Overseas oil investment projects under uncertainty: How to make informed decisions?

Available online at www.sciencedirect.com

ScienceDirect Journal of Policy Modeling 37 (2015) 742–762

Overseas oil investment projects under uncertainty: How to make informed decisions? Lei Zhu a , ZhongXiang Zhang b,∗ , Ying Fan a a

Center for Energy and Environmental Policy Research, Institute of Policy and Management, Chinese Academy of Sciences, Beijing 100190, China b College of Management and Economics, Tianjin University, 92 Weijin Road, Tianjin 300072, China Received 7 January 2015; received in revised form 12 July 2015; accepted 10 August 2015 Available online 22 August 2015

Abstract To better address the complexity of overseas oil investment and help investors to make the informed decision in overseas oil investment, this paper applies real options theory and Monte Carlo simulation to study overseas oil investment evaluation. The model has incorporated not only the uncertainties of oil price and investment cost but also the uncertainties of exchange rate and investment environment. These unique features have enabled our model to be best equipped to evaluate the value of oil overseas investment projects of three oil field sizes (large, medium, small) and under different resource tax systems (royalty tax and production sharing contracts). In our empirical setting, we have selected China as an investor country and Indonesia as an investee country as a case study. Our results show that, although production sharing contract (PSC) is convertible to that of resource royalty system, given the flexibility in PSC, it can to great extent increase the bargaining power of investee country. As there is an important tradeoff between oil resource investee country and overseas oil investor, in medium and small sized oil investment negotiation the oil company should try to increase the cost oil limit in production sharing contract and avoid the term of a windfall profits tax to reduce the investment risk of overseas oil fields. © 2015 Society for Policy Modeling. Published by Elsevier Inc. All rights reserved. Keywords: Overseas oil investment; Project value; Real options; Least squares Monte-Carlo JEL classification: Q41; Q43; Q48; G31; O13; O22; C63

∗ Corresponding author at: College of Management and Economics, Tianjin University, 92 Weijin Road, Tianjin 300072, China. E-mail address: [email protected] (Z. Zhang).

http://dx.doi.org/10.1016/j.jpolmod.2015.08.001 0161-8938/© 2015 Society for Policy Modeling. Published by Elsevier Inc. All rights reserved.

L. Zhu et al. / Journal of Policy Modeling 37 (2015) 742–762

743

1. Introduction Rapid economic growth in emerging economies like China and India has led to a huge increase in oil imports. This has raised great concern regarding their energy security. As a response, these countries have supported the expansion of their national oil companies (NOCs) as an integrated part of energy strategies to address their growing dependence on imported oil. In their view, compared to oil trading, overseas oil investment can provide a more stable oil supply to enable them to ease pressure on their domestic energy supply and can thus offset to some extent the adverse effects of high oil prices on their economies. However, the extent to which this would enhance their energy security is a matter of dispute, because these oil companies do not necessarily send the oil that they produce overseas back to home countries. Instead, they prefer to let market considerations dictate where it is sold (Zhang, 2012a,b). But, there is a great consensus that investments in oil fields overseas help to stabilize the oil prices by pumping more oil out of the fields and enlarging the overall availability of oil on the world market. With the world’s oil use set to rise, accompanied with increasing world’s oil prices, adding new capacities to world oil supplies via oil overseas investment is seen as beneficial to all, and needs thus to be encouraged and appreciated. Thus, the real issue is not about where the oil produced from overseas investment goes. Rather, it is overseas investment itself. While such an investment will benefit oil users, whether it is in the best interests of investors needs a careful evaluation, particularly given large capital investment involved in and a very long duration of an oil investment project. According to Wu (2008), in recent years, as the link between the government and the NOCs has been loosened perceptively, in overseas investment, the NOCs are behaving more like international oil companies (IOCs). Under this condition, one important purpose of overseas oil investment is to improve their poor reserve/production ratios. So the assessment and evaluation of risks are playing significant roles in their overseas oil activities. In conventional investment project evaluation, commodity price uncertainty is always used to reflect the project uncertainty. For oil reserves valuation, oil price is always be used to reflect oil project uncertainty, and in most cases it is the only uncertainty which is considered in oil reserves valuation. As more investments have been made in emerging countries of rich oil and gas resources, the meaning of uncertainty or risk in such investment may be much greater than that of oil price uncertainty. In our opinion, only taking oil price uncertainty into account cannot fully capture the complexity of overseas oil investment. In general, overseas oil investment has the same properties as that of foreign direct investment (FDI) in that the development of overseas oil fields is associated with large capital budgets, a long construction period, and high investment uncertainty. However, the decision process for overseas oil investment is more complex than that of FDI. As operational factors like product mix and reserve replacement efforts have lost some of their historical influence on the investment process (Mohn & Misund, 2009), some other uncertainty factors should be considered in valuing overseas oil investments as the development is mainly carried out through international or national oil companies with the added complication of the investee country’s resource tax system. Therefore, to more accurately evaluate the risk and benefits of overseas oil investment, a proper overseas oil investment evaluation method is needed for oil companies (both NOCs and IOCs) to address these uncertainty factors and consider their complex interrelationships, including the impact of different resource taxes. This paper applies real options theory to establish an overseas oil investment evaluation model based on Monte Carlo simulation, and is solved by the Least Squares Monte-Carlo method (LSM). Several uncertainty factors most relevant to overseas oil investment are considered, including not only the oil price and investment cost, but also the exchange rate and investment environment. The model can evaluate the value of oil fields of

744

L. Zhu et al. / Journal of Policy Modeling 37 (2015) 742–762

different sizes (large, medium, small) and under different resource tax systems (royalty tax and production sharing contracts). The remainder of the paper is organized as follows. Section 2 provides a literature review. Section 3 describes our model. Section 4 undertakes a case study of an overseas oil investment project under a variety of uncertainty factors and different resource tax systems. Also we present a cross-country comparison as sensitivity analysis. Section 5 presents main conclusions and suggestions for further research. 2. Literature review For overseas oil investment, in recent years, a lot of literature has focused on the challenges from NOCs to IOCs, especially the NOCs in some emerging countries located in Asian countries (Mitchell & Lahn, 2007). Ma and Andrews-Speed (2006), Xu (2007), Wu (2008) all pointed out that since 2002 the link between the government and the NOCs had loosened, as the NOCs are increasingly behaving like IOCs in overseas investment. As the assessment and evaluation of risk have played significant roles in overseas oil investment, both NOCs and IOCs need to pay more attention to the uncertainties associated with overseas oil projects, not only the oil price, but also some other external factors related to overseas oil investment. The real options approach is well suited to estimate future uncertainty in natural-resource investments. Theories of irreversible investment have motivated a number of studies of the naturalresource evaluations (Brennan & Schwartz, 1985; Paddock et al., 1988; Smith & Nau, 1995; Smith & McCardle, 1996,1999). In recent years, Schwartz and Trolle (2010) had developed a model for pricing expropriation risk in oil projects. The model was used to investigate, under the uncertainty of oil price, the option value that the government had to expropriate oil resource from the oil company during oil project development period. Fan and Zhu (2010) applied real options theory to overseas oil investment by adding an investment-environment factor to oil-resource valuation, under oil-price, exchange-rate, and investment-environment uncertainties, different countries’ oilinvestment situation can be compared by using the option value index (OVI) calculated by their model. As their model was based on closed-form solution and infinite time horizon, however, it cannot be used in overseas oil project evaluation. In this paper, we have considered multiple risk factors to reflect the project uncertainty in oil resource evaluation. Multifactor models of project evaluation have been proposed being much more successful in capturing the behavior of asset value, as well as taking into account the different types of flexibilities available to decision makers when managing their projects (Kulatilaka, 1995; Bernardo & Chowdry, 2002; Hsu & Schwartz, 2003). The introduction of multifactor models into real option models with many interacting flexibilities increases the difficulty of solving them, making traditional numerical approaches inadequate. So some sort of computer-based simulation procedures for solving American options has been applied to solving multifactor financial options (Tilley, 1993; Barraquand & Martineau, 1995; Raymar & Zwecher, 1997; Broadie & Glasserman, 1997; Andersen, 2000). In this paper, with several uncertainty factors considered, we applied LSM method to solve our model. LSM method was developed for valuing American options and was based on Monte Carlo simulation and least squares regression (Longstaff & Schwartz, 2001; Schwartz, 2004). It is viewed as one of the most promising new approaches in solving complex financial options (Clement, Lamberton, & Protter, 2002; Moreno and Navas, 2003; Stentoft, 2004; Cortazar, Gravet, & Urzua, 2008). The most relevant work related to this paper is Fan and Zhu (2010). Compared to their work, we have established a new overseas oil investment evaluation model with several significant

