Institutional Ownership and Corporate Investment Performance

Abstract

This paper examines the role of institutional investors in improving firm performance through the channel of corporate investment decisions. We find that the interaction effect between institutional ownership and capital expenditures is significantly related to firm performance. We examine this relationship for different types of institutional investors, and find that investment advisors are most effective monitors in improving firm performance through corporate investment. Moreover, we find that the monitoring role of institutional investors becomes more important when internal governance is weak. Institutional ownership and other forms of corporate governance mechanisms (including CEO incentive compensation and control, shareholder right provisions, and board of director monitoring) operate as substitutes, rather than complements, in improving capital expenditure decisions. Copyright © 2012 ASAC. Published by John Wiley & Sons, Ltd.

Résumé

Cet article examine le rôle que les investisseurs institutionnels jouent dans l'amélioration de la performance des entreprises à travers les décisions de placement des sociétés. Il montre que l'effet d'interaction entre la propriété institutionnelle et les dépenses en capital est fortement lié à la performance d'entreprise. L'examen de cette relation pour différents types d'investisseurs institutionnels révèle que les conseillers en investissement sont les mieux placés pour veiller à l'amélioration de la performance d'entreprise à travers les placements de la société. Il ressort des résultats que le rôle de surveillance attribué aux investisseurs institutionnels gagne en importance lorsque la gouvernance interne devient faible. La propriété institutionnelle et les autres formes de mécanismes de gouvernance d'entreprise (y compris le contrôle des PDG et les compensations qui leur sont accordées, les dispositions relatives au droit des actionnaires, le suivi du conseil d'administration) fonctionnent non pas comme de suppléments, mais plutôt comme des substituts dans l'amélioration des décisions de dépense en capital.Copyright © 2012 ASAC. Published by John Wiley & Sons, Ltd.

The rising importance of institutional investors is evident as they become more visibly active in influencing major corporate decisions and hence firm performance. Their activities range from improving corporate control and governance practice to improving the industry's capacity and firm's investment competitiveness. Nevertheless, prior studies on institutional ownership's effect on firm performance are inconclusive. Shleifer & Vishny (1986), McConnell & Servaes (1990), Smith (1996), Del Guercio & Hawkins (1999), and Gompers & Metrick (2001) found a positive relation between institutional ownership and firm performance. In contrast, Agrawal & Knoeber (1996), Karpoff, Malatesta, & Walkling (1996), Duggal & Millar (1999), and Faccio & Lasfer (2000) found no such significant relation.

1The mixed evidence might be a result of the failure to take into account the specificity of corporate decisions, the heterogeneity of institutional investors, and the interactions between institutional investors and other corporate governance mechanisms.

We set out to examine the following three connected questions that are relatively under-studied in the literature. First, can institutional investors improve firm value through the channel of corporate investment policies? In principle, institutional investors can influence firm value through different channels of corporate activities. While existing literature investigates the impact of institutional investors on other corporate activities (Bushee, 1998; Gaspar, Massa, & Matos, 2005; Chen, Harford, & Li, 2007), this paper examines the impact of institutional investors on firms’ capital expenditure decisions, which represent an important component of corporate investment policies and a key determinant of firm value.

Second, do different types of institutional investors have differential impacts on corporate investment performance? This study considers that institutional investors are heterogeneous economic agents when performing governance and influencing corporate decisions. We examine different types of institutional investors (including investment advisors, hedge funds, pension funds, banks, and insurance companies) based on their differences in incentives and monitoring effectiveness.

Third, how do institutional investors interact with other internal governance mechanisms? This study examines how the effectiveness of external governance mechanisms (institutional ownership) changes with different internal governance mechanisms (including managerial incentives and control, corporate governance provisions, and board characteristics).

Methodologically, this paper features a new empirical approach to answer these questions based on the hypothesis that firms make better investment decisions and improve performance when institutional investors can provide monitoring activities. To examine how monitoring by institutional investors affect corporate investment decision and ultimately firm performance, we estimated the effect of the interaction between institutional ownership and capital expenditures on firm performance. To disentangle the monitoring role of institutional investors from other possible functions, we examine such effect with different categories of institutional investors (by identifying institutional investor types that are better monitors) and with stratifications of firm types (by identifying firm types that are more prone to governance problems).

Our dataset, obtained from COMPUSTAT, CRSP, Thomson Ownership Database, ExecuComp, and IRRC RiskMetrics, comprises a sample of US firms in the NYSE, AMEX, and NASDAQ for the period between 1997 and 2006. Based on this sample, we document the following sets of new findings.

First, we found that the interaction between institutional ownership and capital expenditures has a significant and positive impact on future firm performance. We also found that a high level of institutional ownership is significantly related to improved firm performance through institutional investors’ monitoring and influence over capital investment decisions.

Second, we found that different types of institutional investors play distinctive roles in influencing capital expenditure decisions. Consistent with the monitoring hypothesis, we demonstrate that more independent and informed investors, such as investment advisors, are better suited to monitor and improve firm performance through capital expenditure decisions. In contrast, we found that short-term investors such as hedge funds, and less independent grey investors, such as insurance companies and banks, do not improve firm performance through capital expenditure decisions.

Third, we found consistent evidence that the monitoring role of institutional investors becomes more important when internal governance is weak. Institutional ownership, particularly ownership by investment advisors, has a more significant positive impact on corporate investment performance for firms with: (a) CEOs’ incentives that are not well-aligned with firm value; (b) limited internal governance provisions that support shareholder rights; and (c) weak governance provided by the board of directors. These findings suggest that institutional ownership operates as a substitute for other forms of governance within a firm.

Lastly, our findings are robust with respect to alternative samples and tests. We conducted further robustness test to address the potential endogeneity problem by showing that institutional investors can lead to better investment performance after firms have experienced deterioration in prior performance.

Overall, our findings contribute to the literature by demonstrating the importance of institutional investors in improving firm performance through the channel of corporate investment decisions. While the literature provides little consensus on whether institutional ownership and other corporate governance mechanisms are complements or substitutes to each other, we provide new evidence that institutional investors play a more important monitoring role when internal governance is weak. Lastly, our findings shed light on understanding what type(s) of institutional investors are more effective in improving firm performance. We found that informed and independent institutional investors are more effective monitors. The remainder of the paper is organized as follows. We first present the literature review and hypotheses followed by a discussion of the empirical framework and data. We then present our main findings and interpretations and conclude the paper with a discussion of the academic and practical implications of our study.

Literature Review and Hypotheses

Institutional Ownership and Firm Performance

Existing literature suggests that a key activity for institutional investors is monitoring management (Shleifer & Vishny, 1986) and that effective monitoring can improve firm performance by reducing agency costs (Admati, Pfleiderer, & Zechner, 1994; Jensen & Meckling, 1976; Shleifer & Vishny, 1986).2 In contrast, other studies, including Kahn & Winton (1998) and Noe (2002), suggest that institutional investors may choose to benefit from the information asymmetry from outsiders and avoid the cost of activism. This paper focuses on institutional investors’ influence on corporate investment decisions, a research direction which is relatively under-studied.3 Firms’ investing activities are generally more subject to managerial discretion, but are also more observable for outsiders; thus they may receive more attention from institutional investors than operating and financing activities and hence serve as an important channel through which institutional investors influence firm value. Through equity ownership, institutional investors can effectively acquire information, perform monitoring functions, and thereby improve corporate investment policies. As such, we hypothesize:

H 1. An increasing level of institutional ownership is significantly related to improved firm performance through capital expenditure decisions.

