Get 20M+ Full-Text Papers For Less Than $1.50/day. Start a 14-Day Trial for You or Your Team.

Learn More →

Stock liquidity and capital allocation efficiency: Evidence from Chinese listed companies

Stock liquidity and capital allocation efficiency: Evidence from Chinese listed companies China Journal of Accounting Studies, 2014 Vol. 2, No. 3, 228–252, http://dx.doi.org/10.1080/21697213.2014.959413 Stock liquidity and capital allocation efficiency: Evidence from Chinese listed companies a b,c Jiacai Xiong * and Dongwei Su School of Accountancy, Jiangxi University of Finance and Economics, Nanchang 330013, b c China; Department of Finance, Jinan University, Guangzhou 510632, China; Research Institute of Finance, Jinan University, Guangzhou 510632, China Based on market microstructure theories and evidence, this paper investigates the relationship between stock liquidity and capital allocation efficiency using Chinese listed companies from 1998 to 2011. This paper finds that stock liquidity helps improve investment efficiency, mitigating both overinvestment and underinvestment. This finding is robust to numerous sensitivity analyses, including controls for endo- geneity and for the other known determinants of investment efficiency, the choice of the measure of stock liquidity and investment efficiency. Further analysis shows that stock liquidity improves corporate capital allocation efficiency by reducing agency costs and increasing the information content of share prices. Keywords: capital allocation efficiency; market microstructure; overinvestment; stock liquidity; underinvestment 1. Introduction As one of the three basic financial decisions that are made by firms, investment has become a long-debated issue in corporate finance. Investment links financing activities with capital allocation as well as determines the future cash flow and profitability of firms, thereby serving as the basis for enterprise growth. At a macroeconomic level, investment is among the three engines that pull an economy (namely, consumption, investment and export-import). However, a large body of literature finds that Chinese listed companies have low investment efficiency (e.g., Tang, Zhou, & Ma, 2007; Zhang & Song, 2009). Low-level development and redundant development are widespread among the steel, iron, and building materials industries, creating significant excess capacity, whereas the agricultural, environment, and high-technology sectors receive an insufficient amount of investment (Lv & Zhang, 2011). Because investment is among the major forces that have driven the economic growth of China over the past decades, mitigating underinvestment and overinvestment, as well as improving investment efficiency, must be investigated in order to boost the country’s economy. In the perfect world of Modigliani and Miller (1958), an investment policy is solely dependent on the investment opportunities of a firm as measured by Tobin’s Q (Tobin, 1969). However, several types of friction and distorting forces, such as information asymmetries and agency costs, may prevent firms from making optimal investments (Stein, 2003). Myers and Majluf (1984) argue that information asymmetries between *Corresponding author. Email: xiongjc-p@163.com Paper accepted by Tong Yu. © 2014 Accounting Society of China China Journal of Accounting Studies 229 managers and shareholders may lead to adverse selection and subsequently to underin- vestment. With regard to agency costs, the separation of ownership and control creates a conflict of interest between shareholders and managers (Jensen & Meckling, 1976). Managers are also trying to make investments to pursue their own private interests, leading to managerial empire-building and overinvestment (Jensen, 1986; Stulz, 1990). Previous studies have found that superior corporate governance is useful in improving the investment efficiency of Chinese listed firms (Tang et al., 2007; Xin, Lin, & Wang, 2007; Li, 2009; Lv & Zhang, 2011), whereas government intervention distorts the investment behaviour of these firms and subsequently leads to investment inefficiency (Chen, Sun, Tang, & Wu, 2011; Cheng, Xia, & Yu, 2008; Zhong, Ran, & Wen, 2010). However, we have not found any published material relating to the effect of stock liquidity on investment efficiency of Chinese listed firms. The positive effect of market liquidity on firm investment is strongly backed by evi- dence. Theoretical models predict that market liquidity facilitates the formation of a toehold stake (Kyle & Vila, 1991), permits non-blockholders to intervene and become blockholders (Maug, 1998), promotes highly efficient management compensation (Holmstrom & Tirole, 1993), strengthens the exit threats of blockholders (Admati & Pfleiderer, 2009; Edmans, 2009), and stimulates trade by informed investors, thereby improving investment decisions through highly informative stock prices (Khanna & Sonti, 2004). Empirical evidence shows that firms with higher liquidity maintain a lower cost of equity capital (Amihud & Mendelson, 1986; Amihud, 2002), a higher investment level (Munoz, 2013), and a higher firm value (Fang, Noe, & Tice, 2009). Therefore, liquidity may be positively related to investment efficiency. Although many studies relate liquidity to corporate governance and financial decisions, we have not found any published study relating to the effect of stock liquidity on the investment efficiency of Chinese listed firms. This study aims to provide empirical evidence and shed light on the relationship between liquidity and investment efficiency by using a sample of Chinese firms that were publicly listed between 1998 and 2011. We find that stock liquidity helps improve investment efficiency, mitigating both overinvestment and underinvestment. This result holds for multiple measures of liquid- ity and investment efficiency, and is robust to controlling for the other known determi- nants of investment efficiency. We also investigate the causes behind the beneficial effects of liquidity, such as curbing the opportunistic behaviour of managers and increasing the information content of share prices. This paper enriches the literature on stock liquidity. Early studies, such as those by Amihud (2002), Amihud and Mendelson (1986) and Pastor and Stambaugh (2003), exam- ine the effect of stock liquidity on asset pricing. However, there has been a growing inter- est in studying the relationship between stock liquidity and the decision of firms. These studies include evidence that firms with greater stock liquidity are more likely to use equity financing, resulting in lower financial leverage (Lipson & Mortal, 2009), that stock liquid- ity interacts with firm investment (Munoz, 2013), and that firms with liquid stocks have better performance (Fang et al., 2009). We contribute to this literature by relating stock liquidity to investment efficiency. We use the expected investment model of Richardson (2006) to compute overinvestment and underinvestment as well as to investigate the rela- tionship between liquidity and investment efficiency. We find that stock liquidity helps improve investment efficiency, mitigating both overinvestment and underinvestment. Additionally, our study sheds light on the mechanisms through which stock liquidity can affect firm decisions and outcomes. Fang et al. (2009) find liquidity enhances firm per- formance by increasing the information content of market prices and performance- 230 Xiong and Su sensitive managerial compensation. We explicitly investigate the causes behind the benefi- cial effects of liquidity, and find that stock liquidity improves corporate capital allocation efficiency by reducing agency costs and increasing the information content of share prices. The rest of the paper is organised as follows. Section 2 discusses the institutional background, reviews the existing literature on investment efficiency and on the role of stock liquidity in investment decisions, and develops testable hypotheses. Section 3 describes the research design, models, variable measures, and the sample. Sections 4 and 5 present empirical evidence on whether and why stock liquidity affects investment inefficiency. Section 6 concludes. 2. Institutional background, literature, and hypothesis development 2.1. Institutional background China offers a unique environment for analysing the investment efficiency of firms. First, China has maintained a state-dominated financial system in which government at various levels controls the allocation of financial resources in both the banking and security markets (He, Mao, Rui, & Zha, 2013). The banking system in China comprises the central bank, three policy banks, four large state-owned commercial banks, 10 national joint-stock commercial banks, approximately 90 regional commercial banks, as well as urban and rural credit cooperatives. The four state-owned commercial banks dominate the entire Chinese market (He et al., 2013). According to Cai and Zeng (2012), these four state-owned commercial banks had a market share of approximately 70% from 2002 to 2007. The Chinese stock market is also controlled by the govern- ment. Before 1999, the total annual number of initial public offerings (IPOs) in China was subject to a quota system. The central government set a quota for the quantity of equities or the number of firms to be listed annually before this period. Before 1999, a company that intended to list was required to be selected by a provincial government or ministry with a quota before asking the China Securities Regulatory Commission (CSRC) for approval. In July 1999, the IPO quota system was replaced by an authori- sation system. Investment banks are allowed to nominate firms for public listing and their nominations are screened by an independent listing committee of the CSRC. The independent listing committee assesses the qualifications of a to-be listed company. In addition, government-guided financial resource allocation usually favours a few large-scale state-owned enterprises (SOEs) and these SOEs may encounter soft budget constraints, while smaller SOEs and most non-SOEs cannot easily obtain financing from the state-controlled financial system (Allen, Qian, & Qian, 2005; Brandt & Li, 2003). This phenomenon of bank discrimination remains in place currently. Therefore, the latter group of enterprises suffers from serious financial constraints and subse- quently faces underinvestment. Second, most Chinese listed firms are business units that have been carved out from large SOEs and are controlled by government-related entities. As the controlling share- holder, the Chinese government often plays the conflicting dual roles of an SOE owner and administrator of social affairs. As an SOE owner, the government is supposed to benefit from value maximisation. However, to accomplish their social and political goals, such as regional economic development, higher employment rates, and social sta- bility, government leaders are driven to intervene in SOEs by changing the objective functions of SOEs based on their political preferences (Chen et al., 2011). Therefore, these SOEs tend to miss profitable investment opportunities by implementing the plans China Journal of Accounting Studies 231 of the government as well as fails to terminate unsuccessful projects because of their potential conflicts with government policies, which results in investment inefficiency. Third, the typical external governance mechanisms, such as debt, takeover threats, legal protection of investors, and product market competition, have been ineffective because of the political nature of the privatisation process (Su, 2005). Bank loans are traditionally viewed as grants from the state that are designed to bail out failing firms. State-owned banks retain a monopoly in the banking sector, but profit is not their over- riding objective. When a political favour is deemed appropriate, issuing subsidised loans and rescheduling of overdue debt can be arranged with SOEs (these are seen as soft budget constraints) (Su, 2005). In addition, a market for private, non-bank debt is yet to be established. The stock market lacks an active merger or takeover activity to discipline firm management. The capital market also has insufficient information to keep managerial decisions at arm’s length. Furthermore, equity-based executive incen- tive contracts, such as stock options and performance-based stock grants, are rarely used in Chinese listed firms. Wei, Xie, and Zhang (2005) report that senior managers and directors have an average stock holding of only 0.015% for partially privatised SOEs. Inefficient corporate governance may cause the managers to pursue their personal interests, such as building an empire and enjoying a quiet life, which subse- quently leads to investment inefficiency (Jensen, 1986; Jensen & Meckling, 1976). 2.2. Determinants of investment efficiency In a perfect world with a frictionless capital market as envisioned by Modigliani and Miller (1958), capital is allocated in such a way that the marginal product of capital is similar across every project in the economy. Managers obtain financing at the prevailing interest rate and undertake all projects with a positive net present value. Empirically, the investment of a firm should be solely determined by the profitability of its investment as measured by Tobin’s Q (Tobin, 1969). However, several distorting forces in the real world may drive firms to deviate from their optimal investment level, leading to underinvestment and overinvestment. Among these, the most pervasive and important factors affecting the efficiency of capital investment are those that arise from informational asymmetries and agency conflicts (Stein, 2003). With regard to agency costs, the separation of ownership and control creates a con- flict of interest between shareholders and managers, and insufficient monitoring may lead the managers to pursue private objectives that may be in conflict with those of out- side shareholders (Jensen & Meckling, 1976). Jensen (1986) and Stulz (1990) argue that empire-building preferences may cause managers to spend all of their available funds on investment projects, which leads to overinvestment. Blanchard, Lopez-de-Silanez, and Shleifer (1994) examine how a small sample of firms responds to a large cash windfall coming from legal settlements, a source of cash that does not change its investment opportunity set. They find that managers typically spend the cash on acquisitions rather than turning over the windfalls to their shareholders. Besides, the managers who prefer to enjoy a quiet life tend to invest less because investing requires these managers to oversee the investment. When firms expand their existing facilities or start new product lines, the managers are required to fulfil more duties or spread their talents over a wider array of activities. Therefore, managers generally prefer to invest less or to underinvest in the sense that they decide to forgo some projects with positive net present value (Betrand & Mullainathan, 2003). 232 Xiong and Su Models of costly external financing generally predict underinvestment because of the informational asymmetry of adverse selection. Acting in the interest of current shareholders, managers try to sell new shares when their private information suggests that these shares are overpriced; however, rational investors will discount these new issues, thereby discouraging firms with investment opportunities from raising equity capital even at the expense of giving up projects with positive net present value (Myers & Majluf, 1984). These findings suggest that costly external financing brought about by information asymmetries and opportunistic behaviour of managers may lead to overinvestment and underinvestment. As regards the investment efficiency of Chinese listed companies, Zhang and Song (2009) find that the optimal investment ratio is 24.4% of fixed assets at the start of an accounting year, and that 39.26% of the sample companies overinvest, while 60.74% of the sample companies underinvest. Tang et al. (2007) observe widespread overin- vestment among Chinese listed firms from 2000 to 2002. They also identify cash divi- dends, liabilities, and corporate governance as effective restrictions on overinvestment. Xin et al. (2007) investigate the governance effect of executive compensation on capital investment decision by using a sample of Chinese firms that were listed from 2002 to 2004. They find that government control, particularly that over SOEs affiliated with local government and state asset-management bureaux, induces overinvestment when the diligence and talent of executives are not being compensated. Li (2009) suggests that accounting information quality can improve capital allocation efficiency by mitigat- ing adverse selection due to information asymmetry between managers and investors as well as by curbing the incentives of managers to engage in value-destroying activities, such as empire building. By using a sample of Chinese firms listed between 2004 and 2006, Li (2009) finds that those firms with high-quality financial reporting have a reduced tendency to deviate from the predicted investment level, thereby mitigating both overinvestment and underinvestment. Cheng et al. (2008) and Zhong et al. (2010) hypothesise that the Chinese government intervenes in SOEs to accomplish its social and political goals, such as boosting regional economic growth, increasing regional rev- enues and employment rates, and maintaining social stability, which influences the investment behaviour of firms and leads to investment inefficiency. Using Chinese data, Chen et al. (2011) investigate whether government intervention distorts the investment behaviour of firms and leads to investment inefficiency. They find that the sensitivity of investment expenditure to investment opportunities is significantly weaker for SOEs, and that political connections significantly reduce the investment efficiency in these enterprises. Lv and Zhang (2011) find that the managerial stock incentive in Chinese publicly traded firms helps mitigate underinvestment and overinvestment. 2.3. Development of hypothesis As the central element of market microstructure, stock liquidity is generally described as the ability to trade large quantities quickly at low cost and with limited effects on prices. Earlier studies primarily focus on stock liquidity and asset pricing (e.g., Amihud, 2002; Amihud & Mendelson, 1986; Pastor & Stambaugh, 2003), whereas recent studies investigate the effect of stock liquidity on corporate governance and find that stock liquidity helps optimise CEO compensation and enhances the ability of investors to monitor managerial decisions (Adamati & Pfleiderer, 2009; Edmans, 2009; Jayaraman & Milbourn, 2012). These findings introduce a new question of whether stock liquidity affects investment inefficiency. China Journal of Accounting Studies 233 First, stock liquidity has been argued to reduce the cost of equity capital and relax financial constraints through asset pricing and positive feedback mechanisms. Amihud and Mendelson (1986) suggest that stock transaction costs must be considered when valuing equity investment, and that liquidity decreases the required rates of return for equity investment. Amihud (2002), Amihud and Mendelson (1986), and Pastor and Stambaugh (2003) employ different liquidity measures to test the relationship between liquidity and stock return, which is eventually found to be negative. Khanna and Sonti (2004) find that informed traders factor the effect of their trades on managerial behaviour into their trading strategies, which drives them to trade more aggressively. This practice increases the information content of prices, which improves the operating performance and relaxes the financial constraints of firms. The expected return is equal to the discount rate of investment opportunities (Ross, Westerfield, & Jaffe, 2012); thus, firms with liquidity stock maintain a low discount rate, which in turn improves the profitability of investment projects. At the same time, the feedback effect from price improves the operating performance and relaxes the financial constraints of these firms, which subsequently drives these firms to expand their investments and mitigate their underinvestment. Second, stock liquidity improves investment efficiency by helping optimise CEO compensation and enhancing the ability of investors to monitor managerial decisions. Holmstrom and Tirole (1993) and Kyle (1985) model informed traders who optimally select their trading intensity as a function of stock liquidity,thereby determining the amount of private information that is impounded in the stock price. At equilibrium, higher stock liquidity drives increased informed trading to dominate over increased unin- formed trading, which increases the amount of private information that is incorporated into the stock price. With regard to the increased informativeness of the stock price, Holmstrom and Tirole (1993) argue that firms can offer steeper stock-based incentives to senior managers. Jayaraman and Milbourn (2012) find that the pay-for-performance sensitivity of CEOs to stock prices increases along with the liquidity of stock. An optimal managerial contract induces senior managers to explore investment opportu- nities with positive net present value and forgo projects with negative net present value, which subsequently improves firm performance and remuneration. Aggarwal and Samwick (2006) develop a model and empirically analyse the relationships among the incentives from compensation, investment, and firm performance. They find that the pay-for-performance sensitivity of CEOs to stock prices can mitigate managerial shirking and increase capital expenditure. These results, which are consistent with the models of underinvestment, sup- port the argument that managers have private costs of investment (Betrand & Mullainathan, 2003). Lv and Zhang (2011) examine how the managerial stock incentive mechanism of China affects corporate investment and find that such a mechanism helps mitigate inefficient investment, which subsequently mitigates both underinvestment and overinvestment. Moreover, stock liquidity can improve the ability of blockholders to monitor and enhance the power of exit threats. These blockholders monitor and trade with the aim of profiting from the price appreciation caused by their monitoring activities. Maug (1998) concludes that higher stock liquidity facilitates highly effective monitoring by allowing blockholders to purchase additional shares at a price that does not reflect the benefits of intervention. Adamati and Pfleiderer (2009) and Edmans (2009) argue that when the compensation of managers is tied to current stock prices, increased stock liquidity also increases the cost of opportunism to managers by facilitating informed selling. Managers will undertake productive efforts and investments to improve the firm value and dissuade blockholders from exiting. 