L. Zhu et al. / Journal of Policy Modeling 37 (2015) 742–762

745

improvements. First, on the scope of the study, we focus on the investment evaluation of overseas oil project, while the work of Fan and Zhu (2010) uses an option value index (OVI) to compare the oil investment values among different countries. Second, on modeling, we establish a sequential investment real options model based on Monte-Carlo simulation and with finite time horizon. By contrast, in the work of Fan and Zhu (2010), the critical value which resulted by closed-form solution and infinite time horizon can be used to compare oil investment in different countries, but cannot be used for project evaluation. Third, for uncertainty factors, we have considered the uncertainty of oil investment cost, which is highly related to oil geological conditions and extraction technologies. Fourth, in the process of investment environment, we take the uncertainty of investment environment to be reflected in the oil production cost, while the work of Fan and Zhu (2010) takes investment environment as a multiplier in the calculation of oil revenue. It is more realistic to make investment environment reflected in oil production cost, because, in oil project evaluation, the effects of investment environment change may first pass on to oil production cost, then affect oil project cash flows and revenues. Finally, as our model is based on real options and Monte Carlo simulation, the model can better describe complex oil resource tax systems among different investee countries. By contrast, oil resource tax system had not been modeled in the work of Fan and Zhu (2010). 3. The model This paper emphasis on the evaluation of overseas oil investment and does not consider the barriers for oil companies to enter investee (resource) countries. The evaluation includes oil project construction period and development (operation) period. It does not consider the exploration period. And we are focusing on the investment evaluation of overseas oil projects, in which the capital costs of different oil companies have not been considered. The model established here can be used by all the oil companies, either NOCs or IOCs, to evaluate their overseas activities. During the construction of oil project, the oil company can decide whether to continue investment or give up the project according to the new information at each investment stage. It has the right to exercise the abandon option to terminate the oil project in any investment stage. In general, at the initial stage the oil company will sign a contract with local government to specify the oil field development years and total investment amount. After the expiration of the contract, the oil company can either extend the development period in the contract, or transfer the exploitation rights back to local government. Assuming the total period for oil field development is T years, for the purpose of evaluation we divide the T years into N periods, each with a length of t = T/N, and define tn = nt, n = 0, 1, . . . N. 3.1. Modeling uncertainty factors 3.1.1. Oil prices, exchange rate, and investment environment (oil production cost) Changes in oil prices will directly affect the gain of overseas oil investment. Therefore they have a significant impact on oil project decision. Here assuming the international oil price follows a geometric Brownian motion (Pindyck, 1999): dPOil = αP POil dt + σP POil dzP

(1)

where POil is oil price √ in units of U.S. dollar/Barrel; dzP is the independent increments of Wiener process dzP = εP dt, where εP is a normally distributed random variable with mean 0 and

746

L. Zhu et al. / Journal of Policy Modeling 37 (2015) 742–762

standard deviation 1; and αP and σ P represent the drift and variance parameters of the oil price, respectively. Eq. (1) represents the dynamics of the true process for the oil price. For valuation purposes, however, the risk neutral or risk adjusted process for oil price will be used: dPOil = (αP − υP )POil dt + σP POil dzP

(2)

where υC is the risk premium associated with the oil price process. Let X ≡ ln(POil ), and applying Ito’s Lemma yields:   1 dX = αP − σP2 − υP dt + σP dzP (3) 2 In the simulations, the discrete approximation to oil price process in risk-neutral version is:    1 2 1/2 (4) POil (ti+1 ) = POil (ti ) exp αP − σP − υP t + σP (t) εP 2 Both oil prices and overseas investment are denominated in U.S. dollars. Changes in U.S. dollar exchange rates will to some extent affect the oil price and the overseas oil valuation. In previous works of currency options pricing, some scholars let the exchange rate to follow geometric Brownian motion (Biger & Hull, 1983; Geske & Johnson, 1984). Investment environment is a necessary external condition for overseas investment activities. In previous studies of investment environment, most of them depended on quantitative analysis (Stobaugh, 1969; Mun & Ho, 1979). And the most authoritative index system for evaluating different countries’ investment environments is ‘Ease of Doing Business’ scale developed by the World Bank (Djankov, Manraj, McLiesh, & Ramalho, 2005; Djankov, McLeish, & Shleifer, 2007). These indexes of investment environment can hardly be incorporated into investment project evaluation. Nordal (2001) used the real options approach to study the impact of risk in emerging-market countries on foreign direct investment by adding country risk to project valuation1 . Our paper assumes that the investment environment would mainly affect the oil production cost, and we use the uncertainty of oil production cost to represent the uncertainty of investment environment. This treatment implies that, on the one hand, the oil production cost can to some extent reflect different countries’ oil quality and geographical diversity; on the other hand, that the uncertainty of future oil production cost is caused by the changes in investment environment. Here also assuming that the exchange rate and investment environment follow geometric Brownian motions, and their risk neutral process with discrete approximation forms are:    1 2 1/2 (5) SE (ti+1 ) = SE (ti ) exp αS − σS − υS t + (t) εS 2    1 2 1/2 (6) COil (ti+1 ) = COil (ti ) exp αC − σC − υC t + (t) εC 2 where SE and COil are the exchange rate between investor country’s currency and U.S. dollar, and oil production cost in units of U.S. dollar/Barrel; εS and εC are normally distributed random variables with mean 0 and standard deviation 1; and αS and σ S represent the drift and variance parameters of the exchange rate, respectively, and αC and σ C represent the drift and variance parameters of the investment environment (oil production cost), respectively. υS and υC are the 1

Nordal (2001) defined a country-state variable, and assuming that this variable followed geometric Brownian motion.

