DETERMINANTS OF THE DEGREE OF INNOVATION IN THE INDUSTRIES LISTED ON THE B3

The study aims to identify the determinants of the degree of innovation of industries listed on the Brazilian Stock Exchange - B3. To this end, it used as sample 129 industries that trade their shares on the B3. From the principal component analysis, it was found that the innovation indicators are combined into two distinct factors. Therefore, it can be said that the degree of innovation of the firms takes into account the combination of indicators referring to the innovation strategies of exploitation and exploration, and, by encompassing two distinct factors, it is valid to mention that the degree of innovation has a two-dimensional character, suggesting the existence of a synergy between the innovation strategies.


Introduction
In emerging markets, innovation emerges as a source of competitive advantage, where innovative strategies implemented by managers meet market demands, providing sustainable growth for firms (Anning-Dorson, 2018).
In this context, a renewal in the organizations is necessary, leading managers to seek new sources of growth for the organization, through investments, especially those linked to innovation. However, such investments direct corporate strategies and the classical literature identifies two generic types: exploitation and exploration. While exploration focuses on generating new opportunities, knowledge, and competencies, exploitation takes advantage of opportunities, knowledge, and competencies already existing in the company (Levinthal & March, 1993;March 1991).
Since the seminal work of March (1991), it has been observed that these two strategies compete for the same set of scarce resources, which often results in the favoring of one strategy over the other. Therefore, the challenge faced by business managers and researchers is to figure out how to leverage a firm's capabilities in a way that allows it to succeed by using both types of innovation strategies simultaneously.
It should be noted that there are arguments in the literature in favor of the interaction of exploration and exploitation innovation strategies because, in this case, their combination can provide a superior competitive advantage compared to when each strategy is used in isolation, emphasizing the importance of better understanding this interaction and its effects (Andriopoulos & Lewis, 2009;Lennerts, Schulze, & Tomczak, 2020).
In this scenario, one realizes the need for the combination of factors related to the exploitation and exploration strategies, creating a degree of innovation. Given the above, the following question arises: which indicators are deterministic of the degree of innovation in industries listed on the Bolsa Brasil Balcão (B3)?
The objective of the present study was, therefore, to identify the determinants of the degree of innovation of industries listed on the B3. It is noteworthy that the sample of 129 industries listed on B3 was intentionally selected since the industrial sector has great relevance for the country's economy. Furthermore, the industry presents a relationship with technology that makes it a pioneer in the use of innovation strategies to obtain a competitive advantage. skills, they create resources that guide the company to growth and competitive differential, facts that ensure the sustainability and elevation of the firm's market value (Rezende, Correia & Gomes, 2017).
The literature points out different classifications/typologies for intangible assets, highlighting the one proposed by  for the fact that these authors elaborate a non-exhaustive classification of these assets, distributing them into human assets, innovation assets, structural assets, and relationship assets. To measure the innovation strategy exploitation, the present study adapt such typology, because one of the components of innovation assets in the study of , the investments in research and development (R&D), are characterized by the literature as being a radical innovation of long term, besides that it aims at new learning to generate new knowledge and have higher risk and uncertain return (Lampert & Kim, 2019), characteristics that are compatible with the innovation strategy exploration.

Innovation strategy exploration and R&D investments
Characterized as the act of experimenting with new technologies or approaches and continuous efforts to achieve breakthrough inventions, the innovation exploration strategy implies a business behavior defined by experimentation and risk-taking. The returns associated with this method are systematically less certain, more variable, and more distant in time, making the firm's earnings more volatile and less predictable by experimentation (March, 1991).
Exploration innovation involves moving away from existing knowledge and experimenting with new technologies and new ways of thinking and acting, to generate inventions, imposing a greater information gap between the company and its stakeholders to protect itself from the unwanted competition, because once such a strategy is successful, it can generate greater growth through these new categories of products and services, resulting in a greater competitive advantage when compared to incremental innovations, which are the result of the exploitation strategy (March, 1991).