234 Xiong and Su Third, improved stock liquidity helps informed parties (speculators) to disguise their private information and to profit from such information (Kyle & Vila, 1991). An increase in the marginal value of information drives speculators to spend more time on monitoring activities. The increased flow of information into the market improves the information content of the stock prices. These stock prices with large amounts of private information can provide managers with a greater quantity of new information about the prospects of their own firms, such as product demand and strategic issues, which in turn affect the investment decisions of managers (Chen, Goldstein, & Jiang, 2007). Moreover, the enhanced informativeness of the stock prices can reduce informa- tion asymmetries between managers and external shareholders as well as enhance the ability of investors to monitor the managers, which in turn improves the capital allocation of the firm (Durnev, Morck, & Yeung, 2004; Yang & Nie, 2010). Yang and Nie (2010) investigate the relationship between stock price informativeness and capital allocation efficiency by using a sample of Chinese companies listed in 2001 and 2008. They find that stock price informativeness is negatively associated with the underinvest- ment and overinvestment of firms that arise from abusing free cash flow. These findings lead to the following hypothesis: H: Firms with higher liquidity show lower investment inefficiency. 3. Research Design 3.1. Model specification By analogy with the work of Xin et al. (2007) and Zhong et al. (2010), we estimate the following multivariate panel data regressions to test the effect of stock liquidity on investment efficiency: INEFFINV ¼ / þ / LIQ þ / FCF þ / MCOST þ / ORECTA þ / SIZE þ / LEV it 0 1 it1 2 it 3 it 4 it 5 it 6 it X X þ / Q þ / INDUSTRY DUMMIES þ / YEAR DUMMIES þ 1 it it 7 k t k t (1) where the dependent variable INEFFINV represents investment inefficiency for firm i it in year t. LIQ represents different proxies of stock liquidity for firm i in year t–1. To it-1 alleviate any concern for potential endogeneity, we use lagged liquidity as an indepen- dent variable. The rest are control variables that may influence investment inefficiency: free cash flow (FCF), agency costs (MCOST), other receivables (ORECTA), size (SIZE), leverage (LEV), Tobin’sQ(Q), industry and year dummies. See Appendix A for definitions of variables. We estimate equation (1) using ordinary least squares (OLS). Following Peterson (2009), we adjust the standard errors for heteroscedasticity and within-firm serial corre- lation using cluster at the firm level. Since our hypothesis predicts that stock liquidity mitigates investment inefficiency, we expect the coefficient estimates for / is signifi- cantly negative. 3.2. Variable measures 3.2.1. Firm investment Richardson (2006) splits total investment I into two main components: TOTAL,t (i) required investment expenditure to maintain assets in place, I , and MAINTAIN China Journal of Accounting Studies 235 (ii) investment expenditure on new projects, I . Following Lv and Zhang (2011), we NEW define total investment, I , as change in gross value of fixed assets, construction TOTAL,t in progress, project material and intangible assets, and I as amortisation and MAINTAIN depreciation. Firm investment (INV) is the difference between total investment I TOTAL and maintenance investment I , scaled by lagged total assets. MAINTAIN Chinese listed companies report cash payments for fixed assets, intangible assets, and other long-term assets, and cash receipts from selling these assets in their cash flow statement; therefore, we use the difference between them as our measure for total investment. Based on this, we construct a new firm investment variable (INV2) as the difference between new total investment and maintenance investment I MAINTAIN (namely, amortisation and depreciation). 3.2.2. Investment inefficiency Conceptually, investment efficiency refers to firms undertaking all, and only, projects with positive net present value. Richardson (2006) and Biddle, Hilary, and Verdi (2009)define investment inefficiency as deviations from expected investment using a model that predicts investment as a function of firm characteristics. Thus, deviations from expected investment represent investment inefficiency. Applying the equation used in Richardson (2006) and Xin et al. (2007), we estimate expected investment according to the following regression: INV ¼ a þ a SIZE þ a LEV þ a CASH þ a AGE þ a RET þ a Q it 0 1 it1 2 it1 3 it1 4 it1 5 it1 6 it1 X X þa INV þ a INDUSTRY DUMMIES þ a YEAR DUMMIES þ l 7 it1 k t it (2) where INV is investment expenditure for firm i in year t. The explanatory variables it include firm size (SIZE), leverage (LEV), the level of cash holdings (CASH), firm age (AGE), past sock return (RET ), Tobin’sQ(Q) and prior level of investment (INV ). it–1 it–1 Appendix A provides definitions of variables. We estimate equation (2) using the panel data firm-fixed effect model. The residu- als from the regression models represent the deviations from expected investment level, and we use the absolute value of these residuals as a firm-specific proxy for investment inefficiency (INEFFINV ). Thus, a higher value means higher inefficiency. A positive it residual means that companies invest above the optimal level. We define the residual as overinvestment (OverINV). A negative residual is regarded as underinvestment. To ease the exposition, we multiply the residual by –1 and define it as underinvestment (UnderINV), so that a higher value indicates more severe underinvestment. 3.2.3. Stock liquidity Liquidity is generally described as the ability to trade large quantities quickly at low cost with little price impact. This description highlights four dimensions to liquidity, namely, trading quantity, trading speed, trading cost, and price impact. Considering data availability and computing cost, we employ five commonly used measures of liquidity. Bid-ask spread (HL). Corwin and Schultz (2012) propose a bid-ask spread estimator based on two ideas. First, daily high (low) prices are almost always buyer (seller) -initi- ated trades. Therefore, the ratio of high-to-low prices for a day reflects both the 236 Xiong and Su fundamental volatility of the stock and its bid-ask spread. Second, the component of the high-to-low price ratio that is attributed to volatility increases proportionately with the length of the trading interval, while the component due to bid-ask spreads stays constant over a short period. In addition, Corwin and Schultz (2012) find that this high-low measure outperforms other commonly used low-frequency measures. There- fore, we use this high-low spread measure (HL) as a stock liquidity measure and com- pute it using two consecutive days as follows: it 2eðÞ 1 HL ¼ (3) it it 1 þ e where 8 9 "# ! pffiffiffiffiffiffiffiffi pffiffiffiffiffi rffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi < 1 o = 2b  b c it;dþj it it it pffiffiffi pffiffiffi a ¼  , b = E ln , it it : ; 3  2 2 3  2 2 it;dþj j¼0 "# ! it;d;dþ1 o o c ¼ ln : H L it it;d it;d it;d;dþ1 denotes the actual high (low) stock price for stock i on day d of year t, o o H L is the high price over days t and t+1. A higher value of HL means it;d;dþ1 it;d;dþ1 lower stock liquidity. Turnover ratio (TOVER). The second measure is the tradable turnover ratio (TOVER), which is defined as the average of daily turnover ratio for a firm during the year. Daily turnover ratio is total shares traded in the day divided by total tradable shares. Considering that a large portion of shares (state shares and legal entity) in a typical Chinese listed company is not allowed to be traded and the total turnover would be biased (Wang & Chin, 2004), we use the tradable turnover ratio. it 1 VOL itd TOVER ¼ (4) it D LNS it itd d¼1 where VOL is the number of shares traded on day d of year t for stock i, LNS is itd itd total tradable A shares, D is equal to the number of days for stock i in year t. it Amihud illiquidity ratio (ILLIQ). The third measure is Amihud’s(2002) illiquidity ratio, which captures Kyle’s(1985) conception of illiquidity, i.e., the response of price to order flow. Specifically, the illiquidity ratio for stock i in year t is equal to it 1 jj r itd ILLIQ ¼  100 (5) it D V it itd d¼1 where r and V are stock i’s return and dollar volume (in millions) on day d in year itd itd t, respectively. D is equal to the number of days for stock i in year t. |r |/V it itd itd measures the price change induced by order flow. Since the Amihud ratio measures illiquidity, the ratio is smaller for more liquid stocks. Amivest liquidity ratio (LR). The Amivest liquidity ratio measures the trading volume associated with a unit change in the stock price. A higher value of LR implies greater market liquidity or depth. The liquidity ratio is defined as follows: it 1 V itd LR ¼  10 (6) it D jj r it itd d¼1 China Journal of Accounting Studies 237 Return reversal measure (GAM). Pastor and Stambaugh (2003) point out that order flow should be accompanied by a return that one expects to be partially reversed in the future if the stock is not perfectly liquid. They argue that therefore the greater the expected reversal for a giver dollar volume, the lower the stock’s liquidity. Thus, we estimate the return reversal measure (GAM) by running the following regression: e e r ¼ h þ / r þ c  Signðr Þ V þ e (7) i;t i;t;d i;t;d i;t;dþ1 i;t;dþ1 i;t i;t i;t;d where r is stock’s excess return above market return on day d. Return reversal mea- i;t;d sure GAM is defined as the absolute value of γ , GAM ¼ c . A higher GAM implies i,t i;t greater price impact and lower liquidity. 3.2.4. Free cash flow Jensen (1986) argues that free cash flow is cash flow beyond what is necessary to finance all projects with positive net present value. Therefore, we define free cash flow as cash flow from operations minus expected investment level: FCF ¼ CFO  INV (8) it it it where CFO represents the cash flow from operations, INV is the expected level of it investment, which is equal to the fitted values of equation (2). 3.2.5. Agency costs Ang, Coles, and Lin (2000) and Singh and Davidson (2003) use a firm’s selling, general, and administrative (SG&A) expenses to capture equity agency costs. They point out that SG&A expenses represent the costs related to the management function and sale of products, includes managerial salaries, rents, insurance, utilities, supplies, and advertising costs. Higher levels of SG&A expenses are close approximations of managerial pay and perquisite consumption. These costs, to a large extent, reflect managerial discretionary expenses and may be a closer proxy for agency costs (Singh & Davidson, 2003, p. 794). Chinese listed companies report selling expenses, and general and administrative expenses in their income statement, therefore, we use the sum of selling expenses and general and administrative expenses scaled by total assets (MCOST) as our measure for agency costs. 3.2.6. Stock price informativeness Following Chen et al. (2007), we use stock price non-synchronicity as stock price informativeness. The construction of stock price informativeness (SYN) consists of two steps. First, we estimate equation (9) to decompose the variation of a stock return into two components: systematic risk and idiosyncratic volatility. r ¼ a þ b r þ c r þ e (9) it;w it mt;w jt;w it;w it it where r is stock return for firm i on week w of year t, r and r represent market it,w mt,w jt,w and industry return, respectively. R reflects the proportion of volatility that is it explained by systematic components. A high value of R indicates that firms’ stock it returns are closely tied to market and industry returns, and are assumed to reflect rela- tively less firm-specific information. Thus, 1–R measures stock price informativeness. it In the second step, we use a logistic transformation of 1–R to construct SYN: it 238 Xiong and Su 2 2 SYN ¼ Ln½ð1  R Þ=R  (10) it it it A higher value of SYN implies greater information content of share prices. 3.2.7. Other control variables Following Xin et al. (2007) and Yang and Nie (2010), we introduce several control variables in our model. As a proxy for tunnelling by a controlling shareholder, we use other receivables scaled by total assets (ORECTA); to control for the size, we use natu- ral logarithm of total assets (SIZE); to measure growth options we include Tobin’sQ (Q) as the sum of market value of tradable shares, book value of non-tradable shares, and liabilities, scaled by book value of total assets. We also control for the industry and year dummies. 3.3. Sample and descriptive statistics Our sample is drawn from firms listed on the Shenzhen and Shanghai stock exchanges over the period from 1998 to 2011. We exclude financial firms because their liabilities are not strictly comparable to those in other industries. The stock price, return, trading volume and accounting information data are extracted from the China Stock Market and Accounting Research (CSMAR) database commercially available from Shenzhen GTA Information Co. Ltd. Ownership and corporate governance data are gathered from China Center for Economics Research (CCER) China stock database. We winsorize each continuous variable at the first and 99th percentiles to mitigate the influence of outliers. Table 1 reports descriptive statistics for the variables, including the mean, standard deviation, median, minimum, and maximum. Firm investment (INV) in the sample has a mean of 0.021 and a median of –0.004, while standard deviation of investment is 0.136, indicating that there is significant variation among firm investment. The mean and standard deviation of the Amihud illiquidity ratio (ILLIQ) are 0.305 and 0.395, respectively, suggesting a great variation between firm stock liquidity. 4. Empirical results 4.1. The effect of stock liquidity on investment inefficiency Table 2 reports panel regression estimates of equation (1) using different stock liquidity measures. In the first column, we use the bid-ask spread (HL) developed by Corwin and Schultz (2012); in the second, the illiquidity ratio (ILLIQ) proposed by Amihud (2002); in the third, the return reversal measure (GAM)defined by Pastor and Stamb- augh (2003); in the fourth, the tradable turnover ratio (TOVER); in the fifth column, the Amivest liquidity ratio (LR). The dependent variable of the regressions is investment inefficiency (INEFFINV) based on INV. As shown in Table 2, when we use illiquidity ratio and return reversal to proxy for stock liquidity, the coefficient estimates for ILLIQ and GAM are all significantly posi- tive at the 1% level, indicating that firms with higher stock liquidity show higher investment efficiency, which supports our hypothesis. The economic magnitude of the liquidity effect is also meaningful. A 10 percentage-point decrease in ILLIQ and GAM will reduce investment inefficiency by 21% and 25%, respectively. China Journal of Accounting Studies 239 Table 1. Descriptive statistics. Variable Observation Mean Standard deviation Median Minimum Maximum INV 16,475 0.021 0.136 –0.004 –0.307 0.710 HL 19,196 0.046 0.013 0.044 0.023 0.083 TOVER 19,354 2.738 2.546 2.013 0.293 16.073 ILLIQ 19,355 0.305 0.395 0.158 0.005 2.286 LR 19,366 2.291 3.526 1.094 0.103 26.096 GAM 19,196 0.102 0.158 0.042 0.000 0.951 Q 19,465 1.683 1.026 1.360 0.812 7.459 CFO 19,685 0.044 0.083 0.043 –0.216 0.279 SIZE 19,687 21.277 1.203 21.137 10.842 28.282 LEV 19,687 0.496 0.290 0.477 0.055 2.253 CASH 19,687 0.178 0.148 0.136 0.003 0.723 AGE 19,690 7.599 4.743 7.000 1.000 19.000 RET 17,723 0.271 0.878 –0.055 –0.746 3.717 MCOST 19,605 0.130 0.244 0.075 0.008 1.956 ORECTA 19,686 0.054 0.091 0.019 0.000 0.554 This table reports the descriptive statistics of the main variables used in our multivariate analysis for our full sample of firms. INV is firm investment. HL is bid-ask spread of Corwin and Schultz (2012), TOVER is turn- over ratio, ILLIQ is Amihud’s(2002) illiquidity measure, LR is Amivest liquidity ratio, and GAM is return reversal developed by Pastor and Stambaugh (2003). Q is Tobin’sq, CFO is cash flow from operations scaled by total assets, SIZE is the natural logarithm of total assets, LEV is total liabilities scaled by total assets, CASH is the ratio of cash holdings to total assets, AGE is the number of years after going public, RET is stock return, MCOST is the sum of selling expenses and general and administrative expenses scaled by total assets, and ORECTA is other receivables scaled by total assets. When we use the tradable turnover ratio (TOVER) as stock liquidity, the coefficient for TOVER is 0.02 and is significant at the 10% level, suggesting a positive association between liquidity and investment inefficiency, which contradicts our hypothesis. The possible reason is that the turnover ratio may reflect investor sentiment. Baker and Stein (2004) argue that high liquidity is a symptom of the fact that the market is domi- nated by irrational investors, and hence is overvalued. Managers who try to boost short-run share price have an incentive to waste resources in projects with negative NPV to cater for current sentiment when stock price is overpriced, with subsequent overinvestment (Polk & Sapienza, 2009). 