1. Introduction
Recently, developed countries are in the process of promoting various environmental regulations to respond to climate change, and ESG management is becoming a widely discussed topic worldwide. ESG management, which has established itself as a new standard in corporate management, includes carbon reduction, social contribution, circular economy, and corporate governance. In recent years, Korean companies have rapidly adopted ESG management to keep up with the demands of the new era, recognizing sustainability management as one of their active management strategies; they are striving to fulfill the corporate social responsibility by protecting customer information and the environment in consideration of various stakeholders and acknowledging the diversity of our employees [
1]. Furthermore, consumers have started to evaluate the company based on the value that the company pursues, not simply based on the products or services. This trend lets ESG emerge and makes it an essential element in assessing the sustainability of corporate [
2]. As ESG became the core of corporate management, international organizations such as the European Commission (EC), the International Organization of Securities Commissions (IOSCO), and the OECD began to understand the current status and problems of ESG evaluation. In addition, the requirements for ESG evaluation agencies to strengthen their transparency are increasing [
3].
Amazon, which has grown into a global conglomerate with an aspiration to sell everything globally, has become a significant culprit in carbon emissions due to high power consumption and environmental pollution related to packaging and transportation in the process of rapid growth. This led them to take attention to ESG. Amazon founder and former CEO Jeff Bezos announced the subject of the “climate pledge” in September 2019, Ref. [
4] and Amazon began to invest USD 2 billion won in clean energy-related companies to implement the climate pledge; it then began to acquire the aspects of a sustainable company by investing in a large-scale fund. In addition, it decided to use 100% renewable energy by 2030 and signed a contract in June 2021 to procure eco-friendly energy at 14 solar and wind power plants in the U.S. and Europe. Amazon is also cooperating with all its partners to achieve its goal of “zero” carbon emissions by 2040 [
4]. By working from various aspects to treat environmental issues, Amazon is transforming into an eco-friendly company, and its reputation is growing as the company that uses the largest amount of renewable energy in the world. This trend is seen not only in Amazon, but in a lot of multinational companies such as Apple, which achieved 100% renewable energy, and Microsoft, which vowed carbon negative, have conducted sustainable management through successful ESG with their superior technologies.
POSCO, which published Korea’s first “environmental report” in 1995, has jointly carried out improvement tasks with business partners since 2004, and then implemented the “performance sharing system” to share results [
5]. In addition, when Chairman Choi Jung-woo took office in 2018, he emphasized the company’s role as a “corporate citizen” who fulfills social responsibility, not just pursuing profits. The company declared “2050 Carbon Neutral” replacing energy sources for steel production with hydrogen instead of coal. This new policy included the new introduction of safe products and solutions that can reduce carbon instead of existing buildings, wood, stone, and aluminum for the first time in Korea, in December 2020, Ref. [
5]. As such, POSCO is striving to respond to global climate change by expanding renewable energy in the medium and long term and actively practicing ESG management. In addition, several Korean manufacturing companies are trying to introduce ESG successfully: Samsung Electronics has expanded ESG management to its partners; SK Group expanded the use of renewable energy; LG Electronics included the environmental burden financially by introducing an internal carbon tax, and Hyundai Motor have planned to make large-scale investments on ESG by 2030.
Social contribution in ESG refers to corporate charity activities, and the Global Reporting Initiative (GRI) standard refers to the performance of social contribution activities as “GRI413-Local Communities” to disclose detailed activities such as social contribution strategies, expenditures, and performance [
6]. In Korea, the annual social value report is published every December by the Federation of Korean Industries; they defined social contribution as companies’ educational, cultural preservation activities, job creation, and social investment through support for the vulnerable. According to an analysis of 220 top Korean companies, social contribution expenditure was about USD 3 trillion won and average expenditure of one company was USD 13.6 billion won; indeed, the total expenditure has increased 14.8% from the previous year [
7]. The above statistic clearly shows that the companies are increasingly interested in realizing social values through investment in social responsibility and environmental responsibility through ESG management, and they are trying out various ways to achieve shared growth with partners [
8].
Recently, ESG research has studied the relationship between the business performance of companies only with rating results announced by evaluation agencies; however, it is not clear whether the company’s ESG activities are reflected in the evaluation. Additionally, the validity and certainty of the evaluation items are not proven delicately. To fill this gap, this study aims to develop the ESG index, a key indicator for examining ESG from a non-financial perspective and evaluating corporate sustainability, based on existing ESG-related research and reports. In addition, we classified the companies into high and low groups according to the ESG index. The impact of CSR investment on management performance was also examined from a financial perspective. This study would be able to play a role as the primary source of data for ESG-related research by developing an ESG index based on corporate sustainability. At the same time, this study has its implication in that we expanded the research field by reinterpreting the relationship between investment and management performance from a financial perspective through comparison of high and low groups; thus, this article can provide some proofs that the management paradigm of corporate social responsibility is now changing, and not only financial but non-financial perspectives also should be highlighted.
