“The effect of related party transactions on R&D investment: Evidence from Korea”

This study aims to investigate the relationship between related party transactions and a firm’s investment in research and development (R&D), as well as the moderating effect of a firm’s financial health on such a relationship. The study applies a fixed-effect panel regression model with a sample of 13,619 Korean listed firms for the period from 2001 to 2020. The results indicate that related party transactions significantly and positively influence a firm’s R&D investment at the 1% level for the study period. Specifically, when related party transactions are divided into operating and non-operating, the re-sults show that only non-operating related party transactions significantly and positively affect firms’ investment in R&D. Moreover, findings report that the effect of related party transactions is stronger for firms with financial distress, lower cash holdings, and in the high-tech industry. The results imply that related party transactions promote a firm’s R&D investment, which is one of the primary business investments that create a firm’s competitive advantage and value. Moreover, the results propose that related party transactions should be carefully evaluated when accessing the firm’s investment behavior.


INTRODUCTION
According to the International Financial Reporting Standard (IFRS), a related party transaction (RPT) is a "transfer of resources, services, or obligations between the reporting entity and a related party" (IASB, 2009, A1268), such as major shareholders, affiliates, or subsidiaries.Since the well-known accounting scandals such as Enron and Parmalat, RPTs have been a prominent topic in the capital market and have drawn significant attention from regulators and academics.Those scandals have been attributed to the extensive use of RPTs to conceal their fraudulent activities, revealing the inherent risk of RPTs.Such crises have obliged regulators and investors to raise concerns about whether conducting RPTs benefits shareholders and the firm itself.
Considering the importance of RPTs, extant research has examined the effects of RPTs on a firm's valuation, performance, and financial reporting (Cheung et al., 2009;Jian. 2003;Nekhili & Cherif, 2011).Nevertheless, there is no consistent evidence about the effects of RPTs.Specifically, there are two prevailing established hypotheses regarding RPTs: the conflict-of-interest and efficient transaction hypotheses.The former considers RPTs as harmful transactions that destroy firm value (Chen et al., 2011;Johnson et al., 2000;Rahman & Nugrahanti, 2021).In contrast, the latter contends that RPTs are efficient and effective transactions under imperfect information, lowering transaction costs and generating an internal capital market where firms can share information and resources (Stein, 1997;Williamson, 1975).These inconclusive views of previous studies clearly imply that it is still an empirical question whether RPTs are efficient contracting mechanisms or harmful transactions.
For the past decade, a firm's R&D activities have expanded considerably with the rapid development of technologies and have become one of the most crucial investments for a significant fraction of listed companies (Ocean Tomo, LLC, 2021).Extant studies show that R&D investments drive economic and firm growth (Kuznets, 1967;Lucas, 1988;Schumpeter, 1939).In this context, extant studies investigate drivers of R&D investments to identify factors that motivate firms to invest in R&D, given the considerable effects R&D investments have on firms and the economy (AlHares, 2020; Baldi & Bodmer, 2018; Geroski & Pomroy, 1990).Although there is considerable interest in R&D investments, only few studies have investigated the influence of RPTs on R&D investment.

