R&D Renewable Project Valuation: a Study of
Applicable Valuation Methods
Master’s Thesis Finance
L.R. Derksen
S2754614
Supervisor: dr. J.H. von Eije
MSc Thesis Finance
Energy Focus Area
University of Groningen, The Netherlands
This paper analyzes the applicability of various valuation models to R&D renewable projects, focusing primarily on the traditional Discounted Cash Flow model and the Real Options Analysis. Since increasingly more companies recognize the importance of the renewable energy transition, the number of R&D projects in the field of energy steadily increases. These R&D renewable projects often require a different valuation approach compared to projects traditionally carried out by energy companies. In this paper, I argue that a combination of the Real Options Analysis and the Discounted Cash Flow model that recognizes growth options and managerial flexibility provides managers with the most accurate representation of the actual project and enhances the understandability of the potential payouts of proposed projects. The challenge for finance departments is to translate this rather complex approach to understandable calculations and advice.
Key words: R&D, valuation, renewable energy, real options, managerial flexibility
1. Introduction
The attention of academic literature as well as public news on the renewable energy transition has rapidly increased over the past few years. The goals set in the Paris Agreement of 2015 illustrate the increasing importance of the energy transition. The Paris Agreement provides a framework for making voluntary pledges, thereby acknowledging the primacy of domestic politics, that can be compared and reviewed internationally, in the hope that the ambition of reducing global warming can be increased through a process of ‘naming and shaming’ (Falkner, 2016). The agreement has been signed by 195 countries. This illustrates the international consensus that climate issues should be dealt with internationally and clarifies the increasing importance of finding solutions in the field of renewable energy sources.
The increasing demand for renewable energy triggers firms to invest in R&D projects related to finding new ‘green’ ways to generate energy. The core business of finance departments within energy firms traditionally was to value and assess investments in projects with much similarities with their existing businesses and with relatively certain outcomes. However, the energy transition requires innovation and enforces these firms to look for other ways to generate energy, encouraging them to invest in R&D projects in the field of renewable energy. The share of renewable energy sources in the total public energy RD&D (Research, Development & Demonstration) has more than doubled over the past decade (International Energy Agency, 2017). The increase in these types of R&D projects requires finance departments to re-evaluate their valuation model and determine whether this model is still accurate for this new kind of projects.
hydrogen and fermentation of gasification of biomass. The projects of this department are very diverse and have high technological and regulatory risks.
The current valuation methodology of Gasunie is mainly focused on large infrastructure projects which have very different characteristics compared with the R&D-type of projects of the New Energy department. The finance department wants to gain more insight into the valuation of these relatively small R&D renewable projects and determine whether their current valuation methodology is still viable for these projects. Whereas current literature primarily focuses on the different valuation methods separately, this paper looks at a broad set of valuation techniques simultaneously and provides practitioners with guidelines regarding the usefulness and limitations of certain approaches.
In this paper, I argue that using a Real Options Analysis in combination with or next to the traditional Discounted Cash Flow method provides the most accurate results when valuing R&D renewable projects. I will analyze the decision-making process at Gasunie and highlight the most important managerial considerations that impact decision-making. In addition, I will construct an example project and show how the traditional DCF model and the Real Options Analysis value this project, to support the advice I provide in this paper.
The main finding of this paper is that the Real Options Analysis is very useful for evaluating R&D renewable projects. The Discounted Cash Flow model used at Gasunie is deemed appropriate for most of the projects at the company. In this paper, I argue that a Real Options Analysis adds value by looking at R&D projects at a different and in a more realistic way. Due to the mathematical complexity of option pricing models, the usefulness of Real Options Analysis might be perceived as limited from a quantitative perspective. However, even the more qualitative and theoretical aspects of the Real Options Analysis might add considerable value in the decision-making process by introducing managerial flexibility during the project’s lifetime.
renewable projects. After that, section 4 shows how the valuation techniques differ in their outcomes with an example project. In section 5, I provide a comprehensive advice for a valuation framework for R&D renewables projects. Section 6 will conclude the paper and provide additional advice for practitioners and suggestions for further research.
2. Theoretical framework
This theoretical framework provides a detailed and comprehensive overview of the valuation techniques that are most widely accepted across academics and deemed appropriate for valuing R&D projects. I will highlight the most important papers on these techniques, explain the most important characteristics and describe their applicability to R&D renewable projects. More specifically, I will briefly outline the essential strengths and weaknesses that practitioners should bear in mind when applying the valuation methods. First, I will deal with the Discounted Cash Flow method, after which I will consider the Decision Tree Analysis and the Real Options Analysis. Finally, I will highlight some other valuation methods that are quite obviously less applicable for the valuation of R&D renewable projects.
2.1.1. Discounted Cash Flow model with CAPM
The most common and broadly accepted method for the valuation of investments is the Discounted Cash Flow (DCF) method. This method uses the weighted average cost of capital (WACC) to discount estimated future cash flows. If the sum of the discounted cash flows is more than the investment costs, the project has a positive Net Present Value (NPV). The standard formula for DCF calculations is:
𝑁𝑃𝑉 = 𝐶𝐹𝑡
(1+𝑟)𝑡+
𝑇𝑒𝑟𝑚𝑖𝑛𝑎𝑙 𝑉𝑎𝑙𝑢𝑒
(1+𝑟)𝑁 (1)
In this formula, investments are included in the cash flows (𝐶𝐹). The Terminal Value (or Continuing Value), which is the value of the project after the explicit forecast period, is often calculated using the Gordon Growth Model. This model is based on the following formula:
𝑇𝑒𝑟𝑚𝑖𝑛𝑎𝑙 𝑉𝑎𝑙𝑢𝑒 =𝐶𝐹𝑓𝑖𝑛𝑎𝑙 𝑦𝑒𝑎𝑟 𝑒𝑠𝑡𝑖𝑚𝑎𝑡𝑖𝑜𝑛 𝑝𝑒𝑟𝑖𝑜𝑑 ∗(1+𝑔)
4 In this formula, 𝑟 corresponds to the discount rate (WACC), and 𝑔 corresponds to
the long-term growth rate of the cash flows. Koller et al. (2015) argue that this formula should be slightly adapted as it contains some hidden, but crucial, assumptions.1 The general NPV decision rule is that all projects with a positive
NPV create value for the firm and thus should be adopted. Another aspect of the Discounted Cash Flow method is the Internal Rate of Return (IRR). The IRR is equal to the discount rate at which the NPV would be zero. The IRR decision rule implies that all projects with an IRR higher than the project-specific WACC add value to the firm and consequently are good projects to pursue. However, using the NPV decision rule can be considered better than using the IRR decision rule. The IRR cannot account for changing discount rates (which might be relevant for long-term projects) and is also not workable for projects with both negative and positive cash flows. Considering that R&D renewable projects often require multiple investments during the lifetime of the project, resulting in a mix of negative and positive cash flows, using the NPV decision rule is arguably more convenient. The Discounted Cash Flow method can be used for valuing businesses, entire firms, buildings, projects and any other asset that generates future cash flows.
