NICOLÁS E. MAGUD
International Monetary Fund
SEBASTIÁN SOSA
International Monetary Fund
Corporate Investment in Emerging Markets:
The Role of Commodity Prices
ABSTRACT We examine how firm-level and country-specific macroeconomic variables determine corporate investment in emerging markets. In particular, we investigate how investment
decisions are affected by changes in country-specific commodity export prices, using firmlevel data from 38 emerging markets for the period 1990–2013. We show that in addition to
the standard firm-level variables (such as expected future profitability, cash flows, leverage,
and new debt flows), commodity export prices play a significant role in driving corporate investment. Moreover, we show that the sharp decline in commodity prices since 2011 has been a
key factor explaining the sizable slowdown in private investment growth during this period,
especially in regions with large net commodity exporters.
JEL Codes: E2, E3, F3, F4
Keywords: Investment, emerging markets, commodity prices, capital inflows
C
ommodity prices have fluctuated widely over the past two decades. The
macroeconomic impact of commodity price swings has been studied
extensively in the literature, both empirically and theoretically. However, empirical studies on the link between commodity prices and corporate
investment in emerging markets are relatively scant, particularly those based
on firm-level data. This paper empirically investigates the determinants of
investment at the firm level in emerging markets, with a special focus on the
role of commodity export prices. As a by-product, the paper examines the
factors behind the post-2011 weakening of private investment in emerging
markets (in particular, commodity export prices) and the differences across
emerging regions.
ACKNOWLEDGMENTS The authors are grateful to Sebnem Kalemli-Ozcan, Hamid Faruqee,
Andre Meier, Gian Maria Milesi Ferretti, Bertrand Gruss, Herman Kamil, Alex Klemm, Samya
Beidas-Strom, Hui Tong, Davide Furceri, and Sergejs Saksonovs for their valuable comments
and suggestions. They also thank Genevieve Lindow and Ben Sutton for their excellent research
assistance.
157
158
E C O N O M I A , Fall 2017
F I G U R E 1 . Real Private Investment and Commodity Export Price Growth, 2004–14
A. Asia excluding China
Percent
35
30
25
20
15
10
5
0
–5
–10
–15
2004
B. Commonwealth of Independent States
Percent
40
Real private investment
20
0
–20
Commodity export price
2006
2008
2010
2012
2014
–40
2004
C. Europe
Percent
30
20
20
10
10
0
0
–10
–10
–20
–20
2004
2006
2008
2008
2010
2012
2014
D. Latin America and the Caribbean
Percent
30
–30
2006
2010
2012
2014
–30
2004
2006
2008
2010
2012
2014
Source: International Monetary Fund (IMF), World Economic Outlook database; Gruss (2014); and IMF staff calculations.
Private investment in emerging markets is highly correlated with (countryspecific) commodity export prices (figure 1). The comovement of private
investment and commodity export prices is especially high in the case of Latin
America and the Caribbean and the Commonwealth of Independent States,
with correlation coefficients of 0.84. This reflects the fact that these regions
include many of the largest commodity exporters. For emerging Europe, the
correlation is also strong (0.82), while it is much lower for emerging Asia
excluding China (0.36). Private investment in emerging markets has also been
highly correlated with capital inflows (figure 2).
We study the determinants of investment in panel regressions that combine firm-level data for about 16,000 listed firms with country-specific
macroeconomic variables—notably, commodity export prices and capital
Nicolás E. Magud and Sebastián Sosa
159
F I G U R E 2 . Real Private Investment Growth and Net Capital Inflows, 2004–14a
B. Commonwealth of Independent States
A. Asia excluding China
Percent
25
20
15
10
5
0
2004
2006
2008
2010
Percent of
GDP
5
4
3
2
1
0
–1
2012 2014
Percent
40
20
0
–20
–40
2004
C. Europe
Percent
30
20
10
0
–10
–20
–30
2004
Percent of
GDP
12
Real private investment
10
8
6
4
2
Capital inflows (right scale)
0
2006 2008 2010 2012 2014
2006
2008
2010
Percent of
GDP
8
6
4
2
0
–2
–4
–6
–8
2012 2014
D. Latin America and the Caribbean
Percent
20
15
10
5
0
–5
–10
–15
–20
–25
2004
Percent of
GDP
4
3
2
1
0
2006
2008
2010
2012
2014
–1
Source: IMF, World Economic Outlook database; and IMF staff calculations.
a. PPP-weighted average. Capital inflows are defined as the balance of the external financial account, in percent of GDP.
inflows—for thirty-eight emerging markets over the period 1990–2013.1 After
identifying the key factors driving firms’ investment decisions in emerging
markets, we shed light on which of these factors have been the main drivers
of the sharp deceleration in corporate investment growth since 2011.
Our study generates four main results. First, we confirm the importance of
what can be dubbed the usual suspects. In line with previous studies in the
literature, we find that emerging market firms’ capital expenditure is positively
associated with expected profitability (proxied by Tobin’s q), cash flows (suggesting the existence of borrowing constraints), and debt flows. It is negatively
1. Table A1 in appendix A presents the list of countries in the sample and the number of
firms in each country.
160
E C O N O M I A , Fall 2017
associated with leverage. Second, and the key contribution of the paper, commodity prices matter. Conditional on the usual suspects, investment is positively associated with changes in country-specific commodity export prices,
and the link is statistically and economically significant. Third, investment by
emerging market firms is also influenced by the availability of foreign (international) financing.
Finally, based on the first three results, we put the magnifying glass on the
most recent event of a fall in commodity prices. Thus, as an extension to our
main contribution, we look into whom to blame for the post-2011 investment
slowdown. Factors vary across emerging market regions, with the sharp adjustment in commodity prices playing a substantial role in commodity exporter
regions (such as Latin America). Another factor was lower expected profitability
of firms, which partly reflects the downward revisions to potential growth in
many emerging markets. The moderation in capital inflows to emerging markets
and increased leverage (particularly in Asia) also played a significant role.
Our paper is related to the extensive empirical literature on the determinants
of corporate investment in emerging markets. It relates to a strand that studies financing constraints, typically relying on Tobin’s q investment models or
Euler investment equations. Most of these studies document the importance
of internal financing for firms’ investment owing to capital markets’ imperfections. Based on this framework, for example, Fazzari, Hubbard, and Petersen
examine the case of U.S. manufacturing firms, while Love and Zicchino study
emerging market companies.2 The sensitivity of investment to cash flows is
particularly strong for smaller firms and for firms in less financially developed economies.3 The use of cash flow as a measure of financial frictions
has been criticized, however.4 This has been addressed by Gilchrist and
Himmelberg, who establish the existence of financial constraints by testing
the significance of investment-to-cash-flow sensitivities beyond the effect of
the so-called fundamental q.5 The latter is essentially a vector autoregression
(VAR) of forecasting equations out of which the expected value of marginal q,
conditional on observed fundamentals (including cash flow), is constructed.
This implies that any additional effect picked up by cash flows should reflect
financial constraints.
2. Fazzari, Hubbard, and Petersen (1988); Love and Zicchino (2006). Hubbard (1998) provides a thorough survey of this literature.
3. Fazzari, Hubbard, and Petersen (2000); Carpenter and Guariglia (2008); (Love, 2003).
4. For example, Kaplan and Zingales (1997); Gomes (2001); Abel and Eberly (2011).
5. Gilchrist and Himmelberg (1995, 1999).
Nicolás E. Magud and Sebastián Sosa
161
We follow this q literature, aware of its possible shortcomings. We use the
q as one important explanatory variable of firm-level investment, but we also
control for other variables to mitigate, to the extent possible, other investment
opportunities that could be misinterpreted as captured by the q.
Harrison, Love, and McMillan document that foreign direct investment
(FDI) flows to emerging markets are associated with a reduction in firms’
financing constraints.6 They examine whether—and to what extent—the
availability of foreign capital helps relax financing constraints in emerging
market firms by combining firm-level data on cash flows with country-specific
capital flows. Forbes also finds that financing constraints relax when capital
account restrictions are eased, as do Gelos and Werner.7 These studies focus
on macroeconomic variables, but only on capital flows and their role in the
relaxation of financial constraints. In contrast, we want to better understand
another key driver of corporate investment in emerging markets: namely,
commodity export prices.
In another related paper, though from a macroeconomic perspective,
Fernández, Gonzales, and Rodríguez show that in emerging markets, business
cycles are strongly influenced by country-specific commodity prices, which
are procyclical.8 Finally, Fornero, Kirchner, and Yany, and Ross and Tashu,
study the link between the terms of trade and investment.9
We contribute to this literature in several ways. First, we analyze the determinants of firms’ investment decisions for a large sample of emerging markets
covering a period of over two decades. This contrasts with previous studies on
investment in emerging markets using firm-level data, which mostly focused
on one country or a small group of countries. Our approach allows us not
only to work with an extensive database, but also to explore (and exploit) the
potential heterogeneity across emerging market regions. Second, in addition
to firm-level data, we include some country-specific macroeconomic variables
in the analysis—notably, commodity export prices. The latter is our main contribution. Finally, as a by-product, we examine the drivers of the post-2011
investment growth slowdown and how the main factors varied across emerging market regions.
The rest of the paper proceeds as follows. The next section presents a theoretical framework to motivate the empirical exercise that follows. Subsequent
6.
7.
8.
9.
Harrison, Love, and McMillan (2004).
Forbes (2007); Gelos and Werner (2002).
Fernández, Gonzales, and Rodríguez (2014).
Fornero, Kirchner, and Yany (2014); Ross and Tashu (2015).
162
E C O N O M I A , Fall 2017
sections describe the empirical approach and present the results, while the
final section provides concluding remarks.
Theoretical Framework
This section presents an augmented q model of investment for a small open
economy, which we use as a framework for the empirical analysis below.
We develop a basic frictionless model to illustrate how commodity prices
can affect investment decisions. Adding frictions to this model is unlikely to
result in different firm-level decisions; however, we test for their impact in
the empirical section below.
The problem of firm i in period t over an infinite horizon is to maximize
the present discounted value of the flow of dividends, Dt, given by
∞ D
Et ∑ t +i 1 ,
i =1 R
(1)
where R represents the gross interest rate. In turn, the firm’s dividend flow
is given by
(2)
Dt = π ( K t , θt ) − pt I t − c ( I t , K t ) ,
where p is the firm’s profit function, Kt the stock of capital, qt the level of
technology, and pt the price of capital in units of domestic goods. It denotes
investment, and c(It, Kt) is a function that captures the adjustment cost of
investment. The profit function is assumed to be increasing in capital and
level of technology, and concave. The adjustment cost of installing new capital is an increasing and convex function in the value of (It /Kt), defined below,
and qt is a stationary first-order Markov process. Given a constant rate of
depreciation, d, the stock of capital equation changes over time, as
(3)
K t +1 = I t + (1 − δ ) K t .
Assume that firms in this small open economy purchase their capital
abroad.10 Since capital is imported, the domestic price of investment depends
on the real exchange rate. In turn, the real exchange rate increases with the
10. Assuming that only a share of the capital stock is imported does not alter the results.
Nicolás E. Magud and Sebastián Sosa
163
country’s terms of trade, that is, the relative price of exports to imports ( pX /pM).
We normalize the real exchange rate, e, to the unit circle, taking a value of
zero when the terms of trade equal their long-run value. Thus, the domestic
price of importing capital is given by
p
pX
p
M
M
(4)
If the terms of trade are at their long-run value (denoted by an overbar),
so is the real exchange rate (equaling zero). In this case we have the typical closed economy example, in which the domestic price of capital equals
one. When the economy’s terms of trade are above their long-run value, the
economy is richer, so the real exchange appreciates (that is, it increases), and
the price of new capital in terms of domestic goods decreases. Likewise, for
terms of trade lower than their long-term value, the economy is poorer, the
real exchange rate depreciates, and the price of investment is higher.
Therefore, the firm’s problem is to maximize equation 1 subject to equations 2–4. The Bellman equation for the firm’s problem is given by
(5)
V ( K t , θt , et ) = max I ,K
t
t +1
p
π ( K t , θt ) − 1 − et X I t
pM
.
1
− c ( I t , K t ) + Et V( K t +1 , θt +1 , et +1 )
R
Equivalently,
(6)
V( K t , θt , et ) = max I
t
p
π ( K t , θt ) − 1 − et X I t − c ( I t , K t )
pM
.
