In this study, we investigate the determinants of credit ratings for companies affiliated with the Korean Chaebols. After controlling for endogeneity, we find that the average credit ratings for Chaebol-affiliated firms are significantly higher than those of non-Chaebol-affiliated firms. Next, we estimate the critical control threshold centrality and position established by Almedia et al. (2011) and apply those measures to the credit ratings of affiliated firms. The results show that affiliated firms with lower centrality measures tend to have higher credit ratings and that the relationship is not linear, suggesting that controlling shareholders use central firms to provide firms that have lower centrality with operational and financial support, that is, propping. Firms with lower centrality have higher credit ratings during a group-wide boom. This suggests that propping is more likely to emerge during an economic boom.

• EndNote
• RefWorks
• Scholar's Aid
• BibTeX

Export Citation Cancel

Starting with the capital asset pricing model (CAPM), a variety of systematic risk factors have been suggested. The performance of equity funds investing primarily in equities is evaluated using these risk factors and serves as the basis for judging a fund manager’s ability. Jensen (1968) evaluates fund performance using the CAPM; Carhart (1997) relies on a four-factor model that includes Fama and French’s (1993) three factors and Jegadeesh and Titman’s (1993) momentum factor; Pastor and Stambaugh (2002a, 2002b) include industry factors; and Barber, Huang, and Odean (2016) include Fama and French’s (2015) profitability and investment factors. As more systematic risk factors are considered, fund performance can be evaluated more precisely, and the systematic risk factor explaining the fund’s abnormal returns can be found. However, Barber, Huang, and Odean (2016) demonstrate that fund investors are most sensitive to CAPM alpha and that more sophisticated investors use more systematic risk factors to assess fund performance. In other words, sophisticated investors may be aware that equity funds are not likely to have abnormal returns in excess of their benchmark returns. Based on Barber, Huang, and Odean (2016), this study reassesses equity fund performance using more information about the fund’s market excess returns, which are most easily accessible to fund investors. This approach enables us to indirectly evaluate fund performance from the perspective of sophisticated investors and identify risk factors that explain the market excess returns of equity funds. Fama and French (2010) argue that the cross-sectional average alpha of U.S. mutual funds is close to zero, while only very extreme performance groups have a significant alpha. Thus, extreme performance is due to luck, not the fund manager’s skill. This study attempts to identify the systematic risk factors that explain extreme performance using market excess return distribution. It examines whether the significance of fund performance changes when more systematic risk factors are used. When equity funds do not earn higher returns than the benchmark returns, performance changes insignificantly and idiosyncratic risk explains the extreme performance when more systematic risk factors are considered. For example, when Fama and French’s (1993) three factors completely explain the performance of a fund with significant market excess returns during the analysis period, a fund investor who evaluates performance using the three-factor model judges the fund managers as having no skill. In addition, fund managers use a widely known strategy in the market rather than adopting various investment strategies. In other words, style investing strategies that do not rely on the three factors do not exceed the market returns on average. This study uses the CAPM, Fama and French’s (1993) three-factor model, and Fama and French’s (2015) five-factor model. This study investigates the effects of idiosyncratic risk on fund performance. Highperforming funds with high idiosyncratic risk are overestimated when only return data are used. Therefore, this study reevaluates the performance using Bayesian inference, which considers the idiosyncratic risk. The Bayesian inference is measured by the weighted average of the actual return data and the prior distribution. The effect of the prior distribution is stronger when the idiosyncratic risk of the actual returns is high. I examine the Bayesian alpha rank of the market excess return groups to see whether a change in performance rank occurs. The absolute value of the Bayesian alpha decreases compared with the OLS alpha due to the effect of prior distribution, but the performance rank does not change if the fund alpha is due to the fund manager’s ability, as all of the funds have the same prior distribution. However, funds with a high idiosyncratic risk have a significant effect on prior distribution, leading to a significant change in performance rank. I use Jones and Shanken’s (2005) Bayesian inference method, which includes the information that the cross-sectional average alpha of the fund is zero. This study identifies several results. First, the average abnormal returns of both the net and gross returns are zero, and the distribution of the returns is close to normal. Second, although performance is ranked according to the market excess returns, only the top 1% (top 4%) of the funds in net returns (gross returns) indicates significant performance. This result indicates that during the analysis period, fund returns are mostly explained by market returns, and extreme returns are due to idiosyncratic risk. Third, the three-factor model completely explains the extremely high returns. Thus, fund managers increase systematic risks (three factors), which are well known in the market. In other words, style investing strategies other than the three factors do not exceed market returns on average during the analysis period. Fourth, the performance rank of the high-performing funds with high idiosyncratic risk changes drastically when Bayesian inference is used. In particular, the effect of Bayesian inference is stronger in the boom market periods, which have a high idiosyncratic risk. Thus, a fund manager’s skill does not result in extreme performance. These findings are empirical and based on fund managers, fund investors, and the development of the fund industry. Although investors’ interest in equity funds weakened after the global financial crisis, the exchange traded fund (ETF) market is growing rapidly. The ETF market is expected to pursue various investment styles while following the market returns with low fund fees.