4.2. The effect of stock liquidity on underinvestment and overinvestment The results in Table 2 raise a further question of whether higher stock liquidity is asso- ciated with a reduction of overinvestment or with a reduction of underinvestment. Based on this, we distinguish underinvestment and overinvestment scenarios. Tables 3 and 4 present results of the estimates of equation (1) using underinvestment (UnderINV) and overinvestment (OverINV) respectively as dependent variables. As shown in Table 3, the coefficient estimates for TOVER and LR are all signifi- cantly negative at the 10% level or above, indicating that firms with higher stock liquidity show lower underinvestment. The results appear economically significant as well. A 10 percentage-point increase in TOVER and LR leads to a decrease in underin- vestment of 1% and 2%, respectively. As shown in Table 4, the coefficient estimates for HL, ILLIQ,and GAM are all significantly positive at 5% or above, suggesting that stock liquidity is negatively associated with overinvestment. In particular, a 10 percentage-point decrease in ILLIQ 240 Xiong and Su Table 2. Regressions of investment inefficiency on stock liquidity. (i) (ii) (iii) (iv) (v) HL ILLIQ GAM TOVER LR LIQ 0.199 0.021*** 0.025*** 0.002* 0.000 it-1 (0.139) (0.004) (0.006) (0.001) (0.000) FCF –0.000*** –0.000*** –0.000*** –0.000*** –0.000*** (0.000) (0.000) (0.000) (0.000) (0.000) ORECTA –0.001 –0.004 –0.003 0.002 –0.001 (0.009) (0.011) (0.011) (0.011) (0.011) SIZE 0.001 0.004*** 0.002 0.002 (0.001) (0.002) (0.002) (0.002) Q 0.008*** 0.009*** 0.009*** 0.008*** 0.007*** (0.001) (0.001) (0.001) (0.001) (0.001) Intercept 0.048* –0.017 0.032 0.044 0.089*** (0.027) (0.037) (0.035) (0.035) (0.006) Obs. 13,867 13,870 13,867 13,870 13,880 R 0.046 0.049 0.047 0.0450 0.045 F 17.46 12.57 12.28 11.61 11.84 This table reports results from regressing investment inefficiency on stock liquidity. All the estimates have been carried out using OLS. All models include year and industry fixed effects and constant term. Robust standard errors that are clustered by firm are present in parentheses under the coefficients. HL is bid-ask spread of Corwin and Schultz (2012), TOVER is turnover ratio, ILLIQ is Amihud’s(2002) illiquidity measure, LR is Amivest liquidity ratio, and GAM is return reversal developed by Pastor and Stambaugh (2003). FCF is free cash flow scaled by total assets, ORECTA is other receivables scaled by total assets, SIZE is the natural logarithm of total assets, LEV is total liabilities scaled by total assets, and Q is Tobin’s q. ***, **, and *denote significance at the 1%, 5%, and 10% levels, respectively. and GAM will reduce overinvestment by 39% and 42%, respectively. The results of Tables 3 and 4 show that stock liquidity improves investment efficiency, mitigating both overinvestment and underinvestment. 4.3. Robustness checks In this section, we perform several additional robustness tests of the reported results. 4.3.1. Endogeneity issues The OLS regressions are prone to the potential endogeneity problem, as improved investment efficiency may increase stock liquidity. To alleviate this concern, we use lagged stock liquidity to estimate the regression equation. In this subsection, we replace it with a second lag of stock liquidity and re-estimate equation (1). The results using a second lag of stock liquidity are similar to those previously reported, as displayed in Panel A of Table 5, supporting the hypothesis that higher stock liquidity helps improve investment efficiency. 4.3.2. Influence of funding constraints Brunnermeier and Pedersen (2009) show that stock liquidity reflects the funding avail- ability in capital markets, therefore, stock liquidity can influence firm investment through funding constraints. To distinguish this channel, following Goyenko and Ukhov (2009), we control for the growth rate of M2 and the Shanghai Interbank offered Rate China Journal of Accounting Studies 241 Table 3. Regressions of underinvestment on stock liquidity. (i) (ii) (iii) (iv) (v) HL ILLIQ GAM TOVER LR LIQ 0.023 0.002 0.004 –0.001* –0.002*** it-1 (0.121) (0.002) (0.005) (0.001) (0.000) FCF –0.089*** –0.089*** –0.089*** –0.089*** –0.114*** (0.009) (0.009) (0.009) (0.009) (0.009) ORECTA 0.007 0.007 0.007 0.008 0.037*** (0.009) (0.009) (0.009) (0.009) (0.009) SIZE –0.016*** –0.016*** –0.016*** –0.017*** (0.001) (0.001) (0.001) (0.001) Q 0.009*** 0.009*** 0.009*** 0.009*** 0.016*** (0.001) (0.001) (0.001) (0.001) (0.001) Intercept 0.416*** 0.421*** 0.412*** 0.428*** 0.044*** (0.024) (0.024) (0.023) (0.023) (0.005) N 7,856 7,859 7,856 7,859 7,864 R 0.191 0.191 0.191 0.192 0.159 F 37.30 37.72 37.36 37.57 30.65 This table reports results from regressing underinvestment on stock liquidity. All the estimates have been carried out using OLS. All models include year and industry fixed effects and constant term. Robust standard errors that are clustered by firm are present in parentheses under the coefficients. HL is bid-ask spread of Corwin and Schultz (2012), TOVER is turnover ratio, ILLIQ is Amihud’s(2002) illiquidity measure, LR is Amivest liquidity ratio, and GAM is return reversal developed by Pastor and Stambaugh (2003). FCF is free cash flow scaled by total assets, ORECTA is other receivables scaled by total assets, SIZE is the natural logarithm of total assets, LEV is total liabilities scaled by total assets, and Q is Tobin’s q. ***, **, and *denote significance at the 1%, 5%, and 10% levels, respectively. (Shibor) of 7 days (RATE). As shown in Panel B of Table 5, the coefficients on M2 are significantly negative at the 10% level, suggesting that monetary expansions relax fund- ing constraints and mitigate underinvestment. More importantly, our results remain unchanged after accounting for funding constraints. Stock liquidity is negatively associ- ated with investment inefficiency. 4.3.3. Influence of information environment Considering that stock liquidity reflects information asymmetries in the capital market, and high-quality financial reporting increases investment efficiency (Biddle et al., 2009; Li, 2009), a credible alternative interpretation of our results is that they may capture the effect of information asymmetries. To exclude this explanation, we re-estimate equation (1) after controlling for the absolute value of discretionary accruals using the performance matched discretionary accrual model (Kothari, Leone, & Wasley, 2005). Panel C of Table 5 reports the results. The coefficient estimates forjj DA are significant and positive at the 1% level, indicating that financial reporting quality mitigates invest- ment inefficiency, which is consistent with Biddle et al. (2009). More importantly, our previous findings are not affected. Firms with higher liquidity show higher investment efficiency. 4.3.4. Alternative measure of firm investment Following Chen et al. (2011) and Lv and Zhang (2011), we define cash payments for fixed assets, intangible assets, and other long-term assets minus cash receipts from 242 Xiong and Su Table 4. Regressions of overinvestment on stock liquidity. (i) (ii) (iii) (iv) (v) HL ILLIQ GAM TOVER LR LIQ 0.575** 0.039*** 0.042*** 0.005*** 0.000 it-1 (0.280) (0.008) (0.014) (0.001) (0.001) FCF –0.067*** –0.073*** –0.072*** –0.073*** –0.035 (0.023) (0.023) (0.023) (0.023) (0.023) ORECTA –0.096*** –0.107*** –0.099*** –0.086*** –0.112*** (0.023) (0.023) (0.023) (0.023) (0.025) SIZE 0.013*** 0.017*** 0.014*** 0.015*** (0.002) (0.002) (0.002) (0.002) Q –0.001 0.001 0.000 0.000 –0.008*** (0.003) (0.003) (0.003) (0.003) (0.003) Intercept –0.155*** –0.247*** –0.161*** –0.183*** 0.156*** (0.050) (0.053) (0.049) (0.051) (0.014) N 6,011 6,011 6,011 6,011 6,016 R 0.075 0.080 0.076 0.075 0.066 F 10.20 10.52 10.39 10.33 7.139 This table reports results from regressing overinvestment on stock liquidity. All the estimates have been car- ried out using OLS. All models include year and industry fixed effects and constant term. Robust standard errors that are clustered by firm are present in parentheses under the coefficients. HL is bid-ask spread of Corwin and Schultz (2012), TOVER is turnover ratio, ILLIQ is Amihud’s(2002) illiquidity measure, LR is Amivest liquidity ratio, and GAM is return reversal developed by Pastor and Stambaugh (2003). FCF is free cash flow scaled by total assets, ORECTA is other receivables scaled by total assets, SIZE is the natural loga- rithm of total assets, LEV is total liabilities scaled by total assets, and Q is Tobin’s q. ***, **, and *denote significance at the 1%, 5%, and 10% levels, respectively. selling these assets as total investment. The new firm investment measure is equal to total investment minus maintenance investment (namely, amortisation and depreciation), and then scaled by total assets. Based on this, we re-estimate expected investment model (2) and investigate the relationship between stock liquidity and investment ineffi- ciency. Our results, as shown in Panel D of Table 5, are similar to those previously reported. Stock liquidity is negatively associated with investment inefficiency. 4.3.5. Tobit models We use the absolute value of residuals of investment model (2) to measure underinvest- ment and overinvestment. Since the distribution of underinvestment and overinvestment are bounded to [0, +∞), the OLS would be biased, thus we use the Tobit model to esti- mate the panel regression of equation (1). The untabulated results are similar to the results reported in Tables 3 and 4. 4.3.6. Investment efficiency based on group classification The mean value of the residuals of the expected investment model (2) is zero, implying that on average investment inefficiency is zero. However, most firms may deviate from their optimal investment level, which contradicts the assumption of model (2). To alle- viate this concern, following Xin et al. (2007) and Biddle et al. (2009), we sort firms each year into three groups, based on the residuals from equation (2). Firm-year obser- vations in the bottom group (i.e., the most negative residuals) are classified as under- investing, while observations in the top group (i.e., the most positive residuals) are China Journal of Accounting Studies 243 Table 5. Stock liquidity and investment inefficiency: robustness checks. (i) (ii) (iii) (iv) (v) HL ILLIQ GAM TOVER LR Panel A: Lagged two periods stock liquidity and investment inefficiency LIQ 0.182 0.018*** 0.035*** 0.002*** 0.000 it-2 (0.130) (0.003) (0.007) (0.001) (0.001) N 13,205 13,302 13,205 13,301 13,311 R 0.044 0.047 0.046 0.044 0.043 Panel B: Stock liquidity and investment inefficiency controlled for M2 and interest rates LIQ 0.185 0.020*** 0.024*** 0.001* 0.000 it-1 (0.131) (0.004) (0.006) (0.001) (0.000) M2 –0.001** 0.000 –0.001* –0.001** –0.001 (0.001) (0.000) (0.000) (0.000) (0.000) RATE –0.002 0.002 –0.001 –0.003 –0.001 (0.003) (0.003) (0.003) (0.003) (0.003) N 13,867 13,870 13,867 13,870 13,880 R 0.046 0.049 0.047 0.046 0.045 Panel C: Stock liquidity and investment inefficiency controlled for information opaque LIQ 0.240* 0.020*** 0.022*** 0.002* 0.000 it-1 (0.131) (0.004) (0.006) (0.001) (0.000) DA 0.046*** 0.044*** 0.044*** 0.047*** 0.047*** jj (0.005) (0.011) (0.011) (0.011) (0.011) N 13,838 13,840 13,838 13,840 13850 R 0.050 0.053 0.051 0.050 0.050 Panel D: Stock liquidity and investment inefficiency based on INV2 LIQ –0.0370 0.006*** 0.006* 0.001** 0.000 it-1 (0.072) (0.002) (0.004) (0.001) (0.000) N 12,193 12,194 12,193 12,194 12,205 R 0.046 0.047 0.047 0.047 0.045 Panel E: Effect of liquidity on investment inefficiency using group classification LIQ 0.291* 0.021*** 0.020*** 0.002** 0.000 it-1 (0.171) (0.004) (0.007) (0.001) (0.001) N 9,307 9,313 9,307 9,313 9,318 R 0.049 0.052 0.049 0.049 0.048 Panel F: Alternative investment efficiency model LIQ 0.176 0.013*** 0.015** 0.003*** –0.001*** it-1 (0.118) (0.004) (0.006) (0.000) (0.000) N 15,471 15,600 15,471 15,599 15,610 R 0.057 0.058 0.057 0.059 0.057 This table reports results of robustness checks. All models include other control variables, constant term, year and industry fixed effects. Other control variables are omitted for brevity. Robust standard errors that are clus- tered by firm are present in parentheses under the coefficients. HL is bid-ask spread of Corwin and Schultz (2012), TOVER is turnover ratio, ILLIQ is Amihud’s(2002) illiquidity measure, LR is Amivest liquidity ratio, and GAM is return reversal developed by Pastor and Stambaugh (2003). M2 is the growth rate of M2, RATE is the Shanghai Interbank offered Rate (Shibor) of 7 days, and DA is the discretionary accruals using Kothari et al. (2005) model. ***, **, and * denote significance at the 1%, 5%, and 10% levels, respectively. 244 Xiong and Su classified as over-investing. We then estimate equation (1) using the absolute value of residuals that belong to the underinvestment and overinvestment group. The results, as presented in Panel E of Table 5, are similar to those previously reported. 4.3.7. Alternative investment efficiency model Following Biddle et al. (2009), we measure investment efficiency using a parsimonious expected investment model, which predicts investment level based on growth opportu- nities (as measured by Tobin’s Q) and use the absolute value of residuals as a proxy for investment inefficiency. The model is described in equation (11): X X INV ¼ a þ a Q þ a INDUSTRY DUMMIES þ a YEAR DUMMIES þ l it 0 1 it1 k t it k t (11) The results of estimating equation (1) using this investment efficiency proxy are simi- lar to those previously reported, as displayed in Panel F of Table 5. Higher stock liquidity helps mitigate investment inefficiency, mitigating both overinvestment and underinvestment. 5. How does liquidity improves investment efficiency? In this section, we perform several tests to examine the various explanations for why liquidity improves investment efficiency from the perspective of financial constraints, agency costs and information content of share prices. According to Baron and Kenny (1986) and Wen, Chang, Hau, and Liu (2004), we propose the following recursive regression equations to test the mediation effects of financial leverage, agency costs and information content of share prices: 8 P P P INEFFINV ¼ / þ / LIQ þ / Control þ / INDUSTRY þ / YEAR þ 1 ð12Þ > it 0 1 it1 j j k t it j k t P P P > MEDIATOR ¼ h þ h LIQ þ h Control þ h INDUSTRY þ h YEAR þ s ð13Þ it 0 1 it1 j j k t it j k t P P 0 0 0 0 0 > INEFFINV ¼ / þ / LIQ þ / MEDIATOR þ / Control þ / INDUSTRY it it1 it j 0 1 2 j k > k þ / YEAR þ x ð14Þ t it The rationale behind this method is as follows. We first fit equation (12) and obtain the coefficient estimates for LIQ . A statistically significant ϕ suggests that stock liquid- it–1 1 ity influences investment efficiency. We then estimate equations (13) and (14). If both θ and / are statistically significant, then the relation between stock liquidity and investment efficiency is intermediated by the mediating variable. Furthermore, when 0 0 the coefficient estimate for / of equation (14) is significant, an insignificant / means 2 1 that the effect of liquidity on investment inefficiency is mediated entirely by the medi- ating variable. If either θ or / is statistically insignificant, we can use the Sobel (1986) test to check the mediation effect by examining the significance of the product 0 0 of coefficients h / . The standard error of h / , derived by the multivariate delta 1 1 2 2 qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi 2 2 0 2 method (Sobel, 1986), is equal to S ¼ h S þðÞ / S . 1 0 h / 2 1 h 2 / 1 2 China Journal of Accounting Studies 245 Table 6. Mediating effect tests. (i) (ii) (iii) (iv) (v) HL ILLIQ GAM TOVER LR Panel A: The mediation effects of capital structure (MEDIATOR=LEV) ϕ 0.199 0.021*** 0.025*** 0.002* 0.000 (0.139) (0.004) (0.006) (0.001) (0.000) θ 0.979* 0.097*** 0.205*** –0.001 –0.003*** (0.519) (0.015) (0.026) (0.002) (0.001) / –0.002 –0.005 –0.003 –0.002 0.000 (0.004) (0.005) (0.005) (0.005) (0.005) / 0.262* 0.020*** 0.021*** 0.002** 0.000 (0.140) (0.004) (0.006) (0.001) (0.000) Sobel test –0.483 0.312 –0.988 0 –0.598 Panel B: The mediation effects of agency costs (MEDIATOR=MCOST) ϕ 0.199 0.021*** 0.025*** 0.002* 0.000 (0.139) (0.004) (0.006) (0.001) (0.000) θ 0.008 0.031*** 0.040** –0.003*** –0.006*** (0.162) (0.011) (0.019) (0.001) (0.001) / 0.021*** 0.022*** 0.022*** 0.021*** 0.019*** (0.004) (0.004) (0.004) (0.004) (0.004) / 0.262* 0.020*** 0.021*** 0.002** 0.000 (0.140) (0.004) (0.006) (0.001) (0.000) Sobel test 0.049 –2.605 2.508 –3.724 1.966 Panel C: The mediation effects of stock price informativeness (MEDIATOR=SYN) ϕ 0.199 0.021*** 0.025*** 0.002* 0.000 (0.139) (0.004) (0.006) (0.001) (0.000) θ 2.896*** –0.112*** –0.196*** 0.022*** 0.006** (0.871) (0.020) (0.039) (0.003) (0.003) / –0.006*** –0.006*** –0.006*** –0.007*** –0.007*** (0.001) (0.001) (0.001) (0.001) (0.001) / 0.262* 0.020*** 0.021*** 0.002** 0.000 (0.140) (0.004) (0.006) (0.001) (0.000) Sobel test –2.908 –5.063 4.094 –1.923 3.853 This table reports results of the mechanisms through which liquidity mitigates investment inefficiency. All the estimates have been carried out using OLS. Robust standard errors that are clustered by firm are present in parentheses under the coefficients. HL is bid-ask spread of Corwin and Schultz (2012), TOVER is turnover ratio, ILLIQ is Amihud’s(2002) illiquidity measure, LR is Amivest liquidity ratio, and GAM is return reversal developed by Pastor and Stambaugh (2003). ***, **, and * denote significance at the 1%, 5%, and 10% levels, respectively. 5.1. The mediating role of capital structure We begin by positing that the association between stock liquidity and investment effi- ciency relates to the relaxation of financial constraints. Myers and Majluf (1984) point out that when managers have more information than outside shareholders, informed managers are motivated to time the issuance of overpriced shares, and rational investors will discount new issues regardless of their quality. This means that firms may refuse to raise equity capital even if that means passing up positive net present value projects, with the consequence of underinvestment. 246 Xiong and Su Prior studies find that firms with higher liquidity stock maintain a lower cost of equity capital (Amihud, 2002; Amihud & Mendelson, 1986), and that firms can make use of higher liquidity to issue stock at low cost. The relaxation of financial constraints will mitigate underinvestment and improve investment efficiency. Since firms with higher liquidity stock are more likely to use equity financing, lead- ing to relatively lower usage of debt, we can examine the cost of capital mechanism by linking stock liquidity with capital structure. Panel A of Table 6 reports regression estimates of equations (12) to (14) using capi- tal structure as a mediating variable. In the equation for capital structure determination, the coefficient estimates for HL, ILLIQ, and GAM are all significantly positive at the 10% level or above, while the coefficient for LR is negative and significant at the 1% level, indicating a significant negative relation between leverage and liquidity. However, the coefficient estimates for mediating variables (LEV) in equation (14) are all insignifi- cant, thus, we use the Sobel (1986) test to examine the mediation effect. The last row of Panel A reports the results of the Sobel test. We find that except for column (iii), all other tests are insignificant at the 5% level, indicating that the mediation effect of financial leverage is not significant. The reason why financial leverage fails to serve as a mediating variable is that the stock market in China is controlled by the government and that the China Securities Regulatory Commission uses an earnings threshold as a criterion for seasoned equity offerings (He et al., 2013). The earnings threshold prevents financially constrained com- panies from issuing stock in the seasoned equity market, even though they want to take advantage of higher liquidity. 5.2. The mediating role of agency costs This section examines the mediating role of agency costs. Conflicts of interest between managers and outside shareholders and a lack of monitoring may lead managers to pur- sue their own interests by making investments that are not in the best interests of share- holders (Jensen & Meckling, 1976). Jensen (1986) predicts that managers will grow their firm beyond the optimal size, resulting in overinvestment. However, managers pre- ferring to enjoy the quiet life are likely to invest less and forgo some positive net pres- ent value projects, with subsequent underinvestment (Betrand & Mullainathan, 2003). Prior studies argue that stock liquidity can reduce agency costs and improve invest- ment efficiency by permitting more efficient managerial compensation (Holmstrom & Tirole, 1993), facilitating blockholder intervention (Maug, 1998), and enhancing the power of exit threats (Adamati & Pfleiderer, 2009; Edmans, 2009). In this manner, the agency costs may mediate the relationship between stock liquidity and investment efficiency. Panel B of Table 6 reports regression estimates of equations (12) to (14) using agency costs as a mediating variable. In the regression of agency costs on stock liquid- ity, the coefficients on ILLIQ and GAM are significantly positive at the 5% level, while coefficient estimates for TOVER and LR are negative and significant at the 1% level, indicating a negative association between stock liquidity and agency costs. Moreover, the coefficients on MCOST are all significantly positive at the 1% level when control- ling for the effect of stock liquidity, indicating that firms with higher agency costs show lower investment efficiency. According to Baron and Kenny (1986) and Wen et al. (2004), the significance of θ and / indicates that the mediating effect of agency cost is statistically significant. This conclusion is confirmed by the Sobel test in the last row China Journal of Accounting Studies 247 of Panel B of Table 6. The results of the Sobel test suggest that, except for column (i), the mediating effects of agency costs are significant at the 1% level. 