The effect of related party transactions on firm performance: The moderating role of political connection in Indonesian banking

Supatmi; T., Sutrisno; Saraswati, Erwin; Purnomosidhi, Bambang Article The effect of related party transactions on firm performance: The moderating role of political connection in Indonesian banking Verslas: Teorija ir praktika / Business: Theory and Practice Provided in Cooperation with: Vilnius Gediminas Technical University (VILNIUS TECH) Suggested Citation: Supatmi; T., Sutrisno; Saraswati, Erwin; Purnomosidhi, Bambang (2019) : The effect of related party transactions on firm performance: The moderating role of political connection in Indonesian banking, Verslas: Teorija ir praktika / Business: Theory and Practice, ISSN 1822-4202, Vilnius Gediminas Technical University, Vilnius, Vol. 20, pp. 81-92, https://doi.org/10.3846/btp.2019.08 This Version is available at: https://hdl.handle.net/10419/247961 Standard-Nutzungsbedingungen: Die Dokumente auf EconStor dürfen zu eigenen wissenschaftlichen Zwecken und zum Privatgebrauch gespeichert und kopiert werden. Sie dürfen die Dokumente nicht für öffentliche oder kommerzielle Zwecke vervielfältigen, öffentlich ausstellen, öffentlich zugänglich machen, vertreiben oder anderweitig nutzen. Sofern die Verfasser die Dokumente unter Open-Content-Lizenzen (insbesondere CC-Lizenzen) zur Verfügung gestellt haben sollten, gelten abweichend von diesen Nutzungsbedingungen die in der dort genannten Lizenz gewährten Nutzungsrechte. Terms of use: Documents in EconStor may be saved and copied for your personal and scholarly purposes. You are not to copy documents for public or commercial purposes, to exhibit the documents publicly, to make them publicly available on the internet, or to distribute or otherwise use the documents in public. If the documents have been made available under an Open Content Licence (especially Creative Commons Licences), you may exercise further usage rights as specified in the indicated licence. https://creativecommons.org/licenses/by/4.0/ Copyright © 2019 The Authors. Published by VGTU Press. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.. VERSLAS: TEORIJA IR PRAKTIKA / BUSINESS: THEORY AND PRACTICE ISSN 1648-0627 / eISSN 1822-4202 http://btp.press.vgtu.lt 2019 20: 81–92 https://doi.org/10.3846/btp.2019.08 To link to this article: https://doi.org/10.3846/btp.2019.08 transactions engaged by the Indonesian listed firms (mean more than 60 percen relative to the total corporate actions). Studies on the impact of RPTs on firm performance still do not offer conclusive results. According to Gordon et al. (2004) RPTs enables firms to minimize transaction costs with other parties so it is called the efficient transaction hypothesis, or facilitates propping up, i.e., enhancing firm performance (Cheung et al. 2009). Tsai et al. (2015), Utama et al. (2010), and Utama and Utama (2014) empirically support this argument. On the contrary, the adverse effect of RPTs, or the conflict of interest hypothesis, enables majority shareholders to inflict a loss to minority shareholders (Gordon et al. 2004) or to facilitate tunneling, i.e., the expropriation of firm assets for their interests (Cheung et al. 2009). Jian and Wong (2003), Gordon et al. (2004), Cheung et al. (2009), Wahab etal. (2011), and Nekhili and Cherif THE EFFECT OF RELATED PARTY TRANSACTIONS ON FIRM PERFORMANCE: THE MODERATING ROLE OF POLITICAL CONNECTION IN INDONESIAN BANKING SUPATMI1, SUTRISNO T.2, Erwin SARASWATI3, Bambang PURNOMOSIDHI4 1 ,2, 3, 4Faculty of Economics and Business, Brawijaya University, MT. Haryono Street 165, Malang, 65145, Indonesia E-mails: [email protected] (corresponding author); [email protected]; [email protected]; [email protected] Received 18 December 2018; accepted 23 January 2019 Abstract. This study aims to examine the effect of related party transactions (RPTs) on banks’ performance and investigates political connections as moderator in their causal relationship. Our sample is 40 Indonesian banks listed on the Indonesian Stock Exchange for the years 2013–2016 with 160 observations as panel data. Based on panel data regression test, our results demonstrate that account receivables-related RPTs have a positive effect on banks’ profitability and its market performance (Tobin’s Q), but there are consequences of high operating costs and the risk of non-performing loans. Banks receive more funds from their related parties (account payables-related RPTs), banks exhibit higher capital capability and lower market performance. Further, the political connection index in banks significantly affect banks’ capability, liquidity, efficiency, and market value through RPTs. This result indicates that political