• EndNote
• RefWorks
• Scholar's Aid
• BibTeX

Export Citation Cancel

Given the wide variety of equity funds in the public offering market, fund investors find it difficult to select funds that match their investment preferences. Therefore, equity fund investors pay greater attention to funds that are recommended by sales professionals from banks or securities companies. They rely on brokers’ recommendations when selecting equity funds instead of conducting rigorous analysis. Therefore, brokers’ fund recommendations carry great importance from the perspective of investor protection. However, due to the limitations of data collection, research on this topic has not yet been conducted in Korea. Analysis of fund recommendations would have valuable implications for investor protection. Studies of equity funds have paid much attention to overall fund performance, fee structures, or the role of fund distributors (e.g., Won, 2009; Kim, 2011; Kho and Baek, 2013). However, although performance analysis of fund recommendations is important, research has paid little attention to the topic. To the best of our knowledge, no study has considered the real performance of recommended funds in Korea, as fund brokers do not voluntarily disclose data related to fund sales or the return performance of recommended funds. As each broker sells tens of funds, it is difficult to estimate the return performance of the funds of each broker. However, fund evaluation companies have recently begun to build data on funds recommended by each vendor, so analysis of the performance of recommended funds is possible. We empirically analyze the determinants of brokers’ fund recommendations and the performance of recommended funds in two ways. First, we analyze the factors that determine brokers’ fund recommendation decisions. Second, we analyze the future performance of the recommended funds. To confirm our findings, we use propensity score matching to compare the performances by matching recommended and non-recommended funds with similar characteristics. To investigate possible conflicts of interest in the fund recommendations, we focus on the levels of fees and commissions and the relationships between brokers and fund managers. We pay attention to the possibility that the seller, in recommending a fund to a customer, acts on behalf of the seller or affiliated fund manager rather than prioritizing the customer’s benefit. The results of this study are as follows. Funds with front-end loads and/or funds of asset management firms with no affiliated brokerage companies are more likely to receive recommendations from brokers. Furthermore, funds of large asset management companies receive more recommendations than those of medium or small management companies. In addition, new funds with short operating periods are selected as recommended funds, and new funds from affiliate management firms are more likely to be recommended. The performance of those recommended funds turns out to be inferior. Based on the results of analysis of the profitability of broker-recommended funds, we find that recommended funds do not perform better than funds that are not recommended by brokers. This result does not support the general perception that brokers have fund-selection ability. However, no evidence supports the view that the inferior performance of broker-recommended funds is caused by the self-interest of the broker who pursues the profit of a subsidiary asset management company. Recommended funds of affiliated asset management companies record better performance than those of non-affiliated companies. In particular, when the seller recommends a newly launched fund of the affiliated management company, its return performance is better. Thus, we conjecture that the inferior performance of a recommended fund is not due to the broker’s intention to subsidize an affiliated asset management company. This result suggests that a broker, through an information transfer from affiliated fund managers, plays a role in mitigating the information asymmetry of newly launched funds. It also supports the view that the affiliate broker analyzes the newly launched fund more accurately. Our study shows that broker-recommended funds generally exhibit inferior return performance. This result suggests that fund sellers are required to improve their fund selection ability to enhance the protection of fund investors. Sellers must improve their fund analysis capabilities through internal research and workforce reinforcement. Our study has some limitations in the sense that it covers a relatively short period due to the insufficient accumulation of fund recommendation data. Further research should be conducted to review the robustness of the analytical results over a wider analysis period after equity fund data are accumulated more broadly.