L. Zhu et al. / Journal of Policy Modeling 37 (2015) 742–762

747

risk premium associated with the exchange rate and investment environment processes. And this paper also considers the correlation among U.S. exchange rate, investment environment, and oil price, ρPS denotes the correlation coefficient between U.S. exchange rate and oil price2 , ρPC denotes the correlation coefficient between oil price and investment environment (oil production cost), and ρCS denotes the correlation coefficient between U.S. exchange rate and investment environment (oil production cost). 3.1.2. Investment cost As a large-scaled project with sequential investment, once the oil company starts to invest in overseas oil field, at the initial stage, assuming KInv is the expected total investment cost for project construction, the total investment remaining at time ti is KInv (ti ). The investment expenditure of each time period is defined as IInv . As overseas oil investment is highly related to international oil prices, so here IInv is set as a linear function of oil price, IInv (ti ) = iPOil (ti ), where i is oil project investment rate. It means the investment expenditure of each period will increase as the oil prices rises so it can speed up the completion of the project. Because the capital budget of overseas oil investment is quite large, such a large investment is inevitably facing uncertainties (e.g., the uncertainties of exploration technology and oil field geological condition). These uncertainties cannot be perfectly foreseen at the initial stage and will cause the changes in the remaining investment at each period. That makes the actual investment amount different from the capital budget specified in the contract. Here assuming the remaining total investment KInv is uncertain in order to reflect the uncertainty of overseas oil investment cost (Dixit & Pindyck, 1994), KInv follows the controlled diffusion process: dKInv = −IInv dt + β[IInv KInv ]0.5 dx

(7)

where β is a scale parameter representing √ the uncertainty surrounding KInv ; and dx is the indepenrandom variable dent increment of Wiener process dx = ε dt, where ε is a normally distributed  2   , with mean 0 and standard deviation 1. The variance of KInv is Var(KInv ) = β2 / 2 − β2 KInv whereby uncertainty of oil investment cost reduces as KInv decreases (Dixit & Pindyck, 1994), which can be viewed as, along with the investment continuing, the oil company is becoming more experienced with dealing with such uncertainties. In addition, as we have discussed before, the uncertainty in KInv mainly comes from geological and extraction technology uncertainties, so dx is assumed to be uncorrelated with the market portfolio, thus investment cost uncertainty will not have a risk premium associated to it. In the simulations, the discrete approximation to remaining investment cost process is: KInv (ti+1 ) = KInv (ti ) − iPOil (ti )t + β[iPOil (ti )KInv (ti )]1/2 (t)1/2 εx

(8)

This model assumes that the switching expenditure IInv is a linear function of oil price POil . As there does not exist any adjustment cost or other cost related to the changes of investment expenditure IInv , the investment rule has a bang–bang solution at any time before the oil investment is completed (Majd & Pindyck, 1987). Therefore, the optimal investment expenditure amount will

2 Theoretically, when the U.S. dollar falls, the dollar-denominated crude oil price is lower against other currencies, which pushes up the crude oil price to some extent, and vice versa (Fan et al., 2008). This point is also argued by Reboredo (2012), which showed that the correlation analysis suggested that the increase in oil prices is weakly associated with U.S. dollar depreciation, and vice versa. But he also mentioned the dependency increased noticeably with the onset of the financial crisis, mainly for oil-exporting countries.

748

L. Zhu et al. / Journal of Policy Modeling 37 (2015) 742–762

either be IInv = 0 or IInv = IInvmax (Majd & Pindyck, 1987; Dixit & Pindyck, 1994; Schwartz, 2004). At the initial stage, the oil company will either take maximum switching rate imax to do the oil investment or abandon the project. Therefore, under the condition of the optimal investment rule, i = imax . Because KInv is uncertain, the time needed to complete the oil investment is uncertain, too. τ The actual oil investment cost can be known only after the investment has been completed, i=0 IInv (ti ), where τ is the actual time it takes to finish overseas oil investment. Here we have separated the uncertainties during oil project construction period from the uncertainties during project development (operation) period. In our view, the uncertainty of investment cost will play a major role during the construction period, as it is highly related to geological and extraction technology uncertainties, while the uncertainty of investment environment will play a major role during the operation period, as the operation of an overseas oil project will be greatly affected by investee country’s investment environment (which is to some extent the same to that of foreign direct investment projects). It should be acknowledged that both two uncertainty factors will play roles during the whole oil construction and development (operation) period, but their effects are not the same at the same periods. 3.2. Overseas oil investment valuation Assuming in overseas oil field development period, the crude oil production at tn period is QOil (tn ), and all of the crude oil produced at tn period can be sold at the same period. Assuming the crude oil production capacity is constant in each period. r is the interest rate, differing across countries. Investee country’s resource tax system has been added into the valuation. Oil resource tax systems can be divided into two major categories, including resource royalty and production sharing contract (PSC). Furthermore, some countries also levy windfall profits tax in domestic oil field according to oil prices change. These three resource tax systems have been modeled in our cash flow calculations. 3.2.1. Operational value of overseas oil project After overseas oil investment has been completed, the project starts producing oil. At any time ti in oil development period, CF(ti ) is the cash flow that the oil company can obtain through oil production and sale. Cash flows under three resource tax systems are modeled as follows: (1) Under resource royalty system, the cash flow CF1 (ti ) that the oil company can obtain is represented as: CF1 (ti ) = [POil (ti ) · QOil (ti ) · (1 − Tax1 ) − COil (ti ) · QOil (ti )] · (1 − Tax2 ) · SE (ti ) (9-1) where Tax1 and Tax2 are the resource royalty and income tax rate of investee country, respectively. (2) Under production sharing system, the cash flow CF2 (ti ) can be represented as:

CF2 (ti ) = POil (ti ) · (QOil (ti ) − (QOil (ti ) − clOil ) · gGov ) − COil (ti ) · QOil (ti ) (9-2) ·(1 − Tax2 ) · SE (ti ) where clOil is the cost oil limit under the PSC; gGov is the share of investee country’s government in profit oil at each period.

L. Zhu et al. / Journal of Policy Modeling 37 (2015) 742–762

749

(3) If the investee country has levied windfall profits tax, under existing production sharing system, the cash flow CF3 (ti ) can obtain can be represented as:

CF3 (ti ) =

⎧ ⎪ ⎨ ⎪ ⎩

 if POil (ti ) > PTax3 ,

CF2 (ti ) −

if POil (ti ) ≤ PTax3 ,

CF2 (ti )

(POil (ti ) − PTax3 ) · QOil (ti ) ·Tax3 · (1 − Tax2 ) · SE (ti )

 (9-3)

where Tax3 is investee country’s windfall profits tax according to oil prices change, which is equal to special oil income levy tax by some oil producing countries (e.g. Venezuela). The government will levy the tax only when the oil price is large than threshold price PTax3 of windfall profits tax. Once overseas oil investment is completed, at any time ti in oil development period, the operational value for the oil company that continues to operate the oil project should be the sum of discount cash flows from ti to the end of development period, which can be represented as: VOil (ti ) =

N 

e−r(tn −ti ) CF(tn )

(10)

n=i

3.2.2. Investment value of overseas oil project The oil project would not generate any cash flow during construction. So the cash flows calculated in construction period can to some extent be viewed as anticipated cash flows. Under the option analysis framework, in oil project construction period, the oil company owns the abandon option. At the oil investment completed time τ, the investment value of oil project is equal to that of operational value: FOil (τ) = VOil (τ)

