fima_1115 fima2009v2.cls (1994/07/13 v1.2u Standard LaTeX document class) 10-7-2010 18:2
FIMA fima_1115 Dispatch: 10-7-2010 CE: RMQ
Journal MSP No. No. of pages: 27 PE: Gannon
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
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
Financial Flexibility, Investment Ability
and Firm Value: Evidence from Firms
with Spare Debt Capacity
Maria-Teresa Marchica and Roberto Mura
∗
We document, for the first time, that a conservative leverage policy directed at maintaining
financial flexibility can enhance investment ability. Our analysis reveals that following a period
of low leverage, firms make larger capital expenditures and increase abnormal investment. We
find that these new investments are financed through new issues of debt. The impact of financial
flexibility is both statistically significant and economically sizeable. Further, long-run performance
tests reveal that financially flexible firms not only invest more, but also invest better. Our results
are consistent with the view that financial flexibility in the form of untapped reserves of borrowing
power is a crucial missing link in capital structure theory.
There is a puzzling empirical regularity in the capital structure literature. Many fir ms appear
to borrow less than the dominant theories predict. In his influential paper, Graham (2000) finds
that “Paradoxically, large, liquid, profitable firms with low expected distress costs use debt
conservatively.” He also reports that this conservative behavior appears to be persistent. Similar
issues are discussed, among others, by Minton and Wruck (2001) and Strebulaev and Yang (2008).
Recent survey evidence has shed some light on this matter (Bancel and Mittoo, 2004; Brounen,
De Jong, and Koedijk, 2004; Graham and Harvey, 2001; Pinegar and Wilbricht, 1989). These
studies suggest that it is financial flexibility that primarily drives chief finance officers’ leverage
choices. Respondents say that flexibility is very important in enabling their companies to un-
dertake investment in the future, when asymmetric information and contracting problems might
otherwise force them to forego profitable growth opportunities. In other words, companies may
adopt a conservative leverage policy to maintain “substantial reserves of untapped borrowing
We thank Bill Christie (editor) and the anonymous referee whose comments have significantly helped to improve the paper.
We are indebted to Michael Brennan, Harry and Linda DeAngelo, Marie Dutordoir, Mara Faccio, Annalisa Ferrando,
Andrea Gamba, Ian Garrett, Marc Goergen, Alessandra Guariglia, John Hutton, Evangelos Kharalambakis, Meziane
Lasfer, WeiMin Liu, Kasper Nielsen, Aydin Ozkan, Ser-Huang Poon, Norman Strong, Alex Taylor, Alex Triantis and
Sergey Tsyplakov for helpful discussions. We also thank all the participants in the FIRS 2008, FMA USA 2008, FMA
Europe 2008, EFA USA 2007, FMA Europe 2007, EFMA 2006, and FMA USA 2006 “Top Ten Percent” Special Session,
and PFN 2006 meetings for their insightful comments. We are also grateful to the participants in the seminar series at
European Central Bank, Cass Business School, Manchester Business School, the Nottingham Department of Economics,
Sheffield Management School and the University of Verona for their useful suggestions. Kind help from David Roodman
of the Center for Global Development, Roger Walsh of Bureau Van Dijk, Derek Rouse of Hemscott Plc Support Team and
Francesco Cerlienco of Reuters is also acknowledged. We are also grateful to Wendy Jennings, Pam Losefsky, and Alison
Walters for editorial help. The usual disclaimer applies.
∗
Maria-Teresa Marchica is a Lecturer in Finance at the Manchester Business School, University of Manchester, Manch-
ester, England M13 9PL, UK. Roberto Mura is a Lecturer in Finance at the Manchester Business School, University of
Manchester, Manchester, England M13 9PL, UK.
Financial Management
•
Winter 2010
•
pages 1339 - 1365
fima_1115 fima2009v2.cls (1994/07/13 v1.2u Standard LaTeX document class) 10-7-2010 18:2
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
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
1340 Financial Management
r
Winter 2010
power” (Modigliani and Miller, 1963) allowing them to access the capital market in the event of
positive shocks to their investment opportunity set.
However, as yet, little is known about financial flexibility. How can we identify financially
flexible (FF) firms? Does financial flexibility really improve the investment ability of companies?
Ultimately, is this strategy value-enhancing for firms? This paper provides empirical evidence to
address these questions.
First, we identify “FF” firms by focusing on firms with spare debt capacity (SDC).
