The effect of institutional blockholders' short-termism on firm innovation: Evidence from the Korean market

Pacific-Basin Finance Journal 57 (2019) 101188

Contents lists available at ScienceDirect

Pacific-Basin Finance Journal journal homepage: www.elsevier.com/locate/pacfin

The effect of institutional blockholders' short-termism on firm innovation: Evidence from the Korean market Sanggyu Kanga,b, Chune Young Chunga, a b

⁎,1

T

, Dong-Soon Kima

School of Business Administration, Chung-Ang University, Seoul, Republic of Korea Pricewaterhousecoopers International Ltd., Republic of Korea

A R T IC LE I N F O

ABS TRA CT

Keywords: Firm innovation R&D Institutional blockholding Corporate governance Chaebol Emerging market

This study examines how the behavior of institutional blockholders affects investment in research and development (R&D) in Korean firms. Contrary to the monitoring view that institutional investors promote firms' R&D, the results of this study indicate that institutional blockholders have a significantly negative influence on R&D investment. More importantly, when we decompose institutional blockholdings according to institutions' national origins and investment horizons, we find that firms with higher foreign short-term blockholdings spend significantly less on R&D than other firms do. These results show that when the local characteristics of the Korean market (i.e., a predominance of owner-manager firms and weak corporate governance) prevail, institutional monitoring is not effective, and, thus, institutional blockholders tend to focus on the short term. Overall, in the Korean market, stronger short-termism hinders institutional monitoring and negatively affects firms' R&D investments.

JEL classification: F21 G32 G34 O31 O32 O33

1. Introduction Schumpeter's theory of creative destruction as “the essential fact about capitalism” (Schumpeter, 1942) remains valid. Innovation determines long-term economic growth at the country level (Solow, 1957), is the main engine of growth at the firm level (Aghion et al., 2013), and is an essential element of a firm's competitive advantage (Helfat, 2000) and sustainability (Honoré et al., 2015). Thus, decisions on investment in research and development (R&D), the primary source of innovation, are important for firms to remain innovative and sustainable. However, such decisions are complicated and risky owing to the high level of uncertainty they entail (Chen et al., 2016; Rong et al., 2017). The literature on institutional monitoring shows a positive link between institutional investors and R&D investments and innovation. For example, Aghion et al. (2013) show that when incentive contracts do not fully motivate managers to overcome the risk of being fired owing to the failure of an R&D project, active monitoring by large outside shareholders plays a role in insulating managers against early failure. Bushee (1998) documents that when institutional ownership is high, managers are less likely to reduce investments in R&D, suggesting that institutional monitoring decreases managerial myopia. At the same time, the internationally diversified portfolios (Luong et al., 2017) and better access to foreign capital markets (Kwon and Park, 2018) of foreign

⁎

Corresponding author. E-mail address: bizfi[email protected] (C.Y. Chung). 1 The standard disclaimer rules apply. All remaining errors are our own. https://doi.org/10.1016/j.pacfin.2019.101188 Received 29 October 2018; Received in revised form 28 June 2019; Accepted 6 August 2019 Available online 07 August 2019 0927-538X/ © 2019 Elsevier B.V. All rights reserved.

Pacific-Basin Finance Journal 57 (2019) 101188

S. Kang, et al.

institutional blockholders protect firms against the failure of R&D projects, encouraging managers to invest in R&D. Foreign blockholders, as resource and knowledge providers, encourage R&D investments in resource-poor firms by providing advanced technological resources and boosting R&D activities (Choi et al., 2011). However, few studies examine this relationship within the context of corporate governance and heterogeneous institutional blockholders. Under different governance environments and monitoring roles of institutional investors, the results of previous studies may vary. First, poor corporate governance may decrease the incentive to invest in and monitor a firm. This is because institutional blockholders have fewer chances to intervene directly in management when owner-managers have absolute power over corporate decisions. According to Edmans (2014) and Edmans and Holderness (2017), direct interventions by blockholders are difficult to implement in some cases, such as when controlling shareholders use corporate resources to support their preferred directors in a proxy fight or to reduce the incentives for small blockholdings to monitor the firm. Furthermore, even when these incentives are sufficient, blockholders' small shareholdings reduce their voting power. In summary, when institutional investors find it difficult to monitor firms through direct intervention and run the risk of expropriation by controlling shareholders, they may sell their holdings. This tendency can be stronger for foreign blockholders owing to their information disadvantages. Institutional blockholders prefer firms with good corporate governance, and this preference originates from three sources. First, in a cost–benefit framework, it is beneficial for institutional blockholders to invest in firms with good governance quality. The costs of monitoring firms with weak governance can be greater than the benefits. Firms with weak corporate governance are less likely to provide disclosure and transparency. This opaque information environment increases the costs of information acquisition. Second, in a risk–return framework, investments in firms with weak governance quality do not guarantee a fair return. Insiders gain abnormal returns, and institutional blockholders run the risk of expropriation and self-dealing problems. Thus, the expected return on investments in firms with weak governance decreases. Third, institutional blockholders prefer firms with high market liquidity because they cannot sell their large shareholdings without price pressure under low market liquidity. Firms with good corporate governance show higher market liquidity and narrower spreads. The preference for better corporate governance is stronger for transient institutional blockholders because they trade frequently and have higher portfolio turnovers. In summary, corporate governance is an important factor when institutional blockholders invest in firms. Second, institutional investors show heterogeneous investment objectives and strategies, resulting in different investment horizons and monitoring roles; that is, not all institutions actively monitor the firms they hold. Long-term institutions are likely to monitor actively owing to their long-term investment strategies, whereas short-term institutions are less likely to monitor firms because of their own short-term concerns. Furthermore, only long-term institutions benefit sufficiently to cover the cost of monitoring (Chen et al., 2007). Thus, short-term blockholders do not act as monitoring institutions and are reluctant to invest in long-term projects such as R&D, suggesting short-termism. To examine the monitoring incentives of heterogeneous institutional blockholders on R&D investments under different strengths of corporate governance, we focus on the Korean market. The Korean market provides a unique research environment to explore this relationship. Divergent governance environments in the Korean market stem from chaebols (the Korean term for conglomerates controlled by an owner or family members). Historically, large firms have been provided with governmental benefits and subsidies to boost the economy in Korea. These firms, typically owned and controlled by a founder and his or her family, have involved into chaebols. Chaebols have used the pyramid ownership structure and cross-shareholdings to have the absolute power over firm and voting rights in excess of their cash flow rights. Under the chaebol system, the ownership has been heavily concentrated to the family owner-manager and separated from the control. Because chaebols have been influential in the country's economy as well as on the ownership structure and governance of Korean firms, the concentrated ownership and the disparity between ownership and control were pervasive, leading to the nationwide weak corporate governance in Korea. The distinctive ownership structure and conflicts of interests between the family owner-manager and minority shareholders exacerbates the agency problems in Korea (Hwang et al., 2013). Previous studies show that minority shareholders, in contrast to large shareholders, are not well protected in the Korean market and, thus, suffer from an agency problem (e.g., Bae et al., 2002; Choi et al., 2018). When the separation of ownership and control is severe and the controlling shareholders are secured, the expropriation is more likely to happen (Joh, 2003). Chaebols are prone to engaging in self-interested behavior at the expense of the interests of outside shareholders (Choi et al., 2014). Byun et al. (2011), also, find that the information asymmetry is positively associated with ownership concentration. Thus, a better corporate governance mechanism could be more important for corporate investment decisions in Korean firms. In addition, chaebols and their affiliates account for substantial portions of both financial and nonfinancial corporations in Korea. Their great influence engenders extensive ties with regulatory and non-regulatory monitoring institutions in Korea, rendering these institutions pressure-sensitive to managerial actions (Liu et al., 2018). According to Gillan and Starks (1998), pressure-sensitive institutions are likely to go along (or collude) with management to preserve current or potential business relationships. Thus, they often seek to avoid taking explicit actions against incumbent executives even when they are dissatisfied with inferior managerial decisions. In contrast, pressure-insensitive institutions have limited business relationships with firms and, thus, are more likely to exert monitoring efforts. This setting stands in stark contrast to the case of developed markets, such as the United States, where pervasive evidence of active monitoring has been found (McCahery et al., 2016). R&D and innovation are critical factors that have led to Korea's international competitive advantage and recent significant economic growth. Chaebols such as Samsung, LG, and Hyundai invested in R&D, and their cutting-edge technology and innovation became the main drivers of the Korean economy. Thus, examining R&D activities in the Korean market is an important research topic at both the country and the firm level. For practitioners and regulators in Korea, understanding which institutional investors promote R&D investments and the corporate governance environment in which these investors operate is worthwhile. 2

Pacific-Basin Finance Journal 57 (2019) 101188

S. Kang, et al.

We find that institutional blockholdings are negatively related to R&D intensity. Furthermore, our results show that short-term blockholders' ownership is significantly and negatively related to R&D intensity but that long-term blockholders show no significant relation. More importantly, firms with a higher proportion of foreign short-term blockholders spend less on R&D than those with domestic short-term blockholders do. These results support the argument that institutional blockholders who invest in the Korean market show short-termism and that this tendency is stronger when blockholders are short-term and foreign than when they are longterm and domestic. Blockholders' short-termism fosters managerial myopia and results in decreased R&D investment, which is consistent with the findings of Bushee (1998, 2001). This study contributes to the existing literature in several ways. First, our findings add to the literature on R&D and innovation by suggesting a negative influence of institutional blockholders on R&D investment. Ownership by institutional (foreign) investors in developed countries (emerging markets) is increasing, and institutional investors are emerging as major players in the financial market. As such, a large body of literature examines the influence of institutional (foreign) investors on R&D and innovation. However, previous studies do not consider the heterogeneity of institutional blockholders. Decomposing institutional blockholders according to investment horizon and national origin complements these studies by showing that heterogeneous institutional blockholders play significantly different roles in firms' R&D investment behavior in various corporate governance environments. In addition, the effects of blockholders on R&D have been studied primarily in the context of advanced countries (e.g., the United States) or transitioning economies (e.g., China). However, the behavior of blockholders might differ in emerging markets owing to the characteristics of such markets, although little evidence exists to test this hypothesis. This study provides an extended understanding of the relationship between blockholders and R&D investments by considering the local characteristics of the Korean market. The remainder of the paper is organized as follows. In section 2, we review the related literature and develop the main hypotheses. Section 3 describes the overall sample and presents the descriptive statistics for the main variables. The main empirical results are discussed in Section 4. Section 5 concludes the paper. 2. Related literature and hypothesis development 2.1. Institutional blockholders and R&D intensity A stream of literature describes the positive influence of monitoring by institutional investors on firms' R&D activities and innovation performance. For instance, both Eng and Shackell (2001) and Wahal and McConnell (2000) find a positive relation between institutional investor ownership and R&D investment. Using patent data, Aghion et al. (2013) support the career concern model, where institutional monitoring insulates managers from the risk of harming their reputations and being fired for bad earnings. They suggest that this insulation encourages managers to innovate. Whereas Aghion et al. (2013) describe institutional investors as homogeneous, Kim et al. (2017) consider the heterogeneity of institutions and suggest that long-term institutional monitoring encourages firms' innovation activities. Focusing on Chinese firms, Choi et al. (2011) and Rong et al. (2017) find a positive influence of institutional ownership on firms' innovation. For the Korean market, the prior literature finds mixed results. Kim et al. (2008) find that both domestic and foreign institutional ownership negatively moderate the relation between financial slack and R&D. Using patent data for the period 2000–2003, Choi et al. (2012a) find that both institutional ownership and foreign institutional ownership positively influence technological innovation performance. However, examining the period 1999–2008, Yoo and Rhee (2013) find that foreign investors have a positive relation with R&D investment, whereas institutional ownership has a negative relation. These studies do not identify which institutional investors have positive or negative effects on R&D investments or the monitoring channel through which they exert their influence. Thus, to identify this channel, we decompose total blockholdings into two components: active monitoring blockholdings and nonmonitoring blockholdings. However, before doing so, we examine the factors that may affect institutional blockholders' investment decisions related to Korean firms, namely, the local characteristics related to corporate governance. Corporate governance, along with the traditional proxies for risk, is one of the main factors that investors consider when making investment decisions (Giannetti and Simonov, 2006). Li et al. (2006) show that countries with strong macro governance characteristics (i.e., strong shareholder rights, effective legal enforcement, and extensive financial disclosure) have more prevalent institutional blockholdings. This result is consistent with the findings of Ferreira and Matos (2008), who show that institutional investors invest less in firms with weak corporate governance (i.e., firms largely held by insiders) and prefer firms from countries with good disclosure standards. Similarly, Chung and Zhang (2011) find that institutional ownership has a statistically and economically positive relation with corporate governance quality. Firms with stronger governance attract more institutional investors, and the percentages of shares owned by institutions are larger. This property justifies the higher blockholdings found in firms with good governance characteristics. Understanding the rationale for the preferences of institutional investors for good corporate governance provides meaningful clues to how institutional blockholders affect firms' R&D activities. First, from the point of view of monitoring benefits and costs, it is optimal for institutional blockholders to invest more in firms with good governance quality. Institutional investors monitor firms only when the monitoring benefits outweigh the costs; thus, not all institutional investors monitor (Chen et al., 2007). Bushee and Noe (2000) find that firms with higher disclosure rankings have greater institutional shareholdings and suggest that the rationale for this finding is that higher disclosure reduces the costs of monitoring firm performance. In a similar vein, Li et al. (2006) state that reduced monitoring costs in countries with strong macro governance environments attract institutional investors and increase the propensity of institutional monitoring. Firm transparency nurtured by good corporate governance reduces the costs of information acquisition, which enhances institutional monitoring (Boone and White, 2015). 3