Different Types of Institutional Investors

As Sherman, Beldona, & Joshi (1998) argue, past research is unable to establish a consistent relation between institutional investors and firm decisions because institutional investors are assumed to be a homogenous group possessing the same objectives and behaviours. Different institutional investors may play different roles in providing monitoring and governance activities (Almazan, Hartzell, & Starks, 2005; Brickley, Lease, & Smith, 1988; Chen et al., 2007). Chen et al. (2007) showed that independent institutions with long-term investments specialize in monitoring during takeover. Cornett, Marcus, Saunders, & Tehranian (2007) found that institutional investors with fewer potential business ties to target firms are better monitors in improving firm's operating performance.4 However, the existing evidence on which type(s) of institutional investors are most effective in performing governance is still mixed. As such, we develop the following hypothesis:

H 2. Among different types of institutional investors, informed and independent investors should be more effective in improving capital expenditure decisions.

Corporate Governance Mechanisms

As Denis & McConnell (2003) argue, both internal governance mechanisms and institutional ownership help improve corporate governance. Nonetheless, whether institutional ownership and other corporate governance mechanisms operate as substitutes or complements in improving firm performance remains imperfectly understood. As Shleifer & Vishny (1986) argue, poor governance presents a better opportunity for institutional investors to improve firm performance through their advisory and monitoring ability. We developed the following hypothesis that highlights the nexus between internal governance mechanisms and institutional ownership:

H 3. The positive association between institutional ownership and corporate investment performance is stronger for firms with poor internal governance.

By testing H3, we examined whether internal governance (such as managerial incentives and control, corporate governance provisions, and board characteristics) enhances or weakens the monitoring effect of institutional investors.

Empirical Framework and Data

We analyzed a sample of US publicly traded firms, excluding finance and utilities, from NYSE, AMEX, and NASDAQ over the period of 1997 to 2006. We obtained annual financial data from COMPUSTAT and CRSP. The percentage equity stake data for institution investors as of June-end of every year were obtained from the Thomson Ownership Data, which has data available from year 1997.5 Institutional ownership data were then matched to the following fiscal year financial data from COMPUSTAT and CRSP. We then merged the dataset using ExecuComp (which mainly covers the S&P 1500 firms plus other publicly traded companies and provides executive compensation data from company's annual proxy (DEF14A SEC form)) and IRRC RiskMetrics (which covers the S&P 1500 firms and additional companies for some years and provides data on board characteristics and other corporate governance mechanisms). Our final sample, with the joint intersection of Thomson Ownership Database, ExecuComp, and IRRC RiskMetrics, has 2,249 firms with 14,055 firm-year observations for non-missing values of all key variables over the period of 1997 to 2006.

To examine how institutional owners influence firm decision-making and ultimately firm performance, we took a new empirical approach examining the influence of the interaction between institutional ownership and capital expenditures on firm performance. We estimated the following regression model to examine how institutional ownership affects the profitability of corporate investment.

• (1)

We considered three different measures of firm performance as the dependent variable (Perform_i,t+1 in Equation (1)): (a) Tobin's Q, computed as the market value of equity plus the book value of assets minus the book value of equity, scaled by book assets; (b) residual earnings to lagged total assets to measure the return to shareholders above the required return on capital (i.e., D'Mello & Shroff, 2000; Dong, Hirshleifer, Richardson, & Teoh, 2006 for similar measures);6 and (c) returns on assets (ROA) (see Cornett et al., 2007). Although we obtained similar results with different measures of firm performance, Tobin's Q will be our main measure reported as it is widely used in the literature as a measure of firm value (i.e., Chung & Jo, 1996; Shleifer & Vishny, 1986).

We examined the influence of institutional ownership (IO_it in Equation (1)) on firm performance. Two measures of IO_it were used in the analysis: (a) the equity stakes of institutional investors (i.e. the fraction of the shares institutional investors own); and (b) the indicator variable for the high level of institutional ownership (i.e. the indicator = 1 when institutional ownership of a given firm-year observation is greater than the median institutional ownership for a given year and 0 otherwise).

We used capital expenditures scaled by start-of-year book assets to measure corporate investment activities (Inv_it in Equation (1)). A similar measure of corporate investments has been widely used by authors such as Baker, Stein, & Wurgler (2003). The effect of capital expenditures on firm performance, measured by Tobin's Q, depends on firm characteristics such as corporate structure and governance (Chen, 2006). As indicated by McConnell & Muscarella (1985) and Woolridge & Snow (1990), there is a variation in the economic value embodied in capital expenditures. Capital expenditures do not necessarily create positive economic value. They may in fact have a negative economic impact if the investment is wasteful (Chen & Ho, 1997; Lang, Stulz, & Walkling, 1991).

Most prominently, we examined the interaction term between institutional ownership and corporate investments (IO_it * Inv_it in Equation (1)) to study the impact of institutional investors on firm performance through the channel of corporate investments (as stated in H1). While the variable (IO_it) can verify Sherman et al. (1998), p.173 question on whether institutional investors invest in firms based on past or future performance, our interest lies in the estimation of the coefficient on the interaction term (IO_it * Inv_it). To ensure that the interaction effect is due to higher institutional ownership, we examined the interaction effect between the high institutional ownership indicator and capital expenditures. This enabled us to examine whether capital expenditures enhance the firm value at a higher institutional ownership level.

We included other control variables to take fundamental firm characteristics (Control_it in Equation (1)) into account: (a) analyst coverage, defined as the number of analysts covering the firm, as the measure of firm's information asymmetry and external monitoring (Chang, Dasgupta, & Hilary, 2006); (b) cash flow, measured by the natural log of the firm's operating cash flow, as the measure of firms’ internal funds and profitability;7 (c) dividend-to-book ratio, defined as the ratio of dividends to book value of equity, a control variable that may predict firm profitability and performance; (d) leverage ratio, defined as long-term debts divided by the sum of long-term debts and book equities; (e) firm size, computed by taking the natural log of market capitalization, as the control for the firm size or the scale effect; and (f) firm age, defined as the natural log of firm age, as a measure of the corporate life-cycle effect on investment decisions. All but control variable (d) are similar control variables of fundamental firm characteristics used by finance literature in predicting firm performance (see e.g., Chung & Jo, 1996; Gompers & Metrick, 2001). We used (d) to control for the influences of financing decisions on firm performance.8 In addition, we included the year dummy variables and Fama-French 12-industry dummy variables (excluding finance and utilities).

The regressions for Equation (1) are estimated with clustered standard errors adjusted for intra-group correlation. The robust cluster variance estimator is robust to misspecification and within-cluster correlation.9

To demonstrate the institutional investors’ monitoring role on the corporate investment mechanism and test H2, we examined the differential effects of different categories of institutional investors on corporate investment performance.10 Following Ferreira & Matos (2008), we classified institutions according to the potential for business ties to a corporation as either an independent or grey institution. Independent institutional investors include hedge funds and research companies, whereas grey institutions consist of banks and insurance companies. As such, we examined the following four different categories of institutional investors (in the order of independency) identified in the Thomson Ownership Database:

1. Investment advisors (including mutual fund companies), which are considered better equipped and more active and independent institutional investors for providing effective monitoring functions than other institutional investors.11 As suggested by Almazan et al. (2005) and Chen et al. (2007), investment advisors are independent institutional investors and in a better position, in terms of tools and motivations, to exert positive impacts on corporate decisions.
2. Hedge funds can also be considered independent investors. On one hand, activist hedge funds are better informed and increasingly engage in monitoring (Brav, Jiang, Partnoy, & Thomas, 2008; Clifford, 2008). On the other hand, hedge funds are less regulated and likely to be less effective monitors due to their active trading.12 As such, whether hedge funds are effective monitors of corporate investment decisions is subject to our empirical investigation.
3. Pension funds are more divergent and may not have a significant impact on firm performance through influencing corporate investments. Existing studies by Del Guercio & Hawkins (1999) suggest that while some pension funds act as independent investors, others exhibit the features of grey investors.
4. Banks and insurance companies, often considered as grey investors, are less likely to provide independent monitoring or effective corporate governance.