This study has been divided into five parts: the
Section 1 described the background and purpose of this study. The
Section 2 examines the concept and status of ESG from a non-financial point of view and ESG from a financial point of view based on previous studies and literature. The
Section 3 presents the framework and three methodologies for ESG index development and establishes a hypothesis on the relationship between CSR investment and business performance. The
Section 4 describes the relationship between the results of the ESG index development and the management performance of each group.
Section 5 deals with the implications of this study and future research directions.
3. Research Methods and Design
3.1. ESG Index Development
One of the management indicators was to reduce the defect rate in the production process, avoid injuries to workers, and achieve the goal of products produced in the traditional manufacturing-based industry. In terms of business performance, maximizing efficiency from an economic point of view was the company’s survival strategy; however, as ICT develops, industrial efficiency from a mere economic standpoint alone cannot preempt the competition and cannot protect the company from the threat of suppliers, competitors, and customers. Thus, the importance of environmentally responsible management, socially responsible management, and transparent management is growing and emphasized, which is the basic framework of ESG management [
25]. Since economic and social values occur together, corporate actions interact with the benefit of society, and responsible social activities are becoming the driving force of corporations.
In addition, ESG management is becoming a globally emerging issue in terms of its role as an important factor in assessing companies’ sustainability [
26]. The study aims to develop the ESG index evaluating sustainability with 49 companies issuing a sustainability report among the top 150 Korean companies. ESG keywords were extracted from the CEO messages in Sustainability Report from 2013 to 2020, and those were used to calculate the ESG index.
Figure 3 is a framework for the development of ESG indices. Using the CEO’s greeting from the report as the raw material, we calculated weights using the TF-IDF technique compared to the ESG keyword, the result of the researcher’s previous study. TF-IDF is a more reliable technique than relying on simple frequency as a weight representing how important a word is within a particular document using the frequency of appearance. In other words, it is a technique obtained by multiplying the frequency and frequency of reverse documents with keywords that appear in a large number of documents [
27]. In addition, the characteristics of ESG words were extracted using NMF (Non-Negative Matrix Factorization), a document classification technique. NMF is a technique for decomposing the extracted characteristic matrix into two matrices, W (weighted) and H (characteristic) [
28]. It has the advantage of being quickly decomposed by using features that do not have negative elements in the data. The extraction method for the NMF technique was carried out with three methodologies.
The extraction method for the NMF technique was carried out with three main techniques. Data collection was downloaded directly from each company’s website. Next, text mining was analyzed using Linux_R. In addition, we developed algorithms using R packages tm and NMFN.
- (1)
NMF_mm multiplicative update method
As a basic method of the NMF technique, it is a technique applied using the product of the matrix (Equation (1)). The matrix is updated by Equations (2) and (3) and is repeated until satisfied.
- (2)
NMF_als—alternating least-squares method
Alternating least-squares involves calculating a single feature vector using a least-squares function (Equation (4)) in a way that repeatedly calculates and obtains a solution, and normalization is applied to prevent overfitting in the process of minimizing errors. In order to calculate the user feature vector until the optimal solution is obtained, it refers to a method of first fixing the item feature vector as a constant and solving it with the least number of squares.
- (3)
NMF_prob—multinomial method
The polynomial method multiplies the weight (Equation (5)) and the characteristic (Equation (6)) using a diagonal matrix and is calculated by applying the following equation.
When using the diagonal matrix of P, the weight (W) consists of a product of itself, a diagonal matrix, and the number of factors (k), and the characteristic (H) consists of a diagonal matrix and its product, calculated as the matrix product of Equation (1).