The WACC consists of the weighted sum of the opportunity cost of equity and the opportunity cost of debt of comparable risky projects (with the same systematic risk). There exist various methods to calculate the cost of equity. Probably the best known and most used method is the Capital Asset Pricing Model (CAPM), as introduced by Sharpe (1964) and Lintner (1965) and built upon Markowitz’s model of portfolio choice (1952). More than 50 years later, the model is still widely used in corporate finance and is also the one used for the valuation of projects at Gasunie. The Capital Asset Pricing Model is perceived attractive, because it has powerful and understandable predictions about how to measure risk and the link between expected return and risk (Fama and French, 2004). The model consists of a risk-free rate, a beta that reflects the asset’s sensitivity with respect to the market, and the expected market return. The betas of comparable firms are often
1 Koller et al. (2015) use the following to calculate the Continuing Value:
𝐶𝑜𝑛𝑡𝑖𝑛𝑢𝑖𝑛𝑔 𝑉𝑎𝑙𝑢𝑒𝑡=
𝑁𝑂𝑃𝐿𝐴𝑇𝑡+1(1− 𝑅𝑂𝑁𝐼𝐶𝑔 )
used to estimate the sensitivity to the market of the firm or project. The general formula to calculate the cost of equity using CAPM is the following:
𝑅̅𝑒 = 𝑅𝑓+ 𝛽(𝑅̅𝑚− 𝑅𝑓) (3)
where 𝑅̅𝑒 corresponds to the expected cost of equity of the asset, 𝑅𝑓 is equal to the
risk-free rate (often corresponds to the yield on government bonds of countries with a perceived low risk, such as interest on German government bonds), 𝑅𝑚 is the expected return on the market (often corresponds to the long term history of equity returns in a good proxy market) and 𝛽 corresponds to the volatility of the specific asset with respect to the market. Instead of taking the spot rate of 10 or 20-year German government bonds as the risk-free rate, Gasunie often takes the two or five-year averages. This provides a relatively stable risk-free rate 𝑅𝑓 in the cost of equity with a long-term perspective, which corresponds to the long-term investment decisions Gasunie generally makes. The threat of this is to not be fully up-to-date with respect to the most recent economic developments, as their impact slightly diminishes when using averages. However, it prevents the company from making investment decisions that highly depend on the use of the spot rate, which changes very frequently. For investment decisions with a short horizon, using the spot rate or using some shorter period for averaging spot rates is expected to be more accurate. To estimate the return on the market, Gasunie often uses the average historical return on the indexes that are expected to be closely aligned with the market the specific asset is in. Gasunie generally focuses thereby on Europe, as that is where its business is located.
The CAPM is the most widely accepted and used model for DCF calculations. In the following section, I will briefly outline an extension of the CAPM that sometimes is used to get more accurate valuations.
2.1.2. Discounted Cash Flow model with other discount rate calculations
There have been various extensions of the CAPM that try to better estimate the cost of equity part of the WACC and therefore improve the accuracy of the Discounted Cash Flow method for valuation. In 1992, Eugene Fama and Kenneth French introduced their famous three-factor model, which basically added two factors to the Capital Asset Pricing Model. In their model, they account for their finding that in general value stocks and small-cap stocks outperform the market. Their empirical results showed very strong statistical significance, making their three-factor model well-known and moderately accepted across the economic community. Following the three-factor model, investors in a company require a higher risk premium if its stock returns are correlated with those of small companies or high book-to-market companies (Koller et al., 2015). The general formula for the required rate of return using the three-factor model is the following:
𝑅̅𝑒 = 𝑅𝑓+ 𝛽1(𝑅̅𝑚− 𝑅𝑓) + 𝛽2(𝑆𝑀𝐵) + 𝛽3(𝐻𝑀𝐿) (4)
where SMB stands for a Small-Minus-Big, which is the return of a portfolio which is long in relatively small stocks and short in relatively large stocks, and HML stands for High-Minus-Low, which is the return of a portfolio which is long in stocks with a high book-to-market ratio (value stocks) and short in stocks with a low book-to-market ratio (growth stocks). 𝑅̅𝑒 is equal to the expected cost of equity
of the asset, 𝑅𝑓 corresponds to the risk-free rate, 𝑅𝑚 is the expected return on the market and the betas correspond to the volatility of the asset with respect to the three factors in the model.
the HML factor and the SMB factor. I will not further elaborate on these factors, since it is outside the scope of this paper and is not regarded to be of vital importance for the valuation of R&D renewable projects.