Pacific-Basin Finance Journal 57 (2019) 101188

S. Kang, et al.

In addition, the preferences need to be understood within a risk–return framework. The extraction of private benefits by corporate insiders is one of main sources of distortion in corporate investments (Giannetti and Simonov, 2006). Governance problems and expropriation by insiders prevent institutions from investing in firms. Insiders are expected to have informational advantages because they have access to private information, which they can exploit to gain abnormal returns. However, outsiders, who do not have informational advantages, gain normal returns and only enjoy security benefits on a pro rata basis (Giannetti and Simonov, 2006). The risk of expropriation and self-dealing problems in firms with weak governance means that investors expect different rates of return (Chung and Zhang, 2011). Institutions investing in firms with weak corporate governance do not earn a fair return relative to their risk (Gompers et al., 2003), but insiders do. Hence, the expected return depends on firms' governance quality (Giannetti and Simonov, 2006). Stock liquidity is another important factor considered by institutional investors when selecting a portfolio. Institutional investors selling their shares impacts the price of the shares. Hence, they cannot sell their stakes without price pressure, and this pressure is stronger in the case of large shareholdings. Corporate governance affects stock liquidity. Chung et al. (2010) find that firms with better corporate governance show higher stock market liquidity and narrower spreads. They suggest that the reduced information asymmetry between insiders and outsiders as a result of good corporate governance reduces the probability of information-based trading and spreads. Because transient institutional investors trade aggressively and have high portfolio turnovers based on their short-term trading strategies, liquidity is especially important to such institutions. Bushee and Noe (2000) suggest that transient institutions invest more in firms with better disclosure practices because such practices reduce the price impact of trades, which lowers trading costs and makes it easier to exit and realize a gain. Considering the size of the block, institutional blockholders are likelier to prefer firms with high market liquidity. Firms in emerging markets are often characterized as having heavily concentrated ownership (La Porta et al., 1999; La Porta et al., 1998), with the expropriation of minority shareholders by large shareholders (Dharwadkar et al., 2000) and weak corporate disclosure and financial opacity (Fan et al., 2011). Many firms in Korea are controlled by founders and/or their families, who have considerable power over their firms despite their smaller shareholdings. The disparity between ownership and control is a stylized fact in Korean firms (Kim et al., 2010), exacerbating agency problems, such as expropriation and self-dealing (Choi et al., 2014). In addition, board members who are assigned based on the recommendations of controlling shareholders represent their interests, which tend not to be aligned with those of outside shareholders (Choi et al., 2012b). In these cases, active monitoring by institutional blockholders is not effective. The presence of owner-managers with absolute power may deter direct intervention with management and, even when intervention is possible, the monitoring costs may be excessive. Monitoring blockholders do not have opportunities to influence firms' decisions owing to the predominance of ownermanagers. In addition, weak investor protection and poor corporate disclosure in emerging markets decrease the incentives for institutional investors to invest in and monitor such firms. Hence, institutional blockholders are less likely to invest and are likelier to sell their shares than to intervene directly in a firm's management. Thus, we argue that institutional blockholders that invest in Korean firms tend to have a short-term strategy (short-termism) rather than a long-term focus. Hypothesis 1. Institutional blockholdings are negatively related to the R&D intensity of Korean firms. 2.2. Foreign institutional blockholders and R&D intensity Although foreign institutional investors share several characteristics with institutional investors, they contrast with domestic institutional investors in that they have internationally diversified portfolios (Luong et al., 2017) and better access to capital markets in other countries (Kwon and Park, 2018). Thus, foreign institutional blockholders are less vulnerable to failures of R&D projects in the firms in which they have invested. According to Aghion et al. (2013), this strength related to risk tolerance protects managers from potential career and reputation risks arising from failures of R&D investments, which stimulates R&D investment. Foreign institutional blockholders act as a cushion against possible failures of R&D projects. On the one hand, in emerging markets, foreign institutional blockholders provide the advanced resources and knowledge necessary for R&D and innovation. Chen et al. (2014) state that foreign firms have both codified technological knowledge, such as patents and research reports, and tacit knowledge obtained from global markets and experience associated with R&D. Global networks of foreign institutions serve as a bridge linking domestic and foreign firms, enabling them to exchange knowledge and experience (Luong et al., 2017). In addition, cross-border mergers and acquisitions (M&As) promoted by foreign institutions enhance knowledge spillovers across countries, which facilitate R&D and innovation activities (Luong et al., 2017). Guadalupe et al. (2012) find improvements in innovation and labor productivity after foreign acquisitions of domestic firms. Collectively, foreign institutional blockholders facilitate knowledge spillovers by implanting advanced knowledge and promoting global business networks and crossborder M&As, which enhance firms' R&D activities. On the other hand, foreign institutional blockholders do not always improve R&D activities. The transfer of knowledge from foreign multinational enterprises (MNEs) may not encourage subsidiaries to invest in R&D (Un and Cuervo-Cazurra, 2008). Technologies developed by a parent firm and its subsidiaries decrease the need for R&D investments by other subsidiaries. They find that subsidiaries of foreign MNEs invest less in external R&D, suggesting that the transfer of knowledge and technology from the parent firm are a substitute for external R&D. Kwon and Park (2018) find that R&D intensity is negatively related to firms with > 50% ownership by a parent firm as well as to firms whose businesses are tied to that of a domestic parent firm. Moreover, foreign institutional blockholders that focus on financial interests and are better suited to stock-picking are not interested in long-term investments such as R&D, and instead pursue short-term interests. These facts collectively imply that the R&D activities of foreign4

Pacific-Basin Finance Journal 57 (2019) 101188

S. Kang, et al.

owned firms differ notably from those of domestically owned firms. At the same time, the preference for firms with good corporate governance is stronger for foreign institutional investors. Prior studies state that foreign institutions avoid investing in firms with poor corporate governance. Foreign institutional investors have smaller shareholdings in firms that have a high ratio of control to the cash flow rights of the principal shareholders (Giannetti and Simonov, 2006), are domiciled in countries with poor investor protection and disclosure practices (Leuz et al., 2009), and have high control–ownership disparity in Korea (Lee and Cho, 2016). Miletkov et al. (2014) find that the positive relation between board independence and foreign shareholdings is stronger in countries with poor investor protections. In addition, foreign institutional investors are likelier to be at risk of being exploited by insiders or other local investors owing to information disadvantages. Equity home bias has long been recognized in the capital market. Information asymmetry, or the higher cost of information, is one of several factors contributing to this phenomenon. The informational disadvantages of foreign institutional investors (Choe et al., 2005; Kang and Stulz, 1997) weaken their ability to predict the future stock returns of firms (Baik et al., 2010). Ferreira et al. (2017) find evidence that when information asymmetry is likely to be higher, such as in more opaque countries or countries with less efficient markets, local institutional investors do possess informational advantages. Similarly, in the Korean market, foreign institutional investors do not have superior information to local investors around earnings announcements (Park et al., 2014). Within the risk–return framework, foreign investors do not expect to gain a fair return on the prices traded by local investors because of the risk posed by information asymmetry and, thus, invest less. Even if they invest in foreign stocks, from the point of view of the benefits and costs of monitoring, greater effort is needed to monitor poorly governed firms than to monitor well-governed firms, thus incurring higher monitoring costs. This property implies that foreign investors incur higher monitoring costs than do locals, who can better identify problems related to a poor governance structure. In summary, foreign institutional investors are reluctant to invest in weakly governed firms, and their reluctance is greater than that of local investors owing to their informational disadvantages. Hypothesis 2. The negative effect of institutional blockholdings on R&D intensity is stronger among foreign institutional blockholders than among domestic institutional blockholders. 2.3. Heterogeneous institutional blockholders and R&D intensity The investment horizon of institutional blockholders is one of the identifiers that distinguish monitoring blockholders from nonmonitoring blockholders. Varying investment objectives, strategies, and horizons result in institutional investors having different incentives to interfere in a firm's management. Institutional monitoring is well-grounded in large ownership positions and a long-term horizon. The large shareholdings of blockholders limit their ability to trade, which promotes their monitoring of the firm. In addition, their long-term investment horizon encourages them to restrain managers' short-term focus and instead promote long-term value. Thus, R&D investment decisions are affected by the extent to which institutional blockholders engage in the management of a firm. The final hypothesis extends to the finding that long-term institutional blockholders have a greater incentive to monitor firms than short-term institutional blockholders do, taking into account the short-termism of foreign blockholders. Presumably, if institutional blockholders that invest in Korean firms show short-termism, regardless of their national origin, the presence of long-term institutional blockholders that can potentially monitor firms may not help discipline managers in terms of investing in R&D. Similarly, short-term institutional blockholders who pursue short-term trading rather than long-term investing are likelier to have a short-term focus. In other words, the monitoring incentives of institutional blockholders are hindered by short-termism. Considering that foreign institutional blockholders are likelier to have a short-term view, it is reasonable to expect that foreign short-term blockholders will be more reluctant to invest in R&D than other types of investors will. Hypothesis 3. The negative effect of institutional blockholding on R&D intensity is stronger among foreign short-term blockholders than among other types of investors. 3. Sample The sample data consist of all firms listed on the Korea Stock Exchange. The financial and accounting data are taken from DataGuidePro, provided by FnGuide. FnGuide is a leading provider of financial data in South Korea. We obtain institutional ownership data for all common stocks from FnOwnership, also offered by FnGuide. In Korea, the Financial Supervisory Service mandates that all investors that have more than a 5% share of a firm report their equity positions. Following Fama and French (2006), all financial firms are excluded from the sample data set because financial firms have much higher leverage than nonfinancial firms do. We collect data on common stocks for each fiscal year from March 2005 to September 2015. The final sample comprises 11,535 firmyear observations representing 1513 distinct firms. Each firm in the sample is classified into one of 16 industries, including R&Dintensive industries such as biotech and pharmaceuticals, according to the Korean Standard Statistic Classification. The dependent variable employed in the tests is R&D intensity (hereafter, “RNDS”), following Eng and Shackell (2001) and Choi et al. (2012a). RNDS is measured as annual R&D expenditure divided by sales. We use annual RNDS as a proxy for R&D activity. Following the R&D and innovation literature, RNDS is assigned a value of zero when firms do not report R&D expenditures. Although spending on R&D does not directly reflect innovation (Dalziel et al., 2011), it is required for firms to innovate. Managing R&D allocation and spending is a key element of success, especially for firms in R&D-intensive industries. Furthermore, we investigate how ownership structures influence R&D activities and spending by examining how institutional blockholders introduce various R&D 5

