do institutional investors monitor management? evidence from the relationship between institutional...

Please cite this article in press as: Chung, C. Y., & Wang, K. Do institutional investors monitormanagement? Evidence from the relationship between institutional. . . . North American Journal ofEconomics and Finance (2014), http://dx.doi.org/10.1016/j.najef.2014.10.001

ARTICLE IN PRESSG ModelECOFIN 470 1–31

North American Journal of Economics and Finance xxx (2014) xxx–xxx

Contents lists available at ScienceDirect

North American Journal ofEconomics and Finance

Do institutional investors monitormanagement? Evidence from the relationshipbetween institutional ownership and capitalstructure

Chune Young Chunga,∗, Kainan Wangb

Q2Q1

a School of Business Administration, Chung-Ang University, Seoul 156-756, South KoreaQ3b Department of Finance, College of Business and Innovation, University of Toledo, Toledo, OH 43606, USA

a r t i c l e i n f o

Article history:Received 23 April 2014Received in revised form 8 October 2014Accepted 9 October 2014Available online xxx

Keywords:Corporate governanceInstitutional monitoringHeterogeneous institutionsCapital structureSuboptimal leverage

JEL classification:G20G33G32G34

a b s t r a c t

We examine the dynamic relations between institutional owner-ship and a firm’s capital structure. We find that a firm’s leveragedecreases when institutional ownership increases. This resultimplies that a firm reduces its debt level as institutional investorssubstitute for the monitoring role of debt. More importantly, wefind that a firm’s suboptimal leverage decreases when the institu-tional ownership increases, and institutional ownership decreaseswhen a firm’s suboptimal leverage increases. This finding showsthat institutions not only effectively monitor a firm’s capital struc-ture but they also passively sell their shares when dissatisfied withit. In addition, we find that the monitoring evidence on a firm’sleverage and suboptimal leverage are more pronounced when theinstitutional investors are less likely to have business relationshipswith a firm or the information asymmetry is high in the market.

© 2014 Published by Elsevier Inc.

1. Introduction

The role of institutional investors in influencing firm management has become increasingly impor-tant, as the aggregate institutional ownership has substantially grown over the past decades. The

∗ Corresponding author. Tel.: +82 2 820 5544.E-mail address: [email protected] (C.Y. Chung).Q2

http://dx.doi.org/10.1016/j.najef.2014.10.0011062-9408/© 2014 Published by Elsevier Inc.

1

2

3

4

5

6

7

8

9

10

11

12

13

14

15

16

17

18

19

20

21

22

23

24

25

26

27

28

29

dx.doi.org/10.1016/j.najef.2014.10.001

dx.doi.org/10.1016/j.najef.2014.10.001

http://www.sciencedirect.com/science/journal/10629408

mailto:[email protected]

dx.doi.org/10.1016/j.najef.2014.10.001



2 C.Y. Chung, K. Wang / North American Journal of Economics and Finance xxx (2014) xxx–xxx

literature has documented several ways through which institutional investors may affect a firm’sfinancial decisions, including mergers and acquisitions (Ferreira, Massa, & Matos, 2010), payout pol-icy (Grinstein & Michaely, 2005), executive compensation (Hartzell & Starks, 2003), CEO turnover(Parrino, Sias, & Starks, 2003), earnings management (Chung, Firth, & Kim, 2002; Hsu & Koh, 2005;Wang, 2014), risk taking behavior (Chan, Lin, Chang, & Liao, 2013), and hedging policy (Tai, Lai, &Lin, 2014). In this paper, we examine the interrelationship between institutional ownership and afirm’s capital structure choices. Specifically, we investigate whether institutional ownership and firmleverage influence each other.

In theory, institutional investors may actively influence a firm’s capital structure. In studying agencycosts and sources, Jensen and Meckling (1976) argue that although debt helps reduce the agency costsof free cash flow, it also incentivizes managers to make distorted investment decisions. In contrast,institutional investors as large shareholders have strong incentives to enhance firm value. So theymay substitute for debt to mitigate agency costs. In this case, we should expect a negative relationshipbetween institutional ownership and a firm’s total leverage. Alternatively, institutional ownership anddebt may complement each other to reduce agency costs. Institutional investors who have sufficientvoting power to influence corporate decisions may pressure managers to make dividend payments,which likely leads to the need of future debt financing (La Porta, Lopez-De-Silanes, Shleifer, & Vishny,2000). As a result, management discretion is limited and the agency costs of free cash flow are reduced.If this holds true, we should expect a positive relationship between institutional ownership and a firm’stotal leverage. In addition, the trade-off theory of capital structure suggests that firms have a targetleverage at which firm value is maximized, and thus, any deviation from that target level will reducefirm value. If institutional investors can effectively monitor a firm’s capital structure, we should expecta negative relationship between institutional ownership and a firm’s deviant or suboptimal leverage.

Institutional investors may also passively influence firm management by liquidating their shares.The literature has shown that institutional investors “vote with their feet” when dissatisfied withmanagement. Their selling creates downward pressure on the stock price, which may subsequentlyfacilitate a change in management. Parrino et al. (2003) find greater reductions in institutional owner-ship in the year prior to forced CEO turnovers than in voluntary CEO turnovers. If institutional investorssell their shares when dissatisfied with a firm’s capital structure, we should expect a negative rela-tionship between institutional ownership and a firm’s suboptimal leverage.

To capture the dynamic relations between institutional ownership and a firm’s capital structure,we consider both contemporaneous and lagged relationships. The contemporaneous relationshipassumes that both firm and institutional investors are well informed and can make rapid decisions,whereas the lagged relationship posits that a quick change in capital structure is difficult due to mar-ket imperfections and that it takes time for institutional investors to adjust their ownership stakedue to liquidity shocks. Given the documented heterogeneity of institutional investors and firm man-agement, both relationships are possible and therefore deserve careful examination1. Specifically, westudy how concurrent firm characteristics, including leverage, are interrelated with concurrent insti-tutional ownership in the contemporaneous relationship, and examine how lagged firm information,including lagged leverage and institutional ownership, affects concurrent institutional ownership andleverage in the lagged relationship.

We use the three-stage least squares model to examine the above interrelationships between insti-tutional ownership and a firm’s capital structure. We find that both contemporaneous and laggedmodel specifications produce quantitatively similar results. In particular, we find that institutionalownership is negatively related to total leverage. We also show that an increase in a firm’s insti-tutional ownership leads to a decrease in its suboptimal leverage, and that a decrease in a firm’ssuboptimal leverage leads to an increase in its institutional ownership. Although the results for totaland suboptimal leverage are similar, the implications are different. The total leverage result impliesthat institutional investors can substitute for debt to exert monitoring efforts but does not indicatewhether that monitoring is effective in changing the firm’s management decision. The suboptimal

1 Our empirical results show that the contemporaneous relationship explains the data better and thus suggests that firmsand institutional investors may be well informed and little restricted by market imperfections.

30

31

32

33

34

35

36

37

38

39

40

41

42

43

44

45

46

47

48

49

50

51

52

53

54

55

56

57

58

59

60

61

62

63

64

65

66

67

68

69

70

71

72

73

74

75

76

77

78

79

dx.doi.org/10.1016/j.najef.2014.10.001



C.Y. Chung, K. Wang / North American Journal of Economics and Finance xxx (2014) xxx–xxx 3

leverage result, showing how institutional investors affect or respond to a firm’s suboptimal capi-tal structure, provides direct evidence that institutional monitoring not only serves as an alternativemechanism of firm monitoring but is also effective in changing a firm’s unsatisfactory debt policy.

To further confirm the role of institutional monitoring in firm leverage, we consider two specialcases in which the institutional monitoring effects are expected to be dominant. First, we followBrickley, Lease, and Smith (1998) and Almazan, Hartzell, and Starks (2005) and divide the institu-tional investors in our sample into pressure sensitive institutions and pressure insensitive institutions.Pressure sensitive institutions such as banks and insurance companies are likely to go along (or col-lude) with management to preserve current or potential business relationships. By contrast, pressureinsensitive institutions such as mutual funds and independent investment advisors, with weaker busi-ness relationships with the firm, are more likely to exert monitoring efforts. Moreover, we examinehow institutional monitoring changes throughout the business cycle. We hypothesize that firm man-agers will have more incentives to engage in opportunistic activities when the levels of informationasymmetry and uncertainty are high, such as in economic contractions. During such periods, institu-tional monitoring could be more effective in mitigating a firm’s agency costs. Our empirical resultsare consistent with both hypotheses. Overall, our findings add to our knowledge of the role of insti-tutional investors in corporate governance. We provide empirical evidence that institutions can notonly actively improve a firm’s capital structure but may also passively respond to a suboptimal capitalstructure by “voting with their feet.”

This paper makes several contributions to the literature. First, while numerous studies have exam-ined the monitoring role of institutional investors, little attention has been received for the impact ofinstitutional ownership on a firm’s capital structure. This paper takes a step toward filling this gap byinvestigating the dynamic relationship between institutional ownership and a firm’s capital structure.Second, this paper makes a methodological contribution. Specifically, by using leverage deviation, aneffectiveness measure of institutional monitoring, this paper provides direct evidence of the role ofinstitutional investors in improving a firm’s capital structure management. Third, this paper providesan analysis on which institutional investors monitor and when the monitoring occurs, which adds toour knowledge of the mechanism of institutional monitoring. Lastly, the findings in this paper willhelp practitioners select investment targets. Practitioners who are concerned about a firm’s capitalstructure management may use the change in institutional ownership as a key variable in makinginvestment decisions.

The remainder of the paper is organized as follows. The next section reviews the literature anddevelops the testable hypotheses. The third section discusses the sample and methodology. The fourthsection provides the empirical results. The final section concludes the paper.

2. Related literature and testable hypotheses

2.1. Related literature

The agency theory of Jensen and Meckling (1976) suggests that managers may pursue their ownbenefits by engaging in opportunistic behaviors rather than maximizing shareholders’ wealth. Jensen(1986) argues that using debt may mitigate this agency problem, as the firm must make periodicpayments of interest and principal, which could otherwise have been used by managers for value-decreasing activities. However, the literature also suggests that too much debt may lead to increasedagency costs. In particular, Jensen and Meckling (1976) discuss the risk-shifting or asset substitutionproblem, as they view equity as a call option on the firm’s assets. They show that a high level of debtprovides an incentive for equity holders to undertake high-risk projects. In addition, Myers (1977)argues that a high level of debt keeps equity holders from investing in positive net present value (NPV)projects. This underinvestment problem occurs when equity holders bear the costs of an investmentbut realize only part of the benefits. In a rational expectation setting, debt holders are aware of therisk-shifting and underinvestment problems of equity holders and thus demand a higher requiredreturn on debt, which increases the costs of debt financing.

Previous work on institutional monitoring shows that large shareholders such as institutionalinvestors may have a strong incentive to monitor firm management because they can obtain benefits

80

81

82

83

84

85

86

87

88

89

90

91

92

93

94

95

96

97

98

99

100

101

102

103

104

105

106

107

108

109

110

111

112

113

114

115

116

117

118

119

120

121

122

123

124

125

126

127

128

129

130

dx.doi.org/10.1016/j.najef.2014.10.001




from enhanced firm value (Shleifer & Vishny, 1986; Huddart, 1993; Karpoff, Malatesta, & Walkling,1996; Gillan & Starks, 1998; Maug, 1998). Institutions may exert their monitoring effort through“institutional activism.” Carleton, Nelson, and Weisbach (1998) show that institutional investors alsoinfluence corporate governance through private negotiations. Gillan and Starks (2000) show thatUS institutional investors can pressure management by presenting shareholder proposals at corpo-rate meetings. Chung et al. (2002) argue that institutional investors constrain firm manipulation ofearnings. In addition, Hsu and Koh (2005) and Wang (2014) further show that the role of institu-tional investors in monitoring accruals managements depends on the types of institutional investorsinvolved.

Another strand of literature suggests that the costs of active monitoring could be greater than itsbenefits (Coffee, 1991; Bhide, 1994). Given the increased liquidity of the US market, it may be easierfor institutions to simply sell their shares when dissatisfied with management. Thus, institutionalinvestors can passively, or indirectly, monitor firms by “voting with their feet.” Parrino et al. (2003)find that institutions reduce their shares in the year prior to forced CEO turnovers. However, Smith(1996) and Del Guercio and Hawkins (1999) show that such institutional investor activism does nothave a significant impact on a firm’s performance. Webb, Beck, and McKinnon (2003) document therestrictions imposed by institutional investor participation on corporate governance.

Given the costs of debt monitoring, institutional investors may substitute for debt as an alternativemonitoring force. Alternatively, institutional investors may play a complementary role with debt toreduce the agency costs between managers and shareholders. Both theories imply that the level ofinstitutional ownership affects a firm’s debt level. Michaely and Vincent (2013) find that institutionalownership reduces firm leverage. La Porta et al. (2000) show that institutional investors may induce thefirm to issue debt so that managers’ discretion is limited. Given the costs of institutional monitoring, afirm’s debt level may also influence institutional ownership. Institutional investors may prefer certainfirm characteristics, which may link institutional ownership to a firm’s debt level. Del Guercio (1996)and Gompers and Metrick (2001) show that institutions tend to prefer large and safe stocks. Bennett,Sias, and Starks (2003) find that institutional investors have changed their investment preference fromlarge and safe stocks to small and risky stocks over time.

More importantly, a firm’s capital structure is related to firm value. The trade-off theory of capitalstructure hypothesizes that the optimal debt level is determined by financial distress costs and thetax benefits of debt (Modigliani & Miller, 1963). A refinement of the trade-off theory also considersthe agency costs of debt (Jensen & Meckling, 1976; Myers, 1977). If a firm has an optimal debt level,institutional investors may have incentives to adjust the firm’s leverage toward that optimal level.Moreover, institutional investors may passively sell their shares when dissatisfied with a firm’s sub-optimal debt policy and when monitoring costs are greater than its benefits. Therefore, the trade-offtheory also suggests a simultaneous relationship between institutional ownership and a firm’s debtlevel.

2.2. Testable hypotheses

We investigate the potential interrelationship between institutional ownership and a firm’s capitalstructure. In particular, we focus on the monitoring role of institutional investors in firm managementrelated to debt policy. To investigate whether the monitoring by institutional investors substitutesfor or complements to the monitoring by debt holders and whether institutional investors show pref-erence for a firm’s debt level, we simultaneously estimate the interrelationship between changes inaggregate institutional ownership and changes in a firm’s debt level.

