we consider the next hypothesis:

H3: CEOs’ risk-taking positively mediates the relationship between CFL and technological innovation.

Finally, in the organizational innovation context, the logic of interdependency implies that CEOs’ characteristics, such as the CFL and risk-taking, are likely to contribute to changes in the organizational resources and capabilities, namely, MCS, and vice versa. Previous research considered separately how organizational factors (Crossan et al., 1999) and CEO characteristics (Mumford and Licuanan, 2004) contribute to innovation performance. Therefore, new empirical research would be desirable regarding how the interplay and reciprocal influence between organizational elements and CEOs’ features affect innovation performance (Crossan and Apaydin, 2010). Based on our evaluation of the contents of the preceding subsections, we conclude that the interdependency between organizational factors and CEOs’ features might offer even more significant potential to enhance our understanding of SMEs’ technological innovation.

Therefore, based on the former sub-sections, CFL has a positive impact on MCS, which has a positive effect on CEOs’ risk-taking. Similarly, CEOs’ risk-taking positively impacts technological innovation outcomes. Therefore, we hypothesize the following:

H4: The relationship between CFL and technological innovation is sequentially mediated by MCS and CEOs’ risk-taking.

Methods

This section is devoted to (1) explaining how the information was gathered, providing information on the sample used in the estimation and the population it belongs to; (2) defining the variables included in the research model and the measures considered based on previous literature; and (3) introducing the statistical technique used in the estimation.

Sample and data

The selection of the sample was made considering the regional structure of Spanish SMEs according to official business information (DIRCE) provided by the National Statistical Institute. According to Spanish National Statistics Institute, SMEs are very relevant in Spain because the official figures indicated that they account for 99% of the total population of firms, 62% of gross value added, and 66% of employment (DIRCE, 2020). The population of SMEs in Spain with at least one employee accounts for 1,307,634 (Ministerio de Industria, Comercio y Turismo, 2022).

Therefore, the selection process was based on the stratified sampling principles in finite populations, considering the segmentation according to industry and size. When a firm refused to collaborate, a replacement was made on similar randomly selected companies. We design a qualitative research questionnaire as an instrument to collect the information. The questionnaire was developed to include constructs previously tested and based on previous literature. All items of the questionnaire have been built through closed-ended questions following a Likert’s scale ranging from 1 to 5 points to measure the degree of agreement from participants.

The fieldwork to collect the data took place in 2017. Finally, the valid answers corresponded to 310 managers of Spanish SMEs. We control for possible bias in the responses from participants. Non-response bias was controlled by comparing the first round of questionnaires collected (15% of the sample) with those responding to the follow-up (15% of the sample). In that sense, t-Student showed that responses were not significantly different between the two groups for any variable (Nwachukwv et al., 1997). No nonresponse biases were found. Additionally, common method variance bias was tested following Podsakoff and Organ’s (1986) proposal. As a result, the common method variance was not significant.

Responses show that 31.6% of the sample comprises SMEs in the manufacturing sector, 19.35% in construction, 20% in retail, and 29% in services. Of the sample, 41.9% are small-sized firms with fewer than 25 employees, and 22.3% are medium-sized firms. We consider SMEs following the EU recommendation issued in 2003 (EU Commission Recommendation concerning the definition of micro-, small-, and medium-sized enterprises 2003/361/EC) (Table 1).

Table 1 Sample characteristics

Definition of variables

The scales used in this research have been previously validated by the literature (Table 2) and are based on a five-point Likert’s scale ranging from one (strongly disagree) to five (strongly agree).