5.3. The role of stock price informativeness This section examines whether stock liquidity stimulates the entry of informed investors who increase the informativeness of the stock price and then improve investment effi- ciency. Holmstrom and Tirole (1993) predict that stock liquidity enables an informed trader to disguise private information and profit from it. The increased informativeness of the stock price may contain information that managers do not have. The information in turn can guide a manager in making corporate decisions, such as the decision on cor- porate investment, improving investment efficiency (Chen et al., 2007). Moreover, the increased informativeness of the stock prices reduces information asymmetries between managers and outside shareholders, and increases the ability of investors to monitor management, which in turn improves the capital allocation of the firm (Durnev et al., 2004; Yang & Nie, 2010). Hence, the positive effect of liquidity on investment efficiency should be proportionally mediated by stock price informativeness. Panel C of Table 6 reports regression estimates of equations (12) to (14) using stock price informativeness as a mediating variable. In the regression of stock price informativeness on liquidity, the coefficients on ILLIQ and GAM are significant and negative at the 1% level, while coefficients estimates for TOVER and LR are positive and significant at the 5% level, suggesting liquidity stimulates the entry of informed investors and increases the information content of share prices. Moreover, the coeffi- cients on SYN are all significantly negative at the 1% level when controlling for the effect of stock liquidity, indicating that stock price informativeness helps improve investment efficiency. The significance of θ and / suggests that the relation between stock liquidity and investment efficiency is mediated by stock price informativeness. The results of the Sobel test confirm that the mediating effects of stock price informa- tiveness are significant at the 1% level. 6. Conclusion and policy implications This paper investigates the effect of stock liquidity on firm investment efficiency in a sample of Chinese listed non-financial firms from 1998 to 2011. We find that invest- ment inefficiency, as measured by the deviation from the optimal investment level, is negatively related to stock liquidity. Moreover, if we distinguish between overinvest- ment and underinvestment, we find that stock liquidity mitigates both overinvestment and underinvestment. This finding holds for multiple measures of stock liquidity and investment efficiency and is robust to controlling for other known determinants of investment efficiency and for alternative explanations. Furthermore, we document the channels by which stock liquidity relates to investment efficiency: reducing agency costs and increasing the information content of share prices. Overall, our results are consistent with the idea that stock liquidity helps improve investment efficiency, miti- gating both overinvestment and underinvestment. This paper finds that stock liquidity helps improve investment efficiency. The policy implication from our empirical results is that firms and government should take actions to improve the stock liquidity of firms. We note that, from prior research, Ascioglu, Hegde, Krishnan, and McDermott (2012) and Chung, Elder, and Kim (2010) find that corporate governance and financial reporting quality help improve stock liquidity. 248 Xiong and Su Consequently, we link their findings to ours and propose that firms should reform their corporate governance and strengthen their information disclosure to improve stock liquidity and hence improve investment efficiency. We also suggest that, to improve investor protection and market liquidity, the Chinese government should continue to reform ownership structure and corporate governance, strengthen information disclosure and step up its actions against insider trading. Acknowledgements We thank Haijian Zeng, Liansheng Wu (Joint editor), Jason Xiao (Joint editor), Pauline Weetman (Language editor), Jigao Zhu, Yi Wen, and two anonymous referees for many constructive com- ments that have helped to improve the quality of the paper. We also thank Yongfu Yang for his comments at the China Journal of Accounting Studies conference in Guangzhou. Su gratefully acknowledges financial support from the National Natural Science Foundation of China (Grant No. 71173090), National Social Science Foundation of China (Grant No. 11AJY013), the Guang- dong Pearl River Scholar Fund, the Guangdong Natural Science Foundation (Grant No.S2011010004257) and the Fundamental Research Funds for the Central Universities (12JNYH001). Notes 1. The preceding discussion explores how stakeholders, such as shareholders, creditors, and managers, learn from stock liquidity information when making financial decisions and moni- toring managerial behaviour. Liquidity may also influence firm investment through the avail- ability of funding. Brunnermeier and Pedersen (2009) show that margin spiral and loss spiral may be mutually reinforced, which leads to liquidity spirals. When the margin is positively associated with market illiquidity, a funding shock to the speculators reduces the market liquidity, which leads to higher margins and further tightens the funding constraints of specu- lators. Funding shock and increased funding constraints may drive speculators to sell their initial position, which decreases prices further. The loss spiral and margin spiral reinforce each other, which implies a larger total effect and leads to liquidity spirals or financial crises. Goyenko and Ukhov (2009) find that ‘micro’ or transaction liquidity is closely related to ‘macro’ liquidity or money flow. Illiquidity generally increases as the monetary policy is tightened. Therefore, stock liquidity reflects the availability of funding. The funding con- straint is relaxed when the liquidity is high. This condition helps firms alleviate underinvest- ment by raising their capital and expanding their investments. We thank an anonymous referee for pointing out this idea. 2. These two papers propose a similar regression to test the effect of government control and executive compensation on investment inefficiency. We use their basic regression model and similar control variables to investigate the effect of stock liquidity on investment inefficiency. 3. Since firm investment is influenced by firm-specific unobservable factors, such as corporate culture and managerial characteristics, we use the panel data firm-fixed effect model to esti- mate equation (2). We also fit equation (2) using the OLS method and test the hypothesis. The results are similar to those reported in Table 3 to Table 5. 4. The correlation coefficient between Amivest liquidity ratio (LR) and firm size (SIZE)isas high as 0.6, therefore we exclude firm size (SIZE) when we use the Amivest liquidity ratio (LR) as a liquidity measure. 5. The independent variables of the capital structure determination equation include lagged stock liquidity (LIQ ), firm size (SIZE), Tobin’sQ(Q), tangibility (TANG), non-debt tax it-1 shields (NDTS), return on assets (ROA), firm age (AGE), and industry and year dummies. 6. MacKinnon, Lockwood, Hoffman, West, and Sheets (2002) argue that the tests for the medi- ating effect based on normal theory can yield inaccurate confidence limits and significance tests. They find that alternative tests based on the asymmetric distribution of the product of two normally distributed variables outperform traditional methods. The tables of critical val- ues can be downloaded from http://www.public.asu.edu/~davidpm/ripl/methods.htm. China Journal of Accounting Studies 249 7. The rest of the control variables include firm size (SIZE), leverage (LEV), Tobin’sQ(Q), free cash flow (FCF), the percentage shareholdings of the largest shareholder (TOP1), the sum of percentage shareholdings by second- to fifth-largest shareholders (TOP2_5), government con- trol (STATE), board size (BOARD), board independence (IND), the duality of Chairman and CEO (DUAL), and industry and year dummies. 8. The independent variables of the stock price informativeness determination equation include lagged stock liquidity (LIQ ), firm size (SIZE), leverage (LEV), Tobin’ sQ(Q), return on it–1 assets (ROA), the percentage shareholdings of the largest shareholder (TOP1), the sum of percentage shareholdings by second- to fifth-largest shareholders (TOP2_5), government con- trol (STATE), board size (BOARD), board independence (IND), the duality of Chairman and CEO (DUAL), the marketisation index (MKT), and industry and year dummies. References Admati, A., & Pfleiderer, P. (2009). The ‘Wall Street Walk’ and shareholder activism: Exit as a form of voice. Review of Financial Studies, 22, 2645–2685. Aggarwal, R., & Samwick, A. (2006). Empire builders and shirkers: Investment, firm perfor- mance and managerial incentives. Journal of Corporate Finance, 12, 489–515. Allen, F., Qian, J., & Qian, M.J. (2005). Law, finance, and economic growth in China. Journal of Financial Economics, 77,57–116. Amihud, Y. (2002). Illiquidity and stock returns: Cross-section and time-series effects. Journal of Financial Markets, 5(1), 31–56. Amihud, Y., & Mendelson, H. (1986). Asset pricing and the bid-ask spread. Journal of Financial Economics, 17, 223–249. Ang, J.S., Coles, R., & Lin, J.W. (2000). Agency costs and ownership structure. Journal of Finance, 58(1), 211–221. Ascioglu, A., Hegde, S.P., Krishnan, G.V., & McDermott, J.B. (2012). Earnings management and market liquidity. Review of Quantitative Finance and Accounting, 38(2), 257–274. Baker, M., & Stein, J.C. (2004). Market liquidity as a sentiment indicator. Journal of Financial Markets, 7, 271–299. Baron, R.M., & Kenny, D.A. (1986). The moderator-mediator variable distinction in social psychological research: Conceptual, strategic, and statistical considerations. Journal of Personality and Social Psychology, 51, 1173–1182. Betrand, M., & Mullainathan, S. (2003). Enjoying the quiet life? Corporate governance and managerial preferences. Journal of Political Economy, 111, 1043–1075. Biddle, G.C., Hilary, G., & Verdi, R.S. (2009). How does financial reporting quality relate to investment efficiency. Journal of Accounting and Economics, 48,112–131. Blanchard, O., Lopez-de-Silanez, F., & Shleifer, A. (1994). What do firms do with cash wind- falls? Journal of Financial Economics, 36, 337–360. Brandt, L., & Li, H.B. (2003). Bank discrimination in transition economies: Ideology, informa- tion, or incentives? Journal of Comparative Economies, 31, 387–413. Brunnermeier, M.K., & Pedersen, L.H. (2009). Market liquidity and funding liquidity. Review of Financial Studies, 22(6), 2201–2238. Cai, W.X., & Zeng, C. (2012). Market competition, property-right reform, and the lending activi- ties of commercial banks. Journal of Financial Research, (2), 73–87 (in Chinese). Chen, Q., Goldstein, I., & Jiang, W. (2007). Price informativeness and investment sensitivity to stock price. Review of Financial Studies, 20, 619–650. Chen, S., Sun, Z., Tang, S., & Wu, D. (2011). Government intervention and investment efficiency: Evidence from China. Journal of Corporate Finance, 17, 259–271. Cheng, Z., Xia, X., & Yu, M. (2008). Government intervention, pyramidal shareholding, and overinvestment of SOEs affiliated to local government. Management World, (9), 37–47 (in Chinese). Chung, K.H., Elder, J., & Kim, J. (2010). Corporate governance and liquidity. Journal of Financial and Quantitative Analysis, 45, 265–291. Corwin, S.A., & Schultz, P. (2012). A simple way to estimate bid-ask spreads from daily high and low price. Journal of Finance, 67, 719–760. Durnev, A., Morck, R., & Yeung, B. (2004). Value-enhancing capital budgeting and firm-specific stock return variation. Journal of Finance, 59(1), 65–105. 250 Xiong and Su Edmans, A. (2009). Blockholder trading, market efficiency, and managerial myopia. Journal of Finance, 64, 2481–2513. Fang, V., Noe, T. H., & Tice, S. (2009). Stock market liquidity and firm value. Journal of Financial Economics, 94, 150–169. Goyenko, R.Y., & Ukhov, A.D. (2009). Stock and bond market liquidity: A long-run empirical analysis. Journal of Financial and Quantitative Analysis, 44(1), 189–212. He, J., Mao, X.Y., Rui, M., & Zha, X.L. (2013). Business groups in China. Journal of Corporate Finance, 22, 166–192. Holmstrom, B., & Tirole, J. (1993). Market liquidity and performance monitoring. Journal of Political Economy, 101, 678–709. Jayaraman, S., & Milbourn, T.T. (2012). The role of stock liquidity in executive compensation. The Accounting Review, 87, 537–563. Jensen, M.C. (1986). Agency costs of free cash flow, corporate finance and takeovers. American Economic Review, 76, 323–329. Jensen, M.C., & Meckling, W. (1976). Theory of the firm: Managerial behavior, agency cost and ownership Structure. Journal of Financial Economics, 3, 305–360. Khanna, N., & Sonti, R. (2004). Value creating stock manipulation: Feedback effect of stock prices on firm value. Journal of Financial Markets, 7, 237–270. Kothari, S.P., Leone, A.J., & Wasley, C.E. (2005). Performance matched discretionary accrual measures. Journal of Accounting and Economics, 39(1), 163–197. Kyle, A.S. (1985). Continuous auctions and insider trading. Econometrica, 53, 1315–1335. Kyle, A.S., & Vila, J. (1991). Noise trading and takeovers. Rand Journal of Economics, 22, 54–71. Li, Q. (2009). Accounting information quality and capital allocation efficiency: Empirical evi- dence of Chinese listed corporations. Nankai Business Review, (12), 115–124 (in Chinese). Lipson, M.L., & Mortal, S. (2009). Liquidity and capital structure. Journal of Financial Markets, 12,611–644. Lv, C., & Zhang, H. (2011). The effect of the stock option plans on corporate investment behav- iors. Management World, (11), 118–126 (in Chinese). Mackinnon, D.P., Lockwood, C.M., Hoffman, J.M., West, S.G., & Sheets, V. (2002). A compari- son of methods to test the significance of mediation and other intervening variable effects. Psychological Methods, 7,83–104. Maug, E. (1998). Large shareholders as monitors: Is there a trade-off between liquidity and con- trol? Journal of Finance, 53(1), 65–98. Modigliani, F., & Miller, M.H. (1958). The cost of capital, corporate finance and the theory of investment. American Economic Review, 48, 261–297. Munoz, F. (2013). Liquidity and firm investment: Evidence from Latin America. Journal of Empirical Finance, 20,18–29. Myers, S.C., & Majluf, N.S. (1984). Corporate financing and investment decisions when firms have information that investors do not have. Journal of Financial Economics, 13, 187–221. Pastor, L., & Stambaugh, R.R. (2003). Liquidity risk and expected stock returns. Journal of Political Economy, 111, 642–685. Petersen, M. (2009). Estimating standard errors in finance panel data sets: Comparing approaches. Review of Financial Studies, 22, 435–480. Polk, C., & Sapienza, P. (2009). The stock market and corporate investment: A test of catering theory. Review of Financial Studies, 22(1), 187–216. Richardson, S. (2006). Over-investment of free cash flow. Review of Accounting Studies, 11, 159–189. Ross, S.A., Westerfield, R.W., & Jaffe, J. (2012). Corporate finance. New York: McGraw-Hill/ Irwin. Singh, M., & Davidson, W.N. (2003). Agency costs, ownership structure and corporate gover- nance mechanisms. Journal of Banking and Finance, 27, 793–816. Sobel, M.E. (1986). Some new results on indirect effects and their standard errors in covariance structure models. Sociological Methodology, 16, 159–186. Stein, J. (2003). Agency, information and corporate investment. In G.M. Constantinides, M. Harris, and R.M. Stulz (Eds.). Handbook of the economics of finance. Amsterdam, the Netherlands: North Holland. China Journal of Accounting Studies 251 Stulz, R. (1990). Managerial discretion and optimal financing policies. Journal of Financial Eco- nomics, 26,3–27. Su, D. (2005). Corporate finance and state enterprise reform in China. China Economic Review, 16,118–148. Tang, X., Zhou, X., & Ma, R. (2007). Empirical research on overinvestment behavior and its restriction system in China’s listed companies. Accounting Research, (7), 44–52 (in Chinese). Tobin, J. (1969). A general equilibrium approach to monetary theory. Journal of Money, Credit and Banking, 1(1), 15–29. Wang, C., & Chin, S. (2004). Profitability of return and volume-based investment strategies in China’s stock market. Pacific-Basin Finance Journal, 12, 541–564. Wei, Z., Xie, F., & Zhang, S. (2005). Ownership structure and firm value in China’s privatized firms: 1991-2001. Journal of Financial & Quantitative Analysis, 40,87–108. Wen, Z. L., Chang, L., Hau, K.T., & Liu, H.Y. (2004). Testing and application of the mediating effects. Acta Psychologica Sinica, 36, 614–620 (in Chinese). Xin, Q., Lin, B., & Wang, Y. (2007). Government control, executive compensation and capital investment. Economic Research Journal, (8), 110–122 (in Chinese). Yang, J., & Nie, S. (2010). Stock price informativeness and capital allocation efficiency of corpo- rate. Journal of Management Science, (23), 81–90 (in Chinese). Zhang, G., & Song, X. (2009). Measuring inefficient investment by listed companies in China: Overinvestment or underinvestment. Accounting Research, 5,69–77 (in Chinese). Zhong, H., Ran, M., & Wen, S. (2010). Government intervention, insider Control and corporate investment. Management World, (7), 98–108. (in Chinese).Appendix A. Appendix A. Definitions of variables. Variables Definition Panel A: Main variables of interest Firm investment (INV) Change in gross value of fixed assets, construction in progress, project material and intangible assets minus amortization and depreciation, scaled by lagged total assets. Firm investment 2 (INV2) Cash payments for fixed assets, intangible assets, and other long-term assets from the cash flow statement minus cash receipts from selling these assets and amortization and depreciation, scaled by lagged total assets. Investment inefficiency The absolute value of residuals of expected investment model (INEFFINV) (2) using INV as dependent variable. Overinvestment (OverINV) The positive residuals of expected investment model (2) using INV as dependent variable. Underinvestment (UnderINV) The absolute value of negative residuals of expected investment model (2) using INV as dependent variable. Panel B: Stock liquidity variables Bid-ask spread (HL) See equation (3). Turnover ratio (TOVER) Total shares traded in the day divided by total tradable shares, see equation (4). Illiquidity ratio (ILLIQ) Absolute value of stock return divided by dollar volume, see equation (5). Liquidity ratio (LR) Price change induced by order flow, see equation (6). Return reversal (GAM) The absolute value of coefficient estimate for signed dollar volume in equation (7). (Continued) 252 Xiong and Su Appendix A.(Continued). Variables Definition Panel C: Control variables Firm size (SIZE) The natural logarithm of total assets. Leverage (LEV) Total liabilities scaled by total assets. Tobin’sQ(Q) The sum of market value of tradable shares, book value of non-tradable shares, and liabilities, scaled by book value of total assets. Cash holding (CASH) The ratio of cash holdings to total assets. Stock return (RET) Stock return. Firm age (AGE) The numbers of years since the firm listed. Cash flow (CFO) Cash flow from operations scaled by total assets. Free cash flow (FCF) Cash flow from operations minus expected investment level scaled by total assets. Agency costs (MCOST) The sum of selling expenses and general and administrative expenses scaled by total assets. Tunnelling (ORECTA) Other receivables scaled by total assets. Stock price informativeness Stock price non-synchronicity, see equations (9) and (10). (SYN) Tangibility (TANG) The ratio of tangible fixed assets to total assets. Non-debt tax shields (NDTS) Depreciation scaled by total assets. Return on assets (ROA) Net income divided by total assets. Earning quality (jj DA ) The absolute value of discretional accruals using the performance matched discretionary accrual model (Kothari et al., 2005). Panel D: Ownership and corporate governance variables % of shares by top shareholder The percentage shareholdings of the largest shareholder. (TOP1) nd th % of shares by 2 to 5 The sum of percentage shareholdings by second to fifth largest shareholders (TOP2_5) shareholders. Government control (STATE) STATE is equal to 1 if it is ultimately controlled by the government. Board size (BOARD) The number of directors sitting on the board. Board independent (IND) Ratio of the number of independent directors to board size. Duality (DUAL) A dummy variable, which is equal to 1 if the CEO is also the Chairman. http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png China Journal of Accounting Studies Taylor & Francis

