The investment-uncertainty relationship in the oil and gas industry

Contents lists available at ScienceDirect

Resources Policy

journal homepage: www.elsevier.com/locate/resourpol

The investment-uncertainty relationship in the oil and gas industry

Maryam Ahmadi

^a,∗

, Matteo Manera

^b

, Mehdi Sadeghzadeh

^c

aUniversity of Milan-Bicocca, Italy

bUniversity of Milan-Bicocca and Fondazione Eni Enrico Mattei (FEEM), Italy

cInstitute for Management and Planning Studies (IMPS), Iran

A R T I C L E I N F O Keywords:

Oil market Investment Uncertainty SVAR-GARCH JEL classification:

C3G11 Q41Q43

A B S T R A C T

Recent studies of the oil market demonstrate endogeneity of the oil price by modelling it as a function of consumption and precautionary demands and producers’ supply. However, studies analysing the effect of oil price uncertainty on investment do not disentangle uncertainties raised by underlying components which play a role in the oil market. Accordingly, this study uses a new approach to investigate the relationship between investment and uncertainty for a panel of U.S. firms operating in the oil and gas industry. We decompose oil price volatility to be driven by structural shocks that are recognized in the oil market literature, over and above other determinants, in order to study whether the investment uncertainty relationship depends on the drivers of uncertainty. Our findings suggest that oil market uncertainty lowers investment only when it is caused by global (consumption) oil demand shock. Stock market uncertainty is found to have a negative effect on investment with a year of delay. The results suggest no positive relationship between irreversible investment and uncertainty, but interestingly, a positive relation exists for reversible investment. This finding is in line with the option theory of investment and implies that the irreversibility effect of increased uncertainty dominates the traditional convexity effect.

1. Introduction

Investment decisions have three characteristics: first, the cost of investment is at least partially irreversible; second, there is uncertainty over future profits; and, third, investors can postpone their decisions to get additional information and reduce uncertainty (Dixit and Pindyck (1994)). The orthodox theory of investment, based on the assumption of reversible investment, is built upon net present value of future income of the investment. Postulated on reversible investment and convexity of marginal product of capital, this theory claims that an increase in un- certainty may raise investment (Abel (1983) and Hartman (1972)). The option theory of investment, on the other hand, builds on the as- sumption of irreversible investment. Taking into account the role of uncertainty in the timing of investment, this theory finds a negative effect of uncertainty on investment.

There are various sources for uncertainty, which are different across industries. Examples are: exchange rate uncertainty; output price un- certainty; and input price uncertainty. One important source of un- certainty in the oil and gas industry is about the price of oil. The effects of oil price changes on investment are analysed in many studies (e.g.

Glass and Cahn (1987) and Uri (1980)). The general finding is that oil price fluctuations are important for investment decisions at the ag- gregate level, as well as for single firms operating in the oil and gas sector. Some studies which relate oil price volatility to investment find that increases in oil price uncertainty raise the value of waiting. Hence, firms postpone their investment decisions when they face increased uncertainty (e.g. Bernanke (1983), Mohn and Misund (2009), Ratti and Yoon (2011) and Kellogg (2014)). However, the cited studies, as well as other contributions which investigate the relationship between oil price uncertainty and investment, consider the price of oil exogenous with respect to the macroeconomic variables. Based on this assumption, the construction of oil price variance estimates by averaging the squared residuals obtained from modelling the oil price conditional mean is an approach widely employed in empirical studies to measure uncertainty (e.g. Sadorsky (2008) and Henriques and Sadorsky (2011)). Conversely, there is a consensus in recent studies to consider the oil price as en- dogenous (see, e.g. Hamilton (2009), Kilian (2009), Dvir and Rogoff (2010), Alquist and Kilian (2010) and Kilian and Murphy (2014)). For instance, Hamilton (2009), Kilian (2009) and Baumeister and Kilian (2017) argue that the endogeneity of the price of oil with respect to the

https://doi.org/10.1016/j.resourpol.2019.101439

Received 13 May 2019; Received in revised form 21 June 2019; Accepted 26 June 2019

∗Corresponding author. University of Milan-Bicocca, Piazza dell’Ateneo Nuovo, 1, 20126, Milano, Italy.

E-mail addresses:maryam.ahmadi@unimib.it(M. Ahmadi),matteo.manera@unimib.it(M. Manera),sadmehdi@gmail.com(M. Sadeghzadeh).

0301-4207/ © 2019 Elsevier Ltd. All rights reserved.

T

macroeconomy is essential to study the effects of the oil price on key economic variables. Along this line of thought, Kilian and Murphy (2014) propose a structural decomposition of oil price shocks into their underlying different supply and demand components.

¹

Fig. 1 represents the historical decomposition of real oil price changes driven by relevant structural shocks.

²

The figure conveys the idea which supports our study: oil price dynamics is historically driven by a combination of aggregate demand and speculative shocks, which dominate the oil supply shocks. Accordingly, investors would react more significantly to shocks from the demand side of the oil market when deciding about irreversible investment.

The aim of this paper is to analyse the relationship between un- certainty and investment in the oil and gas industry, taking into account the underlying factors that drive oil price changes, along with other factors represented in the literature as determinants of firms' investment. Stock market volatility is also incorporated in the analysis in order to control for the effects of stock market un- certainty, which is an additional source of uncertainty that affects firms’ investment decisions. A Structural Vector Autoregressive with time-varying conditional variances (SVAR-GARCH) model is em- ployed to describe the global market for crude oil and the stock market. We employ the same structural model as Ahmadi et al.

(2016), although we depart from the assumption of time-invariant variances of the structural shocks.

It has been recognised by researchers that the uncertainty of fi- nancial time series, as measured by their variances, changes over time.

One of the most popular tools that has emerged to model non-constant variances is the Generalised Autoregressive Conditional Heteroscedasticity (GARCH) and its various extensions (for a review, see Bollerslev et al. (1992)). Our SVAR-GARCH model identifies five structural shocks: namely, oil supply shocks; global demand shocks;

speculative demand shocks; other oil market-specific shocks; and stock market shocks. Accordingly, we measure uncertainty by estimating the time varying volatilities of each structural shock. We then feed the (annualised) estimated volatility series into a Tobin's Q model of in- vestment, which relates the investment of a firm to its stock market valuation, measured as the ratio of the market value of a firm to the replacement value of its assets.

³

One advantage of our approach is that it accounts for two different sources of uncertainty that affect a firm's investment: uncertainty brought about by the oil market, and uncertainty related to the stock market. Although some studies have reported a connection between these two markets, as yet there is no consensus about the direction of their casual relation.

⁴

Moreover, according to Kilian and Park (2009), it is widely accepted by researchers that, since the 1970s, oil prices have been responding to the same economic factors that affect the stock market. Moreover the effect of stock market uncertainty on investment is also studied in the literature. Bulan (2005) studies a panel of U.S.

companies and finds that the volatility of stock returns has a strong negative effect on firms' investment and this effect is robust to the in- clusion of Tobin's Q or cash flow. Mohn and Misund (2009), using a panel of oil and gas companies and find that stock market uncertainty reduces investment in the long term. However, for some groupings of companies, in a panel of U.S. companies, Shaanan (2005) finds a po- sitive and significant impact of stock market uncertainty on investment.