Table 2 Indicators and constructs

FL can be considered as “A combination of awareness, knowledge, skill, attitude and behaviour necessary to make sound financial decisions and ultimately achieve individual financial wellbeing” (OECD 2018, p.4). Additionally, according to Bay et al. (2014, p. 42), FL “is seen either as (1) an individual capability that can be acted upon in relation to experience, vocabulary and skills (the autonomous model), or (2) a socially situated issue where financial literacy in itself must always be debated (the situated model).” The CFL scale was adopted from García-Pérez-de-Lema et al. (2021). This construct is calculated in the estimation as a reflective composite, in line with the proposal made by García-Pérez-de-Lema et al. (2021), which considers different dimensions of the same construct that can be related (level of information of the manager about the evolution of the economy, the alternative financing funds, the financial assets to invest in, and the financial information about the company).

To contextualize the concept of MCS, we have chosen a definition recognized by previous literature: “management control systems are the formalized procedures and systems that use information to maintain or alter patterns in organizational activity… these systems broadly include formalized procedures for such things as planning, budgeting, environmental scanning, competitor analyses, performance reporting and evaluation, resource allocation and employee rewards” (Simons, 1990, p.128). In the research, the MCS measure is based on the proposal of Duréndez et al. (2016). This construct is also considered a reflective composite, as it includes the use of ERP systems, cost accounting, budget control, financial analysis, strategic planning, internal audits, and quality controls.

We followed Covin and Slevin (1989) and Yang (2012) to account for risk-taking propensity, considering a factor construct built of four items. The four items comprise the following information: (1) I bet on R&D leadership; (2) I have a strong propensity for high-risk projects; (3) I think that knowing the environment, courageous and wide-ranging actions are necessary to achieve the objectives of the company; (4) When faced with uncertain decision-making, I usually take a courageous and aggressive stance to maximize the likelihood of exploiting potential opportunities.

Finally, to consider innovation in SMEs, we opted for a subjective measure as it is more appropriate for innovation in SMEs (Hughes, 2001). Consequently, in this investigation, a subjective output innovation perspective is considered using the scale used by Madrid-Guijarro et al. (2021). Respondents were asked to indicate the firm’s situation concerning its competitors with regard to the number of new products launched and processes implemented, the firm’s pioneering nature concerning introducing new products and processes, and the rapid response to their competitors’ product/service and process innovations.

In the model, control variables, such as firm age, size (number of employees), and activity sector (three dummy variables: construction, trade, and service industry, manufacturing the hidden category), have been introduced following previous literature (Choi et al., 2011; Duran et al., 2016; Kammerlander et al., 2015; Werner et al., 2018).

Statistical technique

We estimate the proposed model using PLS path modeling, a variance-based SEM method (Roldán and Sánchez-Franco, 2012), SmartPLS 3.3.3. PLS simultaneously assesses the reliability and validity of the variables (outer model) and the estimation of the paths among these constructs (inner model) (Barroso et al., 2010). The PLS technique is appropriate in this research because: (1) No specific distribution is required in the indicators in PLS (Chin, 2010); (2) The proposed research model has great complexity as we pay attention to the existence of mediating effects (Hair et al., 2017); (3) The model makes use of reflective composite constructs (Sarstedt et al. 2016; Chin, 1998). As in our model, one construct (risk-taking propensity) is a factor. In this case, we use the consistent PLS estimation algorithm (Dijkstra and Henseler, 2015).

After verifying the psychometric properties of the outer model (measurement model), we continued estimating the structural model proposed in Fig. 1. Control variables, such as firm age, industry, and size, have also been introduced into the analysis.

Fig. 1

The conceptual model

Data analysis

Model assessment

We consider Dijkstra–Henseler’s indicator (Rho_A), composite reliability (CR), factor loadings, and average variance extracted (AVE) to assess the reliability, convergent, and discriminant validities of reflective composite and factor constructs (Table 3). These indicators are chosen because reflective composite and factors are characterized by important correlations among their indicators (see Table 8 in Appendix) (Dijkstra and Henseler, 2015). The measurement model assessment shows acceptable outcomes. In this sense, Cronbach’s alpha levels vary between 0.768 and 0.867, and Dijkstra–Henseler’s indicator (Rho_A) ranges between 0.777 and 0.896, exceeding both indicators’ acceptable threshold of 0.7. Similarly, CR also reaches the appropriate threshold varying between 0.849 and 0.910.