Our choice of these institutional investor groups was also motivated by Bushee (2001), who categorized institutions based on two criteria respectively—investment horizon (transient, dedicated, and quasi-indexers) and fiduciary standards (bank trusts, insurance companies, investment advisors including mutual fund companies, and pensions and endowments). As noted in Bushee (2001), prior studies have found that the categorization based on fiduciary standards highlights significant differences across institutions in preferences for certain firm characteristics like size and growth potential (Abarbanell, Bushee & Raedy, 2003; Del Guercio, 1996; Lang & McNichols, 1997). Moreover, Bushee (2001) suggested that banks and pensions act as transient institutions while insurance companies and investment advisors share similarity with dedicated institutions and quasi-indexers. Further, our choice of these institutional investor groups is consistent with Almazan et al. (2005) who classified institutions into active monitors (investment advisors and investment companies) and passive monitors (bank trusts, insurance companies, and pension funds).

We considered different corporate governance metrics and mechanisms to test H3 and examined whether institutional ownership and other forms of corporate governance mechanisms operate as substitutes or complements in improving capital expenditure decisions. First, we looked at three different measures for managerial incentive, compensation, and control.

1. CEO Long-Term Incentive Plan. This variable can strengthen a firm's corporate governance and is collected from ExecuComp. Stein (1989) and Garvey, Grant, & King (1999), and Jensen (2005) argue that myopic CEOs with short-horizon objectives are more likely to make value-destroying investment decisions. As such, firms without CEO long-term incentive plan are expected to have a higher potential for agency problems.
2. Executive Severance Agreements. Executive severance agreements assure high-level executives of their positions or compensation and are not contingent upon a change in control (unlike Golden or Silver Parachutes). As used by Gompers, Ishii, & Metrick (2003), Executive severance agreements can weaken a firm's corporate governance. This variable is collected from IRRC RiskMetrics.
3. CEO serving as a director on the board. When a CEO serves as a board director, it creates the agency problem of evaluating and compensating CEO performance, which ultimately reduces the CEOs and the company's performance (Jensen, 1993). As such, this variable can weaken a firm's internal control and governance mechanism. This variable is collected from ExecuComp.

Next, we identified key corporate governance provisions used in the literature and governance mechanisms that are related to shareholder rights and intervention and are likely to impact a firm's key policies such as corporate investment decisions. Cai, Garner, & Walkling (2009) studied board of director election and suggested that cumulative voting, confidential voting, and staggered boards are important factors for corporate governance.

1. Cumulative Voting. According to Gompers et al. (2003), this governance provision allows shareholders to concentrate their votes and helps minority shareholders elect directors. This provision increases shareholder rights and has a positive effect on internal governance. Bhagat & Brickley (1984) found that the elimination of cumulative voting reduces shareholder wealth. However, Pozen (2003) reported that only a small fraction of firms (9.2%) allows cumulative voting. This variable is collected from IRRC RiskMetrics.
2. Confidential Voting. Also known as secret ballot, voting confidentially for either an independent third party or employees sworn to secrecy are used to count proxy votes, and management usually agrees not to look at individual proxy cards (Gompers et al., 2003). Confidential voting can help eliminate potential conflicts of interest for fiduciaries voting shares on behalf of others, and increase shareholder rights. Pound (1988) argues that confidential voting enables shareholders to vote against management when doing so would maximize share value. Cai et al. (2009) found that directors at firms with confidential voting receive fewer votes, suggesting shareholders are more willing to vote against directors when their identities are protected. This variable is collected from IRRC RiskMetrics.
3. Staggered Board. According to Gompers et al. (2003), a staggered board (or classified board) is one in which the directors are placed into different classes and serve overlapping terms. A staggered board can delay outside shareholders’ gain of control and intervention, and hence insulate entrenched managers from market discipline. Bebchuk & Cohen (2005) argue that staggered boards are the most powerful takeover protection and are associated with lower firm valuation. Gompers et al. (2003), Faleye (2007), and others found that staggered boards weaken corporate governance and have a negative impact on firm value.13 This variable is collected from IRRC RiskMetrics.
4. Number of Board Meetings occurred each year. According to Jensen (1993), active and attentive boards are those who monitor CEO performance. The frequency of board meetings could indicate the board of directors’ intervention in a CEO's operational decision making. This variable is collected from ExecuComp.

In addition to managerial incentives and board characteristics, Table 1 provides the descriptive statistics of the key variables. As shown in Table 1, total institutional investors have an average of 43% of total equity ownership of the firms in our sample. Among different types of institutional investors, investment advisors have largest equity stake (average of 30% of total equity ownership of the firm), followed by hedge funds (4%), pension funds (3%), and bank and insurance companies (2%). For fundamental firm characteristics, Table 1 shows that the average ratio of capital expenditures to total assets is 0.06. For firm performance measures, the average Tobin's Q is 2.15. As for financing decisions and dividends, the average leverage ratio is 0.27, and the average dividend-to-book ratio is 0.02. Also, the average log(market capitalization) is 7.08 and the average of log(firm age) is 2.52. The standard deviations of these variables suggest that our sample covers a wide selection of firms with different characteristics. For CEO, shareholder provisions, and board of director characteristics, we found that a low percentage of observations with internal governance mechanisms such as CEO long-term incentive plan, cumulative voting, and confidential voting. The firms have a median of six board meetings per year.

Table 1. Descriptive Statistics

This table reports the descriptive statistics of institutional ownership, firm fundamental characteristics, firm performance, and CEO and board of director characteristics. Firm fundamental characteristics include: capital expenditures (scaled by start-of-year book assets), analyst coverage (defined as the number of analysts covering the firm), cash flow (measured by the natural log of the firm's cash and cash equivalents), dividends ratio (defined as dividends-to-book), leverage ratio (calculated as long-term debts divided by the sum of long-term debts and book equities), firm size (defined as the natural log of market capitalization), and firm age (defined as the natural log of firm age). All variables are winsorized at the 1^st and 99^th percentile to reduce the influence of outliers. Firm performance measures include stock performance (measured by Tobin's Q computed as the market value of equity plus the book value of assets minus the book value of equity, scaled by book assets), and operating performance (measured by residual earnings to lagged assets, and returns on assets (ROA)). CEO and board of director characteristics include: CEO equity shares (percentage of total shares of the firm), and the number of board of director meetings. The statistics for variables are obtained from a sample of 2,249 US firms publicly traded in the NYSE, AMEX and NASDAQ during the years 1997 to 2006 (excluding financial and utility firms) and 14,055 firm-year observations.