This motivates the inclusion of stock market uncertainty to the analysis, by relating the U.S. stock returns to the global crude oil market, within our SVAR-GARCH model. This helps us, differently from previous stu- dies, to account for the pure effects of stock market uncertainty from which the effects of oil market shocks are excluded. Consequently, stock market uncertainty is measured by the volatility of aggregate stock market return shocks which are uncorrelated with oil market-related shocks.

Many empirical studies augment the Tobin's Q model for investment with additional explanatory variables, such as different measures of uncertainty (see, for example, Mohn and Misund (2009), Henriques and Sadorsky (2011) and Abel (2018)). In this paper, the Tobin's Q model is augmented by the firm's cash flow

⁵

and oil price uncertainty driven by structural shocks to the price of oil, as well as by aggregate stock market uncertainty.

The rest of the paper proceeds as follows. Section 2 provides a brief review of the relevant literature. In Section 3 the data are described.

The Structural VAR model and the augmented Tobin's Q model of in- vestment are set out in Section 4. Section 5 presents and discusses the estimation results. Section 6 concludes.

2. Literature review

The relationship between investment and uncertainty has been analysed in numerous studies. A number of theoretical studies apply the standard neoclassical investment model that predicts a positive relationship between uncertainty and investment. According to this model, a firm invests when the present value of the project's expected cash flow is greater than its cost. Oi (1961), Hartman (1972) and Abel (1983), under the assumptions of risk neutrality, perfect com- petition and constant returns to scale, find a positive effect of output price uncertainty on investment. This strand of the literature does not, however, consider the important characteristic of the invest- ment decisions, that is, investment is at least partly irreversible (Chirinko (1993)).

The other strand of the literature highlights the importance of ir- reversibility. The idea is that the combination of irreversibility and uncertainty gives value to postponing investment decisions similar to a financial call option. Therefore, there is an opportunity cost for in- vesting today rather than waiting for more information about pro- spective returns.

⁶

According to the option theory, firms invest if the net present value of investment is greater than the option value of waiting (Favero et al. (1992)). This suggests that there is an irreversibility ef- fect, under which the value of the waiting option rises with uncertainty, and it dominates any positive impact of uncertainty on the marginal profitability of capital (Carruth et al. (1998), and Bond and Cummins (2004)).

Some studies show that investment theory based on irreversibility may also imply a positive effect of uncertainty on investment in certain situations. Sarkar (2000) shows that, in the case of a low-risk low- growth firm, such as one in a regulated industry like the utilities sector, a moderate increase in uncertainty has a positive effect on investment.

Akdogu and MacKay (2008) provide evidence that, when competition is not perfect, as in oligopolistic industries, an increase in uncertainty may spur investment. According to Bolton et al. (2014), this is also the case for a financially constrained firm.

Employing various proxies for uncertainty, several authors report a negative and significant relationship between investment and

1This structural model of the global oil market has been widely used in the literature, for example,Baumeister and Kilian (2015),Baumeister and Kilian (2016), Kilian and Lee (2014), Kilian (2017) and Herrera and Rangaraju (2018).

2For a full structural model specification, refer to the methodological section of this paper.

3This ratio is called Tobin's Q and captures the firm's opportunity cost of capital investment.

4See e.gCharles and Gautam (1996),Sadorsky (1999),Park and Ratti (2008), Odusami (2009)andSinghal and Ghosh (2016).

5Abel (2018), in a study of the relationship between investment, Q and cash flow, shows that cash flow has explanatory power for investment and this effect is even larger for faster growing firms that are likely to be financially con- strained.

6See, e.g.,Cukierman (1980),Bernanke (1983),McDonald and Sigel (1986), Pindyck (1991),Dixit and Pindyck (1994), andChirinko (1993).

M. Ahmadi, et al. Resources Policy 63 (2019) 101439

uncertainty for U.S. firms. A few examples of these proxies are: vola- tility of exogenous variables such as output prices and wages (Huizinga (1993) and Ghosal and Loungani (1996)); exchange rate (Campa (1993)) and stock market return (Bulan (2005)); as well as endogenous variables such as: sale growth rates (Ogawa and Suzuki (2010)); future profits (Bond and Cummins (2004)); and expected volatility of the fu- ture price (Kellogg (2014)). However, some authors report that the significance of the negative relationship is not robust when Tobin's Q or other important factors, such as cash flow, are included. This finding may be due to a negative correlation between Tobin's Q and a given measure of uncertainty (Leahy and Whited (1996)). Hayashi (1982) shows that, for the case of perfectly competitive markets and constant returns to scale technology, average Q, the ratio of the maximised value of the firm to the replacement cost of its capital stock, would be a sufficient statistic for investment rates. Tobin's Q, however, measures the maximised value of the firm by its stock market valuation, and assumes that the stock market valuation captures all information about expected future profitability. However, as Bond et al. (2004) show, if the Hayashi conditions are not satisfied, or if stock market valuations are influenced by factors other than the present discounted value of expected profits, Tobin's Q would not capture all relevant information about the expected future profitability of current investment. This motivates some empirical studies based on the Q model augmented with other explanatory variables of interest.

⁷

Several studies assess the importance of energy price fluctuations in investment decisions. Uri (1980) develops a simple model and finds that the price of energy is important in explaining investment decisions at the aggregate level and, more importantly, for energy intensive in- dustries. Bernanke (1983) shows that, when oil price uncertainty in- creases, firms postpone their irreversible investment when they have to choose between energy-efficient and energy-inefficient capital. Ratti and Yoon (2011) estimate an error correction model of capital stock adjustment with data on U.S. manufacturing firms. Their results suggest that stability in energy prices would be conducive to greater stability in firm-level investment. Rafiq et al. (2009), in a study of Thai economy and by using a vector auto-regression model, show that oil price vola- tility has negative effect on investment. Similar result found by Wang et al. (2017) in a study of the effects of oil price uncertainty on China's economy and Phan et al. (2019) in a global level. Using a large set of data, which spans a wide range of countries, Phan et al. (2019) also find that the negative effect is stronger in the crude oil producers group than for crude consumers.

The relationship between uncertainty and investment in the oil and gas industry, as a special case, has attracted academic interest at both micro and aggregate levels. However, the number of studies working on the effects of uncertainty on investment in oil and gas fields is rather small, and the empirical findings are mixed. At the aggregate level, Favero et al. (1992) develop a theoretical model and derive the de- terminants of the decision to develop an oil field. They evaluate the importance of the variables suggested by theory to explain the length of development lags of U.K. oil and gas fields. Their results imply that the volatility of prices positively(negatively) affects the duration of

Fig. 1. Historical decomposition of real oil price change.

7See, e.g.,Mohn and Misund (2009),Henriques and Sadorsky (2011)and Abel (2018).

investment appraisal when prices are low(high). Hurn and Wright (1994), using data from oil fields of the North Sea, analyse the effects of expected oil price, variance of oil price and the level of reserves, on the lag between discovery of a field and the decision to develop the field.

They find that the expected oil price and the level of reserves are im- portant in influencing the appraisal duration, while the variance is not.