Table 3 Measurement model

The AVE shows evidence of the convergent validity of the measurement model. In this sense, this indicator ranges between 0.585 and 0.698, exceeding 0.5 (Hair et al., 2009). Furthermore, loads are higher than 0.7 except for one item that belongs to the MCS construct, which has a load equal to 0.648 (higher than 0.5) (Arzubiaga et al., 2019; Hair et al., 2017).

The heterotrait-monotrait (HTMT) ratio of Henseler et al. (2016) is used to assess the discriminant validity among constructs. Table 4 reports that the HTMT between each pair of variables varies from 0.357 to 0.690, with bootstrapping analysis demonstrating that the HTMTs are significantly lower than 1, thereby verifying the discriminant validity of the variables (Henseler et al., 2016). Therefore, we find evidence that all the constructs in the proposed model are distinctive ones.

Table 4 Discriminant validity. Fornel & Larcker and HTMT ratio

Structural model estimation and results

As the estimations of the path coefficients are based on OLS, the antecedent variables of each of the endogenous constructs must avoid multicollinearity. Table 5 shows that all inner VIF values are under 5. Therefore, our model has no multicollinearity problems (Hair et al., 2017).

Table 5 Structural model estimation (PLSc)

Chin (2010) proposed that the structural model can be evaluated considering the algebraic sign, the significance and magnitude of the path coefficients, and the adjusted coefficient of determination values. The significant paths (t-values higher than 1.64) oscillate between 0.360 and 0.675. However, unexpected findings are revealed in two paths that are not significant in the model. These paths are the relationship between CFL and innovation and the one between CFL and risk-taking. The adjusted coefficient of determination values is high, ranging between 0.32 and 0.5 (Hair et al., 2011; Henseler et al., 2009). Stone-Geisser Q² (Chin, 2010; Sarstedt et al., 2014) is also an indicator used to evaluate the model’s predictive power. Q² values larger than zero for redundancy indicate that a model has predictive power for a certain endogenous construct. In this model, Q² for endogenous variables ranges from 0.102 and 0.225.

The results show nonsignificant values for the relationship between CFL and innovation (path: − 0.048, t-value: 0.543) and CFL and risk-taking (path: − 0.009, t-value: 0.083).

The findings show evidence favoring the relationship between CFL and MCS as the path is positive and significant (path: 0.674, t-value: 13.881). Therefore, the more information the manager has about finance in different aspects, the higher the likelihood that the company implements tools related to MCS. The results also verify that MCS is a key variable to promote firm innovation (path: 0.320; t-value: 3.249) and a positive attitude toward risk-taking (path: 0.490; t-value: 5.370). Consequently, these positive relationships advocate for the mediating effect exerted by MCS. The analysis of the indirect effect reports positive and significant effects. Thus, MCS fully mediates the relationship between CFL and innovation (indirect effect: 0.216; t-value: 3.197) and CFL and risk-taking (indirect effect: 0.330; t-value: 4.780), verifying H1 and H2. The full mediation is confirmed as the total effects that CFL exerts on innovation are significant (total effect: 0.283, t-value: 3.966), whereas the direct effect is not. The same reasoning is valid for the full mediation exerted by MCS in the relationship between CFL and risk-taking. The results relating to H1 and H2 align with recent literature defending that FL is an essential driver of budgeting and financial planning (Rostamkalaei et al., 2022). The results regarding H1 also agree with a bit earlier study proving that MCS practices positively affect innovation (Henri and Wouters, 2020; Feranita et al., 2021). Then, the results for H2 concur with the findings obtained by Braumann et al. (2020), showing that the use of budgets and performance measures increases risk awareness.