• EndNote
• RefWorks
• Scholar's Aid
• BibTeX

Export Citation Cancel

Until recently, the loan portfolios of savings banks in Korea have tended to concentrate on project financing with higher risks than the conventional loan portfolios of commercial banks. However, these concentrated loan portfolios and other disruptive behaviors led to massive bankruptcies among savings banks in 2011. In the past, many savings banks have experienced a similar bankruptcy process, largely due to the expansion of saving banks’ assets through mergers, high risk project financing loan projects, the illegal acts of large shareholders and executives, and the increasing loan concentration of high risk lenders. In particular, the concentration of real estate-related loans became a major management risk factor for savings banks in the 2000s, particularly in the deteriorating operating environment caused by the collapse of the real estate market after the global financial crisis. Furthermore, saving banks’ dependence on high-risk loans, such as project finance loans and unsecured personal loans, has intensified in recent years. This is a major factor in the failure of Korean savings banks in the event of external economic shocks. However, due to the limited data collection issue, only a few empirical studies have considered the relationship between the loan concentration ratio of savings banks and bank insolvency. The backdrop to the 2011 insolvency crisis of the savings banks is their large-scale growth and organizational change into a business group, which progressed rapidly up until 2010. Although mergers between savings banks were limited before December 2005, savings banks were already experiencing financial distress at that time due to the household credit crisis in 2003. The number of large savings banks with an asset size of over 1 trillion Korean won increased from 6 in June 2005 to 29 by the end of 2010. As of June 2010, 31 subsidiaries had been opened by 11 savings bank groups. The expansion and integration of these savings banks led to aggressive top-line expansion strategies, including high-risk lending operations and expansions. Over the 2011 to 2017 period, the average total assets of savings banks declined between 2011 and mid-2014 due to the failure of 31 savings banks after 2011, but the assets of the remaining savings banks rapidly increased after 2014. Research on savings bank credit concentration has been limited due to the lack of data, and most recent studies have analyzed only commercial banks or specific types of loans. Beck and Jonghe (2013) described the positive effects of loan concentration on commercial banks. The authors argued that commercial banks accumulate sufficient knowledge and experience in the field by focusing on loans to specific sectors over many years. This suggests that banks improve their ability to inspect and supervise loans and can reduce potential risks by identifying and addressing credit-related issues in a timely manner. Tabak, Fazio, and Cajueiro (2011) also argued that the concentration of lending to specific sectors in Brazilian banks shows that banks can enhance supervision, increase expertise in those areas, and lower bank risk while improving profitability. In this paper, we empirically examine the impact of concentrated loan portfolios on bank failures using the methodology of logistic regression. Our analysis shows that for saving banks the more concentrated the loan portfolio and the larger the number of mega loans (i.e., loans above 1 billion won), the higher the likelihood of failure. We also find that the more loans to relatively high risk borrowers and to borrowers with lower credit ratings in both the corporate and household sectors, including project financing and unsecured personal loans, the higher the probability of failure. However, mortgages with relatively good collateral turn out to have a significant negative relationship with savings bank insolvency.

• EndNote
• RefWorks
• Scholar's Aid
• BibTeX

Export Citation Cancel

Shareholder activism in public pension funds, asset management firms, and foreign hedge funds have led to an increasing number of calls for dividend payouts, so listed firms on the Korean Stock Exchange should urgently prepare an optimal dividend policy for raising corporate value in the mid- to long-term. However, developing such a policy has not previously been discussed or analyzed. In this study a method of deriving the optimal dividend policy of a firm to maximize its value is suggested. Based on the arguments (Fama and Babiak, 1968; Marsh and Merton, 1987; Lee, 1996, etc.) that cash dividends rely on permanent earnings and on the empirical findings (Fama and French, 1998; Pinkowitz, Stulz, and Williamson, 2006 and etc.) that cash dividends contribute to an increase in corporate value, the effect of cash dividends, which are driven by the permanence of earnings, on corporate values is analyzed. In the Beveridge-Nelson method earnings are divided into permanent and temporary earnings. Empirical results show that as the proportion of cash dividend relative to permanent earnings increases, corporate value increases significantly. Second, it is possible to derive the optimal level of dividend for a company, as the proportion of cash dividend relative to the permanent earnings is nonlinear and takes a reverse U form. Specifically, we find that corporate value increases with the cash dividend proportion of permanent earnings up to the point of 41.47%, and it decreases as the percentage increases beyond this point. These results imply that domestic firms must realize their optimal dividend policy by considering the persistence of earnings, and that institutional investors need to monitor whether a firm’s dividend policy at a micro level is aimed at achieving its optimal dividend level, which can maximize its shareholder value.