For the sake of completeness in this paper, I will outline one other method that is sometimes used for DCF calculations. This method is the Arbitrage Pricing Theory (APT), which was first introduced by Ross (1976). He proposed this model as an alternative to the mean variance Capital Asset Pricing Model, mainly because of false assumptions in the CAPM. Whereas the CAPM assumes market efficiency, the APT allows for market inefficiency, which is a more realistic assumption. The APT model uses multiple factors to determine the required rate of return. The corresponding general formula is the following:
𝑅𝑖 = 𝑅𝑓+ 𝛽1𝐹1+ 𝛽2𝐹2 + ⋯ + 𝛽𝑁𝐹𝑁+ 𝜀 (5)
𝑅𝑖 is the required return on asset 𝑖, 𝐹𝑛 is the risk premium associated with factor 𝑛 for the average asset in the market and 𝛽𝑁 is the sensitivity of the specific asset 𝑖 to the factor 𝑛. The 𝜀 corresponds to random noise and is on average equal to zero, as one generally assumes that investors can hold well-diversified portfolios (Koller et al., 2015). The APT may add some accuracy to the DCF model, but presumably does not work for R&D renewable projects. Koller et al. (2015, pp. 306) state: “In practice, implementation of the model has been tricky, as there is little agreement about how many factors there are, what the factors represent, or how to measure the factors”. It requires extensive research to estimate the sensitivity of a project’s cash flows to a broad set of relevant factors. For R&D renewable projects, one could argue that obtaining accurate estimates for all these betas is even impossible, making the Arbitrage Pricing Theory not applicable for this kind of projects.
This might not look like a very strong argument at first, but the other models require significant additional calculation work and are harder to explain to people with a non-finance background. It is hard to estimate the costs of switching precisely, but the additional benefits are not proven either. Switching to either the Fama-French three-factor model or the Carhart four-factor model is also not likely to have a very significant impact on the valuation. These models might add some accuracy, but this benefit most likely does not outweigh the aforementioned costs. Moreover, estimating the additional betas will be very hard for R&D renewable projects with no historical data and very high uncertainty. My advice to Gasunie would therefore be to stay with the CAPM for Discounted Cash Flow calculations. In the next section, I will outline the most important strengths and limitations of the Discounted Cash Flow model that practitioners should be aware of when using this model to value R&D renewable projects.
2.1.3. Strengths and limitations of the Discounted Cash Flow model
First of all, with all valuation methods the following rule applies: garbage in, garbage out. This basically means that quality inputs will result in a low-quality output of the model. Therefore, estimating accurate inputs is very important for the quality of the valuation. This certainly applies to the Discounted Cash Flow model. Koller et al. (2015) state that the Discounted cash flow model is the most accurate and flexible model for valuing projects, but the accuracy of the model depends on the estimated inputs.
clear decision criteria. If the NPV is positive, the project creates value. Next to that, project proposals are easy to compare based on this decision rule.
results is to do sensitivity analyses or a Monte Carlo simulation. A Monte Carlo simulation provides more insight in the probabilities of different outcomes and helps to understand the possible scenarios. It does however still not account for managerial flexibility during the project. I will show later in this paper that using a Real Options Analysis as supplement to the Discounted Cash Flow model provides a more sophisticated and accurate manner to overcome this problem. But first, I will describe the Decision Tree Analysis and analyze its applicability to R&D renewable projects in the next section.
2.2.1. Decision Tree Analysis
The Decision Tree Analysis (DTA) has traditionally been developed and used to capture the concept of managerial flexibility, which lacks in the Discounted Cash Flow model. It provides a graphical representation that shows the decisions management has to make and how they impact the value of the project. With a Decision Tree Analysis, one needs to evaluate all possible outcomes and assign probabilities to them. These outcomes are contingent on previous events. The decision tree shows the impact of the decision on the likelihood of the project’s outcomes. In addition, DTA incorporates some managerial flexibility by allowing to include expanding or abandoning the project into the decision tree. Figure 1 shows a simple example of a decision tree with two decision moments as it could look like for R&D project valuation.
The squares represent decision moments and the circles represent (intermediate) outcomes. A key aspect of the decision tree is that it displays how new information (success or failure) results in new decisions to be made. One should acknowledge that a DTA for R&D renewable project valuation is often more complex and involves a greater number of nodes than the one displayed in Figure 1. To calculate the value of the project, DTA builds upon the principles of the Discounted Cash Flow model. Each cash flow displayed in the decision tree should be multiplied by its probability and being discounted by the risk adjusted discount rate. There is some disagreement in the academic literature on whether or not one discount rate can be used to discount the cashflows in the entire decision tree. Making a certain choice might impact the risk associated with the project and therefore the discount rate should be adjusted for the risk profile associated with each scenario.
Decision Tree Analysis is not an alternative to the Discounted Cash Flow model, but rather extends the model. It provides more insight into the contingent decisions managers make, but the outcomes are based on DCF calculations. For this reason, I will also not include the DTA in the example project of section 4. The aim of that section is to show how Real Options Analysis differs from traditional DCF calculations (on which the DTA also is based), as I consider the Real Options Analysis to be often the most viable way of looking at R&D renewable projects. But first, I will outline some important strengths and limitations of the Decision Tree Analysis.
2.2.2. Strengths and limitations of the Decision Tree Analysis
A first strength of the DTA is that it is a very transparent graphical representation of the possible scenarios of a business case and thus provides managers with clear insight in the possible impact of the decisions they make. Moreover, it also indicates that there must be new information based on which the management can decide in the future. The DTA thus can incorporate some flexibility, which was lacking in the DCF model. Its calculations are not more complex than those of a traditional Discounted Cash Flow model and should therefore be understandable for people with a non-finance background.
be multiplied with its individual probability, before discounting the payouts. This can be a very difficult task, especially for projects with a long time horizon. Project managers might overstate the probabilities of success, just to make it look like that what they are doing is valuable for the firm. The probabilities in the DTA do however have a significant impact on the value of the project and misprediction regarding these probabilities can lead to making wrong decisions. Therefore, applying the DTA in an organization requires a strict and comprehensive framework of auditing and motivating project managers to correctly predict the probabilities. Next to that, the probabilities should not be based upon the expectations of one person, but reflect the expectations of various people having different roles within the project. The R&D renewable projects of Gasunie have already gone through the first phases of research. This will make it easier to accurately estimate the probabilities for different scenarios, as the uncertainty regarding the outcome of the project has already slightly decreased. Another limitation of the Decision Tree Analysis is that it uses the same discount rate for the entire decision tree, leading to incorrect results according to Makropoulou (2011). He states that the value of the option to invest, which is related to future managerial decisions, does not have the same riskiness as the value of the project cash flows. Assuming that there are future decisions to be made, based on new information, implies that the option to invest (or expand in the example of Figure 1) should have considerable lower riskiness, and thus a lower discount rate, as it limits downside risk. I will come back at this point later in this paper. The following section outlines the Real Options Analysis and discusses its applicability to the valuation of R&D renewable projects.
