A Test of the Free Cash Flow Hypothesis: The Impact of Increased
Institutional Holdings on Firm Characteristics*
SIGITAS KARPAVICIUS† and FAN YU‡
July 19, 2011
ABSTRACT
This paper tests the free cash flow hypothesis and analyzes the impact of the increased
institutional ownership on firm characteristics. Institutional ownership of U.S. equities increases
from 7.3% in 1980 to 45.7% in 2009. Greater institutional ownership reduces the agency
problem of free cash flow. We find that the increased institutional ownership results in the lower
leverage and payout that consequently lead to greater cash holdings and firm value. The results
support the free cash flow hypothesis and provide an alternative explanation why firms hold so
much cash and why debt and payout ratios decrease during the last 30 years.
Key words: Agency problem; Free cash flow hypothesis; Institutional ownership; Cash holdings;
Capital structure; Payout policy
JEL classification: G23; G32; G35
*
We thank Jarrad Harford and seminar participants at the Shanghai University of Finance and Economics and 2011
China International Conference in Finance for their helpful comments and suggestions.
†
Corresponding author. Address: Flinders Business School, Flinders University, GPO Box 2100, Adelaide SA 5001,
Australia. E-mail: [email protected].
‡
Foster School of Business, Box 353200, University of Washington, Seattle, WA 98195, USA.
1
The average cash-to-assets ratio for U.S. industrial firms doubles from 1980 to 2009. Classical
agency theory predicts that corporate managers with substantial free cash flow are more likely to
invest in negative net present value (NPV) projects even if paying out cash is better for
shareholders (Jensen (1986), Stulz (1990)). Jensen (1986) suggests using debt and cash payout to
control the agency problem associated with excess cash flow accessible to managers. These two
mechanisms help prevent such firms from wasting resources on low-return projects. The passive
monitoring has its costs: cash constraint and cost of raising external capital (Jensen and Meckling
(1976), Myers and Majluf (1984)), overleverage (Campello (2006)), agency costs associated with
debt (Myers and Majluf (1984)), and underinvestment (Myers (1977)).
Meanwhile, the average institutional ownership of U.S. industrial firms increases almost seven
times (from 7.3% in 1980 to 45.7% in 2009). Prior literature suggests that the presence of
institutional investors is associated with lower information asymmetry, better corporate
governance, and lower agency costs (see Hartzell and Starks (2003), Szewczyk, Tsetsekos, and
Varma (1992), Brous and Kini (1994), Velury and Jenkins (2006), O’Neill and Swisher (2003)).
The dramatic change in ownership structure gives us an excellent opportunity to analyze its
impact on controlling the agency problem associated with excess cash flow. The goal of this
paper is to test the free cash flow hypothesis and investigate the impact of increased institutional
holdings in corporate equities on cash balances and on two mechanisms that reduce agency costs
of excess cash flow – leverage and payout (the sum of dividends and share repurchase).
2
The empirical evidence on free cash flow hypothesis is mixed. Lang, Stulz, and Walkling (1991)
find support for free cash flow hypothesis by analyzing a sample of U.S. successful tender offers
from 1980 to 1986. They report that bidder returns are significantly negatively related to cash
flow for bidders with low Tobin’s q but not for high Tobin’s q bidders. However, Gregory
(2005) uses UK takeovers of listed domestic companies during the period 1984 to 1992 and finds
no support for free cash flow hypothesis. Griffin (1988) analyses the petroleum industry during
the period 1979 to 1985 and finds support for the hybrid free cash flow model. Lehn and Poulsen
(1991) analyze the source of stockholder gains in going private transactions. The authors find
that the major source of the gains is the mitigation of agency problems associated with free cash
flow. Lang and Litzenberger (1989) analyze dividend announcements and provide the support
for the free cash flow hypothesis. In contrast, Howe, He, and Kao (1992) analyze tender offer
share repurchase and specially designated dividend announcements and find no support for free
cash flow theory. Richardson (2006) finds evidence that over-investment is concentrated in firms
with the highest levels of free cash flow supporting free cash flow hypothesis.
A relatively small sample size is the common drawback of most of these studies. For example,
Lang, Stulz, and Walkling (1991) have totally 101 observations in their sample; Griffin (1988)
uses the panel data set for 25 firms; the sample size of Lehn and Poulsen (1989) is 236
observations; the sample in Gregory (2005) consists of 217 observations. Bathala, Moon, and
Rao (1994) test their hypotheses using 516 observations. However, the sample of Richardson
(2006) covers 58,053 firm-year observations. Our paper uses a much larger sample that consists
3
of more than 140,000 observations. It spans over three decades and covers most of Compustat
firms.
Another stream of literature focuses on the increasing cash balances of industrial firms. Recent
literature documents that the increased cash holdings are in line with the rational behavior of a
firm. Opler et al. (1999) find that firms with better growth prospects and riskier cash flow tend to
hold more cash. Bates, Kahle, and Stulz (2009) point out that the cash increase is due to the
changes in firm characteristics. They find that the increasing risk in cash flow and the greater
importance of research and development (R&D) expense relative to capital expenditure
(CAPEX) requires firms to hold more cash. The literature suggests that if a firm cannot take a
full advantage of the growth opportunities, it risks being predated and losing its market share.
For example, Chevalier (1995) investigates supermarket leveraged buyouts (LBOs). She finds
that the prices decrease in the local market following an LBO if the rival firms are not highly
leveraged while the prices rise if rival firms are also highly leveraged. Haushalter, Klasa, and
Maxwell (2007) report that firms hold more cash and use more derivatives if they share a larger
proportion of their growth opportunities with rivals.
Further, Faulkender and Wang (2006) find that additional cash is most highly valued by
shareholders of firms with low levels of cash holdings; however, the value of additional cash
diminishes in the level of cash holdings. Foley et al. (2007) argue that the tax costs associated
with repatriations contribute to the magnitude of cash holdings. Harford, Mansi, and Maxwell
(2008) find that cash holdings increase with stronger corporate governance.
4
We argue that concentrated institutional ownership, measured as the ownership controlled by
five largest institutional investors, is an alternative monitoring mechanism for agency problem.
The results show that institutional monitoring has partially substituted debt and payout as the
increase in institutional holdings leads to the lower debt and payout ratios. As institutions are
good monitors, the decreased debt and payout result in greater cash balances rather than to
investments in negative NPV projects. Further, cash reserves are positively affected by greater
institutional holdings. In the analysis, we control for the predation risk and still find that the cash
holdings are higher for firms with greater institutional ownership. At last, we show that greater
cash balances enhance firm value. It is consistent with shareholder wealth maximization. The
results are statistically and economically significant and robust for both high-tech and non hightech firms. This study provides empirical support for free cash flow hypothesis and helps explain
the evolution of leverage, cash balances, and payout ratio during the last 30 years.
The rest of the paper is structured as follows. Section I develops testable hypotheses. Section II
describes the sample. Obtained results are detailed in Section III. Finally, Section IV concludes.
I. Hypotheses Development
We start from the free cash flow hypothesis. It assumes that managers want to invest all the
available funds even in negative NPV projects. This conflict is not likely to be resolved by
5
contracts based on cash flow and investment expenditure. The use of debt can decrease the free
cash flow available to managers through repayment to debtholders (Jensen and Meckling (1976),
Jensen (1986)). Jensen (1986) suggests also using dividends. Similarly, by paying to
shareholders, a firm’s free cash flow decrease. Managers are less likely to invest in value
destroying projects if they do not have sufficient funds. Alternative mechanism for controlling
inefficient investment is through monitoring. We use ownership controlled by five largest
institutional investors (Top5 holdings) as a proxy for monitoring.1
Empirical studies suggest that institutions are good monitors. Carleton, Nelson, and Weisbach
(1998) use a private database consisting of the correspondence between TIAA-CREF and 45
firms it contacted about governance issues between 1992 and 1996, to analyze the process of
private negotiation between financial institutions and the companies they attempt to influence.2
They find that at least 87% of the firms took actions. A survey conducted by McCahery, Starks,
and Sautner (2010) found that the majority institutions that responded to their survey are willing
to engage in shareholder activism. Chen, Harford, and Li (2007) use acquisition decisions to
reveal monitoring and find that firms with concentrated holdings of independent long-term
institutions are more likely to make withdrawal of bad bids. Hartzell and Starks (2003) find that
1
As a robustness check, we use total institutional holdings as the proxy for monitoring and our results are even
stronger.
2
TIAA-CREF is the abbreviation of Teachers Insurance and Annuity Association - College Retirement Equities
Fund.
6
firms with higher concentrated institutional holdings are associated with lower level of CEO
compensation and higher pay-for-performance sensitivities.3
If monitoring by concentrated institutional investors can substitute higher leverage and higher
payout ratio as controlling mechanism, then firms with higher Top5 holdings will have lower
leverage and lower payout ratio. This leads to our first two hypotheses:
Hypothesis 1: Higher ownership controlled by five largest institutional investors will be
associated with lower leverage.
Hypothesis 2: Higher ownership controlled by five largest institutional investors will be
associated with lower payout ratio.
The debt and payout policies are insufficient to discourage the managers not to engage into lowreturn projects. It is likely that firms still invest in negative NPV projects, but less than in
absence of debt and payout. The monitoring and pressure by the institutional investors might
discourage firm management to invest in negative NPV projects. Thus, we expect a positive
relationship between institutional ownership and cash holdings:
Hypothesis 3: Higher ownership controlled by five largest institutional investors will be
associated with higher cash holdings.
In presence of good monitoring, lower debt and payout ratios mechanically lead to greater cash
3
However, prior studies also show that institutional investors do not always have influence on a firm’s corporate
governance. Karpoff, Malatesta, and Walkling (1996) find no persuasive evidence that shareholder proposals
increase firm value. However, it might be due to the actions behind the door before the initiation of shareholder
proposal.
7
balances rather than to investment in negative NPV projects. Our next two hypotheses are as
follows:
Hypothesis 4: There will be a negative relationship between cash holdings and leverage.
Hypothesis 5: There will be a negative relationship between cash holdings and payout
ratio.
It is costly to use debt and payout to reduce agency costs of free cash flow. Both mechanisms
reduce firm’s financial flexibility. A firm must forego some good projects if they require quick
response or the external financing is too costly for the firm (Myers and Majluf (1984)). Firm
value is hurt by insufficient internal funds. Besides, if a firm shares a large portion of investment
opportunities with its rivals, it risks being predated and losing market share if it cannot make
sufficient investment. Haushalter, Klasa, and Maxwell (2007) find inter- and intra-industry
evidence that the extent of the interdependence of a firm’s investment opportunities with rivals is
positively associated with its use of derivatives and the size of its cash holdings. Campello
(2006) shows that debt taking can both boost and hurt firm performance: moderate debt taking is
associated with relative-to-rival sales gains; and high indebtedness leads to product market
underperformance. Further, Dittmar and Mahrt-Smith (2007) show that good corporate
governance improves the value of cash reserves and so enhances firm value. We would expect
that a firm which adopts better monitoring is more likely to enhance its value by increasing its
cash holdings:
Hypothesis 6: Cash holdings will be positively associated with firm value.
8
Figure 1 illustrates our hypotheses.
[Insert Figure 1 here]
II. Data
Our initial sample is drawn from Compustat. It covers the period 1980 through 2009. We
eliminate financial firms (with Standard Industrial Classification (SIC) codes 6000-6999) since
they have different capital structure and their cash balances might be subject to the regulatory
authority. We also exclude public utility firms (with SIC codes 4900-4999) because they operate
in regulated industries and their financing and capital structure decisions might be impacted by
the changes in the regulatory environment. To be included in the sample, firms must have
positive book value of assets (Compustat item AT), positive sales (Compustat item SALE),
positive common shares outstanding (Compustat item CSHO), positive closing share price at the
end of the fiscal year (Compustat item PRCC_F), and be incorporated in the United States.