connection strengthens the effects of RPTs on banks’ performance. Although this study has limited information in determining political connections and has not considered macroeconomic conditions, these findings imply that political connection plays an important role in banks’ performance in Indonesia. Keywords: related party transactions, political connection, firm performance, market performance, banks, Indonesia. JEL Classification: G21, L25, M41. Introduction Agency problem distorts reported earnings (Watts and Zimmerman 1986) and eventually leads to the discrepancy between accounting-based financial performance and market performance. One of the transactions that potentially imply high agency problem is related party transactions (RPTs). With the ownership structure dominated by family ownership (Claessens et al. 2000, Habib et al. 2017), Indonesian firms are more susceptible to expropriation by majority shareholders on minority shareholders’ interests through RPTs (Friedman et al. 2003). Firms from the same group, especially in Asia and more specifically in Indonesia (Utama et al. 2010), even regularly engage in these transactions (Wong et al. 2015). The Indonesian Stock Exchange corporate action data for the years 2012–2016 demonstrates that RPTs or affiliated transactions are the most frequent (2011) demonstrate the tendency of firms engaging in RPTs to expropriate minority shareholders’ interests and to manage earnings. However, Pozzoli and Venuti (2014), Magdalena and Dananjaya (2015), and Downs et al. (2016) find that RPTs do not affect firm performance. This research aims to find empirical evidence of the effect of RPTs on the banks performance in Indonesia. This study to confirm the previous studies on the effects of RPTs on firm performance. These previous studies mainly use non-financial firms as their sample. Thus, this study uses the Indonesian banking firms as the sample. We select banks as our sample for the following reasons. Firstly, although in general the banking industry is highly regulated, the majority shareholders of the Indonesian banks (including the government) frequently expropriate, thus indicating the low quality of the banks’ corporate governance (Surifah 2013). Secondly, banks play a significant role in the Indonesian economy. Indonesia itself is more a bank-centered economy than a market-centered one. Further, this study also aims to investigate the role of political connection in moderating the effect of RPTs on banks performance. Fisman (2001) argue that political connection is a valuable resource for many firms, so it can affect firms’ transactions or performance. When members of the board of executive directors, members of the board of non-executive directors, or shareholders of a firm are parts of the ruling government or the head of states, their firms will likely enjoy various benefits and facilities in engaging transactions (Faccio 2006). According to the resource dependence theory, political connection provides positive effects for firms, including banks. However, the agency theory argues that political connection is detrimental to banks. Indonesia has been applying the multi-party system since the collapse of the New Order era. Consequently, firms can maintain their political connection not only by being closely related to presidents or the families of presidents, but also their relationships with politicians from various political parties, including those who hold various positions in parliament, ministries, military service, local governments, together with their close families (Habib et al. 2017). Indonesia as an emerging market, also offers a unique research context for this issue because of its relatively weak investor protection (Leuz and Oberholzer-Gee 2006, Habib et al. 2017) that provides more significant opportunities for expropriation relative to countries with stronger investor protection (Wong et al. 2015). In addition to adding empirical evidence to agency theory in Indonesia, this study also contributes to providing a different perspective in measuring political connections. According to our knowledge there is still very limited research that measures the level of political connections, especially for Indonesian case. This study will measure political connections in the bank by considering the level of connectivity, and not only the existence or number of politically connected personnel in banks. For regulators, especially the Financial Accounting Standards and the Financial Services Authority, this research can be a consideration in determining banking governance policies, especially in terms of the composition of the company’s board. This research uses 40 Indonesian banks listed on the Indonesian Stock Exchange for the years 2013–2016 with 160 observations as panel data. Based on hypothesis testing using panel data regression, the research hypothesis is supported. However, this study has limited information in tracing political connected personnel in banks and ignoring macroeconomic conditions during the study period. This paper is organized as follows. The second part presents the theoretical analysis and develops the research hypotheses. Next, the third part discusses research methods used in this study. The fourth part offers and discusses the results. Lastly, the fifth part concludes and presents the research implications and the suggestion for future research. 