However, although companies adopt the exploration innovation strategy to create and maintain competitive advantage, they also need to understand how to manage the level of uncertainty and risk involved in innovation activity (Barbosa et al., 2020), moreover, such typology is present in the innovation strategies of companies committed to investments focused on Research and Development (R&D) (D'Este, Marzucchi & Rentocchini, 2018).
According to Fabris, Camargo, Russo, & Zayas-Castro (2015), R&D are activities that differ in their purpose, knowledge bases, people involved, and management styles. The authors emphasize that research activity is understood as an instrument or tool for the discovery of new basic knowledge or applied knowledge, while development activity is oriented toward the application of new knowledge to obtain practical results.
In this context, R&D investments are related to the innovation capacity of firms, since the accumulation of this capacity occurs through corporate learning mechanisms, providing feasibility of innovation projects that the company aims to develop, and, consequently, increase its sales and boost the company's economic results (Ramos & Zilber, 2015).
By investing in R&D, companies enhance expertise in fields that represent new opportunities and contribute to the development of internal competencies aimed at developing new products and processes, which contributes to a radical departure from the organization's current knowledge base focusing on the Vol :-10 No-4, 2022 International Educative Research Foundation and Publisher © 2022 pg. 145 long term (D'Este et al., 2018). Thus, scientific knowledge is the one who serves as a resource for the implementation of the exploration strategy, as it can increase the effectiveness in the search for innovative and valuable solutions, helping researchers to experiment with new skills and technologies besides avoiding unnecessary efforts (Geiger & Makri, 2006).
That said, the literature suggests that the use of both strategies, exploitation and exploration, may be the key to achieving the benefits expected by firms. Thus, the studies by Chen (2017) Lennerts et al. (2020) analyzed the interaction of exploitation and exploration innovation strategies as a means by which firms come to ensure their long-term survival and prosperity.

Exploitation and exploration in previous studies
This section presents a brief overview of the interaction between exploitation and exploration based on recent studies about the themes, and finally presents the study hypothesis. Chen (2017) when examining the relationship between exploitation and exploration, evaluates the difficulties of combining both innovation strategies and identifies that there is a way to maintain a balance between the strategies. Thus, the control of radical innovation (exploration) with incremental innovation (exploitation) is achieved as a new way to overcome these difficulties, besides that by helping companies to manage both strategies, they ensure the sustainability of the business.
With a sample of 101 European small and medium-sized enterprises, Bauer et al. (2018) also performed the same relational analysis between exploitation and exploration, concluding that there is a synergistic effect between the two, where the demand for continuous strategic renewal indicates that firms need to balance activities that may be conflicting for resource allocation, competencies, and learning mechanisms of the two innovation strategies.
The study by Luger et al. (2018) analyzed the evolution of investments in firms' exploitation and exploration strategies and their long-term performance outcomes. The authors found that the dynamic exploitation-exploration balance combines capability building processes with capability change processes and warn that firms benefit from such a relationship thereby achieving superior performance.
In Solís-Molina et al. (2018), the exploitation and exploration relationship is analyzed in 281 manufacturing firms in Colombia, understanding that the combination of both strategies is important to raise the performance of firms.
Through a survey of 227 Brazilian software development companies, Severgnini et al. (2019) also concluded that the convergence of both strategies extends the organization's results. The authors suggest verifying the exploitation and exploration balance through a degree of innovation. Lennerts et al. (2020) argue that the interaction of exploitation and exploration is a source of competitiveness, pointing out that both strategies are equally important for a firm's long-term growth and survival, but stress that maintaining a proper balance in the use of these strategies is crucial to raising the firm's performance.
Given this empirical evidence, the research hypothesis is presented: the innovation strategies of exploitation and exploration are determinants of the degree of innovation in the Brazilian industries listed on B3.

Methodology
The study gathers a population of 129 industries listed on B3. For the data survey, the years from 2013 to 2019 were considered, which include the previous and subsequent three-year periods, taking into account the New Legal Framework for Science, Technology and Innovation -Law no. 13,243/2016. It should be noted that the triennium 2017 to 2019 corresponds to the most current one at the beginning of this research. Therefore, the study comprised the analysis of 903 observations.