Fama and French (2001) report that the population of firms has changed over time. The
proportion of small firms with low profitability but high growth opportunities has increased. It is
likely that these firms are from high-tech sector. Thus, we control for industry (high-tech vs. non
9
high-tech) in the analysis by including high-tech dummy in the models. Consistent with
TechAmerica, we use 45 SIC codes to define the high-tech industry.4
Table I presents the number of all firms, high-tech firms, and non high-tech firms in each year.
The last column reports the high-tech firms ratio (the number of high-tech firms over the number
of all firms in each year). We find that the ratio of high tech firms increases from 14.7% to
24.9% during the sample period.
[Insert Table I here]
Table I presents the evolution of institutional ownership during the sample period. Institutional
ownership is the percentage of shares held by institutional investors. We assume that firms not
covered by Thomson Reuters have no institutional investors. We winsorize institutional
ownership at one to avoid non-meaningful numbers. The mean (median) institutional ownership
increases from 7.3% (0.0%) in 1980 to 45.7% (47.9%) in 2009. In addition, we report the
institutional ownership for high-tech and non high-tech firms. We find that the institutional
ownership is quite similar for both sectors over the whole sample period. We also find that the
mean (median) value of Top5 holdings increases from 4.8% (0.0%) to 21.4% (23.0%) from 1980
4
TechAmerica is a U.S. technology trade association. It was formed from the merger of AeA (formerly known as the
America Electronics Association), the Cyber Security Industry Alliance (CSIA), the Government Electronics &
Information Technology Association (GEIA), and the Information Technology Association of America (ITAA) in
2009. 45 SIC codes can be retrieved from http://www.techamerica.org/sic-definition.
10
to 2009. The substantial changes in the ownership structure over the three decades should have
impacted firms’ financing and capital structure decisions.
The increase in institutional holdings can be explained by the general increase in the financial
assets of institutional investors over the sample period. Financial assets of institutional investors
in the U.S.A. increase from $11.2 trillion in 1995 to $24.2 trillion in 2007 (in constant 2000 U.S.
dollars). This corresponds to 140.8% and 211.2% of GDP respectively (Gonnard, Kim, and
Ynesta (2008)). Thus, we can assume that the increase in institutional ownership is exogenous.
Nevertheless, for robustness, we still control for the possible endogeneity.
Table II presents the evolution of cash balances over time. We use two measures of cash
balances: book cash ratio (cash and short-term investments (Compustat item CHE) over book
value of assets) and market cash ratio (cash and short-term investments over market value of
assets).5 To mitigate the impact of outliers and errors, we winsorize the values of both cash ratios
at the tails of 0.5% and 99.5%. We find the substantial increase in cash holdings over time as
illustrated in Table II. Market (book) cash ratio increases from 7.6% to 13.2% (from 10.6% to
22.6%) over the sample period. The median values are smaller but have similar dynamics. We
also divide our sample into non high-tech firms and high-tech firms. The evolution of cash
balances for both subsamples is similar.
[Insert Table II here]
5
Market value of assets = book value of assets – common equity (Compustat item CEQ) + common shares
outstanding * closing share price at the end of the fiscal year.
11
Table III presents the dynamic of leverage over the sample period. We measure leverage using
book leverage (debt over book value of assets) and market leverage (debt over market value of
assets).6 We winsorize the values of both debt ratios at the tails of 0.5% and 99.5%. We find that
the mean (median) market leverage decreases from 23.1% to 15.9% (from 19.8% to 9.4%)
during the 1980-2009 period. However, the evolutions of mean and median book leverage are
quite different: mean book leverage slightly increases (from 26.9% to 28.7%) whereas median
book leverage decreases from 24.5% to 15.7%. Consistent with the prior empirical studies, we
find that high-tech firms tend to have lower debt ratios than non high-tech firms.
[Insert Table III here]
Next, we compute the payout ratio. It is a sum of common stock dividends (Compustat item
DVC) and absolute value of the difference between purchase of common and preferred stock
(Compustat item PRSTKC) and preferred stock redemption value (Compustat item PSTKRV)
divided by market value of equity (common shares outstanding * closing share price at the end of
the fiscal year). We find that this variable has a lot of outliers; thus, we winsorize it at the tails of
5% and 95%. The evolution of the payout ratio over the sample period is shown in Table IV. The
mean (median) payout ratio decreases from 5.5% to 3.1% (from 2.3% to 0.1%) during the last
6
Debt is the sum of long-term debt (Compustat item DLTT) and debt in current liabilities (Compustat item DLC).
12
three decades. We find that the dynamic of payout ratio is impacted by non high-tech industries
as the median payout ratio for high-tech firms is 0. This is consistent with the findings of Fama
and French (2001) who report that firms have become less likely to pay dividends and the
proportion of dividend payers decreases, due in part to the changing firm characteristics.
The descriptive statistics show the negative trend for debt and payout ratios; however, cash
holdings tend to increase over the sample period. This provides the initial support for our
hypotheses. However, we find that the ratios based on book value of assets and ratios based on
market value of assets have different evolutions over time. One possible explanation is the
decreasing book-to-market ratio (book value of assets divided by market value of assets). Table
IV reports the book-to-market ratio, winsorized at the tails of 1% and 99%, in each year. The
mean (median) book-to-market ratio is 0.851 (0.895) in 1980 and decreases to 0.696 (0.683) in
2009. We also find that on average, high-tech firms have lower book-to-market ratio than non
high-tech firms; however, the gap between the two ratios erodes over time. Thus, the decreasing
book-to-market ratio is indeed one of the possible explanations for the differences between the
ratios based on book value of assets and ratios based on market value of assets.
[Insert Table IV here]
13
III. Results
In this section, we test our hypotheses. First of all, we analyze the impact of institutional
ownership on firms’ leverage. Then we investigate whether payout ratio is affected by the
increase in institutional ownership. Further, we test the individual and combined effects of the
changes in Top5 holdings, leverage, and payout ratio on firms’ cash balances. At last, we test
whether greater cash balances enhance firm value.
A. The Impact of Institutional Ownership on Leverage
To test whether there is a negative relationship between leverage and concentrated institutional
ownership, we estimate the regressions similar to those used in Chang and Dasgupta (2009),
Fama and French (2002), Flannery and Rangan (2006), and Lemmon, Roberts, and Zender
(2008). Specifically, our benchmark models are:
Market leveraget   0  1  Top5 holdings t   2  HT dummy t   3Ln  Assets t   4  B/M t
 5  EBIT/Assets t   6  PPE/Assets t   7  R&D/Assets t   7  R&D dummy t   t ;
Book leveraget   0  1  Top5 holdings t   2  HT dummy t   3Ln  Assets t   4  B/M t
 5  EBIT/Assets t   6  PPE/Assets t   7  R&D/Assets t   7  R&D dummy t   t ,
14
(1)
(2)
where HT dummy is equal to one if a firm is from the high-tech industries and zero otherwise.
Assets denotes book value of assets. B/M is book-to-market ratio. EBIT is earnings before
interests and taxes (the sum of income before extraordinary items (Compustat item IB), interest
and related expense (Compustat item XINT), and income taxes (Compustat item TXT)). PPE is
net property, plant, and equipment (Compustat item PPENT). R&D is research and development
expense (Compustat item XRD). R&D dummy is equal to one when R&D expense is unreported
in Compustat and zero otherwise. To reduce the impact of outliers and potential errors in
Compustat, we winsorize variables EBIT/Assets and R&D/Assets at the tails of 1% and 99%.
Further, PPE/Assets is winsorized so that it is between zero and one. The models include year
fixed effects.7 The standard errors are corrected for clustering at the firm level.
Table V presents the results. Model 1 and Model 2 show the results for Equations (1) and (2),
respectively. We find that both leverage measures are negatively impacted by concentrated
institutional ownership. The results are statistically and economically significant. The average
Top5 holdings have increased by 16.6 p.p. (0.214 – 0.048 = 0.166) over the sample period. The
coefficient estimate of Top5 holdings for market leverage is approximately –0.139. Thus, the
impact of the increase in Top5 holdings on average market leverage is –0.023 (0.166 * (–0.139)
= –0.023), ceteris paribus. It accounts for one third of the average decrease in market leverage as
the mean market leverage has decreased by 0.072 (0.159 – 0.231 = –0.072). The coefficient
estimate of Top5 holdings for book leverage is approximately –0.175. Similarly, the impact on
7
In our main models, we do not include industry fixed effects as it is likely that the effect of the industry might have
changed during 30-year time period. In other words, the impact of a particular industry in 1980 might be different
from the impact in 2009. For robustness, we repeat all our empirical tests with industry fixed effects defined by twodigit SIC codes and find similar results.
15
book leverage is –0.029 (0.166 * (–0.175) = –0.029) whereas the mean book leverage has
increased by 0.018 (0.287 – 0.269 = 0.018) during the 1980-2009 period. We also find that hightech firms have less debt. The signs and significance of the coefficients of other control variables
are similar to those reported in previous studies (see, for example, Chang and Dasgupta (2009),
Fama and French (2002), and Flannery and Rangan (2006)). The results support our Hypothesis
1.
[Insert Table V here]
It is possible that leverage and institutional ownership are interrelated with each other. To control
for endogeneity, we follow Harford, Mansi, and Maxwell (2008) and first estimate Model 3 and
Model 4 whose dependent variables are lead values of market leverage and book leverage. The
results are similar to those for Model 1 and Model 2. Secondly, we include lagged values of
leverage into the models (Model 5 and Model 6). In this specification, the coefficient estimates
for institutional ownership are still positive and statistically significant; however, their values
have become smaller.8
8
Another approach to control for endogeneity is two-stage least squares. However, the suitability of this method
depends on the availability of instrumental variables. Unfortunately, empirical studies that analyze institutional
ownership and firm capital structure (as well as cash holdings and payout policy) use similar control variables. Thus,
in our paper, we do not use two-stage least squares models. Recent empirical study by Harford, Mansi, and Maxwell
(2008) discusses this issue in more details.
16
Franzen, Rodgers, and Simin (2009) report a negative relationship between off-balance sheet
lease financing and leverage. So it is entirely conceivable that the decrease in leverage is
observed as it has been substituted by the greater off-balance sheet lease financing. Following
Franzen, Rodgers, and Simin (2009), we construct a proxy for lease financing, Lease/Assets. It is
equal to the present value of non-cancelable operating leases divided by book value of assets.9 If
lease financing is unreported or missing on Compustat, we assume it is 0. We winsorize variable
Lease/Assets at the tails of 1% and 99%. Model 7 re-estimates Model 1 with the variable
Lease/Assets. We find the off-balance sheet lease financing does not impact leverage. The other
coefficient estimates are the same as in Model 1. To conclude, Table V provides the convincing
results that one of the reasons why leverage has decreased over the sample period is the
substantial increase in institutional ownership.
B. The Impact of Institutional Ownership on Payout Ratio
In this section, we test the impact of concentrated institutional ownership on firms’ payout ratio.
We estimate the model similar to one used in Fama and French (2001) and Fenn and Liang
(2001):
9
Franzen, Rodgers, and Simin (2009) compute the present value of non-cancelable operating leases as the
discounted sum of lease payments (Rental Expense (Compustat item XRENT) + 1/1.1 * Rental Commitments
Minimum 1st Year (Compustat item MRC1) + 1/(1.1)2 * Rental Commitments Minimum 2nd Year (Compustat item
MRC2) + 1/(1.1)3 * Rental Commitments Minimum 3rd Year (Compustat item MRC3) + 1/(1.1)4 * Rental
Commitments Minimum 4th Year (Compustat item MRC4) + 1/(1.1)5 * Rental Commitments Minimum 5th Year
(Compustat item MRC5), where 1.1 is a discount factor).