1. Theoretical analysis and hypotheses development Related party transaction and firm performance This research aims to test the effect of RPTs on the Indonesian bank performance and the moderating effect of political connection in this relationship. According to Villalonga and Amit (2006), the agency conflict between principals and agents in Jensen and Meckling (1976) represents the type I agency conflict that is more pervasive in a more dispersed corporate ownership. Meanwhile, in many Asian countries such as Indonesia, the type II agency conflict between majority and minority shareholders is more prevalent. In this context, majority shareholders expropriate minority shareholders through various methods, such as RPTs. Claessens et al. (2000) find that family ownership dominates the ownership structure of Indonesian firms. Weak investor protection and law enforcement enable the majority shareholders of Indonesian firms to exploit RPTs to expropriate the minority shareholders (Habib etal. 2017). RPTs tend to be more conflict of interest than efficient transaction, or be more tunneling type transactions than propping up type. Many Indonesian firms, including banks, carry out RPTs (Utama et al. 2010). Firms can engage in RPTs that are related to assets, liabilities, sales, or expenses. Assets-related RPTs in banks often take the form of account receivable (credit allocation), while liabilities-related RPTs take the form of clients’ saving, either current accounts or savings accounts. Because banks’ RPTs are closely related to banks’ primary activities, the amount of RPTs affects financial and market performance. Thus, investors likely consider RPTs in their decision making (Jian and Wong 2003). Sari et al. (2016) demonstrate that relative to firms with dispersed ownership, Indonesian listed firms with 82 Supatmi et al. The effect of related party transactions on firm performance: the moderating role of political... concentrated ownership tend to engage in tunneling-type transactions. In a similar vein, using Indonesian listed firms as their sample, Utama et al. (2010), Utama and Utama (2014), Tambunan et al. (2017) indicate that RPTs negatively affect firm performance. Several studies find that both the number and the amount of RPTs negatively affect firm performance, such as Jian and Wong (2003) in China, Gordon et al. (2004) in the US, Lin et al. (2010) in Taiwan, and Wahab et al. (2011) in Malaysia. Thus, RPTs tend to be an expropriati on to minority shareholders. The findings confirm Cheung et al. (2009) who show that in the Chinese context, majority shareholders use RPTs more for the tunneling purpose to expropriate minority shareholders than for propping up. Based on the arguments, we propose the following hypothesis: H1: Related party transactions negatively affect banks’ performance. Related party transaction, political connection, and firm performance This study also predicts that political connection moderates the effect of RPTs on banks performance. The quality of corporate governance, especially the role of the board of directors (BOD), affects the extent of expropriation using RPTs (Pizzo 2013). However, according to Hillman (2005), studies on BOD mainly focus on the composition and characteristics of BOD and not on the linkage types of BOD, especially the political capital or political connection of the board. Political connection is a firm’s valuable resource (Fisman 2001) that affects its performance or transactions. When a firm’s executive directors, non-executive directors, or shareholders are parts of governments or heads of states, it is likely that the firm will enjoy various facilities and easiness in their transactions (Faccio 2006). This study uses two theoretical perspectives in analyzing political connection, i.e. the resource dependence theory and agency theory. Based on the resource dependence theory, the presence of politicians in BOD is a form of corporate political strategy to minimize the external environmental uncertainty of the firm, especially from the