2.3.1. Real Options Analysis
adaptive behavior. Real options exist when there could be some news that will possibly arrive in the future which may affect decision-making. They are often abundant within firms and projects, but not recognized as such. Copeland and Antikarov (2001) point out the following five types of real options:
- Option to abandon the project - Option to expand the project
- Option to extend the life of the project - Option to defer a project
- Options to contract or scale back
Some investments create future opportunities to invest, rather than creating cash flows itself. When considering real options with R&D renewable projects, I will primarily focus on the option to move on from the pilot phase to the production phase of a project, since this is presumably the most important type of real option for Gasunie. This real option can be viewed as a growth option or an expansion option. Later on in this paper I will also briefly highlight the option to defer an investment. All of the real options that I will describe have the structure of an American call option with no dividends (exercisable at any time up to the expiration date). The value of a call option at the expiration date is equal to:
𝐶 = 𝑀𝑎𝑥(𝑆 − 𝑋, 0) (6)
If the firm takes this first investment decision, it implicitly acquires the right (but not the obligation) to continue with the project when this first phase ends (Schneider et al., 2008). This option arises with every new phase of the R&D project. Real Options Analysis builds upon recognizing and valuing learning and adaptive behavior. When an organization wants to use a Real Options Analysis and thus also value the implied managerial flexibility, the organization should structure its projects in a way that flexibility during the project actually exists. If such flexibility in practice is actually not present and projects just proceed until the point that it has either succeeded or failed, applying a Real Options Analysis would not be the accurate approach for the valuation. In practice, there often exists some reluctance to abandon projects when there already have been made some monetary commitments. This is a behavioral aspect within decision-making, which I will further elaborate on in section 5.
Real options can be either shared or proprietary. Shared real options can be exercised by more than one party, whereas proprietary real options are held by a single firm. To make a real option have significant economic value, there has to be some exclusivity. When the option does not provide some exclusive rights, everyone in the market can exercise the option. Holding the option provides no economic benefits in such cases. Since the R&D renewable projects at Gasunie are technically innovative and presumably hard to replicate quickly by other parties once the application turns out to be successful, I consider the real options present at these projects as proprietary real options. Therefore, the options do have significant economic value, which makes it worthwhile to actually value them.
There are several ways to cover real option value in the valuation process. The first and most easy one is simply doing a DCF analysis of the expected cash flows and assessing the real options qualitatively. In this way, ROA is valuable through gaining awareness of the options that an investment generates, which then may be decisive when the DCF analysis alone does not provide convincing results for the management. However, being able to actually calculate the real option value of a project is of course more beneficial. The first and probably easiest way to do so is using a decision tree in combination with some scenario planning. The idea here is that you outline the possible outcomes of the underlying project and determine their individual probabilities, identify under which scenarios you will not exercise the option, after which you can calculate the NPV of the remaining scenarios. This kind of real option valuation can also be done using Monte Carlo Simulation. With a Monte Carlo Simulation, one can model the probability of the set of different outcomes of the underlying (the project in this case). Subsequently one removes the scenarios under which you will not exercise the option, after which the value of the option can be calculated. Other ways to calculate real option value build upon models and calculations for financial options. The most common model based on partial differential equations and used to value financial and real options is the Black & Scholes option pricing model (Luehrman, 1998). This model was introduced by Black and Scholes in 1973. The equations they introduced can be found in Appendix A. This equation, which calculates the value of a call and put options on stocks, has six input variables:
- The exercise price (or strike price) of the stock - The current stock price
- The time to expiration of the option - The volatility of the stock
- The risk-free rate
- The dividend yield of the stock
I did not include the sixth variable, the dividend yield, since with the real options present in R&D renewable projects the underlying asset generally does not generate any cash flows before the expiration date of the option. With regard to the value of the underlying asset, I adopted the Market Asset Disclaimer (MAD), as suggested by Copeland and Antikarov (2001). With financial options, the underlying asset is often traded in the market and its value is thus publicly available. With R&D renewable projects, it is very challenging to find another traded asset that is perfectly correlated with the considered project and has the exact same payouts (a twin security). Copeland and Antikarov (2001) advocate that in such cases the best and unbiased estimate of the underlying asset’s value is the present value of the project without flexibility. This present value can be obtained by using the Discounted Cash Flow model.
The volatility of the real option is hard to estimate when there are no historical data available and there does not exist a twin security. Newton et al. (2004) state that the project under consideration only exists if the R&D is exercised, and therefore no historical or current estimate of volatility is possible. The volatility can be estimated using the judgment of the practitioner, but this is of course very prone to errors. Although a perfect twin security may be hard to find, one could identify a batch of relatively similar securities or previously completed projects to get a good estimate of the potential range of the volatility. Another way is determining the main sources of the volatility and identifying the range of potential outcomes. Once identified low-high scenarios, one can obtain the
Table 1. Input variables Black-Scholes formula financial options and real options
Financial option Real option
Exercise price Expected investment cost to exercise the option
Value underlying asset The present value of the project’s cash flows without flexibility (DCF)
Time to expiration Time to the decision on whether or not to exercise the
option
Volatility of the stock Volatility of the expected cash flows / riskiness of the project
Risk-free rate Time value of money
volatility by calculating it in a statistical computer program. With all these ways of determining volatility, doing a sensitivity analysis will help to understand the impact of small changes in the volatility. The Black-Scholes model assumes that the volatility of the underlying asset is constant during the entire lifetime of the option and that the price of the underlying asset moves in small increments (no jumps in the price), as the value of the underlying asset follows a Geometric Brownian Motion (GBM). In addition, the Black-Scholes model has the following three implicit assumptions:
- You have to be able to buy and sell the underlying asset. - You have to be able to buy and sell the option.
- You have to be able to borrow and lend at the risk-free rate.
These assumptions build on the notion of replicating portfolios that drive option value. They hold for most financial options, but are less obvious for real options. The more you have to loosen these assumptions, the less likely it is that the Black-Scholes option pricing model provides a fair estimate of the value of the option. Therefore, when applying the Black-Scholes model to value a real option, you have to accept that the estimates are less precise in comparison with estimates for financial options, due to the difficulty of arbitrage. When keeping this in mind however, the model can still be applied to real option valuation.