17
Payout ratiot   0  1  Top5 holdings t   2  HT dummy t   3Ln  Assets t   4  B/M t
 5  Assets growth t   6  EBIT/Assets t   7  Book leverage t   t ,
(3)
where Assets growth is the annual growth rate of book value of assets. A variable Assets growth
is winsorized so that it is not greater than 1. The model includes year fixed effects and standard
errors are corrected for clustering at the firm level.
Model 1 of Table VI presents the results for Equation (3). We find that the impact of institutional
ownership on payout ratio is negative and statistically significant supporting our Hypothesis 2.
The change in average payout ratio over the sample period is –0.024 (0.031 – 0.055 = –0.024).
The coefficient estimate for Top5 holdings is approximately –0.06; therefore, the effect of the
increase in Top5 holdings on average payout ratio is –0.01 (–0.06 * 0.166 = –0.01) and it
accounts for almost 42% of the change in average payout ratio over the sample period. Thus, the
results are economically significant. We find that payout ratio tends to be smaller for high-tech
firms. It is consistent with Fama and French (2001) study which reports that small firms with low
profitability and strong growth opportunities are less likely to pay dividends. Consistent with
Fama and French (2001), we also find that larger, low-growth, and firms with greater book-tomarket ratio tend to have higher payout ratio. However, we find that profitability is negatively
related to payout ratio and it is in contrast to Fama and French (2001), presumable because Fama
and French (2001) use only dividends in their analysis. The results show that book leverage is
positively related to payout ratio. This implies that debt and payout policies are not substitutes in
mitigating agency costs of free cash flow but rather complements. One might argue that book
18
leverage is endogenous because many empirical studies use dividends as one of the independent
variables for explaining leverage and find that dividend payers tend to have less debt (see, for
example, Lemmon, Roberts, and Zender (2008)). Thus, Model 2 re-estimates Model 1 without
book leverage. The coefficient estimates are consistent with those of Model 1. For robustness,
we re-estimate Model 1 and Model 2 with lead values of payout ratio as the dependant variable
(see Model 3 and Model 4). The results are consistent with the previous findings.
[Insert Table VI here]
C. The Determinants of Cash Holdings
To test Hypotheses 3 and 4, we estimate the regressions similar to those used in Opler et al.
(1999) and Bates, Kahle, and Stulz (2009). Specifically, our benchmark models are:
Market cash ratiot   0  1  Top5 holdings t   2  HT dummy t   3Ln  Assets t
 4  NWC/Assets t   5  Industry sigma t   6  FCF/Assets t   7  B/M t
 8  CAPEX/Assets t   9  R&D/Assets t  10  R&D dummy t  11  Book leverage t
12  Dividend dummy t  13  Debt issuance/Assets t  14  Equity issuance/Assets t
15  Acquisitions/Assets t   t ;
19
(4)
Book cash ratiot   0  1  Top5 holdings t   2  HT dummy t   3 Ln  Assets t
 4  NWC/Assets t   5  Industry sigma t   6  FCF/Assets t   7  B/M t
 8  CAPEX/Assets t   9  R&D/Assets t  10  R&D dummy t  11  Book leverage t
(5)
12  Dividend dummy t  13  Debt issuance/Assets t  14  Equity issuance/Assets t
15  Acquisitions/Assets t   t ,
where NWC/Assets is the net working capital scaled by total assets (the difference between
working capital (Compustat item WCAP) and cash and short-term investments divided by book
value of assets). Industry sigma is the mean of the standard deviations of cash flow (operating
income before depreciation (Compustat item OIBDP) – interest and related expense – income
taxes) to book value of assets ratio over 10 years (if there are at least three observations) for
firms in the same industry, as defined by the two-digit SIC code. FCF/Assets is free cash flow
(operating income before depreciation – interest and related expense – income taxes – common
stock dividends (Compustat item DVC)) to book value of assets ratio. CAPEX/Assets is capital
expenditures (Compustat item CAPX) to book value of assets ratio. Dividend dummy is equal to
one if common stock dividends are positive and zero otherwise. Debt issuance/Assets is the
difference between long-term debt issuance (Compustat item DLTIS) and long-term debt
reduction (Compustat item DLTR) divided by book value of assets. Equity issuance/Assets is the
difference between sale of common and preferred stock (Compustat item SSTK) and purchase of
common and preferred stock divided by book value of assets. Acquisitions/Assets is acquisitions
(Compustat item AQC) divided by book value of assets. The variables FCF/Assets,
20
CAPEX/Assets, Debt issuance/Assets, and Equity issuance/Assets are winsorized at the tails of
1% and 99%. NWC/Assets is winsorized so that it is greater than –1.
We include NWC/Assets into the models as it is entirely conceivable that one type of current
assets (cash) substituted other types of current assets (net working capital). Industry sigma
controls for cash flow risk. We expect that firms operating in the riskier industries hold more
cash (see Opler et al. (1999)). Further, we expect that cash holdings increase with FCF/Assets,
Debt issuance/Assets, and Equity issuance/Assets; however, are negatively affected by greater
Acquisitions/Assets and CAPEX/Assets. R&D/Assets is a proxy for growth opportunities. We
expect that firms with better growth opportunities hold more cash. The models also include year
fixed effects. The standard errors are corrected for clustering at the firm level.
Table VII presents the results. Model 1 shows the coefficient estimates where the dependent
variable is market cash ratio (Equation (4)) and Model 2’s dependent variable is book cash ratio
(Equation (5)). We find that cash holdings are positively related to Top5 holdings. The results
are statistically and economically significant. The average market cash ratio has increased by
0.056 (0.132 – 0.076 = 0.056) and the average book cash ratio has increased by 0.120 (0.226 –
0.106 = 0.120) over the sample period. The coefficient estimate of Top5 holdings for market
cash ratio is 0.043. Thus, the impact of the increase in Top5 holdings on market cash ratio is
0.007 (0.043 * 0.166 = 0.007). It corresponds to 13% of the increase in the average market cash
ratio during the last 30 years, ceteris paribus. The coefficient estimate of Top5 holdings for book
cash ratio is 0.089. The impact of the increase in Top5 holdings on market cash ratio is 0.015
21
(0.089 * 0.166 = 0.015). It accounts for 12% of the change in the average book cash ratio during
the last 30 years. Thus, the results support our Hypothesis 3.
[Insert Table VII here]
We also find a negative and statistically significant relationship between cash balances and
leverage. This supports our Hypothesis 4 as lower leverage implies greater cash holdings. We
find that high-tech firms hold more cash on average. The sign and significance of other variables
are similar to those documented in prior studies (see, for example, Opler et al. (1999) and Bates,
Kahle, and Stulz (2009)).
Model 3 and Model 4 re-estimate Model 1 and Model 2 using Payout ratio as the additional
independent variable. We document the negative relationship between cash holdings and payout
ratio. The result is consistent across both models and supports our Hypothesis 5.
For robustness, we re-estimate Model 1 and Model 2 with lead values of cash holdings as the
dependent variables (see Model 5 and Model 6). The results are similar to those reported earlier
and support our hypotheses that cash holdings increase with institutional ownership and decrease
with leverage. At last, we investigate whether our conclusion holds after controlling for the effect
of predation risk. Following Haushalter, Klasa, and Maxwell (2007), we replicate Model 1 using
Herfindahl-Hirschman Index as the additional independent variable. Herfindahl-Hirschman
Index is a measure of product market competition and is calculated using sales data of individual
22
firms in the same industry, as defined by the four-digit SIC code.10 Model 7 in Table VII shows
the coefficient estimates for the regression. We find that institutional ownership and leverage are
still significant determinants of cash holdings after controlling for Herfindahl-Hirschman Index.
However, in contrast to Haushalter, Klasa, and Maxwell (2007), the results show that firms
operating in more competitive industries hold more cash.11
To conclude, the results presented in Table VII support our Hypotheses 3, 4, and 5. We show
that the changes in cash ratios are due in part to the changes in institutional ownership, leverage,
and payout policy. Further, we test Hypothesis 6.
D. The Impact on Firm Value
The results above support our first five hypotheses. However, it does not imply that we find
support for free cash flow hypothesis. We argue that firms should rationally increase their cash
holdings if agency problem of free cash flow is reduced. As the goal of firm management is to
 Sales
HHI 
( Sales )
2
i
10
Herfindahl-Hirschman Index (HHI) is computed as follows:
i
2
, where Salesi denotes sales
i
i
of firm i in a particular industry.
11
As a robustness check, we also use Herfindahl-Hirschman Index calculated assuming that industry is defined by
the two-digit SIC code. In this specification, we find that coefficient estimate for Herfindahl-Hirschman Index is
positive but insignificant. We also re-estimate the models using Book cash ratio as the dependent variable. We find
that the coefficient estimate for Herfindahl-Hirschman Index is negative and statistically significant, disregarding
how we compute Herfindahl-Hirschman Index. Results are available upon request.
23
maximize shareholder value, this rational increase in cash holdings should eventually lead to
greater firm value. In this section, we test this issue (Hypothesis 6).
We use Tobin’s q as a proxy for firm value. Then we estimate the following model:
Qt   0  1  Top5 holdings t   2  HT dummy t   3Ln  Assets t   4  Book leverage t
 5  Book cash ratio t   6  EBIT/Assets t   7  PPE/Assets t  8  CAPEX/Assets t
(6)
10  Dividend dummy t   t ,
where Q is Tobin’s q (market value of assets divided by book value of assets) winsorized at tails
of 1% and 99%. The selection of independent variables is based on the prior studies (see, for
example, Coles, Daniel, and Naveen (2008), Kalcheva and Lins (2007)). The model includes
year fixed effects. Standard errors are corrected for clustering at the firm level.
Model 1 in Table VIII reports the results for Equation (6). We find that cash holdings are
positively associated with firm value proxied by Tobin’s q after controlling for firm
characteristics. It supports Hypothesis 6 that greater cash balances enhance firm value. The
results suggest that greater institutional ownership further increases the firm value. In addition,
the results show that high-tech firms are more likely to have a greater Tobin’s q. Leverage
enhances firm value, presumably due to the additional risk associated with debt and additional
monitoring provided by debtholders. We also find the negative relationship between firm size
and Tobin’s q; however, dividends and capital expenditures tend to improve firm value.
[Insert Table VIII here]
24
Model 2 re-estimates Model 1 using lead values of Tobin’s q as the dependent variable. The
results qualitatively are the same as those in Model 1 and further support Hypothesis 6. We reestimate Model 2 using Tobin’s q as an additional independent variable. Model 3 in Table VIII
presents the results that are consistent with our previous findings except the coefficient estimate
for Top5 holdings is insignificant. In Model 4, we replace Dividend dummy with Payout ratio
and test whether decrease in payout ratio improves firm value. We find that the coefficient
estimate for Payout ratio is negative and statistically significant providing support for our
prediction.
When firm’s cash holdings are low, an increase in institutional concentration (monitoring) has
larger impact on the firm value. With the increase in monitoring (Top5 holdings), we would
expect that a firm will use its cash reserves more effectively. Therefore, we would expect that
each dollar has a higher value. However, Faulkender and Wang (2006) find that the value of
additional cash diminishes in the level of cash. Thus, the monitoring effect is subject to
diminishing marginal returns implying that the monitoring effect is greater when the initial cash
balance is smaller, and vice versa. In other words, keeping the same monitoring level, each
incremental unit of cash will have smaller impact on the improving firm value. Thus, at last we
test whether monitoring effect is indeed non-linear. Model 5 re-estimates Model 1 with the
interaction term of Top5 holdings and Book cash ratio. We expect that the coefficient estimate
for Top5 holdings will be positive and the coefficient estimate for the interaction term will be
25
negative. The results presented in Table VIII show that the marginal effect of Top5 holdings on
cash value is decreasing with the increase of cash reserves. The results support our view.
It is possible that the impact of Top5 holdings on firm and cash values is not instantaneous.