(11)

In any period before the oil investment is completed, if the investment needed is higher than expected oil project value, the oil company will exercise the abandon option to terminate the project to prevent more losses. The investment value of oil project can be denote as FOil (ti ), which depends on the expected cash flows after oil investment has been completed and the cost needed to complete oil project investment. So at period ti before the oil investment is completed, the value of oil investment is: 

 FOil (ti ) = max 0, Eti e−r(ti+1 −ti ) FOil (ti+1 ) − IInv (ti ) (12) where Eti [∗] means, at period ti , the expected value for oil company to continue to hold abandon option. As aforementioned, the investee country’s government may have penalty for oil company to abandon the project in investment stage. That will also be a default loss for the company to bear. With the penalty, at period ti before the oil investment is completed, the value of oil investment can be rewritten as: 

 FOil (ti ) = max −Pen, Eti e−r(ti+1 −ti ) FOil (ti+1 ) − IInv (ti ) (13) where Pen is the penalty that the oil company should pay for abandoning investment project. The penalty may occur in some oil development contracts. However, because of data limitations, while we incorporate this parameter in our model, this has not been considered in our empirical study.

750

L. Zhu et al. / Journal of Policy Modeling 37 (2015) 742–762

We need to estimate the value of Eq. (13) in order to estimate the value of oil project. As the expected value Eti e−r(ti+1 −ti ) FOil (ti+1 ) is hard to determine, we apply Least Squares Monte Carlo (LSM) method to compute the expected value and oil project value. See details of the solution processes in Appendix A1. 4. A case study We select an oil company in China as the overseas investor and Indonesia in Southeast Asia as the oil resource investee as the case study for the real options evaluation model discussed in Section 3. Indonesia is the largest oil importer to China in Asia, and in recent years, China’s NOCs have carried out a series of oil activities in Indonesia. Now the China National Petroleum Corporation (CNPC) is among the largest oil developers in Indonesia. We apply the model to evaluate whether China should invest in an overseas oil investment project in Indonesia, taking into consideration oil price, exchange rate, investment environment, and oil investment cost uncertainties, and the impacts of different resource tax systems. 4.1. Model parameters Table 1 shows the parameter values of the model. The data is based on the year 2006. It should be noted that it is difficult to quantify the fluctuation of an investee country’s investment environment. We use the consumer price index (CPI) as a reflection of a country’s degree of policy stability in accordance with Fan and Zhu (2010). Some oil-investee countries have highly unstable policies, which often lead to huge price fluctuations and deterioration of the investment environment. Therefore, the investee country’s CPI volatility is used as a proxy variable to reflect changes in its investment environment. Because of the lack of comprehensive overseas oil investment data, we refer to the research of Blake and Roberts (2006), who suggest evaluating three typical sized oil fields (large, medium, small). All the parameters and their assumed values are given in Appendix A2. Fig. 1 shows the changes of these uncertainty factors in 250 out of 5000 simulation paths. A large sample of random routing Monte Carlo simulation can simulate the result of every possible change in the uncertainty factors. We also consider the correlations between these uncertainty factors in our model to better quantify the impacts of the uncertainties on the value of the overseas oil investment. The value of overseas oil projects of different field sizes can be calculated by the LSM, taking the set parameter values into the model and simulating changes for each uncertainty factor based on their initial values. For simplicity, we use O-L, O-M, and O-S to denote oil fields with large, medium, and small recoverable reserves, respectively. 4.2. Results and discussions In this section, we simulate the impacts of different resource tax systems on the evaluation of three sized overseas oil projects. Under the royalty system, a fixed percentage agreed between the government and the oil company is charged on the gross oil production. The PSC system is similar to the royalty system except that the rate is applied after consideration of production costs (cost oil limit). With foreign investors investing in oil resource development, by implementing different resource tax systems, the investee country’s government can also be viewed as one of the stakeholders in the development of the oil fields. The resource tax system is of great significance

40 20

b (Paths: 250 of 5000)

4 2030

2025

2020

2015

2010

2005

0

1000 500 0

c

2020

8

1500

2015

12

2010

16

2000

2005

20

Residual Investment Cost (millions US Dollars)

d (Paths: 250 of 5000)

300.00

Oil Price

(Paths: 250 of 5000) Development Cost

9.00 8.00 7.00 6.00 5.00 4.00 3.00 2.00 1.00 0.00

Exchange Rate

250.00 200.00 150.00 100.00 50.00 2030

2025

2020

2015

2010

2005

0.00

RMB:1 Dollar

Exchange Rates (RMB:1 Dollar)

(Paths: 250 of 5000)

US Dollar/Barrel

2030

2005

a

2030

0

2030

2025

2020

2015

2010

0

60

2025

300

80

2025

600

100

2020

900

751

2015

1200

2010

Development Cost (US Dollar/Barrel)

1500

2005

Oil Prices (US Dollar/Barrel)

L. Zhu et al. / Journal of Policy Modeling 37 (2015) 742–762

e Fig. 1. (a) Oil prices simulation (Paths: 250 of 5000) (b): oil production cost simulation (Paths: 250 of 5000) (c) exchange rate simulation (Paths: 250 of 5000) (d) residual investment cost simulation (Paths: 250 of 5000) (e) single simulated path of oil prices, production cost, and exchange rate.

to the investee country as such kind of tax revenue is highly linked to its fiscal policy (El Anshasy & Bradley, 2012). The PSC system predominates in Indonesia, the investee country in our study. Here we adopt Eq. (9-2) in project cash flow calculation. Considering the randomness of the samples in Monte Carlo simulation, in order to have a more accurate result, we calculate five seeds for each value of oil investment under different parameter settings. Each seed has a result based on 5000 simulation paths using LSM. The average of the five seeds is taken as the value of the oil investment under each parameter setting.

752

L. Zhu et al. / Journal of Policy Modeling 37 (2015) 742–762

Table 1 Oil project values with different seeds for large oil field. Oil field size: large (O-L)

Project Value (Millions RMB) Project value of equivalent oil production capacity (Millions RMB/Millions Barrels per year) Percentage of paths abandoned Project completion period (years)

Seed 1

Seed 2

Seed 3

Seed 4

Seed 5

Average

17,107.15

16,284.52

16,610.38

16,151.83

17,252.64

16,681.30

1425.60

1357.04

1384.20

1345.99

1437.72

1390.11

0.180%

0.140%

0.100%

0.180%

0.120%

0.144%

3.69

3.69

3.69

3.67

3.64

3.68

Take O-L as an example. As shown in Table 2, taking into consideration the four uncertainty factors, the oil project value of O-L lies between 16,151.83 and 17,252.64 million RMB, with a mean of 16,681.30.million RMB, and the expected construction time is 3.68 years. The oil investment is abandoned in only 0.144% of paths, implying that the investment risk of O-L is small. 4.2.1. Production sharing contract The project values of the three different sizes of oil fields (O-L, O-M, O-S) have been calculated given the same production sharing rate as in Appendix A1, against which other cases that consider the aforementioned uncertainty factors are measured. As shown in Table 2, for the overseas oil investment, the project values of O-L and O-M are much larger than that of O-S. The project value of equivalent oil production capacity in O-S is only 7.88% and 13.93% to that of O-L and O-M, respectively. There are two main reasons for this. First, for equivalent oil production capacity, the investment needed in O-S is much larger than that of O-L and O-M (for equivalent oil production capacity, the extra investment needed in O-S would be an increase of 147.40% and 48.61% relative to that of O-L and O-M, respectively). Second, the investment risks of O-L and O-M are lower than that of O-S. The percentages of paths abandoned in O-L and O-M are 0.144% and 0.360%, respectively, which are much smaller than that of 19.727% for O-S. So O-L and O-M are the preferred choices in making overseas oil investment over O-S of a greater risk. Table 2 Oil project values with resource production sharing contract. Production sharing contract