The Black-Scholes option pricing model only applies to European-style options (only exercisable at the expiration date). However, Merton (1973) shows that the differences in the values of European and American options can only be rationally explained when the underlying pays dividends or when there are unfavorable exercise price changes. When considering a real R&D call option, if there is no income from the project during the R&D phase (or pilot phase), the two values will be identical (Newton et al., 2004). Therefore, assuming that the exercise price is fixed, the Black-Scholes option pricing model can also be used for American call options with no dividends. The Black-Scholes model uses a lognormal distribution, which is skewed to the right, to determine option prices. With a lognormal distribution, a variable can only have positive values. This makes sense, since the value of an option can never be negative.
option value, I will not deeply analyze the exact calculations. However, I included the functions and equations of the Black-Scholes option pricing model in Appendix A of this paper.
To make the DCF model and the ROA comparable with each other, one thing has to be clarified. The Black-Scholes model for option valuation implicitly discounts the exercise price (the investment cost in our example) at the risk-free rate, whereas the operating cash flows are being discounted at the WACC. For financial options this makes sense, since the exercise price of a financial option is known for certain. For Real Options Analysis, this is a conservative approach, since the investment costs in this case have a lower discount factor than the operating cash flows. To make the DCF model and the ROA comparable, one could arguably discount the cash flows in both models in the same way. The question which discount rate for the investment cost is most accurate depends on the individual project. I would advise to show graphically how the option value differs across various discount rates for the exercise price. I will show an example of this in section 4. A limitation of the Black-Scholes option pricing model is that it cannot be used for compound options, which basically is an option on an option. Compound options thus have two expiration dates and two exercise prices. The value of the second option impacts the value of the first option. In the case of R&D renewable projects, a compound option could arise when there are several R&D phases distinguishable before the production phase arises. See Figure 2 for a graphical example of a simple option and a compound option in the case of an R&D project.
In this figure, entering a new phase of the project requires an additional investment (the exercise price). The pricing of compound options builds upon the equations introduced by Robert Geske (1979) and Ariel Rubinstein (1991). The calculations involved with compound options are significantly more complex than those of simple options. To keep things understandable and considering that the Real Options Analysis is currently not used at Gasunie, I will in this paper only proceed with simple options.
A second method that is used to quantify real option value is the Binomial Lattice model. This is a tree diagram that shows how the asset value can develop during the lifetime of the option. During each interval in the tree, the asset price can move up by factor u with a probability p. It can also move down by factor d with probability 1-p. A simple example of this is shown in Figure 3.
The Binomial Lattice model is less restrictive in its assumptions than the Black-Scholes model, since the practitioner can alter the inputs at each step in the binomial tree. Therefore, this model can deal with a changing volatility, in contrast to the Black-Scholes model. However, the model can be quite complex and extensive, depending on the number of steps included. When no alterations are made at individual nodes, the model will yield the same outcome as the Black-Scholes option pricing model. Therefore, I will continue with the Black-Black-Scholes option pricing model in the rest of this paper.
For a Real Options Analysis to be applicable and valuable, there are some criteria that need to be met. First, there needs to be a significant amount of uncertainty involved with the project’s outcomes. When there is no uncertainty, managerial flexibility would not be important, as all future optimal decisions can be determined today. Of course, such a situation is not realistic, but the main point I want to stress here is that the value of the real option increases with the uncertainty related with the project. This feels quite counterintuitive, as with most other valuation techniques uncertainty lowers the value of the project (through a higher discount rate). With real options however, an increase in volatility results in an increase in the range of possible future values for the underlying (Allessandri et al., 2004). When uncertainty (and thus the variance of the expected cash flows of the project) is high, having the right but not the obligation to do certain actions is more valuable than implicitly committing to future decisions today. The holder of the option cannot lose more than he paid for the option. The second criterion for the applicability of Real Options Analysis is that the assumed managerial flexibility actually needs to be present. Assuming that managers are flexible in their decision-making during the project when in practice re-evaluations do not take place leads to inaccurate valuations. Therefore, Real Options Analysis requires managers to regularly re-evaluate the options they have regarding the project. When it turns out that the project proceeds without interim evaluations and optionality does not play a role, Real Options Analysis would not be appropriate. In such cases, I would advise to stick to the traditional Discounted Cash Flow model. In addition, practitioners need to have the ability to recognize the most important real options involved with projects. This requires organizations to change the way they look at projects.
calculations and reasoning. Thirdly, there have been published relatively few case studies. Those that were published often were very specific in their examples. Most case studies of the Real Options Analysis focus on the pharmaceutical industry. As a result, the debate about the Real Options Analysis remains rather academic.
2.3.2. Strengths and limitations of the Real Options Analysis
Just as any other valuation technique, Real Options Analysis has its strengths and limitations. This section will briefly discuss those that are vital to acknowledge when proceeding with this method.
The main strength of the Real Options Analysis for the valuation of projects is that it incorporates the concept of managerial flexibility when there is new relevant information. The core aspect of options is that they can be exercised or not (e.g. proceed to the next phase of the project or not). Whereas the traditional DCF model implicitly assumes that there is no managerial flexibility, the Real Options Analysis does incorporate this concept, which presumably is a more realistic approach for many projects. A second advantage of the Real Options Analysis is that it also can be applied in a more qualitative way, when certain data inputs required for estimating the real option value are limited. Recognizing optionality within a project can make practitioners look at the project in a different and more realistic way. According to Triantis and Borison (2001), Real Options Analyses create an important way of thinking for companies, making the method already valuable when used qualitatively. Of course, Real Options Analysis is most valuable when it can be quantified, but using it qualitatively already provides beneficial insights. Next to that, a Real Options Analysis can quite easily be repeated during the lifetime of the project. This is primarily useful when new information arises which changes future expectations.
within a project and focus on those that most likely significantly impact the value of the project. A last and probably most significant limitation of the Real Options Analysis is that the models that are used to model real options are simplified and require some non-realistic assumptions. As a result, the outcomes should be interpreted and not taken for granted. This underlines the idea that ROA is a complement to the DCF model rather than a substitute.