Model 6 re-estimates Model 5 using lead values of Tobin’s q as the dependent variable. The
results are economically and statistically similar.
Then we perform a sensitivity analysis. We calculate hypothetical lead values of Tobin’s q using
the coefficient estimates of Model 6 with different values of Top5 holdings and Book cash ratio,
and mean values of other variables. Then we calculate the difference between the computed
numbers and the hypothetical lead value of Tobin’s q that is computed using the coefficient
estimates of Model 6 and mean values of all variables including Top5 holdings and Book cash
ratio. The positive (negative) difference shows a greater (lower) firm value. Table IX presents
the results. We find that greater cash holdings improve firm value at any level of Top5 holdings.
However, the effect of Top5 holdings on firm value is nonlinear. In the last column (Diff.) of
Table IX, we calculate the incremental impact on firm value when Book cash ratio increases
from 0 to 0.4. We find that cash balances have greater effect on firm value when concentrated
institutional ownership is smaller. Top5 holdings enhance firm value when Book cash ratio is
less than 0.3 or 157% of the mean of Book cash ratio.12 This suggests that in most cases greater
Top5 holdings improve firm value. However, if Book cash ratio is greater than 0.3 then there is a
negative relationship between firm value and concentrated institutional ownership.
12
The mean of Book cash ratio is 0.175.
26
[Insert Table IX here]
In summary, we show that the increased institutional ownership translates into the lower leverage
and payout ratio that consequently lead to greater cash holdings and firm value. The results
provide strong support for the free cash flow hypothesis and help explain the evolution of
leverage, cash holdings, and payout ratio during the last 30 years.
E. Robustness Checks
We perform several robustness checks.13 First of all, we re-estimate all models separately for
high-tech and non high-tech firms as one might argue that these two sectors have evolved
differently over the sample period. However, the results for both types of firms are similar to
those previously reported and further support our hypotheses. We also repeat our empirical tests
with industry fixed effects defined by two-digit SIC codes. All the results hold.
Then we re-estimate all models using total institutional ownership instead of Top5 holdings and
get similar results.
13
The untabulated results are available upon request.
27
In all models, we use ln(Assets) as our firm size proxy. The sample spans over a 30-year period.
Thus, one might argue that our results are systematically biased as firm size tends to increase
over time. We repeat all our tests using the percentile of book value of assets as a proxy for firm
size. The results are essentially unchanged.
Opler et al. (1999) and Haushalter, Klasa, and Maxwell (2007) run the regressions where the
dependent variable is the natural logarithm of the sum of cash and short-term investments
divided by book assets minus cash and short-term investments. Similarly, Harford, Mansi, and
Maxwell (2008) use the natural logarithm of cash-to-sales ratio as a proxy for cash holdings. For
robustness, we repeat all the tests using the natural logarithm of book cash ratio and market cash
ratio as the dependent variables and find similar results.
At last, to make sure that endogeneity is not affecting our results, we estimate three-stage least
squares model. The dependent variables are book leverage, payout ratio, and book cash ratio.
The potential endogenous variables are book leverage and payout ratio.14 The instrumental
variable for book leverage is net property, plant, and equipment scaled by book value of assets.
The instrumental variable for payout ratio is assets growth. The results are similar to those
reported in Tables V-VII except the coefficient estimate for Top5 holdings in Book cash ratio
equation is significant only at 0.115 level (see Table X).15 Thus, the results support our
Hypotheses 1-5.
14
It is also likely that Tobin’s q (inverse Book-to-market ratio) might be endogenous. However, we do not include
Tobin’s q equation into the simultaneous equation model as all the exogenous control variables in the Tobin’s q
equation are also included in the other models. Thus, Tobin’s q would be unidentified.
15
If we replace Top5 holdings with total institutional ownership, we get that the coefficient estimate for total
28
[Insert Table X here]
IV. Conclusion
This paper tests the free cash flow hypothesis and documents the impact of the dramatic increase
in institutional ownership on key firm characteristics. We argue that greater institutional
ownership, measured as the ownership controlled by five largest institutional investors, reduces
the agency problem of free cash flow. To test our hypothesis, we use a large data sample that
spans over a 30-year time period.
The results reveal the channels of value creation. We find that the increased institutional
ownership substitutes other mechanisms that reduce agency problem associated with excess cash
flow. Thus, we observe the decrease in debt and payout ratios. Due to the effective monitoring of
institutional investors, lower debt and payout ratios lead to greater cash holdings rather than to
the value-destroying investments. At last, greater cash balances reduce underinvestment and
predation risks and thus increase firm value. All our tests support these findings.
The results of this paper contribute to our better understanding of the role of institutional
investors in monitoring firm managers and in the process of shareholder wealth maximization.
institutional ownership is significant at 0.001 level and consistent with our main results.
29
The presence of institutional investors enhances firm value directly and indirectly (via greater
cash holdings and reduced underinvestment and predation risks).
The sample that spans over 30-year time period provides an excellent opportunity to investigate
the long-term impact of the change in ownership structure and improved monitoring on key firm
characteristics. Our paper contributes to three different strands of literature. First of all, our
results show that one of the reasons for decreasing payout over time is increased institutional
ownership.16 Secondly, we show that the change in ownership structure is one of the reasons for
decreasing leverage. Thirdly, we provide the alternative explanation for the increased cash
holdings.17 We argue that cash balances increase due to improved monitoring. This suggests that
firms hold less than optimal cash in absence of effective monitoring by shareholders.
To conclude, this paper supports the free cash flow hypothesis. We find that the dramatic
increase in institutional ownership (from 7.3% to 45.7%) during the period 1980 through 2009
positively affects cash holdings of U.S. firms; however, the impact of institutional ownership on
leverage and payout ratio is negative. The results are robust to a number of alternative
specifications.
16
The recent papers that concern this issue include DeAngelo, DeAngelo, and Skinner (2004), Fama and French
(2001), and Grullon and Michaely (2002).
17
Bates, Kahle, and Stulz (2009), Faulkender and Wang (2006), Harford, Mansi, and Maxwell (2008), Haushalter,
Klasa, and Maxwell (2007), and Opler et al. (1999) investigate this issue.
30
REFERENCES
Bates, Thomas W., Kathleen M. Kahle, and René M. Stulz, 2009, Why do U.S. firms hold so
much more cash than they used to? Journal of Finance 64, 1985–2021.
Bathala, Chenchuramaiah T., Kenneth P. Moon, and Ramesh P. Rao, 1994, Managerial
ownership, debt policy, and the impact of institutional holdings: An agency perspective,
Financial Management 23, 38–50.
Bolton, Patrick, and David S. Scharfstein, 1990, A theory of predation based on agency problems
in financial contracting, American Economic Review 80, 93–106.
Brous, Peter A., and Omesh Kin, 1994, The valuation effects of equity issues and the level of
institutional ownership: Evidence from analysts’ earnings forecasts, Financial
Management 23, 33–46.
Campello, Murillo, 2006, Debt financing: Does it boost or hurt firm performance in product
markets?, Journal of Financial Economics 82, 135–172.
Carleton, Willard T., James M. Nelson, and Michael S. Weisbach, 1998, The Influence of
institutions on corporate governance through private negotiations: Evidence from TIAACREF, The Journal of Finance 53, 1335–1362.
Chang, Xin, and Sudipto Dasgupta, 2009, Target behavior and financing: How conclusive is the
evidence?, Journal of Finance 64, 1767–1796.
Chen, Xia, Jarrad Harford, and Kai Li, 2007, Monitoring: Which institutions matter? Journal of
Financial Economics 86, 279–305.
31
Chevalier, Judith A., 1995, Do LBO supermarkets charge more? An empirical analysis of the
effects of LBOs on supermarket Pricing, Journal of Finance 50, 1095–1112.
Coles, Jeffrey L., Naveen D. Daniel, and Lalitha Naveen, 2008, Boards: Does one size fit all,
Journal of Financial Economics 87, 329–356.
DeAngelo, Harry, Linda DeAngelo, and Douglas J. Skinner, 2004, Are dividends disappearing?
Dividend concentration and the consolidation of earnings, Journal of Financial
Economics 72, 425–456.
Dittmar, Amy and Jan Mahrt-Smith, 2007, Corporate governance and the value of cash holdings,
Journal of Financial Economics 83, 599–634.
Fama, Eugene F., and Kenneth R. French, 2001, Disappearing dividends: Changing firm
characteristics or lower propensity to pay?, Journal of Financial Economics 1, 3–43.
Fama, Eugene F., and Kenneth R. French, 2002, Testing trade‐off and pecking order predictions
about dividends and debt, Review of Financial Studies 15, 1–33.
Faulkender, Michael, and Rong Wang, 2006, Corporate financial policy and the value of cash,
Journal of Finance 61, 1957–1990.
Fenn, George W., and Nellie Liang, 2001, Corporate payout policy and managerial stock
incentives, Journal of Financial Economics 60, 45–72.
Flannery, Mark J., and Kasturi P. Rangan, 2006, Partial adjustment toward target capital
structures, Journal of Financial Economics 79, 469–506.
Foley, C. Fritz, Jay C. Hartzell, Sheridan Titman, and Garry Twite, 2007, Why do firms hold so
much cash? A tax-based explanation, Journal of Financial Economics 86, 579–607.
32
Franzen, Laurel, Kimberly Rodgers Cornaggia, and Timothy T. Simin, 2009, Capital structure
and the changing role of off-balance-sheet lease financing, Working paper.
Gonnard, Eric, Eun Jung Kim, and Isabelle Ynesta, 2008, Recent trends in institutional investors
statistics. Financial Market Trends 95, 1–22.
Gregory, Alan, 2005, The long run abnormal performance of UK acquirers and the free cash flow
hypothesis, Journal of Business Finance & Accounting 32, 777–814.
Griffin, James M., 1988, A test of the free cash flow hypothesis: Results from the petroleum
industry, The Review of Economics and Statistics 70, 76–82.
Grullon, Gustavo, and Roni Michaely, 2002, Dividends, share repurchases, and the substitution
hypothesis, Journal of Finance 57, 1649–1684.
Harford, Jarrad, Sattar Mansi, and William Maxwell, 2008, Corporate governance and a firm’s
cash holdings, Journal of Financial Economics 87, 535–555.
Hartzell, Jay C., and Laura T. Starks, 2003, Institutional investors and executive compensation,
Journal of Finance 58, 2351–2374.
Haushalter, David, Sandy Klasa, and William F. Maxwell, 2007, The influence of product market
dynamics on a firm's cash holdings and hedging behavior, Journal of Financial
Economics 84, 797–825.
Jensen, Michael C., 1986, Agency costs of free cash flow, corporate finance, and takeovers,
American Economic Review 76, 323–329.
Jensen, Michael C., and William H. Meckling, 1976, Theory of the firm: Managerial behavior,
agency costs and ownership structure, Journal of Financial Economics 3, 305–360.
33
Kalcheva, Ivalina, and Karl V. Lins, 2007, International evidence on cash holdings and expected
managerial agency problems, Review of Financial Studies 20, 1087–1112.
Karpoff, Johnathon M., Paul H. Malatesta, and Ralph A. Walkling, 1996, Corporate governance
and shareholder initiatives: Empirical evidence, Journal of Financial Economics 42, 365–
396.
Keith, M. Howe, Jia He, and G. Wenchi Kao, 1992, One-time cash flow announcements and free
cash-flow theory: Share repurchases and special dividends, Journal of Finance 47, 1963–
1975.
Lang, Larry H.P., and Robert H. Litzenberger, 1989, Dividend announcements: Cash flow
signalling vs. free cash flow hypothesis?, Journal of Financial Economics 24, 181–191.
Lang, Larry H.P., René M. Stulz, and Ralph A. Walkling, 1991, A test of the free cash flow
hypothesis: The case of bidder returns, Journal of Financial Economics 29, 315–335.