• EndNote
• RefWorks
• Scholar's Aid
• BibTeX

Export Citation Cancel

A stop-loss strategy is an investment strategy that automatically sells a stock that is suffering a loss when its price reaches a stop price. It is well known that investors actively use these strategies to protect against downward risk in the face of (ultra-) short-term volatility and drastically dropping stock prices. Since a stop-loss strategy uses algorithms to determine when the stock should be sold, investors do not have to follow the market constantly. Therefore, investors can control their behavioral biases and prevent losses in advance, and in practice this strategy is generally known to be useful in improving returns on investment. However, according to Kaminski and Lo (2014), the stop-loss strategy has different effects depending on the return-generating process the stocks follow. If stock returns follow a momentum or a regime-switching process, then a stop-loss strategy can increase portfolio profits and/or reduce volatility. However, if they follow a random-walk or a mean-reverting process, then a stop-loss strategy can lower portfolio profits. This paper analyzes the performance and effectiveness of stop-loss strategies in the Korean stock market by analyzing a general stop-loss strategy and a specific stop-loss strategy. For the general stop-loss strategy, we use the one studied by Lei and Li (2009). For the specific stop-loss strategy, we use the Active Risk-Management Phase 3 of the Korea National Pension Service (NPS). We analyze the effects of the general stop-loss strategy on the stocks listed on the KOSPI and KOSDAQ markets from July 1996 to June 2016. We use block bootstrapping to compare the performances of the stop-loss strategy to the buy-hold strategy, and calculate the efficiency of the stop-loss strategy by calculating the stop-loss premium. To analyze the specific stop-loss strategy embedded in the NPS’s Active Risk-Management Phase 3, we analyze the NPS’s investment universe from January 2011 to December 2017. We compute the cumulative rate of returns after the entry of the Active Risk-Management Phase 3 under an algorithm simulation. Ever since the 2008 global financial crisis, the Korean stock market, like other leading global markets, has experienced ultra-short-term volatility and sudden liquidity droughts in the absence of any material news. Stop-loss orders are a basic risk-management tool under these circumstances. Despite these structural market conditions, the Korea Exchange (KRX) does not offer stop orders or stop-loss orders to investors. Therefore, it is important to determine whether stop-loss strategies can help control the downside risk to investors. This study also addresses how the interaction between the structural conditions of the Korean stock market and the implementation of mandatory stock trading under the risk-management rules affect the investment strategy of the NPS, which is a timely and important topic. The results of our analyses are as follows. First, in terms of performance, we cannot confirm that a general stop-loss strategy is statistically superior to a buy-hold strategy, or vice versa. However, in terms of risk management, a general stop-loss strategy is more effective at reducing volatility than a buy-hold strategy. Second, 75% of the stocks that enter Phase 1 exit and return to a normal status within fourteen trading days. Phase 1 only involves enhanced monitoring of the stock, and stop-loss orders are not applied to these. When a stock enters Phase 2, a report is generated, but no specific action is mandated. Once a stock enters Phase 3, NPS is mandated to consider placing a stop-loss order. We find that in Phase 2 it is beneficial to observe whether the stock return becomes positive or remains stable over the ensuing 2~3 months (60 trading days) or 4~5 months (120 trading days). If the return continues to be positive or stable, NPS should continue to hold the stock and not place a stop-loss order. However, if the return remains negative in Phase 2, it is desirable to sell the stock without waiting for it to enter Phase 3. If a stock enters Phase 3 and NPS still holds it, it is desirable to sell the stock and to invest the sales proceeds in safe assets, and then repurchase the stock if and when it meets the re-entry threshold. Third, as the cumulative return after entering Phase 3 is often negative for up to 240 trading days, the stop-loss strategy does not seem to hurt returns by realizing transitory losses. Any possible increase in volatility, which results from the introduction of static VI or the increase of the price limit, does not appear to adversely affect the stop-loss strategy. Fourth, we find that the general stop-loss strategy in the Korean stock market does not improve performance. However, in the case of active risk management of NPS, it is desirable to sell the stocks prior to or on entering Phase 3. The difference between the two results appears to be largely due to the differences of the specific parameters of the stop-loss strategies. In the end, the effectiveness of stop-loss strategies critically depends on matching the parameters of the strategy to the conditions in the Korean stock market.