government, and eventually to improve firm performance (Hillman 2005). On the other hand, the agency theory argues that politically connected firms are risk takers (Ling et al. 2016), are more likely to experience business failure (Johnson and Milton 2003), and increase the blockholder rent appropriation (Sun et al. 2016). Engaging in RPTs will increase the incentives of BOD to take over (Gordon and Henry 2005) because RPTs potentially inflict a loss for shareholders if firms are better off when engaging non-RPTs than RPTs (Ryngaert and Thomas 2007). Several studies support both theories. According to Faccio (2006), a firm is politically connected if at least one of its top personnel (CEO, the head of board, president, vice president, or board secretary) or large shareholders (defined as those control at least 10% of the firm’s voting rights) is a head of state (president, king/ queen, or prime minister), a governmental minister, or a parliamentary member. A firm is also politically connected if a relative whose family name is the same as the head of state’s or a minister’s is in a top management position or a large shareholder. Lastly, a firm is also considered to have a political connection when one of its top executives or large shareholders has a “friendship” relationship with the head of state, ministers, or members of parliament. Firms directly owned by governments or states exhibit a direct political connection (Adhikari et al. 2006). Ang et al. (2013) and Bona-Sánchez et al. (2014) argue that the presence of ex-politicians in boards also indicates a political connection. Although ex-politicians are not currently involved in politics, they have already accumulated political experience and relationship that are beneficial for their firms. Modifying Faccio (2006) and some previous research to better fit with the Indonesian context, Habib et al. (2017) demonstrate that a firm is politically connected if at least one large shareholder (one who directly or indirectly controls at least 10% the firm’s voting rights) or board member is a current or an ex-member of parliament, a minister or local governmental chief, a senior military officer, or one who is closely related to politicians or a political party. They also extend the relationship with ministers into close relatives (spouses, children, or other family members). The previous literature shows that politically connected firms can either increase or decrease firm performance. Political connection affects the types and intensity of firms’ transactions and eventually affects firm performance, including banks. This study predicts that political connection is more pervasive and plays an increasingly important role for Indonesian firms, especially during the post-New Order era. Su et al. (2013) find that political connection positively affects firm investment. However, the interaction between political connection and RPTs negatively affects firm investment. These findings indicate that firms’ ownership structure plays a significant role in their activities. Liu et al. (2012) demonstrate that political connection in Chinese SOEs facilitates expropriation through RPTs for the interests of the governments and not the firms. Firms often use their political connection to gain special facilities in engaging in transactions, such as RPTs (Habib et al. 2017). Because Indonesia has weak investor protection regulations (Leuz and Oberholzer-Gee 2006, Habib et al. 2017) and many Indonesian firms, including banks, exhibit a “principal-principal conflicts” (Sun et al. 2016), the presence of political connection will moderate the effect of RPTs on banks performance. Based on these arguments, we propose the following hypothesis: H2: Political connection moderates the relationship between related party transactions and firm performance. Business: Theory and Practice, 2019, 20: 81–92 83 2. Research method This study investigates banking firms listed on the Indonesian Stock Exchange for the years of 2013–2016. We select banking firms as our research objects because the banking industry is a highly regulated one and this industry exhibits a considerable amount of expropriation by the controlling shareholders, even including the government that indicates the low quality of its corporate governance (Surifah 2013). Further, the Indonesian economy exhibits a bank-centered characteristic. By using the purposive sampling method, this study generates 160 observation firms for financial performance and 128 observation firms for market performance (information on the stock market price of several banks not be found). We use the secondary data of annual reports and information on the profiles of the sample firms’ board of executive directors