Pacific-Basin Finance Journal 57 (2019) 101188

S. Kang, et al.

activities to the firm. Accordingly, we focus on understanding the levels of R&D expenditures across heterogeneous institutional blockholders. Institutional blockholders are defined as those who own 5% or more of the shares outstanding in a firm, following Chen et al. (2007). This definition is generally accepted by the existing empirical literature (Edmans, 2014). It is typically considered that owning at least 5% of a firm's shares is significant enough to incentivize the monitoring of the firm. However, as the block size increases above 5%, theoretical models show that monitoring increases continuously rather than discontinuously (Edmans, 2014). Although the 5% blockholder definition has been criticized, we use it here to facilitate comparisons with the findings in the literature. To calculate institutional blockholders' ownership (BLOCK_IO), we divide all shares owned by institutional blockholders by the total number of shares outstanding in a firm. Foreign institutional blockholders might have different effects on R&D activities from those of domestic institutional blockholders. Thus, we decompose BLOCK_IO into domestic institutional blockholdings (BLOCK_IO_D) and foreign institutional blockholdings (BLOCK_IO_F). Furthermore, we consider the heterogeneity of institutional investors based on their investment horizons. Institutional investors show heterogeneous investment behavior originating from differences in their investment strategies and objectives, which lead to varying effects on and/or incentives to monitor the firms in which they invest. Specifically, to differentiate ownership by long-term and short-term institutional blockholders from that of total institutional blockholders, we use investor portfolio turnover, following Yan and Zhang (2009). For quarter t, each institution's portfolio turnover is measured using institution k's churn rate, as follows:

CRk, t =

min (Buyk, t , Sellk, t ) Nk Sk , i, t Pi, t + Sk , i, t − 1 Pi, t − 1 ∑i = 1 2

,

where Buyk, t and Sellk, t are the aggregate purchases and sales, respectively, by investor k in quarter t; Pi, t−1 and Pi, t are the share prices for stock i at the end of quarters t − 1 and t, respectively; and Sk, i, t−1 and Sk, i, t are the numbers of shares of stock i held by investor k at the end of quarters t − 1 and t, respectively. Then, we average each institution's churn rate over four quarters, as follows:

AVG _CRk, t =

1 4

3

∑ CRk,t−j . j=0

All institutional blockholders are sorted into three terciles based on their average churn rate, AVG_CR. Blockholders ranked in the top tercile are classified as short-term institutional blockholders, and those ranked in the bottom tercile are classified as long-term institutional blockholders. Short-term (SIO) and long-term institutional ownership (LIO) are defined as the ratio of the number of shares owned by short-term (long-term) blockholders to the total number of shares outstanding. We also include several firm-characteristic variables to control for potential drivers of R&D activities. High leverage induces managers to reduce their R&D expenditures to meet their debt commitments (Bushee, 1998). In addition, a high leverage ratio affects corporate decisions related to innovation resources (Choi et al., 2011). Thus, we measure leverage (LEV) as the book value of debt divided by total assets, and we expect a negative relation between R&D activities and leverage. A cash flow shortage leads small firms to reduce their R&D expenditure (Bushee, 1998). Larger firms have a greater ability to invest in R&D. Following Choi et al. (2011), we measure a firm's size (SIZE) as the natural logarithm of total assets. A firm's life cycle affects innovation (Kim et al., 2017). To control for the effect of age, we include the firm's age (AGE), defined as the natural logarithm of years since the firm was established. Following the innovation literature, we include the return on assets (ROA) as a proxy for a firm's profitability (Choi et al., 2011; Fang et al., 2014), the book-to-market ratio (BM) to control for growth opportunities in firms (Eng and Shackell, 2001; Wahal and McConnell, 2000), and Tobin's Q (TOBQ) as a stock market-based proxy for firm value (Aghion et al., 2013). Finally, we measure asset tangibility (TGBT) as tangible assets scaled by total assets. Table 1 reports the descriptive statistics for the sample. The summary statistics of the major variables are the mean, standard deviation, minimum, 25th percentile, median, 75th percentile, 95th percentile, and maximum. On average, institutional blockholders hold about 4.2% of a firm's total shares outstanding. The average number of shares owned by domestic (foreign) blockholders is 3.1% (0.9%), and domestic blockholders have greater shareholdings than those of foreign blockholders by 2.2%. Note that a large number of firms in Korea are not owned by any blockholders. This result suggests a dispersed ownership structure in which most shareholders might be individual investors or institutional investors with fewer than 5% of shares. Many firms are controlled by owner-managers, and, thus, diffused shareholders cannot exert governance in firms. Institutional blockholders, especially foreign blockholders, are therefore likelier to sell their shareholdings than to actively monitor their firms. Foreign blockholders have far lower shareholdings than domestic blockholders do. For example, in contrast to domestic short-term blockholders, which hold 1.2% of shares outstanding, foreign short-term blockholders hold only 0.1% of outstanding shares. This result suggests that foreign short-term blockholders are likelier to sell their shares than they are to monitor their firms. Table 2 presents the correlations between RNDS and the ownership variables. The lower diagonal shows the Pearson contemporaneous correlation coefficients, and the upper diagonal shows the Spearman contemporaneous correlation coefficients. Institutional blockholders show a significant negative correlation with RNDS, consistent with our hypothesis. Additionally, we find a significant negative correlation between RNDS and SIO. More importantly, the correlation for foreign short-term blockholders is much higher than that for domestic short-term blockholders, suggesting that foreign blockholders might not encourage R&D activities by exercising a monitoring role. The correlations show the interrelation between R&D activities and the ownership variables. However, a causal relation cannot be identified because the variables are all contemporaneous. In addition, the relationship between R&D activities and the ownership variables might be the result of other firm characteristics. Thus, the analysis in the next section focuses on the influence of lagged 6

Pacific-Basin Finance Journal 57 (2019) 101188

S. Kang, et al.

Table 1 Descriptive statistics. Variable

Mean

Std. Dev.

Min

5%

25%

Median

75%

95%

Max

RNDS BLOCK_IO BLOCK_IO_D BLOCK_IO_F LIO LIO_D LIO_F SIO SIO_D SIO_F LEV SIZE AGE ROA TOBQ TGBT BM

0.0151 0.0417 0.0309 0.0092 0.0159 0.0088 0.0052 0.0136 0.0118 0.0014 0.4462 25.8142 3.1514 0.0365 1.1755 0.3148 1.3335

0.0325 0.0802 0.0676 0.0310 0.0522 0.0389 0.0216 0.0358 0.0334 0.0096 0.2054 1.4486 0.6524 0.0801 0.6804 0.1864 0.9935

0.0000 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000 0.0619 22.7073 0.6931 −0.2651 0.4654 0.0055 0.1017

0.0000 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000 0.1215 23.9897 2.0794 −0.1105 0.5838 0.0232 0.2579

0.0000 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000 0.2804 24.8334 2.7081 0.0061 0.7921 0.1740 0.6284

0.0017 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000 0.4481 25.5119 3.2581 0.0386 0.9710 0.3065 1.0847

0.0143 0.0620 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000 0.5991 26.4777 3.6636 0.0775 1.2933 0.4403 1.7338

0.0722 0.2122 0.1738 0.0830 0.1057 0.0637 0.0585 0.1026 0.0970 0.0000 0.7844 28.9199 4.0254 0.1559 2.4867 0.6398 3.2744

0.2220 0.3900 0.3458 0.1670 0.3188 0.2628 0.1204 0.1618 0.1558 0.0723 0.9523 30.4902 4.1897 0.3212 4.7367 0.8144 5.5705

This table presents the mean, standard deviation, minimum, 5th percentile, 25th percentile, median, 75th percentile, 95th percentile, and maximum values of the main variables for the sample for March 2005 to September 2015 after winsorizing at the 1% and 99% levels. Table 2 Correlations. RNDS RNDS BLOCK_IO BLOCK_IO_D BLOCK_IO_F LIO LIO_D LIO_F SIO SIO_D SIO_F

−0.0462⁎⁎⁎ −0.0432⁎⁎⁎ −0.0144 −0.0169⁎ −0.0186⁎⁎ 0.0019 −0.0403⁎⁎⁎ −0.0299⁎⁎⁎ −0.0361⁎⁎⁎

BLOCK_IO −0.0501

⁎⁎⁎

0.8767⁎⁎⁎ 0.4723⁎⁎⁎ 0.7149⁎⁎⁎ 0.5718⁎⁎⁎ 0.3857⁎⁎⁎ 0.5541⁎⁎⁎ 0.5187⁎⁎⁎ 0.1882⁎⁎⁎

BLOCK_IO_D

BLOCK_IO_F

⁎⁎⁎

⁎⁎⁎

−0.0352 0.8679⁎⁎⁎ 0.0207⁎⁎ 0.5442⁎⁎⁎ 0.6535⁎⁎⁎ 0.0116 0.5689⁎⁎⁎ 0.6008⁎⁎⁎ 0.0116

−0.0449 0.5006⁎⁎⁎ 0.0742⁎⁎⁎

0.4180⁎⁎⁎ −0.0342⁎⁎⁎ 0.8062⁎⁎⁎ 0.1859⁎⁎⁎ 0.0378⁎⁎⁎ 0.4346⁎⁎⁎

LIO

LIO_D ⁎⁎⁎

−0.0331 0.6063⁎⁎⁎ 0.4061⁎⁎⁎ 0.5241⁎⁎⁎ 0.8132⁎⁎⁎ 0.5074⁎⁎⁎ −0.0081 −0.0104 0.0063

−0.0128 0.4339⁎⁎⁎ 0.5000⁎⁎⁎ −0.0259⁎⁎⁎ 0.7179⁎⁎⁎ −0.0310⁎⁎⁎ −0.0264⁎⁎⁎ −0.0230⁎⁎ −0.0140

LIO_F

SIO ⁎⁎⁎

−0.0351 0.4011⁎⁎⁎ 0.0502⁎⁎⁎ 0.7968⁎⁎⁎ 0.6628⁎⁎⁎ −0.0259⁎⁎⁎ 0.0497⁎⁎⁎ 0.0361⁎⁎⁎ 0.0504⁎⁎⁎

SIO_D ⁎⁎⁎

−0.0454 0.6342⁎⁎⁎ 0.6437⁎⁎⁎ 0.2172⁎⁎⁎ 0.0347⁎⁎⁎ −0.0133 0.0634⁎⁎⁎ 0.9392⁎⁎⁎ 0.3303⁎⁎⁎

SIO_F ⁎⁎⁎

−0.0339 0.5902⁎⁎⁎ 0.6804⁎⁎⁎ 0.0634⁎⁎⁎ 0.0275⁎⁎⁎ −0.0115 0.0504⁎⁎⁎ 0.9292⁎⁎⁎

−0.0375⁎⁎⁎ 0.2240⁎⁎⁎ 0.0322⁎⁎⁎ 0.4530⁎⁎⁎ 0.0298⁎⁎⁎ −0.0095 0.0544⁎⁎⁎ 0.3535⁎⁎⁎ 0.0068

0.0035

This table presents the correlations between R&D intensity and the main ownership variables. The figures below (above) the diagonal are Pearson (Spearman) contemporaneous correlation coefficients. ⁎⁎⁎, ⁎⁎, and ⁎ indicate statistical significance at the 1%, 5%, and 10% levels, respectively.

blockholders' ownership on R&D activities, controlling for firm characteristics. 4. Empirical results 4.1. Baseline estimation In this section, we use a cross-sectional regression to examine the impact of institutional blockholders on firms' R&D activities after controlling for other confounding factors. To examine the influence of institutional blockholdings on a firm's R&D activities, we estimate the following baseline model:

R&D intensityi, t = α + β1 ∙Institutional Blockholdingsi, t − 1 + β2 ∙Leveragei, t − 1 + β3 ∙Sizei, t − 1 + β4 ∙Agei, t − 1 + β5 ∙Return on Asseti, t − 1 + β6 ∙Tobin′Qi, t − 1 + β7 ∙Asset Tangibilityi, t − 1 + β8 ∙Book to Market ratioi, t − 1 + β9 ∙Industry Dummyi + β10 ∙Year Dummyt + εi, t .