LITERATURE REVIEW AND HYPOTHESES
In terms of the effect of RPTs, there are two contradicting perspectives supported by two different hypotheses from prior literature.According to the conflict-of-interest hypothesis, RPTs are harmful transactions intended to extract resources from minority shareholders.For instance, Berkman et al. (2009) demonstrate that controlling shareholders expropriate wealth by issuing loan guarantees to firms in which they hold large shares or are controlled by them.Similarly, Johnson et al. (2000) argue that controlling shareholders "tunnel" minority shareholders' wealth through transactions between related parties.Cheung  By examining the relationship between RPTs and R&D investment, this study intends to provide novel insight into the impact of RPTs.Moreover, this study investigates the moderating effect of a firm's financial condition on such a connection.Based on the prior literature reviews and arguments, this study anticipates a significant association between RPTs and a firm's R&D investment.In addition, the study predicts that the impact of RPTs will be stronger for financially distressed firms.This study, therefore, set the following hypotheses: H1a: There is a negative relationship between RPTs and a firm's R&D investment (Conflictof-interest hypothesis).
H1b: There is a positive relationship between RPTs and a firm's R&D investment (Efficient transaction hypothesis).
H2: The effect of RPTs on R&D investment is stronger for financially distressed firms.
In the model, RD denotes the firm's R&D intensity, which captures the total R&D expenditures of a firm.RPT is the firm's total amount of RPTs, which is the independent variable.This study categorizes RPTs as operating (OP_ RPT ) and non-operating (NON_OP_ RPT ) RPTs in order to verify the RPT type that significantly impacts a firm's R&D investment.Operating RPTs present transactions of services, materials, or goods, and non-operating RPTs capture the transactions related to fixed and investment assets (Kang et al., 2014).The model includes firm characteristics that may potentially influence R&D investments.Each variable in the model is described in detail in Appendix A. Standard errors are adjusted to confirm robustness within-firm cluster correlations (Petersen, 2009).Furthermore, this study employs a fixed effect model to account for the industry and year fixed effects.Lastly, the lead-lag test model is adopted to address the influence of omitted variables and causality issues by incorporating lagged independent and control variables.
To test the second hypothesis, samples are separated according to Altman's (1968) Z-score, which measures the degree of a firm's financial health.Specifically, this study divides sample firms into three subsamples depending on Z-score: (

RESULTS
The empirical results for hypotheses 1a and 1b are presented in Table 3.As shown in Panel A of Table 3 Notes: (1) All variables are defined in the appendix.( 2) All continuous variables are winsorized at the 1% and 99% levels.
( Panel B of Table 4 illustrates the results when sample firms are divided into two subgroups based on the median of the industry.The results indicate that the association between RPTs and R&D investment is statistically significant.In addition, the results indicate that this connection is stronger for firms with Altman's Z-score below the median of the industry.For firms with a low Z-score, the coefficient of RPT is 1.4446 and statistically significant at the 1% level.Similarly, the coefficient of RPT (0.6731) for firms with high Altman's (1968) Z-score is also significant and positive (p-value = 0.0909).However, the coefficient of RPT is bigger for firms with low Z-scores than for those with high Z-scores.Statistically, their differences are significant at the 1% level.The findings corroborate the efficient transaction hypothesis, which states that RPTs influence a firm's investment in R&D by establishing and developing the internal capital market among related parties.ples into two subsamples based on a company's cash holdings.Table 5 displays the test results on the influence of RPTs on R&D investment based on the firm's cash holdings: firms with large cash holdings and firms with low cash holdings based on the median of the industry.The findings indicate that the regression coefficient of RPT for firms with low cash holdings is 1.4719 and statistically significant at the 1% level.Moreover, the coefficient of RPT for firms with lower cash holdings is stronger than that of firms with large cash holdings (0.8579, p-value < 0.01).In addition, they are significantly different at the 1% level, indicating that the influence of RPTs is stronger for firms with low cash holdings than those with large cash holdings.
Prior research indicates that the marginal value of internal finance is greater for companies in the high-tech industry.For example, Hu et al. (2017) argue that information asymmetry is high for companies in the high-tech industry, increasing the firm's cost of eternal capital and creating difficulties for investors in evaluating the firm's value.Additionally, Myers and Majluf (1984) claim that firms in the high-tech industry maintain significant information asymmetry strategically in order to preserve the return on R&D investment and enhance future competitiveness.Hence, firms in the high-tech industry have to finance their R&D investment through internal finance (Chen & Lee, 2018; Himmelberg & Petersen, 1994).
Given that firms in the high-tech industry are more likely to rely on internal finance or internal capital markets to support R&D activities, the positive relationship between RPTs and firms' R&D investments would be stronger for firms in the high-tech industry than those in the low-tech industry.To test this argument, this study conducts the additional analysis by dividing the samples into two subgroups.Following Kile and Phillips (2009) Table 6 reveals that the coefficient of RPT for firms in the high-tech industry is 1.0424, showing a positive and statistically significant value at the 1% level.Moreover, it is greater than the coefficient for firms in the low-tech industry, which is also positive (0.6671) and significant at the 1% level.Furthermore, their difference is statistically significant at the 1% level, demonstrating that the  Additionally, the study conducts the change analysis by separating operating and non-operating RPTs from the total RPTs.Panel B of Table 7 shows that changes in non-operating RPTs are statistically significant and positive, indicating that non-operating RPTs are positively related to changes in the R&D investment of the firm.This demonstrates that the main result still holds even after addressing the causality issues.