1
+ Et V I t + (1 − δ ) K t , θt +1 , et +1
R
{
}
164
E C O N O M I A , Fall 2017
Optimizing over the control variable It, while Kt is the state variable, implies
the following first-order condition:
(7)
pX
1
1
1 − et
+ cI ( I t , K t ) = R Et V( K t +1 , θt +1 , et +1 ) = R Et qt +1 .
p
M
On the right-hand side of equation 7, as usual in the literature, we define
Tobin’s q as the discounted shadow price of capital—marginal q—which
equals the replacement cost of capital plus the adjustment cost of installing
new capital, that is, the effective price of new capital. Assume that a constantreturns-to-scale adjustment cost of capital is given by
2
1 I
c ( I t , K t ) = b t − µ K t ,
2 Kt
(8)
in which µ denotes the investment-to-capital ratio in steady state, which is
associated with no adjustment costs. Intuitively, µK is the level of investment
necessary to maintain a constant stock of capital in the steady state. Substituting equation 8 in equation 7, we get
(9)
1
It
pX
1
− µ = Et V( K t +1 , θt +1 , et +1 ) = Et qt +1 .
1 − et
+ b
R
pM
R
Kt
Rearranging equation 9 yields
(10)
p
It 1 1
= Et qt +1 + et X − 1 + µ,
Kt b R
pM
which shows the standard positive association between Tobin’s q and investment. As shown in the literature, an increase in marginal q (that is, a higher
shadow price of capital, which implies a larger present discounted value of
the flow of dividends, as shown below), causes the firm to optimally increase
investment. The latter can be shown by using the envelope condition out of
equation 6:
(11)
qt = π K ( K t , θt ) − cK ( I t , K t ) +
1
(1 − δ ) Et [ qt +1 ].
R
Nicolás E. Magud and Sebastián Sosa
165
Updating (11) one period, forwarding it, taking expectations as of period t,
applying the law of iterated expectations and substituting back in (11), and
finally iterating forward and using the transversality condition yields
(12)
i
∞ 1 − δ
π K ( K t + i , θt + i ) − cK ( I t + i , K t + i ) ,
VK ( K t , θt ) = Et ∑
i = 0 R
which shows that the marginal value of an additional unit of capital should
equal the discounted flow of marginal profits, net of adjustment costs.
Crucially for our empirical analysis, equation 10 also shows that, all
else equal, an improvement in the terms of trade (that is, the relative
price of exports to imports) results in real appreciation, which increases
investment—consistent with the lower costs of importing capital—and
vice versa. Appendix B presents the phase diagram corresponding to the
saddle-path equilibrium and the effects of (transitory and permanent)
terms-of-trade shocks.
Econometric Approach
Based on the model presented in the previous section, we estimate a panel
regression model of investment with time and firm-level fixed effects, combining firm-level data and country-specific macroeconomic variables to identify the main determinants of corporate investment in emerging markets. The
analysis focuses on factors that, for theoretical reasons, are thought to affect
firms’ investment decisions. These factors include firm-specific variables such
as expected future profitability, cash flows, cost of debt, leverage, and debt
flows. We also include country-specific macroeconomic variables—notably
commodity export prices, but also net capital inflows and uncertainty. We
then look at the recent deceleration of private investment growth in emerging
markets to examine the key factors explaining the slowdown and the main
differences across emerging market regions.
Empirical Model
Our empirical specification is a variation of the traditional Tobin’s q investment model, augmented to include other possible determinants identified in
the literature on corporate investment. In a neoclassical model, the marginal
benefit from an extra unit of investment and the cost of capital should be
166
E C O N O M I A , Fall 2017
sufficient statistics to explain investment behavior. The q theory of investment basically reformulates the neoclassical theory, such that firms’ investment decisions are based on the ratio between the market value of the firm’s
capital stock and its replacement cost.11 Much of the literature on corporate
investment published over the last decades, however, highlights the importance of financing constraints. In the presence of financial frictions, access to
external financing for investment projects that would in principle be profitable may be limited. Therefore, firms’ investment decisions would be determined not only by investment opportunities, but also by the availability of
internal funds.
Evidence of financial constraints is largely based on the sensitivity of
investment to different measures of internal funds—typically cash flow
or cash stock. A firm’s higher dependence on internal funding is interpreted as a sign of tighter financial constraints.12 However, this interpretation of the correlation between cash flow and investment as evidence of
financial constraints is far from uncontroversial. A strand of the literature
argues that rather than financing constraints, the relationship between cash
flows and investment may reflect the correlation between cash flow and
investment opportunities that are not captured well by traditional measures
of investment opportunities, in particular Tobin’s q. Nevertheless, a number
of studies address these criticisms, and most empirical studies continue to
use the investment-to-cash-flow sensitivity as a measure of financial frictions.13 We also follow this approach, using both cash flow measures and
Tobin’s q.
Beyond corporate financial indicators, we also include key country-specific
macroeconomic variables that may affect corporate investment. Specifically,
we consider commodity export prices (which drive the terms of trade), capital
inflows, and uncertainty. We estimate linear panel regressions allowing for
both time and firm-level fixed effects.14 Given that our specification contains
both firm-level and country-level data, we use clustered (by country) robust
standard errors to address the risk of standard-error bias. As is common in the
literature, we use the lagged dependent variable as an additional explanatory
11. Tobin (1969); Hayashi (1982). For instance, investment would increase whenever the
value of q is larger than one, an indicator that the present discounted value of the flow of
expected dividends outweighs the replacement cost of capital.
12. See, for example, Fazzari, Hubbard, and Petersen (1988); Blanchard, Rhee, and Summers
(1993); and Fazzari, Hubbard, and Petersen (2000).
13. For example, Gilchrist and Himmelberg (1995, 1999); Carpenter and Guariglia (2008).
14. As discussed later, the results are robust to also allowing for country fixed effects.
Nicolás E. Magud and Sebastián Sosa
167
variable. Thus, the baseline specification, consistent with equation 10 above,
is as follows:
(13)
I ic,t
I ic,t −1
CFic,t
=α+λ
+ β1Qic,t + β 2
+ β3 LEVic,t −1
K ic,t −1
K ic,t − 2
K ic,t −1
+ β4
∆DEBTic,t
+ β5 INTic,t + β6 Picx,t −1 + β 7KI c,t
K ic,t −1
+ β8 UNCc,t + di + d t + ε ic,t ,
where the subscripts (ic,t) stand for firm i in country c during period t. I is fixed
investment (excluding inventories) and K the stock of capital. The variable q
represents the standard Tobin’s q, where average q, measured as the firm’s
price-to-book-value ratio, is used as a proxy for (unobservable) marginal q.15
CF denotes the firm’s cash flow; LEV is leverage; DDEBT is the change in
total debt since the previous period; and INT is a measure of the firm’s cost of
capital, to account for the opportunity cost of funds. KI denotes (net) capital
inflows; Px, (the log difference of) the commodity export price index; and
UNC, aggregate uncertainty. The variables di and dt represent firm and trend
(or alternatively time, see discussion below) fixed effects. Finally, e is the
error term.
Intuitively, this specification is based on the idea that investment is determined by the flow of (discounted) future dividends. As shown in equation 10
above, we should expect a positive coefficient associated with q, indicating
that firms that expect to be more profitable should invest more, which is
a common finding in the literature. As also discussed above, the cash flow
coefficient should exhibit a positive sign if firms face financial constraints,
since firms would need to rely on internal funds to finance investment projects. Debt stock and debt flows, in turn, are expected to have opposite effects
on corporate investment. While higher leverage is expected to be negatively
associated with investment, the flow of debt would be positively related to
capital expenditure because financing investment is one of the main reasons to
incur new debt. A higher cost of debt, in turn, is expected to be associated with
lower investment. Regarding the country-level variables, commodity export
prices are expected to be positively related to capital spending. Net capital
inflows should also be positively related to corporate investment, because
15. See Hayashi (1982) for a discussion of the conditions under which the two measures
are equivalent.
168
E C O N O M I A , Fall 2017
they may play a role in relaxing firms’ financing constraints in emerging
markets.16 Finally, economic theory predicts that higher uncertainty should be
associated with lower investment as firms enter a wait-and-see mode, especially to the extent that investment decisions are irreversible.17
Data
We use annual firm-level data from Worldscope. The sample includes 16,000
publicly traded firms from thirty-eight emerging markets, covering the period
1990–2013. Table A1 in appendix A presents the list of countries in the sample and the number of firms per country.18 The number of firms varies significantly across countries as well as across time, with a smaller number in most
countries during the first half of the 1990s.19
f i r m - l e v e l d a t a . Investment (I) is measured as the purchase of fixed
assets by the firm. The stock of capital (K) is measured as the total net value
of property, plant, and equipment. Tobin’s q is given by average q. Cash flow
(CF) is computed as the firm’s net profits from operating activities; leverage
(LEV) is measured as the ratio of total debt obligations to total assets; new
debt (DDEBT) is defined as the change in total debt obligations since the
previous period; and the cost of funds (INT) is defined as the firm’s effective
interest rate paid on total debt obligations.
To avoid the presence of outliers and coding errors that would bias the estimation, observations with inconsistent data are dropped from the sample.20 The
country-specific distribution is then calculated for each of the variables, and
16. Harrison, Love, and McMillan (2004).
17. See, for instance, Bloom, Bond, and van Reenen (2001); Magud (2008); Baum,
Caglayan, and Talavera (2008); and Dixit and Pindyck (1994). More recently, Li, Magud, and
Valencia (2015) document how firm heterogeneity matters in the response of investment
to interest rate versus uncertainty shocks, as the balance sheet dimension can identity if either
a financial channel or a wait-and-see channel dominates the firm’s investment reaction to
the shock.
18. We consider countries that were classified as emerging markets at the start of the sample.
19. The share of total private investment accounted for by corporate investment ranges,
for example, between 70 and 75 percent across countries in Latin America and the Caribbean (although disaggregated data are not available for many countries). Moreover, the recent
downturn has mainly been driven by corporate investment (although residential investment has
also been trending downward in some countries). The firm-level data in the sample represent
about 12 percent of aggregate private investment (in the national accounts), with correlation
coefficients varying by country but averaging over 30 percent.
20. For example, negative book values for the capital stock, debt, or the price-to-book value
of equity.
Nicolás E. Magud and Sebastián Sosa
169
T A B L E 1 . Summary Statistics
Variable
Investment/capital stock (t-1)
q
Cash flow/capital stock (t-1)
Leverage
Interest expense ratio
Change in debt/capital stock (t-1)
Commodity export price growth
Capital inflows/GDP
No. observations
Mean
Standard deviation
389,977
435,454
410,693
493,919
355,256
357,397
367,748
497,058
0.25
1.81
0.06
0.68
0.08
0.27
4.32
-0.49
1.46
1.59
4.67
1.05
0.08
6.69
13.18
5.39
Source: Authors’ calculations.
the bottom and top 5 percent of each variable’s observations are excluded from
the analysis. Table 1 reports the summary statistics for the firm-level data.21
m a c r o e c o n o m i c d a t a . We use the country-specific gross commodity export
price indexes constructed by Gruss.22 Capital inflows (measured using the
financial account balance, in percent of GDP) and real GDP series come
from International Financial Statistics and World Economic Outlook, both
published by the International Monetary Fund (IMF). Finally, we use data
from Bloomberg to construct our measure of country-specific uncertainty
based on the average monthly volatility of stock market returns, computed as
the standard deviation of daily stock market returns over a month.
Results
Table 2 reports the results of the baseline specification (equation 13). Columns 1–3 show that all the coefficients for the firm-level variables have the
expected sign and are statistically significant at the one percent level. Following the theoretical model above, the dependent variable is the investment-tocapital ratio (ICR), with the stock of capital lagged one period. Consistent
with the theory and findings in previous empirical studies, Tobin’s q is positively related to investment. Also in line with previous studies, we find robust
evidence of financial constraints, reflected in a positive relationship between
a firm’s cash flow and capital spending. Moreover, more leveraged firms tend
21. Using only listed firms restricts the sample of firms, imposing some limitations to the
data.