At the micro-level, Mohn and Misund (2009) estimate the effect of oil price volatility on investment in the international oil and gas industry.

Their results show that Tobin's Q is a poor investment indicator, but uncertainty measures contribute significantly to the explanation of in- vestment. Elder and Serletis (2010) apply a bi-variate GARCH model to study how oil price uncertainty affects investment for the U.S.

economy, and find that increases in oil price volatility reduce aggregate investment. Similar results are found for Canada. Lee et al. (2011) apply a standard investment model to analyse the joint effects of an oil price shock and firm-specific uncertainty on investment, using firm-level panel data. They conclude that an oil price shock has a greater effect on delaying a firm's investment, the greater the uncertainty faced by that firm. Abel (2018) uses a tractable stochastic model to analyse invest- ment, Q and cash flow and shows that Q is not a sufficient statistic for investment and other variables, in particular, cash flow, have additional explanatory power for investment behavior of firms.

This paper analyses the relation between investment and different sources of uncertainty that are identified by their underlying causes for the firms operating in the U.S. oil and gas industry. We consider two sources for uncertainty, oil and stock market uncertainties, and employ a Structural Vector Autoregressive with a time-varying conditional variances (SVAR-GARCH) model to describe the oil and stock markets.

By using this model, we depart from the literature in three ways. First, we take into account the fact that the uncertainty of financial time series, as measured by their variances, is time varying. Second, we take into account, for the first time, the endogeneity of the oil price in the investment-uncertainty relationship, and estimate the oil price un- certainty driven by each of the oil market structural shocks. Third, we identify the stock market uncertainty from which the effects of oil market shocks are excluded.

Our focus is on the oil and gas industry, since uncertainty about the price of oil is one of the most important sources of uncertainty in this industry, which connects the oil and stock markets. We believe that, since the oil price is not exogenous with respect to oil market partici- pants and macroeconomic fundamentals,

⁸

the oil and gas industry re- presents a natural laboratory for the application of our new approach to the analysis of the investment-uncertainty relationship.

3. Data

Our data are in the form of an unbalanced panel of U.S. oil and gas companies ( =

i

1,

…

,

N

) for the period 1974 to 2018 ( =

t

1,

…

,

T

) drawn from COMPUSTAT. It includes the following annual variables: market value of equity, long-term debt, total assets, capital expenditure, short- term liabilities, short-term assets, income before extraordinary items, depreciation, and amortization. Firm investment is proxied by capital expenditure on property, plant, and equipment. Following the litera- ture, the Tobin's Q is measured as the sum of market value and long- term debt, divided by lagged total assets

⁹

:

= + +

Q market value of equity preferred stock debt total assets

( )

it it

1.

it 1

Our firm investment measure is constructed as the ratio of capital expenditure to capital stock, and capital stock is measured by total assets (Mohn and Misund (2009)). Cash flow is measured as the sum of

income before extraordinary items and depreciation over capital stock, written as:

= +

Cashflow income before extraordinary items depreciation total assets

( )

it it

1.

it 1

We clean our data using a number of sample selection rules. First, observations with Tobin's Q values out of the 99% confidence interval are dropped. Second, any observation for which there is one or more variable/s missing is screened. At the end, we keep firms with at least five observations in our sample. After this procedure, our sample is left with N = 985 firms.

Global oil market variables include global crude oil production, an updated measure of cyclical fluctuations in global real economic ac- tivity

¹⁰

, real price, and above-ground inventory of oil, which are all available in monthly frequency. Data on global oil production are from the Monthly Energy Review of the Energy Information Administration (EIA). The real price of oil, proxied by the U.S. refiners’ acquisition cost for imported crude oil, is also available from the EIA. The price of oil is deflated by the U.S. consumer price index. Since data on oil inventories for all countries are not available, we follow Hamilton (2009), Kilian and Murphy (2014), and Kilian (2017) to construct global oil inventory data. The series is constructed by scaling total U.S. oil inventories by the

Fig. 2. Monthly volatility of structural shocks.

Table 1

Summary statistics of data.

mean sd min max

0.199 0.372 0.000 36.968

Q 0.996 2.766 −0.964 24.914

Cashflow 0.030 0.656 −3.226 40.960

Observations 12299

1.064 0.367 0.614 2.265

1.855 0.735 0.846 3.779

1.944 0.795 0.819 3.528

3.188 0.392 2.485 4.713

2.528 0.636 1.275 4.019

Observations 511

8See, e.g.,Hamilton (2009),Kilian (2009), andKilian (2008).

9In terms of COMPUSTAT item labels, the relationship is. =Q (PRCC C_ CSHO+PSTKL+(LCT ACT)+DLTT AT)/ ( 1) 1.

10This measure of global real economic activity, introduced byKilian (2009) and updated byKilian (2019), captures the global business cycle, and is used to measure consumption demand for oil and all industrial commodities.

M. Ahmadi, et al. Resources Policy 63 (2019) 101439

ratio of the OECD petroleum stocks to the U.S. petroleum stocks, which are available from the EIA.

The U.S. aggregate stock market returns are obtained from the Center for Research in Security Prices (CRSP), and are related to a value-weighted market portfolio including NYSE, AMEX and NASDAQ stocks. The real stock returns are constructed by subtracting the Consumer Price Index (CPI) inflation from the log of aggregate stock returns.

The oil market variables, in addition to the stock market returns, form the vector of variables for the SVAR-GARCH analysis. The vola- tility series extracted from estimating the SVAR-GARCH model (see Fig. 2) are annualised by averaging monthly volatilities prior to each year.

¹¹

Table 1 presents the descriptive statistics for the whole data set.

The graph in Fig. 3 shows the annual mean of the investment capital ratio of all firms versus demand and supply shocks volatilities. The patterns of variability associated with these series suggest a more pro- nounced negative correlation of the investment measure with the vo- latility of the demand shock, especially after 2000. The correlation of the mean investment capital ratio with the demand shock volatility is

−0.46, while the value of the correlation coefficient with the supply shock volatility is 0.27.

4. Methodology

Our empirical methodology follows a two-step procedure. First, we estimate the volatility series of the oil market and stock market shocks within a time-varying volatility Structural VAR (SVAR-GARCH) fra- mework. Since the firm-level data are annual, volatility series are an- nualised, such that each annual volatility is the average of the monthly volatilities. Second, an augmented Tobin's Q model of investment is estimated using firm-level data and annualised volatility series.

4.1. Measures of uncertainty

A SVAR-GARCH model is employed to estimate the oil price un- certainty driven by each of the oil market structural shocks as well as by the stock market shocks. SVAR-GARCH models have been widely ap- plied by researchers for the identification of structural shocks or for testing over-identification in a VAR context.

¹²

The assumption of con- ditional heteroskedasticity allows for statistical identification without the need for conventional exclusion restrictions. However, this purely statistical identification procedure may not provide the researcher with economically interpretable results, which makes it difficult to interpret the shocks as being structural VAR shocks. Our methodology addresses this issue by applying a SVAR-GARCH framework which uses a set of identifying restrictions generally adopted in the literature, and departs from the conventional framework by assuming time-varying variances of structural shocks.