and board of non-executive directors. We generate our data from www. idx.co.id and other related websites that contain information on the political connection of the board members. For the dependent variable (firm/bank perfomance/FP), we use Tobin’s Q (TOBINS) as the proxy of market performance and ROE (return on equity), NPL (non-performing loan), OER (operational efficiency ratio – a ratio between operating expenses and operating revenues), LDR (loan to deposit ratio), and CAR (capital adequacy ratio) as the proxies of financial performance. We use these ratios because they are commonly used to measure the health of banks. For RPTs as our independent variable, we use the RPTs classification of Cheung et al. (2009) that defines tunneling as the ratio between total account receivables and total assets (AR) and propping up as the ratio between total account payables and total liabilities (AP). We then subtract the AR RPT with the AP RPT and divide the difference with total equity to measure the level of RPT. We measure political connection (PC) as the moderating variable with the political connection score index (PCIDX) as the natural logarithm of the political connection score. We define political connection by adopting Habib et al. (2017). Political connection score is measured by summing the score of a bank’s presence of political connection based on the positions and status. In scoring the bank’s presence of a political connection, we modify Tao et al. (2017) to better fit with the Indonesian context. Based on the organizational structure of ministries and government institutions in Indonesia, career or position can be divided into seven levels. Therefore a politically connected board will be scored according to the position held in the government agency. The politically connected board is still in position or will be scored between 2 (lowest) to 8 (highest), while board which is a former government official will be scored between 1 (lowest) to 7 (highest). Zero score is given if the bank has no political connection. Appendix A show example of determining the index of political connections and political connection score. This study includes blockholder ownership, managerial ownership, firm size, and audit firm size as the control variables. We measure blockholder ownership (OWN) as the percentage of share ownership of individual or institution with the largest ownership and managerial ownership (MNJ) as the proportion of share ownership owned by managers to total shares. Firm size (SIZE) is measured by taking the natural logarithm of total assets, and audit firm size (AUD) is measured by developing a dummy variable that is equal to one if the audit firm is a Big four members and zero otherwise. This research uses panel data which is a combination of cross-section data and time series data so that hypothesis testing using panel data regression, and here is the regression model: We perform sensitivity tests by measuring political connection in (1) the number of politically connected senior personnel (PCTOT), and (2) the dummy variable (PCDUM) that is equal to one if a bank has a political connection and zero otherwise. 3. Results and discussion Table 1 indicates that during the years 2013–2016 the Indonesian banks engaged more account payables-related RPTs than the account receivables-related RPTs. This finding suggests that the Indonesian banks involved in RPTs more for propping up than for tunneling. Further, the result is in contrast with the Chinese evidence (Cheung et al. 2009). These results are in line with the data that demonstrate that our sample on average exhibit good performance, as indicated by the positive ROE and Tobin’s Q of more than one. These findings also imply that banks utilize their capital effectively and investors value these banks higher than their book values. Meanwhile, according to other performance indicators (NPL, OER, LDR, and CAR), our sample banks qualify as healthy banks according to the Bank of Indonesia’s regulation. In general, the Indonesian banks tend to be politically connected and to use Big 4 audit firms. The mean value of the largest share ownership (blockholder) is 56.6%, while the average managerial ownership is relatively low (3.12%). Most banks even do not exhibit any managerial ownership. From 110 firm-year observations, the most prevalent form of political connection is the governmental ownership, followed by the political connection of banks’ senior personnel with political parties, military, and ex-ministers. The average political connection score is 7.1, and the average political connection score index is 1.4. Meanwhile, on average the 84 Supatmi et al. The effect of related party transactions on firm