• EndNote
• RefWorks
• Scholar's Aid
• BibTeX

Export Citation Cancel

Since the Korean currency crisis of 1997, Korean firms have followed the compensation system used in the U.S. With this shift, pay for executives has risen sharply and the pay multiple has increased. We investigate the effect of this increased pay multiple on the overall operating and stock performance of Korean firms. Our data comprise 15,438 firm-year observations in 1,745 non-financial firms listed on two Korean stock markets from 2004 to 2016. Researchers in the field of human resource management find that relative pay has greater relevance than absolute pay in shaping the behaviors or attitudes of lower-level employees. Accordingly, we conjecture that enlarged pay multiples have a negative effect on Korean firms, as Korea has a relatively more egalitarian culture and a less elaborate evaluation system. In addition, relative deprivation theory posits that employees feel deprived when they perceive that their executives are overpaid. Employees may accept that high pay for an executive is deserved if that executive has high managerial ability. However, the employees may feel that executives with low managerial ability are paid too much. We therefore estimate the managerial ability of executives by adopting the method suggested by Demerjian, Lev, and McVay (2012), and we divide our sample firms into high and low managerial ability groups. We hypothesize that the effects of pay multiple on firm performance differ according to whether the firms have high or low managerial ability. Our empirical results are as follows. First, pay multiple negatively affects the overall operating and stock performance of most Korean firms. This finding implies that most employees think their executives’ pay is unjustifiably high, and it should be reduced. This sentiment seems to support relative deprivation theory. Second, the negative impact of pay multiple occurs only in the low managerial ability group. In the high managerial ability group, increased pay multiples have a positive effect. This finding indicates that employees in firms with low-ability managers feel deprived because of their executives’ excessive pay, whereas employees in firms with high-ability managers accept that their executives’ increased levels of pay are deserved. Third, the squared terms of pay multiple have negative coefficients. In other words, the negative effect of pay multiple becomes greater as pay multiple increases in the low managerial ability group. Meanwhile, this result indicates that the positive effect of pay multiple changes negatively as pay multiple increases in the high managerial ability group. In quintile analyses, we confirm that this non-linearity applies for both high and low managerial ability groups. In general, the negative effect becomes greater for each higher quintile of pay multiple in the low managerial ability group. We find that there is an inverted U-shaped relationship between pay multiple and firm performance in the high managerial ability group, which implies the existence of the optimal pay multiple. Fourth, we use an instrumental variable to check for potential endogeneity due to reverse causality between pay multiples and firm performances. Our empirical results are robustly confirmed. Finally, we further analyze various sub-samples of manufacturing and non-manufacturing firms, chaebol and non-chaebol firms, and Korea Stock Exchange (KSE) as compared to KOSDAQ firms. We obtain consistent results for all of the subsamples. In summary, we show that pay multiple has a negative effect on firm performance. However, this negative effect only occurs in low managerial ability firms, where the negative impact becomes more significant as the pay multiple grows larger. These findings are consistent with relative deprivation theory. This paper examines executive-employee pay multiples, rather than pay gaps among executives or employees. By linking the pay multiple to firm performance, we show that it is important to adopt a pay structure that includes both executives and employees. Furthermore, we contribute meaningful evidence concerning relative deprivation theory by analyzing how managerial ability is related to the relationship between pay multiple and firm performance. Our findings indicate that to enhance firm performance, executive pay structures need to reflect managerial ability.