Project value (Millions RMB) Project value of equivalent oil production capacity (Millions RMB/Millions Barrels per year) Percentage of paths abandoned Project completion period (years)

Oil field size Large (O-L)

Medium (O-M)

Small (O-S)

16,681.30 1390.11

4542.24 757.04

337.45 112.48

0.144% 3.68

0.360% 3.58

19.727% 3.27

L. Zhu et al. / Journal of Policy Modeling 37 (2015) 742–762

753

Table 3 Oil project values with resource royalty system. Resource tax rate: 26.67%

Project value (Millions RMB) Project value of equivalent oil production capacity (Millions RMB/Millions Barrels per year) Percentage of paths abandoned Project completion period (years)

Oil field size Large (O-L)

Medium (O-M)

Small (O-S)

16,668.91 1389.08

4531.54 755.26

319.35 106.45

0.200% 3.68

0.640% 3.58

20.780% 3.26

4.2.2. Resource royalty system In mathematical calculation, resource royalty system is to great extent convertible to that of production sharing contract. By letting Eq. (9-1) equals to that of Eq. (9-2), the relationship between resource tax rate Tax1 , cost oil limit clOil , and government share gGov can be derived. Given the PSC parameters of Indonesia in Appendix A2, the equivalent resource tax rate is 26.67%. Taking Eq. (9-1) into the model, and with resource tax rate as 26.67%, the results are shown in Table 3. Given the equivalent resource tax rate to that of PSC, the project values of the three sizes of the oil fields are nearly the same as that in Table 2. For oil investee country’s government, resource tax rate should be defined by legislation, and always taken as a unified rate to all kinds of oil fields. It can ensure a given share of benefits for local government in its oil developing, but the lack of flexibility has limited the government bargaining power with foreign investors. Compared to resource royalty system, as PSC would be negotiated among each of the individual oil projects, it is not required by legislation. Although PSC may have some general rules, the local government is able to adjust and negotiate the production sharing rate with foreign investors, and such flexibility can to great extent increase the local government’s bargaining power in its oil development. 4.2.3. Ladder production sharing rates Following the discussions above, in recent years, oil resource investee countries have introduced different production sharing rates according to oil field production capacity to encourage foreign oil companies to develop their medium or small sized oil fields. As the production sharing rate (cost oil limit) is related to oil field quality, here we keep the investee country’s cost oil limit in the O-M unchanged, but change the rate in the O-S from 2.00 to 2.40 million barrels/year (an increase) and in the O-L from 8.00 to 6.00 million barrels/year (a decrease). As our results are shown in Table 4, when the cost oil limit rate increases from 66.67% to 80.00%, the project value of O-S increases by 217.63%. And the O-S’s investment risk has been reduced to 3.2%, which makes it more attractive for investment. And with the decrease of cost oil limit in O-L (from 66.67% to 50%), although the investment risk of the O-M is still low, its project value decreases significantly (a nearly one third decrease in value), meantime the project value of equivalent oil production in the O-L being only slightly larger than that in the O-M. So the ladder production sharing rates can narrow the diversity between different oil fields, and increase the bargaining power of local government. On the one hand, by setting a lower cost oil limit rate, the investee country can seek greater benefits with large sized oil fields. On the other hand, by setting a higher cost oil limit rate, it can make its small sized oil fields more attractive to foreign investment.

754

L. Zhu et al. / Journal of Policy Modeling 37 (2015) 742–762

Table 4 Oil project values with resource ladder production sharing rate. Ladder production sharing rate

Project value (Millions RMB) Project value of equivalent oil production capacity (Millions RMB/Millions Barrels per year) Percentage of paths abandoned Project completion period (years)

Oil field size Large (O-L)

Medium (O-M)

Small (O-S)

10,447.92 870.66

4542.24 757.04

1071.85 357.28

0.320% 3.70

0.360% 3.58

3.200% 3.47

Table 5 Oil project values under windfall profits tax. Windfall profit tax investee country

Project value (Millions RMB) Project value of equivalent oil production capacity (Millions RMB/Millions Barrels per year) Percentage of paths abandoned Project completion period (years)

Oil field size Large (O-L)

Medium (O-M)

Small (O-S)

13,202.16 1100.18

2548.72 424.79

92.79 30.93

0.600% 3.66

0.880% 3.58

53.720% 2.91

4.2.4. Windfall profits tax When the oil price is high, some countries also levy a windfall profits tax according to a given level of oil price. Windfall profits tax is kind of “contingent tax claim” of investee country, and can be viewed as another option for investee country to keep more benefits in domestic oil resource developing. Here the threshold price of windfall profits tax is set at US$70/barrel and the tax rate at 20%. By adopting Eq. (9-3) in project cash flow calculation, the results are shown in Table 5. With windfall profits tax implemented, the project values of the three sizes of oil fields decrease, especially that of the O-S whose percentage of paths abandoned significantly rises to 53.720%, which has been increased by 172.32% compared to that in Table 2. Although the project values of the O-L and O-M have decreased, their investment risks to great extent remain at a low level. Therefore, in overseas oil investment, small sized oil fields will be most affected by the levy of windfall profits tax. As the model is based on Monte Carlo simulation using a large sample, the model can better describe complex oil resource tax systems among different investee countries. Therefore, the model has good applicability for an evaluation of overseas oil investments. 4.3. Cross-country comparisons The discussions in Section 4.2 are all based on a case study of a single country (Indonesia). Here we will calculate the project values of O-L, O-M, and O-S in different countries to undertake a cross-country comparison. The comparison can be viewed as another kind of sensitivity analysis of the parameters adopted in our model. We choose 5 countries (including Indonesia) among 26 target countries defined in the work of Fan and Zhu (2010), which are all with rich oil resources

L. Zhu et al. / Journal of Policy Modeling 37 (2015) 742–762

755

in different continents. For each country, the initial investment cost and production capacity of three typical sized oil fields are all set as the same values (which also refers to Blake and Roberts, 2006), and each country’s specific data (e.g. oil production cost, investment environment, etc) are all come from Fan and Zhu (2010). As resource tax system had not been taken into account in Fan and Zhu (2010), here we assume, except Indonesia, all the other countries have royalty tax systems, and the tax rates are set at 30%. Such royalty tax rate level can to some extent have the same offset effect on these countries’ oil revenue to that of PSC in Indonesia. Taking the all the parameter values into the model, we can then get the simulation results for different sized oil fields in different countries. For comparison, we also show the ranking results of these countries based on OVI conducted by Fan and Zhu (2010). As shown in Table 6, with the same trend to the evaluation results in Indonesia, in each country, the project values of O-L and O-M are all much larger than that of O-S, as well as the investment risks of O-L and O-M are lower than that of O-S. And if we compare the project value or project value of equivalent oil production capacity among these countries in the same sized oil fields, we can see that the countries with high OVI rankings in Fan and Zhu (2010) are having high project value in our results. For example, the project values of O-L, O-M, and O-S in Canada are all larger than the values in Venezuela, Indonesia, and Kuwait, respectively, as the ranking results of Canada is also higher than these countries in Fan and Zhu (2010). It should be pointed out that, if we compare the investment risks among these countries in the same sized oil fields, we can see that for O-L and O-M, the investment risks in these countries are quite low (the investment risks of O-L and O-M among these countries are all less than 1.000%). For O-S, although the investment risks are larger than that of O-L and O-M, they are still at a low