Our baseline SVAR specification is taken from Kilian and Murphy (2014). We depart from this specification in two ways. First, following Ahmadi et al. (2016), we include stock market returns in our structural model to account for the effects of stock market uncertainty on the firm- level investment. The advantage of this approach is that we identify the

Fig. 3. Firm investment capital ratio versus annualized supply shock and demand shock volatilities.

11To match the frequency of the volatility series with firm-level data, we aggregate monthly series to obtain annual series. We are aware of the in- formation we lose when aggregating the monthly volatilities. To cope with this problem, we have applied quarterly firm-level data to reduce the information loss caused by data aggregation. Although quarterly firm-level data are highly seasonal, our results do not change overall. Another possibility for avoiding data aggregation is to apply the Mixed Data Sampling (MiDaS) approach, which allows for estimation where regressors have different frequencies.

Unfortunately, we are not aware of a consistent MiDaS estimator that in- corporates the Two-step System GMM methodology.

12See, e.g.,Lanne and Saikkonen (2007),Lutkepohl and Milunovich (2016), andLutkepohl and Netsunajev (2017).

stock market shocks from which the effects of oil market shocks are excluded. Second, we estimate the model under the assumption that the structural shocks have time-varying volatility. Our specification is.

¹³

= + +

= = =

= + +

A y = A y

E E H E t s

h h

( ) 0, ( ) , ( ) 0,

t i i t i t

t t t t t s

i t i i i t i i t

0 241

, 2, 1

, 1

(1)

where y

_t

is the vector of endogenous variables including the percentage change in global oil production, the global real economic activity, first difference of oil inventory, the first difference of real oil price and real stock market returns;

t

is the vector of structural shocks, namely, oil supply shocks, global demand shocks, oil speculative demand shocks, other oil market shocks and stock market shocks. The i th structural shock has a GARCH(1,1) conditional variance, h

i t,

, which is the i th diagonal element of the conditional covariance matrix H

_t

. The struc- tural shocks are estimated from the analogous reduced form VAR model:

= + +

=

y_t B y e

i i t i t

1 24

(2)

=

= = =

B A A i

e A A D z S z

, , ,

i i

t t t t t t

01 01

01 1 2

The VAR residuals have the following time-varying covariance matrix:

= = = = =

E e e

(

_{t t}

)

E A

(

01_{t t}A

)

A E

(

_{t t}

)

A A D At S St t Vt 01

01

01 01

01

The structural shocks are identified from the reduced form VAR model, =

et A01t

, by imposing short-run restriction on A

0

. Following Kilian (2009) and Ahmadi et al. (2016), the matrix of short-run re- strictions is:

= e

e e e

e

a a a

a a a

a a a a

a a a a a

0 0 0 0

0 0 0

0 0

0

tglobal oil production global real activityt

real price of oilt tglobal oil inventory

real stock returnt

oil supply shockt global demand shockt

speculative shockt other oil shockst stock market shockt 1

2 4 3

5

11

21 22

31 32 33

41 42 43 44

51 52 53 54 55

1 2

3 4

5

(3)

The identifying restrictions are based on four assumptions. First, within a month, changes in global oil production do not respond to oil demand shocks. This assumption is made because adjustment in oil production plans is very costly. Second, when the increase in the oil price is caused by speculative demand or other oil market shocks, it affects global real economic activity with at least one month of delay.

Third, within a month, the real price of oil responds to oil supply, consumption demand, and speculative demand shocks. Finally, oil market variables are predetermined with respect to stock market re- turns, while stock returns are affected by different oil price shocks.

4.2. The investment model

The effects of uncertainty on firm-level investment are estimated using an augmented Tobin's Q model (Tobin (1969)). This model relates investment to the firm's stock market valuation, which reflects the present discounted value of expected future profits. The Q-model of investment is represented by the following relationship

¹⁴

:

= + +

I

K a

bQ

1 ,

it it it

(4)

where Q

it

is the ratio of the market value of the ith firm to the re- placement value of its assets; I

_it

is the firm's gross investment; K

_it

_{is the} firm's net capital stock;

it

is a random error term; and a and b are the structural parameters of the adjustment cost function. To take into ac- count the factors reported in the literature which affect investment, we augment Tobin's Q model with additional explanatory variables, in- cluding stock market uncertainty (

stock

) and oil market uncertainty, driven by oil supply shocks (

supply

); global demand shocks (

demand

);

speculative demand shocks (

spec

); and other oil market shocks (

other

).

These variables are included in the model to capture and compare local (i.e. industry) and global (i.e. market) risks brought about by different sources of uncertainty. Following Henriques and Sadorsky (2011) and Abel (2018), firms' cash flow is also included among the explanatory factors. In order to avoid serial correlation in the error terms,

t

is as- sumed to follow an AR(1) process (see Mohn and Misund (2009)):

= +

,

it it 1 it

(5)

with

it

representing a white noise error term. Substituting the addi- tional variables in equation (4) and incorporating (5), we obtain the following dynamic investment model:

= + + + + +

+ + +

I

K a I

K bQ

bQ CF CF

B X B X

(1 ) 1

,

it it it it it it

t t it

1 1 1 1 1

1 1 1

(6)

where X

t

is a vector of variables including

_supply

,

_demand

,

_spec

,

_other

and

stock

. However, it is more convenient to estimate the unrestricted

version of (8), represented as:

= + + + + + +

+ +

I

K b b I

K b Q b Q b CF b CF B X

B X

.

it it it it it it t

t it

0 1

1 2 3 1 3 4 1 1

2 1

(7)

The empirical specification (7) relates the firms' investment capital ratio to its one-period lag, Tobin's Q, cash flow, oil price volatility series, and stock market volatility.

4.2.1. Panel unit root tests

To determine the order of integration of the variables involved in the investment model, two panel unit root tests were carried out. The reason to choose panel unit root tests is that these statistics have a higher power to reject the null hypothesis of a unit root than traditional tests such as the Augmented Dickey-Fuller (ADF) test, especially in shorter time spans. The Levine et al. (2002)'s panel unit root test pos- tulates that the unit roots among the cross-sections are homogeneous.

This assumption allows for different lag orders across cross sections.

Hence, the null hypothesis is that each individual time series has a unit root, and the alternative is that at least one time series is stationary.

Conversely, the Im et al. (2003) test allows for heterogeneity across cross-sectional unit roots and specifies an ADF regression for each cross section. The null hypothesis is that there is a unit root, while the al- ternative is that the fraction of individual units following stationary processes is non-zero.

The results of the tests are presented in Table 2 and suggest strong

13Applying 24 months of lags is consistent withHamilton and Herrera (2004) andKilian and Park (2009)who argue that allowing for a high lag order is crucial to capture the transmission of the structural shocks in the oil market.

They provide evidence that moving cycles in the oil market are very slow and a low number of lags would fail to capture the whole dynamics of the cycle. The alternative way of setting the lag order is by testing the goodness of fit using information criteria. However, some researchers (see, e.g.,Leeb and Potscher (2006), andIvanov and Kilian (2005)) argue against the validity of such methods especially when there is a prior on the number of lags. However, ac- cording toHamilton and Herrera (2004), there are strong claims about the value of lag order in the oil market based on prior studies, and the AIC estimates would make a lower bound (Ahmadi et al. (2016)).