• EndNote
• RefWorks
• Scholar's Aid
• BibTeX

Export Citation Cancel

In this paper, I analyze household portfolios with and without household debts, each of which is supposedly subject to distinct financial target according to household characteristics and financial status, respectively. Four issues were identified through the 2011~2013 Household Financial Survey. First, households with debts have more financial assets in addition to real assets than those without debts, which include stocks, bonds, insurance, and mutual funds. Households with debts also own more real estate and durables such as automobiles and real assets, which is statistically significant. This result implies that households with debts invest aggressively in asset allocation and in specific assets, unlike those without debts. Second, disposable income is a main interest variable and reduces the share of financial assets at a low level but increases the share at a high level, irrespective of household debts. In contrast, At a low level, the net worth increases real assets, whereas it decreases at a high level. Thus, we conclude that households tend to diversify the portfolio risk of real assets through financial markets. Third, as previously found, household debts reduce household financial assets but reinforce the holdings of real assets, which is also prevalent in the sample of households with excessive debts. Finally, in terms of generations, middle-aged and older households with debts are more likely to prefer stocks, and significantly more likely than younger households.

• EndNote
• RefWorks
• Scholar's Aid
• BibTeX

Export Citation Cancel

This study examines how the business relationship between IPO underwriters and institutional investors affects the holding periods of IPO stocks allocated to institutional investors by underwriters. Studies such as Reuter (2006), Binay, Gatchev, and Prinsky (2007), and Chung and Kim (2015) have shown that the business relationship between underwriters and institutional investors affects underwriters’ initial allocation of IPO stocks. Chemmanur, Hu, and Huang (2010) argue that underwriters allocate more IPO stocks to institutional investors because institutional investors tend to hold stocks longer, especially for weaker post-issue demand IPOs, helping underwriters’ efforts to stabilize prices in the after-IPO market period. However, it is well known that on average, institutional investors sell initially allocated IPO stocks quite soon after they have been listed in the market. For example, Aggarwal (2003) and Boehmer, Boehmer, and Fishe (2006) report that in the U.S. market, flipping—that is, selling initially allocated IPO stocks within a couple of days after their listings —is common and it is mostly conducted by institutional investors. An analysis by the Financial Supervisory Service in Korea of 142 Korean IPOs listed in the Jan. 2008 to Sept. 2010 period found that on average, 64.7% of IPO shares were allocated to institutional investors, who sold 34.2% of them during the first trading day and 48% of them during the first 4 weeks after the listing. In addition, Dongyang Securities, examining 119 Korean IPOs from the Jan. 2010 to Jun. 2011 period, report that the return on IPO stocks falls rapidly after their listing, and the average closing price of the IPO stocks in one year falls to only 77% of their closing prices on the first trading day. They blame the institutional investors’ flipping of IPO stocks for the rapid price drop. It is thus evident that the length of time that institutional investors hold initially allocated IPOs in the after-market is an important issue for IPO firms and underwriters who wish to stabilize prices and protect their reputation capital. We therefore argue, based on the findings from prior studies, that the holding period can be affected by the business relationship between underwriters and institutional investors. To examine the effect of the business relationship between underwriters and institutional investors on institutional investors’ holding period of initially allocated IPOs, we use monthly mutual fund holding information and IPOs listed in the KOSDAQ market for the 2002 to 2012 period. To the best of our knowledge, this is the first study to examine the holding period of IPO stocks by institutional investors in Korea. Following Chung and Kim (2015), we measure the strength of a business relationship in three ways: (1) the total investment amount by each mutual fund in IPOs underwritten by each underwriter, (2) the number of participation by each mutual fund in IPOs underwritten by each underwriter, and (3) the average investment of each mutual fund in each IPO underwritten by an underwriter for the two years prior to the target IPO. We find that mutual funds sell, on average, about 48% of the IPO stocks within 30 calendar days of the listing. However, funds that have 1 (2)-year business relationships with the underwriter hold the IPO stocks, on average, 22.67 (17.12) calendar days longer than those without a relationship. We also show that the strength of the relationship matters: the stronger the relationship is, the longer the holding period. Tobit analyses confirm Aggarwal’s (2003) finding that institutional investors tend to sell more underpriced IPOs (hot IPOs) more quickly. We, however, show that when institutional investors have a strong business relationship with underwriters, they seem to hold hot IPOs longer than when they have a relatively weaker relationship. Combined with the results from Chung and Kim (2015), these results imply that underwriters allocate more favorable IPO stocks to institutional investors with whom they have business relationships, and institutional investors hold the IPO stocks longer, helping the underwriters stabilize prices in the after-IPO market. In addition, we find that funds that are affiliated with underwriters hold IPO stocks for shorter periods than unaffiliated funds. We also find that the business relationship has a negative effect on the holding periods of affiliated funds.