(1)

In this model, the subscripts i and t denote a firm and year, respectively, and industry and year dummies are included because R& D activities in firms depend on the industry and time. All independent variables are lagged by 1 year. For instance, RNDS is measured in 2010, and the institutional blockholdings and control variables are measured in 2009. It takes time for institutional blockholders to monitor firms and to influence the decision process associated with R&D activities. This time-lagged approach not only considers the time taken to monitor a firm's activities but also avoids reverse causality. The causality describing whether institutional blockholders encourage or discourage firms' spending on R&D might be reversely inferred as blockholders preferring firms with high or low R&D intensity. We use lagged blockholders' ownership to estimate firms' future R&D activities. However, reverse causality cannot be completely addressed. To mitigate this concern, we revisit the reverse causality problem in the robustness tests. For the institutional blockholdings variable in the baseline model, Model 1 employs BLOCK_IO to examine the influence of total 7

Pacific-Basin Finance Journal 57 (2019) 101188

S. Kang, et al.

Table 3 Influence of heterogeneous institutional blockholdings on firms' R&D activities.

Constant BLOCK_IOt-1

Model 1

Model 2

Model 3

Model 4

-0.88⁎⁎⁎ (-3.20) -0.01⁎ (-1.81)

-0.79⁎⁎⁎ (-3.34)

-0.76⁎⁎⁎ (-2.95)

-0.87⁎⁎⁎ (-3.08)

-0.01⁎ (-1.80) -0.00 (-0.31)

BLOCK_IO_Dt-1 BLOCK_IO_Ft-1 LIOt-1

0.00 (0.48) -0.02⁎⁎ (-2.25)

SIOt-1 LIO_Dt-1

-0.03⁎⁎⁎ (-9.54) 0.00 (0.69) -0.01⁎⁎⁎ (-7.30) -0.06⁎⁎⁎ (-6.10) 0.01⁎⁎⁎ (5.25) -0.01⁎⁎⁎ (-3.56) -0.00⁎⁎⁎ (-4.28) Yes Yes 11,535 0.16

-0.03⁎⁎⁎ (-9.51) 0.00 (0.52) -0.01⁎⁎⁎ (-7.27) -0.06⁎⁎⁎ (-6.02) 0.01⁎⁎⁎ (5.25) -0.01⁎⁎⁎ (-3.58) -0.00⁎⁎⁎ (-4.21) Yes Yes 11,535 0.16

0.00 (0.33) 0.01 (0.49) -0.02 (-1.47) -0.08⁎⁎⁎ (-3.38) -0.03⁎⁎⁎ (-9.55) 0.00 (0.55) -0.01⁎⁎⁎ (-7.26) -0.06⁎⁎⁎ (-6.01) 0.01⁎⁎⁎ (5.29) -0.01⁎⁎⁎ (-3.58) -0.00⁎⁎⁎ (-4.21) Yes Yes 11,535 0.16

0.16

4.47⁎⁎

5.20⁎⁎

LIO_Ft-1 SIO_Dt-1 SIO_Ft-1 LEVt-1 SIZEt-1 AGEt-1 ROAt-1 TOBQt-1 TGBTt-1 BMt-1 Industry dummies Year dummies Observations R-squared

-0.03⁎⁎⁎ (-9.52) 0.00 (0.72) -0.01⁎⁎⁎ (-7.29) -0.06⁎⁎⁎ (-6.11) 0.01⁎⁎⁎ (5.25) -0.01⁎⁎⁎ (-3.55) -0.00⁎⁎⁎ (-4.28) Yes Yes 11,535 0.16

Test for the equality of coefficient estimates

institutional blockholdings on R&D activities. In Model 2, we employ BLOCK_IO_D and BLOCK_IO_F to consider the different monitoring incentives of institutional blockholders by national origin. Model 3 uses LIO and SIO to examine the different influences of institutional blockholders with different investment horizons on R&D investments. Lastly, in Model 4, we include LIO_D, LIO_F, SIO_D, and SIO_F to determine the influence of institutional blockholders by national origin and investment horizon simultaneously. Table 3 shows the results of the four estimation models for the relationship between institutional blockholdings and RNDS, controlling for other firm characteristics. We adjust the standard errors by clustering at the firm level.2 In Model 1, 1-year lagged institutional blockholdings are significantly and negatively related to a firm's R&D intensity, consistent with the hypothesis. The results suggest that institutional blockholders who invest in Korean firms are less likely to affect a firm's R&D activities and are likelier to sell their shares rather than engage in active monitoring. When we consider the national origins of institutional blockholders, the results in Model 2 do not support the hypothesis, but the coefficients on BLOCK_IO_D and BLOCK_IO_F are both negative. Even domestic institutional blockholders, who are considered to have a positive influence on R&D investments, have a significantly negative effect, supporting the short-termism theory. The results of Model 3 show that short-term institutional blockholdings are significantly and negatively related to RNDS but that the influence of long-term institutional blockholdings is not significant. In addition, the bottom part of Table 3 presents the results of the test for the equality of the blockholding coefficient estimates. The results show that the F-statistics of the test for the equality of the estimated coefficients on LIO and SIO are significant at the 5% level, suggesting a significant difference in the coefficients on LIO and SIO. Long-term institutional blockholders, who are considered to monitor actively, do not motivate firms to invest in R&D. Short-

2 In undocumented analyses, we re-estimate the models using the Fama–MacBeth cross-sectional regression approach (Fama and MacBeth, 1973) and the two-way clustered standard errors approach (Petersen, 2009). The overall results are consistent with the main results and support the hypotheses.

8

Pacific-Basin Finance Journal 57 (2019) 101188

S. Kang, et al.

term institutional blockholders dislike firms investing in R&D and show short-termism. These findings support the hypothesis that the monitoring incentives of institutional blockholders decrease with stronger short-termism. Thus, long-term blockholders do not monitor actively, and short-term blockholders aggravate their myopic behavior. Taken together, institutional blockholders have a negative influence on firms' R&D activities, and this tendency is stronger for short-term blockholders than it is for long-term blockholders. When we differentiate the influence of institutional blockholders on R&D by national origin and investment horizon simultaneously, only foreign short-term institutional blockholdings are significantly and negatively related to firms' R&D intensity (see the results of Model 4 in Table 3). Long-term institutional blockholders do not have a significant positive influence on R&D intensity. The coefficient for foreign short-term blockholders is about four times higher than that for domestic short-term blockholders. The results of the test for the equality of the estimated coefficients on SIO_D and SIO_F show that this difference is statistically significant. These findings show that the short-term focus of institutional blockholders is increased by the short-termism of foreign institutional blockholders. In terms of economic significance, every 1% increase in foreign short-term blockholdings reduces R&D investments by 0.08% relative to sales per year. Considering that the mean of RNDS is 1.5% (see Table 1), this amount is equivalent to about 5% of the mean of annual R&D investments relative to sales in Korean firms. In summary, we find that institutional blockholders do not actively engage in firms' R&D activities.3 More importantly, foreign short-term institutional blockholders are incentivized to discourage firms from investing in R&D. Overall, in the Korean market, the monitoring incentive of institutional blockholders has weakened, but the myopic behavior incentive has strengthened. These results are contrary to those of Rong et al. (2017) that independent institutional investors (mutual funds) promote R&D productivity. In the Korean market, institutional monitoring is not effective, and institutional investors engage in myopic investment behavior, consistent with the results of Bushee (1998, 2001), who finds that the short-term focus of institutional investors leads to managers to cut their R &D spending to meet short-term earnings goals. We argue that the local characteristics of Korean firms have led to these changes in the monitoring role of institutional blockholders and short-termism. To further investigate why the active monitoring channel is not effective and institutional blockholders are short-term oriented, we examine the corporate governance environment and focus on how the preference for good governance quality affects institutional blockholders' engagement in firms' R&D spending. 4.2. Additional tests Institutional blockholders consider the governance quality of a firm before investing. To capture the internal corporate governance of a firm, we introduce three governance variables: KCGS, BOD_SIZE, and IND_PCT. Here, KCGS measures the overall internal governance quality of a firm, and BOD_SIZE and IND_PCT capture the composition of the firm's board of directors. The Korea Corporate Governance Service evaluates the comprehensive governance practices of firms listed on the KOSPI and selected KOSDAQ firms on an annual basis and reports a total corporate governance score in each case. This firm-level corporate governance score is measured for the categories of shareholder protection, board of directors, corporate disclosure, auditing organization, and earnings distribution. BOD_SIZE is the natural logarithm of the total number of directors, and IND_PCT is the percentage of independent directors on the overall board. If the negative relationship between institutional blockholding and R&D intensity is driven by institutional blockholders' preferences for good governance quality, it should be more evident in firms with poorer governance practices. To investigate this hypothesis, we divide the sample into two subsamples: a high KCGS sample and a low KCGS sample. For each sample, we estimate Model 4 and provide the results in Table 4. Whereas the negative influence of short-term blockholdings on R&D intensity is still statistically significant for firms with a low KCGS, it is not statistically significant for firms with a high KCGS. In the low KCGS sample, the negative relationship is again stronger for foreign blockholders than it is for domestic blockholders, implying that the preference for strong corporate governance is more prevalent for foreign blockholders than it is for domestic blockholders. These results suggest that the overall quality of corporate governance influences institutional blockholders' behavior related to firms' R&D activities. Not only does poor corporate governance make institutional monitoring ineffective, it also makes foreign institutional blockholders susceptible to expropriation by the owner-manager. Thus, foreign short-term blockholders are likelier to avoid R&D investment and to be more short-term focused than domestic short-term blockholders are. To examine the influence of heterogeneous institutional blockholdings on firm R&D activities when we control for corporate governance quality, we estimate the following model (see Table 5):

RNDSi, t = α + β1 ∙LIO _Di, t − 1 + β2 ∙LIO _Fi, t − 1 + β3 ∙SIO _Di, t − 1 + β4 ∙SIO _Fi, t − 1 + β5 ∙KCGSi, t − 1 + β6 ∙BOD _SIZEi, t − 1 + β7 ∙IND _ PCTi, t − 1 + β8 ∙LEVi, t − 1 + β9 ∙SIZEi, t − 1 + β10 ∙AGEi, t − 1 + β11 ∙ROAi, t − 1 + β12 ∙TOBQi, t − 1 + β13 ∙TGBTi, t − 1 + β14 ∙BMi, t − 1 + β15 ∙Industry Dummyi + β16 ∙Year Dummyt + εi, t .