22. Gruss (2014).
170
E C O N O M I A , Fall 2017
T A B L E 2 . Baseline Resultsa
Explanatory variable
ICR (t-1)
q
(1)
(2)
(3)
(4)
(5)
(6)
0.0967***
(0.0126)
0.0207***
(0.0045)
0.0966***
(0.0124)
0.0200***
(0.0044)
0.0069***
(0.0019)
0.1070***
(0.0154)
0.0190***
(0.0045)
0.0125***
(0.0023)
-0.0337***
(0.0035)
-0.0793***
(0.0273)
0.0033***
(0.0009)
0.0949***
(0.0188)
0.0182***
(0.0045)
0.0117***
(0.0022)
-0.0324***
(0.0029)
-0.0712**
(0.0283)
0.0029***
(0.0010)
0.0004***
(0.0001)
0.0929***
(0.0187)
0.0178***
(0.0043)
0.0117***
(0.0022)
-0.0318***
(0.0030)
-0.0685**
(0.0292)
0.0029***
(0.0010)
0.0005***
(0.0001)
0.0023***
(0.0007)
8.8320***
(1.0170)
8.8720***
(0.9980)
9.8130***
(0.9630)
9.3580***
(0.8300)
9.1640***
(0.8540)
0.0905***
(0.0191)
0.0176***
(0.0042)
0.0115***
(0.0021)
-0.0318***
(0.0031)
-0.0663**
(0.0298)
0.0029***
(0.0010)
0.0005***
(0.0001)
0.0024***
(0.0007)
-1.39e–06
(1.31e–06)
8.7850***
(0.9890)
94,183
0.030
16,512
38
94,157
0.036
16,511
38
83,327
0.059
15,102
38
63,799
0.051
12,262
36
63,799
0.053
12,262
36
62,632
0.052
12,190
36
Cash flow
Leverage (t-1)
Interest expense ratio
(t-1)
Change in debt
Commodity export
price (t-1)
Net capital inflows
Uncertainty
Constant
Summary statistic
No. observations
R2
No. firms
No. countries
** Statistically significant at the 5 percent level.
*** Statistically significant at the 1 percent level.
a. The dependent variable is the investment-capital ratio (ICR), with the stock of capital lagged one period.The regressions control for time
and firm-level fixed effects. Robust standard errors (clustered by country) are in parentheses.
to exhibit lower investment in the following period, while an increase in debt
is associated with higher capital expenditure. Finally, the coefficient on the
cost of debt is negative, as expected.
We then introduce the country-specific macroeconomic variables (table 2,
columns 4–6). The magnitude and significance of the coefficients of Tobin’s q,
cash flow, leverage, cost of debt, and change in debt do not change. We find
robust evidence that an increase in a country’s commodity export prices is
associated with higher investment in firms in that country. This result is consistent with previous studies that document the positive impact of improving
terms of trade on investment even beyond firms in the export sector.23 It also is
23. For example, Fornero, Kirchner, and Yany (2014) for Chile; Ross and Tashu (2015) for
Peru.
Nicolás E. Magud and Sebastián Sosa
171
consistent with Fernández, Gonzales, and Rodríguez, who document that,
on average, emerging markets are commodity exporters and country-specific
commodity prices are procyclical.24 The impact of commodity export prices
could be transmitted through direct channels affecting commodity sectors (and
other sectors, such as manufacturing and services, related to commodities) or
indirectly through income effects affecting aggregate demand and activity in
other sectors, as well.25
Investment in emerging market firms is also influenced by the availability of foreign (cross-border) financing. The larger the net capital flows an
emerging market economy receives, the larger its firms’ capital expenditure.
Both coefficients (on commodity export prices and capital inflows) are positive and strongly statistically significant. Interestingly, we do not find market
uncertainty to be a significant determinant of capital expenditure at the firm
level. This result is consistent with previous studies showing that although
uncertainty has a negative effect on investment, the effect generally disappears
when Tobin’s q is introduced.26
The estimated coefficients are not only statistically but also economically significant in most cases. A one-standard-deviation change in each
of the main independent variables would be associated with the following
changes in the investment-to-capital ratio (in percentage points): Tobin’s q:
2.9; cash flow: 5.3; leverage: 3.3; change in debt: 1.9; commodity export
growth: 0.63; and capital inflows: 1.4 (see figure 3). As indicated in table 1,
the investment-to-capital ratio has a mean of 0.25 and a standard deviation
of 1.46.
We then explore whether the overall results are mostly explained by one
emerging market region or if they hold across regions. Table 3 reports the
results of splitting the sample by regions. The results for the main explanatory
variables hold for most regions.27 In particular, the coefficient on commodity
export prices is positive and statistically significant for all regions.
Extension: The Post-2011 Private Investment Weakening
Private investment exhibited strong growth in emerging markets in the period
2003–11, except in 2009, when the global financial crisis hit. After peaking
24. Fernández, Gonzales, and Rodríguez (2014).
25. See Druck, Magud, and Mariscal (2015).
26. For example, Leahy and Whited (1996).
27. An exception is emerging Europe, where a few regressors (such as cash flow, leverage,
and cost of debt) show the correct sign but are not statistically significant.
172
E C O N O M I A , Fall 2017
F I G U R E 3 . Investment-Capital Ratio Response to One-Standard-Deviation Shock
to Independent Variables
Percentage points
6
5
4
3
2
1
0
Tobin’s q
Cash flow
Leverage
Change
in debt
Commodity
export prices
Capital
inflows
Source: IMF staff calculations.
in 2011, however, investment growth has gradually slowed (figure 4). Most
emerging market regions have shared a similar pattern of investment dynamics, with strong growth in the precrisis period, a sharp contraction in 2009 followed by a rapid and strong recovery, and a sustained deceleration since 2011.
The latter was particularly pronounced in emerging Europe, where growth has
stalled, and “Other” economies, where it actually turned negative in 2014.
But, which of the factors identified above play the biggest role in explaining
the recent investment deceleration? Have the key factors varied across emerging market region? To answer these questions, we add to the equation a dummy
variable (RECENT) that takes the value of one for all observations during the
post-2011 period. Here, we control for time effects through a time trend rather
than year dummy variables (to mitigate multicolinearity problems).28 We also
add interaction terms, interacting the RECENT dummy variable with the main
factors determining investment, in order to assess whether the marginal effect
of any of the latter changed in the most recent period—both in the full sample
and for each region. Specifically, we estimate the following specification:
28. Analysis of time effects through year dummy variables points to a clear downward
trend, which supports the substitution for a time trend in the regression.
Nicolás E. Magud and Sebastián Sosa
173
T A B L E 3 . Regional Decompositiona
Explanatory variable
ICR (t-1)
q
Cash flow
Leverage (t-1)
Interest expense ratio (t-1)
Change in debt
Commodity export price (t-1)
Net capital inflows
Uncertainty
Constant
Summary statistic
No. observations
R2
No. firms
No. countries
Full sample
(1)
LAC
(2)
Asia
(3)
Europe
(4)
Other
(5)
0.0905***
(0.0191)
0.0176***
(0.0042)
0.0115***
(0.0021)
-0.0318***
(0.0031)
-0.0663**
(0.0298)
0.0029***
(0.0010)
0.0005***
(0.0001)
0.0024***
(0.0007)
-1.39e–06
(1.31e–06)
8.785***
(0.989)
0.1900***
(0.0353)
0.0129***
(0.0030)
0.0136**
(0.0051)
-0.0450***
(0.0089)
-0.0114
(0.0214)
0.0026*
(0.0013)
0.0006**
(0.0002)
0.0019
(0.00104)
-9.01e–07**
(3.39e–07)
1.844
(1.349)
0.0787***
(0.0221)
0.0162**
(0.0051)
0.0191***
(0.0039)
-0.0329***
(0.0035)
-0.0803*
(0.0402)
0.0027*
(0.0014)
0.0005***
(0.0001)
0.0024**
(0.0009)
-4.66e–06**
(1.61e–06)
9.360***
(1.119)
0.0776**
(0.0310)
0.0230***
(0.0059)
0.00137
(0.0012)
-0.0133
(0.0089)
0.0026
(0.0768)
0.0008
(0.0016)
0.0004***
(6.81e–05)
0.0040***
(0.0012)
-2.90e–06
(2.29e–06)
9.094*
(5.046)
0.1520***
(0.0357)
0.0268***
(0.0020)
0.00839***
(0.0011)
-0.0291*
(0.0119)
-0.1330*
(0.0604)
0.0075**
(0.0021)
-5.00e–05
(0.0004)
0.0012
(0.0012)
7.98e–06
(5.47e–06)
14.04***
(2.912)
47,506
0.049
8,894
10
6,404
0.044
1,615
13
62,632
0.052
12,190
36
4,622
0.085
775
7
4,100
0.142
906
6
* Statistically significant at the 10 percent level.
** Statistically significant at the 5 percent level.
*** Statistically significant at the 1 percent level.
a. The dependent variable is the investment-capital ratio (ICR), with the stock of capital lagged one period.The regressions control for time
and firm-level fixed effects. Robust standard errors (clustered by country) are in parentheses.
(14)
I ic,t
I ic,t −1
CFic,t
=α+λ
+ β1Qic,t + β 2
+ β3 LEVic,t −1
K ic,t −1
K ic,t − 2
K ic,t −1
+ β4
∆DEBTic,t
+ β5 INTic,t −1 + β6 P icx ,t −1 + β 7 KI c ,t
K ic,t −1
+ δRECENT + ηh RECENT × X th + di + d t + ε ic,t ,
CFic,t
∆DEBTic,t x
for X th =
, LEVic,t −1 ,
, Pic,t −1, KI c,t , respectively.
K ic,t −1
K ic,t −1
174
E C O N O M I A , Fall 2017
F I G U R E 4 . Real Private Investment Growth, 2001–14
Percent
40
30
20
10
0
–10
Latin America and the Caribbean
Europe
Asia excl. China
Other
–20
–30
–40
2002
2004
2006
2008
2010
2012
2014
Source: IMF, World Economic Outlook database; and IMF staff calculations.
Table 4 presents the results for the full sample. The coefficient on the
RECENT dummy variable is negative and statistically significant, pointing
to weaker corporate investment during this period (column 1), while all the
regressors (both firm-level and country-specific macroeconomic variables)
retain their sign and statistical significance. Regarding the interaction terms,
financial constraints relaxed in the recent slowdown (column 3), while the negative relationship between leverage and firm-level investment became stronger
(column 4). At the same time, firms’ investment sensitivity to changes in capital inflows and debt flows weakened in the post-2011 period (columns 5–6).
To focus on the contribution of each factor in each emerging market region
during the recent slowdown we run specification 14 for each region’s firms
separately. The results are shown in the appendix (tables A2–A4). Notably,
corporate investment has become more sensitive to commodity export prices
in Latin America and less so in emerging Asia (columns 5–6 in table A2),
while leverage’s role in explaining investment increased in emerging Asia and
dropped in Latin America (columns 1–2 in table A3). Finally, the sensitivity to q increased in emerging Europe (column 7 in table A3), while in Asia
the relationship between capital inflows and firm-level investment weakened
(column 6 in table A4).
Recent
Uncertainty
Net capital inflows
Commodity export price (t-1)
Change in debt
Interest expense ratio (t-1)
Leverage (t-1)
Cash flow
q
ICR (t-1)
Explanatory variable
0.0907***
(0.0191)
0.0175***
(0.0043)
0.0114***
(0.0021)
-0.0316***
(0.0031)
-0.0638**
(0.0293)
0.0029***
(0.0010)
0.0004***
(9.46e–05)
0.0025***
(0.0007)
-2.11e–06
(1.39e–06)
-0.0084*
(0.0042)
(1)
0.0905***
(0.0191)
0.0170***
(0.0045)
0.0114***
(0.0021)
-0.0317***
(0.0031)
-0.0644**
(0.0299)
0.0029***
(0.0010)
0.0004***
(9.39e–05)
0.0025***
(0.0007)
-2.02e–06
(1.34e–06)
-0.0136*
(0.0069)
(2)
(3)
0.0911***
(0.0191)
0.0174***
(0.0043)
0.0130***
(0.0021)
-0.0315***
(0.0031)
-0.0639**
(0.0294)
0.0029***
(0.0010)
0.0004***
(9.42e–05)
0.0025***
(0.0007)
-2.13e–06
(1.37e–06)
-0.0065
(0.0040)
T A B L E 4 . The Role of the Main Factors in the Post–2011 Slowdowna
0.0906***
(0.0191)
0.0174***
(0.0043)
0.0115***
(0.0021)
-0.0313***
(0.0031)
-0.0638**
(0.0292)
0.0029***
(0.0010)
0.0004***
(9.94e–05)
0.0025***
(0.0007)
-2.24e–06
(1.46e–06)
-0.0072
(0.0044)
(4)
0.0904***
(0.0191)
0.0174***
(0.0043)
0.0114***
(0.0021)
-0.0315***
(0.0031)
-0.0641**
(0.0294)
0.0029***
(0.0010)
0.0004***
(9.40e–05)
0.0026***
(0.0007)
-2.09e–06
(1.36e–06)
-0.0097**
(0.0039)
(5)
0.0909***
(0.0189)
0.0174***
(0.0043)
0.0116***
(0.0021)
-0.0312***
(0.0032)
-0.0638**
(0.0293)
0.0033***
(0.0009)
0.0004***
(9.41e–05)
0.0025***
(0.0007)
-2.09e–06
(1.39e–06)
-0.0076*
(0.0042)
(6)
0.0906***
(0.0191)
0.0175***
(0.0043)
0.0114***
(0.0021)
-0.0316***
(0.0031)
-0.0637**
(0.0291)
0.0029***
(0.0010)
0.00037***
(9.65e–05)
0.0025***
(0.0007)
-2.29e–06
(1.63e–06)
-0.0069
(0.0046)
(continued)
(7)
62,632
0.052
12,190
36
7.809***
(0.960)
(1)
62,632
0.052
12,190
36
7.868***
(0.966)
0.0038
(0.0026)
(2)
62,632
0.053
12,190
36
7.799***
(0.959)
-0.0075**
(0.0036)
(3)
62,632
0.052
12,190
36
7.710***
(0.941)
-0.0045*
(0.0025)
(4)
62,632
0.052
12,190
36
7.862***
(0.978)
-0.0013***
(0.0004)
(5)
62,632
0.052
12,190
36
7.809***
(0.968)
-0.0023*
(0.0012)
(6)
62,632
0.052
12,190
36
0.0004
(0.0005)
7.691***
(0.936)
(7)
* Statistically significant at the 10 percent level.