Table 6 Cross-country comparison results. Cross-country comparison

Norway—Europe Large (O-L) Medium (O-M) Small (O-S) Canada—North American Large (O-L) Medium (O-M) Small (O-S) Kuwait—Middle East Large (O-L) Medium (O-M) Small (O-S) Venezuela—S. & Cent. America Large (O-L) Medium (O-M) Small (O-S) Indonesia—Asia

OVI ranking results in Fan and Zhu (2010)

Our results Project value (Millions RMB)

Project value of equivalent oil production capacity (Millions RMB/Millions Barrels per year)

26,091.22 9069.21 2137.22

2174.27 1511.53 712.41

0.360 0.660 1.040

21,363.66 6803.88 1070.24

1780.30 1133.98 356.75

0.660 0.620 2.680

15,251.27 3541.52 14.17

1270.94 590.25 4.72

0.120 0.620 17.240

14,112.85 3426.61 156.62

1176.07 571.10 52.21

0.340 0.360 37.660

Percentage of paths abandoned (%)

4

11

20

25

21

756

L. Zhu et al. / Journal of Policy Modeling 37 (2015) 742–762

level in the countries with good investment environment3 (the investment risk of O-S is 1.040% in Norway, and 2.680% in Canada). And in the other three countries in which the investment environment is not as good as Norway or Canada, the investment risks of O-S are much larger than that of O-L and O-M (the investment risk of O-S is 17.240% in Kuwait, and 37.660% in Venezuela). So it can be inferred that the investment risks of O-S in Canada and Norway are much smaller than that of the other three countries mainly because these two countries have more stable investment environment. 5. Conclusions With the world’s oil use set to rise, accompanied with increasing world’s oil prices, adding new capacities to world oil supplies via oil overseas investment is seen as beneficial to all, and needs thus to be encouraged and appreciated. However, making overseas oil investment is a complex process, and a number of uncertainty factors play important roles in overseas oil development activities. Thus, the evaluation of overseas oil project should take into consider not only the uncertainties of oil price, but also multiple other uncertainties. To help investors to make the informed decision in overseas oil investment, our paper first establishes an overseas oil investment evaluation model. Given that real options analysis is considered to better reflect the flexibility and impacts of uncertainty factors on the value of overseas oil investment, our model has considered a number of uncertainty factors by applying real options analysis and is solved by the Least Squares Monte-Carlo (LSM) method. We then employ the model to study and compare the values of three typical sized oil fields (large, medium, small) in overseas oil investment. Finally, we have undertaken a cross-country comparison to examine the applicability of basic trends generated from Indonesia. Our model has several advantages in overseas oil investment project evaluation. It has good applicability as it is based on Monte Carlo simulation and solved by the LSM, by which both the investment risk and project value can be calculated. It is also easy to simulate different resource tax systems in our model. It can take multiple uncertainty factors into account. In this paper we have considered four uncertainty factors (oil price, investment cost, exchange rate, and investment environment) and their interrelationships. Also we take overseas oil investment as a multi-stage investment decision problem so that the investment option during oil investment stage has been taken into consideration. These unique features have enabled our model to be best equipped to evaluate the value of oil overseas investment projects. With the simulation of different oil resource tax systems, our analysis shows that there is an important tradeoff between oil resource investee country and overseas oil investor. On the one hand, to encourage foreign oil companies to develop their medium or small seized oil fields, oil resource investee countries prefer to adjust their resource tax systems to balance the resource valuation diversity among different size oil fields, in particular by means of production sharing rates. On the other hand, with oil price fluctuation in recent years, to obtain more oil development benefits, some investee countries also levy a windfall profits tax so that more oil revenues can remain in their country. Therefore, in the oil investment negotiation between the oil company and investee country, the oil company should first try to the extent possible to increase the cost oil limit in PSC for O-M and O-S in order to obtain more benefits in the development of such fields. Our results show that if the

3 The investment environment discussed here is mainly referring to the index from “Ease of Doing Businessdeveloped ¨ by World Bank (Djankov et al., 2005, 2007), as well as the OVI ranking results from Fan and Zhu (2010).

L. Zhu et al. / Journal of Policy Modeling 37 (2015) 742–762

757

cost oil limit increases, the project value of O-S increases. Second, as small sized oil fields are more sensitive to windfall profits tax, the oil company also needs to avoid the term of a windfall profits tax in oil development contract in order to further reduce the investment risk of overseas small sized oil fields. Finally, our cross-comparison among five investee countries from different continents shows that the basic trends among three typical sized oil fields in each country remain unchanged. This to some extent suggests the applicability of our model in different countries. Also our results show that the investment risks of O-S in Canada and Norway are much smaller than the other three countries because of their more stable investment environment. But the oil resources in these countries are in depletion. The oil companies have no choice but to carry out overseas activities in some developing countries without good investment environment. Therefore, our model developed here is helpful for oil companies to evaluate the risks and benefits of overseas oil investment. Our model has incorporated a number of uncertainty factors to better reflect the reality of overseas oil investment. However, making overseas oil investment is a complex decision process. Although the real options model established in this paper adds extra functionality over existing models such as the NPV method, there exist some limitations. First, the model does not consider the uncertainty of oil production capacity, or the relationship between output and the investment environment. In general, nearly all oil fields will to a varying degree suffer production decline, and also oil output may to some extent decline as the investment environment deteriorates. Second, a lot of oil companies are involved in overseas oil exploration activities. Therefore, how to combine the exploration process into our model is also an important issue. These issues are examples of interesting issues that need to be addressed in our future work.

Acknowledgments This article is the revised and shortened version of our lengthy paper “An Evaluation of Overseas Oil Investment Projects under Uncertainty Using a Real Options Based Simulation Model”, Nota di Lavoro 83.2011, Fondazione Eni Enrico Mattei, Milan, Italy. This study is supported by the National Natural Science Foundation of China under Grant Nos. 71273253, 70825001 and 71133005. It has benefited from helpful comments from Antonio Maria Costa and four anonymous referees. That said, the views expressed here are those of the authors. The authors bear sole responsibility for any errors and omissions that may remain.