(2)

The results presented in Table 5 (see the left column, “Corporate Governance Quality”) are consistent with the previous results. 3 Our findings might be affected by selection bias because the dependent variable, RNDS, is only reported when firms report R&D expenditures. In this setting, the sample is not selected randomly, resulting in potential selection bias. To address this concern, we retest our four baseline models using Heckman's (1979) selection model. In undocumented tables, we again find that foreign short-term blockholders have a significantly negative influence on R&D. In addition, when we re-estimate the baseline models for the subsample with both positive R&D intensity and lagged institutional blockholdings only, the overall results are consistent with the main results.

9

Pacific-Basin Finance Journal 57 (2019) 101188

S. Kang, et al.

Table 4 Influence of heterogeneous institutional blockholdings on firms' R&D activities, depending on overall corporate governance.

Constant LIO_Dt−1 LIO_Ft−1 SIO_Dt−1 SIO_Ft−1 LEVt−1 SIZEt−1 AGEt−1 ROAt−1 TOBQt−1 TGBTt−1 BMt−1 Industry dummies Year dummies Observations R-squared

Lower KCGS

Higher KCGS

−0.00 (−0.00) −0.01 (−1.13) −0.01 (−0.65) −0.02⁎⁎⁎ (−2.87) −0.04⁎ (−1.77) −0.01⁎⁎ (−2.18) 0.00 (0.24) −0.00 (−0.22) −0.02⁎⁎⁎ (−3.04) 0.00⁎⁎⁎ (2.84) −0.00 (−1.44) −0.00⁎ (−1.71) Yes Yes 2137 0.11

−0.00 (−0.27) 0.01 (0.66) −0.02 (−1.23) −0.02 (−1.39) −0.02 (−0.66) −0.02⁎⁎⁎ (−3.68) 0.00 (0.69) 0.00 (1.04) −0.05⁎⁎⁎ (−2.97) 0.01⁎ (1.91) −0.00 (−0.94) −0.00⁎ (−1.84) Yes Yes 1976 0.17

This table reports the estimation results for the regression of heterogeneous institutional blockholdings on R&D intensity, depending on KCGS. The lower (higher) KCGS column shows the results for firms with a lower (higher) KCGS than the median value. t-statistics are provided in parentheses and are clustered at the firm level. ⁎⁎⁎, ⁎⁎, and ⁎ indicate statistical significance at the 1%, 5%, and 10% levels, respectively.

Even after controlling for the corporate governance effects on R&D activities, we find that the negative relation between institutional blockholdings and R&D intensity is strongest for foreign short-term institutional blockholders. Furthermore, the positive coefficient on KCGS shows that good governance quality leads to an increase in the level of R&D investment, which is consistent with the agency view that good governance practices align managers' short-term interests with shareholders' long-term interests and, thus, encourage R&D investment. The negative relation between board size and R&D intensity supports the view that a lack of cohesiveness of larger boards requires more compromises to reach an agreement, and, thus, the decisions of larger boards are relatively less extreme (Cheng, 2008). Additionally, to consider the uniqueness of the Korean market, we investigate the effect of chaebols on the relationship between heterogeneous institutional blockholdings and R&D intensity. In the ownership structures of chaebol-affiliated firms, in which the firms are largely controlled by their (family) owner-managers, institutional blockholders do not have sufficient opportunity to exert governance and, thus, are encouraged to have short-term interests and to avoid long-term investments such as R&D. However, the chaebol structure does not only have a negative effect on R&D. Historically, chaebol-affiliated firms have become the main players in the country's economic growth as a result of government initiatives to boost the economy and to achieve exportoriented development. Under government protection, chaebols have been provided with various benefits and subsidies and encouraged to invest in R&D. As a result, chaebols have become global players and continue to invest in R&D to maintain their competitive advantage. For example, Samsung has been a top-ranked firm for several years in terms of the number of US patents. Furthermore, according to the Annual Report 2018 by the European Patent Office, Samsung and LG ranked third and fourth, respectively, and Korea was among the top 10 countries in terms of patent applicants. In addition, the wealth of owner-managers is tied to their firm's performance. Thus, they are motivated to care about the long-term profitability of the firm rather than focusing on managerial myopia. Accordingly, owner-managers' interests are aligned with those of shareholders, and the agency problem is less evident in family-run business groups (James, 1999). In summary, because chaebols might invest in R&D for their own interests, R&D in chaebol-affiliated firms might not be influenced by institutional blockholders' monitoring if chaebols have sufficient incentive to be innovative. To test the chaebol effect, we include a chaebol dummy (Chaebol_D), which takes a value of one if a firm is a chaebol affiliate and a value of zero otherwise. Furthermore, we add several interaction terms between various institutional blockholdings and the chaebol dummy. The model is estimated as follows: 10

Pacific-Basin Finance Journal 57 (2019) 101188

S. Kang, et al.

Table 5 Influence of heterogeneous institutional blockholdings on firms' R&D activities, controlling for corporate governance quality and considering the chaebol effect.

Constant LIO_Dt−1 LIO_Ft−1 SIO_Dt−1 SIO_Ft−1 KCGSt−1 BOD_SIZEt−1 IND_PCTt−1

Corporate governance quality

Chaebol effect

−0.01 (−0.55) −0.01 (−0.65) −0.01 (−0.40) −0.02⁎⁎ (−2.28) −0.11⁎⁎⁎ (−2.63) 0.00⁎⁎⁎ (3.47) −0.00⁎ (−1.73) −0.00 (−1.04)

−0.02 (−1.43) −0.02 (−1.43) 0.01 (0.39) −0.02⁎⁎ (−2.02) −0.13⁎ (−1.83) 0.00⁎⁎⁎ (3.61) −0.00⁎ (−1.96) −0.00 (−0.49) 0.02 (0.90) −0.08⁎⁎ (−2.08) 0.01 (0.29) 0.05 (0.58) −0.00⁎ (−1.69) −0.01⁎⁎⁎ (−3.90) 0.00 (1.21) 0.00 (0.74) −0.03⁎⁎⁎ (−3.12) 0.01⁎⁎⁎ (2.76) −0.00 (−0.64) −0.00⁎⁎ (−2.25) Yes Yes 2131 0.17

LIO_Dt−1 × Chaebol_Dt−1 LIO_Ft−1 × Chaebol_Dt−1 SIO_Dt−1 × Chaebol_Dt−1 SIO_Ft−1 × Chaebol_Dt−1 Chaebol_Dt−1 LEVt−1 SIZEt−1 AGEt−1 ROAt−1 TOBQt−1 TGBTt−1 BMt−1 Industry dummies Year dummies Observations R-squared

−0.01⁎⁎⁎ (−3.94) 0.00 (0.30) 0.00 (0.79) −0.03⁎⁎⁎ (−2.89) 0.01⁎⁎⁎ (2.81) −0.00 (−0.70) −0.00⁎ (−1.80) Yes Yes 2131 0.16

The left column of the table (Corporate Governance Quality) reports the estimation results for the regression of heterogeneous institutional blockholdings on R&D intensity, controlling for corporate governance variables. The right column of the table (Chaebol effect) reports the estimation results for the regression of heterogeneous institutional blockholdings on R&D intensity, including the chaebol interaction terms. t-statistics are provided in parentheses and are clustered at the firm level. ⁎⁎⁎, ⁎⁎, and ⁎ indicate statistical significance at the 1%, 5%, and 10% levels, respectively.

RNDSi, t = α + β1 ∙LIO _Di, t − 1 + β2 ∙LIO _Fi, t − 1 + β3 ∙SIO _Di, t − 1 + β4 ∙SIO _Fi, t − 1 + β5 ∙KCGSi, t − 1 + β6 ∙BOD _SIZEi, t − 1 + β7 ∙IND _ PCTi, t − 1 + β8 ∙ (LIO _Di, t − 1 × Chaebol _Di, t − 1) + β9 ∙ (LIO _Fi, t − 1 × Chaebol _Di, t − 1) + β10 ∙ (SIO _Di, t − 1 × Chaebol _Di, t − 1) + β11 ∙ (SIO _Fi, t − 1 × Chaebol _Di, t − 1) + β12 ∙Chaebol _Di, t − 1 + β13 ∙LEVi, t − 1 + β14 ∙SIZEi, t − 1 + β15 ∙AGEi, t − 1 + β16 ∙ROAi, t − 1 + β17 ∙TOBQi, t − 1 + β18 ∙TGBTi, t − 1 + β19 ∙BMi, t − 1 + β20 ∙Industry Dummyi + β21 ∙Year Dummyt + εi, t .

(3)

Table 5 provides the estimation results when we consider the chaebol effect (see the right column, “Chaebol effect”). The coefficient on the foreign short-term blockholding term remains significantly negative. In contrast, the coefficient on the foreign shortterm blockholding interaction term is not statistically significant. Furthermore, the effect of foreign short-term blockholders investing in chaebol-affiliated firms is 0.05(1) − 0.00 = 0.05, which is positive. The negative relation between foreign short-term blockholdings and R&D intensity disappears in the case of chaebol-affiliated firms. In sum, R&D in chaebol-affiliated firms is not affected by the short-termism of foreign short-term institutional blockholders. 11

Pacific-Basin Finance Journal 57 (2019) 101188

S. Kang, et al.

In contrast to chaebol-affiliated firms, non-chaebol firms have not been provided with the same level of support from the government. In addition to a lack of resources and knowledge, poor corporate governance in non-chaebol firms reduces monitoring incentives and increases short-termism. Combining the results that short-termism disappears in firms with good corporate governance and in chaebol-affiliated firms suggests that chaebol-affiliated firms do possess good governance compared with non-chaebol firms. As global players, chaebol-affiliated firms have been required to continuously improve their corporate governance. To meet this expectation, chaebol-affiliated firms have tried to improve corporate transparency and satisfy good disclosure standards. However, nonchaebol firms operate under relatively less strict governance standards and are monitored by fewer outside investors. Thus, in nonchaebol firms, the need to improve corporate governance is lower, and the opportunities to exert governance by monitoring are not sufficient owing to stronger owner-managers. Overall, the findings shown in Table 5 suggest that the heterogeneous characteristics of chaebol-affiliated firms and non-chaebol firms lead to a fundamentally different influence of institutional blockholders on R&D activities. In summary, when the monitoring channel is not effective owing to the existence of absolute ownership, institutional blockholders exert a negative influence on firms' R&D spending. This result is consistent with the findings of Choi et al. (2012a) and Rong et al. (2017). Choi et al. (2012a) do not find a positive relationship between ownership concentration and state ownership and technological innovation performance. Similarly, Rong et al. (2017) document that the positive effect of monitoring institutional ownership on firm innovation performance does not exist among majority state-owned enterprises (SOEs), suggesting that SOEs impede the effective external governance provided by institutional monitoring. They also find that, owing to the small proportion of ownership, the positive effect on firm innovation is significant only when qualified foreign institutional investors rely on mutual funds (i.e., independent institutional investors). We build upon these studies, which do not consider corporate governance, by suggesting that weak corporate governance under absolute ownership fosters short-termism and hinders monitoring. When absolute ownership limits opportunities to provide governance through monitoring, institutional blockholders no longer play a monitoring role to promote R&D activities. Instead, they engage in myopic behavior by encouraging managers to reduce R&D investment and boost short-term earnings, supporting the findings of Bushee (1998, 2001). Thus, institutional investors prefer firms with good corporate governance. 4.3. Robustness tests In this section, we perform three robustness checks of the primary findings in the previous sections. First, we analyze institutional blockholders' monitoring behavior by liquidity level in the Korean market. In general, institutional blockholders face a dilemma when they sell their blockholdings owing to the block size. In the case of low liquidity, institutional blockholders find it difficult to exit their positions when they are not satisfied with the management of a firm. This difficulty motivates institutional blockholders to monitor actively. However, in the Korean market, foreign short-term blockholders have strong short-termism and a significantly negative influence on firms' R&D. Hence, we suppose that the negative relation between foreign short-term blockholding and R&D intensity is still prominent in the case of low liquidity. We substantiate this argument using Amihud's (2002) illiquidity ratio as a proxy for a firm's level of liquidity. We calculate the annual average of Amihud's (2002) daily illiquidity ratios as follows: D