To better understand the monitoring role of institutional investors, we also analyze two interestingsubsamples. First, we divide institutional investors into two groups: pressure sensitive institutionsand pressure insensitive institutions. Brickley et al. (1998), discussing the differences between thetwo groups, argue that pressure insensitive institutions such as banks and insurance companies areless likely to have investment and business relationships with the firms they invest in. They find thatinstitutional investor activism in proxy voting is more evident in pressure insensitive institutions.Almazan et al. (2005) document that institutional investor activism in executive compensation ismore pronounced in pressure insensitive institutions. In addition, we divide the sample into economic

131

132

133

134

135

136

137

138

139

140

141

142

143

144

145

146

147

148

149

150

151

152

153

154

155

156

157

158

159

160

161

162

163

164

165

166

167

168

169

170

171

172

173

174

175

176

177

178

179

180

181

182

dx.doi.org/10.1016/j.najef.2014.10.001




contraction and expansion periods because institutional investors’ monitoring influence should bestronger when information asymmetry is high in contractions.

The trade-off theory of capital structure suggests that any deviation from a firm’s target leverage isharmful to firm value. To investigate whether institutional investors reduce a firm’s deviated leverageand whether the deviated leverage affects a firm’s institutional ownership, we simultaneously examinethe relationship between the change in aggregate institutional ownership and the change in a firm’sleverage deviation, defined as the absolute value of difference between a firm’s total leverage and itspredicted leverage. To obtain the predicted leverage, we use the tobit regression model, similar toKayhan and Titman (2007) and Harford, Klasa, and Walcott (2009). Similar to the first set of tests, wealso consider various institution types and business cycles because institutional activism is expectedto be more prominent in pressure insensitive institutions or during contractions.

The above discussion leads to the following testable hypotheses:

Hypothesis 1. Changes in institutional ownership have a significant effect on changes in a firm’stotal leverage, as the monitoring roles of institutional investors and debt are either substitutes orcomplements to one another.

Hypothesis 2. Hypothesis 1 is more evident in pressure insensitive institutions or in economiccontractions, as the monitoring role should be stronger in these situations.

Hypothesis 3. Changes in institutional ownership have a negative effect on changes in a firm’sleverage deviation, as institutional investors improve a firm’s capital structure management.

Hypothesis 4. Changes in a firm’s leverage deviations have a negative effect on changes in institu-tional ownership, as institutional investors vote with their feet.

Hypothesis 5. Hypotheses 3 and 4 are more evident in pressure insensitive institutions or in economiccontractions, as the monitoring effect should be greater in these circumstances.

Hypotheses 1 and 2 examine whether institutional investors substitute for or complements to debtto monitor firm management, which would provide indirect evidence of institutional monitoring.Hypotheses 3–5 analyze the effectiveness of institutional monitoring on a firm’s suboptimal capitalstructure management, which would be direct evidence of institutional monitoring.

3. Sample and methodology

3.1. Data

Our data come from three sources. We obtain quarterly institutional holdings from theCDA/Spectrum institutional ownership (13F) database. Under Section 13F of the Securities ExchangeAct of 1934, all institutional investors with $100 million or more under management are required toreport all equity positions greater than 10,000 shares, or $200,000, to the SEC at the end of each quarter.We obtain share prices, stock returns, firm capitalizations, and SIC codes from the Center of Researchin Security Prices (CRSP) database. We exclude all financial (SIC codes 6000–6999) and utilities firms(SIC codes 4900–4949) because these industries are highly regulated. We obtain annual accountinginformation from the COMPUSTAT database. We obtain the Fama-French 49-industry classificationsfrom Kenneth French’s website2. The sample period is from 1985 to 2008.

The CDA/spectrum institutional ownership data divide institutional investors into five types: banktrust departments, insurance companies, mutual funds, independent investment advisors, and otherunclassified institutional investors. To consider heterogeneous monitoring incentives among insti-tutions, we further classify each institution into two types: we group bank trust departments andinsurance companies as pressure sensitive institutions and mutual funds and independent invest-ment advisors as pressure insensitive institutions. We expect institutional monitoring to be moreevident in pressure insensitive institutions.

2 We thank Kenneth French for providing the 49 industry classification data (http://mba.tuck.dartmouth.edu/pages/faculty/ken.french/data library.html).

183

184

185

186

187

188

189

190

191

192

193

194

195

196

197

198

199

200

201

202

203

204

205

206

207

208

209

210

211

212

213

214

215

216

217

218

219

220

221

222

223

224

225

226

227

dx.doi.org/10.1016/j.najef.2014.10.001




We identify the economic contraction and expansion periods using the National Bureau of EconomicResearch (NBER) business cycle3. We expect institutional monitoring to be more effective duringeconomic contractions due to high information asymmetry.

3.2. Institutional ownership and leverage definition

We measure institutional ownership as the fraction of a firm’s shares held by institutional investorsand total leverage as a firm’s total debt divided by its total assets. We calculate leverage deviation asthe absolute difference between the total leverage and the expected, or target, leverage.

We compute the target leverage using the methodology of Kayhan and Titman (2007), Frank andGoyal (2009), and Harford et al. (2009). Specifically, we use the following model to obtain the targetleverage for firm i in year t:

Leveragei,t = ̨ + ˇ1 × MBi,t−1

+ ˇ2 × EBITDi,t−1 + ˇ3 × R&Di,t−1 + ˇ4 × R&D Di,t−1 + ˇ5×

PPEi,t−1+ˇ6 × SEi,t−1+ˇ7 × SIZE Si,t−1+ˇ8 × Asset betai,t−1+�Industry×Industry Di+�Year×Year Dt + i,t

(1)

where M/B is the market-to-book ratio; EBITD is the earnings before interest, taxes, and depreciationscaled by total assets; R&D is research and development expenses scaled by sales; R&D D is a dummyvariable equal to 1 if a firm does not report any research and development expenses and 0 otherwise;PPE is the net property, plants, and equipment scaled by total assets; SE is the selling expenses scaled bysales; SIZE S is the natural log of sales; Asset beta is the weighted average beta of equity and debt; andIndustry and Year are two dummy variables controlling for industry and year effects. All explanatoryvariables are selected based on recommendations in the literature and are updated annually. Wesummarize the relevant literature in the next subsection.

We use the tobit regression model to estimate Eq. (1) because the leverage ratios lie between 0and 1. The target leverage is obtained as the predicted value of the leverage in Eq. (1). We provide thecoefficients for estimating the target leverage in the Appendix.

3.3. Explanatory variables for target leverage

The literature suggests a relationship between growth opportunities and a firm’s leverage choice.For example, Myers (1977) argues that high growth firms with high debt levels can have an underin-vestment or debt-overhang problem because managers seeking to increase shareholder wealth mayforego positive NPV projects when most of the value accrues to debt holders. In addition, Jensen (1986)and Stulz (1990) suggest that higher levels of debt are preferred for low growth firms with large freecash flows to discourage managers from overinvesting in risky projects. We use the market-to-bookratio (M/B) as a proxy for firms’ growth opportunities. A higher M/B ratio suggests that a firm tendsto have higher growth opportunities. However, a higher M/B may also capture stock mispricing. Forinstance, Baker and Wurgler (2002) show that firms with a higher M/B may reduce debt levels as theyissue more equity to exploit the high stock price.

We use EBITD (earnings before interest, taxes, and depreciation/total assets) as the profitabilitymeasure. Even though the trade-off theory implies that more profitable firms should have more debtbecause their expected financial distress costs are low, empirical findings suggest a negative relation-ship between profitability and leverage. Several explanations for this relationship are possible. Thepecking order theory in Myers (1977) implies that highly profitable firms are more likely to have lowQ4debt levels because they have more internal funds. Market frictions could be another cause. Fischer,Heinkel, and Zechner (1989) show that high debt adjustment costs may result in lower debt levelsfor more profitable firms. Alternatively, Flannery and Rangan (2006) and Kayhan and Titman (2007)argue that firms may strategically keep their debt levels low in order to deter competitors’ entry into

3 This information is available at http://www.nber.org/cycles.html.

228

229

230

231

232

233

234

235

236

237

238

239

240

241

242

243

244

245

246

247

248

249

250

251

252

253

254

255

256

257

258

259

260

261

262

263

264

265

266

267

268

269

dx.doi.org/10.1016/j.najef.2014.10.001




their lines of business; their argument relies on the hypothesis that higher profitability implies marketdominance.

We use R&D (research and development expenses/sales), R&D D (1 if a firm has no R&D expenses and0 otherwise), SE (selling expenses/sales), and PPE (net property, plant, and equipment/total assets)to capture the nature of the assets employed in the firm. The trade-off theory suggests that firmswith higher distress costs should use less debt, while the pecking order theory implies that firmswith greater information asymmetry should choose debt over equity for external financing decisions.Because higher R&D may lead to higher distress costs and more debt-related agency costs, the trade-offtheory implies a negative relationship between R&D and leverage. Moreover, higher R&D investmentsmay be associated with greater information asymmetry. Thus, the pecking order theory predicts apositive relationship between R&D and leverage. Because many firms do not report R&D in theirfinancial statements, we also include the dummy variable R&D D to control for any possible R&Deffects. Because higher PPE implies more tangible assets, which subsequently leads to lower distresscosts, the trade-off theory suggests a positive relationship between PPE and leverage. By contrast,higher SE is usually linked to higher distress costs, especially in economic contractions, implying anegative relationship between SE and leverage.

Size S (natural log of sales) and Asset beta (equity/total assets × equity beta) are also related tofirm debt levels. The trade-off theory implies that larger firms face lower risks due to their greaterdiversifications and can thus hold more debt. By contrast, the pecking order theory suggests a negativerelationship between firm size and debt because small firms are more influenced by informationasymmetry. Asset beta measures a firm’s systematic risk. Because firms with high Asset beta face highdistress costs, a negative relationship between Asset beta and leverage is expected.

A firm’s target debt level can vary across industries and years (Lemmon, Roberts, & Zender, 2008).In addition, a firm’s debt policy can be influenced by exogenous changes in government policies,regulations, or market conditions. To control for these effects, we include Industry D (industry dummy),based on the Fama-French 49 industry classifications, and Year D (year dummy) as two additionalexplanatory variables.

3.4. Empirical model

To identify the simultaneous relationship between institutional ownership and a firm’s capitalstructure and to deal with the potential interrelations with other firm characteristics, we use a mul-tivariate regression model with a simultaneous system of equations. We examine two importanteconomic questions: how changes in institutional ownership affect the changes in leverage-relatedmeasures (total leverage and leverage deviation) and how changes in leverage-related measures affectchanges in institutional ownership.

We consider both contemporaneous and lagged model specifications. The contemporaneousmodel is:

�Leverage (deviation)i,t = ˛0 + ˛1 × MBi,t

+ ˛2 × EBITDi,t + ˛3 × R&Di,t + ˛4 × R&D Di,t + ˛5×

PPEi,t + ˛6 × SEi,t + ˛7 × SIZE Si,t + ˛8 × Asset betai,t + ˛9 × Industry Di + ˛10 × Year Dt + ˛11×

�Ownershipi,t + �i,t

(2)

�Ownershipi,t = ˇ0 + ˇ1 × Agei,t + ˇ2 × Betai,t + ˇ3 × Dividend yieldi,t + ˇ4×

Firm-specific riski,t + ˇ5 × Lag returni,t + ˇ6 × Pricei,t + ˇ7 × Standard deviationi,t + ˇ8 × Size Ei,t + ˇ9×

Turnoveri,t + ˇ10 × Industry Di + ˇ11 × Year Dt + ˇ12 × �Leverage (deviation)i,t + vi,t

(3)

where �leverage (deviation) is the change in leverage (deviation) over the year, �ownership is thechange in institutional ownership over the year, and � and � are error terms.

We use the same set of firm characteristic variables used in Eq. (1) to explain the changes in leverage(deviation) and nine other variables to explain the changes in institutional ownership. Although thefirm characteristic variables in Eq. (2) are the same as those in Eq. (1), the model specifications are

270

271

272

273

274

275

276

277

278

279

280

281

282

283

284

285

286

287

288

289

290

291

292

293

294

295

296

297

298

299

300

301

302

303

304

305

306

307

308

309

310

311

312

dx.doi.org/10.1016/j.najef.2014.10.001




different. For example, we always use the one-year lagged firm characteristics to compute the leveragedeviation in Eq. (1).

We use beta, standard deviation of returns, and firm-specific volatility to measure risk; we use firmsize, firm age, and dividend yield to capture institutional investors’ preference, and we use share priceand share turnover to measure liquidity. We use lagged returns to control for institutional investors’momentum trading. These characteristic variables are common and well-documented in the literature.We define these firm characteristics similarly to Del Guercio (1996), Falkenstein (1996), Gompers andMetrick (2001), and Bennett et al. (2003). These characteristics are considered to reflect institutionalinvestors’ heterogeneous preferences, distinguishing them from individual investors. In particular,we compute a firm’s beta using rolling regressions based on Fama and French (1992). We estimate afirm’s standard deviation as the natural logarithm of the standard deviation of monthly returns overthe previous 24 to 60 months, depending on data availability. We compute firm-specific volatility asthe natural logarithm of 1 plus the firm-specific risk. Following Campbell, Lettau, and Xu (2001), wecompute firm-specific risk as the average daily squared firm-specific returns over a month, in whichthe firm-specific return is defined as the difference between a firm’s daily return and its industryreturn. We use the Fama-French 49 industry classifications to identify each firm’s industry. We thencompute the daily industry return as the equal-weighted portfolio return, including all firms withineach industry. Firm size is the natural logarithm of equity capitalization (Size E). Firm age is thenatural logarithm of the number of months for which the firm has CRSP return data. Dividend yieldis the natural logarithm of 1 plus the average monthly dividend yield over the previous 12 months.Share price is the natural logarithm of 1 plus the month-end share price. Share turnover is the naturallogarithm of 1 plus the ratio of the monthly volume to the number of shares outstanding. Laggedreturn is the buy-and-hold return over the previous six months. By construction, the firm characteristicvariables in Eq. (2) are obtained annually whereas those in Eq. (3) are updated monthly.

Because institutional ownership data are available every quarter, we match each firm’s leveragemeasure with its most recent quarterly institutional ownership. For example, if a firm has a fiscal yearend month in February, �Leverage (deviation)i,t is computed as the difference in leverage (deviation)between February year t − 1 and February year t; �Ownershipi,t is calculated as the difference ininstitutional ownership between December year t − 2 and December year t − 1. The firm characteristicvariables in Eq. (3) are calculated as the monthly averages occurring from December year t − 2 toDecember year t − 14.