14Derivation of the linear Q model of investment from a theoretical model is provided byBond and Van Reenen (2007)andMohn and Misund (2009).

M. Ahmadi, et al. Resources Policy 63 (2019) 101439

evidence against the presence of unit root in our panel data set,

¹⁵

meaning that all variables are stationary in levels. Accordingly, we do not need to apply the cointegration test, and our data are appropriate to adopt the System GMM (SGMM) methodology.

4.2.2. Model estimation

The augmented Tobin's Q model of investment in equation (7) re- lates the ratio of investment to a firm's capital to its own one-period lag, Tobin's Q, and cash flow, as well as to our measures of oil price and stock market uncertainty. The model uses a panel data of N = 985 U.S.

firms in the oil and gas industry over a period of T = 44 years. The model is estimated using two methods: panel estimation with firm fixed effect and the two-step System Generalized Method of Moments (SGMM) proposed by Blundell and Bond (1998). The motivation to apply SGMM is four-fold. First, it allows us to control for possible en- dogeneity between the model's variables and the predetermined re- gressors that are independent of current disturbances, but can be in- fluenced by past disturbances.

¹⁶

Second, when some series in the model have near unit root properties, according to Bond and Cummins (2004) and Blundell and Bond (1998), the lagged levels of the regressors, used in the difference GMM model, may be poor instruments for the first- differenced regressors. The level of persistence of the variables in our model is shown in Table 3.

The SGMM method includes level equations in the estimated system of equations. This method applies lags of both first-differences and le- vels of the dependent and predetermined variables as instruments to improve efficiency, comparable to Arellano and Bond (1991)'s differ- ence GMM estimator (Mohn and Misund (2009)). Third, the two-step estimator is asymptotically more efficient than the one-step estimator, and the standard covariance matrix is robust to panel-specific auto- correlation and heteroskedasticity.

¹⁷

Fourth, since our model in- corporates a generated vector of variances among the regressors, in- ference based on standard t tests is generally biased (see Pagan (1984), and Pagan (1986)). On the contrary, the SGMM approach avoids the inferential issues raised by the generated risk regressors (Pagan and

Ullah (1988)).

It is very important to test for autocorrelation when the GMM method is applied on panels, and lags are treated as instruments.

Arellano and Bond (1991) suggest an appropriate test for first- and second-order autocorrelation in first-differenced errors. If the residuals are independent and identically distributed, the first-order test statistic is significant, meaning that the first-differenced errors are first-order serially correlated. The higher-order serial correlations should be in- significant. To test for over-identifying restrictions, Arellano and Bond (1991)'s version of the Sargan test could incorrectly reject the null hypothesis in the presence of heteroskedasticity. Hence, following Rodman (2005), we use the Hansen statistic to test for over identifying- restrictions.

5. Estimation results

We report two sets of estimation results. Table 4 presents the results of a set of linear models, where the average firm investment capital ratio is the dependent variable. Similarly, cash flows and Tobin's Q are averaged annually and treated as explanatory variables. Averaging variables across firms cancels out the noise among firms in making investments decisions, and could serve as an aggregate measure in analysis.

¹⁸

The first three columns report estimation results based on conven- tional Tobin's Q specifications. The last columns report the results for the specification augmented with volatility measures. The empirical findings suggest a high explanatory power for the lagged Tobin's Q.

Most important, out of five volatility measures, only the volatility of global demand shock negatively affects the aggregate investment ca- pital ratio. This result is robust to the inclusion of lag dependent vari- able and is in line with the oil market literature, where demand shock is reported as the main driver of the oil price. In other words, investors mainly monitor demand shifts when deciding about new investment projects.

The results from estimating investment model (10) are reported in Table 5. The first column refers to panel estimation, where robust standard errors are reported, and firm fixed effects are controlled for.

The last two columns present the results obtained with the SGMM method. For both SGMM specifications, robust standard errors are re- ported and all volatility measures are treated as instrument variables. In the second column, the lagged dependent variable, cash flow and the Tobin's Q are treated as endogenous, and in the last column the lagged Tobin's Q and cash flow are also assumed to be endogenous.

Evidence on the sign and the significance of the coefficient of the lagged investment-capital ratio, reported in the first row of Table 5, confirms a positive and significant impact of lag dependent variable.

The contemporaneous cash flow significantly raises investment only in the third specification of the model, while a positive significant re- sponse of investment to the lagged cash flow is robust with larger coefficients in the all specifications. This result supports the finding in Chen (2016), suggesting that when firms face investment frictions such as high adjustment cost and investment planning lag, their capital ex- penditures are more sensitive to the lagged cash holdings than to the contemporaneous cash flows.

¹⁹

The results show that Tobin's Q and its lag positively affect investment.

The estimated coefficients of different uncertainty measures, re- ported in Table 5, show that investment-uncertainty relationship greatly depends on the cause of uncertainty. When uncertainty is driven by shocks to oil supply, speculative demand, or oil market-specific demand, there is no significant effect of uncertainty on investment. On the contrary, when uncertainty is brought about by global

Table 2

Panel unit root tests.

without trend with trend

IPS LL IPS LL

−1.989** −11.012*** −5.518* −19.340***

−9.217*** −21.689*** 5.862 −8.388***

7.607 7.362 −16.083*** −21.193***

−2.621*** −3.385*** 0.827 4.199

−10.091*** −14.266* −8.649*** −16.591***

Q −7.763*** −10.561*** −5.640*** −11.527***

−6.3028*** −16.5863*** −8.4223*** −5.2720***

Notes: ***, ** and * indicate statistical significance at the 1%, 5% and 10%

levels, respectively.

15Since both theIm et al. (2003)andLevine et al. (2002)tests require a balanced panel dataset, we restrict our analysis to the sample period starting from 1998.

16The lagged dependent variable and the Tobin's Q ratio are both considered to be endogenous in the model. There is a consensus among researchers on the endogeneity bias for the lagged dependent variable in panel data models (see, e.g.,Arellano (2003)). Moreover, our Tobin's Q measure may not be strictly exogenous since it incorporates market valuation in the numerator (Mohn and Misund (2009)).

17The Arellano-Bond GMM estimators can be implemented as one- and two- step estimators. The two-step estimator is asymptotically more efficient, but the standard errors are downward-biased (Blundell and Bond (1998)). A finite- sample correction of the covariance matrix of two-step GMM estimators pro- posed byWindmeijer (2005)is adopted in our paper, which results in larger standard errors that are much more reliable in finite samples.

18Aggregation is carried out by averaging firm level values across each year.

19They argue that by making investments with realized cash holdings, firms avoid the costs that are incurred when investing with uncertain cash flows.

(consumption) demand shocks, investment significantly decreases on impact. This negative effect, however tends to be positive after one year from the shock. This result is consistent with the option theory that increased uncertainty raises the option value of waiting to get more information or to wait until the uncertainty is resolved (Pindyck (1991) and Dixit and Pindyck (1994)). After one year, given that the source of uncertainty is shocks to the consumption demand for oil, further in- creases in uncertainty about existing oil reserves becoming depleted, dominate the value of waiting option. Consequently, new exploration and development will need to replace oil reserves. Stock market un- certainty is found to negatively affect investment decisions in the oil and gas industry with a one-year lag, although the concurrent effect is positive. This finding could be due to diversified firm investment in non-oil sectors, where investment is at least partially reversible (see Lamont (1997) for further discussion). Table 5 shows how the results are robust across different estimation techniques.