Considering the literature review on the subject, Chart 1 presents the main indicators related to innovation modalities, their operationalization, and data sources.
It is noteworthy that this study adapts the classification of intangibles proposed by , considering R&D expenditures as a long-term radical innovation, for having higher risk and uncertain return (D'Este et al., 2018;Jia, 2018;Lampert & Kim, 2019;Solís-Molina et al., 2018). Thus, in this study, the indicators patents and trademarks measure the exploitation innovation strategy, while the Innovation Index is a proxy for the exploration innovation strategy.
Through the descriptive statistics it is presented the sample characteristics, taking into account the aforementioned indicators, as well as is analyzed through the Multiple Correspondence Analysis -MCA the association between the indicators related to the innovation modalities. According to Fávero, Belfiore, Silva, & Chan (2009), the MCA is a technique that displays the associations between a set of non-metric categorical variables in a perceptual map, thus allowing a visual examination of any pattern or data structure. To this end, categorizations were created based on the quartiles of the indicators referring to innovation, thus resulting in three categories or levels: low, medium, and high, as presented in Chart 2. Vol :-10 No-4, 2022 International Educative Research Foundation and Publisher © 2022 pg. 147 It is noteworthy that in the 1st and 2nd quartile indicators the maximum value is zero, so both quartiles made up the low-level categorization.

Chart 2 -Categorization of innovation indicators
Then, to identify the deterministic factors of the degree of innovation of the companies, the statistical technique Factor Analysis was used, which through the method of Principal Component Analysis (PCA) identifies the explanatory factors of the relationship of a set of variables (Corrar, Dias Filho & Paulo, 2014).
The technique applied considered as the number of significant correlations between the variables at least 0.30 and the Measure of Sampling Adequacy (MSA) greater than or equal to 0.50 in all variables. In the general model, the Kaiser-Meyer-Olkin (KMO) test was used to measure 0.50 for all variables, including the correlation matrix (Bartlett's Test). In the end, the number of factors extracted was based on the eigenvalue of each factor, where factors with an eigenvalue greater than one unit were considered, and to identify the variables that explain the factors, the factorial loads with an absolute value greater than 0.50 with factors rotated using the varimax method were considered.

Results and discussion
First of all, the study sample is characterized by the innovation indicators of annual patents, trademarks, and R&D, as shown in Table 1. In Table 1 presents the percentage of companies studied that made use of each mode of innovation. Given its results, on average, the innovation modality most used by companies in the period under analysis occurred through R&D, as 37.10% of the companies analyzed made such investments. In 2015 and 2018 were the years in which companies made the most use of innovation, since, on average, 33.59% of the firms adopted some modality of innovation in their context. The modality via patents stood out in 30.23% of the companies in 2015, while innovation through trademarks stood out in 2018, where 39.53% of the firms in the sample used such modality. Taking into account the New Legal Framework for Science, Technology and Innovation -Law No. 13,243/2016, only in 2018, there was an increase in the number of companies that adopted one of the innovation modalities, with the trademarks being the most prominent. After the publication of the New Legal Framework, there was a decrease in the number of firms that used the R&D strategic typology. The opposite behavior was observed in innovation via patents, but with lower percentages about the periods before the New Framework, a fact that can be justified by the gain in the robustness of Brazilian legislation regarding the development of innovative activities. It follows, then, that about innovation, there is a preference in companies for the exploration strategy practice, aiming at the development of new products and processes, focusing on the long term (D'Este et al., 2018).