We also consider a lagged model with the following specification:

�Leverage (deviation)i,t = ˛0 + ˛1 × MBi,t−1

+ ˛2EBITDi,t−1 + ˛3 × R&Di,t−1 + ˛4×

R&D Di,t−1 + ˛5 × PPEi,t−1 + ˛6 × SEi,t−1 + ˛7 × Size Si,t−1 + ˛8 × Asset betai,t−1 + ˛9×

Industry Di + ˛10 × Year Dt + ˛11 × �Ownershipi,t−1 + �i,t

(4)

�Ownershipi,t−1 = ˇ0 + ˇ1 × Agei,t−2 + ˇ2 × Betai,t−2 + ˇ3 × Dividend yieldi,t−2 + ˇ4×

Firm-specific riski,t−2 + ˇ5 × Lag returni,t−2 + ˇ6 × Pricei,t−2 + ˇ7 × Standard deviationi,t−2+

ˇ8 × Size Ei,t−2 + ˇ9 × Turnoveri,t−2 + ˇ10 × Industry Di + ˇ11 × Year Dt−1+ˇ12×�Leverage (deviation)i,t−2+vi,t−1

(5)

Eqs. (4) and (5) are identical to Eqs. (2) and (3), except that �ownership is at year t − 1 and allexplanatory variables are one-year lagged. Because firm leverage and institutional ownership areavailable at different frequencies, in Eq. (4), we match each firm’s leverage measure with the mostrecent quarterly institutional ownership that is at least one year prior to the firm’s fiscal year end.For example, if a firm has a fiscal year-end month in February, �Leverage (deviation)i,t is computedfrom February year t − 1 to February year t, and �Ownershipi,t−1 is calculated from March year t − 2to December year t − 25. Similarly, �Leverage (deviation)i,t−2 is calculated from February year t − 3 to

4 We use monthly averages because we do not know the firm characteristics in which month affect the changes in institutionalownership over a year.

5 Notice that, in this particular example of a lagged model, �Ownershipi,t−1 is calculated over nine months instead of oneyear to avoid the time overlapping between �Ownershipi,t−1 and �Leverage (Deviation)i,t−2. In about 20% of the observations

313

314

315

316

317

318

319

320

321

322

323

324

325

326

327

328

329

330

331

332

333

334

335

336

337

338

339

340

341

342

343

344

345

346

347

348

349

350

351

352

353

dx.doi.org/10.1016/j.najef.2014.10.001




Table 1Summary statistics.

Mean S.D. 10% Median 90%

Panel A: institutional ownership (%)Bank trust departments 0.0628 0.0667 0.0032 0.0439 0.1462Insurance companies 0.0242 0.0371 0.0045 0.0121 0.0626Pressure sensitive institutions 0.0871 0.0416 0.0185 0.0646 0.1201Mutual funds 0.0639 0.0810 0.0054 0.0313 0.1766Independent investment advisers 0.2170 0.1685 0.0267 0.1862 0.4475Pressure insensitive institutions 0.2810 0.1445 0.0143 0.2386 0.4127Unclassified institutions 0.0245 0.0353 0.0087 0.0107 0.0644

Total 0.3928 0.2394 0.0508 0.3591 0.7810

Panel B: leverage-related characteristics (%)Leverage 0.3564 0.2394 0.0814 0.3218 0.6909Target leverage 0.3601 0.1378 0.1924 0.3713 0.5072Leverage–target leverage −0.0568 0.1910 −0.7863 −0.0481 0.6620Leverage deviation 0.5240 0.1659 0.0728 0.3853 0.8264

This table reports the sample statistics for the principle variables. Panel A shows the percentage of institutional ownership (inaggregate and by type), and Panel B shows the leverage-related firm characteristics for 40,963 observations over the periodfrom 1985 to 2008. Pressure sensitive institutions include bank trust department and insurance companies, and pressureinsensitive institutions include mutual funds and independent investment advisors. Leverage is total debt divided by totaldebt plus market value of equity. Target leverage is the predicted leverage from the tobit regression model documented inthe Appendix. Leverage–target leverage is the difference between total leverage and target leverage. Leverage deviation is theabsolute value of difference between total leverage and target leverage.

February year t − 2. The firm characteristic variables in Eq. (5) are computed in the month end priorto �Ownershipi,t−1. In the previous example, the firm characteristics are obtained in February yeart − 2.

We use the three-stage least squares (3SLS) model to estimate the system of equations. By design,�ownership and �leverage (deviation) on the right hand side of each equation are endogenous vari-ables. We estimate the system of equations in two steps. First, following Wooldridge (2002), we obtainthe predicted values of �leverage (deviation) and �ownership using their respective exogenousvariables. We use the tobit regression model to estimate the predicted values, because �leverage(deviation) and �ownership are restricted to lie between −1 and 1. We then replace the dependentvariables �leverage (deviation) and �ownership with their predicted values in the system of equa-tions. In the second step, we use the seemingly unrelated regression (SUR) model to estimate thesystem of equations and obtain the coefficients6.

3.5. Descriptive statistics

Panel A in Table 1 reports the sample statistics of institutional ownership (in aggregate and by type).We find that institutional investors held 39% of each firm’s shares outstanding. Pressure sensitiveinstitutions such as bank trust departments and insurance companies held about 9% of each firm’sshares, accounting for about 23% of all institutional shares. Pressure insensitive institutions such asmutual funds and independent investment advisors accounted for approximately 28% of each firm’sshares, and for about 71% of all institutional shares.

Fig. 1 shows substantial changes in institutional ownership in aggregate and by type over our sam-ple period. The figure reveals two striking trends. First, average institutional ownership has increased

in our sample, firm leverage and institutional ownership are obtained at different dates (i.e., when the firm has a fiscal year-endmonth not in March, June, September, or December).

6 SUR allows the error terms in different equations to be cross-correlated, providing higher estimation efficiency. In arobustness test, we also consider the two-step GMM estimation in the second stage to deal with heteroscedasticity andcross-correlation. Our main findings remain the same.

354

355

356

357

358

359

360

361

362

363

364

365

366

367

368

369

370

371

372

373

374

dx.doi.org/10.1016/j.najef.2014.10.001




0%

10%

20%

30%

40%

50%

60%

70%

80%

1985 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008

Ave

rge

% o

f ow

ners

hip

by in

stit

utio

ns

Year

Total Instit uti onal Ow nership

Press ure Se nsitive Instit uti ons

Press ure Insensiti ve Instit ution s

Fig. 1. Average % of ownership by total Institutions, pressure sensitive institutions, and pressure insensitive institutions.

significantly in the US, from about 23% in 1985 to about 68% in 2008. This not only shows institutionalinvestors’ increasing importance in the stock market but also implies their growing active partici-pation in corporate decisions, which provides a solid ground for our study. Second, the increase ininstitutional ownership is primarily driven by pressure insensitive institutions, which are perceivedto be more informed and less likely to go along with management decisions (Brickley et al., 1998;Almazan et al., 2005). This implies that, on average, institutional monitoring may have become moredominant and important.

Panel B in Table 1 reports the sample statistics of total leverage, target leverage, and leveragedeviation. We find that an average firm in our sample finances approximately 35% of its asset throughdebt. There is substantial variation in debt levels, suggesting that each firm may have a different debtpolicy and/or that it may vary over time. A firm’s target leverage is 36%, which is very close to the totalleverage. The leverage deviation is right skewed and has a mean of 52%, larger than that for either totalor target leverage. This result suggests that the costs and benefits of adjusting the leverage deviationcould be significantly different across firms, possibly due to market imperfections.

Table 2 reports the univariate relationships between institutional ownership (in aggregate and bytype) and leverage-related firm characteristics. We find a statistically significant negative relation-ship between institutional ownership and total leverage, holding for all institution types and with aslightly larger magnitude for pressure insensitive institutions. These results suggest that institutionalinvestors may substitute for debt as a monitoring vehicle for firm management. Similar results arefound for the relationship between institutional ownership and leverage deviation. Because lever-age deviation proxies for a suboptimal debt policy, this result implies that institutional investorsmay actively improve a firm’s capital structure through monitoring efforts. However, the causalityof these relationships should be interpreted with caution because of the potential endogeneity ofinstitutional ownership. For example, institutional investors are likely to prefer firms with low totalleverage and small leverage deviation; if so, we would almost certainly find a negative correlationbetween institutional ownership and total leverage as well as leverage deviation. For the target lever-age, we find a negative correlation with the aggregate institutional ownership, a positive correlationwith the ownership of pressure sensitive institutions, and a negative correlation with the ownershipof pressure insensitive institutions. This result suggests that institutional ownership could be anotherimportant factor determining a firm’s target leverage. As institutional ownership may be endogenous,this result could also mean that a firm’s target leverage affects institutional ownership.

375

376

377

378

379

380

381

382

383

384

385

386

387

388

389

390

391

392

393

394

395

396

397

398

399

400

401

402

403

404

405

dx.doi.org/10.1016/j.najef.2014.10.001

Please cite

this

article in

press

as: C

hu

ng,

C.

Y.,

& W

ang,

K.

Do

institu

tional

investors

mon

itorm

anagem

ent?

Eviden

ce from

the

relationsh

ip betw

een in

stitution

al. .

. .

North

Am

erican Journal

ofEconom

ics and

Finance (2014),

http

://dx.d

oi.org/10.1016/j.najef.2014.10.001

AR

TIC

LE

IN P

RE

SS

G M

odelEC

OFIN

470 1–31

C.Y. Chung,

K.

Wang

/ N

orth A

merican

Journal of

Economics

and Finance

xxx (2014)

xxx–xxx

11

Table 2Correlations between institutional ownership and leverage-related firm characteristics.

Total Bank trustdepartments

Insurancecompanies

Mutualfunds

Independentinvestmentadvisers

Unclassifiedinstitutions

Pressuresensitiveinstitutions

Pressureinsensitiveinstitutions

Leverage −0.101 (0.000)*** −0.051 (0.000)*** −0.029 (0.000)*** −0.083 (0.000)*** −0.082 (0.000)*** −0.061 (0.000)*** −0.050 (0.000)*** −0.105 (0.000)***

Target leverage −0.020 (0.000)*** 0.093 (0.000)*** 0.054 (0.000)*** −0.056 (0.000)*** −0.050 (0.000)*** 0.023 (0.000)*** 0.092 (0.000)*** −0.068 (0.000)***

Leverage deviation −0.105 (0.000)*** −0.075 (0.000)*** −0.060 (0.000)*** −0.058 (0.000)*** −0.090 (0.000)*** −0.084 (0.000)*** −0.081 (0.000)*** −0.095 (0.000)***

This table presents the Pearson correlation coefficients between institutional ownership (in aggregate and by type) and leverage-related firm characteristics. Pressure sensitive institutionsinclude bank trust department and insurance companies, and pressure insensitive institutions include mutual funds and independent investment advisors. Leverage is total debt dividedby total debt plus market value of equity. Leverage deviation is the difference between total leverage and target leverage. Target leverage is the predicted leverage from the tobit regressionmodel documented in the Appendix. The p-values are reported in parentheses.Q6

dx.doi.org/10.1016/j.najef.2014.10.001




Table 3Relationship between changes in institutional ownership and changes in leverage.

Model Contemporaneous Lagged Contemporaneous Lagged

Dep. variable �Leverage �Ownership

Intercept 0.051 (0.000)*** −0.013 (0.002)*** Intercept 0.044 (0.000)*** 0.011 (0.175)�Ownership −1.671 (0.000)*** −0.997 (0.000)*** �Leverage −0.049 (0.026)** 0.346 (0.000)***

Asset beta −0.011 (0.000)*** 0.022 (0.000)*** Age −0.007 (0.000)*** −0.004 (0.001)***

EBITD −0.016 (0.000)*** 0.055 (0.000)*** Beta −0.002 (0.082)* −0.006 (0.000)***

M/B −0.003 (0.000)*** 0.008 (0.000)*** Dividend yield −1.703 (0.000)*** −1.255 (0.000)***

PPE −0.007 (0.025)** −0.008 (0.024)** Firm-specific risk −0.283 (0.000)*** 0.036 (0.115)R&D −0.004 (0.038)** −0.003 (0.470) Lag return 0.073 (0.000)*** 0.008 (0.000)***

R&D D 0.005 (0.001)*** 0.006 (0.000)*** Price −0.001 (0.341) −0.012 (0.000)***

SE −0.001 (0.894) 0.001 (0.827) Standard deviation 0.031 (0.008)*** 0.0367 (0.005)***

Size S −0.001 (0.096)* −0.002 (0.000) Size E 0.001 (0.091)* 0.005 (0.000)***

Turnover −0.010 (0.000)*** −0.027 (0.000)***

Adjusted-R2 0.148 0.028 0.032 0.006N 40,963 40,456 40,963 40,456

This table reports the coefficient estimates for both the contemporaneous and lagged relationships between changes in institu-tional ownership and changes in total leverage from 1985 to 2008. Leverage is total debt divided by total debt plus the marketvalue of equity, and �Leverage is the change in leverage over the year. �Ownership is the change in institutional ownershipover the year. EBITD is earnings before interest, taxes, and depreciation divided by assets. M/B is market value of asset dividedby book value of asset. PPE is net property, plants, and equipment divided by assets. R&D is research and development expensesdivided by sales. R&D D is equal to 1 if a firm does not report the research and development expenses. SE is selling expensesdivided by sales. Size E and Size S are the natural logarithm of the market value of equity and sales, respectively. Beta is fromthe rolling regressions based on Fama and French (1992). Standard deviation is the natural logarithm of the monthly standarddeviation of returns over the previous 24 to 60 months. Asset beta is weighted average beta of equity and debt. Firm-specificrisk is the natural logarithm of 1 plus the average daily squared firm-specific returns over a month, and the firm-specific returnis the difference between a firm’s daily return and its industry return. Age is the natural logarithm of the number of months forwhich the firm has CRSP return data. Dividend yield is the natural logarithm of 1 plus the average monthly dividend yield overthe previous 12 months. Price is the natural logarithm of 1 plus the month-end share price. Turnover is the natural logarithm of1 plus the ratio of monthly volume to number of shares outstanding. Lag Return is the buy-and-hold return over the previoussix months. Though not reported, an industry dummy based on the Fama-French 49 industry classifications and a year dummyare also included in the model. p-Values are reported in parentheses.***, **, and * indicate statistical significance at the 1%, 5%, and 10% levels, respectively.

4. Empirical results

4.1. Indirect evidence of institutional monitoring: Institutional monitoring as a substitute for debtmonitoring

4.1.1. Relationship between changes in institutional ownership and changes in total leverageTable 3 reports the coefficient estimates for both the contemporaneous and lagged relationships

between changes in institutional ownership and changes in total leverage7. We find that firms arelikely to reduce their total leverage as institutional investors increase their shares during the concur-rent and previous year, suggesting that institutional investors may substitute for debt to engage infirm monitoring. The substitution of increased institutional monitoring for reduced debt monitoringmay help reduce the agency costs of risk-shifting or underinvestment problems. Although this findingis consistent with Bathala, Moon, and Rao (1994), our methodology is more robust in identifying thesubstitutability of institutional monitoring for debt. In particular, our methodology relies on exten-sive institutional ownership data and considers both the contemporaneous and lagged relationshipsbetween institutional ownership and a firm’s capital structure.