However, the results indicate that there is no significant on-impact positive relationship between uncertainty and investment. Consistent with the option theory, this result shows that the irreversibility effect of increased uncertainty dominates the traditional convexity effect. The results resemble some previous results from other studies that Tobin's Q is not a sufficient statistic to explain firms' investment behaviour and, indeed, there is an important role for uncertainty in the investment decision-making process of firms (Rafiq et al. (2009), Mohn and Misund

(2009), Henriques and Sadorsky (2011), Wang et al. (2017), Abel (2018) and Phan et al. (2019)). Moreover, inconclusive sign and size of the response of investment to uncertainty

²⁰

, found in the literature, indicates the importance of decomposing volatility into the components driven by supply and demand shocks.

The results from Arellano and Bond (1991)'s post-estimation speci- fication tests, the last rows of Table 5, show that there is no significant evidence of second-order serial correlation in the first-differenced er- rors. The result of the Hansen test does not reject the validity of the over-identifying restrictions.

6. Conclusions

In this paper we have studied the relationship between investment and uncertainty in a panel of U.S. firms operating in the oil and gas industry. To achieve this, we estimate a Tobin's Q model of investment, augmented with measures of industry-level and aggregate stock market uncertainty from a SVAR-GARCH model of the global oil market. We contribute to the literature in three ways. First, we take into account that the uncertainty of financial time series, as measured by their var- iances, is time varying. Second, we take into account, for the first time,

Table 3

Results from regressing each variable on its one-period lag.

I

K Q CashF supply demand oil specific stock

I( 1)

K . .515***(.030)

Q ( 1). .399***(.022)

CashF ( 1) 0.15*** (0.011)

supply( 1) . .864***(.004)

demand( 1) .560***(.030)

oil specific( 1) .510***(.0354)

stock( 1). .605***(.030)

Notes: ***, ** and * indicate statistical significance at the 1%, 5% and 10% levels, respectively. Standard error are reported in parentheses.

Table 4

Results of the regression of the aggregate level equation This table presents the results of a set of linear models, where the average firm investment capital ratio is the dependent variable. Similarly, cash flows and Tobin's Q are averaged annually and treated as explanatory variables. The first three columns report estimation results based on conventional Tobin's Q specifications. The last columns report the results for the specification augmented with volatility measures.

(1) (2) (3) (4) (5) (6)

0.254 0.0936 0.272 0.130

(0.205) (0.457) (0.117) (0.249)

Cshflw −0.0156^∗∗∗ −0.0148^∗∗∗ −0.0170^∗∗∗ −0.0184^∗∗∗ −0.0178^∗∗∗ −0.0189^∗∗∗

(0.000) (0.000) (0.000) (0.000) (0.000) (0.000)

Q 0.0132 0.0131 −0.0159 0.00828 0.00731 −0.0155

(0.088) (0.083) (0.222) (0.323) (0.386) (0.222)

Q(-1) 0.0489^∗∗∗ 0.0439^∗∗∗

(0.000) (0.000)

0.0103 0.00925 0.0153

(0.628) (0.653) (0.421)

−0.0613^∗∗ −0.0647^∗ −0.0317

(0.008) (0.011) (0.191)

0.0287 0.0264 0.00910

(0.071) (0.116) (0.548)

0.00183 0.00477 −0.00224

(0.870) (0.666) (0.830)

−0.0000102 −0.00297 −0.00305

(0.999) (0.667) (0.540)

Constant 0.188^∗∗∗ 0.135^∗∗∗ 0.147^∗∗∗ 0.183^∗∗ 0.145 0.166^∗∗

(0.000) (0.001) (0.000) (0.003) (0.051) (0.001)

Observations 43 42 42 43 42 42

0.155 0.214 0.566 0.323 0.384 0.608

Notes: ***, ** and * indicate statistical significance at the 1%, 5% and 10% levels, respectively. p-values are reported in parentheses.

20For more detailed discussion, refer to section2.

M. Ahmadi, et al. Resources Policy 63 (2019) 101439

the endogeneity of the oil price in the investment-uncertainty re- lationship and proxy oil market uncertainty by considering the under- lying causes behind oil price fluctuations. Third, we identify the stock market uncertainty from which the effects of oil market shocks are excluded.

The results show that the impact of oil market uncertainty on in- vestment largely depends on the causes behind uncertainty. When un- certainty results from shocks to oil supply, speculative demand, or other oil market-specific demands, there is no significant impact on firms' investment decisions. If oil price uncertainty is driven by global (con- sumption) demand shock, current investment responds negatively to oil price uncertainty. Therefore, the consumption demand component of the oil price is the main oil market variable which affects investment in the oil and gas industry. Stock market uncertainty has a positive and significant impact on firms' investment decisions. However, the effect becomes negative for the lagged stock market uncertainty. We interpret this finding by assuming that firms increase their diversified investment in the non-oil sectors, where investment is not sunk by increased stock market uncertainty. Our main results are robust across different esti- mation techniques. We confirm the view that Tobin's Q is not a

sufficient factor to explain firms' investment behaviour. Conversely, there is an important role for uncertainty in the investment decision- making process of a firm, which is consistent with the option theory of investment.

The policy makers in governments or industries can apply the findings of this paper when there is uncertainty about the price of oil and the price is not stable. According to the findings of this paper, in order to take the right and relevant policy, they should take into ac- count the reason behind this situation. In other words, government should mainly monitor demand shifts to pursue its policy to boost in- vestment and growth in the economy. This is very important also for investors when they decide to invest in the oil sector or to diversify their investment in non-oil sector.

Acknowledgments

The first and third authors acknowledge financial support from the Department of Economics, Management and Statistics of the University of Milan-Bicocca, Italy. Previous versions of the paper have been pre- sented in seminars organized by: the Department of Economics, Management and Statistics, University of Milano-Bicocca, Italy; EMEE 2018, 11th International Workshop on “Empirical Methods in Energy Economics”, Fondazione Eni Enrico Mattei, Milano, Italy; the Department of Environmental Science and Policy, University of Milan, Italy; and SIS 2018, 49th Scientific meeting of the Italian Statistical Society, University of Palermo, Italy. The authors are indebted to seminar participants for insightful comments and suggestions.

References

Abel, A.B., 1983. Optimal investment under uncertainty. Am. Econ. Rev. 73 (1), 228–233.

Abel, B., 2018. The effects of q and cash flow on investment in the presence of mea- surement error. J. Financ. Econ. 128 (2), 363–377.

Ahmadi, M., Manera, M., Sadeghzadeh, M., 2016. Global oil market and the U.S. stock returns. Energy 114, 1277–1287.

Akdogu, E., MacKay, P., 2008. Investment and competition. J. Financ. Quant. Anal. 43 (2), 299–330.