Finally, risk-taking attitude positively affects firm innovation (path: 0.359; t-value: 5.278). This result involves a double mediating effect on the relationship between CFL and innovation (total effect: 0.283, t-value: 3.966). The effect of CFL on innovation is transmitted using the MCS and the risk attitude of the firm, verifying H4. This finding is in line with recent studies showing independently that CFL is an important driver of MCS implementation (Rostamkalaei et al., 2022), MCS influences risk-taking (Braumann et al., 2020), and risk-taking has a significant effect on innovation performance (Giaccone and Magnusson 2022). However, our result refined the former works by interacting with the former research models and showing that MCS and CEOs’ risk-taking play a mediating role in the understanding of the relationship between CFL and innovation performance. The key variable in this mediating relationship is MCS as the direct relationship between CFL and risk-taking attitude is not significant when MCS is included in the model. Thus, H3 is not confirmed. The result related to H3 does not support previous research, suggesting that risk-taking is a transmission mechanism for the effect of FL on innovation (Liu et al., 2021). A possible explanation of this result is that risk-taking per se does not always produce an increase in innovation. Nevertheless, when CEOs’ risk-taking is promoted by an organizational factor, namely, the implementation of MCS, it leads to a higher innovation performance (Giaccone and Magnusson, 2022), thereby confirming H4. Table 5 shows only the significant paths related to the control variables. In this sense, the results show that firm size in terms of the number of employees negatively affects risk-taking, whereas construction firms develop less technological innovation compared with manufacturing companies. Figure 2 shows the main results.

Fig. 2

Structural model results

Control for endogeneity and nonlinearity

Assessment of endogeneity

Our assessment of potential endogeneity follows Hult et al.’s (2018) systematic procedure, starting with the application of Park and Gupta’s (2012) Gaussian copula approach, using the latent variable scores of the original model estimation as input. We first verified if the variables, which potentially exhibit endogeneity, were non-normally distributed. That is, we run the Kolmogorov–Smirnov test with Lilliefors correction (Sarstedt and Mooi, 2019) on the latent variable scores of FL and MCS from the PLS path model. The construct latent variables were built considering the weighted average of indicators using the weights that provide consistent PLS. The results show that none of the constructs had normally distributed scores, allowing us to proceed with Park and Gupta’s (2012) Gaussian copula approach. The results in Table 6 show that none of the Gaussian copulas was significant (p value > 0.05). We consequently concluded that endogeneity was not present in this study, which supported the robustness of the structural model results (Hult et al., 2018; Sarstedt et al., 2020).

Table 6 Assessment of endogeneity test using the Gaussian copula approach

Assessment of nonlinear effects

We used Ramsey’s (1969) RESET on the latent variable scores extracted after the convergence of the original model’s PLS-SEM algorithm to test for potential nonlinearities in the structural model relationships. We found that neither the partial regression of MCS on FL (F (3.305) = 1.01; p = 0.389), the one related to RT on MCS and FL (F (3.304) = 1.32; p = 0.267), nor the partial regression of I on RT, MCS, and FL (F (3.303) = 0.92; p = 0.430) is subject to nonlinearities. Next, we included interaction terms to represent the quadratic effects among (1) FL on MCS; (2) FL and MCS on RT; and (3) FL, MCS, and RT on I. The results indicate that neither of the nonlinear effects is significant (Table 7). We, therefore, conclude that the linear effects model is robust.

Table 7 Assessment of nonlinear effects

Discussion and conclusion

The findings of this study add new evidence to very recent previous literature connecting CFL and technological innovation in SMEs. The importance of organizational management systems when analyzing technological innovation outcomes of SMEs is reflected in the results of this study, which show a full mediation effect exerted by implementing MCS within the organization. This full mediation effect of MCS, either through its use alone or through its impact on CEOs’ risk-taking, on the relationship between CFL and the technological innovation of SMEs plays an essential role in the utilization of the MCS.