We also find that institutional ownership increases when total leverage has increased during thelast year. This may be evidence that institutional investors prefer firms with reduced monitoring costs

7 For a robustness analysis, we incorporate the lagged leverage into the model to capture the adjustment of leverage docu-mented in Lemmon et al. (2008). We find that our main results are supportive of the adjustment of leverage. The results areavailable upon request.

406

407

408

409

410

411

412

413

414

415

416

417

418

419

420

421

dx.doi.org/10.1016/j.najef.2014.10.001




due to an increase in debt monitoring. Table 1 shows that an average firm in our sample has a totalleverage lower than its target leverage. Because a firm’s leverage is related to its value based on thetrade-off theory, this result could also mean that institutional investors prefer firms with a higher valuedue to increased leverage. Moreover, we find that institutional ownership declines as the concurrenttotal leverage increases, perhaps because increases in total leverage convey a bad signal to institutionalinvestors, who may have low incentives to perform monitoring if the associated costs are greater thanits benefits.

4.1.2. Relationship between changes in institutional ownership and changes in total leverage forpressure sensitive and insensitive institutions

If institutional investors substitute for debt to influence management activities, we should expectthe substitutability to be more pronounced in institutions that are more informed and/or less likely togo along with management. To validate this conjecture, we divide the aggregate institutional owner-ship into pressure sensitive institutional ownership and pressure insensitive intuitional ownership.Panels A and B in Table 4 report the coefficient estimates for the interrelationships between changes ininstitutional ownerships and changes in total leverage for pressure sensitive institutions and pressureinsensitive institutions, respectively. To test the statistical significance of the difference in coefficientsbetween the two groups of institutions, we employ the Wald test for the key coefficients by simulta-neously estimating the models for pressure sensitive and insensitive institutions in the third stage of3SLS. The Wald test results are given in Panel C in Table 4.

Panel A shows that firms decrease their total leverage when pressure sensitive institutions holdmore shares, or when they held more shares in the last year. Similar results are found for pressureinsensitive institutions in Panel B. Most importantly, we find that the magnitude of coefficients forpressure insensitive institutions is larger than that for pressure sensitive institutions. For example,in the contemporaneous relationship, the coefficient of the change in institutional ownership forpressure insensitive institutions is −4.16, more than twice as large in absolute value as is the coefficientfor pressure sensitive institutions. Panel C shows that the magnitude differences in the coefficientsof interest between pressure sensitive and insensitive institutions are statistically significant at the1% level for all model specifications. These results imply that the substitution-monitoring effect ismore evident when institutional investors have more information and/or are more independent frommanagement.

4.1.3. Relationship between changes in institutional ownership and changes in leverage throughoutthe business cycle

The agency theory of firms hypothesizes that managers have incentives to maximize their ownbenefits by engaging in opportunistic behaviors. The incentives could be higher in the presence ofa wider information asymmetry between the manager and shareholders; in such a case, the firmshall face high agency costs. This implies that institutional monitoring could be more beneficial whenhigh information asymmetry and uncertainty surround the firm. To examine whether informationasymmetry and uncertainty levels impact institutional monitoring, we consider two sets of economictime periods: contractions and expansions. The literature shows that a firm’s information asymmetryusually increases during economic contractions. We use an interaction term (the contraction dummymultiplied by the change in ownership) to capture the incremental effect of institutional monitoringwhen the level of information asymmetry is high. The contraction dummy (Con D) is equal to 1 if thekey variables are obtained in economic contractions.

Table 5 presents the results. We are primarily interested in the coefficients of the interaction termswith the contraction dummy. We find that the interaction term for the change in institutional owner-ship is negative and statistically significant at the 1% level in both contemporaneous and lagged models,meaning that, in contractions, firms reduce their leverage as institutional ownership increases, or whenit had increased in the previous period. This result implies that institutional monitoring may inten-sify when the agency costs are high. This finding, together with the result for pressure insensitiveinstitutions, amplifies the evidence for the substitutability of institutional debt monitoring.

In addition, we find that the interaction terms for the change in total leverage are negative andstatistically significant at the 1% level, suggesting that during economic contractions, institutional

422

423

424

425

426

427

428

429

430

431

432

433

434

435

436

437

438

439

440

441

442

443

444

445

446

447

448

449

450

451

452

453

454

455

456

457

458

459

460

461

462

463

464

465

466

467

468

469

470

471

472

473

dx.doi.org/10.1016/j.najef.2014.10.001

Please cite

this

article in

press

as: C

hu

ng,

C.

Y.,

& W

ang,

K.

Do

institu

tional

investors

mon

itorm

anagem

ent?

Eviden

ce from

the

relationsh

ip betw

een in

stitution

al. .

. .

North

Am

erican Journal

ofEconom

ics and

Finance (2014),

http

://dx.d

oi.org/10.1016/j.najef.2014.10.001

AR

TIC

LE

IN P

RE

SS

G M

odelEC

OFIN

470 1–31

14

C.Y. Chung,

K.

Wang

/ N

orth A

merican

Journal of

Economics

and Finance

xxx (2014)

xxx–xxx

Table 4Relationships between changes in institutional ownership and changes in leverage for pressure sensitive and insensitive institutions.



Panel A: pressure sensitive institutionsIntercept 0.046 (0.000)*** −0.017 (0.000)*** Intercept 0.013 (0.000)*** 0.010 (0.000)***

�Ownership −1.756 (0.000)*** −0.782 (0.070)* �Leverage −0.005 (0.602) 0.051 (0.001)***

Asset beta −0.016 (0.000)*** 0.022 (0.000)*** Age −0.003 (0.000)*** −0.003 (0.000)***

EBITD −0.101 (0.000)*** 0.037 (0.000)*** beta −0.001 (0.423) −0.001 (0.046)**

M/B −0.010 (0.000)*** 0.004 (0.000)*** Dividend yield −0.372 (0.006)*** −0.311 (0.011)**

PPE −0.004 (0.074)* −0.003 (0.225) Firm-specific risk −0.055 (0.000)*** 0.016 (0.042)**

R&D −0.001 (0898) −0.002 (0.271) Lag return 0.019 (0.000)*** 0.004 (0.000)***

R&D D −0.003 (0.001)*** 0.007 (0.000)*** Price 0.001 (0.396) −0.001 (0.060)*

SE 0.001 (0.191) 0.001 (0.186) Standard deviation 0.014 (0.002)*** 0.011 (0.009)***

Size S 0.001 (0.000)*** −0.001 (0.496) Size E −0.001 (0.743) 0.001 (0.027)**

Turnover −0.001 (0.031)** −0.004 (0.000)***

Adjusted-R2 0.070 0.019 0.013 0.003N 40,963 40,456 40,963 40,456



Panel B: pressure insensitive institutionsIntercept 0.048 (0.000)*** −0.017 (0.000)*** Intercept 0.028 (0.000)*** 0.010 (0.064)*

�Ownership −4.156 (0.000)*** −2.013 (0.004)*** �Leverage −0.032 (0.049)** 0.193 (0.000)***

Asset beta −0.017 (0.000)*** 0.022 (0.000)*** Age −0.003 (0.000)*** −0.002 (0.018)**

EBITD −0.099 (0.000)*** 0.037 (0.000)*** Beta −0.002 (0.028)** −0.004 (0.000)***

M/B −0.010 (0.000)*** 0.004 (0.000)*** Dividend yield −0.851 (0.001)*** −0.155 (0.517)PPE −0.006 (0.000)*** −0.003 (0.213) Firm-specific risk −0.225 (0.000)*** 0.018 (0.260)R&D −0.005 (0.779) −0.002 (0.258) Lag return 0.050 (0.000)*** 0.002 (0.076)*

R&D D −0.003 (0.005)*** 0.007 (0.000)*** Price −0.001 (0.275) −0.007 (0.000)***

SE −0.001 (0.196) 0.001 (0.182) Standard deviation 0.029 (0.001)*** 0.029 (0.001)***

Size S 0.002 (0.000)*** −0.001 (0.483) Size E 0.001 (0.474) 0.003 (0.000)***

Turnover −0.006 (0.000)*** −0.014 (0.000)***

Adjusted-R2 0.075 0.018 0.029 0.004N 40,963 40,456 40,963 40,456

dx.doi.org/10.1016/j.najef.2014.10.001

Please cite

this

article in

press

as: C

hu

ng,

C.

Y.,

& W

ang,

K.

Do

institu

tional

investors

mon

itorm

anagem

ent?

Eviden

ce from

the

relationsh

ip betw

een in

stitution

al. .

. .

North

Am

erican Journal

ofEconom

ics and

Finance (2014),

http

://dx.d

oi.org/10.1016/j.najef.2014.10.001

AR

TIC

LE

IN P

RE

SS

G M

odelEC

OFIN

470 1–31

C.Y. Chung,

K.

Wang

/ N

orth A

merican

Journal of

Economics

and Finance

xxx (2014)

xxx–xxx

15

Panel C: difference in coefficients for key variables

Wald test Contemporaneous Lagged

�Ownership in PSI = �Ownership in PII 457.22 (0.000)*** 83.73 (0.000)***

�Leverage in PSI = �leverage in PII 38.64 (0.000)*** 9.77 (0.002)***

This table reports the coefficient estimates for both the contemporaneous and lagged relationships between changes in institutional ownership of pressure sensitive and insensitiveinstitutions and the changes in total leverage from 1985 to 2008. Panel A and Panel B provide the results for pressure sensitive institutions (bank trust department and insurancecompanies) and for pressure insensitive institutions (mutual funds and independent investment advisors), respectively. Panel C reports the Wald test statistic of the difference in keycoefficients between pressure sensitive and insensitive institutions. Leverage is total debt divided by total debt plus the market value of equity, and �leverage is the change in leverageover the year. �Ownership is the change in institutional ownership over the year. EBITD is earnings before interest, taxes, and depreciation divided by assets. M/B is market value of assetdivided by book value of asset. PPE is net property, plants, and equipment divided by assets. R&D is research and development expenses divided by sales. R&D D is equal to 1 if a firm doesnot report the research and development expenses. SE is selling expenses divided by sales. Size E and Size S are the natural logarithm of market value of equity and sales, respectively.Beta is from the rolling regressions based on Fama and French (1992). Standard deviation is the natural logarithm of monthly standard deviation of returns over the previous 24 to 60months. Asset beta is the weighted average beta of equity and debt. Firm-specific risk is the natural logarithm of 1 plus the average daily squared firm-specific returns over a month, andthe firm-specific return is the difference between a firm’s daily return and its industry return. Age is the natural logarithm of the number of months for which the firm has CRSP returndata. Dividend yield is the natural logarithm of 1 plus the average monthly dividend yield over the previous 12 months. Price is the natural logarithm of 1 plus the month-end share price.Turnover is the natural logarithm of 1 plus the ratio of monthly volume to the number of shares outstanding. Lag Return is the buy-and-hold return over the previous six months. Thoughnot reported, an industry dummy based on the Fama-French 49 industry classifications and a year dummy are also included in the model. p-Values are reported in parentheses.***, **, and * indicate statistical significance at the 1%, 5%, and 10% levels, respectively.

dx.doi.org/10.1016/j.najef.2014.10.001




Table 5Relationship between changes in institutional ownership and changes in leverage throughout the business cycle.



Intercept 0.051 (0.000)*** −0.015 (0.000)*** Intercept 0.036 (0.000)*** −0.001 (0.985)�Ownership −1.549 (0.000)*** −0.781 (0.000)*** �Leverage −0.013 (0.548) 0.362 (0.000)***

Con D 0.038 (0.000)*** 0.085 (0.000)*** Con D 0.001 (0.729) −0.044 (0.000)***

�Ownership × Con D −0.684 (0.000)*** −0.961 (0.000)*** �Leverage× Con D

−0.346 (0.000)*** −0.428 (0.005)***

Asset beta −0.012 (0.000)*** 0.021 (0.000)*** Age −0.005 (0.000)*** −0.002 (0.182)EBITD −0.019 (0.000)*** 0.050 (0.000)*** Beta −0.001 (0.348) −0.005 (0.000)***

M/B −0.004 (0.000)*** 0.007 (0.000)*** Dividendyield

−1.688 (0.000)*** −1.219 (0.001)***

PPE −0.003 (0.430) −0.004 (0.249) Firm-specificrisk

−0.278 (0.000)*** 0.036 (0.118)

R&D −0.004 (0.032)** −0.003 (0.518) Lag return 0.075 (0.000)*** 0.006 (0.000)***

R&D D 0.004 (0.001)** 0.005 (0.000)*** Price −0.001 (0.250) −0.012 (0.000)***

SE −0.001 (0.868) 0.001 (0.885) Standarddeviation

0.024 (0.046)** 0.027 (0.034)**

Size S −0.001 (0.000)*** −0.003 (0.000)*** Size E 0.001 (0.061)* 0.005 (0.000)***

Turnover −0.009 (0.000)*** −0.024 (0.000)***

Adjusted-R2 0.156 0.044 0.033 0.011N 40,963 40,456 40,963 40,456

This table reports the coefficient estimates for both the contemporaneous and lagged relationships between changes in institu-tional ownership and changes in total leverage from 1985 to 2008. Leverage is total debt divided by total debt plus the marketvalue of equity, and �leverage is the change in leverage over the year. �Ownership is the change in institutional ownershipover the year. EBITD is earnings before interest, taxes, and depreciation divided by assets. M/B is market value of asset dividedby book value of asset. PPE is net property, plants, and equipment divided by assets. R&D is research and development expensesdivided by sales. R&D D is equal to 1 if a firm does not report its research and development expenses. SE is selling expensesdivided by sales. Size E and Size S are the natural logarithm of the market value of equity and sales, respectively. Beta is fromthe rolling regressions based on Fama and French (1992). Standard deviation is the natural logarithm of the monthly standarddeviation of returns over the previous 24 to 60 months. Asset beta is the weighted average beta of equity and debt. Firm-specificrisk is the natural logarithm of 1 plus the average daily squared firm-specific returns over a month, and the firm-specific returnis the difference between a firm’s daily return and its industry return. Age is the natural logarithm of the number of months forwhich the firm has CRSP return data. Dividend yield is the natural logarithm of 1 plus the average monthly dividend yield overthe previous 12 months. Price is the natural logarithm of 1 plus the month-end share price. Turnover is the natural logarithmof 1 plus the ratio of the monthly volume to number of shares outstanding. Lag Return is the buy-and-hold return over theprevious six months. Con D is equal to 1 if �leverage or �Ownership is obtained during economic contractions. The economiccontraction and expansion periods are taken from the National Bureau of Economic Research (NBER) business cycle. Thoughnot reported, an industry dummy based on the Fama-French 49 industry classifications and a year dummy are also included inthe model. p-Values are reported in parentheses.***, **, and * indicate statistical significance at the 1%, 5%, and 10% levels, respectively.

investors sell their shares as total leverage increases, or when it had increased in the last year. Thisresult is intuitive because an increase in total leverage during contractions implies that a firm mayexperience more economic or financial distress. Institutional investors, who are perceived to be moresophisticated and prudent than are individual investors, may view this change as a bad signal and thusliquidate their position.