Alquist, R., Kilian, L., 2010. What do we learn from the price of crude oil futures? J. Appl.

Econom. 25, 539–573.

Arellano, M., 2003. Panel Data Econometrics. Oxford University Press.

Arellano, M., Bond, S.R., 1991. Some tests of specification for panel data: Monte Carlo evidence and an application to employment equations. Rev. Econ. Stud. 58, 277–297.

Baumeister, C., Kilian, L., 2015. Forecasting the real price of oil in a changing world: a forecast combination approach. J. Bus. Econ. Stat. 33 (3), 338–351.

Baumeister, C., Kilian, L., 2016. Forty years of oil price fluctuations: why the price of oil may still surprise us. J. Econ. Perspect. 30 (1), 139–160.

Baumeister, C., Kilian, L., 2017. Lower Oil Prices and the U.S. Economy: Is This Time Different? Brookings Papers on Economic Activity. pp. 287–336.

Bernanke, B.S., 1983. Irreversibility, uncertainty, and cyclical investment. Q. J. Econ. 98 (1), 85–106.

Blundell, R.W., Bond, S.R., 1998. Initial conditions and moment restrictions in dynamic panel data models. J. Econom. 87, 115–143.

Bollerslev, T., Chou, R.Y., Kroner, K.F., 1992. ARCH modeling in finance, A review of the theory and empirical evidence. J. Econom. 52 (5–59).

Bolton, P., Wang, N., Yang, J., 2014. Investment under Uncertainty and the Value of Real and Financial Flexibility. National Bureau of Economic Research NBER Working Paper 20610.

Bond, S., Klemm, A., Newton-Smith, R., Syed, M., Vlieghe, G., 2004. The Roles of Expected Profitability, Tobin's Q and Cash Flow in Econometric Models of Company Investment. Bank of England Working paper no. 222.

Bond, S.R., Van Reenen, J., 2007. Microeconometric models of investment and employ- ment. Handb. Econom. 6 (A), 4417–4498.

Bond, S.R., Cummins, J.G., 2004. Uncertainty and Company Investment, an Empirical Model Using Data on Analysts' Profits Forecasts. Mimeo, Institute for Fiscal Studies.

Bulan, L.T., 2005. Real options, irreversible investment and firm uncertainty, new evi- dence from U.S. firms. Rev. Financ. Econ. 14, 255–279.

Campa, J., 1993. Entry by foreign firms in the United States under exchange rate un- certainty. Rev. Econ. Stat. 75 (4), 614–622.

Carruth, A., Dickerson, A., Henley, A., 1998. What Do We Know about Investment under Uncertainty? Discussion Paper. University of Kent No. 9804.

Charles, M.J., Gautam, K., 1996. Oil and the stock markets. J. Financ. 51 (2), 463–491.

Chen, L., 2016. Corporate investments, cash flows and cash holdings: evidence from the oil industry before and after the financial crisis. Account. Finance Res. 5 (4), 192–213.

Chirinko, R., 1993. Business fixed investment spending: modeling strategies, empirical results, and policy implications. J. Econ. Lit. 31, 1875–1911.

Table 5

Results of estimating the investment model 10 This table reports the results from estimating investment model (10). The first column refers to panel esti- mation. The last two columns present the results obtained with the SGMM method. For both SGMM specifications, all volatility measures are treated as instrument variables. In the second column, the lagged dependent variable, cash flow and the Tobin's Q are treated as endogenous, and in the last column the lagged Tobin's Q and cash flow are also assumed to be endogenous.

Panel with fixed effect System GMM(1) System GMM(2)

0.133^∗∗∗ 0.342^∗∗∗ 0.131^∗

(0.000) (0.000) (0.010)

Cashflow −0.000119 −0.00298 0.000656^∗

(0.829) (0.281) (0.044)

Cashflow(-1) 0.0169^∗ 0.0130^∗ 0.00934^∗

(0.018) (0.028) (0.026)

Q 0.00933^∗∗∗ 0.00409 0.00667^∗∗∗

(0.000) (0.371) (0.000)

Q(-1) 0.00587^∗∗∗ 0.00227 0.00541^∗∗∗

(0.000) (0.422) (0.000)

−0.00212 −0.0166 −0.00820

(0.686) (0.131) (0.187)

Sply

2_ (-1) −0.00629 −0.00142 −0.0129

(0.252) (0.928) (0.092)

−0.0700^∗∗∗ −0.0790^∗∗∗ −0.0733^∗∗∗

(0.000) (0.000) (0.000)

Dmnd

2_ (-1) 0.0333^∗∗∗ 0.0538^∗∗ 0.0438^∗∗∗

(0.000) (0.003) (0.000)

0.00431 0.00525 −0.00303

(0.144) (0.359) (0.326)

Spec

2_ (-1) 0.00441 −0.00894 −0.00151

(0.132) (0.304) (0.760)

−0.000768 −0.00467 −0.00149

(0.751) (0.186) (0.607)

Other

2_ (-1) 0.000357 0.00826 0.00278

(0.885) (0.316) (0.505)

0.00847^∗∗∗ 0.00824^∗∗ 0.00979^∗∗∗

(0.000) (0.003) (0.000)

Stck

2_ (-1) −0.0111^∗∗∗ −0.0173^∗∗∗ −0.0127^∗∗∗

(0.000) (0.000) (0.000)

Constant 0.194^∗∗∗ 0.208^∗∗∗ 0.223^∗∗∗

(0.000) (0.000) (0.000)

AR(1) −9.75 −10.09

(0.000) (0.000)

AR(2) −0.73 −1.19

(0.465) (0.235)

Hansen 5.94 4.53

0.312 0.467

Observations 10379 10379 10379

Notes: ***, ** and * indicate statistical significance at the 1%, 5% and 10%

levels, respectively. p-values are reported in parentheses.

Cukierman, A.W., 1980. The effects of uncertainty on investment under risk neutrality with endogenous information. J. Political Econ. 88 (3), 462–475.

Dixit, A.K., Pindyck, R.S., 1994. Investment under Uncertainty. Princeton, N.J.

Dvir, E., Rogoff, K., 2010. Three Epochs of Oil. Working Paper, Boston College.

Elder, J., Serletis, A., 2010. Oil price uncertainty. J. Money Credit Bank. 42, 1137–1159.

Favero, C.A., Pesaran, M.H., Sharma, S., 1992. Uncertainty and Irreversible Investment:

an Empirical Analysis of Development of Oilfields on the UKCS. Cambridge Working Papers in Economics 9210. Faculty of Economics, University of Cambridge.

Ghosal, V., Loungani, P., 1996. Product market competition and the impact of price uncertainty on investment. J. Ind. Econ. 44 (2), 217–228.

Glass, V., Cahn, E.S., 1987. Energy prices and investment over the business cycle. Energy Econ. 9 (4), 257–264.

Hamilton, J., Herrera, A., 2004. Oil shocks and aggregate macroeconomic behavior: the role of monetary policy. J. Money Credit Bank. 36 (2), 265–286.

Hamilton, J.D., 2009. Causes and Consequences of the Oil Shock of 2007-08. Brookings Papers on Economic Activity, vol. 1. Spring, pp. 215–261.