Lehn, Kenneth, and Annette Poulsen, 1989, Free cash flow and stockholder gains in going
private transactions, Journal of Finance 44, 771–787.
Lemmon, Michael L., Michael R. Roberts, and Jamie F. Zender, 2008, Back to the beginning:
Persistence and the cross-section of corporate capital structure, Journal of Finance 63,
1575–1608.
McCahery, Joseph A., Laura T. Starks, and Zacharias Sautner, 2010, Behind the Scenes: The
Corporate Governance Preferences of Institutional Investors, Working paper.
Myers, Stewart C., 1977, Determinants of corporate borrowing, Journal of Financial Economics
5, 147–175.
34
Myers, Stewart C., and Nicholas S. Majluf, 1984, Corporate financing and investment decisions
when firms have information that investors do not have, Journal of Financial Economics
13, 187–221.
O’Neill, Michele, and Judith Swisher, 2003, Institutional investors and information asymmetry:
An event study of self-tender offers, Financial Review 38, 197–211.
Opler, Tim, Lee Pinkowitz, René M. Stulz, and Rohan Williamson, 1999, The determinants and
implications of corporate cash holdings, Journal of Financial Economics 52, 3–46.
Richardson, Scott, 2006, Over-investment of free cash flow, Review of Accounting Studies 11,
159–189.
Stulz, René M., 1990, Managerial discretion and optimal financing policies, Journal of Financial
Economics 26, 3–27.
Szewczyk, Samuel H., George P. Tsetsekos, and Raj Varma, 1992, Institutional ownership and
the liquidity of common stock offerings, Financial Review 27, 211–225.
Velury, Uma, and David S. Jenkins, 2006, Institutional ownership and the quality of earnings,
Journal of Business Research 59, 1043–1051.
35
Figure 1. Hypotheses. This figure plots our hypotheses. Hypothesis 1 (H1): Higher ownership
controlled by five largest institutional investors (Top5 holdings) will be associated with lower
leverage. H2: Higher ownership controlled by five largest institutional investors will be
associated with lower payout ratio. H3: Higher ownership controlled by five largest institutional
investors will be associated with higher cash holdings. H4: There will be a negative relationship
between cash holdings and leverage. H5: There will be a negative relationship between cash
holdings and payout ratio. H6: Cash holdings will be positively associated with firm value.
36
Table I
Institutional Ownership
This table shows the institutional ownership from 1980 to 2009. The sample consists of all Compustat
firm-year observations during the period 1980 through 2009. We eliminate financial firms (with
Standard Industrial Classification (SIC) codes 6000-6999) and public utility firms (with SIC codes
4900-4999). Firms must have positive assets (Compustat item AT), positive sales (Compustat item
SALE), positive common shares outstanding (Compustat item CSHO), positive closing share price at
the end of the fiscal year (Compustat item PRCC_F) and be incorporated in the United States of
America. Industry (high-tech vs. non high-tech) is defined according to the definition of TechAmerica.
Institutional ownership is the percentage of shares held by institutional investors. We assume that firms
not covered by Thomson Reuters have no institutional investors. N is the number of observations.
High-tech firms ratio is the number of high-tech firms over the number of all firms in each year.
Year
1980
1981
1982
1983
1984
1985
1986
1987
1988
1989
1990
1991
1992
1993
1994
1995
1996
1997
1998
1999
2000
2001
2002
2003
2004
2005
2006
2007
2008
2009
N
3,678
4,194
4,169
4,482
4,517
4,423
4,552
4,670
4,427
4,265
4,195
4,233
4,441
4,768
5,026
5,619
6,126
6,183
5,997
6,066
5,944
5,489
5,075
4,756
4,656
4,481
4,368
4,221
3,925
3,512
All firms
Mean
0.073
0.070
0.077
0.091
0.098
0.112
0.120
0.125
0.132
0.141
0.152
0.160
0.177
0.180
0.198
0.198
0.193
0.216
0.223
0.214
0.223
0.246
0.277
0.304
0.358
0.389
0.423
0.456
0.459
0.457
Median
0.000
0.000
0.000
0.000
0.000
0.003
0.004
0.007
0.013
0.014
0.018
0.018
0.041
0.061
0.074
0.061
0.059
0.084
0.088
0.077
0.075
0.083
0.121
0.176
0.256
0.335
0.388
0.443
0.469
0.479
Non high-tech firms
N
Mean Median
3,136
0.073
0.000
3,529
0.070
0.000
3,467
0.076
0.000
3,628
0.089
0.000
3,600
0.096
0.000
3,477
0.111
0.002
3,569
0.117
0.002
3,647
0.123
0.007
3,451
0.132
0.011
3,327
0.141
0.013
3,271
0.153
0.014
3,317
0.161
0.016
3,467
0.181
0.044
3,704
0.184
0.068
3,920
0.200
0.078
4,289
0.202
0.064
4,609
0.198
0.060
4,615
0.223
0.090
4,467
0.233
0.099
4,445
0.222
0.088
4,285
0.231
0.081
3,966
0.254
0.098
3,672
0.288
0.142
3,441
0.316
0.204
3,361
0.373
0.289
3,252
0.403
0.362
3,217
0.438
0.425
3,118
0.470
0.482
2,920
0.469
0.488
2,639
0.464
0.498
37
High-tech firms
N
Mean
Median
542
0.073
0.000
665
0.070
0.000
702
0.083
0.000
854
0.099
0.000
917
0.106
0.003
946
0.116
0.008
983
0.131
0.014
1,023
0.132
0.011
976
0.130
0.016
938
0.139
0.018
924
0.149
0.032
916
0.156
0.029
974
0.162
0.034
1,064
0.168
0.044
1,106
0.188
0.066
1,330
0.186
0.046
1,517
0.181
0.057
1,568
0.193
0.073
1,530
0.193
0.053
1,621
0.191
0.055
1,659
0.202
0.060
1,523
0.224
0.059
1,403
0.248
0.084
1,315
0.274
0.116
1,295
0.319
0.178
1,229
0.351
0.250
1,151
0.380
0.297
1,103
0.416
0.358
1,005
0.430
0.421
873
0.436
0.449
High-tech
firms ratio
0.147
0.159
0.168
0.191
0.203
0.214
0.216
0.219
0.220
0.220
0.220
0.216
0.219
0.223
0.220
0.237
0.248
0.254
0.255
0.267
0.279
0.277
0.276
0.276
0.278
0.274
0.264
0.261
0.256
0.249
Table II
Mean and Median Cash Ratios from 1980 to 2009
This table shows cash holdings from 1980 to 2009. Market cash ratio is cash and short-term
investments (Compustat item CHE) over market value of assets (book value of assets (Compustat item
AT) – common equity (Compustat item CEQ) + common shares outstanding (Compustat item CSHO) *
closing share price at the end of the fiscal year (Compustat item PRCC_F)). Book cash ratio is cash and
short-term investments over book value of assets. Industry (high-tech vs. non high-tech) is defined
according to the definition of TechAmerica.
Year
1980
1981
1982
1983
1984
1985
1986
1987
1988
1989
1990
1991
1992
1993
1994
1995
1996
1997
1998
1999
2000
2001
2002
2003
2004
2005
2006
2007
2008
2009
Market cash ratio
All firms
Non high-tech firms High-tech firms
Mean Median Mean
Median
Mean Median
0.076 0.043
0.077
0.044
0.069 0.036
0.092 0.048
0.092
0.047
0.094 0.054
0.092 0.050
0.093
0.049
0.087 0.050
0.095 0.056
0.094
0.053
0.099 0.066
0.095 0.049
0.094
0.048
0.099 0.051
0.088 0.046
0.085
0.044
0.098 0.054
0.095 0.048
0.091
0.046
0.109 0.066
0.104 0.050
0.098
0.046
0.127 0.071
0.094 0.044
0.088
0.040
0.116 0.060
0.089 0.041
0.083
0.037
0.109 0.054
0.098 0.043
0.090
0.039
0.123 0.064
0.091 0.044
0.085
0.040
0.112 0.067
0.091 0.046
0.083
0.041
0.119 0.070
0.089 0.046
0.080
0.040
0.118 0.079
0.089 0.042
0.081
0.036
0.114 0.078
0.081 0.039
0.074
0.032
0.105 0.070
0.091 0.045
0.082
0.037
0.119 0.078
0.094 0.044
0.084
0.035
0.121 0.079
0.100 0.041
0.089
0.033
0.133 0.073
0.082 0.035
0.076
0.029
0.101 0.054
0.120 0.045
0.103
0.033
0.165 0.089
0.123 0.050
0.103
0.038
0.173 0.097
0.149 0.062
0.122
0.048
0.218 0.134
0.106 0.058
0.093
0.046
0.140 0.099
0.106 0.062
0.092
0.049
0.143 0.106
0.106 0.062
0.093
0.050
0.141 0.104
0.103 0.060
0.092
0.049
0.135 0.094
0.110 0.061
0.098
0.049
0.144 0.101
0.151 0.080
0.130
0.063
0.210 0.136
0.132 0.085
0.117
0.073
0.177 0.130
38
Book cash ratio
All firms
Non high-tech firms High-tech firms
Mean Median Mean
Median
Mean Median
0.106 0.055
0.103
0.055
0.123 0.056
0.121 0.058
0.114
0.054
0.158 0.092
0.122 0.064
0.115
0.060
0.157 0.092
0.156 0.085
0.142
0.075
0.218 0.144
0.138 0.068
0.130
0.063
0.170 0.090
0.140 0.069
0.131
0.063
0.176 0.104
0.154 0.079
0.145
0.070
0.188 0.119
0.153 0.074
0.140
0.066
0.196 0.122
0.138 0.066
0.127
0.059
0.176 0.098
0.135 0.060
0.126
0.053
0.167 0.091
0.132 0.059
0.121
0.052
0.172 0.094
0.152 0.069
0.141
0.060
0.190 0.122
0.159 0.076
0.145
0.065
0.210 0.133
0.169 0.081
0.148
0.066
0.242 0.173
0.153 0.069
0.134
0.055
0.222 0.164
0.167 0.069
0.140
0.052
0.254 0.186
0.188 0.083
0.160
0.063
0.270 0.206
0.187 0.086
0.159
0.061
0.268 0.212
0.175 0.071
0.149
0.052
0.251 0.185
0.197 0.079
0.161
0.053
0.297 0.225
0.201 0.083
0.166
0.054
0.290 0.225
0.203 0.096
0.171
0.066
0.285 0.222
0.203 0.104
0.168
0.074
0.293 0.244
0.220 0.125
0.187
0.090
0.309 0.261
0.234 0.140
0.203
0.102
0.315 0.277
0.231 0.141
0.204
0.104
0.301 0.266
0.231 0.132
0.208
0.099
0.295 0.244
0.223 0.123
0.200
0.092
0.290 0.236
0.204 0.114
0.182
0.091
0.268 0.215
0.226 0.146
0.205
0.121
0.291 0.250
Table III
Mean and Median Debt Ratios from 1980 to 2009
This table shows leverage from 1980 to 2009. Market leverage is debt (the sum of long-term debt
(Compustat item DLTT) and debt in current liabilities (Compustat item DLC)) over market value of
assets (book value of assets (Compustat item AT) – common equity (Compustat item CEQ) + common
shares outstanding (Compustat item CSHO) * closing share price at the end of the fiscal year
(Compustat item PRCC_F)). Book leverage is debt over book value of assets. Industry (high-tech vs.
non high-tech) is defined according to the definition of TechAmerica.