** Statistically significant at the 5 percent level.
*** Statistically significant at the 1 percent level.
a. The dependent variable is the investment-capital ratio (ICR), with the stock of capital lagged one period. The regressions control for time and firm-level fixed effects. Robust standard errors (clustered by country)
are in parentheses.
Summary statistic
No. observations
R2
No. firms
No. countries
Constant
Recent * commodity export prices
Recent * change in debt
Recent * capital inflows
Recent * leverage (t-1)
Recent * cash flow
Recent * q
Explanatory variable
T A B L E 4 . The Role of the Main Factors in the Post–2011 Slowdowna (Continued)
Nicolás E. Magud and Sebastián Sosa
177
The contribution of each of the determinants to the post-2011 downturn in
the investment-to-capital ratio in the average firm is computed by multiplying
this period’s change in each factor by its corresponding estimated marginal
effect. Based on these regional regressions, the marginal effect of each variable
in the post-2011 period is computed as the sum of the coefficient associated
with that variable and the coefficient on the interaction term (of that variable
with the RECENT dummy), if the latter is statistically significant. Then, this
marginal effect is multiplied by the change in the explanatory variable since
2011 to compute the overall contribution of the latter to the recent slowdown.
Formally, the contribution of each factor X in region j (conditional on being
statistically significant) is given by
(β + η ) ∆ X
h
j
h
j
j
2011–13
CF
∆DEBTj,t x
for X j = j,t , LEV j,t −1 ,
, P j,t −1, KI j,t ,
K j,t −1
K j,t −1
j = LAC, ASIA, EUR, Other.
The recent weakening in business investment in the average firm can largely
be explained by the evolution of its main explanatory factors (figure 5).29 However, our results suggest that the relative contribution of each of the determinants has varied across regions. Lower commodity export prices emerge
as the largest contributor to the slowdown in Latin American and Caribbean
economies. The substantial contributions of weaker commodity prices to the
decline in private investment growth observed since 2011 is not surprising
given the large share of commodity sectors in private investment in this region.
Lower expectations of firms’ future profitability (as measured by Tobin’s q)
were the primary factor behind the weakening of investment in emerging
Europe, other emerging markets, and emerging Asia. This is likely to reflect,
at least partly, the downward revisions to potential growth observed in many
emerging markets during this period, as well as a general sense of leaner
times associated with weaker external demand and tighter global financial
conditions.30
Corporate investment has also been influenced by the declining availability
of international financing in recent years, particularly in emerging Asia. A
29. The sum of the contributions of each variable adds to the fitted value presented in the
figure. Thus, the illustrated fitted value does not include the impact of fixed effects.
30. Potential GDP growth has slowed considerably in emerging markets as a whole, by
about 1.2 percentage points since 2011. See IMF (2015, chap. 3).
178
E C O N O M I A , Fall 2017
F I G U R E 5 . Contributions to the Post–2011 Slowdowna
Percent
50
0
–50
–100
–150
LAC
Asia
Europe
Other
Tobin’s q
Cash flow
Leverage
Change in debt
Net capital inflows
Commodity export prices
EMs total
Interest expense ratio
Source: Authors’ calculations.
a. Relative contribution of each factor to the 2011–13 investment slowdown.
number of economies have seen a moderation in capital inflows since 2012.31
Our firm-level regressions suggest that this explains a nonnegligible share of
the investment slowdown. Higher corporate leverage (presumably increasing
the external finance premium) and lower internal cash flow have also played
a role, especially in Asian emerging markets.32
Robustness
We check the robustness of our results in several ways. First, we estimate
the model using the Arellano-Bond difference-in-differences approach. The
results for the baseline specification remain broadly unchanged (table 5).
Second, we use cash stock rather than cash flow to measure the availability of internal funds. Some previous studies use the cash stock because
they assume it is less likely to be associated with future growth opportunities
31. See IMF (2013, chap. 4; 2014b).
32. The result for leverage is in line with IMF (2014a, chap. 2).
72,049
72,016
23.1700***
(1.0710)
-0.2310***
(0.0075)
0.0151***
(0.0013)
0.00649***
(0.0015)
-0.2330***
(0.0075)
0.0155***
(0.0013)
23.2300***
(1.0790)
(2)
(1)
72,001
23.6700***
(1.0860)
-0.2330***
(0.0075)
0.0151***
(0.0013)
0.00653***
(0.0015)
-0.0801***
(0.0058)
(3)
63,098
22.4900***
(1.1000)
-0.2280***
(0.0080)
0.0139***
(0.0014)
0.0140***
(0.0026)
-0.0800***
(0.0062)
-0.0245
(0.0254)
(4)
63,090
22.3400***
(1.0960)
-0.2280***
(0.0080)
0.0137***
(0.0014)
0.0140***
(0.0025)
-0.0737***
(0.0062)
-0.0233
(0.0255)
0.0026***
(0.00077)
(5)
48,459
17.4000***
(1.2710)
-0.2610***
(0.0094)
0.0132***
(0.0016)
0.0132***
(0.0030)
-0.0714***
(0.0074)
-0.0274
(0.0280)
0.0021***
(0.0008)
0.0005***
(5.09e–05)
(6)
48,459
17.3900***
(1.2710)
-0.2620***
(0.0094)
0.0132***
(0.0016)
0.0131***
(0.0030)
-0.0704***
(0.0073)
-0.0240
(0.0280)
0.0021***
(0.0008)
0.0005***
(5.08e–05)
0.0023***
(0.0003)
(7)
47,742
-0.2620***
(0.0095)
0.0126***
(0.0016)
0.0127***
(0.0030)
-0.0701***
(0.0073)
-0.0289
(0.0285)
0.0021***
(0.0008)
0.0004***
(5.10e–05)
0.0025***
(0.0003)
7.57e–06***
(1.74e–06)
17.1300***
(1.2820)
(8)
*** Statistically significant at the 1 percent level.
a. The dependent variable is the investment-capital ratio (ICR), with the stock of capital lagged one period. The regressions control for time and firm-level fixed effects. Robust standard errors (clustered by country)
are in parentheses.
Summary statistic
No. observations
Constant
Uncertainty
Net capital inflows
Commodity export price (t-1)
Change in debt
Interest expense ratio (t-1)
Leverage (t-1)
Cash flow
q
ICR (t-1)
Explanatory variable
T A B L E 5 . Robustness: Arellano–Bond Specifcationa
180
E C O N O M I A , Fall 2017
T A B L E 6 . Cash Stocka
Explanatory variable
ICR (t-1)
q
(1)
(2)
(3)
(4)
(5)
(6)
0.0967***
(0.0126)
0.0207***
(0.0045)
0.0934***
(0.0134)
0.0201***
(0.0046)
0.00065***
(0.0002)
0.1060***
(0.0164)
0.0196***
(0.0048)
0.0027**
(0.0010)
-0.0390***
(0.0035)
-0.0662**
(0.0282)
0.0036***
(0.0010)
0.0933***
(0.0198)
0.0186***
(0.0047)
0.0024**
(0.0010)
-0.0374***
(0.0030)
-0.0585*
(0.0289)
0.0033***
(0.0012)
0.0005***
(0.0001)
0.0916***
(0.0198)
0.0183***
(0.0045)
0.0024**
(0.0010)
-0.0368***
(0.0032)
-0.0568*
(0.0300)
0.0033***
(0.0012)
0.0005***
(0.0001)
0.0021***
(0.0007)
8.8320***
(1.0170)
8.7800***
(1.0700)
9.7000***
(1.0890)
9.0860***
(0.9670)
8.9470***
(0.9730)
0.0889***
(0.0201)
0.0181***
(0.0044)
0.0023**
(0.0010)
-0.0369***
(0.0033)
-0.0541*
(0.0304)
0.0033***
(0.0012)
0.0005***
(0.0001)
0.0022***
(0.0007)
-1.71e–06
(1.52e–06)
8.5010***
(1.1110)
94,183
0.030
16,512
38
88,273
0.032
15,281
36
79,319
0.056
14,126
36
60,541
0.048
11,414
34
60,541
0.050
11,414
34
59,398
0.048
11,344
34
Cash stock
Leverage (t-1)
Interest expense
ratio (t-1)
Change in debt
Commodity export
price (t-1)
Net capital inflows
Uncertainty
Constant
Summary statistic
No. observations
R2
No. firms
No. countries
* Statistically significant at the 10 percent level.
** Statistically significant at the 5 percent level.
*** Statistically significant at the 1 percent level.
a. The dependent variable is the investment-capital ratio (ICR), with the stock of capital lagged one period.The regressions control for time
and firm-level fixed effects. Robust standard errors (clustered by country) are in parentheses.
than the cash flow measure.33 The results are reported in table 6. Using cash
stock rather than cash flow does not alter the results. Specifically, Tobin’s q,
lagged leverage, the change in debt, commodity export prices, and the availability of foreign financing all have similar coefficients as before, in terms of
both magnitude and statistical significance. Cash stock is also a significant
explanatory variable of firms’ capital spending, with a positive and statistically significant coefficient.
As a third test, we include additional controls (table 7). In particular, real
GDP growth is added as a proxy for aggregate economic activity—with the
33. For example, Harrison, Love, and McMillan (2004). See Love (2003) for further
discussion.
181
Nicolás E. Magud and Sebastián Sosa
T A B L E 7 . Other Robustness Checksa
Explanatory variable
ICR (t-1)
q
Cash flow
Leverage (t-1)
Interest expense ratio (t-1)
Change in debt
Commodity export price (t-1)
Net capital inflows
Uncertainty
Commodity import price (t-1)
(1)
(2)
(3)
0.0904***
(0.0191)
0.0177***
(0.0043)
0.0115***
(0.0021)
-0.0318***
(0.0031)
-0.0661**
(0.0298)
0.0029***
(0.0010)
0.0003**
(0.0001)
0.0024***
(0.0007)
-1.11e–06
(1.44e–06)
0.0002
(0.0003)
0.0912***
(0.0192)
0.0168***
(0.0043)
0.0114***
(0.0021)
-0.0315***
(0.0031)
-0.0622**
(0.0288)
0.0028***
(0.0010)
0.0005***
(9.87e–05)
0.0023***
(0.0006)
-9.56e–07
(1.21e–06)
0.0913***
(0.0190)
0.0180***
(0.0043)
0.0115***
(0.0021)
-0.0324***
(0.0029)
-0.0691**
(0.0291)
0.0029***
(0.0010)
0.0004***
(0.0001)
Real GDP growth (t-1)
-9.69e–07
(1.22e–06)
0.0004***
(2.96e–05)
0.0020***
(0.0001)
-5.09e–06***
(7.43e–07)
0.0010*
(0.0006)
Change in debt (t-1)
Summary statistic
No. observations
R2
No. firms
No. countries
0.0025***
(0.0004)
0.0277***
(0.0004)
0.0038***
(0.0002)
-0.0294***
(0.0007)
0.0849***
(0.0073)
0.0014*
(0.0007)
Net capital inflows (t-1)
Constant
(4)
8.8310***
(1.0210)
8.4800***
(0.9700)
8.7870***
(1.006)
0.0007***
(9.60e–05)
5.3430***
(0.1840)
62,632
0.052
12,190
36
62,632
0.052
12,190
36
62,632
0.050
12,190
36
209,726
0.036
35,047
36
* Statistically significant at the 10 percent level.