Variables	Mean	Median	S.D.	Min.	Max.
A. Institutional ownerships
Total institutional ownerships	0.43	0.51	0.34	0.00	0.93
Investment advisor	0.30	0.34	0.24	0.00	0.73
Bank and insurance companies	0.02	0.01	0.03	0.00	0.48
Hedge fund	0.04	0.04	0.02	0.00	0.13
Pension fund	0.03	0.01	0.04	0.00	0.22
B. Firm fundamental characteristics
Capital expenditures	0.06	0.05	0.06	0.00	0.32
Number of analysts	8.90	7.00	7.52	0.00	46.00
log(cash flows)	4.51	4.46	1.72	−5.12	10.78
Dividends-to-book	0.02	0.00	0.06	0.00	2.00
Leverage ratio	0.27	0.25	0.23	0.00	0.98
log(market capitalization)	7.08	6.91	1.60	−2.95	13.05
log(firm age)	2.52	2.64	0.91	0.00	3.64
C. Firm performance
Tobin's Q	2.15	1.64	1.49	0.53	8.69
Residual earnings to lagged assets	0.05	0.05	0.08	−0.50	0.24
ROA	0.01	0.02	0.10	−0.76	0.20
D. Internal corporate governance
CEO's long-term incentive plan	0.13	0.00	0.34	0.00	1.00
Executive severance agreements	0.08	0.00	0.28	0.00	1.00
CEO as director	0.76	1.00	0.43	0.00	1.00
Cumulative voting	0.09	0.00	0.29	0.00	1.00
Confidential voting	0.11	0.00	0.32	0.00	1.00
Staggered board	0.58	1.00	0.49	0.00	1.00
No. of board meetings	7.10	6.00	3.06	1.00	39.00

Results

Impacts of Institutional Ownerships on Investment-Induced Performance

Table 2 reports the impacts of institutional ownership on a firm's future performance, measured by one-year-ahead Tobin's Q, residual earnings on total assets, and returns on assets. The impact of institutional investors on firm value can be decomposed into two channels. First, the coefficient of institutional ownership (IO_it in Equation (1)) reflects the general relation between institutional ownership and future firm performance. To help with the interpretation, we first estimated the regressions without the interaction term between institutional ownership and capital expenditures to see how adding the interaction term changes the role of the different factors. Models (1) to (3) of Table 2 show that institutional ownership alone has a significant and positive impact on future firm performance. This result is consistent with Gompers & Metrick (2001).

Table 2. The Impact of Institutional Ownerships and Capital Expenditures on Firm Performance

This table presents the impact of institutional ownerships, capital expenditures, and their interactions on future firm performance. Capital expenditures are expressed in percentage terms normalized by one-year lagged total assets. Firm performance measures include both stock performance measured by one-year ahead Tobin's Q (reported in Model (1) and (4)), and operating performances measured by one-year ahead residual earnings to lagged assets (reported in Model (2) and (5)) and one-year ahead ROA (reported in Model (3) and (6)). Firm fundamental characteristics that are used to predict firm performance include: analyst coverage (defined as the number of analysts covering the firm), cash flow (measured by the natural log of the firm's cash and cash equivalents), dividends ratio (defined as dividends-to-book), leverage ratio (calculated as long-term debts divided by the sum of long-term debts and book equities), firm size (defined as the natural log of market capitalization), and firm age (defined as the natural log of firm age). The results are estimated with the Huber/White/Sandwich estimation of robust variance with clustered standard errors adjusted for intra-group correlation (standard error in parenthesis). * indicates significance at the 10 percent level, ** indicates significance at the 5 percent level, and *** indicates significance at the 1 percent level.

	Model (1)	Model (2)	Model (3)	Model (4)	Model (5)	Model (6)
Explanatory variables:	Dependent variables: Tobin's Q	Dependent variables: Residual earnings to total assets	Dependent variables: ROA	Dependent variables: Tobin's Q	Dependent variables: Residual earnings to total assets	Dependent variables: ROA
Capital expenditures	1.885***	0.049**	0.053**	0.939**	0.005	−0.002
	(0.330)	(0.019)	(0.022)	(0.387)	(0.030)	(0.034)
Total institutional ownerships	0.127** (0.053)	0.022*** (0.003)	0.023*** (0.004)	−0.031 (0.071)	0.015*** (0.005)	0.014*** (0.005)
Capital expenditures * Total institutional ownerships				2.493*** (0.760)	0.112** (0.052)	0.137** (0.060)
Number of analysts	0.003	−0.001***	−0.002***	0.003	−0.001***	−0.002***
	(0.004)	0.000	0.000	(0.004)	0.000	0.000
log(cash flows)	−0.276***	0.006***	0.010***	−0.277***	0.006***	0.010***
	(0.020)	(0.001)	(0.001)	(0.020)	(0.001)	(0.001)
Dividends-to-book	2.665***	0.142***	0.182***	2.670***	0.143***	0.182***
	(0.383)	(0.018)	(0.021)	(0.383)	(0.018)	(0.021)
Leverage ratio	−1.526***	−0.084***	−0.033***	−1.517***	−0.084***	−0.033***
	(0.083)	(0.005)	(0.005)	(0.083)	(0.005)	(0.005)
log(market capitalization)	0.475***	0.007***	0.007***	0.476***	0.007***	0.007***
	(0.026)	(0.001)	(0.002)	(0.026)	(0.001)	(0.002)
log(firm age)	−0.218***	−0.002*	−0.001	−0.218***	−0.002*	−0.001
	−0.023	−0.001	−0.001	−0.023	−0.001	−0.001
Constant	0.290*	−0.019*	−0.062***	0.357**	−0.016	−0.058***
	(0.148)	(0.010)	(0.012)	(0.149)	(0.011)	(0.013)
Fama-French 12 industry dummies	Yes	Yes	Yes	Yes	Yes	Yes
Year dummies	Yes	Yes	Yes	Yes	Yes	Yes
Adjusted R^2	0.34	0.13	0.13	0.34	0.13	0.13
N	12685	11955	11955	12685	11955	11955

Second, the positive impact of the interaction term of institutional ownership and capital expenditures (IO_it * Inv_it in Equation (1)) on firm value provides evidence that institutional investors provide monitoring and corporate governance. Note that the partial derivative of Perform_it+1 with respect to Inv_it for equation (1) provides additional insights into the decomposition of effects of capital expenditures on firm performance—that is, ∂ Perform_it+1/ ∂ Inv_it = γ + λ * IO_it where IO_it = 1 if high institutional ownership, γ captures the “baseline” effect of capital expenditures on firm performance, and λ captures the “incremental” effect on firm performance due to institutional monitoring. Hypotheses 1 predicts that λ should be positive and significant, as the monitoring effect of institutional ownership increases the productivity of capital expenditures. Models (4) to (6) of Table 2 show that the interaction effect of total institutional ownership and capital expenditures is positive and significant with respect to one-year-ahead Tobin's Q with the “incremental” effect (λ = 2.493) being positive and significant, and the “overall” effect (γ + λ = 0.939 + 2.493) being positive. Overall, these findings support H1 that institutional investors can add value to the firm through their interaction with capital expenditure decisions.

Different Types of Institutional Investors

Table 3 shows that institutional investors are heterogeneous monitors in improving investment performance. First, Model (1) shows that the interaction effect between investment advisors’ equity ownership and capital expenditures is significant and positive on one-year-ahead Tobin's Q. Model (2) shows that the interaction effect between hedge funds’ equity ownership and capital expenditures is also significantly positive. In contrast, the interaction between capital expenditures and equity ownerships by other categories of institutional investors does not have any significant impact.

Table 3. The Impact of Institutional Ownerships and Capital Expenditures on Firm Performance Separated by Different Types of Institutional Investors

This table presents the impact of institutional ownerships, capital expenditures, and their interactions on future firm performance. The impact is separately examined for different types of institutional investors. Institutional ownerships include equity ownerships by: investment advisors, bank and insurance companies, hedge funds, and pension funds. Capital expenditures are expressed in percentage terms normalized by one-year lagged total assets. Firm performance is measured by one-year ahead Tobin's Q. Firm fundamental characteristics that are used to predict firm performance include: analyst coverage (defined as the number of analysts covering the firm), cash flow (measured by the natural log of the firm's cash and cash equivalents), dividends ratio (defined as dividends-to-book), leverage ratio (calculated as long-term debts divided by the sum of long-term debts and book equities), firm size (defined as the natural log of market capitalization), and firm age (defined as the natural log of firm age). The results are estimated with the Huber/White/Sandwich estimation of robust variance with clustered standard errors adjusted for intra-group correlation (standard error in parenthesis). * indicates significance at the 10 percent level, ** indicates significance at the 5 percent level, and *** indicates significance at the 1 percent level.