performance: the moderating role of political... Indonesian banks have 1-2 politically connected senior personnel. These figures suggest that banks’ politically connected senior personnel hold very high positions as suggested by the average political connection score of 7.1 although on average banks only have 1-2 politically connected senior personnel. Further, Table 2 below displays the simple relationship between the existence of political connection in the Indonesian banks, RPT, and bank performance. Table 2 demonstrates the relationship between the political connection with various indicators of bank performance. Politically connected banks exhibit better ROE, OER, LDR, and Tobin’s Q than non-politically connected banks. These findings indicate that political connection is associated with higher profitability, operating cost efficiency, liquidity, and market value. However, non-politically connected firms exhibit better NPL and CAR than politically connected firms. Further, politically connected firms have higher non-performing loans and lower capital adequacy than non-politically connected banks. Finally, the difference between account receivables-related RPTs and account payables-related RPTs is higher for politically connected banks than non-politically connected banks. These findings support Habib et al. (2017) who show that political connection leads firms to engage in more RPTs. Table 3 demonstrates the results of the Pearson Correlation test for all research variables. The table indicates the positive correlation between account receivables-related RPTs with ROE and the political connection score index and the positive correlation between account-payables RPTs with CAR and the political connection score index. Further, Tobin’s Q is positively correlated with the number of politically connected personnel in firms. This research uses pool data which is a combination of cross-section data and time series data so that hypothesis testing using panel data regression with EViews program. The result of the parameter estimation test finds the hypothesis test using the random effect model (generalized least squares). Table 4 summarizes the hypothesis tests using panel data regression. Table 4 shows that account receivables-related RPTs positively affect ROE but the difference between account receivables-related RPTs and account payables-related RPsT (ARP) negatively affects ROE. The finding implies that higher credit distribution is associated with higher profitability but higher fund receipts will be countered it. The panel shows that account receivables-related RPTs (AR) negatively affect NPL and OER. It implies that there are consequences of high operating costs and the risk of non-performing loans on credit distribution granted to the related parties. Different results are found on Tobin’s Q where account receivables-related RPTs (AR) positively affect Tobin’s Q, but account payables-related RPTs (AP) and the difference between account receivables-related RPT and account payables-related RPT (ARP) negatively affect this performance. These results suggest that more credit allocation to Table 1. Descriptive statistics Research Variables Min Max Mean Std.Dev. AR 0.000 0.139 0.016 0.029 AP 0.001 0.886 0.079 0.118 ARP (4.289) 0.279 (0.364) 0.712 ROE (1.066) 0.394 0.078 0.065 NPL 0.000 8.900 2.066 1.508 OER 7.120 235.200 86.527 22.1394 LDR 5.150 140.720 84.210 15.620 CAR 8.020 87.490 19.977 8.154 SIZE 27.474 34.576 30.894 1.817 OWN 12.460 96.920 56.600 20.667 MNJ 0.000 72.070 3.108 11.474 TOBINS 0.014 4.227 1.374 0.911 PCIDX 0.000 3.332 1.441 1.170 PCTOT 0.000 6.000 1.513 1.542 Research Variable N Percentage AUDITOR Big 4 100 62.5% Non big 4 60 37.5% PCDUM Exist 110 68.8% Not Exist 50 31.3% Variable definitions: AR: account receivables-related RPTs divided by total assets; AP: account payables-related RPTs divided by total liabilities; ARP: the difference between AR and AP divided by total equity; ROE: return on equity; NPL: non-performing loans; OER: operational efficiency ratio (operating expenses divided by operating revenues); LDR: loan to deposit ratio; CAR: capital adequacy ratio; SIZE: firm size; AUD: audit firm size (dummy variable); OWN: the largest shareholder’s ownership; MJN: managerial ownership; PCIDX: political connection score index; PCTOT: the number of politically connected senior personnel in a firm; PCDUM: a dummy variable indicating whether a firm is politically connected (score one) and zero otherwise. Table 2. The mean value of banks’ performance, RPT, and political connection Average Performance (%) ROE NPL OER LDR CAR TOBIN’S Q PC 0.069 2.176 83.644 85.098 19.406 1.445 No PC 0.045 1.824 92.871 82.256 21.235 1.209 Average RPT (%) AR AP