** Statistically significant at the 5 percent level.
*** Statistically significant at the 1 percent level.
a. The dependent variable is the investment-capital ratio (ICR), with the stock of capital lagged one period.The regressions control for time
and firm-level fixed effects. Robust standard errors (clustered by country) are in parentheses.
previous results also holding. Commodity import prices are included as additional regressors, since they may affect the firms’ cost of inputs, particularly
in commodity-importer economies. However, this variable is not statistically
significant with all the other coefficients unchanged. We also lagged capital
inflows and the change in debt to mitigate potential endogeneity problems, and
the results again remain unaltered. In all these alternative specifications, the
182
E C O N O M I A , Fall 2017
T A B L E 8 . Excluding Countries with the Most Firmsa
Explanatory variable
ICR (t-1)
q
(1)
(2)
(3)
(4)
(5)
(6)
0.0850***
(0.0150)
0.0250***
(0.0020)
0.0849***
(0.0148)
0.0243***
(0.0019)
0.00581***
(0.0017)
0.0964***
(0.0197)
0.0231***
(0.0020)
0.0115***
(0.0024)
-0.0301***
(0.0034)
-0.0458*
(0.0242)
0.0033***
(0.0011)
0.0919***
(0.0232)
0.0238***
(0.0019)
0.0110***
(0.0021)
-0.0319***
(0.0037)
-0.0486
(0.0289)
0.0029**
(0.0012)
0.0004***
(0.0002)
0.0898***
(0.0230)
0.0231***
(0.0021)
0.0110***
(0.0021)
-0.0312***
(0.0039)
-0.0438
(0.0280)
0.0029**
(0.0012)
0.0005***
(0.0001)
0.0020***
(0.0007)
7.1560***
(0.8940)
7.2070***
(0.8760)
8.2370***
(0.9810)
8.5020***
(0.9300)
8.2740***
(0.9670)
0.0859***
(0.0232)
0.0230***
(0.0021)
0.0107***
(0.0020)
-0.0313***
(0.0040)
-0.0396
(0.0279)
0.0029**
(0.0012)
0.0005***
(0.0001)
0.0021***
(0.0007)
-3.59e–08
(1.15e–06)
7.8330***
(1.1850)
57,851
0.029
10,372
35
57,837
0.035
10,372
35
50,580
0.061
9,392
35
44,416
0.059
8,558
34
44,416
0.061
8,558
34
43,249
0.059
8,486
34
Cash flow
Leverage (t-1)
Interest expense
ratio (t-1)
Change in debt
Commodity export
price (t-1)
Net capital inflows
Uncertainty
Constant
Summary statistic
No. observations
R2
No. firms
No. countries
* Statistically significant at the 10 percent level.
** Statistically significant at the 5 percent level.
*** Statistically significant at the 1 percent level.
a. The dependent variable is the investment-capital ratio (ICR), with the stock of capital lagged one period.The regressions control for time
and firm-level fixed effects. Robust standard errors (clustered by country) are in parentheses.
positive relationship between commodity export prices and firms’ investment
remains statistically and economically significant.
Fourth, we estimate the model without the countries with the largest number of firms, such as China, Korea, and Taiwan, to rule out the possibility that
these countries are driving the results (table 8). Our results hold when we
exclude these countries from the sample. Although not shown here, results
also hold if we add firm-specific sales as a control.
Fifth, we exclude firms in the lower decile of capital stock levels, to ensure
that they are not biasing the results, and the results remain robust. We also
run quantile regression, with the results again holding. Another extension to
Nicolás E. Magud and Sebastián Sosa
183
check the performance of the model was to control for firm size and for the
degree of internationalization of the firm. Once again, our main results did
not change.34
In our last set of robustness tests, we consider a specification including country fixed effects, and the results remain unaltered. To control for
time effects, we use year dummy variables, which reveal a negative trend in
investment-to-capital ratios. Thus, we then use a trend variable rather than
year dummy variables, and the baseline results do not change.35 Finally, we
also estimate the model including country-time dummy variables instead of
the country-specific macroeconomic variables. The coefficients on the firmlevel variables do not change substantially (in terms of both statistical and
economic significance).36
To sum up, we find that beyond the standard firm-level variables used
to explain investment, country-specific macroeconomic variables—notably
commodity export prices—are important determinants of firms’ investment
decisions, and this result appears to be quite robust.
Concluding Remarks
We find that commodity export prices are key to explaining firm-level
investment decisions, an aspect that appears to have been overlooked in the
past. As commodity export prices rise, private sector firms increase their
investment ratios. This finding is based on an analysis of business investment
34. We find that larger firms and firms that are highly integrated with international financial
markets, all else equal, tend to invest more. These results are available on request.
35. In the extension incorporating the RECENT dummy variable, the trend variable is used
to capture time effects, since having both year dummy variables and the RECENT dummy
variable one would entail identification and interpretation issues.
36. These country-time dummy variables capture time-varying idiosyncratic domestic
factors, which are positively correlated with our country-specific macroeconomic variables—
particularly commodity export prices. Our baseline specification given by equation 13 does not
necessarily capture all possible domestic factors that may influence firms’ investment. This does
not affect the interpretation of our results on commodity export prices, however, since these
are mostly exogenous to the country and most likely are not affected by any other domestic
variables not included in the model. That is, there may be other relevant domestic factors, such
as a political cycle, but this should not be correlated with commodity export prices and therefore
should not be biasing the estimated coefficient of the latter.
184
E C O N O M I A , Fall 2017
using standard panel regression models and firm-level data for about 16,000 firms
for thirty-eight emerging markets over the period 1990–2013. We also include
a simple investment model consistent with this finding.
Moreover, we find that other country-specific macroeconomic variables
such as profitability, debt stocks and flows, the availability of external financing, and financial constraints also affect private-sector investment decisions,
in line with the existing literature. We document which of all these factors
drove the recent episode of weak investment and how the contribution of
each factor varied across regions. Commodity export prices were particularly
important in Latin America and the Caribbean.
Appendix A: Supplemental Tables
T A B L E A 1 . List of Countries and Number of Firms in Sample
Country
No. firms
Country
No. firms
Argentina
Brazil
Bulgaria
Chile
China
Colombia
Croatia
Czech Republic
Egypt
Hungary
India
Indonesia
Israel
Jordan
Kazakhstan
Korea (South)
Lithuania
Malaysia
Mexico
1,073
3,100
1,164
3,103
22,799
753
545
511
1,227
563
17,480
4,355
3,618
1,538
223
17,245
225
12,814
2,096
Morocco
Pakistan
Peru
Philippines
Poland
Romania
Russian Federation
Serbia
Singapore
Slovakia
Slovenia
South Africa
Sri Lanka
Taiwan
Thailand
Turkey
Ukraine
Venezuela
Vietnam
538
2,342
1,436
2,708
3,602
770
4,998
534
7,982
237
361
5,381
1,551
17,997
7,065
2,453
375
378
3,515
Net capital inflows
Commodity export price (t-1)
Change in debt
Interest expense ratio (t-1)
Leverage (t-1)
Cash flow
q
ICR (t-1)
Explanatory variable
0.1900***
(0.0362)
0.0131***
(0.0030)
0.0133**
(0.0049)
-0.0444***
(0.0089)
-0.0096
(0.0224)
0.0026*
(0.0013)
0.0006**
(0.0002)
0.0030**
(0.0009)
LAC
(1)
0.0790***
(0.0221)
0.0160**
(0.0051)
0.0200***
(0.0046)
-0.0327***
(0.0035)
-0.0782*
(0.0391)
0.0027*
(0.0014)
0.0004***
(0.0001)
0.0025**
(0.0009)
Asia
(2)
Europe
(3)
0.0773**
(0.0320)
0.0224***
(0.0057)
0.00654
(0.0037)
-0.0130
(0.0084)
0.00788
(0.0789)
0.0014
(0.0012)
0.0004***
(0.0001)
0.0040***
(0.0012)
Cash flow
0.1500***
(0.0362)
0.0272***
(0.0020)
0.00868***
(0.0012)
-0.0285*
(0.0118)
-0.134*
(0.0572)
0.0075**
(0.0022)
-0.0002
(0.0004)
0.0019
(0.0012)
Other
(4)
0.1920***
(0.0351)
0.0131***
(0.0030)
0.0136**
(0.0051)
-0.0444***
(0.0089)
-0.0113
(0.0214)
0.0026*
(0.0013)
0.0006**
(0.0002)
0.0029**
(0.0009)
LAC
(5)
T A B L E A 2 . Regional Decomposition: Interaction of RECENT with Cash Flow and Commodity Export Pricesa
0.0784***
(0.0219)
0.0160**
(0.0051)
0.0190***
(0.0038)
-0.0326***
(0.0034)
-0.0747*
(0.0376)
0.0027*
(0.0014)
0.0003**
(0.0001)
0.0025**
(0.0009)
Asia
(6)
0.0777**
(0.0314)
0.0229***
(0.0058)
0.0014
(0.0012)
-0.0132
(0.0088)
0.00356
(0.0772)
0.0008
(0.0016)
0.0004***
(0.0001)
0.0040***
(0.0012)
Europe
(7)
Commodity export prices
0.1500***
(0.0361)
0.0264***
(0.0021)
0.0084***
(0.0010)
-0.0283*
(0.0119)
-0.1300*
(0.0581)
0.0076**
(0.0021)
-0.0002
(0.0004)
0.0018
(0.0013)
(continued)
Other
(8)
8.7240***
(0.9790)
-0.0877
(1.9610)
47,506
0.049
8,894
10
-5.26e–06***
(1.49e–06)
-0.0046
(0.0049)
-0.0034
(0.0057)
-1.91e–06**
(6.34e–07)
-0.0228
(0.0121)
0.0092
(0.0061)
4,622
0.087
775
7
Asia
(2)
LAC
(1)
6,404
0.047
1,615
13
8.7060
(6.0600)
-3.15e–06
(3.11e–06)
-0.0006
(0.0118)
-0.0086
(0.0051)
Europe
(3)
4,100
0.145
906
6
10.3600***
(2.5480)
2.04e–06
(4.75e–06)
-0.0272***
(0.0042)
-0.0091**
(0.0026)
Other
(4)
4,622
0.087
775
7
0.0021***
(0.0005)
0.16800
(1.8850)
-1.67e–06**
(6.33e–07)
-0.0095
(0.0133)
LAC
(5)
47,506
0.050
8,894
10
-0.0012***
(0.0003)
8.2090***
(0.9830)
-7.28e–06***
(1.44e–06)
-0.0158**
(0.0058)
Asia
(6)
6,404
0.044
1,615
13
0.0004
(0.0003)
8.8400
(6.1120)
-3.03e–06
(3.18e–06)
0.00035
(0.0128)
Europe
(7)
Commodity export prices
4,100
0.144
906
6
-0.0003
(0.0008)
10.4200**
(2.5870)
2.59e–06
(5.24e–06)
-0.0328***
(0.0060)
Other
(8)
* Statistically significant at the 10 percent level.
** Statistically significant at the 5 percent level.
***S tatistically significant at the 1 percent level.
a. The dependent variable is the investment-capital ratio (ICR), with the stock of capital lagged one period. The regressions control for time and firm-level fixed effects. Robust standard errors (clustered by country)
are in parentheses.