1
We estimate
a leverage equation from which we calculate the predicted level of debt. Since the demand for
financial flexibility is an unobservable factor, it ends up in the residual of the estimated model, and
generates a systematic deviation between observed and estimated leverage. For this reason, we
propose to capture the demand for financial flexibility indirectly, using negative deviations from
estimated target leverage. We classify a firm as FF if it has spare debt capacity for a minimum
number of consecutive years.
Second, we test econometrically whether this degree of financial flexibility has any impact on
investment ability. The prediction is that, in t he presence of market frictions, firms that anticipate
valuable growth options in the future may respond by pursuing a policy of low leverage for a
number of years. In this way, FF firms have enough spare borrowing power to be able to raise
external funds, and to invest more in the years following the conservative financial policy. To
test this hypothesis, we specify a q-model of investment augmented by our FF dummy, plus an
interaction term with cash flow. According to our flexibility argument, the FF dummy should have
a positive and significant impact on capital expenditures. In addition, to the extent that FF firms
can, after a period of conservative borrowing, more easily raise external funds to finance their
projects, their investment ability should be less dependent on internal funds. As a consequence,
we would expect a lower sensitivity of investment to cash flow.
We provide evidence that a conser vative leverage policy can help firms attain a degree of
financial flexibility. Our results indicate that FF companies exhibit enhanced investment ability.
Our tests reveal that an average company that maintains a spare debt capacity policy for three
years (FF3) can increase its capital expenditures by around 37%. Further, our tests reveal that
the longer the period of low leverage, the lower the economic impact of FF status on the firm’s
investment ability. For instance, a company that maintains spare debt capacity for at least six
years typically increases its capital expenditures by around 28%. This may be because the ability
of managers to anticipate future g rowth opportunities decreases the further into the future they
go. These results are robust to the method we follow to classify FF firms.
Also, we find that companies finance new investments by means of positive net debt issues.
This provides strong evidence that companies sacrifice borrowing today to enhance their ability
to seize better growth opportunities in the future. This result is also robust when we take into
account other financing strategies that may achieve financial flexibility, such as a cash policy. For
instance, when we account for the presence of (excess) cash in the investment decisions, or when
we consider leverage net of cash, as in Bates, Kahle, and Stulz (2009), we still find an economic
impact similar to our main results.
Third, we test the long run performance effect of this strategy. A number of papers suggest that
Q1
managers prefer to keep debt ratios low to reduce risk and to protect their undiversified human
1
Firms may achieve financial flexibility in other ways. For instance, Denis and Sibilkov (2009) demonstrate how higher
cash holding helps constrained firms to invest in value-enhancing projects. We do address this issue later in the text and
show that our results are robust even when we control for the effect of internal funds. Powers and Tsyplakov (2008) stress
the role of make-whole call provisions that allow firms to have the flexibility of retiring their debt early. Jagannathan,
Stephens, and Weisbach (2000) underline how stock repurchases (as opposed to cash dividends) allow firms a higher
degree of financial flexibility.
fima_1115 fima2009v2.cls (1994/07/13 v1.2u Standard LaTeX document class) 10-7-2010 18:2
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
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
Marchica & Mura
r
Financial Flexibility, Investment Ability and Firm Value 1341
capital (Fama, 1980), or to alleviate the pressure that comes with interest payment commitments
Q2
(Jensen, 1986). Alternatively, managers may choose to increase debt levels in a manner that allows
them to pursue empire building (Zwiebel, 1996) and to minimize takeover risk (Berger, Ofek,
and Yermak, 1997). In other words, conservative leverage and high investment may be symptoms
of greater agency costs.
To distinguish these various potential influences and to investigate whether a financing policy
aimed at financial flexibility is value-enhancing, the last step of the analysis examines its impact
on both long run performance and post-investment profitability. First, we use both the capital asset
pricing model (CAPM) and the Fama and French (1993) three-factor model to investigate the
behavior of Jensen’s (1986) alpha for FF firms in the long run, after they acquire flexibility status
or, alternatively, after they make abnormal investments. If our flexibility hypothesis is correct,
we expect this policy to be value-enhancing. If, on the other hand, this policy is an expression of
greater agency conflicts, we should find a negative effect on firm performance. Then, we compare
the company’s operating performance before and after the FF status is acquired, and before and
after FF companies make abnormal investments.
Our findings strongly indicate that companies that acquire financial flexibility through spare
debt capacity are not only able to invest more, but also seem to invest better. Long run performance
analysis consistently returns positive and statistically significant results for Jensen’s (1986) alpha
using either the CAPM or the Fama and French (1993) three-factor model up to 60 months
after the FF status identification, irrespective of the method employed to classify FF firms.