ARP PC 0.018 0.080 (0.406) No PC 0.012 0.079 (0.270) Variable definitions see Table 1. Business: Theory and Practice, 2019, 20: 81–92 85 Table 3. The Results of the Correlation Test AR AP ARAP SIZE AUD OWN MNJ ROE NPL OER LDR CAR PCIDX PCTOT PCDUM TOBINS AR Pearson Correlation 1 Sig. (2-tailed) AP Pearson Correlation .207** 1 Sig. (2-tailed) .009 ARAP Pearson Correlation .473** –.037 1 Sig. (2-tailed) .000 .643 SIZE Pearson Correlation .257** .022 .028 1 Sig. (2-tailed) .001 .786 .722 AUD Pearson Correlation .095 .040 .073 .599** 1 Sig. (2-tailed) .230 .619 .358 .000 OWN Pearson Correlation .102 .044 .143 .055 .054 1 Sig. (2-tailed) .198 .577 .072 .486 .494 MNJ Pearson Correlation –.049 –.125 .012 –.277** –.267** .015 1 Sig. (2-tailed) .536 .115 .881 .000 .001 .847 ROE Pearson Correlation .323** .005 .081 .483** .160*.101 –.109 1 Sig. (2-tailed) .000 .945 .306 .000 .044 .204 .169 NPL Pearson Correlation –.029 –.021 –.032 –.012 –.098 .155 –.161*–.167*1 Sig. (2-tailed) .713 .792 .685 .883 .219 .051 .041 .035 OER Pearson Correlation –.127 –.051 –.057 –.357** –.251** .024 .065 –.480** .350** 1 Sig. (2-tailed) .109 .525 .471 .000 .001 .764 .416 .000 .000 LDR Pearson Correlation .038 .029 .031 .210** .343** .259** –.500** .081 .111 –.152 1 Sig. (2-tailed) .633 .716 .695 .008 .000 .001 .000 .307 .163 .055 CAR Pearson Correlation –.012 .203*–.035 –.281** –.238** –.157*.036 –.195*–.245** .005 –.180*1 Sig. (2-tailed) .878 .010 .657 .000 .002 .047 .650 .013 .002 .945 .023 PCIDX Pearson Correlation .246** .156*.142 .566** .281** –.138 –.247** .320** .119 –.179*.053 –.037 1 Sig. (2-tailed) .002 .049 .073 .000 .000 .083 .002 .000 .135 .024 .506 .643 PCTOT Pearson Correlation .006 .056 .050 .382** .099 –.156*–.196*.258** .035 –.183*–.060 .137 .804** 1 Sig. (2-tailed) .935 .481 .528 .000 .214 .049 .013 .001 .660 .021 .450 .084 .000 PCDUM Pearson Correlation .095 –.007 .074 .491** .313** –.244** –.256** .227** .108 –.194*.085 –.104 .833** .663** 1 Sig. (2-tailed) .230 .925 .353 .000 .000 .002 .001 .004 .173 .014 .288 .189 .000 .000 TOBINS Pearson Correlation .129 –.022 .002 –.008 –.133 .014 .001 .340** –.138 –.097 .010 .167*.158 .332** .119 1 Sig. (2-tailed) .113 .785 .985 .918 .103 .862 .987 .000 .091 .237 .904 .040 .053 .000 .146 a. *, **, ***: significance level at 10, 5, and 1%, respectively. b. Variable definitions see Table 1. 86 Supatmi et al. The effect of related party transactions on firm performance: the moderating role of political... related parties is associated with higher market performance. However, when banks receive more funds from their related parties, or when the difference between the fund distribution to and the fund receipts from related parties is higher, banks exhibit lower market performance.The result supports first hypothesis for account payables-related RPTs but not for account receivables-related RPTs. The above findings demonstrate a clear relationship between account payables and receivables, interest expense and revenue, and their impacts on firm profitability, the risk of non-performing loans, operating expense, and market performance. Higher account receivables-related RPTs are associated with higher interest revenue, implying higher net earnings, and market performance. However, when banks engage in more account payables-related RPTs the bank’s capability increases but they incur more interest expense that eventually causes the risk of non-performing loans, net earnings, and market performance to decline. These findings indicate that RPTs do not have different effects as the non-RPTs do. It is less likely that majority shareholders expropriate minority shareholders’ wealth. Our results are different from Jian and Wong (2003) and Cheung et al. (2009) for the Chinese case, and Gordon et al. (2004) in the US which show that majority shareholder exploit RPTs to expropriate minority shareholders that eventually harms firm performance. We explain this difference by pointing to the countries’ different economic characteristics where Indonesia is more bank-centered while these other countries are more capital market-centered. Besides, the different categories of related parties may explain why our results differ from other studies. The data suggest that the account-receivables RPTs of the Indonesian bank are mostly corporate credits and not the personal ones. In a similar vein, the account payables-related RPTs of the Indonesian banks are mostly institutional savings. Different from individuals-related RPTs such as RPTs with executive and non-executive directors, institutions-related RPTs are less likely to cause banks to treat the transactions differently. Most banks in our sample disclose that they treat their account receivables-related RPTs