The results show a first mediation effect of MCS on the relationship between CFL and the technological innovation of SMEs. This result suggests that CEOs of SMEs with the proper financial education and competencies for the funding and investment decision-making process are better prepared for implementing MCS based on budgeting (Rostamkalaei et al., 2022) and another control mechanism within the organization (e.g., Kulathunga et al., 2020). Then, CEOs trained in finance skills (financial planning, cost accounting, financial analysis, cash-flow management, risk analysis, investment options, etc.) would foster the proper usage of MCS, which becomes a condition to stimulate innovation (Henri, 2006). MCS also plays a second significant mediation effect between the CFL and CEOs’ risk-taking behavior of SMEs. This result supports the idea that MCS influences CEOs’ risk propensity, lessening uncertainty in decision-making and improving risk management (Collier et al., 2004). Thus, the mediating effect involves that CFL affects risk-taking behavior through MCS. Finally, the results also indicate that the implementation of MCS in the organization—in addition to its mediating effect in the CFL–technological innovation relationship—exerts another impact on technological innovation. This latest influence may be an indirect mediating result of technological innovation through its effect on CEOs’ risk-taking. Hence, the results align with previous studies identifying the critical role of running MCS in supporting technological innovations (Chenhall and Moers, 2016).

The study makes several contributions to the existing research on FL and innovation in SMEs by exploring the relationships among CFL, implementation of MCS, CEOs’ risk-taking, and technological innovation. The study sheds new light on the effects of CFL on technological innovation in SMEs, exploring the mediating effects of MCS and CEOs’ risk-taking. Our findings show that CFL has a positive impact on MCS, thereby having a positive effect on technological innovation and CEOs’ risk-taking. Thus, the favorable effect on technological innovation is enhanced. These results complement scholars’ previous findings, suggesting that executive FL promotes technological innovation inputs (R&D expenditures) by alleviating financial constraints and bettering risk management (Tian et al., 2020). To further refine the assessment of CFL influencing technological innovation, this study uses objective and subjective qualitative questions to assess technological innovation. This study goes beyond previous research that also analyzed the effect of FL on innovation but measured whether innovation output is involved rather than the level of innovation outcomes achieved (Liu et al., 2021). This study also goes beyond input–output models and extends the results of prior research (García-Pérez-de-Lema et al., 2021) by showing that the use of both MCS and CEOs’ risk-taking plays a mediating role in the understanding of the relationships between CFL and technological innovation in SMEs. Nevertheless, the essential key is in the use of MCS. If the CFL does not materialize in MCS utilization, the CFL will not affect CEOs’ risk-taking and ultimately technological innovation of SMEs. Our study expands our knowledge about the relationship between CFL and technological innovation performance. The study advances this research by testing the connection between identified innovation determinants. Specifically, the study explores the effect of CFL—a leadership feature, its consequences for MCS implementation—an organizational factor, the feedback loop from MCS back to another leadership characteristic—CEOs’ risk-taking, and its final influence on technological innovation performance—an organizational outcome. Thus, this study applied a systemic view of the variables driving innovation performance having a multi-level approach (in this case, individual and organizational levels) Crossan and Apaydin (2010), as demanded by very recent colleagues (Giaccone and Magnusson, 2022), noted that these factors have solid inter-relationships.

UET has previously been used to justify that firm innovation performance results from the idiosyncrasies of the top managers (Hambrick, 2007; Hambrick and Mason, 1984). Top managers’ decision-making regarding innovation often depends on their education, financial background, experience, and values (Smith et al., 1994; Talke et al., 2010). The concept of CFL has received minimal attention (García-Pérez-de-Lema et al., 2021; Li et al., 2021; Tian et al., 2020). Therefore, this study is one of the first to use CEOs’ novel individual-level characteristic, CFL, as a determinant of technological innovation in SMEs, contributing to the further development of UET. Furthermore, this study confirms that specific executive characteristics—namely, the CFL and risk-taking—ensemble well with MCS, following previous studies requiring identifying those certain managerial features that better suit MCS than others (Hiebl 2014). This study also highlights that upper echelon characteristics may interact with organizational resources and capability choices to explain technological innovation, which allows combining UET and RBV, adopting a complementary and enriched multi-theoretic approach (Crossan and Apaydin, 2010).