In this section, we find evidence that institutional investors may substitute for debt to monitor firmmanagement. However, this finding does not indicate whether institutional monitoring improves afirm’s capital structure. In the following sections, we present direct evidence of the role of institutionalmonitoring in capital structure management using the deviated or value-decreasing leverage.

4.2. Direct evidence of institutional monitoring: Active and passive institutional monitoring on thesuboptimal capital structure

4.2.1. Relationship between changes in institutional ownership and changes in leverage deviationWe examine institutional investors’ influence on a firm’s capital structure by focusing on two

relationships. First, we examine whether institutional investors change a firm’s suboptimal capital

474

475

476

477

478

479

480

481

482

483

484

485

486

487

dx.doi.org/10.1016/j.najef.2014.10.001




Table 6Relationship between changes in institutional ownership and changes in leverage deviation.


Dep. variable �Leverage deviation �Ownership

Intercept 0.118 (0.000)*** 0.125 (0.000)*** Intercept 0.092 (0.000)*** 0.100 (0.000)***

�Ownership −0.102 (0.000)*** −0.653 (0.000)*** �Leveragedeviation

−0.383 (0.000)*** −0.498 (0.000)***

Asset beta −0.002 (0.025)** −0.001 (0.546) Age −0.009 (0.000)*** −0.009 (0.000)***

EBITD −0.071 (0.000)*** −0.046 (0.000)*** Beta −0.001 (0.258) −0.002 (0.079)*

M/B −0.002 (0.000)*** −0.002 (0.000)*** Dividendyield

−1.722 (0.000)*** −2.124 (0.000)***

PPE −0.011 (0.000)*** −0.007 (0.006)*** Firm-specificrisk

−0.142 (0.000)*** 0.033 (0.142)

R&D −0.002 (0.059)* 0.002 (0.453) Lag return 0.088 (0.000)*** 0.012 (0.000)***

R&D D 0.004 (0.000)*** 0.005 (0.000)*** Price −0.006 (0.000)*** −0.012 (0.000)***

SE 0.001 (0.546) −0.002 (0.036)** Standarddeviation

0.032 (0.008)*** 0.034 (0.009)***

Size S −0.004 (0.000)*** −0.006 (0.000)*** Size E 0.001 (0.465) 0.002 (0.000)***

Turnover −0.003 (0.111) −0.008 (0.002)***

Adjusted-R2 0.045 0.044 0.032 0.006N 40,963 40,456 40,963 40,456

This table reports the coefficient estimates for both the contemporaneous and lagged relationships between changes in insti-tutional ownership and changes in leverage deviation from 1985 to 2008. Leverage deviation is the absolute value of totalleverage minus target leverage. �Leverage deviation is the change in leverage deviation over the year. �Ownership is thechange in institutional ownership over the year. EBITD is earnings before interest, taxes, and depreciation divided by assets.M/B is market value of asset divided by book value of asset. PPE is net property, plants, and equipment divided by assets. R&D isresearch and development expenses divided by sales. R&D D is equal to 1 if a firm does not report its research and developmentexpenses. SE is selling expenses divided by sales. Size E and Size S are the natural logarithm of the market value of equity andsales, respectively. Beta is from the rolling regressions based on Fama and French (1992). Standard deviation is the naturallogarithm of monthly standard deviation of returns over the previous 24 to 60 months. Asset beta is the weighted average betaof equity and debt. Firm-specific risk is the natural logarithm of 1 plus the average daily squared firm-specific returns over amonth, and the firm-specific return is the difference between a firm’s daily return and its industry return. Age is the naturallogarithm of the number of months for which the firm has CRSP return data. Dividend yield is the natural logarithm of 1 plusthe average monthly dividend yield over the previous 12 months. Price is the natural logarithm of 1 plus the month-end shareprice. Turnover is the natural logarithm of 1 plus the ratio of monthly volume to number of shares outstanding. Lag Return isthe buy-and-hold return over the previous six months. Though not reported, an industry dummy based on the Fama-French 49industry classifications and a year dummy are also included in the model. p-Values are reported in parentheses.***, **, and * indicate statistical significance at the 1%, 5%, and 10% levels, respectively.

structure. Second, we examine how institutional investors respond when a firm has a suboptimalcapital structure. The trade-off theory of capital structure suggests that a firm’s target leverage maybe determined by debt-related costs and the tax benefits of debt (Modigliani & Miller, 1963; Jensen &Meckling, 1976; Myers, 1977). Kayhan and Titman (2007) and Harford et al. (2009) present empiricalevidence of the existence of target leverage and show that a firm’s debt tends to revert back to the targetlevel when it deviates. This implies that leverage deviation may decrease firm value, giving institutionalinvestors a strong incentive to respond. Specifically, institutional investors may reduce the level ofdeviation or increase their ownership in firms with a high level of deviation for subsequent monitoring.Alternatively, they may sell their shares when they are dissatisfied with the level of deviation if thecosts of monitoring exceed the benefits.

Table 6 reports the coefficient estimates for both the contemporaneous and lagged relationshipsbetween changes in institutional ownership and changes in leverage deviation. We find that leveragedeviation as a proxy for suboptimal (value-decreasing) capital structure decreases when institutionalownership increases, or when it had increased in the last year. This result implies that institutionalinvestors actively improve a firm’s debt policy to increase firm value. In addition, we find that insti-tutional ownership decreases when leverage deviation increases, or when it had increased in the lastyear, suggesting that institutional investors sell their shares (“vote with their feet”) when dissatisfiedwith a firm’s capital structure. This result contradicts our earlier conjecture that institutional investorsmay stay with the firm and benefit from subsequent monitoring efforts.

488

489

490

491

492

493

494

495

496

497

498

499

500

501

502

503

504

505

506

dx.doi.org/10.1016/j.najef.2014.10.001




4.2.2. Relationship between changes in institutional ownership and changes in leverage deviation forpressure sensitive and insensitive institutions

Because pressure insensitive institutions appear to be more informed and have fewer incentivesto go along with management decisions, we expect institutional monitoring to be more effectivein these institutions. To examine this matter, we replicate Table 4 by replacing total leverage withleverage deviation and report the results in Table 7 . We find that for both pressure sensitive andinsensitive institutions, the leverage deviation decreases when institutional investors had increasedtheir ownership during the last year. However, the coefficient for pressure insensitive institutionsis more negative than its counterpart for pressure sensitive institutions. For example, in the laggedrelationship, the coefficient for pressure insensitive institutions is −0.99, compared to a correspondingcoefficient of −0.002 for pressure sensitive institutions. As for the contemporaneous relationship, thecoefficient for pressure sensitive institutions is positive but statistically insignificant, whereas a similarcoefficient for pressure insensitive institutions is negative and statistically significant at the 1% level.Panel C reports the Wald test statistics for comparing the magnitude of coefficients between pressureinsensitive institutions and pressure sensitive institutions. We find that all test statistics are positiveand statistically significant at the 5% level or better. These results suggest that a firm’s suboptimalcapital structure is improved when institutional investors are more informed and/or are independentfrom management. Moreover, a firm’s suboptimal capital structure worsens when monitoring byinstitutional investors is restrictive and thus ineffective.

For the effect of changes in leverage deviation on changes in institutional ownership, we findnegative coefficients for both types of institutions and in both contemporaneous and lagged rela-tionships. However, the coefficients are larger for pressure insensitive institutions, suggesting thatinstitutional investors who are more independent from management might be able to sell their sharesmore easily when dissatisfied with management (as no strings are attached). Overall, the results in thissubsection not only show different monitoring intensities among heterogeneous institutions but alsoconfirm that the previous section’s results are driven by active and passive institutional monitoringinfluences.

4.2.3. Other suboptimal leverage characteristics and changes in institutional ownershipOur earlier results suggest that institutional investors simply liquidate their shares when dis-

satisfied with a firm’s capital structure. Such behavior is more evident among pressure insensitiveinstitutions, which are more independent from management and thus have fewer restrictions on theirshare trading. In this subsection, we consider other suboptimal leverage characteristics (besides lever-age deviation) and examine how they affect the changes in institutional ownership. Specifically, westudy how the variances of leverage and leverage deviation over the past 3 years affect the changes ininstitutional ownership. We use the two variance measures to proxy for a firm’s unstable or uncertaincapital structure. If these suboptimal leverage characteristics are important to institutional investors,they should have a negative impact on institutional ownership. In addition, we expect that the neg-ative relationship is more evident in pressure insensitive institutions, as they can more freely tradetheir shares if dissatisfied with the firm’s capital structure.

Table 8 presents the coefficients from the ordinary least square (OLS) model for the relation-ship between changes in institutional ownership and the two variance measures. We include thelagged variables used in the previous tables to help explain the changes in institutional ownership.Because the variance measures may also capture a firm’s financial risk, we include the lagged totalleverage to control for that effect. Table 8 suggests several interesting findings. First, we find a neg-ative relationship between firm leverage and the change in institutional ownership, indicating thatinstitutional investors prefer firms with low financial risk. In addition, we find that the change inaggregate institutional ownership is negatively related to both variance measures, suggesting thatinstitutional investors sell their shares when a firm’s leverage is volatile and uncertain over time.Lastly, we find that the negative relationships are more evident in pressure insensitive institutions.However, this result could also be caused by the informational advantage of pressure insensitiveinstitutions. In descriptive statistics, we learn that most pressure insensitive institutions are indepen-dent investment advisors, who are likely to be more informed than other institutions, and who maythus execute their shares more quickly in the presence of bad management. To address the concern

507

508

509

510

511

512

513

514

515

516

517

518

519

520

521

522

523

524

525

526

527

528

529

530

531

532

533

534

535

536

537

538

539

540

541

542

543

544

545

546

547

548

549

550

551

552

553

554

555

556

557

558

559

dx.doi.org/10.1016/j.najef.2014.10.001

Please cite

this

article in

press

as: C

hu

ng,

C.

Y.,

& W

ang,

K.

Do

institu

tional

investors

mon

itorm

anagem

ent?

Eviden

ce from

the

relationsh

ip betw

een in

stitution

al. .

. .

North

Am

erican Journal

ofEconom

ics and

Finance (2014),

http

://dx.d

oi.org/10.1016/j.najef.2014.10.001

AR

TIC

LE

IN P

RE

SS

G M

odelEC

OFIN

470 1–31

C.Y. Chung,

K.

Wang

/ N

orth A

merican

Journal of

Economics

and Finance

xxx (2014)

xxx–xxx

19

Table 7Relationships between changes in institutional ownership and changes in leverage deviation for pressure sensitive and insensitive institutions.



Panel A: pressure sensitive institutionsIntercept 0.012 (0.000)*** 0.006 (0.001)*** Intercept 0.013 (0.000)*** 0.011 (0.000)***

�Ownership 0.111 (0.117) −0.002 (0.053)* �Leverage deviation −0.043 (0.033)** −0.057 (0.015)**

Asset beta −0.014 (0.001)*** 0.008 (0.000)*** Age −0.003 (0.000)*** −0.003 (0.000)***

EBITD −0.028 (0.000)*** −0.009 (0.017)** Beta −0.001 (0.498) −0.001 (0.456)M/B 0.001 (0.752) −0.006 (0.000)*** Dividend yield −0.358 (0.008)*** −0.365 (0.002)***

PPE −0.004 (0.065)* −0.002 (0.432) Firm-specific risk −0.056 (0.000)*** 0.004 (0.580)R&D −0.001 (0.762) −0.003 (0.070)* Lag return 0.019 (0.000)*** 0.004 (0.000)***

R&D D −0.002 (0.092)* 0.001 (0.912) Price 0.001 (0.581) −0.001 (0.454)SE 0.001 (0.194) −0.001 (0.149) Standard deviation 0.011 (0.014)** 0.013 (0.027)**

Size S 0.001 (0.000)*** 0.001 (0.009)*** Size E 0.001 (0.573) 0.001 (0.206)Turnover 0.001 (0.001)*** −0.003 (0.000)***

Adjusted-R2 0.009 0.011 0.014 0.003N 40,963 40,456 40,963 40,456



Panel B: pressure insensitive institutionsIntercept 0.012 (0.000)*** 0.006 (0.001)*** Intercept 0.029 (0.000)*** 0.012 (0.012)**

�Ownership −0.438 (0.000)*** −0.988 (0.001)*** �Leverage deviation −0.289 (0.000)*** −0.279 (0.001)***

Asset beta −0.014 (0.000)*** −0.008 (0.000)*** Age −0.005 (0.015)** −0.003 (0.002)***

EBITD −0.027 (0.000)*** −0.009 (0.017)** Beta −0.002 (0.024)** −0.002 (0.013)**

M/B 0.001 (0.748) −0.006 (0.000)*** Dividend yield −0.822 (0.002)*** −0.280 (0.244)PPE −0.004 (0.088)* −0.002 (0.432) Firm-specific risk −0.230 (0.000)*** −0.006 (0.666)R&D −0.005 (0.760) −0.003 (0.070)* Lag return 0.056 (0.000)*** 0.003 (0.030)**

R&D D −0.002 (0.091)* 0.001 (0.912) Price −0.002 (0.034)** −0.005 (0.000)***

SE 0.001 (0.194) −0.001 (0149) Standard deviation 0.021 (0.025)** 0.028 (0.002)***

Size S 0.001 (0.000)*** 0.001 (0.009)*** Size E 0.001 (0.680) 0.002 (0.000)***

Turnover −0.007 (0.000)*** −0.012 (0.000)***

Adjusted-R2 0.009 0.011 0.029 0.004N 40,963 40,456 40,963 40,456

dx.doi.org/10.1016/j.najef.2014.10.001

Please cite

this

article in

press

as: C

hu

ng,

C.

Y.,

& W

ang,

K.

Do

institu

tional

investors

mon

itorm

anagem

ent?

Eviden

ce from

the

relationsh

ip betw

een in

stitution

al. .

. .

North

Am

erican Journal

ofEconom

ics and

Finance (2014),

http

://dx.d

oi.org/10.1016/j.najef.2014.10.001

AR

TIC

LE

IN P

RE

SS

G M

odelEC

OFIN

470 1–31

20

C.Y. Chung,

K.