and account-payables RPTs similarly to non-RPTs, especially regarding the interest rate. Account receivables-related RPTs increase banks’ earnings and market performance, thus indicating that the related institutions utilize credits granted to them for good business purposes. Further, account receivables-related RPTs that mostly take the form of credit allocation to related parties is considered productive. In 2014–2016, Indonesia was under the Joko Widodo administration that focus on the infrastructure development. This program requires a huge amount of funds; leading banks, especially the state-owned ones, to financially support this program. These findings support the efficient transaction hypothesis for the account receivables-related RPTs and the conflict of interest hypothesis for the account payables’ related RPTs (Gordon et al. 2004). However, the results contradict Cheung et al. (2009) who demonstrate that Indonesian banks enjoy the benefits of account receivablesrelated RPTs (propping up) while account payables-related RPTs are a tunneling type. Table 4 also suggests that the political connection index moderates the effect of account receivables-related RPTs (AR*PCIDX) on bank performance as measured by OER and Tobins’ Q. The results imply that the political connection index moderates the effects of RPTs on banks’ operating Table 4. The results of the hypothesis testing Research Variables ROE NPL OER LDR CAR TOBINS’Q Constant –0.824 5.591 237.992*** 99.019*** 57.408*** –0.126 AR 1.954** –18.506* –406.359** –61.384 –20.071 29.379*** AP –0.032 –0.914 –18.841 –1.264 8.518 –14.131*** ARP –0.066 0.095 17.423* –6.286 4.389 –2.608*** PCIDX –0.012 0.248 –2.676 0.535 –0.622 0.148 SIZE 0.029*** –0.137 –4.631 –0.844 –1.091* 0.046 AUD 0.011 –0.103 –2.242 5.097 –1.021 –0.416 OWN –0.001 0.012 0.044 0.159** –0.052 0.0001 MNJ 0.001 –0.020 –0.162 –0.628*** –0.028 0.004 AR*PCDIX –0.539 6.738 152.883** 60.519 –4.374 –12.209*** AP*PCIDX 0.092 –1.254 –1.568 –28.701*** 16.899*** 6.221*** ARP*PCIDX 0.028 –0.280 –7.209* –1.641 0.160 1.106*** Adj R20.029 0.011 0.059 0.201 0.113 0.054 F 1.446 1.147 1.900** 4.633*** 2.832*** 1.663* *, **, ***: significance level at 10, 5, and 1%, respectively. Variable definitions see Table 1. Business: Theory and Practice, 2019, 20: 81–92 87 expense efficiency and bank’s market performance. The interaction between political connection and account receivables-related RPTs positively affects OER but negatively affets Tobins’ Q. It means that more banks are conducting account receivables-related RPTs and more have political connections, the higher the banks’ operating expense efficiency and Table 5. The results of sensitivity tests Panel A: Political connection measured with PCTOT Research Variables ROE NPL OER LDR CAR TOBINS’Q Constant –0.592* 3.914 198.387*** 97.293*** 58.514*** 0.565 AR 1.869* –20.458* –376.133** –56.227 –13.051 19.485*** AP –0.025 –0.929 –22.671 –3.573 10.312 –6.269 ARP –0.101* –0.253 10.783 –7.667 8.420*** –1.211*** PCTOT 0.010 0.081 –4.228** 0.265 –0.419 0.099 SIZE 0.020* –0.075 –3.317* –0.740 –1.175** 0.024 AUD 0.010 0.010 –2.611 4.636 –0.358 –0.343 OWN –0.001 –0.0978 0.052 0.139** –0.032 0.001 MNJ 0.001 –0.019 –0.137 –0.629*** –0.026 0.003 AR*PCTOT –0.619 10.331 176.015** 58.436 2.916 –9.121*** A*PPCTOT 0.049 –0.798 4.801 –15.254** 6.569* 2.706*** ARP*PCTOT 0.048* –0.108* –5.165 0.854 –3.505*** 0.528*** Adj R20.033 0.002 0.055 0.197 0.169 0.020 F 1.486 0.906 1.843* 4.557*** 3.950*** 1.237 Panel B: Political connection measured with PCDUM PCDUM (1) Research Variables ROE NPL OER LDR CAR TOBINS (N = 150) (N = 150) (N = 150) (N = 150) (N = 150) (N = 96) Constant –0.850* 6.255 229.324*** 100.862** 44.678** 2.897 AR 0.504 –3.614 –27.191 122.275** –49.000 –1.019 AP 0.179 –0.474 14.816 –105.747*** 61.951*** 0.729 ARP –0.001 –0.182 5.729 –14.411*** 5.967** 0.018 SIZE 0.028* –0.149 –4.616** –0.760 –0.722 –0.039 AUD 0.019 0.391 4.328 4.478 –1.630 –0.436 OWN –0.000 0.008 –0.0197 0.134 –0.047 0.002 MNJ 0.009 –0.050 –0.355 –1.560* –0.446 0.007 PCDUM (0) Research Variables ROE NPL OER LDR CAR TOBINS (N = 50) (N = 50) (N = 50) (N = 50) (N = 50) (N = 32) Constant –0.295 –2.800 189.356** 107.275* 66.494 2.751 AR 0.822 –10.944 –158.592 –87.531 –43.031 22.349* AP –0.062 –0.765 –19.551 4.162 7.646 0.234 ARP –0.017 0.181 1.358 2.579 4.303 –0.284 SIZE 0.012 0.176 –3.511 –1.104 –1.015 –0.057 AUD 0.012 –1.275* –9.519 4.363 –2.126 –0.170 OWN –0.001 0.005 0.252 0.193 –0.203* –0.001 MNJ 0.000 –0.021 –0.149 –0.643*** –0.027 –0.008 a. *, **, ***: significance level at 10, 5, and 1%, respectively. b. Variable definitions see Table 1. 88 Supatmi et al. The effect of related party transactions on firm performance: the moderating role of political...