The Figure 1 shows the number of registered trademarks and patent applications in the analyzed period. It can be seen that the number of registered trademarks was higher than the number of patent applications, especially before the advent of the New Legal Framework for Science, Technology and Innovation. As of 2016, the number of patent applications fell year by year, the opposite situation to that which occurred with trademark registration. Thus, one can see the importance that companies give to trademark registration as a way to improve their marketing strategies by focusing, for example, on incremental improvement of the existing product line with the purpose of satisfying consumer needs, typical of the innovation strategy exploitation (Barbosa et al., 2020). Another important difference concerns the duration of registration because patents are granted for a certain period (usually 20 years), while trademarks are, in principle, granted for an unlimited time, since the trademark holder keeps the trademark by paying renewal fees every 10 years. Despite the New Legal Framework for Science, Technology and Innovation in Brazil, which changes important rules aiming at creating a more dynamic innovation environment, it brings with it the dilemma of bureaucracy, which is also a factor for the reduction in the number of patent registrations.
In turn, Figure 2 contains the investments in research and development by the sample companies in the analyzed period. It can be seen that R&D investments were increasing until the year 2016, demonstrating the interest of firms in improving their ability to develop new products and processes in addition to improving existing ones (Cui, Ding, & Yanadori, 2019). The decline in this type of investment in the year 2017 may be linked to the reduction of government resources passed on to research funding agencies, as well as the enactment of the New Legal Framework for Science, Technology and Innovation in the country. However, it can be seen that firms have resorted to investments in R&D as a way to stand out and remain in the market through the exploration type of innovation strategy. It is notable in Figure 3 the association of the three innovation indicators at a low level. It can also be seen that innovation via trademark registration and innovation via R&D are associated at a high level, while innovation via patent application at a high level is associated with innovation via R&D at a medium level. Thus, the associations demonstrate a greater interest of companies in investing high volumes in R&D to expand their product line, meet consumer needs and strengthen their trademarks. It is noteworthy that these results corroborate the studies of Chen (2017) and Bauer et al. (2018), which suggest a synergy between exploitation and exploration innovation strategies.  Next, the descriptive statistics analysis was performed to verify the behavior of the survey data relative to the sample companies, Table 2. As shown in Table 2, the indicator INOV_QMAR has the highest average, but the variable INOV_P&D presents data with greater dispersion about the average, with a standard deviation equal to 9.5319, in addition to having more heterogeneous data, because it presents a coefficient of variation of 2,718.74%.
Subsequently, Spearman's correlation analysis was performed, since the Kolmogorov-Smirnov test indicated that the study variables do not have a normal distribution. Table 3 presents the results concerning the existence of correlation among the indicators of this research.  Note: (***) Significant at the 1% level.
Source: Research data. Table 3 shows that the innovation indicators are positively related with statistical significance at the 1% level. These results are indications of synergy between exploitation and exploration innovation strategies.
To meet the general objective of the research and test the research hypothesis, the factor analysis is carried out. A priori, it is important to highlight that, according to Corrar et al. (2014), the Principal Components Analysis method does not restrict the sample to the requirement of a normal distribution.
The anti-image correlation matrix among the innovation indicators revealed the first indication of data adequacy to factor analysis. According to this analysis, it was possible to raise the values of the measure of sample adequacy (MSA) for the study indicators. The MSA considers that values below 0.5 indicate that the variable does not fit the structure of the other variables and should therefore be eliminated. However, in this study, all indicators presented MSA with values greater than 0.5.
The communalities values represent the total variance explained by the factors in each variable. Initially, the communalities are 1 and after the extraction process, they start to vary between 0 and 1, being closer to 0 when the common factors explain low or no variance of the variable and 1 when all the variance is explained by all the factors. In this study, from the output resulting from the factor analysis processing by the statistical software, the communalities were greater than 0.666.
Bartlett's test of sphericity, used to verify the hypothesis that the matrix of correlations is the identity matrix, pointed to the rejection of the hypothesis, that is, the matrix of correlations can be the identity matrix, revealing the existence of correlations among the variables and indicating the suitability of the data for the application of factor analysis (Sig. 1%).
The KMO statistic ranges between 0 and 1 and compares the simple correlations with the partial correlations observed between the variables. According to the KMO statistic of the surveyed sample (0.500), the application of factor analysis becomes reasonable, allowing its use (FÁVERO et al., 2009). It is noteworthy that two factors were retained that can explain 77.54% of the variance of the original data.