Wang

/ N

orth A

merican

Journal of

Economics

and Finance

xxx (2014)

xxx–xxx

Table 7(Continued )



�Ownership in PSI = �ownership in PII 9.43 (0.002)*** 13.23 (0.001)***

�Leverage deviation in PSI = �Leverage deviation in PII 8.94 (0.003)*** 5.52 (0.018)**

This table reports the coefficient estimates for both the contemporaneous and lagged relationships between changes in institutional ownership for pressure sensitive and insensitiveinstitutions and changes in leverage deviation from 1985 to 2008. Panel A and Panel B provide the results for pressure sensitive institutions (bank trust department and insurancecompanies) and for pressure insensitive institutions (mutual funds and independent investment advisors), respectively. Panel C reports the Wald test statistic of the difference in keycoefficients between pressure sensitive and insensitive institutions. Leverage deviation is the absolute value of total leverage minus target leverage. �Leverage deviation is the change inleverage deviation over the year. �Ownership is the change in institutional ownership over the year. EBITD is earnings before interest, taxes, and depreciation divided by assets. M/B ismarket value of asset divided by book value of asset. PPE is net property, plants, and equipment divided by assets. R&D is research and development expenses divided by sales. R&D D isequal to 1 if a firm does not report its research and development expenses. SE is selling expenses divided by sales. Size E and Size S are the natural logarithm of the market value of equityand sales, respectively. Beta is from the rolling regressions based on Fama and French (1992). Standard deviation is the natural logarithm of the monthly standard deviation of returnsover the previous 24 to 60 months. Asset beta is the weighted average beta of equity and debt. Firm-specific risk is the natural logarithm of 1 plus the average daily squared firm-specificreturns over a month, and the firm-specific return is the difference between a firm’s daily return and its industry return. Age is the natural logarithm of the number of months for whichthe firm has CRSP return data. Dividend yield is the natural logarithm of 1 plus the average monthly dividend yield over the previous 12 months. Price is the natural logarithm of 1 plusthe month-end share price. Turnover is the natural logarithm of 1 plus the ratio of monthly volume to the number of shares outstanding. Lag Return is the buy-and-hold return overthe previous six months. Though not reported, an industry dummy based on the Fama-French 49 industry classifications and a year dummy are also included in the model. p-Values arereported in parentheses.***, **, and * indicate statistical significance at the 1%, 5%, and 10% levels, respectively.

dx.doi.org/10.1016/j.najef.2014.10.001




Table 8Relationship between changes in institutional ownership and other suboptimal leverage characteristics.

Dep. variable: �ownership Aggregateinstitutions

Pressure sensitiveinstitutions

Pressure insensitiveinstitutions

Panel A: variance of leverage (Variance A)Intercept 0.059 (0.000)*** 0.018 (0.000)*** 0.037 (0.000)***

Variance A −0.143 (0.002)*** −0.023 (0.211) −0.093 (0.011)**

Leverage −0.053 −0.012 −0.038(0.000)*** (0.000)*** (0.000)***

Age −0.006 (0.000)*** −0.002 (0.000)*** −0.003 (0.000)***

Beta −0.003 (0.008)*** −0.001 (0.423) −0.002 (0.006)***

Dividend yield −0.666 (0.032)** −0.291 (0.017)** −0.090 (0.709)Firm-specific risk 0.004 (0.800) 0.012 (0.089)* 0.006 (0.677)Lag return 0.007 (0.000)*** 0.003 (0.000)*** 0.001 (0.000)***

Price −0.011 (0.000)*** −0.001 (0.014)** −0.008 (0.000)***

Standard deviation 0.007 (0.526) 0.006 (0.196) 0.011 (0.209)Size E 0.003 (0.000)*** 0.001 (0.257) 0.002 (0.000)***

Turnover −0.015 (0.000)*** −0.003 (0.000)*** −0.011 (0.000)***

Adjusted-R2 0.012 0.005 0.010N 40,456 40,456 40,456

Dep. variable: �ownership Aggregateinstitutions

Pressure sensitiveinstitutions

Pressure insensitiveinstitutions

Panel B: variance of leverage deviation (Variance B)Intercept 0.059 (0.000)*** 0.018 (0.000)*** 0.038 (0.000)***

Variance B −0.060 (0.019)** −0.001 (0.954) −0.073 (0.044)**

Leverage −0.055 −0.012 −0.039(0.000)*** (0.000)*** (0.000)***

Age −0.006 (0.000)*** −0.002 (0.000)*** −0.003 (0.000)***

Beta −0.003 (0.008)*** −0.001 (0.422) −0.002 (0.006)***

Dividend yield −0.662 (0.032)** −0.290 (0.017)** −0.087 (0.719)Firm-specific risk 0.004 (0.830) 0.012 (0.098)* 0.006 (0.657)Lag return 0.006 (0.000)*** 0.003 (0.000)*** 0.001 (0.436)Price −0.010 (0.000)*** −0.001 (0.015)** −0.008 (0.001)***

Standard deviation 0.007 (0.551) 0.005 (0.209) 0.011 (0.205)Size E 0.003 (0.000)*** 0.001 (0.241) 0.002 (0.000)***

Turnover −0.015 (0.000)*** −0.003 (0.000)*** −0.011 (0.000)***

Adjusted-R2 0.013 0.005 0.010N 40,456 40,456 40,456

This table reports the coefficients from the ordinary least square (OLS) model for the relationship between changes in insti-tutional ownership and other leverage characteristics from 1985 to 2008. The variance of leverage (Variance A) and varianceof leverage deviation (Variance B) over the past 3 years are used as other suboptimal leverage characteristics in Panel A andPanel B, respectively. Pressure sensitive institutions include bank trust department and insurance companies, and pressureinsensitive institutions include mutual funds and independent investment advisors. �Ownership is the change in institutionalownership over the year. Size E is the natural logarithm of the market value of equity. Beta is from the rolling regressions basedon Fama and French (1992). Standard deviation is the natural logarithm of the monthly standard deviation of returns over theprevious 24 to 60 months. Firm-specific risk is the natural logarithm of 1 plus the average daily squared firm-specific returnsover a month, and the firm-specific return is the difference between a firm’s daily return and its industry return. Leverage istotal debt divided by total debt plus the market value of equity. Age is the natural logarithm of the number of months for whichthe firm has CRSP return data. Dividend yield is the natural logarithm of 1 plus the average monthly dividend yield over theprevious 12 months. Price is the natural logarithm of 1 plus the month-end share price. Turnover is the natural logarithm of 1plus the ratio of monthly volume to number of shares outstanding. Lag Return is the buy-and-hold return over the previous sixmonths. Though not reported, an industry dummy based on the Fama-French 49 industry classifications and a year dummy arealso included in the model. p-Values are reported in parentheses.***, **, and * indicate statistical significance at the 1%, 5%, and 10% levels, respectively.

that a 3-year window is too short to capture a firm’s unstable debt policy, we conduct a robustnessanalysis by calculating the two variance measures over the past 5 or 10 years. In unreported tables,we find the results remain unchanged. Overall, the results in Table 8 supplement our earlier find-ings that the leverage-related firm characteristics are important to explain institutional investors’preference.

560

561

562

563

dx.doi.org/10.1016/j.najef.2014.10.001




Table 9Relationship between changes in institutional ownership and changes in leverage deviation throughout the business cycle.



Intercept 0.118 (0.000)*** 0.124 (0.000)*** Intercept 0.078 (0.000)*** 0.085 (0.000)***

�Ownership −0.047 (0.017)** −0.579 (0.000)*** �Leveragedeviation

−0.315 (0.000)*** −0.463 (0.000)***

Con D 0.007 (0.000)*** 0.021 (0.000)*** Con D 0.045 (0.000)*** 0.054 (0.002)***

�Ownership × Con D −0.569 (0.000)*** −0.191 (0.000)*** �Leveragedeviation × ConD

−0.678 (0.000)*** −1.390 (0.000)***

Asset beta −0.002 (0.007)*** −0.001 (0.039)** Age −0.007 (0.000)*** −0.006 (0.000)***

EBITD −0.071 (0.000)*** −0.048 (0.000)*** Beta −0.001 (0.776) −0.001 (0.417)M/B −0.002 (0.000)*** −0.002 (0.000)*** Dividend yield −1.681 (0.000)*** −2.136 (0.000)***

PPE −0.010 (0.000)*** −0.006 (0.018)** Firm-specificrisk

−0.162 (0.000)*** 0.020 (0.371)

R&D −0.002 (0.059)* 0.003 (0.421) Lag return 0.086 (0.000)*** 0.010 (0.000)***

R&D D 0.003 (0.000)*** 0.005 (0.000)*** Price −0.006 (0.000)*** −0.011 (0.000)***

SE 0.001 (0.547) −0.003 (0.033)** Standarddeviation

0.027 (0.025)** 0.023 (0.068)**

Size S −0.004 (0.000)*** −0.006 (0.000)*** Size E 0.001 (0.404) 0.002 (0.000)***

Turnover −0.002 (0.173) −0.004 (0.057)*

Adjusted-R2 0.047 0.046 0.034 0.012N 40,963 40,456 40,963 40,456

This table reports the coefficient estimates for both the contemporaneous and lagged relationships between changes in institu-tional ownership and changes in leverage deviation from 1985 to 2008. Leverage deviation is the absolute value of total leverageminus target leverage. �Leverage deviation is the change in leverage deviation over the year. �Ownership is the change ininstitutional ownership over the year. EBITD is earnings before interest, taxes, and depreciation divided by assets. M/B is marketvalue of asset divided by book value of asset. PPE is net property, plants, and equipment divided by assets. R&D is research anddevelopment expenses divided by sales. R&D D is equal to 1 if a firm does not report its research and development expenses. SEis selling expenses divided by sales. Size E and Size S are the natural logarithm of the market value of equity and sales, respec-tively. Beta is from the rolling regressions based on Fama and French (1992). Standard deviation is the natural logarithm of themonthly standard deviation of returns over the previous 24 to 60 months. Asset beta is the weighted average beta of equity anddebt. Firm-specific risk is the natural logarithm of 1 plus the average daily squared firm-specific returns over a month, and thefirm-specific return is the difference between a firm’s daily return and its industry return. Age is the natural logarithm of thenumber of months for which the firm has CRSP return data. Dividend yield is the natural logarithm of 1 plus the average monthlydividend yield over the previous 12 months. Price is the natural logarithm of 1 plus the month-end share price. Turnover isthe natural logarithm of 1 plus the ratio of monthly volume to number of shares outstanding. Lag Return is the buy-and-holdreturn over the previous six months. Con D is equal to 1 if �Leverage or �ownership is obtained during economic contractions.The economic contraction and expansion periods are taken from the National Bureau of Economic Research (NBER) businesscycle. Though not reported, an industry dummy based on the Fama-French 49 industry classifications and a year dummy arealso included in the model. p-Values are reported in parentheses.***, **, and * indicate statistical significance at the 1%, 5%, and 10% levels, respectively.

4.2.4. Relationship between changes in institutional ownership and changes in leverage deviationthroughout the business cycle

Table 5 shows that the substitution effect of institutional monitoring is more evident during eco-nomic contractions. In this section, we examine whether institutional monitoring is also more effectivein these conditions. We measure the effectiveness of institutional monitoring using leverage devia-tion. Specifically, we estimate the 3SLS model with the interaction terms of the contraction dummywith the changes in institutional ownership and the changes in leverage deviation, after controllingfor other explanatory variables.

Table 9 reports the coefficient estimates. We find that, for both contemporaneous and lagged mod-els, the interaction terms are negative and statistically significant at the 1% level. These results havetwo important implications. First, they suggest that institutional investors improve a firm’s subopti-mal capital structure in contractions, which provides direct evidence that institutional monitoring ismore effective when firms face high agency costs. In addition, they imply that institutional investorssell more shares when the leverage deviation is large in contractions. This result may be caused by

564

565

566

567

568

569

570

571

572

573

574

575

576

577

dx.doi.org/10.1016/j.najef.2014.10.001




the high adjustment costs for capital structure during contractions, when financing costs are usuallyhigh.

Our findings thus far suggest that institutional investors engage in both active and passive monitor-ing and that the monitoring is more evident under two distinct conditions: among pressure insensitiveinstitutions and during economic contractions. Next, we combine these two conditions and examinewhether the time-varying (active and passive) institutional monitoring is more pronounced in pres-sure insensitive institutions during economic contractions. Specifically, we estimate the 3SLS modelwith the interaction terms of the contraction dummy with the changes in institutional ownershipand the changes in leverage deviation, after controlling for other explanatory variables. We split theaggregate institutional ownership into pressure sensitive and pressure insensitive institutions, andsimultaneously estimate the models to compare the results. Table 10 shows that except for one case,all interaction terms are negative. The sole exception lies in the interaction term with the changes ininstitutional ownership for pressure sensitive institutions. Given that pressure sensitive institutionstend to have little or no incentives to monitor a firm’s capital structure, the positive coefficient maynot capture the information related to institutional monitoring. Thus, the observed positive coeffi-cient does not necessarily contradict to our earlier finding but is rather consistent with the hypothesisthat institutional monitoring should be more or only evident among pressure insensitive institutions.This can be further seen by the Wald test results in Panel C which show that pressure insensitiveinstitutions have more negative coefficients than pressure sensitive institutions, and the differencesin magnitude are statistically significant in all cases. These findings imply that institutional monitor-ing generally intensifies during contractions but such intensified monitoring is more pronounced inpressure insensitive institutions due to their independence from management.

4.2.5. Time-series relationships between changes in institutional ownership and changes in leveragedeviation

As shown in Fig. 1, institutional ownership of corporate shares has substantially increased over thepast decades, suggesting that institutional monitoring may also have become more effective. Giventhat pressure insensitive institutions have strong monitoring incentives, we should expect that theincrease in monitoring effectiveness is driven primarily by these institutions. To test this hypothesis,we split the sample into three sub-periods: 1980s, 1990s, and 2000s. For each sub-period, we repli-cate Table 7 and report the coefficient estimates in Table 11. In aggregate institutions, we find thatthe coefficients become more negative for the change in both institutional ownership and leveragedeviation as time progresses, suggesting that institutional monitoring becomes increasingly effectiveover time. In addition, we find that such monitoring activities are most evident in pressure insensitiveinstitutions during the 2000s. These results provide further evidence that institutional investors playan important role in firm’s capital structure management through monitoring, especially during thelast decade.

4.2.6. Is institutional monitoring related to firm performance?Our earlier results suggest that institutional investors may actively engage in monitoring activities.

The motivation for institutional monitoring is to increase firm value. If such monitoring is effective,we should expect a positive relationship between institutional monitoring and firm value. To test thishypothesis, we divide all firms in our sample into two groups: firms in which institutional ownershipincreases and leverage deviation decreases, and firms in which institutional ownership decreasesand leverage deviation increases. The first group represents an improved capital structure throughenhanced monitoring, whereas the second group mimics a scenario in which the capital structureworsens due to weakened institutional monitoring. To ensure our results are robust, we use threemeasures to proxy for firm value: return on assets (ROA), return on equity (ROE), and Tobin’s Q (Q)8.Because these measures may vary significantly across industries, we compute for each measure theindustry-adjusted value by subtracting the industry average value from the firm value. The industry

8 See Table B in Appendix A for the definition of each performance measure in details. In particular, we calculate thesophisticated Tobin’s Q based on Kaplan and Zingales (1997) and Gompers, Ishii, and Metrick (2003).