Year
1980
1981
1982
1983
1984
1985
1986
1987
1988
1989
1990
1991
1992
1993
1994
1995
1996
1997
1998
1999
2000
2001
2002
2003
2004
2005
2006
2007
2008
2009
Market leverage
All firms
Non high-tech firms
Mean Median
Mean
Median
0.231 0.198
0.243
0.213
0.231 0.200
0.244
0.219
0.231 0.195
0.245
0.215
0.187 0.141
0.203
0.161
0.209 0.171
0.224
0.188
0.207 0.166
0.222
0.185
0.205 0.162
0.219
0.180
0.217 0.175
0.231
0.192
0.220 0.176
0.232
0.191
0.221 0.176
0.234
0.192
0.237 0.191
0.251
0.208
0.203 0.146
0.217
0.165
0.180 0.125
0.194
0.143
0.157 0.109
0.171
0.127
0.165 0.119
0.183
0.142
0.165 0.111
0.187
0.143
0.157 0.099
0.177
0.127
0.161 0.100
0.180
0.129
0.192 0.130
0.214
0.164
0.184 0.107
0.216
0.160
0.191 0.112
0.221
0.153
0.183 0.109
0.207
0.143
0.188 0.123
0.209
0.156
0.150 0.084
0.171
0.117
0.132 0.074
0.149
0.104
0.132 0.077
0.146
0.097
0.132 0.077
0.145
0.096
0.147 0.086
0.161
0.107
0.201 0.136
0.219
0.163
0.159 0.094
0.173
0.116
High-tech firms
Mean Median
0.167 0.124
0.163 0.104
0.159 0.104
0.117 0.056
0.152 0.102
0.152 0.101
0.155 0.102
0.168 0.115
0.176 0.127
0.177 0.122
0.190 0.126
0.154 0.086
0.131 0.061
0.106 0.046
0.104 0.044
0.095 0.027
0.098 0.030
0.103 0.032
0.126 0.048
0.096 0.019
0.115 0.026
0.120 0.034
0.133 0.048
0.095 0.020
0.087 0.015
0.093 0.022
0.096 0.029
0.109 0.032
0.148 0.059
0.117 0.044
39
All firms
Mean Median
0.269 0.245
0.263 0.231
0.276 0.237
0.252 0.207
0.267 0.221
0.282 0.235
0.289 0.244
0.289 0.247
0.291 0.248
0.296 0.258
0.293 0.247
0.268 0.220
0.249 0.198
0.233 0.184
0.233 0.189
0.247 0.195
0.247 0.180
0.263 0.192
0.292 0.214
0.295 0.211
0.296 0.190
0.327 0.194
0.327 0.190
0.313 0.179
0.295 0.159
0.291 0.156
0.298 0.158
0.305 0.165
0.328 0.186
0.287 0.157
Book leverage
Non high-tech firms
Mean
Median
0.272
0.250
0.270
0.239
0.283
0.245
0.263
0.219
0.275
0.234
0.291
0.248
0.301
0.259
0.300
0.262
0.300
0.261
0.305
0.270
0.301
0.260
0.279
0.239
0.260
0.221
0.247
0.210
0.252
0.219
0.270
0.230
0.268
0.215
0.284
0.228
0.310
0.251
0.317
0.253
0.323
0.240
0.345
0.238
0.338
0.230
0.327
0.217
0.310
0.196
0.298
0.186
0.300
0.186
0.308
0.195
0.332
0.219
0.290
0.182
High-tech firms
Mean Median
0.250 0.221
0.226 0.171
0.243 0.178
0.204 0.132
0.234 0.171
0.249 0.184
0.247 0.176
0.252 0.187
0.259 0.197
0.264 0.197
0.265 0.181
0.230 0.136
0.209 0.114
0.184 0.089
0.167 0.085
0.171 0.074
0.183 0.067
0.200 0.080
0.239 0.104
0.235 0.077
0.224 0.060
0.278 0.074
0.300 0.079
0.275 0.056
0.255 0.047
0.272 0.056
0.295 0.070
0.296 0.072
0.316 0.090
0.280 0.077
Table IV
Mean and Median Payout and Book-to-Market Ratios from 1980 to 2009
This table illustrates payout and book-to-market ratios from 1980 to 2009. Payout ratio is a sum of
common stock dividends (Compustat item DVC) and absolute value of the difference between
purchase of common and preferred stock (Compustat item PRSTKC) and preferred stock redemption
value (Compustat item PSTKRV) divided by market value of equity (common shares outstanding
(Compustat item CSHO) * closing share price at the end of the fiscal year (Compustat item PRCC_F)).
Book-to-market ratio is book value of assets (Compustat item AT) divided by market value of assets
(book value of assets – common equity (Compustat item CEQ) + common shares outstanding * closing
share price at the end of the fiscal). Industry (high-tech vs. non high-tech) is defined according to the
definition of TechAmerica.
Year
1980
1981
1982
1983
1984
1985
1986
1987
1988
1989
1990
1991
1992
1993
1994
1995
1996
1997
1998
1999
2000
2001
2002
2003
2004
2005
2006
2007
2008
2009
Payout ratio
All firms
Non high-tech firms High-tech firms
Mean Median Mean
Median Mean Median
0.055 0.023
0.059
0.029
0.033 0.002
0.053 0.017
0.057
0.024
0.030 0.000
0.050 0.014
0.055
0.019
0.027 0.000
0.039 0.007
0.043
0.012
0.021 0.000
0.045 0.008
0.050
0.013
0.028 0.000
0.043 0.006
0.047
0.011
0.029 0.000
0.043 0.004
0.047
0.007
0.030 0.000
0.049 0.006
0.051
0.010
0.040 0.000
0.048 0.005
0.050
0.008
0.040 0.000
0.047 0.004
0.049
0.007
0.040 0.000
0.054 0.007
0.056
0.010
0.047 0.000
0.044 0.002
0.047
0.005
0.035 0.000
0.039 0.001
0.041
0.003
0.029 0.000
0.034 0.000
0.037
0.002
0.024 0.000
0.037 0.000
0.040
0.002
0.025 0.000
0.037 0.000
0.039
0.001
0.027 0.000
0.036 0.000
0.039
0.001
0.026 0.000
0.037 0.001
0.038
0.002
0.032 0.000
0.047 0.005
0.048
0.007
0.042 0.000
0.045 0.003
0.050
0.008
0.029 0.000
0.052 0.002
0.056
0.006
0.041 0.000
0.044 0.001
0.046
0.002
0.040 0.000
0.045 0.001
0.046
0.002
0.044 0.000
0.035 0.000
0.035
0.001
0.035 0.000
0.034 0.000
0.035
0.001
0.031 0.000
0.037 0.000
0.038
0.003
0.035 0.000
0.037 0.001
0.037
0.003
0.038 0.000
0.044 0.002
0.044
0.004
0.044 0.000
0.060 0.010
0.061
0.013
0.059 0.004
0.031 0.001
0.030
0.002
0.033 0.000
40
Book-to-market ratio
All firms
Non high-tech firms High-tech firms
Mean Median Mean
Median
Mean Median
0.851
0.895
0.885
0.945
0.656 0.609
0.879
0.906
0.912
0.949
0.702 0.674
0.854
0.873
0.895
0.918
0.649 0.605
0.729
0.733
0.772
0.783
0.546 0.503
0.791
0.809
0.822
0.843
0.667 0.653
0.741
0.755
0.768
0.784
0.642 0.619
0.719
0.726
0.738
0.748
0.649 0.646
0.768
0.780
0.785
0.801
0.707 0.714
0.761
0.778
0.773
0.788
0.717 0.717
0.748
0.753
0.760
0.767
0.705 0.689
0.832
0.834
0.843
0.854
0.794 0.777
0.741
0.733
0.754
0.754
0.695 0.664
0.702
0.684
0.715
0.710
0.656 0.609
0.647
0.628
0.665
0.654
0.581 0.535
0.686
0.673
0.709
0.703
0.604 0.562
0.644
0.618
0.683
0.673
0.517 0.459
0.625
0.602
0.657
0.643
0.528 0.488
0.613
0.587
0.644
0.626
0.519 0.478
0.691
0.678
0.723
0.724
0.595 0.547
0.636
0.619
0.705
0.726
0.446 0.351
0.738
0.710
0.783
0.778
0.624 0.547
0.704
0.667
0.735
0.715
0.622 0.549
0.777
0.766
0.793
0.789
0.736 0.701
0.595
0.576
0.631
0.628
0.502 0.462
0.556
0.541
0.578
0.569
0.499 0.476
0.560
0.543
0.575
0.563
0.520 0.501
0.551
0.540
0.566
0.556
0.512 0.496
0.598
0.576
0.615
0.593
0.550 0.534
0.825
0.810
0.835
0.823
0.799 0.748
0.696
0.683
0.712
0.707
0.647 0.620
Table V
Determinants of Market and Book Leverage
This table presents the results of least squares regressions where the dependent variable is either market
or book leverage. Market leverage is debt (the sum of long-term debt (Compustat item DLTT) and debt
in current liabilities (Compustat item DLC)) over market value of assets (book value of assets
(Compustat item AT) – common equity (Compustat item CEQ) + common shares outstanding
(Compustat item CSHO) * closing share price at the end of the fiscal year (Compustat item PRCC_F)).
Book leverage is debt over book value of assets. Top5 holdings is the ownership controlled by five
largest institutional investors. We assume that firms not covered by Thomson Reuters have no
institutional investors. High-tech dummy equals one if a firm is from high-tech (defined according to
the definition of TechAmerica), zero otherwise. Lease/Assets is equal to the present value of noncancelable operating leases divided by book value of assets. ln(Assets) is the natural logarithm of book
value of assets (in millions of U.S. dollars (converted into 2009 constant dollars using the GDP
deflator)). Book-to-market is book value of assets divided by market value of assets. EBIT/Assets is
earnings before interests and taxes (the sum of income before extraordinary items (Compustat item IB),
interest and related expense (Compustat item XINT), and income taxes (Compustat item TXT)) divided
by book value of assets. PPE/Assets is net property, plant, and equipment (Compustat item PPENT)
divided by book value of assets. R&D/Assets is research and development expense (Compustat item
XRD) divided by book value of assets. R&D dummy is equal to one when R&D expense is unreported
in Compustat and zero otherwise. p-values based on standard errors robust to clustering by firm are
reported in parentheses.
41
Table V (continued)
Dependant variable
Model 1
Market
leveraget
Market leveraget
Book leveraget
Top5 holdingst
High-tech dummyt
–0.139
(0.000)
–0.018
(0.000)
Model 2
Model 3
Model 4
Model 5
Model 6
Book
Market
Book
Market
Book
leveraget
leveraget+1 leveraget+1 leveraget+1 leveraget+1
0.834
(0.000)
0.776
(0.000)
–0.175
–0.134
–0.201
–0.024
–0.074
(0.000)
(0.000)
(0.000)
(0.000)
(0.000)
–0.019
–0.019
–0.017
–0.005
–0.001
(0.001)
(0.000)
(0.004)
(0.000)
(0.663)
Lease/Assetst
ln(Assets)t
Book-to-market ratiot
EBIT/Assetst
PPE/Assetst
R&D/Assetst
R&D dummyt
Year fixed effects
R2
Adjusted R2
Number of observations
0.013
(0.000)
0.127
(0.000)
–0.048
(0.000)
0.153
(0.000)
–0.153
(0.000)
0.032
(0.000)
Yes
0.236
0.236
141,693
0.004
(0.001)
–0.069
(0.000)
–0.315
(0.000)
0.235
(0.000)
–0.372
(0.000)
0.051
(0.000)
Yes
0.207
0.207
141,693
0.012
(0.000)
0.113
(0.000)
–0.048
(0.000)
0.144
(0.000)
–0.177
(0.000)
0.033
(0.000)
Yes
0.209
0.208
126,464
42
0.004
(0.002)
–0.054
(0.000)
–0.314
(0.000)
0.210
(0.000)
–0.380
(0.000)
0.050
(0.000)
Yes
0.168
0.168
126,474
0.001
(0.000)
–0.009
(0.000)
–0.005
(0.000)
0.021
(0.000)
–0.049
(0.000)
0.008
(0.000)
Yes
0.703
0.703
126,346
–0.001
(0.179)
–0.012
(0.020)
–0.076
(0.000)
0.033
(0.000)
–0.089
(0.000)
0.014
(0.000)
Yes
0.574
0.574
126,356
Model 7
Market
leveraget
–0.139
(0.000)
–0.018
(0.000)
–0.003
(0.631)
0.013
(0.000)
0.127
(0.000)
–0.048
(0.000)
0.153
(0.000)
–0.153
(0.000)
0.032
(0.000)
Yes
0.236
0.236
141,693
Table VI
Determinants of Payout Ratio
This table presents the results of least squares regressions where the dependent variable is payout ratio.