	Model (1)	Model (2)	Model (3)	Model (4)
Explanatory variables:	dependent variables: Tobin's Q	dependent variables: Tobin's Q	dependent variables: Tobin's Q	dependent variables: Tobin's Q
Capital expenditures	1.219***	2.897***	1.769***	2.741***
	(0.385)	(0.568)	(0.495)	(0.616)
Investment advisor	−0.074
	(0.110)
Investment advisor * Capital expenditures	3.828*** (1.178)
Banks and insurance		0.167
		(1.013)
Banks and insurance * Capital expenditures		−12.181 (10.738)
Hedge fund			−1.246**
			(0.592)
Hedge fund * Capital expenditures			50.125*** (17.111)
Pension				−6.006***
				(1.207)
Pension * Capital expenditures				−11.155 (14.711)
Number of analysts	0.003	−0.001	−0.001	0.001
	(0.004)	(0.005)	(0.005)	(0.005)
log(cash flows)	−0.325***	−0.335***	−0.335***	−0.318***
	(0.021)	(0.024)	(0.024)	(0.024)
Dividends-to-book	2.135***	1.977***	2.006***	2.012***
	(0.347)	(0.385)	(0.385)	(0.381)
Leverage ratio	−1.469***	−1.483***	−1.489***	−1.503***
	(0.084)	(0.099)	(0.099)	(0.098)
log(market capitalization)	0.501***	0.541***	0.540***	0.547***
	(0.026)	(0.032)	(0.031)	(0.031)
log(firm age)	−0.24	−0.07	−0.03	0.08
	(0.174)	(0.144)	(0.142)	(0.141)
Constant	1.219***	2.897***	1.769***	2.741***
	(0.385)	(0.568)	(0.495)	(0.616)
Fama-French 12 industry dummies	Yes	Yes	Yes	Yes
Year dummies	Yes	Yes	Yes	Yes
Adjusted R^2	0.32	0.33	0.33	0.34
N	12689	9717	9717	9717

Table 4 presents the results using the indicator variable of the high level of institutional ownership as an alternate proxy of institutional monitoring. As previously discussed, we used the interaction effect between the high institutional ownership indicator and capital expenditures to ensure that the interaction effect is due to higher institutional ownership. The result in Table 4 shows that the interaction term for high level of investment advisors’ equity ownership and capital expenditures has a significant and positive impact on one-year-ahead Tobin's Q, with the incremental effect (λ = 1.262) being positive and significant and the overall effect (γ + λ = 1.262 + 1.388) being positive. In contrast, the interaction effect between high level of hedge fund equity ownership and capital expenditures becomes insignificant.

Table 4. The Impact of High Institutional Ownerships and Capital Expenditures on Firm Performance Separated by Different Types of Institutional Investors

This table presents the impact of the high institutional ownership indicator, capital expenditures, and their interactions on future firm performance. The impact is separately examined for different types of institutional investors. The high institutional ownership indicator (the indicator = 1 when institutional ownership of a given firm-year observation is greater than the median institutional ownership for a given year; 0 otherwise) is assessed for: all institutional investors, investment advisors, bank and insurance companies, hedge funds, and pension funds. Capital expenditures are expressed in percentage terms normalized by one-year lagged total assets. Firm performance is measured by one-year ahead Tobin's Q. Firm fundamental characteristics that are used to predict firm performance include: analyst coverage (defined as the number of analysts covering the firm), cash flow (measured by the natural log of the firm's cash and cash equivalents), dividends ratio (defined as dividends-to-book), leverage ratio (calculated as long-term debts divided by the sum of long-term debts and book equities), firm size (defined as the natural log of market capitalization), and firm age (defined as the natural log of firm age). The results are estimated with the Huber/White/Sandwich estimation of robust variance with clustered standard errors adjusted for intra-group correlation (standard error in parenthesis). * indicates significance at the 10 percent level, ** indicates significance at the 5 percent level, and *** indicates significance at the 1 percent level.

	Model (1)	Model (2)	Model (3)	Model (4)	Model (5)
Explanatory variables:	Dependent variables: Tobin's Q	Dependent variables: Tobin's Q	Dependent variables: Tobin's Q	Dependent variables: Tobin's Q	Dependent variables: Tobin's Q
Capital expenditures	1.375***	1.262***	2.010***	1.737***	2.063***
	(0.360)	(0.338)	(0.460)	(0.526)	(0.463)
High institutional ownerships indicator	−0.009 (0.046)
High institutional ownerships indicator * Capital expenditures	1.206** (0.523)
High investment advisor indicator		−0.002
		(0.047)
High investment advisor indicator * Capital expenditures		1.388** (0.541)
High banks and insurance			0.049
			(0.052)
High banks and insurance indicator * Capital expenditures			0.211 (0.557)
Hedge fund				−0.034
				(0.045)
High hedge fund indicator * Capital expenditures				0.662 (0.566)
Pension					−0.144***
					(0.051)
High pension indicator * Capital expenditures					0.077 (0.636)
Number of analysts	0.003	0.003	−0.001	−0.001	0.000
	(0.004)	(0.004)	(0.005)	(0.005)	(0.005)
log(cash flows)	−0.277***	−0.277***	−0.288***	−0.290***	−0.282***
	(0.020)	(0.020)	(0.024)	(0.024)	(0.024)
Dividends-to-book	2.677***	2.679***	2.519***	2.522***	2.515***
	(0.383)	(0.384)	(0.424)	(0.426)	(0.423)
Leverage ratio	−1.523***	−1.521***	−1.529***	−1.530***	−1.542***
	(0.083)	(0.083)	(0.097)	(0.097)	(0.097)
log(market capitalization)	0.478***	0.478***	0.514***	0.522***	0.529***
	(0.026)	(0.026)	(0.032)	(0.031)	(0.031)
log(firm age)	−0.219***	−0.219***	−0.230***	−0.227***	−0.212***
	(0.023)	(0.023)	(0.026)	(0.026)	(0.026)
Constant	0.472***	0.089	0.505***	0.496***	0.304
	(0.139)	(0.143)	(0.151)	(0.151)	(0.195)
Fama-French 12 industry dummies	Yes	Yes	Yes	Yes	Yes
Year dummies	Yes	Yes	Yes	Yes	Yes
Adjusted R^2	0.34	0.34	0.35	0.35	0.35
N	12682	12680	9709	9709	9710

Together, Tables 3 and 4 support H2. These findings further help to disentangle the monitoring hypothesis from others by showing that not all categories of institutional investors can in fact improve investment performance. Although higher firm performance may attract greater institutional ownership in general, only certain types of institutional investors (e.g., independent institutional investors such as investment advisors) can improve firm performance through monitoring.

Interaction between Institutional Ownership and Internal Governance

In this section (Tables 5 and 6) we continue our investigation on how the monitoring role of institutional investors changes with firm's internal governance mechanisms. For a clear and coherent comparison of the differential effects of institutional ownership across different governance conditions, we first present the result for weak governance subsample and then the result for strong governance subsample—both of which are results for each stratified group of internal governance proxy . We also examined the same effect for investment advisors in Panel B of Tables 5 and 6 (as our results in Tables 3 and 4 have established that investment advisors are likely to be most effective monitors).