2.4. Other valuation techniques
Next to the Discounted Cash Flow method, the Decision Tree Analysis and the Real Options Analysis, there are some additional valuation techniques worth mentioning before proceeding with the following sections. Since these methods have major drawbacks regarding their implicit assumptions and limitations or simply do not fit for R&D renewable projects, they will not be considered in the rest of this paper.
The second additional valuation technique that requires some attention is valuation using multiples. This approach is based upon financial ratios and assumes that these ratios correspond to the underlying value of the asset. The method implies that the value of a firm can be determined by looking at comparable firms. Multiples then determine differences in the value of firms. When using multiples to value a project or company, one should be aware of using forward estimates instead of past data. Forward estimates incorporate future expectations better than multiples based on past data (Koller et al., 2015). The main challenges when using multiples for the valuation of projects are identifying appropriate comparable companies or projects that are priced similarly to the one being valued and making the necessary adjustments to the financial numbers that are used to measure the market multiples of those comparable companies or projects as well as those used to value the company of interest (Holthausen and Zmijewski, 2012). It might be very hard to find a peer group with similar projects/companies for R&D renewable projects, because of their unique characteristics. Being able to do so is however of critical importance to come up with a reasonable valuation (Koller et al., 2015). Therefore, I consider valuation using multiples as not viable for valuing R&D renewable projects.
3. Decision making process at Gasunie
3.1. Characteristics of R&D renewable projects requiring valuation
Energy are focused on finding solutions related to renewable energy, and specifically renewable gas. The market for renewable gas is relatively new. The future demand for renewable gas is highly uncertain, partly because of regulatory risks, but also due to uncertain developments in other energy sectors which impact the role of renewable gas in the future energy mix. Besides, developments of other renewable energy sources can impact the value of a specific project. This makes it quite hard to estimate future cash flows.
With R&D renewable type of projects, in general two final outcomes can be distinguished beforehand, which are success and failure. Often the project either fails and won’t reach the production phase or it succeeds, in which case the success is large. At Gasunie, one currently generally distinguishes three possible scenarios when evaluation projects: The so-called base case, which projects a scenario that can be considered as most likely to happen and often is in some way a weighted average of a set of possible scenarios, an upside case, which entails the situation in which the project becomes a great success, and a downside case, which embodies the project’s outcome in case of failure. This way of projecting the possible outcomes of a project seems viable for the vast majority of the projects dealt with at Gasunie, but might not be applicable for R&D renewable projects. With these type of projects, it is hard to distinguish a ‘base case’, since the most likely outcome of the project often is failure. The most likely scenario with R&D projects is often not somewhere in the middle between the upside and the downside scenario. This would however not imply that the project should be turned down, as the magnitude of the potential success can justify a small probability of success. This is often the reason why firms proceed with R&D type of projects. In such cases where in reality there essentially only exist two possible outcomes, I would advise the finance department not to make the standard projection of a base case, upside case and a downside case, as this does not reflect the real set of possible outcomes. I would then advise to only sketch the situation in which the R&D renewable project succeeds and the situation in which the project fails. In that case, the finance department prevents that the Board of Directors is presented with a base case that does not do justice to the actual situation.
3.2. Managerial considerations in the decision-making process
probabilities of success and sometimes significant monetary commitments that are required at the start of the project. When the valuation of a project outlines a negative NPV, the rule of thumb is that the project should not be adopted, as there is no expected value creation. However, it turns out that the Board of Directors in such cases not automatically turns down the project and in some cases gives its permission to proceed with the project. In this section, I will outline some important managerial considerations that influence the decision whether or not to go on with an R&D renewable project.
NPV might add value to other projects, making the sum of value added positive. Note that most of the arguments made here to commit to an R&D project do also apply to the decision to proceed with the project when the business case has a negative financial NPV. The magnitude of most of the aforementioned effects is very hard to determine and particularly hard to estimate beforehand. Since it cannot be expected that the finance department is able to estimate all the parameters necessary to estimate such effects, I will not include this in the rest of the paper. These effects should however still be considered when assessing an R&D renewables project and included in the qualitative analysis of the project.
In addition to the aforementioned reasons to proceed with projects with an expected negative Net Present Value (or IRR<WACC), the concept of optionality also plays an important role. As discussed in section 2.1.3, a drawback of the traditional Discounted Cash Flow methods is not to include the concept of managerial flexibility. Managerial flexibility can play an important role when accepting a project with an expected negative NPV, especially for R&D renewable projects. The first stages of scaling up the R&D project, which is when the valuation comes into play at Gasunie, often require a relatively small investment. Management can decide to do the small investment while keeping in mind that they can decide to abandon the project when larger investments are required. In the meantime, the expected cash flows of the project might have improved (based on new information), by which the project might be creating value for the firm after all.
The previous paragraphs showed that there can be various reasons to proceed with a project while the finance department calculated that a project itself does not create financial value, by having potential (real) option value, societal value or environmental value. In the next section, I shows that the concept of managerial flexibility can be incorporated in valuing R&D renewable projects at Gasunie by looking at the Discounted Cash Flow model and the Real Options Analysis as complements.
4. Example project
various techniques are used to value the project and how the differences in the outcomes arise.
4.1. Set-up example project
At Gasunie, most of the data is collected by other departments in the process of developing a business case, whereby the finance department challenges these data. In addition, some market data are collected via the databases of Bloomberg or other (publicly) available databases. It should be noted that when the finance department is asked to value a certain project, there already has been various stages of small-scale developments. When the R&D project reaches the stage of upscaling and a pilot facility has to be built, the finance department makes a valuation. Since there already have been some stages of development before the R&D project reaches the stage of valuation, the uncertainty with respect to the estimation of the variables has decreased compared to earlier stages of the project.
The example project set up for this paper has the characteristics of a typical R&D renewable project at Gasunie. First, I show how the valuation techniques deal with a one-stage project. Later on, I assume that the project has a ‘pilot phase’ and a ‘production phase’. The pilot phase of the project, in which a certain technique will be tested in a large pilot facility, will in this case take about 4 years. Assuming managerial flexibility, the management can decide to either move on to the production phase or stop with the project. The primary aim of this example project is to show the differences between the DCF model and the ROA.