Payout ratio is a sum of common stock dividends (Compustat item DVC) and absolute value of the
difference between purchase of common and preferred stock (Compustat item PRSTKC) and preferred
stock redemption value (Compustat item PSTKRV) divided by market value of equity (common shares
outstanding (Compustat item CSHO) * closing share price at the end of the fiscal year (Compustat item
PRCC_F)). Top5 holdings is the ownership controlled by five largest institutional investors. We assume
that firms not covered by Thomson Reuters have no institutional investors. High-tech dummy equals
one if a firm is from high-tech (defined according to the definition of TechAmerica), zero otherwise.
ln(Assets) is the natural logarithm of book value of assets (in millions of U.S. dollars (Compustat item
AT) (converted into 2009 constant dollars using the GDP deflator)). Book-to-market is book value of
assets divided by market value of assets (book value of assets (Compustat item AT) – common equity
(Compustat item CEQ) + common shares outstanding (Compustat item CSHO) * closing share price at
the end of the fiscal year (Compustat item PRCC_F)). Assets growth is the annual growth rate of book
value of assets. EBIT/Assets is earnings before interests and taxes (the sum of income before
extraordinary items (Compustat item IB), interest and related expense (Compustat item XINT), and
income taxes (Compustat item TXT)) divided by book value of assets. Book leverage is debt (the sum
of long-term debt (Compustat item DLTT) and debt in current liabilities (Compustat item DLC)) over
book value of assets. p-values based on standard errors robust to clustering by firm are reported in
parentheses.
Dependant variable
Top5 holdingst
High-tech dummyt
ln(Assets)t
Book-to-market ratiot
Assets growtht
EBIT/Assetst
Book leveraget
Year fixed effects
R2
Adjusted R2
Number of observations
Model 1
Payout ratiot
Model 2
Payout ratiot
Model 3
Payout ratiot+1
Model 4
Payout ratiot+1
–0.055
(0.000)
–0.005
(0.000)
0.005
(0.000)
0.016
(0.000)
–0.025
(0.000)
–0.003
(0.012)
0.026
(0.000)
Yes
0.051
0.051
121,635
–0.062
(0.000)
–0.007
(0.000)
0.005
(0.000)
0.014
(0.000)
–0.025
(0.000)
–0.011
(0.000)
–0.056
(0.000)
–0.005
(0.000)
0.005
(0.000)
0.012
(0.000)
–0.015
(0.000)
–0.009
(0.000)
0.023
(0.000)
Yes
0.041
0.041
108,212
–0.061
(0.000)
–0.007
(0.000)
0.005
(0.000)
0.011
(0.000)
–0.015
(0.000)
–0.016
(0.000)
Yes
0.040
0.040
122,011
43
Yes
0.034
0.034
108,538
Table VII
Determinants of Cash Holdings
This table presents the results of least squares regressions where the dependent variable is either market
or book cash ratio. Market cash ratio is cash and short-term investments (Compustat item CHE)
divided by market value of assets (book value of assets (Compustat item AT) – common equity
(Compustat item CEQ) + common shares outstanding (Compustat item CSHO) * closing share price at
the end of the fiscal year (Compustat item PRCC_F)). Book cash ratio is cash and short-term
investments divided by book value of assets. Top5 holdings is the ownership controlled by five largest
institutional investors. We assume that firms not covered by Thomson Reuters have no institutional
investors. High-tech dummy equals one if a firm is from high-tech (defined according to the definition
of TechAmerica), zero otherwise. ln(Assets) is the natural logarithm of book value of assets (in millions
of U.S. dollars (converted into 2009 constant dollars using the GDP deflator)). NWC/Assets is the
difference between working capital (Compustat item WCAP) and cash and short-term investments
divided by book value of assets. Industry sigma is the mean of the standard deviations of cash flow
(operating income before depreciation (Compustat item OIBDP) – interest and related expense
(Compustat item XINT) – income taxes (Compustat item TXT)) to book value of assets ratio over 10
years (if there are at least three observations) for firms in the same industry, as defined by the two-digit
SIC code. HHI is Herfindahl-Hirschman Index and is calculated using sales data of individual firms in
the same industry, as defined by the four-digit SIC code. FCF/Assets is free cash flow (operating
income before depreciation – interest and related expense – income taxes – common stock dividends
(Compustat item DVC)) to book value of assets ratio. Book-to-market is book value of assets divided
by market value of assets. CAPEX /Assets is capital expenditures (Compustat item CAPX) to book
value of assets ratio. R&D/Assets is research and development expense (Compustat item XRD) divided
by book value of assets. R&D dummy is equal to one when R&D expense is unreported in Compustat
and zero otherwise. Book leverage is debt (the sum of long-term debt (Compustat item DLTT) and debt
in current liabilities (Compustat item DLC)) over book value of assets. Payout ratio is a sum of
common stock dividends and absolute value of the difference between purchase of common and
preferred stock (Compustat item PRSTKC) and preferred stock redemption value (Compustat item
PSTKRV) divided by market value of equity (common shares outstanding * closing share price at the
end of the fiscal year). Dividend dummy is equal to one if common stock dividends are positive and
zero otherwise. Debt issuance/Assets is the difference between long-term debt issuance (Compustat
item DLTIS) and long-term debt reduction (Compustat item DLTR) divided by book value of assets.
Equity issuance/Assets is the difference between sale of common and preferred stock (Compustat item
SSTK) and purchase of common and preferred stock divided by book value of assets.
Acquisitions/Assets is acquisitions (Compustat item AQC) divided by book value of assets. p-values
based on standard errors robust to clustering by firm are reported in parentheses.
44
Table VII (continued)
Model 1
Model 2
Model 3
Model 4
Model 5
Model 6
Model 7
Market cash Book cash Market cash Book cash Market cash Book cash Market cash
Dependant variable
ratiot
ratiot
ratiot
ratiot
ratiot+1
ratiot+1
ratiot
Top5 holdingst
0.043
0.089
0.042
0.086
0.032
0.077
0.042
(0.000)
(0.000)
(0.000)
(0.000)
(0.000)
(0.000)
(0.000)
High-tech dummyt
0.024
0.021
0.024
0.021
0.022
0.017
0.023
(0.000)
(0.000)
(0.000)
(0.000)
(0.000)
(0.000)
(0.000)
ln(Assets)t
–0.009
–0.011
–0.009
–0.011
–0.008
–0.010
–0.009
(0.000)
(0.000)
(0.000)
(0.000)
(0.000)
(0.000)
(0.000)
NWC/Assetst
–0.071
–0.109
–0.071
–0.110
–0.062
–0.101
–0.070
(0.000)
(0.000)
(0.000)
(0.000)
(0.000)
(0.000)
(0.000)
Industry sigmat
0.002
0.004
0.002
0.004
0.002
0.004
0.002
(0.000)
(0.000)
(0.000)
(0.000)
(0.000)
(0.000)
(0.000)
HHIt
–0.018
(0.000)
FCF/Assetst
0.006
0.031
0.006
0.031
0.013
0.014
0.006
(0.005)
(0.000)
(0.008)
(0.000)
(0.000)
(0.001)
(0.007)
Book-to-market ratiot
0.119
–0.045
0.119
–0.044
0.079
–0.044
0.119
(0.000)
(0.000)
(0.000)
(0.000)
(0.000)
(0.000)
(0.000)
CAPEX/Assetst
–0.183
–0.403
–0.184
–0.407
–0.196
–0.370
–0.186
(0.000)
(0.000)
(0.000)
(0.000)
(0.000)
(0.000)
(0.000)
R&D/Assetst
0.114
0.330
0.114
0.330
0.105
0.409
0.112
(0.000)
(0.000)
(0.000)
(0.000)
(0.000)
(0.000)
(0.000)
R&D dummyt
–0.014
–0.019
–0.014
–0.019
–0.011
–0.015
–0.014
(0.000)
(0.000)
(0.000)
(0.000)
(0.000)
(0.000)
(0.000)
Book leveraget
–0.116
–0.181
–0.115
–0.180
–0.111
–0.166
–0.115
(0.000)
(0.000)
(0.000)
(0.000)
(0.000)
(0.000)
(0.000)
Payout ratiot
–0.025
–0.066
(0.002)
(0.000)
Dividend dummyt
–0.010
–0.025
–0.010
–0.024
–0.014
–0.026
–0.010
(0.000)
(0.000)
(0.000)
(0.000)
(0.000)
(0.000)
(0.000)
Debt issuance/Assetst
0.065
0.153
0.065
0.153
0.053
0.105
0.065
(0.000)
(0.000)
(0.000)
(0.000)
(0.000)
(0.000)
(0.000)
Equity issuance/Assetst
0.041
0.219
0.041
0.218
0.028
0.090
0.041
(0.000)
(0.000)
(0.000)
(0.000)
(0.000)
(0.000)
(0.000)
Acquisitions/Assetst
–0.117
–0.292
–0.117
–0.293
–0.126
–0.253
–0.116
(0.000)
(0.000)
(0.000)
(0.000)
(0.000)
(0.000)
(0.000)
Year fixed effects
Yes
Yes
Yes
Yes
Yes
Yes
Yes
R2
0.255
0.343
0.255
0.344
0.177
0.295
0.256
Adjusted R2
0.255
0.343
0.255
0.344
0.177
0.295
0.256
Number of observations
117,928
117,928
117,861
117,861
105,140
105,149
117,928
45
Table VIII
Determinants of Firm Value
This table presents the results of least squares regressions where the dependent variable is firm value
proxied by Tobin’s q. Q is Tobin’s q (market value of assets (book value of assets (Compustat item AT)
– common equity (Compustat item CEQ) + common shares outstanding (Compustat item CSHO) *
closing share price at the end of the fiscal year (Compustat item PRCC_F)) divided by book value of
assets). Top5 holdings is the ownership controlled by five largest institutional investors. We assume that
firms not covered by Thomson Reuters have no institutional investors. High-tech dummy equals one if
a firm is from high-tech (defined according to the definition of TechAmerica), zero otherwise.
ln(Assets) is the natural logarithm of book value of assets (in millions of U.S. dollars (converted into
2009 constant dollars using the GDP deflator)). Book leverage is debt (the sum of long-term debt
(Compustat item DLTT) and debt in current liabilities (Compustat item DLC)) over book value of
assets. Book cash ratio is cash and short-term investments (Compustat item CHE) divided by book
value of assets. EBIT/Assets is earnings before interests and taxes (the sum of income before
extraordinary items (Compustat item IB), interest and related expense (Compustat item XINT), and
income taxes (Compustat item TXT)) divided by book value of assets. PPE/Assets is net property,
plant, and equipment (Compustat item PPENT) divided by book value of assets. CAPEX /Assets is
capital expenditures (Compustat item CAPX) to book value of assets ratio. Dividend dummy is equal to
one if common stock dividends (Compustat item DVC) are positive and zero otherwise. Payout ratio is
a sum of common stock dividends and absolute value of the difference between purchase of common
and preferred stock (Compustat item PRSTKC) and preferred stock redemption value (Compustat item
PSTKRV) divided by market value of equity (common shares outstanding * closing share price at the
end of the fiscal year). p-values based on standard errors robust to clustering by firm are reported in
parentheses.