The rotated component matrix, as shown in Table 4, presents the values of the loadings that correlate the variables with the factors after orthogonal rotation and has the objective of extreme the values of the loadings, in such a way that each variable is associated with a factor.   Table 4, it is understood that the indicators or variables of innovation are combined into two distinct factors, where the first factor encompasses the indicators INOV_QPAT and INOV_QMAR, and the second factor comprises the indicator INOV_P&D. Thus, it can be said that the degree of innovation of the firms takes into account the combination of indicators related to the strategies of innovation, exploitation and exploration, and by encompassing two distinct factors it can be said that the degree has the characteristic of being two-dimensional, exactly in the dimensions of exploitation and exploration.
After performing the factor analysis, followed by the definition of the influencing factors, the factor score values were tested through Pearson's correlation, resulting in a zero value for the factors found, since the varimax rotation method generates factors that are orthogonal to each other.
Thus, the finding in this research is in line with other studies on the topic such as in Severgnini et al. (2019), Luger et al. (2018) and Lennerts et al. (2020), where the authors suggest a balance between exploitation-exploration since the combination of capability building processes with capability change processes leads firms to benefit from such a relationship obtaining a superior result.
Therefore, the hypothesis of this study is confirmed, that the innovation strategies of exploitation and exploration are determining factors of the degree of innovation of the Brazilian industries listed on B3, since the combination of radical innovation (exploration) with incremental innovation (exploitation) is a way to overcome difficulties, helping companies to manage both strategies to ensure the sustainability of the business.

Conclusion
The objective of the present research was to identify the determinants of the degree of innovation in industries listed on B3. It was decided to use the indicators patents, trademarks and R&D as proxies of innovation strategies, proceeding with a quantitative analysis through descriptive statistics and multiple correspondence analysis to characterize the sample and factor analysis through the method of principal component analysis to identify the factors that make up the degree of innovation.
The results showed that the innovation modality most used by companies in the period under analysis occurred through R&D investments since 37.10% of the companies making such investments. Nevertheless, the number of registered trademarks was higher than the number of patent applications, especially before the advent of the New Legal Framework for Science, Technology and Innovation. As of 2016, the number of patent applications fell year by year, the opposite situation to that which occurred with trademark registrations. It can be seen, then, that there is a preference among companies for the practice of the exploration strategy, aiming at the development of new products and processes, in addition to obtaining prominence and permanence in the market, focusing on the long term.
It was verified through multiple correspondence analyses the association, at a high level, of innovation via trademark registration with innovation via R&D, while innovation via patent application at a high level is associated with innovation via R&D at a medium level. It can be seen then that the associations demonstrate a greater interest of companies to invest high volumes in R&D to expand their product line, meeting the needs of consumers and strengthening their trademarks. Through Spearman's correlation, it was found that the innovation indicators are positively related with statistical significance at the 1% level. Thus, the findings described here suggest the existence of a synergy between the innovation strategies of exploitation and exploration.
The innovation indicators or variables were combined into two distinct factors. Thus, it can be said that the degree of innovation of firms takes into account the combination of indicators referring to the innovation strategies of exploitation and exploration. Thus, by encompassing two distinct factors, one can say that the degree of innovation has the characteristic of being two-dimensional, exactly in the dimensions of exploitation and exploration.
The findings of this research contribute to the literature by increasing the relative knowledge about the degree of innovation, from the combination of factors related to the innovation strategies of exploitation and exploration, since no studies were identified on the deterministic factors of this degree. The results can provide answers and possibilities for more concrete actions by managers in the decision-making process, since innovation is a strategic factor of the first order in companies.
Despite the methodological rigor adopted, the results were obtained from an intentional sample. Therefore, a deeper investigation with a larger universe of companies is suggested, thus considering the analysis of all companies listed on B3. It is also noteworthy the possibility of a comparative analysis of the degree of innovation of companies from developing countries and companies from developed countries since the latter has more developed capital markets and technologies.