To do a Discounted Cash Flow analysis, one needs to estimate the future cash flows of the project and determine the discount rate. To determine the Weighted Average Cost of Capital (WACC) of the project, which is equal to the discount rate, the following formula is applied:
𝑊𝐴𝐶𝐶 = (𝐸+𝐷𝐸 ) ∗ 𝐶𝑜𝐸 + (1 − 𝑇𝑎𝑥) ∗ (𝐸+𝐷𝐷 ) ∗ 𝐶𝑜𝐷 (7)
E market value of equity
D market value of debt
CoE cost of equity CoD cost of debt
There already has been extensive research at Gasunie aiming to determine the WACC of a so-called ‘green’ project. Therefore, for this example I will simply pick a rough average of the WACCs of renewable projects, which is approximately 8%. The WACC reflects the riskiness of the project and should thus relate to the volatility of the expected cash flows.
4.2. One-phase example project
Next, one should determine the expected cash flows of the project. First, I assume that there is only one phase within the project. The project has an initial investment cost of €13 000 000, after which the project will generate €1 000 000 annually in the first four years. From year five onwards the project generates €50 000 more, after which the cash flows remain constant. The project has an infinite lifetime. With respect to the growth rate, Gasunie states in their Economic and Financial Guidelines that this is normally limited to the partial inflation correction of the revenues, since in the gas infrastructure business rapid (market) growth of the revenues is normally covered in the project period and after that a steady state growth is reached. However, with R&D renewables projects this is slightly more complex. For simplicity in the example project, I assume that the project generates each year the same cash flows and I will therefore not include a growth rate. The expected cash flows of the project are displayed in Table 2.
To calculate the NPV using the traditional DCF model, one should simply discount all future cash flows with the Weighted Average Cost of Capital (WACC) and subtract the investment costs. The present value of all future cash flows is equal to €12 959 394. The NPV of the project then becomes: €12 959 394 − €13 000 000 = −€40 606. Valuing the project using the Real Options Analysis yields the same
Table 2. Expected cash flows of the example project with one phase
Year 0 1 2 3 4 5 6 7 TV
Cash flow (13) 1 1 1 1 1.05 1.05 1.05 13.125
value, as there is only one decision to make and I implicitly assumed it has to be made today. The parameters for the ROA are the following:
Exercise price €13 000 000
Value underlying asset €12 959 394
Time to expiration 0
Volatility of future cash flows No volatility
Time value of money 8%
The time value of money is in this case equal to the WACC. There is no future investment that has to be discounted, so in this case there is no discussion on using various discount rates. The time to expiration is zero, as the decision has to be made today. When the time to expiration is zero, the volatility of future cash flows does not matter, since all relevant decisions are made today. Putting the parameters above in the Black-Scholes formula results in a call option value of €0 and a put option value of €40 606, which corresponds to the NPV calculated before with the Discounted Cash Flow model. Consequently, both valuation methods report that this project does not create value and thus should not be adopted.
Now I shortly want to highlight the option to delay, which sometimes may be present with R&D renewables projects. Consider that the firm has an option (the exclusive right) to do this project not today, but two years from now. For simplicity, I will assume that the expected cash flows and the WACC do not change. One should acknowledge that this may not be true in practice, for example due to the existence of competitors and changing market conditions. Table 3 shows the expected cash flows of the project considering the real delay option of two years.
Table 3. Expected cash flows of the example project with one phase and a two-year real delay option
Year 0 1 2 3 4 5 6 7 8 9 TV
Cash flow (13) 1 1 1 1 1.05 1.05 1.05 13.125
With the traditional DCF model, one would simply discount the future cash flows and the investment costs for two additional years. The NPV of the project then becomes:
€12 959 394
1.082 +
−€13 000 000
1.082 = −€34 813
With the DCF model, you assume that you have to decide today whether or not you are adopting a project in two years. The Real Options Analysis however considers that one has the right to adopt the project in two years, but not the obligation to do so. The manager might also decide not to adopt the project in two years. This flexibility has some value that is not considered with standard DCF calculations. The parameters for the Black-Scholes option pricing model in this case become:
Exercise price €13 000 000
Value underlying asset €11 110 591 (=12 959 394 1.082 )
Time to expiration 2
Volatility of future cash flows 30%
Time value of money 8%
In this case, the volatility of the future cash flows does matter for the value of the option. So far, we have not estimated the volatility yet. I assume that the firm has determined low-high scenarios, or identified a batch of relatively similar securities or previously completed projects, and calculated a volatility of 30%2 (annually).
When one is significantly uncertain about low-high scenarios and the associated volatility, I would advise to use a conservative approach and be aware not to overstate volatility. Putting the variables above in the Black-Scholes option pricing model, we get a call option value of €1 880 191. When using a risk-free interest rate of 0.5% to discount the exercise price (referring to the discussion in section 2.3.1, p. 18), the call option value becomes €1 252 095. Which discount rate is appropriate depends on the degree of uncertainty about the investment costs. So having the exclusive right to do this project within two years has a value somewhere between €1 252 095 and €1 880 191 (when the volatility is equal to
2 A volatility of 30% indicates that there is a 68,27% probability that the true value of the
30%), which is significantly higher than the value calculated with the DCF model. Note that this value is the value of having the option to do the project within 2 years and not the value of actually doing the project within two years. The option provides not an obligation, but a right. When solely applying the DCF model, one would logically not adopt the project. However, having the exclusive right to adopt the project in two years yields a significant positive value. Upward risk is captured in the option value and the value of downward risk is zero, because it can be avoided by simply not exercising the option. Note here that the value of the call option can never be negative, as it is equal to 𝐶 = 𝑀𝑎𝑥(𝑆 − 𝑋, 0). The value of this option largely depends on the volatility. Figure 4 displays the value of the option to delay as a function of the volatility of the expected cash flows.
mechanics of the Real Options Analysis and may help managers in the decision-making. However, one should bear in mind that the cost of capital used to discount the underlying asset correlates with the expected volatility of the underlying asset. A project with a relatively high expected volatility in its cash flows is likely to have a higher cost of capital than a project with a relatively low expected volatility in its cash flows.