46
Table VIII (continued)
Dependant variable
Qt
Top5 holdingst
High-tech dummyt
ln(Assets)t
Book leveraget
Model 1
Qt
Model 2
Qt+1
0.263
(0.001)
0.147
(0.000)
–0.220
(0.000)
1.755
(0.000)
0.177
(0.030)
0.157
(0.000)
–0.214
(0.000)
1.647
(0.000)
2.842
(0.000)
–2.299
(0.000)
–1.213
(0.000)
4.590
(0.000)
0.672
(0.000)
2.384
(0.000)
–2.240
(0.000)
–0.816
(0.000)
2.207
(0.000)
0.598
(0.000)
Yes
0.404
0.404
141,693
Yes
0.353
0.353
126,499
Book cash ratiot* Top5 holdingst
Book cash ratiot
EBIT/Assetst
PPE/Assetst
CAPEX/Assetst
Dividend dummyt
Payout ratiot
Year fixed effects
R2
Adjusted R2
Number of observations
47
Model 6
Model 3
Model 4
Model 5
Qt+1
Qt+1
Qt
Qt
0.534
(0.000)
0.055
0.246
0.642
0.446
(0.208)
(0.003)
(0.000)
(0.000)
0.078
0.052
0.155
0.163
(0.000)
(0.135)
(0.000)
(0.000)
–0.103
–0.148
–0.220
–0.213
(0.000)
(0.000)
(0.000)
(0.000)
0.788
1.726
1.760
1.651
(0.000)
(0.000)
(0.000)
(0.000)
–2.157
–1.551
(0.000)
(0.000)
0.861
2.624
3.085
2.554
(0.000)
(0.000)
(0.000)
(0.000)
–0.937
–2.401
–2.293
–2.235
(0.000)
(0.000)
(0.000)
(0.000)
–0.148
–1.046
–1.211
–0.815
(0.000)
(0.000)
(0.000)
(0.000)
–0.261
3.990
4.573
2.195
(0.033)
(0.000)
(0.000)
(0.000)
0.266
0.660
0.589
(0.000)
(0.000)
(0.000)
–2.505
(0.000)
Yes
Yes
Yes
Yes
0.540
0.407
0.4043
0.3533
0.540
0.406
0.4042
0.3531
126,492
132,220
141,693
126,499
Table IX
Sensitivity Analysis
This table presents the sensitivity analysis. Using the coefficient estimates of Model 6 from Table VIII, we calculate hypothetical lead values
of Tobin’s q using different values of Top5 holdings and Book cash ratio, and mean values of other variables. Then we calculate the
difference between the computed numbers and the hypothetical lead value of Tobin’s q that is computed using mean values of all variables
including Top5 holdings and Book cash ratio. The mean values of Book cash ratio and Top5 holdings as well as the corresponding effects on
firm value are in bold style. The last column (Diff.) shows the impact on firm value than Book cash ratio increases from 0.000 to 0.400.
Top5 holdings
Book cash ratio
0.000
0.025
0.050
0.075
0.100
0.125
0.150
0.175
0.200
0.225
0.250
0.275
0.300
0.325
0.350
0.375
0.400
Diff.
0.000
–0.47
–0.40
–0.34
–0.28
–0.21
–0.15
–0.08
–0.02
0.04
0.11
0.17
0.23
0.30
0.36
0.43
0.49
0.55
1.02
0.025
–0.46
–0.39
–0.33
–0.27
–0.21
–0.14
–0.08
–0.02
0.05
0.11
0.17
0.24
0.30
0.36
0.42
0.49
0.55
1.01
0.050
–0.45
–0.38
–0.32
–0.26
–0.20
–0.14
–0.07
–0.01
0.05
0.11
0.17
0.24
0.30
0.36
0.42
0.48
0.55
0.99
0.075
–0.43
–0.37
–0.31
–0.25
–0.19
–0.13
–0.07
–0.01
0.05
0.11
0.18
0.24
0.30
0.36
0.42
0.48
0.54
0.98
0.100
–0.42
–0.36
–0.30
–0.24
–0.18
–0.12
–0.06
0.00
0.06
0.12
0.18
0.24
0.30
0.36
0.42
0.48
0.54
0.96
0.116
–0.42
–0.36
–0.30
–0.24
–0.18
–0.12
–0.06
0.00
0.06
0.12
0.18
0.24
0.30
0.36
0.42
0.47
0.53
0.95
0.125
–0.41
–0.35
–0.29
–0.24
–0.18
–0.12
–0.06
0.00
0.06
0.12
0.18
0.24
0.30
0.36
0.41
0.47
0.53
0.94
0.150
–0.40
–0.34
–0.28
–0.23
–0.17
–0.11
–0.05
0.01
0.06
0.12
0.18
0.24
0.30
0.35
0.41
0.47
0.53
0.93
0.175
–0.39
–0.33
–0.28
–0.22
–0.16
–0.10
–0.05
0.01
0.07
0.12
0.18
0.24
0.30
0.35
0.41
0.47
0.52
0.91
0.200
–0.38
–0.32
–0.27
–0.21
–0.15
–0.10
–0.04
0.01
0.07
0.13
0.18
0.24
0.29
0.35
0.41
0.46
0.52
0.90
0.225
–0.37
–0.31
–0.26
–0.20
–0.15
–0.09
–0.04
0.02
0.07
0.13
0.18
0.24
0.29
0.35
0.40
0.46
0.51
0.88
0.250
–0.36
–0.30
–0.25
–0.19
–0.14
–0.09
–0.03
0.02
0.08
0.13
0.19
0.24
0.29
0.35
0.40
0.46
0.51
0.87
0.275
–0.35
–0.29
–0.24
–0.19
–0.13
–0.08
–0.03
0.03
0.08
0.13
0.19
0.24
0.29
0.35
0.40
0.45
0.51
0.85
0.300
–0.33
–0.28
–0.23
–0.18
–0.13
–0.07
–0.02
0.03
0.08
0.14
0.19
0.24
0.29
0.35
0.40
0.45
0.50
0.84
0.325
–0.32
–0.27
–0.22
–0.17
–0.12
–0.07
–0.02
0.04
0.09
0.14
0.19
0.24
0.29
0.34
0.39
0.45
0.50
0.82
0.350
–0.31
–0.26
–0.21
–0.16
–0.11
–0.06
–0.01
0.04
0.09
0.14
0.19
0.24
0.29
0.34
0.39
0.44
0.49
0.80
0.375
–0.30
–0.25
–0.20
–0.15
–0.10
–0.05
0.00
0.04
0.09
0.14
0.19
0.24
0.29
0.34
0.39
0.44
0.49
0.79
0.400
–0.29
–0.24
–0.19
–0.14
–0.10
–0.05
0.00
0.05
0.10
0.15
0.19
0.24
0.29
0.34
0.39
0.44
0.48
0.77
0.425
–0.28
–0.23
–0.18
–0.14
–0.09
–0.04
0.01
0.05
0.10
0.15
0.20
0.24
0.29
0.34
0.39
0.43
0.48
0.76
0.450
–0.27
–0.22
–0.17
–0.13
–0.08
–0.03
0.01
0.06
0.10
0.15
0.20
0.24
0.29
0.34
0.38
0.43
0.48
0.74
0.475
–0.26
–0.21
–0.16
–0.12
–0.07
–0.03
0.02
0.06
0.11
0.15
0.20
0.24
0.29
0.33
0.38
0.43
0.47
0.73
48
Table X
Simultaneous Equation Model
This table presents the results of three-stage least squares regression where the dependent variables are
book leverage, payout ratio, and book cash ratio. Book leverage is debt (the sum of long-term debt
(Compustat item DLTT) and debt in current liabilities (Compustat item DLC)) over book value of
assets (Compustat item AT). Payout ratio is a sum of common stock dividends (Compustat item DVC)
and absolute value of the difference between purchase of common and preferred stock (Compustat item
PRSTKC) and preferred stock redemption value (Compustat item PSTKRV) divided by market value
of equity (common shares outstanding (Compustat item CSHO) * closing share price at the end of the
fiscal year (Compustat item PRCC_F)). Book cash ratio is cash and short-term investments (Compustat
item CHE) divided by book value of assets. Top5 holdings is the ownership controlled by five largest
institutional investors. We assume that firms not covered by Thomson Reuters have no institutional
investors. High-tech dummy equals one if a firm is from high-tech (defined according to the definition
of TechAmerica), zero otherwise. ln(Assets) is the natural logarithm of book value of assets (in millions
of U.S. dollars (converted into 2009 constant dollars using the GDP deflator)). NWC/Assets is the
difference between working capital (Compustat item WCAP) and cash and short-term investments
divided by book value of assets. Industry sigma is the mean of the standard deviations of cash flow
(operating income before depreciation (Compustat item OIBDP) – interest and related expense
(Compustat item XINT) – income taxes (Compustat item TXT)) to book value of assets ratio over 10
years (if there are at least three observations) for firms in the same industry, as defined by the two-digit
SIC code. HHI is Herfindahl-Hirschman Index and is calculated using sales data of individual firms in
the same industry, as defined by the four-digit SIC code. FCF/Assets is free cash flow (operating
income before depreciation – interest and related expense – income taxes – common stock dividends)
to book value of assets ratio. Book-to-market is book value of assets divided by market value of assets.
Assets growth is the annual growth rate of book value of assets. EBIT/Assets is earnings before interests
and taxes (the sum of income before extraordinary items (Compustat item IB), interest and related
expense, and income taxes) divided by book value of assets. PPE/Assets is net property, plant, and
equipment (Compustat item PPENT) divided by book value of assets. R&D/Assets is research and
development expense (Compustat item XRD) divided by book value of assets. R&D dummy is equal to
one when R&D expense is unreported in Compustat and zero otherwise. CAPEX /Assets is capital
expenditures (Compustat item CAPX) to book value of assets ratio. Dividend dummy is equal to one if
common stock dividends are positive and zero otherwise. Debt issuance/Assets is the difference
between long-term debt issuance (Compustat item DLTIS) – long-term debt reduction (Compustat item
DLTR) divided by book value of assets. Equity issuance/Assets is the difference between sale of
common and preferred stock (Compustat item SSTK) and purchase of common and preferred stock
divided by book value of assets. Acquisitions/Assets is acquisitions (Compustat item AQC) divided by
book value of assets. p-values are reported in parentheses.
49
Table X (continued)
Dependant variable
Book leveraget
Book leveraget
Payout ratiot
0.042
(0.000)
–0.201
(0.000)
–0.006
(0.020)
–0.003
(0.000)
–0.051
(0.000)
–0.003
(0.000)
0.005
(0.000)
–0.099
(0.000)
0.015
(0.000)
–0.025
(0.000)
0.002
(0.131)
Payout ratiot
Top5 holdingst
High-tech dummyt
Ln(Assets)t
NWC/Assetst
Industry sigmat
HHIt
FCF/Assetst
Book-to-market ratiot
Assets growtht
EBIT/Assetst
PPE/Assetst
R&D/Assetst
R&D dummyt
–0.299
(0.000)
0.398
(0.000)
–0.302
(0.000)
0.045
(0.000)
0.462
(0.000)
–0.001
(0.487)
–0.059
(0.000)
0.005
(0.000)
–0.123
(0.000)
0.249
(0.000)
–0.300
(0.000)
CAPEX/Assetst
Dividend dummyt
Debt issuance/Assetst
Equity issuance/Assetst
Acquisitions/Assetst
Yes
0.197
108,129
Year fixed effects
System weighted R2
Number of observations
50
Book cash ratiot
–0.122
(0.000)
–0.565
(0.000)
0.010
(0.115)
0.015
(0.000)
–0.014
(0.000)
0.019
(0.000)
0.004
(0.000)
–0.066
(0.000)
0.071
(0.000)
–0.040
(0.000)