Stock liquidity and capital allocation efficiency: Evidence from Chinese listed companies

China Journal of Accounting Studies , Volume 2 (3): 25 – Jul 3, 2014

Stock liquidity and capital allocation efficiency: Evidence from Chinese listed companies

Abstract

Based on market microstructure theories and evidence, this paper investigates the relationship between stock liquidity and capital allocation efficiency using Chinese listed companies from 1998 to 2011. This paper finds that stock liquidity helps improve investment efficiency, mitigating both overinvestment and underinvestment. This finding is robust to numerous sensitivity analyses, including controls for endogeneity and for the other known determinants of investment efficiency, the choice...
Loading next page...
 
/lp/taylor-francis/stock-liquidity-and-capital-allocation-efficiency-evidence-from-Ns7tGuxoAb
Publisher
Taylor & Francis
Copyright
© 2014 Accounting Society of China
ISSN
2169-7221
eISSN
2169-7213
DOI
10.1080/21697213.2014.959413
Publisher site
See Article on Publisher Site

Abstract

China Journal of Accounting Studies, 2014 Vol. 2, No. 3, 228–252, http://dx.doi.org/10.1080/21697213.2014.959413 Stock liquidity and capital allocation efficiency: Evidence from Chinese listed companies a b,c Jiacai Xiong * and Dongwei Su School of Accountancy, Jiangxi University of Finance and Economics, Nanchang 330013, b c China; Department of Finance, Jinan University, Guangzhou 510632, China; Research Institute of Finance, Jinan University, Guangzhou 510632, China Based on market microstructure theories and evidence, this paper investigates the relationship between stock liquidity and capital allocation efficiency using Chinese listed companies from 1998 to 2011. This paper finds that stock liquidity helps improve investment efficiency, mitigating both overinvestment and underinvestment. This finding is robust to numerous sensitivity analyses, including controls for endo- geneity and for the other known determinants of investment efficiency, the choice of the measure of stock liquidity and investment efficiency. Further analysis shows that stock liquidity improves corporate capital allocation efficiency by reducing agency costs and increasing the information content of share prices. Keywords: capital allocation efficiency; market microstructure; overinvestment; stock liquidity; underinvestment 1. Introduction As one of the three basic financial decisions that are made by firms, investment has become a long-debated issue in corporate finance. Investment links financing activities with capital allocation as well as determines the future cash flow and profitability of firms, thereby serving as the basis for enterprise growth. At a macroeconomic level, investment is among the three engines that pull an economy (namely, consumption, investment and export-import). However, a large body of literature finds that Chinese listed companies have low investment efficiency (e.g., Tang, Zhou, & Ma, 2007; Zhang & Song, 2009). Low-level development and redundant development are widespread among the steel, iron, and building materials industries, creating significant excess capacity, whereas the agricultural, environment, and high-technology sectors receive an insufficient amount of investment (Lv & Zhang, 2011). Because investment is among the major forces that have driven the economic growth of China over the past decades, mitigating underinvestment and overinvestment, as well as improving investment efficiency, must be investigated in order to boost the country’s economy. In the perfect world of Modigliani and Miller (1958), an investment policy is solely dependent on the investment opportunities of a firm as measured by Tobin’s Q (Tobin, 1969). However, several types of friction and distorting forces, such as information asymmetries and agency costs, may prevent firms from making optimal investments (Stein, 2003). Myers and Majluf (1984) argue that information asymmetries between *Corresponding author. Email: xiongjc-p@163.com Paper accepted by Tong Yu. © 2014 Accounting Society of China China Journal of Accounting Studies 229 managers and shareholders may lead to adverse selection and subsequently to underin- vestment. With regard to agency costs, the separation of ownership and control creates a conflict of interest between shareholders and managers (Jensen & Meckling, 1976). Managers are also trying to make investments to pursue their own private interests, leading to managerial empire-building and overinvestment (Jensen, 1986; Stulz, 1990). Previous studies have found that superior corporate governance is useful in improving the investment efficiency of Chinese listed firms (Tang et al., 2007; Xin, Lin, & Wang, 2007; Li, 2009; Lv & Zhang, 2011), whereas government intervention distorts the investment behaviour of these firms and subsequently leads to investment inefficiency (Chen, Sun, Tang, & Wu, 2011; Cheng, Xia, & Yu, 2008; Zhong, Ran, & Wen, 2010). However, we have not found any published material relating to the effect of stock liquidity on investment efficiency of Chinese listed firms. The positive effect of market liquidity on firm investment is strongly backed by evi- dence. Theoretical models predict that market liquidity facilitates the formation of a toehold stake (Kyle & Vila, 1991), permits non-blockholders to intervene and become blockholders (Maug, 1998), promotes highly efficient management compensation (Holmstrom & Tirole, 1993), strengthens the exit threats of blockholders (Admati & Pfleiderer, 2009; Edmans, 2009), and stimulates trade by informed investors, thereby improving investment decisions through highly informative stock prices (Khanna & Sonti, 2004). Empirical evidence shows that firms with higher liquidity maintain a lower cost of equity capital (Amihud & Mendelson, 1986; Amihud, 2002), a higher investment level (Munoz, 2013), and a higher firm value (Fang, Noe, & Tice, 2009). Therefore, liquidity may be positively related to investment efficiency. Although many studies relate liquidity to corporate governance and financial decisions, we have not found any published study relating to the effect of stock liquidity on the investment efficiency of Chinese listed firms. This study aims to provide empirical evidence and shed light on the relationship between liquidity and investment efficiency by using a sample of Chinese firms that were publicly listed between 1998 and 2011. We find that stock liquidity helps improve investment efficiency, mitigating both overinvestment and underinvestment. This result holds for multiple measures of liquid- ity and investment efficiency, and is robust to controlling for the other known determi- nants of investment efficiency. We also investigate the causes behind the beneficial effects of liquidity, such as curbing the opportunistic behaviour of managers and increasing the information content of share prices. This paper enriches the literature on stock liquidity. Early studies, such as those by Amihud (2002), Amihud and Mendelson (1986) and Pastor and Stambaugh (2003), exam- ine the effect of stock liquidity on asset pricing. However, there has been a growing inter- est in studying the relationship between stock liquidity and the decision of firms. These studies include evidence that firms with greater stock liquidity are more likely to use equity financing, resulting in lower financial leverage (Lipson & Mortal, 2009), that stock liquid- ity interacts with firm investment (Munoz, 2013), and that firms with liquid stocks have better performance (Fang et al., 2009). We contribute to this literature by relating stock liquidity to investment efficiency. We use the expected investment model of Richardson (2006) to compute overinvestment and underinvestment as well as to investigate the rela- tionship between liquidity and investment efficiency. We find that stock liquidity helps improve investment efficiency, mitigating both overinvestment and underinvestment. Additionally, our study sheds light on the mechanisms through which stock liquidity can affect firm decisions and outcomes. Fang et al. (2009) find liquidity enhances firm per- formance by increasing the information content of market prices and performance- 230 Xiong and Su sensitive managerial compensation. We explicitly investigate the causes behind the benefi- cial effects of liquidity, and find that stock liquidity improves corporate capital allocation efficiency by reducing agency costs and increasing the information content of share prices. The rest of the paper is organised as follows. Section 2 discusses the institutional background, reviews the existing literature on investment efficiency and on the role of stock liquidity in investment decisions, and develops testable hypotheses. Section 3 describes the research design, models, variable measures, and the sample. Sections 4 and 5 present empirical evidence on whether and why stock liquidity affects investment inefficiency. Section 6 concludes. 2. Institutional background, literature, and hypothesis development 2.1. Institutional background China offers a unique environment for analysing the investment efficiency of firms. First, China has maintained a state-dominated financial system in which government at various levels controls the allocation of financial resources in both the banking and security markets (He, Mao, Rui, & Zha, 2013). The banking system in China comprises the central bank, three policy banks, four large state-owned commercial banks, 10 national joint-stock commercial banks, approximately 90 regional commercial banks, as well as urban and rural credit cooperatives. The four state-owned commercial banks dominate the entire Chinese market (He et al., 2013). According to Cai and Zeng (2012), these four state-owned commercial banks had a market share of approximately 70% from 2002 to 2007. The Chinese stock market is also controlled by the govern- ment. Before 1999, the total annual number of initial public offerings (IPOs) in China was subject to a quota system. The central government set a quota for the quantity of equities or the number of firms to be listed annually before this period. Before 1999, a company that intended to list was required to be selected by a provincial government or ministry with a quota before asking the China Securities Regulatory Commission (CSRC) for approval. In July 1999, the IPO quota system was replaced by an authori- sation system. Investment banks are allowed to nominate firms for public listing and their nominations are screened by an independent listing committee of the CSRC. The independent listing committee assesses the qualifications of a to-be listed company. In addition, government-guided financial resource allocation usually favours a few large-scale state-owned enterprises (SOEs) and these SOEs may encounter soft budget constraints, while smaller SOEs and most non-SOEs cannot easily obtain financing from the state-controlled financial system (Allen, Qian, & Qian, 2005; Brandt & Li, 2003). This phenomenon of bank discrimination remains in place currently. Therefore, the latter group of enterprises suffers from serious financial constraints and subse- quently faces underinvestment. Second, most Chinese listed firms are business units that have been carved out from large SOEs and are controlled by government-related entities. As the controlling share- holder, the Chinese government often plays the conflicting dual roles of an SOE owner and administrator of social affairs. As an SOE owner, the government is supposed to benefit from value maximisation. However, to accomplish their social and political goals, such as regional economic development, higher employment rates, and social sta- bility, government leaders are driven to intervene in SOEs by changing the objective functions of SOEs based on their political preferences (Chen et al., 2011). Therefore, these SOEs tend to miss profitable investment opportunities by implementing the plans China Journal of Accounting Studies 231 of the government as well as fails to terminate unsuccessful projects because of their potential conflicts with government policies, which results in investment inefficiency. Third, the typical external governance mechanisms, such as debt, takeover threats, legal protection of investors, and product market competition, have been ineffective because of the political nature of the privatisation process (Su, 2005). Bank loans are traditionally viewed as grants from the state that are designed to bail out failing firms. State-owned banks retain a monopoly in the banking sector, but profit is not their over- riding objective. When a political favour is deemed appropriate, issuing subsidised loans and rescheduling of overdue debt can be arranged with SOEs (these are seen as soft budget constraints) (Su, 2005). In addition, a market for private, non-bank debt is yet to be established. The stock market lacks an active merger or takeover activity to discipline firm management. The capital market also has insufficient information to keep managerial decisions at arm’s length. Furthermore, equity-based executive incen- tive contracts, such as stock options and performance-based stock grants, are rarely used in Chinese listed firms. Wei, Xie, and Zhang (2005) report that senior managers and directors have an average stock holding of only 0.015% for partially privatised SOEs. Inefficient corporate governance may cause the managers to pursue their personal interests, such as building an empire and enjoying a quiet life, which subse- quently leads to investment inefficiency (Jensen, 1986; Jensen & Meckling, 1976). 2.2. Determinants of investment efficiency In a perfect world with a frictionless capital market as envisioned by Modigliani and Miller (1958), capital is allocated in such a way that the marginal product of capital is similar across every project in the economy. Managers obtain financing at the prevailing interest rate and undertake all projects with a positive net present value. Empirically, the investment of a firm should be solely determined by the profitability of its investment as measured by Tobin’s Q (Tobin, 1969). However, several distorting forces in the real world may drive firms to deviate from their optimal investment level, leading to underinvestment and overinvestment. Among these, the most pervasive and important factors affecting the efficiency of capital investment are those that arise from informational asymmetries and agency conflicts (Stein, 2003). With regard to agency costs, the separation of ownership and control creates a con- flict of interest between shareholders and managers, and insufficient monitoring may lead the managers to pursue private objectives that may be in conflict with those of out- side shareholders (Jensen & Meckling, 1976). Jensen (1986) and Stulz (1990) argue that empire-building preferences may cause managers to spend all of their available funds on investment projects, which leads to overinvestment. Blanchard, Lopez-de-Silanez, and Shleifer (1994) examine how a small sample of firms responds to a large cash windfall coming from legal settlements, a source of cash that does not change its investment opportunity set. They find that managers typically spend the cash on acquisitions rather than turning over the windfalls to their shareholders. Besides, the managers who prefer to enjoy a quiet life tend to invest less because investing requires these managers to oversee the investment. When firms expand their existing facilities or start new product lines, the managers are required to fulfil more duties or spread their talents over a wider array of activities. Therefore, managers generally prefer to invest less or to underinvest in the sense that they decide to forgo some projects with positive net present value (Betrand & Mullainathan, 2003). 232 Xiong and Su Models of costly external financing generally predict underinvestment because of the informational asymmetry of adverse selection. Acting in the interest of current shareholders, managers try to sell new shares when their private information suggests that these shares are overpriced; however, rational investors will discount these new issues, thereby discouraging firms with investment opportunities from raising equity capital even at the expense of giving up projects with positive net present value (Myers & Majluf, 1984). These findings suggest that costly external financing brought about by information asymmetries and opportunistic behaviour of managers may lead to overinvestment and underinvestment. As regards the investment efficiency of Chinese listed companies, Zhang and Song (2009) find that the optimal investment ratio is 24.4% of fixed assets at the start of an accounting year, and that 39.26% of the sample companies overinvest, while 60.74% of the sample companies underinvest. Tang et al. (2007) observe widespread overin- vestment among Chinese listed firms from 2000 to 2002. They also identify cash divi- dends, liabilities, and corporate governance as effective restrictions on overinvestment. Xin et al. (2007) investigate the governance effect of executive compensation on capital investment decision by using a sample of Chinese firms that were listed from 2002 to 2004. They find that government control, particularly that over SOEs affiliated with local government and state asset-management bureaux, induces overinvestment when the diligence and talent of executives are not being compensated. Li (2009) suggests that accounting information quality can improve capital allocation efficiency by mitigat- ing adverse selection due to information asymmetry between managers and investors as well as by curbing the incentives of managers to engage in value-destroying activities, such as empire building. By using a sample of Chinese firms listed between 2004 and 2006, Li (2009) finds that those firms with high-quality financial reporting have a reduced tendency to deviate from the predicted investment level, thereby mitigating both overinvestment and underinvestment. Cheng et al. (2008) and Zhong et al. (2010) hypothesise that the Chinese government intervenes in SOEs to accomplish its social and political goals, such as boosting regional economic growth, increasing regional rev- enues and employment rates, and maintaining social stability, which influences the investment behaviour of firms and leads to investment inefficiency. Using Chinese data, Chen et al. (2011) investigate whether government intervention distorts the investment behaviour of firms and leads to investment inefficiency. They find that the sensitivity of investment expenditure to investment opportunities is significantly weaker for SOEs, and that political connections significantly reduce the investment efficiency in these enterprises. Lv and Zhang (2011) find that the managerial stock incentive in Chinese publicly traded firms helps mitigate underinvestment and overinvestment. 