Economically, the impact on the returns of an FF firm also seems sizeable. According to the
figures obtained from the Fama and French (1993) three-factor model, FF companies outperform
the market by almost 30 basis points per month, which corresponds to about 7.1% in the first two
years. Similarly, we find that our FF firms outperform the market when we measure their long
run performance after they have made abnormal investments. The economic impact implied by
Jensen’s (1986) alpha suggests an abnormal long run performance of approximately 13.24% in the
first two years. Further, firms experience an increase of more than 18% in operating performance
within two years of the acquisition of the FF status and a staggering increase of 38% in their
profitability within two years after an abnormal investment. This enhances the strength of our
conclusions, and also allows us to rule out the possibility that managerial entrenchment drives
low leverage and higher investments.
Our study contributes to the literature in a number of ways. We provide direct evidence, for the
first time, of the value of financial flexibility to companies by studying the impact this policy
has on investment ability and long run performance. These are ver y important results within
recent developments in the capital structure literature. According to DeAngelo and DeAngelo
(2007), “financial flexibility is the critical missing link for an empirically viable theory [of capital
structure].” We provide very sound evidence that complements this hypothesis. A large fraction of
observed leverage is left unexplained by conventional theories of capital structure. Our inability
to “measure” the FF factor ex ante causes a systematic spread between observed and predicted
leverage. In this way, our work also provides a rationale for debt conservativism that tallies with
the theoretical predictions recently proposed by Almeida, Campello, and Weisbach (2009) and
DeAngelo, DeAngelo, and Whited (2010), and may provide an explanation of the so-called low
leverage puzzle.
The remainder of the paper is organized as follows. In Section I, we describe the data and
present the main hypotheses. In Section II, we present the empirical results and all the robustness
tests performed, while in Section III, we discuss our conclusions.
fima_1115 fima2009v2.cls (1994/07/13 v1.2u Standard LaTeX document class) 10-7-2010 18:2
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
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
1342 Financial Management
r
Winter 2010
I. Data and Hypotheses
A. Data Collection and Sampling
We construct our sample from all UK listed fir ms present in Datastream from 1965 to 2008.
We exclude financial firms as their capital structure is likely to differ from that of the other firms
in our sample. Since both leverage and investment models include the lagged dependent variable
as a regressor, we exclude companies with fewer than two consecutive years of data. Firms with
missing values of relevant variables are also excluded. All data are trimmed at the 1% level to
reduce the impact of outliers. Our results are largely unaltered when we winsorize the data at the
1% level rather than trim it. This results in an unbalanced panel of 47,553 observations for 4,290
firms.
For the performance analysis, we collect monthly returns data for individual stocks in the sample
for the FTSE All Share Index, FTSE Small Cap, and FTSE 100 Indexes to proxy portfolios of
small and large firms, and for the FTSE Global Value and FTSE Global Growth Indexes to proxy
portfolios of high and low book-to-market stocks. We collect the United Kingdom three-month
Treasury bill rates to approximate the risk free interest rate.
In an attempt to control for the agency costs of equity in the leverage equation, we collect,
by hand, information regarding ownership by managers, ownership by nonmanagers, and board
composition for a randomly selected subsample of firms for the period 1991-2001. Data are
collected from the Price Waterhouse Corporate Register (December issue). After screening the
data f or possible errors and matching it with the accounting information from Datastream, we
have an unbalanced panel of 677 firms and 5,660 observations. For this subsample of companies,
we also collect and match information on debt rating (Reuters) and bankr uptcy risk using FAME,
which is part of the Bureau van Dijk Electronic Publishing’s databases, with a coverage specific
to UK firms (Lemmon, Roberts, and Zender, 2008).
2
B. Identification of Financially Flexible Firms
Recent survey studies of capital structure choices provide strong evidence that the single most
important determinant of leverage decisions by firms is the desire to maintain financial flexibility
(Graham and Harvey, 2001; Bancel and Mittoo, 2004). However, as we argued above, since there
is no well-defined measure of flexibility in the literature, this is an unobservable factor that
depends largely on managers’ assessment of future growth options. Consequently, this factor will
end up in the residual of the model, where it will generate systematic deviations between observed
and estimated leverage. For this reason, we propose to indirectly capture the effect of financial
flexibility using deviations from predicted target leverage.