578

579

580

581

582

583

584

585

586

587

588

589

590

591

592

593

594

595

596

597

598

599

600

601

602

603

604

605

606

607

608

609

610

611

612

613

614

615

616

617

618

619

620

621

622

623

624

625

dx.doi.org/10.1016/j.najef.2014.10.001

Please cite

this

article in

press

as: C

hu

ng,

C.

Y.,

& W

ang,

K.

Do

institu

tional

investors

mon

itorm

anagem

ent?

Eviden

ce from

the

relationsh

ip betw

een in

stitution

al. .

. .

North

Am

erican Journal

ofEconom

ics and

Finance (2014),

http

://dx.d

oi.org/10.1016/j.najef.2014.10.001

AR

TIC

LE

IN P

RE

SS

G M

odelEC

OFIN

470 1–31

24

C.Y. Chung,

K.

Wang

/ N

orth A

merican

Journal of

Economics

and Finance

xxx (2014)

xxx–xxxTable 10Relationships between changes in institutional ownership and changes in leverage deviation for pressure sensitive and pressure insensitive institutions throughout the business cycle.



Panel A: pressure sensitive institutionsIntercept 0.003 (0.096)* 0.002 (0.325) Intercept 0.014 (0.000)*** 0.015 (0.000)***

�Ownership −0.119 (0.045)** −0.003 (0.882) �Leverage deviation −0.003 (0.905) −0.046 (0.075)*

Con D 0.008 (0.000)*** 0.008 (0.000)*** Con D 0.006 (0.037)** 0.003 (0.075)*

�Ownership × Con D −0.074 (0.000)*** 0.138 (0.001)*** �Leverage deviation × Con D −0.186 (0.034)** −0.110 (0.063)*

Asset beta −0.012 (0.000)*** 0.007 (0.000)** Age −0.003 (0.000)*** −0.002 (0.000)***

EBITD −0.019 (0.000)*** −0.007 (0.078)* Beta 0.001 (0.882) 0.001 (0.878)M/B 0.007 (0.038)** −0.006 (0.000)*** Dividend yield −0.323 (0.023)** −0.304 (0.014)**

PPE −0.001 (0.569) 0.001 (0.882) Firm-specific risk −0.052 (0.000)*** −0.002 (0.797)R& D −0.001 (0.282) −0.002 (0.262) Lag return 0.019 (0.000)*** 0.004 (0.000)***

R&D D −0.001 (0.328) 0.001 (0.891) Price 0.001 (0.695) −0.001 (0.613)SE 0.001 (0.391) −0.001 (0.111) Standard deviation 0.014 (0.007)*** 0.010 (0.029)**

Size S 0.001 (0.000)*** 0.001 (0.019)** Size E −0.001 (0.439) −0.001 (0.722)Turnover −0.001 (0.016)** −0.002 (0.000)***

Adjusted-R2 0.017 0.014 0.015 0.005N 40,963 40,456 40,963 40,456



Panel B: Pressure insensitive institutionsIntercept 0.003 (0.014)** 0.002 (0.323) Intercept 0.034 (0.000)*** 0.0026 (0.000)***

�Ownership −0.475 (0.000)*** −0.335 (0.000)*** �Leverage deviation −0.002 (0.000)*** −0.194 (0.001)**

Con D 0.009 (0.000)*** 0.008 (0.000)*** Con D 0.003 (0.772) 0.009 (0.000)***

�Ownership × Con D −0.411 (0.000)*** −0.284 (0.000)*** �Leverage deviation × Con D −0.596 (0.000)*** −0.871 (0.000)***

Asset beta −0.012 (0.000)*** 0.007 (0.000)*** Age −0.003 (0.011)** −0.002 (0.002)***

EBITD −0.019 (0.000)*** −0.007 (0.078)* Beta −0.001 (0.095)* −0.001 (0.064)*

M/B 0.001 (0.040)** −0.006 (0.000)*** Dividend yield −0.613 (0.007)*** 0.049 (0.808)PPE −0.001 (0.529) −0.001 (0.881) Firm-specific risk −0.206 (0.000)*** −0.010 (0.476)R&D −0.002 (0.285) −0.002 (0.021)** Lag return 0.054 (0.000)*** 0.003 (0.001)***

R&D D −0.001 (0.320) −0.001 (0.892) Price −0.001 (0.011)** −0.004 (0.000)***

SE 0.001 (0.399) −0.001 (0.122) Standard deviation 0.016 (0.059)* 0.027 (0.001)***

Size S 0.001 (0.001)*** 0.001 (0.019)** Size E −0.001 (0.369) 0.001 (0.013)**

Turnover −0.005 (0.000)*** −0.010 (0.000)***

Adjusted-R2 0.019 0.014 0.029 0.008N 40,963 40,456 40,963 40,456

dx.doi.org/10.1016/j.najef.2014.10.001

Please cite

this

article in

press

as: C

hu

ng,

C.

Y.,

& W

ang,

K.

Do

institu

tional

investors

mon

itorm

anagem

ent?

Eviden

ce from

the

relationsh

ip betw

een in

stitution

al. .

. .

North

Am

erican Journal

ofEconom

ics and

Finance (2014),

http

://dx.d

oi.org/10.1016/j.najef.2014.10.001

AR

TIC

LE

IN P

RE

SS

G M

odelEC

OFIN

470 1–31

C.Y. Chung,

K.

Wang

/ N

orth A

merican

Journal of

Economics

and Finance

xxx (2014)

xxx–xxx

25



�Ownership in PSI = �ownership in PII 3.20 (0.059)* 4.94 (0.059)*

�Ownership × Con D in PSI = �ownership × Con D in PII 5.43 (0.052)* 13.23 (0.001)***

�Leverage deviation in PSI = �Leverage deviation in PII 9.32 (0.001)*** 12.29 (0.001)***

�Leverage deviation × Con D in PSI = �Leverage deviation × Con D in PII 18.64 (0.000)*** 22.15 (0.000)***

This table reports the coefficient estimates for both the contemporaneous and lagged relationships between changes in institutional ownership and changes in leverage deviation from1985 to 2008. Leverage deviation is the absolute value of total leverage minus target leverage. �Leverage deviation is the change in leverage deviation over the year. �Ownership is thechange in institutional ownership over the year. EBITD is earnings before interest, taxes, and depreciation divided by assets. M/B is market value of asset divided by book value of asset.PPE is net property, plants, and equipment divided by assets. R&D is research and development expenses divided by sales. R&D D is equal to 1 if a firm does not report its research anddevelopment expenses. SE is selling expenses divided by sales. Size E and Size S are the natural logarithm of the market value of equity and sales, respectively. Beta is from the rollingregressions based on Fama and French (1992). Standard deviation is the natural logarithm of the monthly standard deviation of returns over the previous 24 to 60 months. Asset betais the weighted average beta of equity and debt. Firm-specific risk is the natural logarithm of 1 plus the average daily squared firm-specific returns over a month, and the firm-specificreturn is the difference between a firm’s daily return and its industry return. Age is the natural logarithm of the number of months for which the firm has CRSP return data. Dividend yieldis the natural logarithm of 1 plus the average monthly dividend yield over the previous 12 months. Price is the natural logarithm of 1 plus the month-end share price. Turnover is thenatural logarithm of 1 plus the ratio of monthly volume to the number of shares outstanding. Lag Return is the buy-and-hold return over the previous six months. Con D is equal to 1 if�Leverage or �Ownership is obtained during economic contractions. The economic contraction and expansion periods are taken from the National Bureau of Economic Research (NBER)business cycle. Though not reported, an industry dummy based on the Fama-French 49 industry classifications and a year dummy are also included in the model. p-Values are reportedin parentheses.***, **, and * indicate statistical significance at the 1%, 5%, and 10% levels, respectively.

dx.doi.org/10.1016/j.najef.2014.10.001

Please cite

this

article in

press

as: C

hu

ng,

C.

Y.,

& W

ang,

K.

Do

institu

tional

investors

mon

itorm

anagem

ent?

Eviden

ce from

the

relationsh

ip betw

een in

stitution

al. .

. .

North

Am

erican Journal

ofEconom

ics and

Finance (2014),

http

://dx.d

oi.org/10.1016/j.najef.2014.10.001

AR

TIC

LE

IN P

RE

SS

G M

odelEC

OFIN

470 1–31

26

C.Y. Chung,

K.

Wang

/ N

orth A

merican

Journal of

Economics

and Finance

xxx (2014)

xxx–xxx

Table 11Time-series regression coefficients of changes in institutional ownership and changes in leverage deviation.

Independent variable Aggregate institutions Pressure sensitive institutions Pressure insensitive institutions N

�Ownership �Leverage deviation �Ownership �Leverage deviation �Ownership �Leverage deviation

Panel A: regression coefficients incontemporaneous relationships

1980s −0.213 (0.001)*** −0.258 (0.006)*** 0.669 (0.051)* −0.028 (0.585) −0.287 (0.001)*** −0.188 (0.007)*** 74621990s −0.297 (0.000)*** −0.288 (0.000)*** 0.112 (0.000)*** −0.019 (0.001)*** −0.390 (0.000)*** −0.274 (0.000)*** 19,1192000s −0.555 (0.000)*** −0.304 (0.010)*** −0.369 (0.000)*** −0.111 (0.005)*** −0.749 (0.001)*** −0.078 (0.038)*** 14,382

Independent variable Aggregate institutions Pressure sensitive institutions Pressure insensitive institutions N

�Ownership �Leverage deviation �Ownership �Leverage deviation �Ownership �Leverage deviation

Panel B: regression coefficients inlagged relationships

1980s 0.080 (0.620) −0.007 (0.962) 0.798 (0.188) −0.059 (0.444) −0.105 (0.033)** 0.113 (0.323) 53711990s −0.212 (0.026)** −0.176 (0.000)*** −0.026 (0.926) −0.156 (0.001)*** −0.608 (0.000)*** −0.196 (0.064)* 18,8992000s −0.414 (0.000)*** −0.201 (0.000)*** −0.730 (0.000)*** −0.018 (0.644) −1.490 (0.000)*** −0.438 (0.000)*** 16,186

This table reports the time-series regression coefficients for both the contemporaneous and lagged relationships between changes in institutional ownership and changes in leveragedeviation during the 1980s, 1990s, and 2000s. Panel A and Panel B provide the results for the contemporaneous and lagged relationships, respectively. Pressure sensitive institutionsinclude bank trust department and insurance companies, and pressure insensitive institutions include mutual funds and independent investment advisors. Leverage deviation is theabsolute value of total leverage minus target leverage. �Leverage deviation is the change in leverage deviation over the year. �Ownership is the change in institutional ownership overthe year. p-Values are reported in parentheses.***, **, and * indicate statistical significance at the 1%, 5%, and 10% levels, respectively.

dx.doi.org/10.1016/j.najef.2014.10.001

Please cite

this

article in

press

as: C

hu

ng,

C.

Y.,

& W

ang,

K.

Do

institu

tional

investors

mon

itorm

anagem

ent?

Eviden

ce from

the

relationsh

ip betw

een in

stitution

al. .

. .

North

Am

erican Journal

ofEconom

ics and

Finance (2014),

http

://dx.d

oi.org/10.1016/j.najef.2014.10.001

AR

TIC

LE

IN P

RE

SS

G M

odelEC

OFIN

470 1–31

C.Y. Chung,

K.

Wang

/ N

orth A

merican

Journal of

Economics

and Finance

xxx (2014)

xxx–xxx

27Table 12Institutional monitoring and firm performance.

Group Aggregate institutions Pressure sensitive institutions Pressure insensitive institutions

�Ownership> 0, �leveragedeviation < 0

�Ownership< 0, �leveragedeviation > 0

Difference �Ownership> 0, �leveragedeviation < 0




Difference

Panel A:contemporaneousfirm performance

ROA 0.033 [0.129] −0.001 [0.145] 0.034 (0.000)*** 0.034 [0.130] 0.009 [0.142] 0.024 (0.000)*** 0.032 [0.126] 0.004 [0.146] 0.028 (0.000)***

�ROA 0.001 [0.091] −0.017 [0.107] 0.019 (0.000)*** 0.001 [0.091] −0.012 [0.103] 0.014 (0.000)*** 0.001 [0.090] −0.016 [0.107] 0.016 (0.000)***

ROE 0.010 [0.613] −0.072 [0.713] 0.082 (0.001)*** 0.047 [1.425] 0.039 [0.754] 0.008 (0.146) 0.047 [0.213] −0.079 [0.127] 0.126 (0.000)***

�ROE 0.005 [0.064] −0.007 [0.073] 0.012 (0.001)*** 0.005 [0.064] −0.008 [0.075] 0.013 (0.000)*** 0.035 [0.065] −0.012 [0.075] 0.047 (0.001)***

Q 0.262 [1.579] −0.414 [1.398] 0.676 (0.001)*** −0.281 [1.481] −0.387 [1.425] 0.106 (0.000)*** 0.281 [1.481] −0.387 [1.425] 0.668 (0.000)***

�Q 0.125 [1.363] −0.040 [1.229] 0.165 (0.001)*** −0.115 [1.332] 0.048 [1.231] −0.163(0.000)***

0.115 [1.332] −0.004 [1.231] 0.119 (0.000)***

N 11,541 8831 10,023 10,492 11,569 8422

Group Aggregate institutions Pressure sensitive institutions Pressure insensitive institutions

�Ownership> 0, �leveragedeviation < 0






Difference

Panel B: lagged firmperformance after�Ownership

ROA 0.032 [0.129] 0.008 [0.136] 0.024 (0.000)*** 0.031 [0.1297] 0.015 [0.1317] 0.015 (0.000)*** 0.031 [0.1302] 0.012 [0.1416] 0.018 (0.000)***

�ROA 0.001 [0.097] −0.009 [0.099] 0.010 (0.000)*** −0.001 [0.098] −0.007 [0.095] 0.007 (0.000)*** −0.001 [0.096] −0.008 [0.101] 0.007 (0.000)***

ROE 0.041 [0.213] 0.013 [0.123] 0.028 (0.001)*** 0.161 [0.110] 0.251 [0.068] −0.089(0.001)***

0.161 [0.110] −0.251 [0.008] 0.412 (0.001)***

�ROE 0.024 [0.043] −0.003 [0.073] 0.027 (0.001)*** 0.009 [0.011] 0.012 [0.008] −0.003(0.001)***

0.015 [0.018] −0.007 [0.009] 0.022 (0.001)***

Q 0.292 [0.431] −0.409 [0.429] 0.701 (0.001)*** −0.311 [1.417] −0.385 [1.337] 0.073 (0.001)*** 0.311 [1.417] −0.385 [1.337] 0.696 (0.000)***

�Q 0.025 [0.240] −0.006 [0.234] 0.031 (0.001)*** −0.030 [1.204] −0.014 [1.206] −0.016 (0.402) 0.030 [1.204] −0.014 [1.206] 0.044 (0.000)***

N 13,673 7585 11,717 8817 13,706 7241

This table reports the comparison results for firm performance between two groups of firms: firms in which institutional ownership increases and leverage deviation decreases(�Ownership > 0 and �leverage deviation < 0) and firms in which institutional ownership decreases and leverage deviation increases (�Ownership < 0 and �leverage deviation > 0).Panel A and Panel B provide the results based on the contemporaneous and lagged firm performance, respectively. Pressure sensitive institutions include bank trust department andinsurance companies, and pressure insensitive institutions include mutual funds and independent investment advisors. Leverage deviation is the absolute value of total leverage minustarget leverage. �Leverage deviation is the change in leverage deviation over the year. �Ownership is the change in institutional ownership over the year. ROA is the ratio of earningsbefore interest and taxes to total assets, ROE is the ratio of earnings before interest and taxes to total equity, and Q is Tobin’s Q defined as Book value of assets plus market value of commonstock less sum of book value of common stock and balance sheet deferred taxes divided by book value of assets. All measures are adjusted by the industry mean. p-Values are reported inparentheses.***, **, and *Q7

dx.doi.org/10.1016/j.najef.2014.10.001




average is computed based on the Fama-French 49 industry classifications. We compute the changein firm value as the difference in values over the year. We consider both the contemporaneous andlagged relationships between institutional monitoring and change in firm value.