• EndNote
• RefWorks
• Scholar's Aid
• BibTeX

Export Citation Cancel

The cost of capital is one of the key determinants of corporate investment, and according to the q theory framework, it has a negative effect on investment. However, there has been relatively little research on the relation between the cost of capital and corporate investment. Frank and Shen (2016) report the following findings: (1) the cost of capital estimated by asset pricing models such as the CAPM have a positive effect on investment; and (2) the implied cost of capital is negatively related to investment. They provide considerable insight into methods for estimating the cost of equity, but do not give a convincing explanation for their unexpected results. Building on their insight, we use a dataset of listed Korean firms to estimate a firm’s expected return using different measurement methods, and investigate whether the empirical result depends on how the cost of equity capital is measured. The most common approach to estimating a firm’s cost of equity is to calculate the expected return based on the CAPM. The implied cost of equity derived from stock prices and analyst forecasts could be an alternative measure of the expected return for a firm’s stock. In this study, we estimate the implied cost of equity by applying two different valuation models: the Gordon growth model and the abnormal earnings growth model developed by Ohlson and Juettner-Nauroth (2005). The sample dataset covers the 2002 to 2016 period. Following Frank and Shen (2016), we use a g-decomposed model in which investment is formulated as a function of marginal profitability and discount rate for a firm’s capital budgeting decision. According to the model based on the q theory, investment should be negatively related to the weighted average cost of capital, which represents the discount rate. Our findings, however, show that the CAPM-based estimate does not confirm the validity of the theoretical prediction. We consider the effect of macroeconomic factor such as the business cycle on the association between the CAPM-based estimates and corporate investment, as it is possible that firms make different investment decisions at different stages of the business cycle. We divide our sample period into an expansion period and a contraction period, and conduct the same analysis on the subsamples. We find that the positive impact of the CAPM estimate is only significant in the expansion period. We also find that the positive effect is large for the subset of relatively small firms that are listed on the Kosdaq market. The implied cost of capital-based analysis produces different results. We find that the implied cost of capital is negatively linked to investment, perhaps because the implied cost of equity capital (i.e., the return required by shareholders) better reflects investors’ expectations about the riskiness of firm-specific investments, and the information imbedded in the estimate can be useful for managers’ capital budgeting decisions. Next, we focus on the internal capital market effect, which occurs in the Korean business group. Following previous studies, we assume that firms that belong to business groups have easier access to internal capital markets. To examine whether the internal capital market plays a significant role in firms’ investment decisions, we categorize our sample firms into two subsamples: independent firms and group-affiliated firms. Our analyses of the subsamples suggest that the cost of capital can have different effects on firms’ investment expenditures depending on their business group affiliation. More specifically, we find that the negative relation between the weighted average cost of capital and investment is statistically insignificant for group-affiliated firms, whereas for independent firms, we find that investment is negatively affected by the weighted average cost of capital. This evidence is consistent with the notion that the investments of group-affiliated firms with easier access to internal capital markets are less affected by external financing cost than independent firms. In a further analysis, we also find that the investment of group-affiliated firms is negatively correlated with the debt component of the WACC, but not with its equity component. This implies that the internal capital market relieves the financing constraints of firms belonging to business groups, but it does not completely substitute for the external capital market. Finally, we use the expected return estimated by the Fama–French three-factor model as a proxy for the cost of equity. We find that the FF3-based estimate has a negative effect on investment, but the effect is not statistically significant for the whole sample, suggesting that the factor loadings of the Fama-French model can measure firm-specific risk more accurately than the CAPM beta, but they are insufficient to represent the unsystematic risk that individual firms are likely to face with respect to their investment decisions.

• EndNote
• RefWorks
• Scholar's Aid
• BibTeX

Export Citation Cancel