DISCUSSION
The  The result for H2 reveals that the favorable impact of RPTs on R&D investment is stronger for financially distressed firms.This finding lends credence to He et al. (2013), who show that firms with financial difficulties are more prone to rely on internal capital markets.Given that R&D characteristics result in a higher cost of external capital, the role of RPTs in establish-ing and facilitating the internal capital market is important in financing R&D activities.This implies that when firms are financially healthy and have sufficient abilities to finance R&D activities, the positive effect of RPTs on R&D investment may not be influential.By contrast, when firms are financially distressed and face difficulties in accessing external capital, the effect of RPTs will be stronger.
Collectively, in confirming that RPTs have a favorable influence on a company's R&D investment, the results provide new insight into the contradictory evidence on the impact of RPTs.It provides evidence in support of the efficient transaction hypothesis, which states that RPTs are efficient transactions that allow firms to achieve economic efficiency.Furthermore, the results indicate that the financial health of a firm is a significant moderator of the positive relationship between RPTs and R&D investment.

CONCLUSION
The objective of this study is to examine the impact of RPTs on R&D investment.Moreover, this study examines how the financial condition of a firm influences the relationship between RPTs and R&D investment.
The study's primary finding is that RPTs are positively related to a firm's R&D activities.The study also finds that the positive impact of RPTs on R&D activities is more pronounced for financially distressed firms.In addition, the study demonstrates that the impact of RPTs on a company's R&D investment is greater for firms in the high-tech industry and those with low cash holdings.The primary results of this study still hold even after addressing omitted variables and causality issues using the lead-lad test model and change analysis.
The study's findings provide new insight into the contradictory evidence on the effect of RPTs.This extends and advances the relevant literature by demonstrating that RPTs are significantly related to R&D investment, which is crucial to its future growth and long-term sustainability.While majority of research examines the impact of RPTs on firms' value, performance, and financial reporting, the study provides empirical evidence that RPTs are also a major driver of a firm's investment in R&D.Collectively, the study demonstrates that transactions among related parties facilitate R&D investment by helping their financing activities via internal capital markets.Moreover, the study provides the practical implication that market participants, including investors and regulators, should focus on the firm's RPTs to set effective investment decisions and desirable policies on RPTs as they could drive firms' investment in R&D, a crucial component of a firm's future growth and long-term sustainability.
) The numbers in parentheses are p-values.(4) All p-values are based on two-tailed tests.

Table 1 .
Descriptive statistics This study analyzes data on publicly traded Korean companies for the period 2001-2020.The data on the firm's financial information and RPTs are retrieved from the FnGuide and TS2000 databases, which are comparable to Compustat in the United States.From the sample, financial institutions are excluded due to their unique industry features.Moreover, to maintain sample homogeneity, firms without fiscal yearends of December are also removed from the sample.Finally, firms without necessary data for variables are excluded, resulting in a large sample size of 13,619 firm-year observations.Table1summarizes the descriptive statistics for variables used in the model.Each continuous variable is winsorized at the top and bottom 1%.The mean and median values of RD, the dependent variable, are 1.1524 and 0.1551, respectively.The value also indicates that the average of the RPT is 0.2540.OP_ RPT and NON_OP_ RPT have respective mean values of 0.0652 and 0.1868, indicating that the amount of non-operating RPTs is higher than that of operating RPTs.Notes: (1) All variables are defined in the appendix.(2) All continuous variables are winsorized at the top and bottom 1%.(3) All p-values are based on two-tailed tests.