Table 12 reports the comparison of firm performance between the two groups. In aggregate andpressure insensitive institutions, we find consistent evidence that firms with effective institutionalmonitoring show a higher firm value and change in firm value than those with weakened institutionalmonitoring, suggesting that successful institutional monitoring enhances firm value. In addition, wefind that an increase in ownership by pressure sensitive institutions leads to worse performancewhen measured by the change in Tobin’s Q. This result is consistent with the hypothesis that pressuresensitive institutions have little incentives to monitor firm management, and as such, their ownershipchange may not reflect institutional monitoring.

5. Conclusions

Institutional ownership has increased substantially over the past decades in US financial markets.The extant literature indicates that institutional investors have strong incentives to influence man-agement decisions. Even though a strong relationship between institutional ownership and a firm’scapital structure is expected, the empirical evidence and its implications are still unknown. In thispaper, we use the 3SLS model to identify the causal effects of changes in institutional ownership onchanges in leverage as well as changes in leverage deviation. We find that changes in leverage decreasewhen changes in institutional ownership increase, suggesting that a firm reduces its debt level as insti-tutional investors substitute for external debt monitoring. Most importantly, we find that changes inleverage deviation (suboptimal leverage) decrease when changes in institutional ownership increase,and that changes in institutional ownership decrease when changes in leverage deviation increase.These results amplify the evidence that institutional investors play both active and passive monitoringroles (i.e., through institutional activism and “voting-with-feet”) in capital structure management. Ourresults also show that institutional monitoring is more pronounced when institutions are less likely togo along with management decisions and/or when firms face high information asymmetry. This studythus offers unique evidence that institutional investors actively improve a firm’s capital structure andpassively sell their shares when dissatisfied with the firm’s capital structure.

Acknowledgement

We would like to thank Editor Hamid Beladi and two anonymous referees for their highly con-structive comments.

Appendix A.

(Tables A and B).Q5

Table AEstimation of expected leverage.

Model

Intercept 0.306 (0.000)***

Asset beta −0.103 (0.000)***

EBITD −0.305 (0.000)***

M/B −0.010 (0.000)***

PPE 0.073 (0.006)***

R&D −0.005 (0.021)**

R&D D 0.036 (0.000)***

SE 0.001 (0.036)**

626

627

628

629

630

631

632

633

634

635

636

637

638

639

640

641

642

643

644

645

646

647

648

649

650

651

652

653

654

655

656

657

658

dx.doi.org/10.1016/j.najef.2014.10.001




Table A (Continued )

Model

Size S 0.022 (0.000)***

N 53,507

This table summarizes the coefficients from the tobit regression model predicting firms’ expected, or target, leverage from 1985to 2008. Leverage is total debt divided by total debt plus the market value of equity. Asset beta is the weighted average beta ofequity and debt. EBITD is earnings before interest, taxes, and depreciation divided by assets. M/B is market value of asset dividedby book value of asset. PPE is net property, plants, and equipment divided by assets. R&D is research and development expensesdivided by sales. R&D D is equal to 1 if a firm does not report its research and development expenses. SE is selling expensesdivided by sales. Size S is the natural logarithm of sales. Though not reported, an industry dummy based on the Fama-French49 industry classifications and a year dummy are also included in the model. p-Values are reported in parentheses.***, **, and * indicate statistical significance at the 1%, 5%, and 10% levels, respectively.Q8

Table BList of variables.

Variable Description

Age Natural logarithm of the number of months for which the firm has CRSP return dataAsset beta Weighted average beta of equity and debtBeta Equity betaCon D Dummy variable equal to 1 if during economic contractions and 0 otherwiseDividend yield Natural logarithm of 1 plus the average monthly dividend yield over the previous 12 monthsEBITD Earnings before interest, taxes, and depreciation divided by assetsFirm-specific return Difference between a firm’s daily return and industry returnFirm-specific risk Natural logarithm of 1 plus the average daily squared firm-specific returns over a monthLag return Buy-and-hold return over the previous six monthsM/B Sum of market value of common stock plus book value of debt divided by book value of assetOwnership Number of shares held by institutional investors divided by number of shares outstanding�Ownership Change in institutional ownership over the yearPPE Net property, plants, and equipment divided by assetsPrice Natural logarithm of 1 plus the month-end share priceROA Earnings before interest and taxes divided by total assetROE Earnings before interest and taxes divided by book value of equityR&D Research and development expenses divided by salesR&D D Dummy variable equal to 1 if a firm does not report the research and development expensesSE Selling expenses divided by salesSize E Natural logarithm of the market value of equitySize S Natural logarithm of salesStandard deviation Natural logarithm of the monthly standard deviation of returns over the previous 24 to 60

monthsTobin’s Q Book value of assets plus market value of common stock less sum of book value of common

stock and balance sheet deferred taxes divided by book value of assetsTurnover Natural logarithm of 1 plus the ratio of monthly volume to number of shares outstandingVariance A Variance of total leverage over the past 3 yearsVariance B Variance of leverage deviation over the past 3 yearsLeverage Total debt divided by total debt plus the market value of equityTarget leverage Predicted leverage from the tobit regression modelLeverage deviation Absolute value of difference between leverage and target leverage�Leverage Change in total leverage over the year�Leverage deviation Change in leverage deviation over the year�Ownership Change in institutional ownership over the year

Appendix B. Supplementary data

Supplementary data associated with this article can be found, in the online version, athttp://dx.doi.org/10.1016/j.najef.2014.10.001.

References

Almazan, A., Hartzell, J. C., & Starks, L. T. (2005). Active institutional shareholders and costs of monitoring: Evidence fromexecutive compensation. Financial Management, 34, 5–34.

659

660

661

662

663

664

dx.doi.org/10.1016/j.najef.2014.10.001

http://dx.doi.org/10.1016/j.najef.2014.10.001




Baker, M., & Wurgler, J. (2002). Market timing and capital structure. Journal of Finance, 57, 1–32.Bathala, C. T., Moon, K. P., & Rao, R. P. (1994). Managerial ownership, debt policy, and the impact of institutional holdings: An

agency perspective. Financial Management, 23, 38–50.Bennett, J. A., Sias, R. W., & Starks, L. T. (2003). Greener pastures and the impact of dynamic institutional preferences. Review of

Financial Studies, 16, 1203–1238.Bhide, A. (1994). Efficient markets, deficient governance: US securities regulations protect investors and enhance market

liquidity. But do they alienate managers and shareholders? Harvard Business Review, 72, 128–140.Brickley, J., Lease, R., & Smith, C. (1998). Ownership structure and voting on antitakeover amendments. Journal of Financial

Economics, 20, 267–292.Campbell, J., Lettau, B. M., & Xu, Y. (2001). Have individual stocks become more volatile? An empirical exploration of idiosyncratic

risk. Journal of Finance, 56, 1–43.Carleton, W. T., Nelson, J. A., & Weisbach, M. S. (1998). The influence of institutions on corporate governance through private

negotiations: Evidence from Tiaa-cref. Journal of Finance, 53, 1335–1362.Chan, C.-C., Lin, B.-H., Chang, Y.-H., & Liao, W.-C. (2013). Does bank relationship matter for corporate risk-taking? Evidence from

listed firms in Taiwan. North American Journal of Economics and Finance, 26, 323–338.Chung, R., Firth, M., & Kim, J. (2002). Institutional monitoring and opportunistic earnings management. Journal of Corporate

Finance, 8, 29–48.Coffee, J. C. (1991). Liquidity versus control: the institutional investor as corporate monitor. Columbia Law Review, 91, 1277–1368.Del Guercio, D. (1996). The distorting effects of the prudent-man laws on institutional equity investments. Journal of Financial

Economics, 40, 31–62.Del Guercio, D., & Hawkins, J. (1999). The motivation and impact of pension fund activism. Journal of Financial Economics, 52,

293–340.Falkenstein, E. (1996). Preferences for stock characteristics as revealed by mutual fund portfolio holdings. Journal of Finance, 51,

111–135.Fama, E. F., & French, K. R. (1992). The cross-section of expected stock returns. Journal of Finance, 47, 427–465.Ferreira, M. A., Massa, M., & Matos, P. (2010). Shareholders at the gate? Institutional investors and cross-border mergers and

acquisitions. Review of Financial Studies, 23, 601–644.Fischer, E. O., Heinkel, R., & Zechner, J. (1989). Dynamic capital structure choice: Theory and tests. Journal of Finance, 44,

41–57.Flannery, M. J., & Rangan, K. P. (2006). Partial adjustment toward target capital structures. Journal of Financial Economics, 79,

469–506.Frank, M. Z., & Goyal, V. K. (2009). Capital structure decisions: Which factors are reliably important? Financial Management, 38,

1–37.Gillan, S. L., & Starks, L. T. (1998). A survey of shareholder activism: Motivation and empirical evidence. Contemporary Finance

Digest, 2, 10–34.Gillan, S. L., & Starks, L. T. (2000). Corporate governance proposals and shareholder activism: The role of institutional investors.

Journal of Financial Economics, 57, 275–305.Gompers, P. A., & Metrick, A. (2001). Institutional investors and equity prices. Quarterly Journal of Economics, 116, 229–259.Gompers, P. A., Ishii, J., & Metrick, A. (2003). Corporate governance and equity prices. Quarterly Journal of Economics, 118,

107–155.Grinstein, Y., & Michaely, R. (2005). Institutional holdings and payout policy. Journal of Finance, 60, 1389–1426.Harford, J., Klasa, S., & Walcott, N. (2009). Do firms have leverage targets? Evidence from acquisitions. Journal of Financial

Economics, 93, 1–14.Hartzell, J. C., & Starks, L. T. (2003). Institutional investors and executive compensation. Journal of Finance, 58, 2351–2374.Hsu, G., & Koh, P. (2005). Does the presence of institutional investors influence accruals management? Evidence from Australia,

Corporate Governance: An International Review, 13, 809–823.Huddart, S. (1993). The effect of a large shareholder on corporate value. Management Science, 39, 1407–1421.Jensen, M. C., & Meckling, W. H. (1976). Theory of the firm: Managerial behavior, agency cost, and ownership structure. Journal

of Financial Economics, 3, 305–360.Jensen, M. C. (1986). Agency cost of free cash flow, corporate finance, and takeovers. American Economic Review, 76, 323–329.Kaplan, S., & Zingales, L. (1997). Do investment-cash flow sensitivities provide useful measures of financing constraints?

Quarterly Journal of Economics, 112, 169–216.Karpoff, J. M., Malatesta, P. H., & Walkling, R. A. (1996). Corporate governance and shareholder initiatives: Empirical evidence.

Journal of Financial Economics, 42, 365–395.Kayhan, A., & Titman, S. (2007). Firms’ histories and their capital structures. Journal of Financial Economics, 83, 1–32.La Porta, R., Lopez-De-Silanes, F., Shleifer, A., & Vishny, R. (2000). Agency problems and dividend policies around the world.

Journal of Finance, 55, 1–33.Lemmon, M. L., Roberts, M. R., & Zender, J. F. (2008). Back to the beginning: Persistence and the cross-section of corporate capital

structure. Journal of Finance, 63, 1575–1608.Maug, E. (1998). Large shareholders as monitors: Is there a trade-off between liquidity and control? Journal of Finance, 53,

65–98.Michaely, R., & Vincent, C. (2013). Do institutional investors influence capital structure decisions? Cornell University (Unpublished

manuscript)Modigliani, F., & Miller, M. (1963). Corporation income taxes and the cost of capital: A correction. American Economic Review,

53, 433–443.Myers, S. C. (1977). Determinants of corporate borrowing. Journal of Financial Economics, 5, 147–175.Parrino, R., Sias, R. W., & Starks, L. T. (2003). Voting with their feet: Institutional ownership changes around forced CEO turnover.

Journal of Financial Economics, 68, 3–46.Shleifer, A., & Vishny, R. W. (1986). Large shareholders and corporate control. Journal of Political Economy, 94, 461–478.Smith, M. P. (1996). Shareholder activism by institutional investors: Evidence from CalPERS. Journal of Finance, 51, 227–252.

665

666

667

668

669

670

671

672

673

674

675

676

677

678

679

680

681

682

683

684

685

686

687

688

689

690

691

692

693

694

695

696

697

698

699

700

701

702

703

704

705

706

707

708

709

710

711

712

713

714

715

716

717

718

719

720

721

722

723

724

725

726

727

728

729

730

731

732

733

734

dx.doi.org/10.1016/j.najef.2014.10.001




Stulz, R. M. (1990). Managerial discretion and optimal financing policies. Journal of Financial Economics, 26, 3–27.Tai, V. W., Lai, Y.-H., & Lin, L. (2014). Local institutional shareholders and corporate hedging policies. North American Journal of

Economics and Finance, 28, 287–312.Wang, M. (2014). Which types of institutional investors constrain abnormal accruals? Corporate Governance: An International

Review, 22, 43–67.Webb, R., Beck, M., & McKinnon, R. (2003). Problems and Limitations of institutional investor participation in corporate gover-

nance. Corporate Governance: An International Review, 11, 65–73.Wooldridge, J. (2002). Econometric analysis of cross section and panel data. Cambridge: MIT Press (Chapter 8).

735

736

737

738

739

740

741

742

dx.doi.org/10.1016/j.najef.2014.10.001

do institutional investors monitor management? evidence from the relationship between institutional...

Documents