Table 5. The Impact of Institutional Ownerships and Capital Expenditures on Firm Performance Separated by Managerial Incentives and Control

This table reports the interaction effect of institutional ownership and capital expenditures on firm performance with different groups of managerial incentives and control, for overall institutional ownership (Panel A) and investment advisors (Panel B). The regressions are estimated for firms grouped by variables of CEO incentives, compensation and control. The median of these characteristics is used to partition the sample into high- and low sub-samples. CEO characteristics include: (i) CEO's long-term incentive plan; (ii) executive severance agreements (which assure high-level executives of their positions or some compensation and are not contingent upon a change in control); and (iii) CEO as board of director. Capital expenditures are expressed in percentage terms normalized by one-year lagged total assets. Firm performance is measured by one-year ahead Tobin's Q. Firm fundamental characteristics are identical to those in other tables but their coefficients are not reported here. The results are estimated with the Huber/White/Sandwich estimation of robust variance with clustered standard errors adjusted for intra-group correlation (standard error in parenthesis). * indicates significance at the 10 percent level, ** indicates significance at the 5 percent level, and *** indicates significance at the 1 percent level.

Panel A. Total Institutional Ownership
	Model (1)	Model (2)	Model (3)	Model (4)	Model (5)	Model (6)
	CEO's long-term incentive plan		Executive severance agreements		CEO as director
	No (weak governance)	Yes (strong governance)	Yes (weak governance)	No (strong governance)	Yes (weak governance)	No (strong governance)
Explanatory variables: Capital expenditures (Invit)	1.687*** (0.493)	1.866* (1.105)	1.313 (1.778)	1.028** (0.440)	1.785*** (0.471)	0.786 (0.528)
High institutional ownership indicator (IOit)	−0.031 (0.058)	0.053 (0.081)	−0.328** (0.153)	0.042 (0.051)	−0.019 (0.054)	0.006 (0.081)
High institutional ownership indicator * Capital expenditures	1.339** (0.669)	−0.223 (1.123)	5.252** (2.409)	0.693 (0.573)	1.207* (0.627)	0.548 (0.855)
Adjusted R^2	0.36	0.49	0.39	0.37	0.37	0.27
No. of Obs.	8453	1292	838	8671	9670	3012
Panel B. Equity Ownershipby Investment Advisors
	Model (1)	Model (2)	Model (3)	Model (4)	Model (5)	Model (6)
	CEO's long-term incentive plan		Executive severance agreements		CEO as Director
	No (weak governance)	Yes (strong governance)	Yes (weak governance)	No (strong governance)	Yes (weak governance)	No (strong governance)
Explanatory variables: Capital expenditures (Invit)	1.389*** (0.454)	2.374** (1.077)	1.068 (1.774)	0.857** (0.403)	1.548*** (0.434)	0.843* (0.510)
High investment advisor indicator (IOit)	−0.018 (0.060)	0.136* (0.079)	−0.334** (0.157)	0.045 (0.053)	−0.001 (0.055)	−0.015 (0.078)
High investment advisor indicator * Capital expenditures	1.837*** (0.686)	−1.001 (1.077)	5.642** (2.418)	0.975* (0.580)	1.587** (0.640)	0.212 (0.767)
Adjusted R^2	0.36	0.49	0.39	0.37	0.37	0.27
No. of Obs.	8452	1291	838	8668	9668	3012

Table 6. The Impact of Institutional Ownerships and Capital Expenditures on Firm Performance Separated by Corporate Governance Provisions and Mechanisms

This table reports the interaction effect of institutional ownership and capital expenditures on firm performance with different groups of internal corporate governance provisions and mechanisms, for overall institutional ownership (Panel A) and investment advisors (Panel B). The governance provisions and mechanisms include: (i) cumulative voting (a provision that can increase shareholder rights by allowing shareholders to allocate their total votes in any manner desired and helps minority shareholders to elect directors); (ii) confidential voting (a provision that helps eliminate potential conflicts of interest for fiduciaries voting shares on behalf of others, and reduces pressure by management on shareholder-employees or shareholder-partners); (iii) staggered board (the mechanism in which the directors are placed into different classes and serve overlapping terms); and (iv) number of board meetings (the measure which indicates the board of directors’ intervention in a CEO's operational decision making). Capital expenditures are expressed in percentage terms normalized by one-year lagged total assets. Firm performance is measured by one-year ahead Tobin's Q. Firm fundamental characteristics are identical to those in other tables but their coefficients are not reported here. The results are estimated with the Huber/White/Sandwich estimation of robust variance with clustered standard errors adjusted for intra-group correlation (standard error in parenthesis). * indicates significance at the 10 percent level, ** indicates significance at the 5 percent level, and *** indicates significance at the 1 percent level.

Panel A. Total Institutional Ownership
	Model (1)	Model (2)	Model (3)	Model (4)	Model (5)	Model(6)	Model (7)	Model(8)
	Cumulative voting		Confidential voting		Staggered board		Number of board meetings
	No (weak governance)	Yes (strong governance)	No (weak governance)	Yes (strong governance)	Yes (weak governance)	No (strong governance)	Low (weak governance)	High (strong governance)
Explanatory variables: Capital expenditures (Invit)	1.238*** (0.459)	−1.439 (1.052)	0.822* (0.459)	1.057 (1.082)	1.543** (0.604)	0.219 (0.629)	1.329* (0.702)	2.206*** (0.654)
High institutional ownership indicator (IOit)	0.008 (0.052)	0.048 (0.150)	0.010 (0.052)	−0.101 (0.116)	−0.067 (0.061)	0.126 (0.079)	−0.073 (0.086)	0.043 (0.068)
High institutional ownership indicator * Capital expenditures	1.043* (0.607)	1.774 (1.790)	0.960 (0.621)	1.535 (1.457)	1.370* (0.779)	0.588 (0.884)	2.470** (1.054)	0.075 (0.812)
Adjusted R^2	0.37	0.35	0.37	0.45	0.38	0.35	0.38	0.36
No. of Obs.	8595	914	8458	1051	5531	3978	3581	4393
Panel B. Equity Ownership by Investment Advisors
	Model (1)	Model (2)	Model (3)	Model (4)	Model (5)	Model(6)	Model (7)	Model(8)
	Cumulative voting		Confidential voting		Staggered board		Number of board meetings
	No (weak governance)	Yes (strong governance)	No (weak governance)	Yes (strong governance)	Yes (weak governance)	No (strong governance)	Low (weak governance)	High (strong governance)
Explanatory variables: Capital expenditures	1.050** (0.425)	−1.411 (1.125)	0.598 (0.429)	1.305 (1.189)	1.299** (0.574)	0.139 (0.582)	1.269* (0.693)	1.887*** (0.591)
High investment advisor indicator (IOit)	0.008 (0.053)	0.092 (0.173)	0.007 (0.053)	−0.064 (0.117)	−0.093 (0.063)	0.167** (0.082)	−0.038 (0.086)	0.068 (0.070)
High investment advisor indicator *Capital expenditures	1.364** (0.610)	1.613 (1.890)	1.357** (0.618)	1.146 (1.418)	1.793** (0.785)	0.692 (0.902)	2.514** (1.029)	0.609 (0.811)
Adjusted R^2	0.37	0.36	0.37	0.45	0.38	0.35	0.38	0.36
No. of Obs.	8592	914	8455	1051	5529	3977	3580	4392

Managerial incentives and control

Table 5 reports the interaction effect of institutional ownership and capital expenditure decisions on firm performance with different groups of firm's internal governance measures that proxy for managerial incentives, compensation, and control. In Panel A of Table 5, Models (1) and (2) show that the interaction effect between high institutional ownership and capital investments has a significant and positive impact on one-year-ahead Tobin's Q for firms without CEO long-term incentive plan (a proxy for managerial myopia and weak internal governance). In contrast, the interaction effect is insignificant for firms with CEO long-term incentive plan. Panel B provides similar results for investment advisor. These findings suggest that when firms have no long-term managerial incentives and managers are likely to make value-destroying investment decisions (Garvey et al., 1999; Jensen, 2005; Stein, 1989), institutional investors, with large equity controlling stakes and greater focus on the long-term value of the firm, become important to provide monitor and enhance firm performance.