2.3. Development of hypothesis As the central element of market microstructure, stock liquidity is generally described as the ability to trade large quantities quickly at low cost and with limited effects on prices. Earlier studies primarily focus on stock liquidity and asset pricing (e.g., Amihud, 2002; Amihud & Mendelson, 1986; Pastor & Stambaugh, 2003), whereas recent studies investigate the effect of stock liquidity on corporate governance and find that stock liquidity helps optimise CEO compensation and enhances the ability of investors to monitor managerial decisions (Adamati & Pfleiderer, 2009; Edmans, 2009; Jayaraman & Milbourn, 2012). These findings introduce a new question of whether stock liquidity affects investment inefficiency. China Journal of Accounting Studies 233 First, stock liquidity has been argued to reduce the cost of equity capital and relax financial constraints through asset pricing and positive feedback mechanisms. Amihud and Mendelson (1986) suggest that stock transaction costs must be considered when valuing equity investment, and that liquidity decreases the required rates of return for equity investment. Amihud (2002), Amihud and Mendelson (1986), and Pastor and Stambaugh (2003) employ different liquidity measures to test the relationship between liquidity and stock return, which is eventually found to be negative. Khanna and Sonti (2004) find that informed traders factor the effect of their trades on managerial behaviour into their trading strategies, which drives them to trade more aggressively. This practice increases the information content of prices, which improves the operating performance and relaxes the financial constraints of firms. The expected return is equal to the discount rate of investment opportunities (Ross, Westerfield, & Jaffe, 2012); thus, firms with liquidity stock maintain a low discount rate, which in turn improves the profitability of investment projects. At the same time, the feedback effect from price improves the operating performance and relaxes the financial constraints of these firms, which subsequently drives these firms to expand their investments and mitigate their underinvestment. Second, stock liquidity improves investment efficiency by helping optimise CEO compensation and enhancing the ability of investors to monitor managerial decisions. Holmstrom and Tirole (1993) and Kyle (1985) model informed traders who optimally select their trading intensity as a function of stock liquidity,thereby determining the amount of private information that is impounded in the stock price. At equilibrium, higher stock liquidity drives increased informed trading to dominate over increased unin- formed trading, which increases the amount of private information that is incorporated into the stock price. With regard to the increased informativeness of the stock price, Holmstrom and Tirole (1993) argue that firms can offer steeper stock-based incentives to senior managers. Jayaraman and Milbourn (2012) find that the pay-for-performance sensitivity of CEOs to stock prices increases along with the liquidity of stock. An optimal managerial contract induces senior managers to explore investment opportu- nities with positive net present value and forgo projects with negative net present value, which subsequently improves firm performance and remuneration. Aggarwal and Samwick (2006) develop a model and empirically analyse the relationships among the incentives from compensation, investment, and firm performance. They find that the pay-for-performance sensitivity of CEOs to stock prices can mitigate managerial shirking and increase capital expenditure. These results, which are consistent with the models of underinvestment, sup- port the argument that managers have private costs of investment (Betrand & Mullainathan, 2003). Lv and Zhang (2011) examine how the managerial stock incentive mechanism of China affects corporate investment and find that such a mechanism helps mitigate inefficient investment, which subsequently mitigates both underinvestment and overinvestment. Moreover, stock liquidity can improve the ability of blockholders to monitor and enhance the power of exit threats. These blockholders monitor and trade with the aim of profiting from the price appreciation caused by their monitoring activities. Maug (1998) concludes that higher stock liquidity facilitates highly effective monitoring by allowing blockholders to purchase additional shares at a price that does not reflect the benefits of intervention. Adamati and Pfleiderer (2009) and Edmans (2009) argue that when the compensation of managers is tied to current stock prices, increased stock liquidity also increases the cost of opportunism to managers by facilitating informed selling. Managers will undertake productive efforts and investments to improve the firm value and dissuade blockholders from exiting. 234 Xiong and Su Third, improved stock liquidity helps informed parties (speculators) to disguise their private information and to profit from such information (Kyle & Vila, 1991). An increase in the marginal value of information drives speculators to spend more time on monitoring activities. The increased flow of information into the market improves the information content of the stock prices. These stock prices with large amounts of private information can provide managers with a greater quantity of new information about the prospects of their own firms, such as product demand and strategic issues, which in turn affect the investment decisions of managers (Chen, Goldstein, & Jiang, 2007). Moreover, the enhanced informativeness of the stock prices can reduce informa- tion asymmetries between managers and external shareholders as well as enhance the ability of investors to monitor the managers, which in turn improves the capital allocation of the firm (Durnev, Morck, & Yeung, 2004; Yang & Nie, 2010). Yang and Nie (2010) investigate the relationship between stock price informativeness and capital allocation efficiency by using a sample of Chinese companies listed in 2001 and 2008. They find that stock price informativeness is negatively associated with the underinvest- ment and overinvestment of firms that arise from abusing free cash flow. These findings lead to the following hypothesis: H: Firms with higher liquidity show lower investment inefficiency. 3. Research Design 3.1. Model specification By analogy with the work of Xin et al. (2007) and Zhong et al. (2010), we estimate the following multivariate panel data regressions to test the effect of stock liquidity on investment efficiency: INEFFINV ¼ / þ / LIQ þ / FCF þ / MCOST þ / ORECTA þ / SIZE þ / LEV it 0 1 it1 2 it 3 it 4 it 5 it 6 it X X þ / Q þ / INDUSTRY DUMMIES þ / YEAR DUMMIES þ 1 it it 7 k t k t (1) where the dependent variable INEFFINV represents investment inefficiency for firm i it in year t. LIQ represents different proxies of stock liquidity for firm i in year t–1. To it-1 alleviate any concern for potential endogeneity, we use lagged liquidity as an indepen- dent variable. The rest are control variables that may influence investment inefficiency: free cash flow (FCF), agency costs (MCOST), other receivables (ORECTA), size (SIZE), leverage (LEV), Tobin’sQ(Q), industry and year dummies. See Appendix A for definitions of variables. We estimate equation (1) using ordinary least squares (OLS). Following Peterson (2009), we adjust the standard errors for heteroscedasticity and within-firm serial corre- lation using cluster at the firm level. Since our hypothesis predicts that stock liquidity mitigates investment inefficiency, we expect the coefficient estimates for / is signifi- cantly negative. 3.2. Variable measures 3.2.1. Firm investment Richardson (2006) splits total investment I into two main components: TOTAL,t (i) required investment expenditure to maintain assets in place, I , and MAINTAIN China Journal of Accounting Studies 235 (ii) investment expenditure on new projects, I . Following Lv and Zhang (2011), we NEW define total investment, I , as change in gross value of fixed assets, construction TOTAL,t in progress, project material and intangible assets, and I as amortisation and MAINTAIN depreciation. Firm investment (INV) is the difference between total investment I TOTAL and maintenance investment I , scaled by lagged total assets. MAINTAIN Chinese listed companies report cash payments for fixed assets, intangible assets, and other long-term assets, and cash receipts from selling these assets in their cash flow statement; therefore, we use the difference between them as our measure for total investment. Based on this, we construct a new firm investment variable (INV2) as the difference between new total investment and maintenance investment I MAINTAIN (namely, amortisation and depreciation). 3.2.2. Investment inefficiency Conceptually, investment efficiency refers to firms undertaking all, and only, projects with positive net present value. Richardson (2006) and Biddle, Hilary, and Verdi (2009)define investment inefficiency as deviations from expected investment using a model that predicts investment as a function of firm characteristics. Thus, deviations from expected investment represent investment inefficiency. Applying the equation used in Richardson (2006) and Xin et al. (2007), we estimate expected investment according to the following regression: INV ¼ a þ a SIZE þ a LEV þ a CASH þ a AGE þ a RET þ a Q it 0 1 it1 2 it1 3 it1 4 it1 5 it1 6 it1 X X þa INV þ a INDUSTRY DUMMIES þ a YEAR DUMMIES þ l 7 it1 k t it (2) where INV is investment expenditure for firm i in year t. The explanatory variables it include firm size (SIZE), leverage (LEV), the level of cash holdings (CASH), firm age (AGE), past sock return (RET ), Tobin’sQ(Q) and prior level of investment (INV ). it–1 it–1 Appendix A provides definitions of variables. We estimate equation (2) using the panel data firm-fixed effect model. The residu- als from the regression models represent the deviations from expected investment level, and we use the absolute value of these residuals as a firm-specific proxy for investment inefficiency (INEFFINV ). Thus, a higher value means higher inefficiency. A positive it residual means that companies invest above the optimal level. We define the residual as overinvestment (OverINV). A negative residual is regarded as underinvestment. To ease the exposition, we multiply the residual by –1 and define it as underinvestment (UnderINV), so that a higher value indicates more severe underinvestment. 3.2.3. Stock liquidity Liquidity is generally described as the ability to trade large quantities quickly at low cost with little price impact. This description highlights four dimensions to liquidity, namely, trading quantity, trading speed, trading cost, and price impact. Considering data availability and computing cost, we employ five commonly used measures of liquidity. Bid-ask spread (HL). Corwin and Schultz (2012) propose a bid-ask spread estimator based on two ideas. First, daily high (low) prices are almost always buyer (seller) -initi- ated trades. Therefore, the ratio of high-to-low prices for a day reflects both the 236 Xiong and Su fundamental volatility of the stock and its bid-ask spread. Second, the component of the high-to-low price ratio that is attributed to volatility increases proportionately with the length of the trading interval, while the component due to bid-ask spreads stays constant over a short period. In addition, Corwin and Schultz (2012) find that this high-low measure outperforms other commonly used low-frequency measures. There- fore, we use this high-low spread measure (HL) as a stock liquidity measure and com- pute it using two consecutive days as follows: it 2eðÞ 1 HL ¼ (3) it it 1 þ e where 8 9 "# ! pffiffiffiffiffiffiffiffi pffiffiffiffiffi rffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi < 1 o = 2b  b c it;dþj it it it pffiffiffi pffiffiffi a ¼  , b = E ln , it it : ; 3  2 2 3  2 2 it;dþj j¼0 "# ! it;d;dþ1 o o c ¼ ln : H L it it;d it;d it;d;dþ1 denotes the actual high (low) stock price for stock i on day d of year t, o o H L is the high price over days t and t+1. A higher value of HL means it;d;dþ1 it;d;dþ1 lower stock liquidity. Turnover ratio (TOVER). The second measure is the tradable turnover ratio (TOVER), which is defined as the average of daily turnover ratio for a firm during the year. Daily turnover ratio is total shares traded in the day divided by total tradable shares. Considering that a large portion of shares (state shares and legal entity) in a typical Chinese listed company is not allowed to be traded and the total turnover would be biased (Wang & Chin, 2004), we use the tradable turnover ratio. it 1 VOL itd TOVER ¼ (4) it D LNS it itd d¼1 where VOL is the number of shares traded on day d of year t for stock i, LNS is itd itd total tradable A shares, D is equal to the number of days for stock i in year t. it Amihud illiquidity ratio (ILLIQ). The third measure is Amihud’s(2002) illiquidity ratio, which captures Kyle’s(1985) conception of illiquidity, i.e., the response of price to order flow. Specifically, the illiquidity ratio for stock i in year t is equal to it 1 jj r itd ILLIQ ¼  100 (5) it D V it itd d¼1 where r and V are stock i’s return and dollar volume (in millions) on day d in year itd itd t, respectively. D is equal to the number of days for stock i in year t. |r |/V it itd itd measures the price change induced by order flow. Since the Amihud ratio measures illiquidity, the ratio is smaller for more liquid stocks. Amivest liquidity ratio (LR). The Amivest liquidity ratio measures the trading volume associated with a unit change in the stock price. A higher value of LR implies greater market liquidity or depth. The liquidity ratio is defined as follows: it 1 V itd LR ¼  10 (6) it D jj r it itd d¼1 China Journal of Accounting Studies 237 Return reversal measure (GAM). Pastor and Stambaugh (2003) point out that order flow should be accompanied by a return that one expects to be partially reversed in the future if the stock is not perfectly liquid. They argue that therefore the greater the expected reversal for a giver dollar volume, the lower the stock’s liquidity. Thus, we estimate the return reversal measure (GAM) by running the following regression: e e r ¼ h þ / r þ c  Signðr Þ V þ e (7) i;t i;t;d i;t;d i;t;dþ1 i;t;dþ1 i;t i;t i;t;d where r is stock’s excess return above market return on day d. Return reversal mea- i;t;d sure GAM is defined as the absolute value of γ , GAM ¼ c . A higher GAM implies i,t i;t greater price impact and lower liquidity. 3.2.4. Free cash flow Jensen (1986) argues that free cash flow is cash flow beyond what is necessary to finance all projects with positive net present value. Therefore, we define free cash flow as cash flow from operations minus expected investment level: FCF ¼ CFO  INV (8) it it it where CFO represents the cash flow from operations, INV is the expected level of it investment, which is equal to the fitted values of equation (2). 3.2.5. Agency costs Ang, Coles, and Lin (2000) and Singh and Davidson (2003) use a firm’s selling, general, and administrative (SG&A) expenses to capture equity agency costs. They point out that SG&A expenses represent the costs related to the management function and sale of products, includes managerial salaries, rents, insurance, utilities, supplies, and advertising costs. Higher levels of SG&A expenses are close approximations of managerial pay and perquisite consumption. These costs, to a large extent, reflect managerial discretionary expenses and may be a closer proxy for agency costs (Singh & Davidson, 2003, p. 794). Chinese listed companies report selling expenses, and general and administrative expenses in their income statement, therefore, we use the sum of selling expenses and general and administrative expenses scaled by total assets (MCOST) as our measure for agency costs. 3.2.6. Stock price informativeness Following Chen et al. (2007), we use stock price non-synchronicity as stock price informativeness. The construction of stock price informativeness (SYN) consists of two steps. First, we estimate equation (9) to decompose the variation of a stock return into two components: systematic risk and idiosyncratic volatility. r ¼ a þ b r þ c r þ e (9) it;w it mt;w jt;w it;w it it where r is stock return for firm i on week w of year t, r and r represent market it,w mt,w jt,w and industry return, respectively. R reflects the proportion of volatility that is it explained by systematic components. A high value of R indicates that firms’ stock it returns are closely tied to market and industry returns, and are assumed to reflect rela- tively less firm-specific information. Thus, 1–R measures stock price informativeness. it In the second step, we use a logistic transformation of 1–R to construct SYN: it 238 Xiong and Su 2 2 SYN ¼ Ln½ð1  R Þ=R  (10) it it it A higher value of SYN implies greater information content of share prices. 3.2.7. Other control variables Following Xin et al. (2007) and Yang and Nie (2010), we introduce several control variables in our model. As a proxy for tunnelling by a controlling shareholder, we use other receivables scaled by total assets (ORECTA); to control for the size, we use natu- ral logarithm of total assets (SIZE); to measure growth options we include Tobin’sQ (Q) as the sum of market value of tradable shares, book value of non-tradable shares, and liabilities, scaled by book value of total assets. We also control for the industry and year dummies. 3.3. Sample and descriptive statistics Our sample is drawn from firms listed on the Shenzhen and Shanghai stock exchanges over the period from 1998 to 2011. We exclude financial firms because their liabilities are not strictly comparable to those in other industries. The stock price, return, trading volume and accounting information data are extracted from the China Stock Market and Accounting Research (CSMAR) database commercially available from Shenzhen GTA Information Co. Ltd. Ownership and corporate governance data are gathered from China Center for Economics Research (CCER) China stock database. We winsorize each continuous variable at the first and 99th percentiles to mitigate the influence of outliers. Table 1 reports descriptive statistics for the variables, including the mean, standard deviation, median, minimum, and maximum. Firm investment (INV) in the sample has a mean of 0.021 and a median of –0.004, while standard deviation of investment is 0.136, indicating that there is significant variation among firm investment. The mean and standard deviation of the Amihud illiquidity ratio (ILLIQ) are 0.305 and 0.395, respectively, suggesting a great variation between firm stock liquidity. 4. Empirical results 4.1. The effect of stock liquidity on investment inefficiency Table 2 reports panel regression estimates of equation (1) using different stock liquidity measures. In the first column, we use the bid-ask spread (HL) developed by Corwin and Schultz (2012); in the second, the illiquidity ratio (ILLIQ) proposed by Amihud (2002); in the third, the return reversal measure (GAM)defined by Pastor and Stamb- augh (2003); in the fourth, the tradable turnover ratio (TOVER); in the fifth column, the Amivest liquidity ratio (LR). The dependent variable of the regressions is investment inefficiency (INEFFINV) based on INV. As shown in Table 2, when we use illiquidity ratio and return reversal to proxy for stock liquidity, the coefficient estimates for ILLIQ and GAM are all significantly posi- tive at the 1% level, indicating that firms with higher stock liquidity show higher investment efficiency, which supports our hypothesis. The economic magnitude of the liquidity effect is also meaningful. A 10 percentage-point decrease in ILLIQ and GAM will reduce investment inefficiency by 21% and 25%, respectively. China Journal of Accounting Studies 239 Table 1. Descriptive statistics. Variable Observation Mean Standard deviation Median Minimum Maximum INV 16,475 0.021 0.136 –0.004 –0.307 0.710 HL 19,196 0.046 0.013 0.044 0.023 0.083 TOVER 19,354 2.738 2.546 2.013 0.293 16.073 ILLIQ 19,355 0.305 0.395 0.158 0.005 2.286 LR 19,366 2.291 3.526 1.094 0.103 26.096 GAM 19,196 0.102 0.158 0.042 0.000 0.951 Q 19,465 1.683 1.026 1.360 0.812 7.459 CFO 19,685 0.044 0.083 0.043 –0.216 0.279 SIZE 19,687 21.277 1.203 21.137 10.842 28.282 LEV 19,687 0.496 0.290 0.477 0.055 2.253 CASH 19,687 0.178 0.148 0.136 0.003 0.723 AGE 19,690 7.599 4.743 7.000 1.000 19.000 RET 17,723 0.271 0.878 –0.055 –0.746 3.717 MCOST 19,605 0.130 0.244 0.075 0.008 1.956 ORECTA 19,686 0.054 0.091 0.019 0.000 0.554 This table reports the descriptive statistics of the main variables used in our multivariate analysis for our full sample of firms. INV is firm investment. HL is bid-ask spread of Corwin and Schultz (2012), TOVER is turn- over ratio, ILLIQ is Amihud’s(2002) illiquidity measure, LR is Amivest liquidity ratio, and GAM is return reversal developed by Pastor and Stambaugh (2003). Q is Tobin’sq, CFO is cash flow from operations scaled by total assets, SIZE is the natural logarithm of total assets, LEV is total liabilities scaled by total assets, CASH is the ratio of cash holdings to total assets, AGE is the number of years after going public, RET is stock return, MCOST is the sum of selling expenses and general and administrative expenses scaled by total assets, and ORECTA is other receivables scaled by total assets. When we use the tradable turnover ratio (TOVER) as stock liquidity, the coefficient for TOVER is 0.02 and is significant at the 10% level, suggesting a positive association between liquidity and investment inefficiency, which contradicts our hypothesis. The possible reason is that the turnover ratio may reflect investor sentiment. Baker and Stein (2004) argue that high liquidity is a symptom of the fact that the market is domi- nated by irrational investors, and hence is overvalued. Managers who try to boost short-run share price have an incentive to waste resources in projects with negative NPV to cater for current sentiment when stock price is overpriced, with subsequent overinvestment (Polk & Sapienza, 2009). 