3.2. CSR Investment and Business Performance
A financial statement containing financial information show the company’s current status, and the company’s financial performance can be confirmed by increasing sales and profits and rising stock prices. In addition, investors mainly focused on financial factors when measuring corporate value before the importance of non-financial factors emerged recently; thus, corporate value has been evaluated by quantitative and short-term indicators such as financial statements. In addition, traditionally, the size of management compensation has been determined based on market value and profitability, and accounting information in financial statements has been mainly used when determining compensation by measuring management performance [
29]. Among them, donations in financial statements are voluntary expenditures by companies regardless of business activities and mean that they are used for the public interest, such as education, politics, culture, and social welfare. Indicators used in management performance include net profit per share, which indicates market performance; net income, which indicates profitability; sales and operating profit, and returns on assets. Therefore, in this study, donations are set as CSR investments, and the following hypothesis is established to empirically confirm the relationship between CSR investment and management performance by dividing the company’s management performance by ROA (Return On Assets; ROA), Return on Equity (Return On Equity; ROE), and Return on Sales (Return On Sales; ROS).
Hypothesis 1. CSR investment will have a significant positive (+) effect on a company’s management performance.
In order to correctly estimate the effect of CSR investment on the management performance of a company, other exogenous variables are controlled. In this study, the company’s size, the presence or absence of a chaebol, ownership type, and advertising cost were considered control variables. In addition, CSR investment donations were divided into total assets to confirm the ratio of donations for the year, and the company’s management performance was set as the year (t) and the following year (t + 1) to secure causality and confirm the Time-lag effect.
4. Research Results
In order to develop the ESG index, weights were extracted using the TF-IDF technique, and after extracting the characteristic values of words using NMF, the formula of the research method was used. The ESG index of 339 data of 49 companies was calculated from 2013 to 2020, and the ESG index of each company was confirmed by applying three indices. Method (1) NMF_mm used for calculation will be described by naming NMF_mm ESG Index 1, Method (2) NMF_als ESG Index 2, Method (3) NMF_prob ESG Index 3. The minimum value of ESG index 1 was 0.362, the maximum value was 6.594, and the standard deviation was 1.060. The minimum value of ESG index 2 was 4.756, the maximum value was 31.609, and the standard deviation was 4.821. Finally, the minimum value of ESG index 3 was 7.253, and the maximum value was 129.627, and the standard deviation was 21.229. In addition, the average value for each index was calculated using the developed ESG index, and ESG index 1 was 3.748, ESG index 2 was 18.980, and ESG index 3 was 53.542, whose average was 25.578. Samsung Heavy Industries had the lowest ESG index at 13.536. Lotte Shopping had the highest ESG index, followed by POSCO Energy with 36.928.
In addition, to empirically analyze the relationship between CSR investment and management performance, this study used 339 data from 2013 to 2020 and 290 data from t + 1 for 49 companies that published sustainability reports. Based on the average values of ESG indices 1, 2, and 3, all 17 companies that appeared above average were divided into stubborn groups and 14 companies that appeared below average were divided into low groups. The descriptive statistics and correlation coefficients of the variables used in the empirical analysis are presented in
Table 2 (high group) and
Table 3 (low group). The relationship between CSR investment and management performance in this study is consistent with the hypothesis predicted.
The high group showed insignificant results in the correlation analysis, and the low group showed a significant correlation in the relationship between CSR investment and business performance. Thus, only the low group supported Hypothesis 1 through correlation analysis. Additional testing was performed using OLS linear regression analysis.
Table 4 presents a result of regression analysis with high group. CSR investments in companies in financial performance did not have a significant influence on ROA, ROE, ROS both at t year and t + 1 year.
Table 5 shows the result of regression analysis with the low group. The relationship of CSR investment and management performance was confirmed. As for year t, ROA (β = 0.447,
p < 0.001), ROE (β = 0.264,
p < 0.05), and ROS (β = 0.367,
p < 0.01) were all significantly influenced by CSR investment significantly positively (+). F value of regression equation with ROA as dependent variable was 9.409 (
p < 0.001), and the modified R
2 was 0.294. F value of model with ROE was 5.638 (
p < 0.01) and the modified R
2 was 0.187. F value with ROS model is 3.560 (
p < 0.001) and the modified R
2 was 0.112.
As for t + 1 year, CSR investment was a significant variable in ROA (β = 0.408, p < 0.001), ROE (β = 0.253, p < 0.05), and ROS (β = 0.444, p < 0.001) models. F value of ROA model is 7.529 (p < 0.001) and the modified R2 was 0.273. As for ROE model, F value was 4.486 (p < 0.01), and the modified R2 was 0.167. F value of ROS model was 3.879 (p < 0.01), and the adjusted R2 was 0.142.
In recent years, the relationship between ESG evaluation and business performance has been the central issue in this field. The current study developed ESG indices through text mining. In addition, we divided the CSR investment into high and low groups for comparative analysis and confirmed that the low group achieved positive and short-term performance in CSR investment.