Equally important, Models (3) and (4) of Table 5 show that the interaction effect between high institutional ownership and capital investments is significant and positive for firms with executive severance agreements (which weakens a firm's internal governance). Panel B of Table 5 provides similar results for investment advisors. Consistent with findings in Models (1) and (2), these results suggest that institutional monitoring is more important for firms with weak internal governance. Models (5) and (6) of Table 5 explore whether CEOs’ serving as board directors affects the interaction effect between high institutional ownership and capital investments. CEOs serving as board directors reflect the CEO's ability to influence the board of directors (Bebchuk & Fried, 2003), which then weakens internal governance. Our result in Panel B of Table 5 suggests that the positive interaction effect between high equity ownership of investment advisor and capital expenditures becomes greater and more significant when CEOs weaken internal governance by serving as board directors.

Corporate governance provisions and mechanisms

Table 6 shows that the interaction effect between institutional ownership and capital expenditures varies with respect to internal governance provisions, which signifies shareholders’ power over management. These governance provisions include cumulative voting and confidential voting. In Panel A of Table 6, Models (1) and (2) show that the interaction term for overall institutional ownership and capital expenditures has a significant and positive impact on one-year-ahead Tobin's Q for firms without cumulative voting. In contrast, such interaction effect is not significant for firms with cumulative voting. Furthermore, Models (3) and (4) in Panel B show that the interaction effect of high equity ownership of investment advisor and capital expenditures on one-year-ahead Tobin's Q is significant and positive for firms without confidential voting. Models (5) to (8) of Table 6 focus on how the interaction effect between institutional ownership and capital expenditures varies with board effectiveness and monitoring. Models (5) and (6) show that the interaction effect for high equity ownership of investment advisor and capital expenditures is significant and positive for firms with staggered boards (which weakens internal governance).

Lastly, Models (7) and (8) of Table 6 present the interaction effect between institutional ownership and capital investments for firms with a low (vis-à-vis high) frequency of board meetings. As previously discussed, the frequency of board meetings may proxy for the boards’ intervention in CEOs’ operational decision making. Our results in Panels A and B reveal that the interaction effect between high equity ownership of investment advisor (or all institutional investors) and capital investments on firm performance is positive and significant only for firms with less frequent board meetings (less than the median of six meetings per year).14 Overall, our findings in Tables 5 and 6 provide strong evidence in favour of H3. Institutional investors, particularly investment advisors, becomes more prominent in enhancing corporate investment performance when: (a) CEOs’ incentives are not well-aligned with firm value; (b) internal governance provisions that promote shareholder rights are limited; and (c) internal monitoring by board of directors becomes ineffective. These findings provide important empirical evidence to support the theory by Jensen (1993) that active investors who play an active role in monitoring and strategic direction of the firm are important to a well-functioning governance system.

Robustness

Monitoring Effect of Institutional Investors

We conducted a robustness test to address the potential endogeneity problem in the relation between institutional ownership and firm performance. Table 7 shows that the interaction effect between total institutional ownership (especially equity ownership of investment advisors) and capital expenditures is significant and positive for firms with prior lower performance (negative change in Tobin's Q). This finding suggests that institutional investors do not buy into companies where the performance had already improved, and that the relationship between institutional ownership and capital expenditure performance is not the result of institutional investors’ response to past good performance. Indeed, our finding reveals that institutional investors are likely to observe opportunities in underperforming firms with weak governance. They make a difference by performing an appropriate monitoring role and improve firm's subsequent performance. Our results are consistent with Chung, Firth, & Kim (2002) and Hartzell & Starks (2003) that outside shareholders are becoming increasingly active in corporate governance, especially in underperforming firms.

Table 7. The Impact of Institutional Ownerships and Capital Expenditures on Firm Performance Separated by Lagged Change in Tobin's Q

This table presents the impact of the high institutional ownership indicator, capital expenditures, and their interactions on future firm performance with different groups of lagged changes in Tobin's Q. The impact is separately examined for overall institutional ownership and investment advisors. Capital expenditures are expressed in percentage terms normalized by one-year lagged total assets. Firm performance is measured by one-year ahead Tobin's Q. Firm fundamental characteristics that are used to predict firm performance include: analyst coverage (defined as the number of analysts covering the firm), cash flow (measured by the natural log of the firm's cash and cash equivalents), dividends ratio (defined as dividends-to-book), leverage ratio (calculated as long-term debts divided by the sum of long-term debts and book equities), firm size (defined as the natural log of market capitalization), and firm age (defined as the natural log of firm age). The results are estimated with the Huber/White/Sandwich estimation of robust variance with clustered standard errors adjusted for intra-group correlation (standard error in parenthesis). * indicates significance at the 10 percent level, ** indicates significance at the 5 percent level, and *** indicates significance at the 1 percent level.

	Model (1)	Model (2)	Model (3)	Model (4)
	Dependent variables: Tobin's Q		Dependent variables: Tobin's Q
	Lagged change in Q < 0	Lagged change in Q > 0	Lagged change in Q < 0	Lagged change in Q > 0
Capital expenditures	1.625***	1.874***	1.410***	1.919***
	(0.409)	(0.626)	(0.378)	(0.585)
High institutional ownerships indicator	−0.048	−0.008
	(0.050)	(0.063)
High institutional ownerships indicator * Capital expenditures	1.046* (0.562)	1.299 (0.819)
High investment advisor indicator			−0.062	0.031
			(0.052)	(0.064)
High investment advisor indicator * Capital expenditures			1.464** (0.593)	1.168 (0.836)
Number of analysts	0.003	0.014**	0.003	0.013**
	(0.005)	(0.006)	(0.005)	(0.006)
log(cash flows)	−0.220***	−0.284***	−0.219***	−0.285***
	(0.022)	(0.030)	(0.022)	(0.030)
Dividends-to-book	2.256***	2.596***	2.251***	2.611***
	(0.405)	(0.447)	(0.405)	(0.450)
Leverage ratio	−1.349***	−1.643***	−1.345***	−1.643***
	(0.093)	(0.105)	(0.092)	(0.105)
log(market capitalization)	0.407***	0.461***	0.406***	0.462***
	(0.029)	(0.036)	(0.029)	(0.036)
log(firm age)	−0.158***	−0.290***	−0.159***	−0.289***
	(0.025)	(0.033)	(0.025)	(0.033)
Constant	0.388**	1.442***	0.159	1.885***
	(0.159)	(0.221)	(0.175)	(0.234)
Fama-French 12 industry dummies	Yes	Yes	Yes	Yes
Year dummies	Yes	Yes	Yes	Yes
Adjusted R^2	0.32	0.37	0.32	0.38
N	5335	5365	5334	5365

Alternative Sample

As a robustness test, we examined the result of Equation (1) with a larger sample not subject to the data requirements from ExecuComp and IRRC RiskMetrics. This larger sample covers the years 1990 to 2006 and comprises 6,165 firms, giving a total of 27,751 firm-year observations with no missing values with regard to the key control variables. Table 8 shows that the interaction term between total institutional ownership (equity ownership of investment advisors) and capital expenditures remains significant and positive on future firm performance. The consistent results from the larger sample robustly confirm the monitoring role of institutional investors.