Appendix A. Appendix A1: LSM based model solution The LSM method was developed for valuing American options and is based on Monte Carlo simulation and least squares regression (Longstaff & Schwartz, 2001; Schwartz, 2004). In the model developed here, the oil company has the abandon option before overseas oil investment is completed. And the oil company will evaluate the decision to abandon the oil investment at each discrete time point. The detail of the solution procedure is as follows. Take G as simulation paths, for any path g, conditional on not having abandoned oil project before, at the final date of operational period (time N, the last stage of operational period), the value of the oil project is given by the boundary condition:

758

L. Zhu et al. / Journal of Policy Modeling 37 (2015) 742–762

WOil (g, tN ) = CF(g, tN )

(A-1)

At any period ti , for those paths along which the investment has been completed, the value of the oil project is computed recursively by: WOil (g, ti ) = e−r(ti+1 −ti ) · WOil (g, ti+1 ) + CF(g, ti )

(A-2)

For those paths along which the investment is not completed, the conditional expected value of continuation is estimated by regression. The dependent variable is the discounted value of oil project at ti+1 period, e−r(ti+1 −ti ) · WOil (g, ti+1 ), and the independent variable is the oil project ˆ Oil (g, ti ) can be estimated by polynoanticipated cash flow at period ti . The fitted value W 4 ˆ Oil (g, ti ) with the mial regression. Comparing the conditional expected value of oil project W investment expenditure IInv (ti ), then:  WOil (g, ti ) =

ˆ Oil (g, ti ) < IInv (g, ti ) W

0,

ˆ Oil (g, ti ) − IInv (g, ti ), W

(A-3)

ˆ Oil (g, ti ) ≥ IInv (g, ti ) W

The recursion proceeds by rolling back in time and repeating the procedure until the exercise decisions at each possible exercise time along each path have been determined. The value of the oil project is then computed by starting at time zero, moving forward along each path until the final observation date of a given period or until the first stopping time occurs, discounting the resulting cash flows to time zero, and taking the average over all the paths to get the project value of overseas oil field with abandon option. For more discussion on the method used here, see Schwartz (2004). ⎛

N 

−r(tn −t0 )

τg 

⎞ −r(tn −t0 )

e CF(g, tn ) − e IInv (g, tn ), ⎟ ⎜ ⎟ ⎜ n=τ +1 n=1 ⎟ ⎜ g G ⎜ ⎟  1 ⎟ ⎜ if path g is not abandoned Exp VOil (t0 ) = ⎟ ⎜ ⎟ ⎜ or G g=1 ⎜ ⎟ τg ⎟ ⎜  ⎠ ⎝ −r(tn −t0 ) − e IInv (g, tn ), if path g is abandoned

(A-4)

n=1

where at any path g, if the oil project is abandoned, τ g is the abandon period in path g, else τ g is the investment completed period in path g. LSM method described has been implemented in Matrix Laboratory (MATLAB). Appendix B. Appendix A2: The parameters and the assumed values of the model

4 Laguerre polynomials are applied in this regression with nine terms used in the implementation of the algorithm. The fitted value of this regression is the best linear unbiased estimator of the conditional expectation (Longstaff and Schwartz, 2001; Schwartz, 2004).

Model symbol

Unit

Value (investee country)

Note

Oil field recoverable reserves-large (O-L) Oil field recoverable reserves-medium (O-M) Oil field recoverable reserves-small (O-S) Production capacity of oil field-large Production capacity of oil field-medium Production capacity of oil field-small Oil prices Oil prices drift rate Oil prices standard deviation rate

ROil

300 150 75 12 6 3 60 0.02 30%

The data of three typical sized oil fields refer to the work of Blake and Roberts (2006)

POil αP − υP σP

million barrels million barrels million barrels million barrels/year million barrels/year million barrels/year US dollar/barrel /year /year

Exchange rate

SE

US dollar:RMB

1.00:8.00

Exchange drift rate Exchange standard deviation rate

αS − υ S σS

/year /year

–0.005 7.55%

Oil production cost

COil

US dollar/barrel

6.64

Oil production cost drift rate Oil production cost standard deviation rate

αC − υC σC

/year %/year

0.01 17.85%

Total investment cost of oil field-large

KInv

million US dollar

1730

million US dollar/year

1440 1070 550 470 360 0.5

Total investment cost of oil field-medium Total investment cost of oil field-small Initial annual investment-large Initial annual investment-medium Initial annual investment-small Investment uncertainty

QOil

IInv

β

WTI 2006 yearly average oil spot price has been used in this work Set by this study The data refers to the estimation of oil price volatility from Fan and Zhu (2010) 2006 yearly average exchange rate between US dollar and RMB has been used in this work Set by this study The data refers to the estimation of exchange rate volatility from Fan and Zhu (2010) Oil production cost is derived from the International Energy Agency (IEA) (2003). In this work, the inflationary cost index published by International Energy Agency (IEA) (2006) was used to estimate oil production cost in 2006 Set by this study The data refers to the estimation of consumer price index volatility from Fan and Zhu (2010) The investment data of three typical sized oil fields refer to the work of Blake and Roberts (2006). In this work the investment costs have been adjusted to 2006 with inflationary cost index published by International Energy Agency (IEA) (2006). And for equivalent oil production capacity, the investment cost needed in small and middle sized oil fields are larger than that of large sized oil field, which will increase by 66.47% and 147.40% to that of large sized oil field Set by this study

759

Here refers to the settings in the research of Schwartz (2004), Dixit and Pindyck (1994)

L. Zhu et al. / Journal of Policy Modeling 37 (2015) 742–762

Parameter

760

Appendix B (Continued) Parameter

Model symbol

Correlation between exchange rate and oil price Correlation between exchange rate and oil develop cost

Unit

Note

ρPS

−0.6998

ρCS

−0.5490

Correlation between oil price and oil develop cost Resource royalty

ρPC

0.6874

Tax1

0.00%

Income tax

Tax2

30.00%

Windfall profits tax

Tax3

0.00%

Cost oil limit-large Cost oil limit-medium Cost oil limit-small Share of government-large Share of government-medium Share of government-small Riskfree rate

clOil gGov

r

/year

8 4 2 80.00% 80.00% 80.00% 7.99%

It is estimated through the calculation between oil price and exchange rate historical data. See details in Fan and Zhu (2010) It is estimated through the calculation between investee country’s CPI and exchange rate historical data. See details in Fan and Zhu (2010) It is estimated through the calculation between oil price and investee country’s CPI historical data. See details in Fan and Zhu (2010) The investee country does not levy resource royalty in oil fields development Income-tax data have been obtained from the foreign-investment database of the Ministry of Commerce of China The investee country does not levy windfall profit tax in oil fields development. And we will discuss the case of windfall profit tax in results and discussions Refer to previous oil production sharing contracts, Here we set the cost oil limit is 2/3 of total oil production

Trigger oil price for windfall profits tax

PTax

US dollar/barrel

0

Development period

T

Year

2006–2030

Time step size in simulations Number of simulations

t G

year

1 5000

million barrels/year million barrels/year million barrels/year

Refer to previous oil production sharing contracts, Here we set the government’s share of profit oil is 80% Investee country’s long-term deposit interest rate is used as a risk-free rate, see details in Fan and Zhu (2010) The investee country does not levy windfall profit tax in oil fields development. And we will discuss the case of windfall profit tax in results and discussions As the years of the contract in overseas oil development always last 20–25 years, therefore we set the development period at 25 years, which can be divided into oil project construction and oil field operation periods In general, the simulation results will start to convergence as paths exceed 1000, so the number of paths simulated in different scenarios are set at 5000

L. Zhu et al. / Journal of Policy Modeling 37 (2015) 742–762

Value (investee country)