Amihudi =

∑t =i,1y

|Ri, y, d | Voli, y, d

Di, y

,

where Ri, y, d is the return on stock i on day d of year y; Voli, y, d is the dollar value of the trading volume of stock i on day d of year y; and Di, y is the number of days for stock i in year y. Based on this ratio, firms are sorted into five quintiles for each year, where firms in the fifth (first) quintile are grouped into the bottom (top) liquidity subsample. For each sample, we re-estimate Model 4 and provide the results in Table 6. We again find that foreign short-term blockholders have a significantly negative influence on R&D intensity. However, the coefficient for the bottom liquidity subsample is about twice that for the top liquidity subsample. This finding supports our previous result that even at low liquidity levels, the short-termism of foreign short-term institutional blockholders dominates their monitoring incentives and has a negative influence on firms' R&D activities. Second, we revisit the issue of reverse causality. The findings thus far could potentially arise because foreign institutional blockholders with short-term investment horizons prefer to invest in firms that do not invest in R&D or in non-R&D-intensive-firms. To address this concern, we employ a two-stage least squares model. In the first-stage regression, we use an instrument to model institutional blockholdings. The predicted institutional blockholdings are then used in the second-stage regression to examine the relationship between institutional blockholdings and R&D intensity. Following Kim et al. (2016) and Liu et al. (2018), we use two variables as instruments: TV and DIV. TV is defined as the natural logarithm of the annual average daily trading volume multiplied by the closing price of a given firm in a given year (Ferreira and Matos, 2008). Bennett et al. (2003), Gompers and Metrick (2001), and Parrino et al. (2003) show that institutional investors prefer to invest in liquid firms. The international finance literature (e.g., Brennan and Cao, 1997; Grinblatt and Keloharju, 2000) also shows that foreign institutional investors tend to be momentum traders and prefer to hold more liquid stocks. DIV is an indicator variable equal to one if a firm pays any dividend in a given year. Ferreira and Matos (2008) find that foreign institutional investors prefer to invest in dividend-paying firms. Kim et al. (2016) show that domestic and foreign institutional ownership are both positively related to whether firms pay high dividends. The first panel of Table 7 reports the estimation results for the first-stage regressions with the various institutional blockholding variables as the dependent variables. The second panel of Table 7 reports the estimation results for the second-stage regressions of R& 12

Pacific-Basin Finance Journal 57 (2019) 101188

S. Kang, et al.

Table 6 Influence of heterogeneous institutional blockholdings on firms' R&D activities, depending on Amihud's (2002) illiquidity ratio.

Constant LIO_Dt−1 LIO_Ft−1 SIO_Dt−1 SIO_Ft−1 LEVt−1 SIZEt−1 AGEt−1 ROAt−1 TOBQt−1 TGBTt−1 BMt−1 Industry dummies Year dummies Observations R-squared

Bottom liquidity

Top liquidity

0.02 (0.81) −0.00 (−0.04) −0.01 (−0.22) −0.00 (−0.09) −0.13⁎⁎⁎ (−2.88) −0.03⁎⁎⁎ (−5.73) 0.00 (0.06) −0.01⁎⁎⁎ (−3.80) −0.05⁎⁎⁎ (−3.58) 0.01⁎⁎ (2.57) −0.01⁎⁎ (−2.32) −0.00 (−1.33) Yes Yes 2363 0.16

0.00 (0.12) 0.01 (0.48) −0.08⁎⁎⁎ (−2.70) −0.00 (−0.08) −0.07⁎ (−1.78) −0.04⁎⁎⁎ (−5.11) 0.00 (0.67) −0.00⁎ (−1.68) −0.06⁎⁎⁎ (−3.33) 0.01⁎⁎⁎ (3.57) −0.00 (−0.65) −0.00 (−1.27) Yes Yes 2224 0.19

This table reports the estimation results for the regression of heterogeneous institutional blockholdings on R&D intensity in firms with different liquidities. The first (second) column shows the results for the firms with the highest (lowest) Amihud (2002) illiquidity ratios. t-statistics are provided in parentheses and are clustered at the firm level. ⁎⁎⁎, ⁎⁎, and ⁎ indicate statistical significance at the 1%, 5%, and 10% levels, respectively.

D intensity on the instrumented institutional blockholding variables. In the second panel, the Cragg-Donald Wald F-statistics from the first-stage regressions confirm the validity of our instruments, as the statistics are significant at the 1% level, suggesting that our instruments correlate strongly with the instrumented regressors. The Wu-Hausman test reported in the second panel fails to reject the null hypothesis that our instruments are exogenous. Consistent with our main results, we find that the coefficients on BLOCK_IOt−1 and SIO_Ft−1 are significantly negative, corroborating the finding that foreign short-term institutional blockholders have a strong short-term focus and, thus, are less likely to invest in R&D. Lastly, we perform additional endogeneity tests to ensure that our findings are not due to potential reverse causality between institutional blockholdings and measures of firm R&D intensity. To address the reverse causality concern, we employ the instrumental variable approach of Appel et al. (2016). Appel et al. (2016) study the influence of passive institutional investors on firm governance by examining institutional ownership for a subset of Russell 1000 and Russell 2000 stocks. They make the innovative argument that the bottom 250 stocks in the Russell 1000 and the top 250 stocks in the Russell 2000 have similar firm characteristics (i.e., market capitalizations) but different institutional ownership levels owing to the value-weighted index tracking. Specifically, institutions benchmarking the Russell 1000 (2000) assign a smaller (larger) weight to the bottom (top) stocks in the Russell 1000 (2000), resulting in exogenous variation in institutional ownership among the two groups of stocks. However, this argument may be affected by the aggregate relative weighting of these indices because institutions may invest disproportionately in the two indices. The KOSPI 100 and KOSPI 200 are the best-known Korean large-capitalization benchmarks used by index funds and the options market. The KOSPI 200 consists of 200 firms that are selected from the common stocks listed on the KOSPI based on factors such as market representability, industry representability, and liquidity. Relative to firms not included in this index, KOSPI 200 firms are predominantly large chaebol affiliates for which corporate governance could be a more important issue for investors. The KOSPI 100 further selects the top 100 stocks by market capitalization within the KOSPI 200. Because the KOSPI 100 consists of the largest 100 stocks within the KOSPI 200, institutions tracking these indices place a larger weight on KOSPI 100 stocks than they do on other KOSPI 200 stocks. Importantly, we expect differences in institutional ownership between the two groups of stocks that are unlikely to be attributed solely to differences in institutional investors' preferences. Like the identification strategy of Appel et al. (2016), our strategy benefits from the overweighting of KOSPI 100 stocks owing to the existence of both indices. We examine the impact of institutional blockholdings on firms' R&D intensities for stocks included in the KOSPI 100 and KOSPI 200. Following the identification strategy described by Appel et al. (2016), we calculate the instrumented institutional ownership as 13

Pacific-Basin Finance Journal 57 (2019) 101188

S. Kang, et al.

Table 7 Influence of heterogeneous institutional blockholdings on firms' R&D activities (two-stage least squares estimation). First stage

BLOCK_IOt

BLOCK_IO_Dt

BLOCK_IO_Ft

LIOt

SIOt

LIO_Dt

LIO_Ft

SIO_Dt

SIO_Ft

Intercept

−0.1507⁎⁎⁎ (−5.76) 0.0088⁎⁎⁎ (6.36) 0.0060⁎⁎⁎ (3.25) −0.0072⁎⁎⁎ (−4.76) 0.0129⁎⁎⁎ (4.07) −0.1932⁎⁎ (−2.01) 0.0464⁎⁎⁎ (3.04) −0.1146⁎⁎⁎ (−10.31) 0.0018 (1.30) −0.0579⁎⁎⁎ (−2.93) Yes Yes 10,345 0.1120

−0.1424⁎⁎⁎ (−5.80) 0.0049⁎⁎⁎ (3.76) 0.0038⁎⁎ (2.18) −0.0055⁎⁎⁎ (−3.90) 0.0110⁎⁎⁎ (3.72) −0.0998 (−1.11) 0.0473⁎⁎⁎ (3.30) −0.1046⁎⁎⁎ (−10.03) 0.0039⁎⁎⁎ (2.95) −0.0617⁎⁎⁎ (−3.33) Yes Yes 10,345 0.0935

−0.1283 (−0.73) 0.0039⁎⁎⁎ (6.54) 0.0022⁎⁎⁎ (2.79) −0.0017⁎⁎ (−2.55) 0.0018 (1.33) −0.0933⁎⁎ (−2.24) −0.0009 (−0.14) −0.0100⁎⁎ (−2.08) −0.0021⁎⁎⁎ (−3.38) 0.0038 (0.44) Yes Yes 10,345 0.0362

−0.1289⁎⁎⁎ (−4.99) 0.0077⁎⁎⁎ (5.65) 0.0057⁎⁎⁎ (3.13) −0.0066⁎⁎⁎ (−4.38) 0.0117⁎⁎⁎ (3.75) −0.1595⁎ (−1.68) 0.0582⁎⁎⁎ (3.86) −0.1139⁎⁎⁎ (−10.37) 0.0014 (0.99) −0.0637⁎⁎⁎ (−3.27) Yes Yes 10,345 0.1136

−0.1252⁎⁎⁎ (−5.17) 0.0040⁎⁎⁎ (3.15) 0.0035⁎⁎ (2.07) −0.0049⁎⁎⁎ (−3.46) 0.0104⁎⁎⁎ (3.56) −0.0910 (−1.02) 0.0573⁎⁎⁎ (4.05) −0.1057⁎⁎⁎ (−10.26) 0.0036⁎⁎⁎ (2.71) −0.0648⁎⁎⁎ (−3.54) Yes Yes 10,345 0.0942

−0.1237 (−0.33) 0.0037⁎⁎⁎ (6.14) 0.0022⁎⁎⁎ (2.71) −0.0017⁎⁎⁎ (−2.59) 0.0013 (0.94) −0.0684⁎ (−1.65) 0.0009 (0.14) −0.0082⁎ (−1.71) −0.0022⁎⁎⁎ (−3.52) 0.0011 (0.13) Yes Yes 10,345 0.0380

0.1316⁎ (1.76) 0.9914⁎⁎⁎ (28.74) −0.0071⁎⁎⁎ (−4.42) −0.0028⁎⁎⁎ (−4.91) 0.0073⁎⁎⁎ (4.79) 0.8673⁎⁎⁎ (46.21) −0.0566⁎⁎⁎ (−59.64) −0.7851⁎⁎⁎ (−48.89) 0.0038⁎⁎⁎ (22.26) −0.0066⁎⁎⁎ (−38.61) Yes Yes 10,345 0.0942