Table 8. The Impact of Institutional Ownerships and Capital Expenditures with Larger Sample

This table presents the impact of institutional ownerships, capital expenditures, and their interactions on future firm performance with the larger sample, which comprises 6,165 firms and a total of 27,751 firm-year observations. Capital expenditures are expressed in percentage terms normalized by one-year lagged total assets. Firm performance measures include both stock performance measured by one-year ahead Tobin's Q (reported in Model (1) and (4)), and operating performances measured by one-year ahead residual earnings to lagged assets (reported in Model (2) and (5)) and one-year ahead ROA (reported in Model (3) and (6)). Firm fundamental characteristics that are used to predict firm performance include: analyst coverage, cash flow, dividends/payout ratio, leverage, firm size, and firm age. The results are estimated with the Huber/White/Sandwich estimation of robust variance with clustered standard errors adjusted for intra-group correlation (standard error in parenthesis). * indicates significance at the 10 percent level, ** indicates significance at the 5 percent level, and *** indicates significance at the 1 percent level.

	Model (1)	Model (2)	Model (3)	Model (4)	Model (5)	Model (6)
Explanatory variables:	Dependent variables: Tobin's Q	Dependent variables: Residual earnings to total assets	Dependent variables: ROA	Dependent variables: Tobin's Q	Dependent variables: Residual earnings to total assets	Dependent variables: ROA
Capital expenditures	0.370***	−0.022	−0.021	0.368***	−0.023	−0.023
	(0.135)	(0.014)	(0.015)	(0.137)	(0.014)	(0.015)
Total institutional ownerships	−0.01 (0.051)	0.017*** (0.004)	0.014*** (0.004)
Total institutional ownerships * Capital expenditures	3.064*** (0.589)	0.191*** (0.040)	0.202*** (0.044)
Investment advisor				0.002	0.030***	0.027***
				(0.076)	(0.005)	(0.006)
Investment advisor * Capital expenditures				4.388*** (0.878)	0.275*** (0.057)	0.299*** (0.062)
Number of analysts	0.014***	−0.001***	−0.001***	0.013***	−0.001***	−0.002***
	(0.003)	0.000	0.000	(0.003)	0.000	0.000
log(cash flows)	−0.195***	0.008***	0.010***	−0.195***	0.008***	0.010***
	(0.010)	(0.001)	(0.001)	(0.010)	(0.001)	(0.001)
Dividends-to-earnings	2.045***	0.150***	0.192***	2.057***	0.154***	0.195***
	(0.293)	(0.017)	(0.020)	(0.293)	(0.017)	(0.020)
Leverage ratio	−1.027***	−0.063***	−0.012***	−1.025***	−0.063***	−0.012***
	(0.047)	(0.003)	(0.004)	(0.047)	(0.003)	(0.004)
log(market capitalization)	0.344*** (0.013)	0.003*** (0.001)	0.001 (0.001)	0.343*** (0.013)	0.002*** (0.001)	0 (0.001)
log(firm age)	−0.113***	0.003***	0.004***	−0.114***	0.002***	0.004***
	(0.012)	(0.001)	(0.001)	(0.012)	(0.001)	(0.001)
Constant	0.715***	−0.018***	−0.054***	0.716***	−0.018***	−0.054***
	(0.077)	(0.006)	(0.007)	(0.077)	(0.006)	(0.007)
Fama-French 12 industry dummies	Yes	Yes	Yes	Yes	Yes	Yes
Year dummies	Yes	Yes	Yes	Yes	Yes	Yes
Adjusted R^2	0.29	0.1	0.11	0.29	0.1	0.11
N	27356	26678	26678	27356	26678	26678

Discussion

Summary

This study finds evidence that institutional ownership improves firm performances through capital expenditure decisions. Among different types of institutional investors, investment advisors perform a better monitoring function in improving the quality of capital expenditure decisions. Moreover, institutional investors play a more important role when internal governance is weak.

Contributions to Scholarship

This paper sheds light on the importance of institutional investors in improving firm performance through their monitoring roles and the channel of corporate investment. Our findings are in harmony with the existing literature (e.g., Almazan et al., 2005; Chen et al., 2007; Chung, Fung, & Hung, 2012; Jensen, 1993; Cornett et al., 2007) that informed and independent institutional investors can provide monitoring and improve corporate performance.

Applied Implications

Our findings highlight the importance of informed and independent institutional investors who can perform effective monitoring and enhance firm performance. Our findings also provide policy implications on the interactions between institutional ownership and other corporate governance mechanisms. A firm should encourage institutional investors to play an active role in corporate decisions and monitoring when internal governance is weak.

Limitations and Future Research Directions

This study excludes other sets of corporate decisions wherein institutional investors may perform monitoring and governance. Also, our sample excludes recent 2007–2009 global financial crises where the monitoring role of institutional investors might have changed. Further research into these directions will be important to understand the nexus between institutional investors and firm performance.

Notes

• 1

  Wahal (1996), Gillan and Starks (2000), and Davis and Kim (2007) suggested that institutional investors are unable to improve firm performance.

• 2

  Existing studies suggest that institutional investors have direct impact on firm value (Chen, Harford, & Li, 2007; Cornett, Marcus, Saunders, & Tehranian, 2007; Hartzell & Starks, 2003) and that institutional ownership affects market valuation of the firm (Gompers & Metrick, 2001; Smith, 1996).

• 3

  Previous studies have shown that investment can deviate from its optimal policy under asymmetric information and agency conflicts, and thus jeopardize firm value and overall shareholder returns (Jensen, 1986; Jensen & Meckling, 1976;Myers & Majluf, 1984).

• 4

  In addition, Almazan, Hartzell, and Starks (2005), Chen, Harford, and Li (2007), and Ferreira and Matos (2008) suggested that independent institutional investors are active monitors.

• 5

  Thomson reports the security holdings of institutional investors with greater than $100 million of securities under discretionary management.

• 6

  As an alternative measure of firm performance, we used residual (abnormal) earnings, which were estimated based on the Edwards-Bell-Ohlson (EBO) model. Following Dong, Hirshleifer, Richardson, and Teoh 2006), Residual Earnings are computed as: Net Income – CAPM returns * lagged book value of equity. This measures the return to shareholders above the required return on capital.

• 7

  As noted in Cleary, Povel, and Raith (2007), operating cash flow, rather than free cash flow, should be used to rule out the concern of endogeneity—particularly negative effect of investment on internal funds. Further, our analysis focuses on firms that are not subject to financial distress and our sample does not include firms with negative cash flows.

• 8

  All variables are winsorized at the 1st and 99th percentiles to reduce outliers’ influence.

• 9

  Petersen (2009) showed that when the residuals are correlated across firms or across time, OLS standard errors can be biased. Our results are reported using the Petersen (2009) method of clustered standard errors adjusted for intrafirm correlation that have the correct estimates of standard errors, and hence the correct inference on the significance of estimated coefficients.

• 10

  If the event of institutional investors on investment performance mainly coming from institutional investors unlikely to provide monitoring and governance, the result does not support the monitoring role of institutional investors.

• 11

  In the US, investment advisors are subject to a complex registration process by the Securities and Exchange Commission (SEC) and owe their clients an ongoing fiduciary duty to provide full and complete disclosure and to invest with their clients’ best interests.

• 12

  Hedge funds have exemption in many jurisdictions from regulations and can undertake a wider range of investments that are more aggressive in generating higher total returns.

• 13

  Cai, Garner, and Walkling (2009) found that removing staggered boards can enable the firm to remove poorly performing directors.

• 14

  Further examination (results not reported here) reveals that the interaction effect of high equity ownership by investment advisors and capital investments is greater and more significant before the endorsement of SOX Act in 2002. The introduction of SOX since July 2002 has strengthened firms’ internal control and disclosure requirement (see, e.g., Li, Pincus, & Rego, 2008).