L. Zhu et al. / Journal of Policy Modeling 37 (2015) 742–762

761

References Andersen, L. (2000). A simple approach to the pricing of Bermudian swaptions in the multi-factor libor market model. Journal of Computational Finance, 3, 5–32. Barraquand, J., & Martineau, D. (1995). Numerical valuation of high dimensional multivariate American securities. Journal of Financial and Quantitative Analysis, 30(3), 301–320. Bernardo, A., & Chowdry, B. (2002). Resources, real options and corporate strategy. Journal of Finance, 63, 211–234. Biger, N., & Hull, J. (1983). The valuation of currency options. Financial Management, 12(1), 24–28. Blake, A., & Roberts, M. C. (2006). Comparing petroleum fiscal regimes under oil price uncertainty. Resources Policy, 31, 95–105. Brennan, M. J., & Schwartz, E. S. (1985). Evaluating natural resource investments. Journal of Business, 58, 135–157. Broadie, M., & Glasserman, P. (1997). Pricing American-style securities using simulation. Journal of Economics Dynamics and Control, 21(8), 1323–1352. Clement, E., Lamberton, D., & Protter, P. (2002). An analysis of a least squares regression method for American option pricing. Finance and Stochastics, 6(4), 449–471. Cortazar, G., Gravet, M., & Urzua, J. (2008). The valuation of multidimensional American real options using the LSM simulation method. Computers & Operations Research, 35(1), 113–129. Dixit, A. K., & Pindyck, R. S. (1994). Investment under uncertainty. Princeton NJ: Princeton University Press. Djankov, S., Manraj, D., McLiesh, C., & Ramalho, R. (2005). Doing business indicators: Why aggregate, and how to do it. In World Bank working paper. World Bank, Washington DC. Djankov, S., McLeish, C., & Shleifer, A. (2007). Private credit in 129 countries. Journal of Financial Economics, 84, 265–590. El Anshasy, A. A., & Bradley, M. D. (2012). Oil prices and the fiscal policy response in oil-exporting countries. Journal of Policy Modeling, 34(5), 605–620. Fan, Y., Zhang, Y. J., Tsai, H. T., & Wei, Y. M. (2008). Estimating ‘value at risk’ of crude oil price and its spillover effect using the GED-GARCH approach. Energy Economics, 30(6), 3156–3171. Fan, Y., & Zhu, L. (2010). A real options based study on overseas oil investment and its application in China’s oversea oil investment. Energy Economics, 32, 627–637. Geske, R., & Johnson, H. E. (1984). The American put option valued analytically. The Journal of Finance, 39(5), 1511–1524. Hsu, J., & Schwartz, E. S. (2003). A model of R&D valuation and the design of research incentives. Los Angeles: UCLA: Anderson School. International Energy Agency (IEA). (2003). World energy investment outlook. Paris: OECD/IEA. International Energy Agency (IEA). (2006). World energy outlook. Paris: OECD/IEA. Kulatilaka, N. (1995). Operating flexibilities in capital budgeting: Substitutability and complementarity in real options. In L. Trigeorgis (Ed.), Real options in capital investments: New contributions. New York, NY: Praeger. Longstaff, F. A., & Schwartz, E. S. (2001). Valuing American options by simulation: A simple least square approach. Review of Financial Studies, 14(1), 113–147. Ma, X., & Andrews-Speed, P. (2006). The overseas activities of China’s national oil companies: Rationale and outlook. Minerals & Energy—Raw Materials Report, 21(1), 17–30. Majd, S., & Pindyck, R. S. (1987). Time to build, option value, and investment decisions. Journal of Financial Economics, 18(1), 7–27. Mitchell, J., & Lahn, G. (2007). Oil for Asia. London: The Royal Institute of International Affairs. Mohn, K., & Misund, B. (2009). Investment and uncertainty in the international oil and gas industry. Energy Economics, 31, 240–248. Moreno, M., & Navas, J. F. (2003). On the robustness of least-squares Monte Carlo (LSM) for pricing American derivatives. Review of Derivatives Research, 6(2), 107–128. Mun, K., & Ho, S. (1979). Foreign investment in Hong Kong. In T. Z. Lin, R. P. Lee, & U. E. Simonis (Eds.), Hong Kong: Economic, social, and political studies in development, with a comprehensive bibliography. New York, NY: M.E. Sharpe. Nordal, K. B. (2001). Country risk, country risk indices and valuation of FDI: A real options approach. Emerging Markets Review, 2, 197–217. Paddock, J. L., Siegel, D. R., & Smith, J. L. (1988). Option valuation of claims on real assets: the case of offshore petroleum leases. Quarterly Journal of Economics, 103, 479–508. Pindyck, R. S. (1999). The long-run evolution of energy prices. In MIT-CEEPR working papers. Available at: http://mit.dspace.org/handle/1721.1/45087.

762

L. Zhu et al. / Journal of Policy Modeling 37 (2015) 742–762

Raymar, S., & Zwecher, M. (1997). A Monte Carlo valuation of American call options on the maximum of several stocks. Journal of Derivatives, 5(1), 7–23. Reboredo, J. C. (2012). Modelling oil price and exchange rate co-movements. Journal of Policy Modeling, 34(3), 419–440. Schwartz, E. S. (2004). . Patents and R&D as real options (vol 33) Economic notes by Banca Monte dei Paschi di Siena SpA. Schwartz, E. S., & Trolle, A. B. (2010). Pricing expropriation risk in natural resource contracts—A real options approach. In William Hogan, & Federico Sturzenegger (Eds.), The natural resource trap. Cambridge, MA: MIT Press. Smith, J. E., & Nau, R. F. (1995). Valuing risky projects: Option pricing theory and decision analysis. Management Science, 41(5), 795–816. Smith, J. E., & McCardle, K. F. (1996). Valuing oil properties: Integrating option pricing and decision analysis approaches. Operations Research, 46, 198–217. Smith, J. E., & McCardle, K. F. (1999). Options in the real world: Lessons learned in evaluating oil and gas investments. Operations Research, 47, 1–15. Stentoft, L. (2004). Assessing the least squares Monte-Carlo approach to American option valuation. Review of Derivatives Research, 7(2), 129–168. Stobaugh, R. (1969). How to analyze foreign investment climates. Harvard Business Review, 9, 100–108. Tilley, J. A. (1993). Valuing American options in a path simulation model. Transactions of the Society of Actuaries, 45, 42–56. Wu, K. (2008). China’s overseas oil and gas investment: Motivations, strategies, and global impact. Oil, Gas & Energy Law, 6(1), 1–9. Xu, X. (2007). Chinese NOCs’ overseas strategies: Background, comparison and remarks. Joint Baker Institute/Japan Petroleum Energy Center Policy report. Houston: The James A. Baker III Institute for Public Policy, Rice University. Zhang, Z. X. (2012a). The overseas acquisitions and equity oil shares of Chinese national oil companies: A threat to the west but a boost to China’s energy security? Energy Policy, 48, 698–701. Zhang, Z. X. (2012b). Why are the stakes so high? Misconceptions and misunderstandings in China’s quest for energy security. In Huw McKay, & Ligang Song (Eds.), Rebalancing and sustaining growth in China (pp. 329–356). Canberra: Australian National University E-Press.