In the first step of the analysis, we identify firms with spare debt capacity using Frank and
Goyal’s (2009) baseline model, which includes median industry leverage, market-to-book ratio,
size, tangibility, profitability, and expected inflation. Frank and Goyal (2009) use the data from
the Livingston Survey to proxy for expected inflation. Similar data are not available for the entire
time series for the United Kingdom and, as such, we use T-bills instead. As Frank and Goyal
(2009) state “[ ...] replacing expected inflation with the Treasury bill rate is unlikely to matter
since they are highly correlated.”
We follow t he seminal work of Arellano and Bond (1991) by first differencing the model,
and then using suitable lagged levels of the dependent variable as instruments. We estimate all
leverage models using the GMM-SYS methodology. This allows us to control simultaneously for
2
We thank Francesco Cerlienco from Reuters for kindly providing the data.
fima_1115 fima2009v2.cls (1994/07/13 v1.2u Standard LaTeX document class) 10-7-2010 18:2
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
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
Marchica & Mura
r
Financial Flexibility, Investment Ability and Firm Value 1343
the endogeneity of the regressors and for fixed effects that may be correlated with the explanatory
variables (Blundell and Bond, 1998; Lemmon et al., 2008). Inclusion of the lagged dependent
variable allows for the firm’s targeting behavior.
3
The baseline model we estimate is the following:
LEV
it
= α
1
LEV
it−1
+ β
1
Industry Leverage + β
2
Mtbv + β
3
Size + β
4
Tangibility
+ β
6
Expected Inflation + η
i
+ η
t
+ u
it
.
(1)
We then compare the fitted values from the regression analysis with the actual values and define
as SDC those firms that exhibit a negative deviation between actual and predicted leverage. As
discussed above, we expect the systematic component of these deviations to be due to the
unobserved effect of financial flexibility in the leverage model. To minimize the impact of noise,
we require the deviation to be larger than 10%. We perform a number of robustness tests in which
we require a minimum deviation of either 5% or 25%. Alternatively, we follow Harford (1999)
and require observed leverage to be 1.5 standard deviations lower than the predicted value.
Finally, to classify a firm as FF, we require it to have SDC for a minimum number of consecutive
periods. This ensures that we are indeed observing a policy, not just a transitory shock to the capital
structure of the firm. As a baseline specification, we use FF3, which is the FF dummy that takes
a value of one when we observe at least three consecutive periods in which the firm is classified
as SDC. There is no theoretical rationale for choosing a specific time length. Therefore, to
assess whether the results are sensitive to the choice of time horizon, we use a number of different
proxies, from a minimum of two to a maximum of six consecutive years of leverage conservatism.
This approach is not free of drawbacks. The most serious is that the choice of leverage model
may affect the estimated target and the deviation from it. This, in turn, would influence the
classification of firms and the subsequent investment results. To minimize the possibility that
the results are driven by the choice of a specific leverage model, we take two important steps.
First, we perform numerous robustness tests using different leverage models. In particular, we
try to control for other factors that may allow firms to attain a degree of financial flexibility, the
most important of which is financial slack. Second, we follow an alternative approach similar
to Minton and Wr uck (2001) and classify firms as low leverage when their debt ratio is in the
bottom 20% of the distribution.
4
We refer to this as the percentile methodology.
C. Financial Flexibility and Investment Ability
In their seminal paper, Modigliani and Miller (1963) noted that despite the existence of some
tax advantages for debt financing, firms tend not “to use the maximum possible amount of
debt in their capital structure” due to limitations by lenders leading to “the need for preserving
flexibility.” In the modified version of the pecking order theory (Myers, 1984), firms have two
main reasons to restrain themselves from issuing debt: 1) to avoid the costs of financial distress,
and 2) to maintain financial slack. Taking these ideas as a starting point, we test the hypothesis
that in the presence of market frictions, firms that anticipate valuable growth options in the future
may respond by pursuing a policy of low leverage for a number of years. As in Myers (1984),
reserves of borrowing power enable FF firms to raise external funds and to invest more in the
years following conservative financial policy.
3
Numerous survey studies corroborate the idea that firms have a target capital structure. Graham and Harvey (2001)
report that 37% of US firms have a flexible target debt ratio, while a further 35% have a stricter target. Bancel and Mittoo
(2004) and Brounen et al. (2004) report similar figures for the United Kingdom.
4
Mikkelson and Partch (2003) classify as “high cash” those companies that hold more than 25% of their assets in cash
and equivalents. See also Iona, Leonida, and Ozkan (2004) for UK firms.