To sum up, our findings confirm that the empirical support of a direct link between CFL and firm innovation outcomes is inconsistent. From these results, we can conclude that it is impossible to assume a pure, simple relationship between CFL and technological innovation in SMEs without considering at least a crucial variable that affects this relationship, namely, the use of MCS.

There are also important implications for SMEs. Directors and owners should be aware of the need to develop formalized training programs and human resource policies to foster FL within executives and managers. Acquired finance skills of CEOs would help promote innovation in SMEs by improving the use of MCS and boosting risk-taking behavior. Thus, this study suggests that FL, through the mediating roles of MCS (direct and indirect) and risk-taking attitude, makes SMEs pioneers when introducing new products/services/processes and having a higher speed to develop innovations. Furthermore, managers should promote the implementation of MCS in SMEs whose use is still underdeveloped because the adequate use of formalized MCS also promotes the correct risk-taking behavior of CEOs, which helps to foster innovation. This study also has implications for industry associations. According to the results of this article, leaders of business associations should promote financial education and learning programs to foster CFL and employees and drive SMEs toward innovation.

Regarding implications for public authorities, the need to promote FL through the education of CEOs of SMEs should be highlighted. Thus, public policies to extend finance knowledge and skills to employees, or at least middle responsible executives, through university education and specific programs, become necessary to improve the decision-making process (corporate risk management) and promote innovation in SMEs. Particularly, colleges and universities should regularly organize FL competitions, courses, and similar events to promote financial education (Tian et al. 2020).

The study also has some limitations that provide avenues for future research. The results are focused on the case of SMEs, so they cannot be extrapolated to large or even quoted companies. In that respect, formal R&D processes that facilitate innovation and governance models of large companies should condition the conceptual model outlined. Therefore, extending the research to large companies would be interesting. Considering that most SMEs are family firms (Price et al., 2013), future studies should consider how family influence and family character of CEOs can affect the FL—innovation relationship. It can be expected that the family CEO maintains a lower FL level as, traditionally, previous literature confirmed that they have less training and professionalized profile (De Kok et al., 2006). These new research hypotheses can offer different findings from those obtained for SMEs. Additionally, this research is based on cross-sectional data with new evidence collected through primary sources, so its scope is limited. Further research can be planned with a longitudinal methodology to reach more robust results. Another relevant issue for future research is the institutional setting of analyzed companies. Cross-cultural differences can influence the study results and provide significant differences among SMEs. Extension of research in distinct institutional contexts would help have a complete and comprehensive overview of SMEs’ behavior regarding FL, innovation, and their determining factors. This extension of research could consider the sources of risk in different organizations of the financial system as the modern banking system and the global financial market have formed a complex network (Kou et al., 2019). This complexity can have a relevant impact on the relationship between FL and innovation, identifying a significant moderating effect. Finally, the lack of FL is one of the key challenges affecting the business growth of women entrepreneurs (Baporikar and Akino, 2020). Previous literature confirmed the moderating role of CEOs’ gender on the financial constraint alleviation–innovation relationship (Ruiz-Palomo et al., 2022). Another interesting research direction related to the effect of FL on technological innovation concerns the impact of gender on the former relationship. Consequently, we suggest examining if the influence of FL on technological innovation is different for SMEs managed by female CEOs than for SMEs managed by male CEOs. Finally, considering the findings obtained by Kou et al. (2021), which highlighted the significant role played by no financial information on bankruptcy prediction for SMEs, the FL variable is a relevant candidate to be introduced in the design of these models.

Appendix

See Table

Table 8 Descriptive and correlations of measures

8.