Table 2
reports the Pearson correlations, indicating that the RD, a dependent variable, positively correlates with the independent variable RPT.It also indicates that RD is positively correlated with non-operating RPTs (NON_OP_ RPT ), while operating RPTs (OP_ RPT ) are not.However, it is difficult to draw an accurate conclusion about the effect of RPTs on a firm's investment in R&D based on the Pearson correlation coefficient.The results of regression analyses are reported in the next section, taking into account all variables used in the analyses.

Table 4
displays the results of the second hypothesis.Panel A of Table4reports the findings of the Z-score-based analysis that separates sample firms into three subgroups.The findings indicate that the effect of RPTs on the firm's investment in R&D is stronger for financially distressed and gray firms.The coefficients of RPT for gray and distressed firms are 1.2845 and 1.4191, respectively, and are statistically sig- nificant at the 1% level.However, for financially safe firms, the coefficient of RPT is not statistically significant, indicating that RPTs have a greater impact on financially distressed firms.

Table 4 .
The effect of RPTs on a firm's R&D investments based on the firm's financial health All variables are defined in the appendix.(2) The numbers in parentheses are p-values.(3) All p-values are based on two-tailed tests.(4) All continuous variables are winsorized at the top and bottom 1%.(5) This table documents the OLS regression results of the relation between Relate Party Transactions and R&D conditional on firms' financial health using Altman's (1968) Z-score.Sample firms are partitioned into two groups based on industry median Z-score.
Notes: (1) All variables are defined in the appendix.(2) The numbers in parentheses are p-values.(3) All p-values are based on two-tailed tests.(4) All continuous variables are winsorized at the top and bottom 1%.(5) This table documents the OLS regression results of the relation between Relate Party Transactions and R&D conditional on firms' financial health using Altman's (1968) Z-score.Sample firms are partitioned into three groups based on Z-score: (1) safe firms with Z-score > 2.99, (2) gray firms with 1.8 < Z-score <= 2.99, and (3) distressed firms with Z-scores < 1.8.Panel B: Subsample analysis based on the industry median of Altman's (1968) Z-score

Table 5 .
Effect of RPTs on firm's R&D investments based on the firm's level of cash holdings

Variable Dependent variable = RD t Difference Test Firms with High Cash holdings Firms with Low Cash holdings Coef.
of RPTs is greater for firms in the hightech industry.The result of the study supports the authors' contention that RPTs facilitate firms' investment in R&D by forming and enhancing the internal capital market within related parties.
Notes: (1) All variables are defined in the appendix.(2)The numbers in parentheses are p-values.(3)All p-values are based on two-tailed tests.(4)All continuous variables are winsorized at the top and bottom 1%.influence

Table 6 .
Effect of RPTs on firm's R&D investments based on the industry type Notes: (1) All variables are defined in the appendix.(2) The numbers in parentheses are p-values.(3) All p-values are based on two-tailed tests.(4) All continuous variables are winsorized at the top and bottom 1%.

Table 7 .
Robustness test: Changes in RPTs and changes in the firm's R&D investments

Variable Dependent variable = ∆RD t Coef. p-value
Notes: (1) All variables are defined in the appendix.(2) The numbers in parentheses are p-values.(3) All p-values are based on two-tailed tests.(4) All continuous variables are winsorized at the top and bottom 1%.
Change analysis by dividing RPTs into operating and non-operating RPTs Notes: (1) All variables are defined in the appendix.(2) The numbers in parentheses are p-values.(3) All p-values are based on two-tailed tests.(4) All continuous variables are winsorized at the top and bottom 1%.

Table 7 (
Kang et al. (2014))est: Changes in RPTs and changes in the firm's R&D investments RPTs, firms are more likely to employ non-operating RPTs to transfer large sums of resources.Moreover, non-operating RPTs involve more discretion and subjective judgment than standard operating activities, allowing firms to share and allocate resources more efficiently(Kang et al., 2014).Hence, non-operating RPTs would be more influential than operating RPTs in facilitating a company's R&D investment.The study finds that only non-operating RPTs significantly affect a firm's investment in R&D, which is in line with the claims and findings in research byFan et al. (2008)andKang et al. (2014).