5. Discussion and Future Research
5.1. Conclusions
The topic of this study is mainly about a recently emerging ESG. ESG index was developed using the ESG keyword of the researcher’s previous study result extracted from the Sustainability Report. ESG index of 49 companies was calculated by year through three methodologies extracted by the NMF technique and based on the average value of each index, 17 companies with an average value or higher were divided into high groups and 14 companies with an average value or lower were analyzed empirically. As a result of regression analysis, the relationship between CSR investment and management performance was not significant with the high group, but the CSR investment and management performance of the low group had a significantly positive (+) effect on both t and t + 1.
From a non-financial point of view, ESG management is based on environmental protection by creating an eco-friendly ecosystem, continuous management of supply chains, global social contribution, and establishment of a diverse and healthy corporate culture. On the other hand, CSR investment refers to the amount of expenditure used to solve problems in the community and improve the quality of life through activities included in social contribution during ESG management. Among them, donations are material contribution activities through direct donations and correspond to charitable donations, so it was confirmed that social contribution activities through direct donations are dominant in the case of low groups. In addition, with the descriptive statistics, the average CSR investment of the high group was 0.105, and the value of the low group was 0.151, indicating that the low group had higher CSR investment than the high group. The low group had a significantly positive (+) effect on CSR investment and corporate management performance because the companies in the low group are more active in CSR investment and tend to focus more on short-term profits. On the other hand, it can be seen that the advertising ratio is high at 0.577 and the low group at 0.012. This shows that the companies in the high group are gradually investing in advertisements rather than CSR investments to appeal to various stakeholders as well as advertisements for products. As such, low groups focus more on ESG from a financial perspective than ESG from a non-financial perspective. Various stakeholders will recognize the importance of mid-to-long-term strategies based on sustainable management rather than short-term performance and practice ESG management from a non-financial perspective.
5.2. Discussion of Implications
The implications of the current study are as below: First, the prior studies mainly used KEJI, ESG, and K-GWP, which are indicators to evaluate sustainability. This study is meaningful in that we developed ESG indexes based on ESG keywords extracted using text mining. Second, a quantitative ESG index was developed by extracting weights through TF-IDF techniques and multiplying the frequency of inverse documents. This provides the basis for generalizing the above three indices. Third, by quantifying the ESG index and comparing the sample companies into high and low groups, the relationship between CSR investment and corporate performance could be confirmed. As a result, companies from low group focused more on ESG from a financial perspective than ESG from a non-financial perspective. Fourth, by dividing a company’s sustainability into financial and non-financial perspectives, basic data was presented to seek efficient management measures that companies should consider when establishing strategies for ESG. Finally, this study presented a new perspective on corporate ESG evaluation and meaningful exploratory implications by analyzing the relationship between CSR investment and corporate management performance.
5.3. Limitations and Future Research
This study examined the company’s ESG management from a financial and non-financial perspective. Unlike previous studies, developing the ESG index through non-quantitative data can be meaningful, but limitations exist in sample data and measurement. First, 339 data from 49 companies that published Sustainable Reports from 2013 to 2020 were used, but the number of samples is quite limited. If it becomes mandatory to publish reports for all companies, more data would be available for researchers. Second, it is necessary to consider both financial and non-financial performance in the impact of CSR investment on corporate management performance. It is necessary to analyze whether the ESG index and CSR investment have a significant effect on Tobin’s Q to confirm the relationship with corporate value, and to study the relationship with ESG by reorganizing key performance indicators inside and outside the company into four perspectives using the balance scorecard (BSC).
Third, we used the size of the company as a control variable in analyzing the relationship between CSR investment and corporate management performance. If we considered the company’s size when developing the ESG index, the index could be much more generalized. In addition, many companies are investing in advertising expenses to raise the company’s image among stakeholders, so it would be meaningful to conduct further research from the perspective of advertising expenses.
Fourth, rather than focusing on CSR investment, it is necessary to check whether it affects a company’s management performance through R&D expenses when analyzing high group companies. For example, CSR investment invests in planting trees in the automobile industry, but ESG management invests in R&D expenses to develop hybrid, hydrogen, and electric vehicles, so it is necessary to analyze the relationship with R&D expenses. Lastly, it will help improve the ESG index if various stakeholders’ opinions on sustainable management are collected, and each company’s focus in ESG management is considered. It will be more meaningful as a reliable ESG index for corporate practitioners from a strategic point of view. Further studies on ESG management should be continued to improve ESG in the business field, not only for large companies, but also for small and medium-sized companies and ventures.