Summary statistic
No. observations
R2
No. firms
No. countries
Constant
Recent * commodity export prices
Recent * cash flow
Recent
Uncertainty
Explanatory variable
Cash flow
T A B L E A 2 . Regional Decomposition: Interaction of RECENT with Cash Flow and Commodity Export Pricesa (Continued)
Net capital inflows
Commodity export price (t-1)
Change in debt
Interest expense ratio (t-1)
Leverage (t-1)
Cash flow
q
ICR (t-1)
Explanatory variable
0.1920***
(0.0350)
0.0131***
(0.0030)
0.0135**
(0.0050)
-0.0454***
(0.0089)
-0.0100
(0.0221)
0.0026*
(0.0013)
0.0006**
(0.0002)
0.0029**
(0.0009)
LAC
(1)
0.0788***
(0.0221)
0.0160**
(0.0052)
0.0191***
(0.0039)
-0.0322***
(0.0035)
-0.0778*
(0.0391)
0.0027*
(0.0014)
0.0004**
(0.0001)
0.0025**
(0.0009)
Asia
(2)
Q
0.0774**
(0.0311)
0.0230***
(0.0059)
0.00138
(0.0012)
-0.0132
(0.0087)
0.00282
(0.0773)
0.0008
(0.0016)
0.0004**
(0.0001)
0.0040***
(0.0012)
Europe
(3)
0.1500***
(0.0361)
0.0264***
(0.0022)
0.00840***
(0.0010)
-0.0282*
(0.0117)
-0.1300*
(0.0584)
0.0076**
(0.0021)
-0.0002
(0.0004)
0.0019
(0.0013)
Other
(4)
T A B L E A 3 . Regional Decomposition: Interaction of RECENT with Q and Leveragea
0.1920***
(0.0346)
0.0139***
(0.0032)
0.0136**
(0.0051)
-0.0443***
(0.0088)
-0.00996
(0.0216)
0.0026*
(0.0013)
0.0005**
(0.0002)
0.0030**
(0.0009)
LAC
(5)
0.0787***
(0.0220)
0.0155**
(0.0053)
0.0190***
(0.0038)
-0.0328***
(0.0035)
-0.0790*
(0.0405)
0.0027*
(0.0014)
0.0004***
(0.0001)
0.0024**
(0.0009)
Asia
(6)
Leverage
0.0771**
(0.0313)
0.0213***
(0.0058)
0.00136
(0.0012)
-0.0138
(0.0087)
0.00298
(0.0775)
0.0008
(0.0016)
0.0004***
(0.0001)
0.0040***
(0.0012)
Europe
(7)
0.1500***
(0.0360)
0.0266***
(0.0017)
0.00840***
(0.0010)
-0.0282*
(0.0118)
-0.1300*
(0.0583)
0.0076**
(0.0022)
-0.0002
(0.0004)
0.0019
(0.0012)
(continued)
Other
(8)
4,622
0.087
775
7
47,506
0.049
8,894
10
8.5680***
(0.9630)
-5.73e–06***
(1.56e–06)
-0.0038
(0.0053)
-0.0065*
(0.0034)
-1.78e–06**
(6.49e–07)
-0.0234*
(0.0111)
0.0102*
(0.0050)
0.0863
(1.9420)
Asia
(2)
LAC
(1)
6,404
0.045
1,615
13
8.4280
(6.0420)
-3.31e–06
(3.27e–06)
-0.0006
(0.0106)
-0.0076
(0.0061)
Europe
(3)
4,100
0.144
906
6
10.3900***
(2.5060)
2.39e–06
(5.18e–06)
-0.0306***
(0.0027)
-0.0024
(0.0098)
Other
(4)
4,622
0.087
775
7
-0.0047
(0.0057)
-0.0949
(1.9390)
-1.96e–06**
(5.82e–07)
-0.0126
(0.0149)
LAC
(5)
47,506
0.049
8,894
10
0.0044
(0.0028)
8.7820***
(0.9910)
-5.11e–06***
(1.53e–06)
-0.0116
(0.00824)
Asia
(6)
Leverage
6,404
0.045
1,615
13
0.0105***
(0.0029)
9.0410
(6.1760)
-2.86e–06
(3.20e–06)
-0.0144
(0.00936)
Europe
(7)
4,100
0.144
906
6
-0.0020
(0.0094)
10.3600**
(2.7380)
2.31e–06
(4.84e–06)
-0.0285
(0.0170)
Other
(8)
* Statistically significant at the 10 percent level.
** Statistically significant at the 5 percent level.
*** Statistically significant at the 1 percent level.
a. The dependent variable is the investment-capital ratio (ICR), with the stock of capital lagged one period. The regressions control for time and firm-level fixed effects. Robust standard errors (clustered by country)
are in parentheses.
Summary statistic
No. observations
R2
No. firms
No. countries
Constant
Recent * q
Recent * leverage (t-1)
Recent
Uncertainty
Explanatory variable
Q
T A B L E A 3 . Regional Decomposition: Interaction of RECENT with Q and Leveragea (Continued)
Net capital inflows
Commodity export price (t-1)
Change in debt
Interest expense ratio (t-1)
Leverage (t-1)
Cash flow
q
ICR (t-1)
Explanatory variable
0.1930***
(0.0348)
0.0131***
(0.0029)
0.0139**
(0.0053)
-0.0444***
(0.0089)
-0.0107
(0.0218)
0.0027
(0.0015)
0.0005**
(0.0002)
0.0029**
(0.0009)
LAC
(1)
0.0790***
(0.0220)
0.0160**
(0.0052)
0.0191***
(0.0038)
-0.0326***
(0.0036)
-0.0780*
(0.0394)
0.0030**
(0.0011)
0.0004***
(0.0001)
0.0025**
(0.0009)
Asia
(2)
Europe
(3)
0.0767**
(0.0320)
0.0228***
(0.0057)
0.0023
(0.0016)
-0.0104
(0.0069)
0.00146
(0.0772)
0.0024
(0.0015)
0.0004***
(0.0001)
0.0040***
(0.0012)
Capital inflows
0.1510***
(0.0358)
0.0263***
(0.0021)
0.0084***
(0.0011)
-0.0283*
(0.012)
-0.1300*
(0.0578)
0.0077**
(0.0023)
-0.0002
(0.0004)
0.0019
(0.0012)
Other
(4)
0.1920***
(0.0348)
0.0131***
(0.0030)
0.0136**
(0.0051)
-0.0443***
(0.0089)
-0.0101
(0.0222)
0.0026*
(0.0013)
0.0005**
(0.0002)
0.0029**
(0.0009)
LAC
(5)
T A B L E A 4 . Regional Decomposition: Interaction of RECENT with Capital Inflows and Change in Debta
0.0786***
(0.0221)
0.0160**
(0.0051)
0.0191***
(0.0039)
-0.0327***
(0.0034)
-0.0772*
(0.0394)
0.0027*
(0.0014)
0.0004***
(0.00012)
0.0026**
(0.0009)
Asia
(6)
Europe
(7)
0.0777**
(0.0311)
0.0230***
(0.0059)
0.0014
(0.0012)
-0.0134
(0.0087)
0.00282
(0.0777)
0.0008
(0.0016)
0.0004***
(0.0001)
0.0040***
(0.0012)
Change in debt
0.1500***
(0.0362)
0.0261***
(0.0023)
0.0084***
(0.0010)
-0.0282*
(0.0117)
-0.1300*
(0.0590)
0.0076**
(0.0021)
-0.0002
(0.0004)
0.0022
(0.0019)
(continued)
Other
(8)
8.7290***
(0.9870)
-0.0678
(1.9760)
47,506
0.050
8,894
10
-5.26e–06***
(1.49e–06)
-0.0048
(0.0051)
-0.0015
(0.0009)
-1.90e–06**
(6.37e–07)
-0.0197
(0.0129)
-0.0032
(0.0067)
4,622
0.087
775
7
Asia
(2)
LAC
(1)
6,404
0.047
1,615
13
8.8170
(6.1700)
-3.13e–06
(3.13e–06)
-0.0023
(0.0115)
-0.0036
(0.0027)
Europe
(3)
4,100
0.144
906
6
10.3800***
(2.5450)
2.28e–06
(4.88e–06)
-0.0311***
(0.0038)
-0.0022
(0.0040)
Other
(4)
4,622
0.087
775
7
0.0006
(0.0073)
-0.0612
(2.0630)
-1.89e–06**
(7.10e–07)
-0.0222
(0.0246)
LAC
(5)
47,506
0.050
8,894
10
-0.0014**
(0.0005)
8.7790***
(1.0080)
-5.11e–06***
(1.49e–06)
-0.0082
(0.0051)
Asia
(6)
Europe
(7)
6,404
0.044
1,615
13
0.0013
(0.0025)
8.6590
(6.1320)
-3.15e–06
(3.15e–06)
-0.0033
(0.0089)
Change in debt
4,100
0.144
906
6
-0.0008
(0.0020)
10.5100***
(2.3650)
2.02e–06
(5.04e–06)
-0.0276*
(0.0126)
Other
(8)
* Statistically significant at the 10 percent level.
** Statistically significant at the 5 percent level.
*** Statistically significant at the 1 percent level.
a. The dependent variable is the investment-capital ratio (ICR), with the stock of capital lagged one period. The regressions control for time and firm-level fixed effects. Robust standard errors (clustered by country)
are in parentheses.
Summary statistic
No. observations
R2
No. firms
No. countries
Constant
Recent * capital inflows
Recent * change in debt
Recent
Uncertainty
Explanatory variable
Capital inflows
T A B L E A 4 . Regional Decomposition: Interaction of RECENT with Capital Inflows and Change in Debta (Continued)
Nicolás E. Magud and Sebastián Sosa
191
Appendix B: Effects of Terms-of-Trade Shocks
We start by replicating equation 11:
(A1)
1
qt = π K K t , θt − cK I t , K t + 1 − δ Et qt +1 .
R
(
)
(
)
(
)
After subtracting qt+1 from both sides, we can rearrange the equation as follows:
(A2)
1− δ + R
Et ∆qt +1 = π K K t , θt − cK I t , K t +
Et qt +1 ,
R
(
)
(
)
where DEtqt+1 = qt+1 - qt. In steady state, DEtqt+1 = 0 holds for
(A3)
Et qt +1 =
R
c I , K − π K K t , θt K t .
1− δ + R K t t
(
)
(
)
Thus, the slope of the EtDqt+1 = 0 line is given by
(A4)
∂ Et qt +1
R
c I , K − π KK K t , θt > 0
=
∂ Kt
1 − δ + R KK t t
(
)
(
)
given that cKK (It, Kt) > 0 and pKK(Kt, qt) < 0.
From equations 10 and 3,
(A5)
1 E q
K t +1 − K t = ∆Κ t +1 = µ − δ + t t +1 + et − 1 .
b
R
(
)
Thus,
(A6)
1 E q
∆Κ t +1 = 0 µ − δ = t t +1 + et − 1 ,
b R
implying a zero slope.
Figure B1 shows the phase diagram, which uses the facts that
(A7)
∂ Et ∆qt +1
∂ Kt
(
)
(
)
= π KK K t , θt − cKK I t , K t < 0
∆Et qt +1=0
192
E C O N O M I A , Fall 2017
F I G U R E B 1 . Phase Diagram
Q
Qt+1 = 0
A’
Kt+1 = 0
A
A’
K
and
(A8)
∂∆K t +1 K t
=
> 0.
∂ Et qt +1 bR
A real appreciation (that is, an increase in et), shifts the Et Dqt+1 = 0 schedule upward, while DKt +1 = 0 remains unaltered. Figures B2 and B3 pre
sent the movements in the phase diagram, together with the dynamics over
time of investment and q, in response to permanent and transitory shocks,
respectively.
Nicolás E. Magud and Sebastián Sosa
193
F I G U R E B 2 . Permanent Increase in the Terms of Trade
Q
B’
Qt+1 = 0
A’
1Kt+1 = 0
B
A
0 Kt+1 = 0
B’
A’
K
Q
K
t
t
F I G U R E B 3 . Transitory Increase in the Terms of Trade
Q
C
B’
Qt+1 = 0
A’
1Kt+1 = 0
B
D
0Kt+1 = 0
B’
A
A’
K
Q
K
t
t
194
E C O N O M I A , Fall 2017
References
Abel, Andrew B., and Janice C. Eberly. 2011. “How Q and Cash Flow Affect Investment without Frictions: An Analytical Explanation.” Review of Economic Studies
78(4): 1179–200.
Baum, Christopher F., Mustafa Caglayan, and Oleksandr Talavera. 2008. “Uncertainty Determinants of Firm Investment.” Economic Letters 98(3): 282–87.
Blanchard, Olivier, Changyong Rhee, and Lawrence Summers. 1993. “The Stock
Market, Profit, and Investment.” Quarterly Journal of Economics 108(1): 115–36.
Bloom, Nicholas, Stephen Bond, and John van Reenen. 2001. “The Dynamics of
Investment under Uncertainty.” Working Paper 01/5. London: Institute for Fiscal
Studies.
Carpenter, Robert, and Alessadra Guariglia. 2008. “Cash Flow, Investment, and Investment Opportunities: New Tests Using U.K. Panel Data.” Journal of Banking and
Finance 32(9):1894–906.