4.2. The effect of stock liquidity on underinvestment and overinvestment The results in Table 2 raise a further question of whether higher stock liquidity is asso- ciated with a reduction of overinvestment or with a reduction of underinvestment. Based on this, we distinguish underinvestment and overinvestment scenarios. Tables 3 and 4 present results of the estimates of equation (1) using underinvestment (UnderINV) and overinvestment (OverINV) respectively as dependent variables. As shown in Table 3, the coefficient estimates for TOVER and LR are all signifi- cantly negative at the 10% level or above, indicating that firms with higher stock liquidity show lower underinvestment. The results appear economically significant as well. A 10 percentage-point increase in TOVER and LR leads to a decrease in underin- vestment of 1% and 2%, respectively. As shown in Table 4, the coefficient estimates for HL, ILLIQ,and GAM are all significantly positive at 5% or above, suggesting that stock liquidity is negatively associated with overinvestment. In particular, a 10 percentage-point decrease in ILLIQ 240 Xiong and Su Table 2. Regressions of investment inefficiency on stock liquidity. (i) (ii) (iii) (iv) (v) HL ILLIQ GAM TOVER LR LIQ 0.199 0.021*** 0.025*** 0.002* 0.000 it-1 (0.139) (0.004) (0.006) (0.001) (0.000) FCF –0.000*** –0.000*** –0.000*** –0.000*** –0.000*** (0.000) (0.000) (0.000) (0.000) (0.000) ORECTA –0.001 –0.004 –0.003 0.002 –0.001 (0.009) (0.011) (0.011) (0.011) (0.011) SIZE 0.001 0.004*** 0.002 0.002 (0.001) (0.002) (0.002) (0.002) Q 0.008*** 0.009*** 0.009*** 0.008*** 0.007*** (0.001) (0.001) (0.001) (0.001) (0.001) Intercept 0.048* –0.017 0.032 0.044 0.089*** (0.027) (0.037) (0.035) (0.035) (0.006) Obs. 13,867 13,870 13,867 13,870 13,880 R 0.046 0.049 0.047 0.0450 0.045 F 17.46 12.57 12.28 11.61 11.84 This table reports results from regressing investment inefficiency on stock liquidity. All the estimates have been carried out using OLS. All models include year and industry fixed effects and constant term. Robust standard errors that are clustered by firm are present in parentheses under the coefficients. HL is bid-ask spread of Corwin and Schultz (2012), TOVER is turnover ratio, ILLIQ is Amihud’s(2002) illiquidity measure, LR is Amivest liquidity ratio, and GAM is return reversal developed by Pastor and Stambaugh (2003). FCF is free cash flow scaled by total assets, ORECTA is other receivables scaled by total assets, SIZE is the natural logarithm of total assets, LEV is total liabilities scaled by total assets, and Q is Tobin’s q. ***, **, and *denote significance at the 1%, 5%, and 10% levels, respectively. and GAM will reduce overinvestment by 39% and 42%, respectively. The results of Tables 3 and 4 show that stock liquidity improves investment efficiency, mitigating both overinvestment and underinvestment. 4.3. Robustness checks In this section, we perform several additional robustness tests of the reported results. 4.3.1. Endogeneity issues The OLS regressions are prone to the potential endogeneity problem, as improved investment efficiency may increase stock liquidity. To alleviate this concern, we use lagged stock liquidity to estimate the regression equation. In this subsection, we replace it with a second lag of stock liquidity and re-estimate equation (1). The results using a second lag of stock liquidity are similar to those previously reported, as displayed in Panel A of Table 5, supporting the hypothesis that higher stock liquidity helps improve investment efficiency. 4.3.2. Influence of funding constraints Brunnermeier and Pedersen (2009) show that stock liquidity reflects the funding avail- ability in capital markets, therefore, stock liquidity can influence firm investment through funding constraints. To distinguish this channel, following Goyenko and Ukhov (2009), we control for the growth rate of M2 and the Shanghai Interbank offered Rate China Journal of Accounting Studies 241 Table 3. Regressions of underinvestment on stock liquidity. (i) (ii) (iii) (iv) (v) HL ILLIQ GAM TOVER LR LIQ 0.023 0.002 0.004 –0.001* –0.002*** it-1 (0.121) (0.002) (0.005) (0.001) (0.000) FCF –0.089*** –0.089*** –0.089*** –0.089*** –0.114*** (0.009) (0.009) (0.009) (0.009) (0.009) ORECTA 0.007 0.007 0.007 0.008 0.037*** (0.009) (0.009) (0.009) (0.009) (0.009) SIZE –0.016*** –0.016*** –0.016*** –0.017*** (0.001) (0.001) (0.001) (0.001) Q 0.009*** 0.009*** 0.009*** 0.009*** 0.016*** (0.001) (0.001) (0.001) (0.001) (0.001) Intercept 0.416*** 0.421*** 0.412*** 0.428*** 0.044*** (0.024) (0.024) (0.023) (0.023) (0.005) N 7,856 7,859 7,856 7,859 7,864 R 0.191 0.191 0.191 0.192 0.159 F 37.30 37.72 37.36 37.57 30.65 This table reports results from regressing underinvestment on stock liquidity. All the estimates have been carried out using OLS. All models include year and industry fixed effects and constant term. Robust standard errors that are clustered by firm are present in parentheses under the coefficients. HL is bid-ask spread of Corwin and Schultz (2012), TOVER is turnover ratio, ILLIQ is Amihud’s(2002) illiquidity measure, LR is Amivest liquidity ratio, and GAM is return reversal developed by Pastor and Stambaugh (2003). FCF is free cash flow scaled by total assets, ORECTA is other receivables scaled by total assets, SIZE is the natural logarithm of total assets, LEV is total liabilities scaled by total assets, and Q is Tobin’s q. ***, **, and *denote significance at the 1%, 5%, and 10% levels, respectively. (Shibor) of 7 days (RATE). As shown in Panel B of Table 5, the coefficients on M2 are significantly negative at the 10% level, suggesting that monetary expansions relax fund- ing constraints and mitigate underinvestment. More importantly, our results remain unchanged after accounting for funding constraints. Stock liquidity is negatively associ- ated with investment inefficiency. 4.3.3. Influence of information environment Considering that stock liquidity reflects information asymmetries in the capital market, and high-quality financial reporting increases investment efficiency (Biddle et al., 2009; Li, 2009), a credible alternative interpretation of our results is that they may capture the effect of information asymmetries. To exclude this explanation, we re-estimate equation (1) after controlling for the absolute value of discretionary accruals using the performance matched discretionary accrual model (Kothari, Leone, & Wasley, 2005). Panel C of Table 5 reports the results. The coefficient estimates forjj DA are significant and positive at the 1% level, indicating that financial reporting quality mitigates invest- ment inefficiency, which is consistent with Biddle et al. (2009). More importantly, our previous findings are not affected. Firms with higher liquidity show higher investment efficiency. 4.3.4. Alternative measure of firm investment Following Chen et al. (2011) and Lv and Zhang (2011), we define cash payments for fixed assets, intangible assets, and other long-term assets minus cash receipts from 242 Xiong and Su Table 4. Regressions of overinvestment on stock liquidity. (i) (ii) (iii) (iv) (v) HL ILLIQ GAM TOVER LR LIQ 0.575** 0.039*** 0.042*** 0.005*** 0.000 it-1 (0.280) (0.008) (0.014) (0.001) (0.001) FCF –0.067*** –0.073*** –0.072*** –0.073*** –0.035 (0.023) (0.023) (0.023) (0.023) (0.023) ORECTA –0.096*** –0.107*** –0.099*** –0.086*** –0.112*** (0.023) (0.023) (0.023) (0.023) (0.025) SIZE 0.013*** 0.017*** 0.014*** 0.015*** (0.002) (0.002) (0.002) (0.002) Q –0.001 0.001 0.000 0.000 –0.008*** (0.003) (0.003) (0.003) (0.003) (0.003) Intercept –0.155*** –0.247*** –0.161*** –0.183*** 0.156*** (0.050) (0.053) (0.049) (0.051) (0.014) N 6,011 6,011 6,011 6,011 6,016 R 0.075 0.080 0.076 0.075 0.066 F 10.20 10.52 10.39 10.33 7.139 This table reports results from regressing overinvestment on stock liquidity. All the estimates have been car- ried out using OLS. All models include year and industry fixed effects and constant term. Robust standard errors that are clustered by firm are present in parentheses under the coefficients. HL is bid-ask spread of Corwin and Schultz (2012), TOVER is turnover ratio, ILLIQ is Amihud’s(2002) illiquidity measure, LR is Amivest liquidity ratio, and GAM is return reversal developed by Pastor and Stambaugh (2003). FCF is free cash flow scaled by total assets, ORECTA is other receivables scaled by total assets, SIZE is the natural loga- rithm of total assets, LEV is total liabilities scaled by total assets, and Q is Tobin’s q. ***, **, and *denote significance at the 1%, 5%, and 10% levels, respectively. selling these assets as total investment. The new firm investment measure is equal to total investment minus maintenance investment (namely, amortisation and depreciation), and then scaled by total assets. Based on this, we re-estimate expected investment model (2) and investigate the relationship between stock liquidity and investment ineffi- ciency. Our results, as shown in Panel D of Table 5, are similar to those previously reported. Stock liquidity is negatively associated with investment inefficiency. 4.3.5. Tobit models We use the absolute value of residuals of investment model (2) to measure underinvest- ment and overinvestment. Since the distribution of underinvestment and overinvestment are bounded to [0, +∞), the OLS would be biased, thus we use the Tobit model to esti- mate the panel regression of equation (1). The untabulated results are similar to the results reported in Tables 3 and 4. 4.3.6. Investment efficiency based on group classification The mean value of the residuals of the expected investment model (2) is zero, implying that on average investment inefficiency is zero. However, most firms may deviate from their optimal investment level, which contradicts the assumption of model (2). To alle- viate this concern, following Xin et al. (2007) and Biddle et al. (2009), we sort firms each year into three groups, based on the residuals from equation (2). Firm-year obser- vations in the bottom group (i.e., the most negative residuals) are classified as under- investing, while observations in the top group (i.e., the most positive residuals) are China Journal of Accounting Studies 243 Table 5. Stock liquidity and investment inefficiency: robustness checks. (i) (ii) (iii) (iv) (v) HL ILLIQ GAM TOVER LR Panel A: Lagged two periods stock liquidity and investment inefficiency LIQ 0.182 0.018*** 0.035*** 0.002*** 0.000 it-2 (0.130) (0.003) (0.007) (0.001) (0.001) N 13,205 13,302 13,205 13,301 13,311 R 0.044 0.047 0.046 0.044 0.043 Panel B: Stock liquidity and investment inefficiency controlled for M2 and interest rates LIQ 0.185 0.020*** 0.024*** 0.001* 0.000 it-1 (0.131) (0.004) (0.006) (0.001) (0.000) M2 –0.001** 0.000 –0.001* –0.001** –0.001 (0.001) (0.000) (0.000) (0.000) (0.000) RATE –0.002 0.002 –0.001 –0.003 –0.001 (0.003) (0.003) (0.003) (0.003) (0.003) N 13,867 13,870 13,867 13,870 13,880 R 0.046 0.049 0.047 0.046 0.045 Panel C: Stock liquidity and investment inefficiency controlled for information opaque LIQ 0.240* 0.020*** 0.022*** 0.002* 0.000 it-1 (0.131) (0.004) (0.006) (0.001) (0.000) DA 0.046*** 0.044*** 0.044*** 0.047*** 0.047*** jj (0.005) (0.011) (0.011) (0.011) (0.011) N 13,838 13,840 13,838 13,840 13850 R 0.050 0.053 0.051 0.050 0.050 Panel D: Stock liquidity and investment inefficiency based on INV2 LIQ –0.0370 0.006*** 0.006* 0.001** 0.000 it-1 (0.072) (0.002) (0.004) (0.001) (0.000) N 12,193 12,194 12,193 12,194 12,205 R 0.046 0.047 0.047 0.047 0.045 Panel E: Effect of liquidity on investment inefficiency using group classification LIQ 0.291* 0.021*** 0.020*** 0.002** 0.000 it-1 (0.171) (0.004) (0.007) (0.001) (0.001) N 9,307 9,313 9,307 9,313 9,318 R 0.049 0.052 0.049 0.049 0.048 Panel F: Alternative investment efficiency model LIQ 0.176 0.013*** 0.015** 0.003*** –0.001*** it-1 (0.118) (0.004) (0.006) (0.000) (0.000) N 15,471 15,600 15,471 15,599 15,610 R 0.057 0.058 0.057 0.059 0.057 This table reports results of robustness checks. All models include other control variables, constant term, year and industry fixed effects. Other control variables are omitted for brevity. Robust standard errors that are clus- tered by firm are present in parentheses under the coefficients. HL is bid-ask spread of Corwin and Schultz (2012), TOVER is turnover ratio, ILLIQ is Amihud’s(2002) illiquidity measure, LR is Amivest liquidity ratio, and GAM is return reversal developed by Pastor and Stambaugh (2003). M2 is the growth rate of M2, RATE is the Shanghai Interbank offered Rate (Shibor) of 7 days, and DA is the discretionary accruals using Kothari et al. (2005) model. ***, **, and * denote significance at the 1%, 5%, and 10% levels, respectively. 244 Xiong and Su classified as over-investing. We then estimate equation (1) using the absolute value of residuals that belong to the underinvestment and overinvestment group. The results, as presented in Panel E of Table 5, are similar to those previously reported. 4.3.7. Alternative investment efficiency model Following Biddle et al. (2009), we measure investment efficiency using a parsimonious expected investment model, which predicts investment level based on growth opportu- nities (as measured by Tobin’s Q) and use the absolute value of residuals as a proxy for investment inefficiency. The model is described in equation (11): X X INV ¼ a þ a Q þ a INDUSTRY DUMMIES þ a YEAR DUMMIES þ l it 0 1 it1 k t it k t (11) The results of estimating equation (1) using this investment efficiency proxy are simi- lar to those previously reported, as displayed in Panel F of Table 5. Higher stock liquidity helps mitigate investment inefficiency, mitigating both overinvestment and underinvestment. 5. How does liquidity improves investment efficiency? In this section, we perform several tests to examine the various explanations for why liquidity improves investment efficiency from the perspective of financial constraints, agency costs and information content of share prices. According to Baron and Kenny (1986) and Wen, Chang, Hau, and Liu (2004), we propose the following recursive regression equations to test the mediation effects of financial leverage, agency costs and information content of share prices: 8 P P P INEFFINV ¼ / þ / LIQ þ / Control þ / INDUSTRY þ / YEAR þ 1 ð12Þ > it 0 1 it1 j j k t it j k t P P P > MEDIATOR ¼ h þ h LIQ þ h Control þ h INDUSTRY þ h YEAR þ s ð13Þ it 0 1 it1 j j k t it j k t P P 0 0 0 0 0 > INEFFINV ¼ / þ / LIQ þ / MEDIATOR þ / Control þ / INDUSTRY it it1 it j 0 1 2 j k > k þ / YEAR þ x ð14Þ t it The rationale behind this method is as follows. We first fit equation (12) and obtain the coefficient estimates for LIQ . A statistically significant ϕ suggests that stock liquid- it–1 1 ity influences investment efficiency. We then estimate equations (13) and (14). If both θ and / are statistically significant, then the relation between stock liquidity and investment efficiency is intermediated by the mediating variable. Furthermore, when 0 0 the coefficient estimate for / of equation (14) is significant, an insignificant / means 2 1 that the effect of liquidity on investment inefficiency is mediated entirely by the medi- ating variable. If either θ or / is statistically insignificant, we can use the Sobel (1986) test to check the mediation effect by examining the significance of the product 0 0 of coefficients h / . The standard error of h / , derived by the multivariate delta 1 1 2 2 qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi 2 2 0 2 method (Sobel, 1986), is equal to S ¼ h S þðÞ / S . 1 0 h / 2 1 h 2 / 1 2 China Journal of Accounting Studies 245 Table 6. Mediating effect tests. (i) (ii) (iii) (iv) (v) HL ILLIQ GAM TOVER LR Panel A: The mediation effects of capital structure (MEDIATOR=LEV) ϕ 0.199 0.021*** 0.025*** 0.002* 0.000 (0.139) (0.004) (0.006) (0.001) (0.000) θ 0.979* 0.097*** 0.205*** –0.001 –0.003*** (0.519) (0.015) (0.026) (0.002) (0.001) / –0.002 –0.005 –0.003 –0.002 0.000 (0.004) (0.005) (0.005) (0.005) (0.005) / 0.262* 0.020*** 0.021*** 0.002** 0.000 (0.140) (0.004) (0.006) (0.001) (0.000) Sobel test –0.483 0.312 –0.988 0 –0.598 Panel B: The mediation effects of agency costs (MEDIATOR=MCOST) ϕ 0.199 0.021*** 0.025*** 0.002* 0.000 (0.139) (0.004) (0.006) (0.001) (0.000) θ 0.008 0.031*** 0.040** –0.003*** –0.006*** (0.162) (0.011) (0.019) (0.001) (0.001) / 0.021*** 0.022*** 0.022*** 0.021*** 0.019*** (0.004) (0.004) (0.004) (0.004) (0.004) / 0.262* 0.020*** 0.021*** 0.002** 0.000 (0.140) (0.004) (0.006) (0.001) (0.000) Sobel test 0.049 –2.605 2.508 –3.724 1.966 Panel C: The mediation effects of stock price informativeness (MEDIATOR=SYN) ϕ 0.199 0.021*** 0.025*** 0.002* 0.000 (0.139) (0.004) (0.006) (0.001) (0.000) θ 2.896*** –0.112*** –0.196*** 0.022*** 0.006** (0.871) (0.020) (0.039) (0.003) (0.003) / –0.006*** –0.006*** –0.006*** –0.007*** –0.007*** (0.001) (0.001) (0.001) (0.001) (0.001) / 0.262* 0.020*** 0.021*** 0.002** 0.000 (0.140) (0.004) (0.006) (0.001) (0.000) Sobel test –2.908 –5.063 4.094 –1.923 3.853 This table reports results of the mechanisms through which liquidity mitigates investment inefficiency. All the estimates have been carried out using OLS. Robust standard errors that are clustered by firm are present in parentheses under the coefficients. HL is bid-ask spread of Corwin and Schultz (2012), TOVER is turnover ratio, ILLIQ is Amihud’s(2002) illiquidity measure, LR is Amivest liquidity ratio, and GAM is return reversal developed by Pastor and Stambaugh (2003). ***, **, and * denote significance at the 1%, 5%, and 10% levels, respectively. 5.1. The mediating role of capital structure We begin by positing that the association between stock liquidity and investment effi- ciency relates to the relaxation of financial constraints. Myers and Majluf (1984) point out that when managers have more information than outside shareholders, informed managers are motivated to time the issuance of overpriced shares, and rational investors will discount new issues regardless of their quality. This means that firms may refuse to raise equity capital even if that means passing up positive net present value projects, with the consequence of underinvestment. 246 Xiong and Su Prior studies find that firms with higher liquidity stock maintain a lower cost of equity capital (Amihud, 2002; Amihud & Mendelson, 1986), and that firms can make use of higher liquidity to issue stock at low cost. The relaxation of financial constraints will mitigate underinvestment and improve investment efficiency. Since firms with higher liquidity stock are more likely to use equity financing, lead- ing to relatively lower usage of debt, we can examine the cost of capital mechanism by linking stock liquidity with capital structure. Panel A of Table 6 reports regression estimates of equations (12) to (14) using capi- tal structure as a mediating variable. In the equation for capital structure determination, the coefficient estimates for HL, ILLIQ, and GAM are all significantly positive at the 10% level or above, while the coefficient for LR is negative and significant at the 1% level, indicating a significant negative relation between leverage and liquidity. However, the coefficient estimates for mediating variables (LEV) in equation (14) are all insignifi- cant, thus, we use the Sobel (1986) test to examine the mediation effect. The last row of Panel A reports the results of the Sobel test. We find that except for column (iii), all other tests are insignificant at the 5% level, indicating that the mediation effect of financial leverage is not significant. The reason why financial leverage fails to serve as a mediating variable is that the stock market in China is controlled by the government and that the China Securities Regulatory Commission uses an earnings threshold as a criterion for seasoned equity offerings (He et al., 2013). The earnings threshold prevents financially constrained com- panies from issuing stock in the seasoned equity market, even though they want to take advantage of higher liquidity. 5.2. The mediating role of agency costs This section examines the mediating role of agency costs. Conflicts of interest between managers and outside shareholders and a lack of monitoring may lead managers to pur- sue their own interests by making investments that are not in the best interests of share- holders (Jensen & Meckling, 1976). Jensen (1986) predicts that managers will grow their firm beyond the optimal size, resulting in overinvestment. However, managers pre- ferring to enjoy the quiet life are likely to invest less and forgo some positive net pres- ent value projects, with subsequent underinvestment (Betrand & Mullainathan, 2003). Prior studies argue that stock liquidity can reduce agency costs and improve invest- ment efficiency by permitting more efficient managerial compensation (Holmstrom & Tirole, 1993), facilitating blockholder intervention (Maug, 1998), and enhancing the power of exit threats (Adamati & Pfleiderer, 2009; Edmans, 2009). In this manner, the agency costs may mediate the relationship between stock liquidity and investment efficiency. Panel B of Table 6 reports regression estimates of equations (12) to (14) using agency costs as a mediating variable. In the regression of agency costs on stock liquid- ity, the coefficients on ILLIQ and GAM are significantly positive at the 5% level, while coefficient estimates for TOVER and LR are negative and significant at the 1% level, indicating a negative association between stock liquidity and agency costs. Moreover, the coefficients on MCOST are all significantly positive at the 1% level when control- ling for the effect of stock liquidity, indicating that firms with higher agency costs show lower investment efficiency. According to Baron and Kenny (1986) and Wen et al. (2004), the significance of θ and / indicates that the mediating effect of agency cost is statistically significant. This conclusion is confirmed by the Sobel test in the last row China Journal of Accounting Studies 247 of Panel B of Table 6. The results of the Sobel test suggest that, except for column (i), the mediating effects of agency costs are significant at the 1% level. 