−0.1233⁎⁎⁎ (−18.26) 0.0027⁎⁎ (2.14) 0.0037⁎ (1.83) −0.0056⁎⁎ (−2.22) −0.0034 (−1.08) −0.1555⁎ (−1.92) 0.0455⁎⁎⁎ (8.41) 0.1497 (1.44) 0.0058 (0.07) 0.1108⁎⁎⁎ (4.82) Yes Yes 10,345 0.0783

−0.1320⁎⁎⁎ (−19.05) 0.0760⁎⁎⁎ (3.48) 0.0090⁎⁎⁎ (6.58) −0.0057⁎ (−1.76) −0.0059 (−0.14) 0.1591⁎ (1.69) 0.0402⁎⁎⁎ (3.75) −0.3305⁎⁎⁎ (−2.61) 0.0042⁎ (1.70) 0.0958⁎⁎⁎ (4.16) Yes Yes 10,345 0.0818

TVt−1 DIVt−1 LEVt−1 SIZEt−1 AGEt−1 ROAt−1 TOBQt−1 TGBTt−1 BMt−1 Industry dummies Year dummies Observations R-squared

Second stage

Model 1

Model 2

Model 3

Model 4

Constant

−1.3533⁎⁎⁎ (−3.21) −2.2287⁎ (−1.91)

−1.4857⁎⁎⁎ (−3.85)

−1.4221⁎⁎⁎ (−3.35)

−1.3455⁎⁎⁎ (−3.61)

BLOCK_IOt−1_Instrumented BLOCK_IO_Dt−1_Instrumented

4.2241 (0.98) −2.3124⁎⁎ (−2.27)

BLOCK_IO_Ft−1_Instrumented LIOt−1

1.1522 (0.57) −2.3122⁎⁎⁎ (−3.17)

SIOt−1 LIO_Dt−1 LIO_Ft−1 SIO_Dt−1 SIO_Ft−1 LEVt−1 SIZEt−1 AGEt−1 ROAt−1 TOBQt−1 TGBTt−1 BMt−1 Industry dummies Year dummies Observations R-squared Instrument validity test Cragg-Donald Wald F-statistics (p-value)

−0.0361⁎⁎⁎ (−8.86) 0.0031⁎ (2.21) −0.0060⁎⁎⁎ (−4.95) −0.0644⁎⁎⁎ (−4.95) 0.0051⁎⁎ (2.93) −0.0094⁎⁎⁎ (−4.40) −0.0029⁎⁎⁎ (−3.80) Yes Yes 10,345 0.4346

−0.0360⁎⁎⁎ (−9.01) 0.0035⁎ (1.99) −0.0059⁎⁎⁎ (−5.33) −0.0661⁎⁎⁎ (−4.52) 0.0050⁎⁎ (2.83) −0.0097⁎⁎⁎ (−4.00) −0.0030⁎⁎⁎ (−3.59) Yes Yes 10,345 0.4342

−0.0366⁎⁎⁎ (−7.84) 0.0035⁎ (1.86) −0.0055⁎⁎⁎ (−7.21) −0.0683⁎⁎⁎ (−4.03) 0.0054⁎⁎⁎ (3.34) −0.0099⁎⁎⁎ (−3.97) −0.0026⁎⁎⁎ (−4.55) Yes Yes 10,345 0.4418

0.1294 (0.31) −0.7829 (−1.31) 1.1274 (1.69) −3.1243⁎⁎ (−2.21) −0.0364⁎⁎⁎ (−7.96) 0.0036⁎ (1.84) −0.0055⁎⁎⁎ (−7.21) −0.0686⁎⁎⁎ (−3.96) 0.0053⁎⁎ (3.17) −0.0100⁎⁎⁎ (−3.90) −0.0028⁎⁎⁎ (−4.03) Yes Yes 10,345 0.4395 34.52

(continued on next page) 14

Pacific-Basin Finance Journal 57 (2019) 101188

S. Kang, et al.

Table 7 (continued) Second stage

Model 1

Model 2

Model 3

Model 4 (< 0.01) 1.73 (0.25)

Wu-Hausman test statistics (p-value)

This table reports the estimation results for the regression of heterogeneous institutional blockholdings on R&D intensity in firms based on a twostage least squares approach. t-statistics are provided in parentheses and are clustered at the firm and year levels, following Petersen (2009). ⁎⁎⁎, ⁎⁎, and ⁎ indicate statistical significance at the 1%, 5%, and 10% levels, respectively.

follows:

Institutional Blockholding Variablesi, t = α + β1· KOSPI 100i, t + β2· Log (SIZE )i, t + β3· Year Dummyt + ɛi, t , where KOSPI100i,t is a dummy variable that takes a value of one if firm i is in the KOSPI 100 index in reconstitution year t (i.e., from mid-June in year t to mid-June in year t + 1) and log(SIZE)i,t is the natural logarithm of the average daily market capitalization over the year prior to the end of April in year t (given that reconstitutions of the KOSPI 100 index are based on market capitalization). The corresponding institutional ownership of firm i is measured at the end of September in year t. The model prediction is used as an instrumental variable for institutional blockholdings in our main analysis. Specifically, following Appel et al. (2016), we relate the instrumented institutional blockholdings in September of each year to the measures of R&D intensity in the following fiscal year. The results are presented in Table 8. We again find strong evidence that foreign institutional blockholders reduce a firm's R&D investment, Table 8 Influence of heterogeneous institutional blockholdings on firms' R&D activities for KOSPI 200 stocks.

Constant BLOCK_IOt−1_Instrumented

Model 1

Model 2

Model 3

Model 4

−1.3013⁎⁎⁎ (−6.19) −0.2433⁎⁎⁎ (−2.88)

−1.2979⁎⁎⁎ (−6.18)

−1.2781⁎⁎⁎ (−5.18)

−1.3011⁎⁎⁎ (−3.34)

−0.3988⁎⁎⁎ (−2.77) −0.7572⁎⁎ (−2.21)

BLOCK_IO_Dt−1_Instrumented BLOCK_IO_Ft−1_Instrumented LIOt−1

0.1982 (0.50) −0.4911⁎⁎ (−2.31)

SIOt−1 LIO_Dt−1 LIO_Ft−1 SIO_Dt−1 SIO_Ft−1 LEVt−1 SIZEt−1 AGEt−1 ROAt−1 TOBQt−1 TGBTt−1 BMt−1 Industry dummies Year dummies Observations R-squared

−0.1812⁎⁎ (−2.50) 0.3983⁎⁎⁎ (5.62) −0.0124⁎⁎⁎ (−5.40) 0.2234 (0.73) 0.0987⁎⁎⁎ (4.29) −0.0298⁎⁎ (−2.47) 0.0345 (1.57) Yes Yes 877 0.0295

−0.2351⁎⁎ (−2.38) 0.2421⁎⁎⁎ (5.57) −0.0987⁎⁎⁎ (−5.51) 0.0423 (0.79) 0.2771⁎⁎⁎ (4.38) −0.1762⁎⁎⁎ (−2.80) 0.0341 (1.58) Yes Yes 877 0.0297

−0.2039⁎⁎ (−2.46) 0.1988⁎⁎⁎ (5.67) −0.0123⁎⁎⁎ (−5.43) 0.0773 (0.63) 0.2987⁎⁎⁎ (3.41) −0.1988⁎⁎⁎ (−2.92) 0.0345 (1.59) Yes Yes 877 0.0301

0.0532 (0.31) 0.1987 (0.55) 0.3341 (1.22) −0.9934⁎⁎ (−2.14) −0.1142⁎⁎ (−2.43) 0.1244⁎⁎⁎ (5.84) −0.0199⁎⁎⁎ (−5.54) 0.0344 (0.66) 0.0988⁎⁎⁎ (4.30) −0.0232⁎⁎⁎ (−3.11) 0.0152 (1.59) Yes Yes 877 0.0296

This table presents the coefficient estimations for the regression of firms' R&D activities on lagged institutional blockholdings for KOSPI 200 stocks. Instrumented institutional blockholdings are calculated using the KOSPI 100 index assignment in each year. t-statistics are provided in parentheses and are clustered at the firm and year levels, following Petersen (2009). ⁎⁎⁎, ⁎⁎, and ⁎ indicate statistical significance at the 1%, 5%, and 10% levels, respectively. 15

Pacific-Basin Finance Journal 57 (2019) 101188

S. Kang, et al.

implying that short-termism hinders institutional monitoring and negatively affects firms' R&D investments. 5. Conclusion Given that R&D is essential to a firm's sustainability and long-term growth, it is imperative to understand which institutional blockholders motivate R&D behavior. To do so, this study investigates the effects of heterogeneous institutional blockholders on firms' R&D intensity. We test whether monitoring by institutional blockholders enhances firms' R&D activities. To test this hypothesis, aggregate institutional blockholder ownership is divided according to the nationalities and investment horizons of institutional blockholders. Contrary to the view that monitoring by institutional investors boosts firms' R&D and/or innovation performance, we find that institutional blockholdings induce managers to have a short-term focus. Most notably, this negative influence is strongest for foreign short-term institutional blockholders. To investigate why active monitoring is not effective and why institutional blockholders exhibit short-termism, we consider overall corporate governance quality using various governance variables. The results indicate that the behavior of institutional blockholders may vary depending on the governance quality of the firms in which they have invested. In the presence of powerful owner-managers and weak governance in Korea, monitoring by institutional blockholders might not be effective. This property induces a short-term focus in institutional blockholders that outweighs the monitoring effect and reinforces the short-termism of institutional blockholders. In the Korean market, short-termism encourages managers to sacrifice R&D to achieve short-term performance goals, consistent with the findings of Bushee (1998, 2001). This study contributes to the existing literature by offering new evidence on the effect of heterogeneous institutional blockholders on firms' R&D activities. In particular, we highlight the importance of the role of internal corporate governance. Building a corporate environment favorable to institutional investors, including transparent corporate decision-making, strong investor protection, and good disclosure standards, will ensure that institutional monitoring is effective, leading to greater R&D investment by firms. Corporate governance is also important for regulators, because R&D at the firm level is fundamental to innovation at the governance level, which demands legal support for a transparent and fair market. Appendix A. Definitions of the variables

Variable

Definition

RNDS BLOCK_IO BLOCK_IO_D BLOCK_IO_F LIO LIO_D LIO_F SIO SIO_D SIO_F LEV SIZE AGE ROA BM TOBQ TGBT KCGS BOD_SIZE IND_PCT Chaebol_D

R&D expenditure divided by sales Total shares owned by institutional blockholders divided by the total number of shares outstanding Institutional ownership by domestic blockholders Institutional ownership by foreign blockholders Institutional ownership by long-term blockholders, as in Yan and Zhang (2009) Institutional ownership by domestic long-term blockholders Institutional ownership by foreign long-term blockholders Institutional ownership by short-term blockholders, as in Yan and Zhang (2009) Institutional ownership by domestic short-term blockholders Institutional ownership by foreign short-term blockholders The ratio of the book value of debt to total assets The natural logarithm of total assets The natural logarithm of years since a firm was established The ratio of operating income to total assets The ratio of the book value to the market value The ratio of the stock market value and total debt to total assets The ratio of tangible assets to total assets The overall corporate governance score, as per the Korea Corporate Governance Service The natural logarithm of the total number of directors The ratio of independent directors to the total number of directors Dummy variable that takes a value of one if a firm is a chaebol affiliate and a value of zero otherwise

References Aghion, P., Van Reenen, J., Zingales, L., 2013. Innovation and institutional ownership. Am. Econ. Rev. 103, 277–304. Amihud, Y., 2002. Illiquidity and stock returns: cross-section and time-series effects. J. Financ. Mark. 5, 31–56. Appel, I.R., Gormley, T.A., Keim, D.B., 2016. Passive investors, not passive owners. J. Financ. Econ. 121 (1), 111–141. Bae, K.H., Kang, J.K., Kim, J.M., 2002. Tunneling or value added? Evidence from mergers by Korean business groups. J. Financ. 57 (6), 2695–2740. Baik, B., Kang, J.-K., Kim, J.-M., 2010. Local institutional investors, information asymmetries, and equity returns. J. Financ. Econ. 97 (1), 81–106. Bennett, J.A., Sias, R.W., Starks, L.T., 2003. Greener pastures and the impact of dynamic institutional preferences. Rev. Financ. Stud. 16 (4), 1203–1238. Boone, A.L., White, J.T., 2015. The effect of institutional ownership on firm transparency and information production. J. Financ. Econ. 117, 508–533. Brennan, M.J., Cao, H.H., 1997. International portfolio investment flows. J. Financ. 52 (5), 1851–1880. Bushee, B.J., 1998. The influence of institutional investors in myopic R&D investment behavior. Account. Rev. 73, 305–333. Bushee, B.J., 2001. Do institutional investors prefer near-term earnings over long-run value? Contemp. Account. Res. 18, 207–246. Bushee, B.J., Noe, C.F., 2000. Corporate disclosure practices, institutional investors, and stock return volatility. J. Account. Res. 38, 171–202.