Dixit, Avinash K., and Robert S. Pindyck. 1994. Investment under Uncertainty.
Princeton University Press.
Druck, Pablo, Nicolás E. Magud, and Rodrigo Mariscal. 2015. “Collateral Damage:
Dollar Strength and Emerging Markets’ Growth.” Working Paper 15/179. Washington: International Monetary Fund.
Fazzari, Steven M., R. Glenn Hubbard, and Bruce C. Petersen. 1988. “Financing
Constraints and Corporate Investment.” Brookings Papers on Economic Activity
1(1998): 141–95.
———. 2000. “Investment-Cash Flow Sensitivities Are Useful: A Comment on
Kaplan and Zingales.” Quarterly Journal of Economics 115(2): 695–705.
Fernández, Andrés, Andrés Gonzales, and Diego Rodríguez. 2014. “Riding the Commodity Roller Coaster Together: Common Factors in Business Cycles of Emerging
Economies.” Working Paper. Bogotá: Bank of the Republic of Colombia.
Forbes, Kristin J. 2007. “The Microeconomic Evidence on Capital Controls: No Free
Lunch.” In Capital Controls and Capital Flows in Emerging Economies: Policies,
Practices, and Consequences, edited by Sebastian Edwards, pp. 171–99. University of Chicago Press for the National Bureau of Economic Research.
Fornero, Jorge, Markus Kirchner, and Andrés Yany. 2014. “Terms of Trade Shocks and
Investment in Commodity-Exporting Economies.” Working Paper 773. Santiago:
Central Bank of Chile.
Gelos, R. Gaston, and Alejandro M. Werner. 2002. “Financial Liberalization, Credit
Constraints, and Collateral: Investment in the Mexican Manufacturing Sector.”
Journal of Development Economics 67(1): 1–62.
Gilchrist, Simon, and Charles P. Himmelberg. 1995. “Evidence on the Role of Cash
Flow for Investment.” Journal of Monetary Economics 36(3): 541–72.
———. 1999. “Investment; Fundamentals and Finance.” In NBER Macroeconomics Annual 1998, vol. 13, edited by Ben S. Bernanke and Julio J. Rotemberg,
pp. 223–74. MIT Press.
Nicolás E. Magud and Sebastián Sosa
195
Gomes, João F. 2001. “Financing Investment.” American Economic Review 91(5):
1263–85.
Gruss, Bertrand. 2014. “After the Boom: Commodity Prices and Economic Growth
in Latin America and the Caribbean.” Working Paper 14/154. Washington: International Monetary Fund.
Harrison, Ann E., Inessa Love, and Margaret S. McMillan. 2004. “Global Capital Flows
and Financing Constraints.” Journal of Development Economics 75(1): 269–301.
Hayashi, Fumio. 1982. “Tobin’s Marginal Q and Average Q: A Neoclassical Interpretation.” Econometrica 50(1): 213–24.
Hubbard, R. Glenn. 1998. “Capital-Market Imperfections and Investment.” Journal
of Economic Literature 36(1): 193–225.
IMF (International Monetary Fund). 2013. World Economic Outlook (WEO), October
2013: Transitions and Tensions. Washington.
———. 2014a. Regional Economic Outlook: Asia and Pacific, April 2014. Sustaining
the Momentum: Vigilance and Reforms. Washington.
———. 2014b. 2014 Spillover Report. IMF Policy Paper. Washington.
———. 2015. World Economic Outlook (WEO), April 2015: Uneven Growth: Shortand Long-Term Factors. Washington.
Kaplan, Steven N., and Luigi Zingales. 1997. “Do Investment-Cash Flow Sensitivities Provide Useful Measures of Financing Constraints?” Quarterly Journal of
Economics 112(1):169–216.
Laeven, Luc. 2003. “Does Financial Liberalization Reduce Financing Constraints?”
Financial Management 32(1) 5–34.
Leahy, John, and Toni Whited. 1996. “The Effects of Uncertainty on Investment:
Some Stylized Facts.” Journal of Money, Credit, and Banking 28(1):64–83.
Li, Delong, Nicolás E. Magud, and Fabián Valencia. 2015. “Corporate Investment in
Emerging Markets: Financing vs. Real Options Channel.” Working Paper 15/285.
Washington: International Monetary Fund.
Love, Inessa. 2003. “Financial Development and Financing Constraints: International
Evidence from the Structural Investment Model.” Review of Financial Studies
16(3): 765–91.
Love, Inessa, and Lea Zicchino. 2006. “Financial Development and Dynamic Investment Behavior: Evidence from Panel VAR.” Quarterly Review of Economics and
Finance 46(2): 190–210.
Magud, Nicolás E. 2008. “On Asymmetric Business Cycles and the Effectiveness of
Counter-Cyclical Fiscal Policies.” Journal of Macroeconomics 30(3): 885–905.
Ross, Kevin, and Melesse Tashu. 2015. “Investment Dynamics in Peru.” In Peru: Staying the Course of Economic Success, edited by Alejandro M. Werner and Alejandro
Santos. Washington: International Monetary Fund.
Tobin, James. 1969. “A General Equilibrium Approach to Monetary Theory.” Journal
of Money, Credit, and Banking 1(1): 15–29.
Copyright of Economia is the property of Brookings Institution Press and its content may not
be copied or emailed to multiple sites or posted to a listserv without the copyright holder's
express written permission. However, users may print, download, or email articles for
individual use.
1.
Zona, F. (2012). Corporate Investing as a Response to Economic Downturn:
Prospect Theory, the Behavioural Agency Model and the Role of Financial Slack.
British Journal of Management, 23, S42�S57. https://doi-org.prxkeiser.lirn.net/10.1111/j.1467-8551.2012.00818.x
2.
MAGUD, N. E., & SOSA, S. (2017). Corporate Investment in Emerging
Markets: The Role of Commodity Prices. Economia, 18(1), 157�195. Retrieved from
http://prx-keiser.lirn.net/login?url=http%3a%2f%2fsearch.ebscohost.com
%2flogin.aspx%3fdirect%3dtrue%26db%3dbth%26AN%3d126541296%26site%3dehost-live
3.
Hyunseob Kim, & Howard Kung. (2017). The Asset Redeployability Channel:
How Uncertainty Affects Corporate Investment. Review of Financial Studies,
30(1), 245�280. https://doi-org.prx-keiser.lirn.net/10.1093/rfs/hhv076
The Asset Redeployability Channel: How
Uncertainty Affects Corporate Investment
Hyunseob Kim
Cornell University
Howard Kung
London Business School
Received August 29, 2014; editorial decision July 13, 2016 by Editor Itay Goldstein.
There are active secondary markets for corporate assets. The aggregated value
of transactions in these markets amounts to more than a $100 billion in a typical
year (e.g., Warusawitharana 2008; Gavazza 2011).1 Importantly, activities
in used asset markets vary widely across industries. For example, the retail
and electronics manufacturing industries exhibit active trading in used asset
markets, while the computer and pulp & paper manufacturing industries show
relatively infrequent transactions. This variation in trading activity suggests that
We thank two anonymous referees and the editor Itay Goldstein for their comments and guidance. We are
grateful for comments from Einar Bakke (discussant), Ravi Bansal, Effi Benmelech, Nittai Bergman, Nick
Bloom, Michael Brandt, Alon Brav, Murillo Campello, Steven Davis, Francisco Gomes, Dirk Hackbarth, Emi
Nakamura, John Graham, Andrew Karolyi, Seoyoung Kim (discussant), Mauricio Larrain, Jean-Marie Meier,
Roni Michaely, Justin Murfin, Manju Puri, Adriano Rampini, David Robinson, Norman Schürhoff (discussant),
S. Viswanathan, Ivo Welch, Toni Whited, and seminar and conference participants at Cornell University, CUNY
Baruch, Duke University, EFA, FMA Asia, SFS Finance Cavalcade, Rutgers University, Shanghai Advanced
Institute of Finance (SAIF), Stanford Institute for Theoretical Economics (SITE), and University of British
Columbia. We also thank Dawoon Kim, Young Jun Song, and Ercos Valdivieso for excellent research assistance.
Kim gratefully acknowledges financial support from the Kwanjeong Educational Foundation. Any errors are our
own. Supplementary data can be found on The Review of Financial Studies web site. Send correspondence to
Hyunseob Kim, Samuel Curtis Johnson Graduate School of Management, Cornell University, Sage Hall, 114
East Ave, Ithaca, NY 14853; telephone: (607) 255-8335. E-mail: hk722@cornell.edu.
1 Using the SDC Platinum mergers and acquisitions database, we find that in 2012, U.S. firms were involved in
more than 6,000 transactions of used asset sales and the total value of trading was $167 billion.
© The Author 2016. Published by Oxford University Press on behalf of The Society for Financial Studies.
All rights reserved. For Permissions, please e-mail: journals.permissions@oup.com.
doi:10.1093/rfs/hhv076
Advance Access publication August 29, 2016
Downloaded from http://rfs.oxfordjournals.org/ at :: on December 23, 2016
This paper examines how uncertainty affects corporate investment under varying degrees of
asset redeployability. We develop new measures of asset redeployability by accounting for
the usability of assets within and across industries. We identify plausibly exogenous shocks
to economic uncertainty by using major economic and political events. We find that after an
increase in uncertainty, firms using less redeployable capital reduce investment more. More
redeployable assets exhibit higher recovery rates and are traded more actively in secondary
markets. Overall, our results suggest that frictions in redeploying assets affect liquidation
values and therefore make firms cautious about investment decisions under uncertainty.
(JEL G31, D22, D92)
The Review of Financial Studies / v 30 n 1 2017
2 In the financial economics and industrial organizations literature, “asset redeployability” (Williamson 1988;
Shleifer and Vishny 1992) is often referred to as “asset salability” or “asset market liquidity.” (e.g., Benmelech
2009; Gavazza 2011). We use the term “asset redeployability” throughout this paper for consistency.
3 Using the number of potential users (or buyers) is consistent with other approaches of measuring liquidity in
financial markets (e.g., Demsetz 1968; Amihud, Mendelson, and Uno 1999). These papers generally find lower
bid-ask spreads for assets with a larger number of shareholders or float.
246
Downloaded from http://rfs.oxfordjournals.org/ at :: on December 23, 2016
frictions affecting the redeployability of assets may vary significantly across
firms due to search costs for potential buyers and sellers, and firms’ financial
constraints, among other factors (e.g., Ramey and Shapiro 2001; Hennessy,
Levy, and Whited 2007; Gavazza 2011). In this paper, we investigate how
variation in asset redeployability—the extent to which assets have alternative
uses—affects firm investment under uncertainty.2
Costs associated with redeploying assets are an important source of
investment irreversibility (i.e., the wedge between purchase and liquidation
values of capital). Notably, costly capital reversibility creates incentives for
firms to delay investment when there is uncertainty regarding profitability (e.g.,
Dixit and Pindyck 1994). Consider an increase in uncertainty in which both
very bad and very good states of the world become equally more likely, but
the mean of the distribution stays the same. With investment irreversibility,
firms are more sensitive to bad states because disinvestment is more costly
than investment. From a real options perspective, higher irreversibility implies
lower asset liquidation values, and therefore such assets offer less protection
against negative outcomes (e.g., Caballero 1991; Bloom 2009). Thus, even
when expected investment opportunities (first moment) stay the same, an
increase in uncertainty regarding the payoffs (second moment) induces more
irreversible firms to purchase more protection by delaying investment than less
irreversible firms.
To test this theoretical prediction, we need to construct a measure of asset
redeployability and to identify changes in uncertainty. We measure asset
redeployability using the Bureau of Economic Analysis (BEA) capital flow
table, which breaks down capital expenditures into a variety of asset categories
for a broad cross-section of industries. First, we compute the asset-level
redeployability score as the proportion of firms (or industries) that use a given
asset. Thus, the redeployability score would be higher when more firms (or
industries) in the economy use a given asset.3 This approach to measuring
redeployability incorporates the notions of asset specificity (e.g., Williamson
1988) and asset market thickness (e.g., Gavazza 2011). As an extension, we
incorporate the financial constraints of potential buyers and the correlation of
output within industries to capture the deleterious effects of potential buyers’
illiquidity on redeployability, especially during economic downturns (Shleifer
and Vishny 1992). Second, we compute the industry-level redeployability
index as the value-weighted average of each asset’s redeployability score.