5.3. The role of stock price informativeness This section examines whether stock liquidity stimulates the entry of informed investors who increase the informativeness of the stock price and then improve investment effi- ciency. Holmstrom and Tirole (1993) predict that stock liquidity enables an informed trader to disguise private information and profit from it. The increased informativeness of the stock price may contain information that managers do not have. The information in turn can guide a manager in making corporate decisions, such as the decision on cor- porate investment, improving investment efficiency (Chen et al., 2007). Moreover, the increased informativeness of the stock prices reduces information asymmetries between managers and outside shareholders, and increases the ability of investors to monitor management, which in turn improves the capital allocation of the firm (Durnev et al., 2004; Yang & Nie, 2010). Hence, the positive effect of liquidity on investment efficiency should be proportionally mediated by stock price informativeness. Panel C of Table 6 reports regression estimates of equations (12) to (14) using stock price informativeness as a mediating variable. In the regression of stock price informativeness on liquidity, the coefficients on ILLIQ and GAM are significant and negative at the 1% level, while coefficients estimates for TOVER and LR are positive and significant at the 5% level, suggesting liquidity stimulates the entry of informed investors and increases the information content of share prices. Moreover, the coeffi- cients on SYN are all significantly negative at the 1% level when controlling for the effect of stock liquidity, indicating that stock price informativeness helps improve investment efficiency. The significance of θ and / suggests that the relation between stock liquidity and investment efficiency is mediated by stock price informativeness. The results of the Sobel test confirm that the mediating effects of stock price informa- tiveness are significant at the 1% level. 6. Conclusion and policy implications This paper investigates the effect of stock liquidity on firm investment efficiency in a sample of Chinese listed non-financial firms from 1998 to 2011. We find that invest- ment inefficiency, as measured by the deviation from the optimal investment level, is negatively related to stock liquidity. Moreover, if we distinguish between overinvest- ment and underinvestment, we find that stock liquidity mitigates both overinvestment and underinvestment. This finding holds for multiple measures of stock liquidity and investment efficiency and is robust to controlling for other known determinants of investment efficiency and for alternative explanations. Furthermore, we document the channels by which stock liquidity relates to investment efficiency: reducing agency costs and increasing the information content of share prices. Overall, our results are consistent with the idea that stock liquidity helps improve investment efficiency, miti- gating both overinvestment and underinvestment. This paper finds that stock liquidity helps improve investment efficiency. The policy implication from our empirical results is that firms and government should take actions to improve the stock liquidity of firms. We note that, from prior research, Ascioglu, Hegde, Krishnan, and McDermott (2012) and Chung, Elder, and Kim (2010) find that corporate governance and financial reporting quality help improve stock liquidity. 248 Xiong and Su Consequently, we link their findings to ours and propose that firms should reform their corporate governance and strengthen their information disclosure to improve stock liquidity and hence improve investment efficiency. We also suggest that, to improve investor protection and market liquidity, the Chinese government should continue to reform ownership structure and corporate governance, strengthen information disclosure and step up its actions against insider trading. Acknowledgements We thank Haijian Zeng, Liansheng Wu (Joint editor), Jason Xiao (Joint editor), Pauline Weetman (Language editor), Jigao Zhu, Yi Wen, and two anonymous referees for many constructive com- ments that have helped to improve the quality of the paper. We also thank Yongfu Yang for his comments at the China Journal of Accounting Studies conference in Guangzhou. Su gratefully acknowledges financial support from the National Natural Science Foundation of China (Grant No. 71173090), National Social Science Foundation of China (Grant No. 11AJY013), the Guang- dong Pearl River Scholar Fund, the Guangdong Natural Science Foundation (Grant No.S2011010004257) and the Fundamental Research Funds for the Central Universities (12JNYH001). Notes 1. The preceding discussion explores how stakeholders, such as shareholders, creditors, and managers, learn from stock liquidity information when making financial decisions and moni- toring managerial behaviour. Liquidity may also influence firm investment through the avail- ability of funding. Brunnermeier and Pedersen (2009) show that margin spiral and loss spiral may be mutually reinforced, which leads to liquidity spirals. When the margin is positively associated with market illiquidity, a funding shock to the speculators reduces the market liquidity, which leads to higher margins and further tightens the funding constraints of specu- lators. Funding shock and increased funding constraints may drive speculators to sell their initial position, which decreases prices further. The loss spiral and margin spiral reinforce each other, which implies a larger total effect and leads to liquidity spirals or financial crises. Goyenko and Ukhov (2009) find that ‘micro’ or transaction liquidity is closely related to ‘macro’ liquidity or money flow. Illiquidity generally increases as the monetary policy is tightened. Therefore, stock liquidity reflects the availability of funding. The funding con- straint is relaxed when the liquidity is high. This condition helps firms alleviate underinvest- ment by raising their capital and expanding their investments. We thank an anonymous referee for pointing out this idea. 2. These two papers propose a similar regression to test the effect of government control and executive compensation on investment inefficiency. We use their basic regression model and similar control variables to investigate the effect of stock liquidity on investment inefficiency. 3. Since firm investment is influenced by firm-specific unobservable factors, such as corporate culture and managerial characteristics, we use the panel data firm-fixed effect model to esti- mate equation (2). We also fit equation (2) using the OLS method and test the hypothesis. The results are similar to those reported in Table 3 to Table 5. 4. The correlation coefficient between Amivest liquidity ratio (LR) and firm size (SIZE)isas high as 0.6, therefore we exclude firm size (SIZE) when we use the Amivest liquidity ratio (LR) as a liquidity measure. 5. The independent variables of the capital structure determination equation include lagged stock liquidity (LIQ ), firm size (SIZE), Tobin’sQ(Q), tangibility (TANG), non-debt tax it-1 shields (NDTS), return on assets (ROA), firm age (AGE), and industry and year dummies. 6. MacKinnon, Lockwood, Hoffman, West, and Sheets (2002) argue that the tests for the medi- ating effect based on normal theory can yield inaccurate confidence limits and significance tests. They find that alternative tests based on the asymmetric distribution of the product of two normally distributed variables outperform traditional methods. The tables of critical val- ues can be downloaded from http://www.public.asu.edu/~davidpm/ripl/methods.htm. China Journal of Accounting Studies 249 7. The rest of the control variables include firm size (SIZE), leverage (LEV), Tobin’sQ(Q), free cash flow (FCF), the percentage shareholdings of the largest shareholder (TOP1), the sum of percentage shareholdings by second- to fifth-largest shareholders (TOP2_5), government con- trol (STATE), board size (BOARD), board independence (IND), the duality of Chairman and CEO (DUAL), and industry and year dummies. 8. The independent variables of the stock price informativeness determination equation include lagged stock liquidity (LIQ ), firm size (SIZE), leverage (LEV), Tobin’ sQ(Q), return on it–1 assets (ROA), the percentage shareholdings of the largest shareholder (TOP1), the sum of percentage shareholdings by second- to fifth-largest shareholders (TOP2_5), government con- trol (STATE), board size (BOARD), board independence (IND), the duality of Chairman and CEO (DUAL), the marketisation index (MKT), and industry and year dummies. References Admati, A., & Pfleiderer, P. (2009). The ‘Wall Street Walk’ and shareholder activism: Exit as a form of voice. Review of Financial Studies, 22, 2645–2685. Aggarwal, R., & Samwick, A. (2006). Empire builders and shirkers: Investment, firm perfor- mance and managerial incentives. Journal of Corporate Finance, 12, 489–515. Allen, F., Qian, J., & Qian, M.J. (2005). Law, finance, and economic growth in China. Journal of Financial Economics, 77,57–116. Amihud, Y. (2002). Illiquidity and stock returns: Cross-section and time-series effects. Journal of Financial Markets, 5(1), 31–56. Amihud, Y., & Mendelson, H. (1986). Asset pricing and the bid-ask spread. Journal of Financial Economics, 17, 223–249. Ang, J.S., Coles, R., & Lin, J.W. (2000). Agency costs and ownership structure. Journal of Finance, 58(1), 211–221. Ascioglu, A., Hegde, S.P., Krishnan, G.V., & McDermott, J.B. (2012). Earnings management and market liquidity. Review of Quantitative Finance and Accounting, 38(2), 257–274. Baker, M., & Stein, J.C. (2004). Market liquidity as a sentiment indicator. Journal of Financial Markets, 7, 271–299. Baron, R.M., & Kenny, D.A. (1986). The moderator-mediator variable distinction in social psychological research: Conceptual, strategic, and statistical considerations. Journal of Personality and Social Psychology, 51, 1173–1182. Betrand, M., & Mullainathan, S. (2003). Enjoying the quiet life? Corporate governance and managerial preferences. Journal of Political Economy, 111, 1043–1075. Biddle, G.C., Hilary, G., & Verdi, R.S. (2009). How does financial reporting quality relate to investment efficiency. Journal of Accounting and Economics, 48,112–131. Blanchard, O., Lopez-de-Silanez, F., & Shleifer, A. (1994). What do firms do with cash wind- falls? Journal of Financial Economics, 36, 337–360. Brandt, L., & Li, H.B. (2003). Bank discrimination in transition economies: Ideology, informa- tion, or incentives? Journal of Comparative Economies, 31, 387–413. Brunnermeier, M.K., & Pedersen, L.H. (2009). Market liquidity and funding liquidity. Review of Financial Studies, 22(6), 2201–2238. Cai, W.X., & Zeng, C. (2012). Market competition, property-right reform, and the lending activi- ties of commercial banks. Journal of Financial Research, (2), 73–87 (in Chinese). Chen, Q., Goldstein, I., & Jiang, W. (2007). Price informativeness and investment sensitivity to stock price. Review of Financial Studies, 20, 619–650. Chen, S., Sun, Z., Tang, S., & Wu, D. (2011). Government intervention and investment efficiency: Evidence from China. Journal of Corporate Finance, 17, 259–271. Cheng, Z., Xia, X., & Yu, M. (2008). Government intervention, pyramidal shareholding, and overinvestment of SOEs affiliated to local government. Management World, (9), 37–47 (in Chinese). Chung, K.H., Elder, J., & Kim, J. (2010). Corporate governance and liquidity. Journal of Financial and Quantitative Analysis, 45, 265–291. Corwin, S.A., & Schultz, P. (2012). A simple way to estimate bid-ask spreads from daily high and low price. Journal of Finance, 67, 719–760. Durnev, A., Morck, R., & Yeung, B. (2004). Value-enhancing capital budgeting and firm-specific stock return variation. Journal of Finance, 59(1), 65–105. 250 Xiong and Su Edmans, A. (2009). Blockholder trading, market efficiency, and managerial myopia. Journal of Finance, 64, 2481–2513. Fang, V., Noe, T. H., & Tice, S. (2009). Stock market liquidity and firm value. Journal of Financial Economics, 94, 150–169. Goyenko, R.Y., & Ukhov, A.D. (2009). Stock and bond market liquidity: A long-run empirical analysis. Journal of Financial and Quantitative Analysis, 44(1), 189–212. He, J., Mao, X.Y., Rui, M., & Zha, X.L. (2013). Business groups in China. Journal of Corporate Finance, 22, 166–192. Holmstrom, B., & Tirole, J. (1993). Market liquidity and performance monitoring. Journal of Political Economy, 101, 678–709. Jayaraman, S., & Milbourn, T.T. (2012). The role of stock liquidity in executive compensation. The Accounting Review, 87, 537–563. Jensen, M.C. (1986). Agency costs of free cash flow, corporate finance and takeovers. American Economic Review, 76, 323–329. Jensen, M.C., & Meckling, W. (1976). Theory of the firm: Managerial behavior, agency cost and ownership Structure. Journal of Financial Economics, 3, 305–360. Khanna, N., & Sonti, R. (2004). Value creating stock manipulation: Feedback effect of stock prices on firm value. Journal of Financial Markets, 7, 237–270. Kothari, S.P., Leone, A.J., & Wasley, C.E. (2005). Performance matched discretionary accrual measures. Journal of Accounting and Economics, 39(1), 163–197. Kyle, A.S. (1985). Continuous auctions and insider trading. Econometrica, 53, 1315–1335. Kyle, A.S., & Vila, J. (1991). Noise trading and takeovers. Rand Journal of Economics, 22, 54–71. Li, Q. (2009). Accounting information quality and capital allocation efficiency: Empirical evi- dence of Chinese listed corporations. Nankai Business Review, (12), 115–124 (in Chinese). Lipson, M.L., & Mortal, S. (2009). Liquidity and capital structure. Journal of Financial Markets, 12,611–644. Lv, C., & Zhang, H. (2011). The effect of the stock option plans on corporate investment behav- iors. Management World, (11), 118–126 (in Chinese). Mackinnon, D.P., Lockwood, C.M., Hoffman, J.M., West, S.G., & Sheets, V. (2002). A compari- son of methods to test the significance of mediation and other intervening variable effects. Psychological Methods, 7,83–104. Maug, E. (1998). Large shareholders as monitors: Is there a trade-off between liquidity and con- trol? Journal of Finance, 53(1), 65–98. Modigliani, F., & Miller, M.H. (1958). The cost of capital, corporate finance and the theory of investment. American Economic Review, 48, 261–297. Munoz, F. (2013). Liquidity and firm investment: Evidence from Latin America. Journal of Empirical Finance, 20,18–29. Myers, S.C., & Majluf, N.S. (1984). Corporate financing and investment decisions when firms have information that investors do not have. Journal of Financial Economics, 13, 187–221. Pastor, L., & Stambaugh, R.R. (2003). Liquidity risk and expected stock returns. Journal of Political Economy, 111, 642–685. Petersen, M. (2009). Estimating standard errors in finance panel data sets: Comparing approaches. Review of Financial Studies, 22, 435–480. Polk, C., & Sapienza, P. (2009). The stock market and corporate investment: A test of catering theory. Review of Financial Studies, 22(1), 187–216. Richardson, S. (2006). Over-investment of free cash flow. Review of Accounting Studies, 11, 159–189. Ross, S.A., Westerfield, R.W., & Jaffe, J. (2012). Corporate finance. New York: McGraw-Hill/ Irwin. Singh, M., & Davidson, W.N. (2003). Agency costs, ownership structure and corporate gover- nance mechanisms. Journal of Banking and Finance, 27, 793–816. Sobel, M.E. (1986). Some new results on indirect effects and their standard errors in covariance structure models. Sociological Methodology, 16, 159–186. Stein, J. (2003). Agency, information and corporate investment. In G.M. Constantinides, M. Harris, and R.M. Stulz (Eds.). Handbook of the economics of finance. Amsterdam, the Netherlands: North Holland. China Journal of Accounting Studies 251 Stulz, R. (1990). Managerial discretion and optimal financing policies. Journal of Financial Eco- nomics, 26,3–27. Su, D. (2005). Corporate finance and state enterprise reform in China. China Economic Review, 16,118–148. Tang, X., Zhou, X., & Ma, R. (2007). Empirical research on overinvestment behavior and its restriction system in China’s listed companies. Accounting Research, (7), 44–52 (in Chinese). Tobin, J. (1969). A general equilibrium approach to monetary theory. Journal of Money, Credit and Banking, 1(1), 15–29. Wang, C., & Chin, S. (2004). Profitability of return and volume-based investment strategies in China’s stock market. Pacific-Basin Finance Journal, 12, 541–564. Wei, Z., Xie, F., & Zhang, S. (2005). Ownership structure and firm value in China’s privatized firms: 1991-2001. Journal of Financial & Quantitative Analysis, 40,87–108. Wen, Z. L., Chang, L., Hau, K.T., & Liu, H.Y. (2004). Testing and application of the mediating effects. Acta Psychologica Sinica, 36, 614–620 (in Chinese). Xin, Q., Lin, B., & Wang, Y. (2007). Government control, executive compensation and capital investment. Economic Research Journal, (8), 110–122 (in Chinese). Yang, J., & Nie, S. (2010). Stock price informativeness and capital allocation efficiency of corpo- rate. Journal of Management Science, (23), 81–90 (in Chinese). Zhang, G., & Song, X. (2009). Measuring inefficient investment by listed companies in China: Overinvestment or underinvestment. Accounting Research, 5,69–77 (in Chinese). Zhong, H., Ran, M., & Wen, S. (2010). Government intervention, insider Control and corporate investment. Management World, (7), 98–108. (in Chinese).Appendix A. Appendix A. Definitions of variables. Variables Definition Panel A: Main variables of interest Firm investment (INV) Change in gross value of fixed assets, construction in progress, project material and intangible assets minus amortization and depreciation, scaled by lagged total assets. Firm investment 2 (INV2) Cash payments for fixed assets, intangible assets, and other long-term assets from the cash flow statement minus cash receipts from selling these assets and amortization and depreciation, scaled by lagged total assets. Investment inefficiency The absolute value of residuals of expected investment model (INEFFINV) (2) using INV as dependent variable. Overinvestment (OverINV) The positive residuals of expected investment model (2) using INV as dependent variable. Underinvestment (UnderINV) The absolute value of negative residuals of expected investment model (2) using INV as dependent variable. Panel B: Stock liquidity variables Bid-ask spread (HL) See equation (3). Turnover ratio (TOVER) Total shares traded in the day divided by total tradable shares, see equation (4). Illiquidity ratio (ILLIQ) Absolute value of stock return divided by dollar volume, see equation (5). Liquidity ratio (LR) Price change induced by order flow, see equation (6). Return reversal (GAM) The absolute value of coefficient estimate for signed dollar volume in equation (7). (Continued) 252 Xiong and Su Appendix A.(Continued). Variables Definition Panel C: Control variables Firm size (SIZE) The natural logarithm of total assets. Leverage (LEV) Total liabilities scaled by total assets. Tobin’sQ(Q) The sum of market value of tradable shares, book value of non-tradable shares, and liabilities, scaled by book value of total assets. Cash holding (CASH) The ratio of cash holdings to total assets. Stock return (RET) Stock return. Firm age (AGE) The numbers of years since the firm listed. Cash flow (CFO) Cash flow from operations scaled by total assets. Free cash flow (FCF) Cash flow from operations minus expected investment level scaled by total assets. Agency costs (MCOST) The sum of selling expenses and general and administrative expenses scaled by total assets. Tunnelling (ORECTA) Other receivables scaled by total assets. Stock price informativeness Stock price non-synchronicity, see equations (9) and (10). (SYN) Tangibility (TANG) The ratio of tangible fixed assets to total assets. Non-debt tax shields (NDTS) Depreciation scaled by total assets. Return on assets (ROA) Net income divided by total assets. Earning quality (jj DA ) The absolute value of discretional accruals using the performance matched discretionary accrual model (Kothari et al., 2005). Panel D: Ownership and corporate governance variables % of shares by top shareholder The percentage shareholdings of the largest shareholder. (TOP1) nd th % of shares by 2 to 5 The sum of percentage shareholdings by second to fifth largest shareholders (TOP2_5) shareholders. Government control (STATE) STATE is equal to 1 if it is ultimately controlled by the government. Board size (BOARD) The number of directors sitting on the board. Board independent (IND) Ratio of the number of independent directors to board size. Duality (DUAL) A dummy variable, which is equal to 1 if the CEO is also the Chairman.

Journal

China Journal of Accounting StudiesTaylor & Francis

Published: Jul 3, 2014

Keywords: capital allocation efficiency; market microstructure; overinvestment; stock liquidity; underinvestment

References