Hartman, R., 1972. The effects of price and cost uncertainty on investment. J. Econ.

Theory 5, 258–266.

Hayashi, F., 1982. Tobin's average q and marginal q: a neoclassical interpretation.

Econometrica 50, 213–224.

Henriques, I., Sadorsky, P., 2011. The effect of oil price volatility on strategic investment.

Energy Econ. 33, 79–87.

Herrera, A.M., Rangaraju, S.K., 2018. The Effect of Oil Supply Shocks on U.S. Economic Activity: what Have We Learned? SSRN working paper, no.3170422,.

Huizinga, J., 1993. Inflation uncertainty, relative price uncertainty, and investment in U.S. Manufacturing. J. Money Credit Bank. 25 (3), 521–549.

Hurn, A.S., Wright, R.E., 1994. Geology or economics? Testing models of irreversible investment using North Sea oil data. Econ. J. 104 (423), 363–371.

Im, K.S., Pesaran, M.H., Shin, Y., 2003. Testing for unit roots in heterogeneous panels. J.

Econom. 115, 53–74.

Ivanov, V., Kilian, L., 2005. A practitioner's guide to lag order selection for VAR impulse response analysis. Stud. Nonlinear Dynam. Econom. 9.

Kellogg, R., 2014. The effect of uncertainty on investment: evidence from Texas oil drilling. Am. Econ. Rev. 104 (6), 1698–1734.

Kilian, L., 2008. Exogenous oil supply shocks: how big are they and how much do they matter for the U.S. economy? Rev. Econ. Stat. 90, 216–240.

Kilian, L., 2009. Not all oil price shocks are alike: disentangling demand and supply shocks in the crude oil market. Am. Econ. Rev. 99, 1053–1069.

Kilian, L., 2017. The impact of the fracking boom on arab oil producers. Energy J. 38 (6), 137–160.

Kilian, L., 2019. Measuring Global Real Economic Activity: Do Recent Critiques Hold up to Scrutiny? . Manuscript. University of Michigan.

Kilian, L., Lee, T.K., 2014. Quantifying the speculative component in the real price of oil:

the role of global oil inventories. Int. Money Financ. 42, 71–87.

Kilian, L., Murphy, D.P., 2014. The role of inventories and speculative trading in the global market for crude oil. J. Appl. Econom. 29 (3), 454–478.

Kilian, L., Park, C., 2009. The impact of oil price shocks on the U.S. Stock Market. Int.

Econ. Rev. 50, 1267–1287.

Lamont, O., 1997. Cash flow and investment: evidence from internal capital markets. J.

Financ. 52 (1), 83–109.

Lanne, M., Saikkonen, P., 2007. A multivariate generalized orthogonal factor GARCH model. J. Bus. Econ. Stat. 25, 61–75.

Leahy, J.V., Whited, T.M., 1996. The effect of uncertainty on investment: some stylized facts. J. Money Credit Bank. 28 (1), 64–83.

Lee, K., Kang, W., Ratti, R.A., 2011. Oil price shocks, firm uncertainty and investment.

Macroecon. Dyn. 15, 416–436.

Leeb, H., Potscher, B.M., 2006. Can one estimate the conditional distribution of post- model-selection estimators? Ann. Stat. 34 (5), 2554–2591.

Levine, A., Lin, C.F., Chu, C.S., 2002. Unit root tests in panel data: asymptotic and finite- sample properties. J. Econom. 108 (1–24).

Lutkepohl, H., Milunovich, G., 2016. Testing for identification in SVAR-GARCH models. J.

Econ. Dyn. Control 73, 241–258.

Lutkepohl, H., Netsunajev, A., 2017. Structural vector autoregressions with hetero- skedasticity: a review of different volatility models. Econom. Stat. 1 (2–18).

McDonald, R., Sigel, D., 1986. The value of waiting to invest. Q. J. Econ. 101 (4), 707–728.

Mohn, K., Misund, B., 2009. Investment and uncertainty in the international oil and gas industry. Energy Econ. 31 (2), 240–248.

Odusami, B.O., 2009. Crude oil shocks and stock market returns. Appl. Financ. Econ. 19 (4), 291–303.

Ogawa, K., Suzuki, K., 2010. Uncertainty and investment: some evidence from the panel data of Japanese manufacturing firms. Jpn. Econ. Rev. 51, 170–192.

Oi, W., 1961. The desirability of price instability under perfect competition. Econometrica 29 (1), 58–64.

Pagan, A., 1984. Econometric issues in the analysis of regressions with generated re- gressors. Int. Econ. Rev. 25 (1), 221–247.

Pagan, A., 1986. Two Stage and Related Estimators and Their Applications. Review of Economic Studies, vol. 53. Oxford University Press, pp. 517–538 4.

Pagan, A., Ullah, A., 1988. The econometric analysis of models with risk terms. J. Appl.

Econom. 3 (2), 87–105.

Park, J., Ratti, R.A., 2008. Oil price shocks and stock markets in the U.S. and 13 European countries. Energy Econ. 30 (5), 2587–2608.

Phan, D., Tran, V., Nguyen, D., 2019. Crude oil price uncertainty and corporate invest- ment: new global evidence. Energy Econ. 77, 54–65.

Pindyck, R., 1991. Irreversibility, uncertainty and investment. J. Econ. Lit. 29, 1110–1148.

Rafiq, S., Salim, R.A., Bloch, H., 2009. Impact of crude oil price volatility on economic activities: an empirical investigation in the Thai economy. Resour. Pol. 34, 121–132.

Ratti, R.A., Yoon, K.H., 2011. Energy price uncertainty, energy intensity and firm in- vestment. Energy Econ. 33 (1), 67–78.

Rodman, D., 2005. xtabond2: Stata Module to Extend Xtabond Dynamic Panel Data Estimator. Center for Global Development, Washington.

Sadorsky, P., 1999. Oil price shocks and stock market activity. Energy Econ. 21, 449–469.

Sadorsky, P., 2008. Assessing the impact of oil prices on firms of different sizes: its tough being in the middle. Energy Policy 33, 3854–3861.

Sarkar, S., 2000. On the investment-uncertainty relationship in a real options model. J.

Econ. Dyn. Control 24 (2), 219–225.

Shaanan, J., 2005. Investment, irreversibility, and options: an empirical framework. Rev.

Financ. Econ. 14, 241–254.

Singhal, S.h., Ghosh, S., 2016. Returns and volatility linkages between international crude oil price, metal and other stock indices in India: evidence from VAR-DCC-GARCH models. Resour. Pol. 50, 276–288.

Tobin, J., 1969. A general equilibrium approach to monetary theory. J. Money Credit Bank. 1, 15–29.

Uri, N., 1980. Energy as a determinant of investment behaviour. Energy Econ. 2 (3), 179–183.

Wang, Y., Xiang, E., Cheung, A., Ruan, W., Hu, W., 2017. International oil price un- certainty and corporate investment: evidence from China's emerging and transition economy. Energy Econ. 61, 330–339.

Windmeijer, F., 2005. A finite sample correction for the variance for linear efficient two- step GMM estimators. J. Econom. 126 (25–51).

M. Ahmadi, et al. Resources Policy 63 (2019) 101439