The resulting redeployability index exhibits considerable variation across the
The Asset Redeployability Channel: How Uncertainty Affects Corporate Investment
4 Based on this measure, the industries with the least redeployable assets include materials, transportation, and
manufacturing-related, while those with the most redeployable assets are service-related.
5 See Figure 1 for fluctuations of measures of aggregate uncertainty over our sample periods.
247
Downloaded from http://rfs.oxfordjournals.org/ at :: on December 23, 2016
industries.4 Given that industry peers are likely to suffer from the same
operational and financial difficulties as the firm liquidating assets (e.g., Shleifer
and Vishny 1992; Ramey and Shapiro 2001), accounting for the salability of
corporate assets across as well as within industries is an important advantage of
our redeployability measures. Firms in industries with a higher redeployability
index exhibit higher recovery rates and are more actively involved in asset
sales.
To identify sudden and dramatic increases in economic uncertainty, we study
major economic and political events, such as Gulf War I in 1990 and the
9/11 terrorist attacks in 2001. As both quantitative and anecdotal evidence
suggests, uncertainty about consumer demand and profitability, which drive
firms’ investment opportunities, increased significantly after these events (e.g.,
Bloom 2009). In addition, the economy-wide nature of these shocks implies
that they are likely to affect economic uncertainty similarly across firms, which
we verify using stock return volatility as a proxy for firm-level uncertainty
(e.g., Leahy and Whited 1996).
Exploiting these major events as shocks to uncertainty and using Compustat
quarterly data on firm-level investment, we find that asset redeployability
has an economically sizable effect on corporate investment when uncertainty
varies. Our baseline estimates indicate that in response to a sharp increase in
uncertainty during Gulf War I and following 9/11,5 a one-standard-deviation
(SD) decrease in our measure of asset redeployability leads to a 0.06- to
0.14-percentage-point decrease in the quarterly investment rate (i.e., capital
expenditure divided by total assets). This magnitude is 5% to 19% of the
median quarterly investment rate and accounts for 37% to 47% of average
drops in investment after the shocks. These estimates imply that the typical
variation in asset redeployability leads to a wide dispersion across firms in
their investment response to fluctuating uncertainty. More generally, using a
panel of firm quarters from 1989 to 2009, we find that the investment of firms
with less redeployable capital decreases (increases) significantly more when
proxies for aggregate economic uncertainty increase (decrease).
Our paper contributes to the empirical literature on investment irreversibility.
Despite a large theoretical literature on irreversible investment (e.g., Bernanke
1983; McDonald and Siegel 1986; Abel and Eberly 1996), empirical papers
have focused on examining the average relation between uncertainty and
investment (e.g., Leahy and Whited 1996; Julio and Yook 2012; Gulen and Ion
2016). However, relatively little attention has been paid to testing important
cross-sectional predictions regarding irreversibility, which is partly due to
The Review of Financial Studies / v 30 n 1 2017
6 Guiso and Parigi (1999) find that the negative effect of uncertainty on the sensitivity of investment to demand
conditions is more pronounced for more irreversible investments using a simple dichotomous measure of the
existence of used asset markets for Italian firms. However, their main focus is on testing the average uncertainty–
investment relationship. Relatedly, Durnev (2012) finds that the investment–stock price sensitivity decreases in
election years compared with non-election years.
248
Downloaded from http://rfs.oxfordjournals.org/ at :: on December 23, 2016
the difficulty of measuring investment irreversibility.6 Our paper differs from
previous research along two key dimensions. First, we develop new measures
of asset redeployability that vary significantly across industries, suggesting that
accounting for this heterogeneity is potentially important for testing the theory.
Second, we use these measures and plausibly exogenous shocks to aggregate
uncertainty to show that there is indeed a wide dispersion in the effect of
uncertainty on investment depending on the redeployability of capital.
Our paper also relates to the growing empirical literature that uses asset
redeployability to measure asset liquidation values in studying firm policies.
This line of research has focused on the role of asset redeployability in capital
structure outcomes such as debt maturity (e.g., Benmelech 2009), cost of
capital (e.g., Benmelech and Bergman 2009; Ortíz-Molina and Phillips 2014),
and leverage (e.g., Campello and Giambona 2013). In addition, Almeida,
Campello, and Hackbarth (2011) and Gavazza (2011) examine the effect of
asset redeployability on asset reallocation through mergers and trading in
secondary markets. Beutler and Grobety (2013) use asset redeployability as
a proxy for liquidation values to examine the sensitivity of industry growth
to collateral values. Our paper is among the first to empirically investigate the
relation between asset redeployability and corporate investment. Also, our asset
redeployability measures are applicable economy-wide rather than for only a
specific industry.
More broadly, our paper connects to the empirical literature testing
neoclassical investment models. Fazzari, Hubbard, and Petersen (1988)
document that investment is strongly correlated with cash flows (especially
for financially constrained firms), which runs contrary to the prediction that
marginal q is a sufficient statistic for investment (e.g., Hayashi 1982). Gomes
(2001) theoretically shows that cash flows should matter for investment
only if q is ignored or measured with errors. Consistent with this result,
Erickson and Whited (2000) find support for the q theory after accounting for
measurement errors in marginal q and using a generalized method of moments
estimator. Chen, Goldstein, and Jiang (2007) document that the degree of
private information in stock prices is important in determining the investmentto-q sensitivity. In line with Abel and Eberly (1996) and Barnett and Sakellaris
(1998), our paper provides suggestive evidence that there are nonconvexities
in investment adjustment costs, which implies a nonlinear relation between q
and investment.
Finally, the results in this paper support the argument of policymakers
and academics that uncertainty has a negative effect on the economy. We
show that uncertainty, combined with limited redeployability of capital, can
The Asset Redeployability Channel: How Uncertainty Affects Corporate Investment
significantly dampen capital accumulation and ultimately economic growth.
Thus, we add to the growing literature examining the impact of uncertainty
on the macroeconomy by identifying a microeconomic channel through which
the effect operates, namely, frictions in reallocating capital.7 In particular, our
findings suggest that firms using less redeployable capital such as transportation
and manufacturing are affected more by an increase in uncertainty relative to
those using more redeployable capital such as wholesale and retail.
1. Measurement and Data
1.1 Redeployability of corporate assets
We construct measures of asset redeployability using the 1997 Bureau of
Economic Analysis (BEA) capital flow table (hereafter referred to as the “BEA
table”). The table breaks down expenditures on new equipment, software, and
structures by 180 assets for 123 industries,8 covering virtually all economic
sectors in the United States.9 While the BEA table is also available for 1982
and 1992, we employ the 1997 table throughout the analysis for the following
reasons. First, it provides the most detailed breakdown of asset categories and
industries.10 Second, it is based on the most relevant and up-to-date information
on asset usage by industry for our main sample periods, which range from 1989
to 2002. Third, by using the same table, we can maintain the consistency of asset
and industry classifications in our analysis. Nonetheless, we find qualitatively
similar results when we use the 1992 capital flow table. In order to reduce
the influence of sector-level heterogeneity on investment, we confine our main
analysis to the manufacturing industries (NAICS 310000–339999) but use the
full set of industries when examining the external validity of the main results
(see Section 3).
1.1.1 Constructing the asset-level redeployability score. We construct the
main firm-level measures of asset redeployability in three steps. In the first
7 See, for example, Bloom (2009), Fernández-Villaverde et al. (2015), and Baker, Bloom, and Davis (2016).
8 Land is not included in the BEA asset categories. However, we find quantitatively similar results when
incorporating expenditures on land into our measure by assuming land has complete redeployability across
industries.
9 The industry classification employed by the BEA is based on the 1997 North American Industry Classification
System (NAICS). Therefore, we match the 123 BEA industries with Compustat firms using the NAICS code.
10 The 1982 (1992) capital flow table is based on 79 (64) industries and 52 (163) assets. In addition, both tables are
based on SIC industry classifications, while the 1997 table is based on the more modern NAICS classifications.
249
Downloaded from http://rfs.oxfordjournals.org/ at :: on December 23, 2016
This section describes how we (i) construct the measures of asset
redeployability based on an economic link in asset usage across industries,
(ii) identify variation in aggregate uncertainty, and (iii) construct firm samples
for empirical analysis.
The Review of Financial Studies / v 30 n 1 2017
step, we define an asset’s “redeployability score” as the sum of weights of
industries that use the asset among the 123 industries in the BEA table. We
employ various measures of the weights for industries to ensure robustness of
the empirical results. In particular, we use one of the following variables as
the weight of an industry in the economy in a given year: (i) one over the total
number of BEA industries, (ii) the number of all public (i.e., Compustat) firms
in an industry over the total number of all public firms, and (iii) the sum of
market capitalization of all public firms in an industry over the sum of market
capitalization across all public firms. The resulting asset-level measure is:
Redeployability Scorea,t =
123
j =1
Ia,j (use)×Valuej,t /
123
V aluej,t ,
(1)
j =1
11 If an industry’s expenditure on a given asset constitutes a negligible fraction (less than 0.1%) of the total
expenditure on the asset in the economy, we set Ia,j (use) to zero. Our empirical results are robust to varying this
cutoff from 0% (no cutoff) to 0.5%.
250
Downloaded from http://rfs.oxfordjournals.org/ at :: on December 23, 2016
where Redeployability Scorea,t is the redeployability of asset a in year t;
Ia,j (use) is a dummy variable equal to one if asset a is used by industry j in a
meaningful amount in the BEA table;11 and Valuej,t is either 1 (equal weight
for each industry), Nj,t (number of firms), or MCAPj,t (market capitalization)
of all Compustat firms in industry j in year t. The redeployability score can
vary as these measures of industries’ weights change over time. In addition,
if the description of an asset implies that it is highly specialized for a given
firm (industry), and thus would have virtually no salability outside the firm
(industry), we set the redeployability score to zero (the industry’s weight).
Examples of specialized assets are “custom computer programming” and “new
industrial plants construction.” The value of the asset-level redeployability
score we compute appears reasonable. For example, using MCAPj,t as the
weight in 1997, the score is 0.66 for “industrial trucks, trailers, and stackers,”
which are used in a wide range of industries, while it is 0.02 for “drilling oil
and gas wells,” which are predominantly used in the oil and gas industries.
To conceptually distinguish between these measures, it is useful to
decompose the notion of asset redeployability into asset specificity and liquidity
components. Broadly, our measures capture the usability or specialization of
capital across different purposes (e.g., industries). Therefore, in the transaction
costs framework of Williamson (1988), our measures relate closely to the
notion of asset specificity. In particular, our measure that only uses an industry
count to categorize capital goods aligns most closely with the degree of asset
specificity. However, the industry count does not fully account for market
thickness (mass of potential buyers), and therefore this measure is less related
to the liquidity dimension. Consequently, in our main measure we account
for the liquidity component by using market capitalization of an industry as a
proxy for market thickness to construct the industry’s weight. For example, an
The Asset Redeployability Channel: How Uncertainty Affects Corporate Investment
1.1.2 Constructing industry- and firm-level redeployability measures. In
the second step for each of these measures, we value-weight the asset-level
redeployability scores in Equation (1) across the 180 assets in the BEA table
to give us an industry-level redeployability index:
Redeployabilityj,t =
180
wj,a ×Redeployability Scorea,t
(2)
a=1
where Redeployabilityj,t is a measure of asset redeployability for industry j
in year t, wj,a represents industry j’s expenditure on asset a divided by its
total capital expenditure from the BEA table, and Redeployability Scorea,t is
the redeployability score of asset a in year t computed as in Equation (1). The
resulting redeployability index represents a relative redeployability ranking of
each industry’s asset composition.
A key advantage of our redeployability measure is that it accounts for the
salability of corporate assets across as well as within industries. While industry
peers are more likely to be higher-valuation potential buyers of capital than
outsiders, they are also more likely to suffer from the same operational and
financial difficulties as the firm liquidating assets (e.g., Shleifer and Vishny
1992; Ramey and Shapiro 2001). Therefore, cross-industry salability is crucial
in measuring asset redeployability.12
12 Consistent with this notion, we find that more than 60% of asset sales in secondary markets occur between firms
in different BEA industries by using asset-level transaction data from the SDC Platinum M&A database from
1980 to 2013. Maksimovic and Phillips (2001) also find that from 1978 to 1992, 37% to 52% of buyers of plants
in asset sales are from outside a given three- or four-digit SIC industry. Similarly, using data on closed plants in
the aerospace industry in the early 1990s, Ramey and Shapiro (2001) find that more than three-quarters of used
asse...
Purchase answer to see full
attachment