16

Pacific-Basin Finance Journal 57 (2019) 101188

S. Kang, et al.

Byun, H.-Y., Hwang, L.-S., Lee, W.-J., 2011. How does ownership concentration exacerbate information asymmetry among equity investors? Pac.-Basin Financ. J. 19, 511–534. Chen, X., Harford, J., Li, K., 2007. Monitoring: which institutions matter? J. Financ. Econ. 86, 279–305. Chen, V.Z., Li, J., Shapiro, D.M., Zhang, X.X., 2014. Ownership structure and innovation: an emerging market perspective. Asia-Pac. J. Manag. 31, 1–24. Chen, C.-J., Lin, B.-W., Lin, Y.-H., Hsiao, Y.-C., 2016. Ownership structure, independent board members and innovation performance: a contingency perspective. J. Bus. Res. 69, 3371–3379. Cheng, S., 2008. Board size and the variability of corporate performance. J. Financ. Econ. 87, 157–176. Choe, H., Kho, B.C., Stulz, R.M., 2005. Do domestic investors have an edge? The trading experience of foreign investors in Korea. Rev. Financ. Stud. 18, 795–829. Choi, S.B., Lee, S.H., Williams, C., 2011. Ownership and firm innovation in a transition economy: evidence from China. Res. Policy 40, 441–452. Choi, S.B., Park, B.I., Hong, P., 2012a. Does ownership structure matter for firm technological innovation performance? The case of Korean firms. Corp. Gov. 20, 267–288. Choi, H.M., Sul, W., Min, S.K., 2012b. Foreign board membership and firm value in Korea. Manag. Decis. 50, 207–233. Choi, H.M., Cho, Y.G., Sul, W., 2014. Ownership-control disparity and foreign investors' ownership: evidence from the Korean stock market. Emerg. Mark. Financ. Trade 50, 178–193. Choi, J.J., Jo, H., Kim, J., Kim, M.S., 2018. Business groups and corporate social responsibility. J. Bus. Ethics 153 (4), 931–954. Chung, K.H., Zhang, H., 2011. Corporate governance and institutional ownership. J. Financ. Quant. Anal. 46, 247–273. Chung, K.H., Elder, J., Kim, J.C., 2010. Corporate governance and liquidity. J. Financ. Quant. Anal. 45, 265–291. Dalziel, T., Gentry, R.J., Bowerman, M., 2011. An integrated agency-resource dependence view of the influence of directors' human and relational capital on firms' R& D spending. J. Manag. Stud. 48, 1217–1242. Dharwadkar, R., George, G., Brandes, P., 2000. Privatization in emerging economies: an agency theory perspective. Acad. Manag. Rev. 25, 650–669. Edmans, A., 2014. Blockholders and corporate governance. Annu. Rev. Financ. 6, 23–50. Edmans, A., Holderness, C., 2017. Blockholders: a survey of theory and evidence. In: Benjamin, E.H., Michael, S.W. (Eds.), The Handbook of the Economics of Corporate Governance, North-Holland, Chapter 8. Eng, L.L., Shackell, M., 2001. The implications of long-term performance plans and institutional ownership for firms' research and development (R&D) investments. J. Acc. Audit. Financ. 16, 117–139. Fama, E.F., French, K.R., 2006. Profitability, investment and average returns. J. Financ. Econ. 82, 491–518. Fama, E.F., MacBeth, J.D., 1973. Risk, return, and equilibrium: empirical tests. J. Polit. Econ. 81, 607–636. Fan, J.P.H., Wei, K.C.J., Xu, X.Z., 2011. Corporate finance and governance in emerging markets: a selective review and an agenda for future research. J. Corp. Finan. 17, 207–214. Fang, V.W., Tian, X., Tice, S., 2014. Does stock liquidity enhance or impede firm innovation? J. Financ. 69, 2085–2125. Ferreira, M.A., Matos, P., 2008. The colors of investors' money: the role of institutional investors around the world. J. Financ. Econ. 88, 499–533. Ferreira, M.A., Matos, P., Pereira, J.P., Pires, P., 2017. Do locals know better? A comparison of the performance of local and foreign institutional investors. J. Bank. Financ. 82, 151–164. Giannetti, M., Simonov, A., 2006. Which investors fear expropriation? Evidence from investors' portfolio choices. J. Financ. 61, 1507–1547. Gillan, S.L., Starks, L.T., 1998. A survey of shareholder activism: motivation and empirical evidence. Contemp. Financ. Dig. 2, 10–34. Gompers, P.A., Metrick, A., 2001. Institutional investors and equity prices. Q. J. Econ. 116 (1), 229–259. Gompers, P., Ishii, J., Metrick, A., 2003. Corporate governance and equity prices. Q. J. Econ. 118, 107–155. Grinblatt, M., Keloharju, M., 2000. The investment behavior and performance of various investor types: a study of Finland's unique data set. J. Financ. Econ. 55 (1), 43–67. Guadalupe, M., Kuzmina, O., Thomas, C., 2012. Innovation and foreign ownership. Am. Econ. Rev. 102, 3594–3627. Heckman, J.J., 1979. Sample selection bias as a specification error. Econometrica 47 (1), 153–161. Helfat, C.E., 2000. Guest editor's introduction to the special issue: the evolution of firm capabilities. Strateg. Manag. J. 21, 955–959. Honoré, F., Munari, F., Van Pottelsberghe de La Potterie, B., 2015. Corporate governance practices and companies' R&D intensity: evidence from European countries. Res. Policy 44, 533–543. Hwang, L.-S., Kim, H., Park, K., Park, R.S., 2013. Corporate governance and payout policy: evidence from the Korean business groups. Pac.-Basin Financ. J. 24, 179–198. James, H.S., 1999. Owner as manager, extended horizons and the family firm. Int. J. Econ. Bus. 6, 41–55. Joh, S.W., 2003. Corporate governance and firm profitability: evidence from Korea before the economic crisis. J. Financ. Econ. 68, 287–322. Kang, J.-K., Stulz, R.M., 1997. Why is there a home bias? An analysis of foreign portfolio equity ownership in Japan. J. Financ. Econ. 46, 3. Kim, H., Kim, H., Lee, P.M., 2008. Ownership structure and the relationship between financial slack and R&D investments: evidence from Korean firms. Organ. Sci. 19, 404–418. Kim, I.J., Eppler-Kim, J., Kim, W.S., Byun, S.J., 2010. Foreign investors and corporate governance in Korea. Pac.-Basin Financ. J. 18, 390–402. Kim, I., Miller, S., Wan, H., Wang, B., 2016. Drivers behind the monitoring effectiveness of global institutional investors: evidence from earnings management. J. Corp. Finan. 40, 24–46. Kim, H.D., Park, K., Roy Song, K., 2017. Do long-term institutional investors foster corporate innovation? Acc. Financ. 59 (2), 1163–1195. Kwon, H.U., Park, J., 2018. R&D, foreign ownership, and corporate groups: evidence from Japanese firms. Res. Policy 47, 428–439. La Porta, R., Lopez-de-Silanes, F., Shleifer, A., Vishny, R.W., 1998. Law and finance. J. Polit. Econ. 106, 1113–1155. La Porta, R., Lopez-de-Silanes, F., Shleifer, A., 1999. Corporate ownership around the world. J. Financ. 54, 471–517. Lee, Y., Cho, M., 2016. Does control-ownership disparity matter to foreign investors in Korea? Int. Rev. Econ. Financ. 44, 219–231. Leuz, C., Lins, K.V., Warnock, F.E., 2009. Do foreigners invest less in poorly governed firms? Rev. Financ. Stud. 22, 3245–3285. Li, D.H., Moshirian, F., Pham, P.K., Zein, J., 2006. When financial institutions are large shareholders: the role of macro corporate governance environments. J. Financ. 61, 2975–3007. Liu, C., Chung, C.Y., Sul, H.K., Wang, K., 2018. Does hometown advantage matter? The case of institutional blockholder monitoring on earnings management in Korea. J. Int. Bus. Stud. 49 (2), 196–221. Luong, H., Moshirian, F., Nguyen, L., Tian, X., Zhang, B., 2017. How do foreign institutional investors enhance firm innovation? J. Financ. Quant. Anal. 52, 1449–1490. McCahery, J.A., Sautner, Z., Starks, L.T., 2016. Behind the scenes: the corporate governance preferences of institutional investors. J. Financ. 71 (6), 2905–2932. Miletkov, M.K., Poulsen, A.B., Wintoki, M.B., 2014. The role of corporate board structure in attracting foreign investors. J. Corp. Finan. 29, 143–157. Park, T.J., Lee, Y., Song, K.R., 2014. Informed trading before positive vs. negative earnings surprises. J. Bank. Financ. 49, 228–241. Parrino, R., Sias, R.W., Starks, L.T., 2003. Voting with their feet: institutional ownership changes around forced CEO turnover. J. Financ. Econ. 68, 3–46. Petersen, M.A., 2009. Estimating standard errors in finance panel data sets: comparing approaches. Rev. Financ. Stud. 22, 435–480. Rong, Z., Wu, X., Boeing, P., 2017. The effect of institutional ownership on firm innovation: evidence from Chinese listed firms. Res. Policy 46, 1533–1551. Schumpeter, J.A., 1942. Socialism, Capitalism and Democracy. Harper and Brothers. Solow, R.M., 1957. Technical change and the aggregate production function. Rev. Econ. Stat. 39, 312–320. Un, C.A., Cuervo-Cazurra, A., 2008. Do subsidiaries of foreign MNEs invest more in R&D than domestic firms? Res. Policy 37, 1812–1828. Wahal, S., McConnell, J.J., 2000. Do institutional investors exacerbate managerial myopia? J. Corp. Finan. 6, 307. Yan, X.M., Zhang, Z., 2009. Institutional investors and equity returns: are short-term institutions better informed? Rev. Financ. Stud. 22, 893–924. Yoo, T., Rhee, M., 2013. Agency theory and the context for R&D investment: evidence from Korea. Asian Bus. Manag. 12, 227–252.

17