4.3. Two-phase example project
Now, consider a project that has two phases, with one initial investment (for the pilot phase) and an additional investment to scale up the project to production level. The expected cash flows of the project are displayed in Table 4.
The NPV of the pilot phase is calculated using the traditional DCF model, as I assume that the uncertainty relatively this phase is relatively low and the investment decision for this phase has to be made today. The NPV of this phase is equal to –€3 343 937. When applying the traditional DCF model also to the production phase, with a WACC of 8% and a growth rate of 0%, this phase yields an NPV of €3 037 808. The total NPV of the project then becomes:
𝑁𝑃𝑉 𝑝𝑖𝑙𝑜𝑡 𝑝ℎ𝑎𝑠𝑒 + 𝑁𝑃𝑉 𝑝𝑟𝑜𝑑𝑢𝑐𝑡𝑖𝑜𝑛 𝑝ℎ𝑎𝑠𝑒 = 𝑁𝑃𝑉 𝑝𝑟𝑜𝑗𝑒𝑐𝑡 −€3 343 937 + €3 037 808 = −€306 129
Applying a Real Options Analysis requires understanding of what the actual option is. As outlined before in section 2.3.1, the option considered here is a growth
Table 4. Expected cash flows of the example project with two phases
Year 0 1 2 3 4 5 6 7 8 9 TV
Pilot phase (5) 0.5 0.5 0.5 0.5
Production
phase (11) 1.2 1.2 1.2 1.2 1.2 15
option. The firm has an option to scale up the project after four years, which requires an additional investment of €11 000 000. However, the firm has not the obligation to do so. The firm might also decide to stop the project after the first four years and not proceed to the production phase, thereby limiting the potential losses. The inputs of the Black-Scholes option pricing model are the following:
Exercise price €11 000 000
Value underlying asset €11 025 448
Time to expiration 4
Volatility of future cash flows 30%
Time value of money 8%
Putting these inputs in the Black-Scholes formula (and thus discounting the exercise price with the WACC), we get a call option value of €4 064 031. Adding this to the NPV of the pilot phase, we get a total NPV of:
𝑁𝑃𝑉 𝑝𝑖𝑙𝑜𝑡 𝑝ℎ𝑎𝑠𝑒 + 𝐶𝑎𝑙𝑙 𝑜𝑝𝑡𝑖𝑜𝑛 𝑜𝑛 𝑝𝑟𝑜𝑑𝑢𝑐𝑡𝑖𝑜𝑛 𝑝ℎ𝑎𝑠𝑒 = 𝑁𝑃𝑉 𝑝𝑟𝑜𝑗𝑒𝑐𝑡 −€3 343 937 + €4 064 031 = €720 094
Therefore, such a figure is a useful tool for the management when complex decisions have to be made.
This section showed with a valuation example what the differences are between the traditional Discounted Cash Flow model, which is the one currently used at Gasunie for all valuations, and one with an add-on of the Real Options Analysis.
5. Valuation guidelines
So far I have stressed the situation at Gasunie and outlined the set of possible valuation techniques with their strengths and limitations. In this section, I will provide practitioners with some guidelines on how to choose the best valuation method for their R&D renewable projects.
5.1. Theoretical valuation framework
When uncertainty about future cash flows is low, the Discounted Cash Flow model is probably the best applicable method. It provides significant benefits in terms of understandability and communicability with the management and other parties. Next to that, its lacking of incorporating option value and managerial flexibility is less important when the outcomes of the project are relatively certain. Building a Real Options framework in such cases would not provide major benefits. Managerial flexibility does not matter for the valuation when the uncertainty about the future cash flows is relatively low. However, as mentioned several times before in this paper, with R&D renewable projects the uncertainty related with potential future cash flows is often high.
With R&D renewable projects, the uncertainty often increases with the time. The cash flows of the first stages can often be estimated quite accurately. Following Mun (2006), I would therefore suggest to use the Discounted Cash Flow model up to the point where uncertainty significantly rises. After this point, using Real Options Analysis would probably be more appropriate. Menegaki (2008) states that the use of risk adjusted discount rates (with a DCF approach) and options theory are complementary approaches and are an improvement over the traditional present value approaches. Therefore, practitioners should determine up to which point they think that a DCF would be sufficient and after which ROA would be more applicable. With respect to R&D renewable projects, I would in general advise to use a Real Options Analysis for the ‘production phase’, when new investments to scale up the project are required. At this point, new information with respect to the potential success of the project is available. To do a Real Options Analysis, one should start with a standard DCF analysis, which then becomes the base of the valuation (referring to the Market Asset Disclaimer). There often consist many sources of uncertainty for this phase, making a Real Options Analysis that allows for learning based on new information and adaptive behavior most appropriate.
5.2. Practical implementation of the valuation framework
reluctant to abandon projects and tend to proceed with the project even though valuations do not support this, which is a behavioral issue. However, ROA implies that the option (e.g. scaling up the project) should be only exercised when the value of the underlying asset is larger than the exercise price. This means that the progress of the R&D renewable project should be evaluated on a regular basis or at predetermined points in time (e.g. just before the expiration of the real option). Such interim evaluations of R&D projects do not only allow for managerial flexibility. They also help to assess if past decisions and assumptions were correct, which can improve the decision-making process for future projects. In addition, introducing post-completion auditing (PCA) for the pilot phase is useful to improve the valuation process and the resulting information may also help to implement the Real Options Analysis in a more accurate way. According to Brantjes et al. (1999), with a PCA the quality of the investment system and future investment decisions can be improved. Therefore, I would advise the finance department to request future research on how this can be implemented at Gasunie.
A Real Options Analysis can also be done during the lifetime of the option. It is possible that during the pilot phase of the project it turns out that, based on new information, the expected costs are too high or expected cash flows are too low. In such cases one might want to reevaluate the business case and assess whether it is still worthwhile to proceed with the project. Ideally, I would recommend to do a Real Options Analysis at every point in time when there is some news that potentially affects the value of the option. Every kind of news that affects the expectations about future cash flows impacts the value of the real option and thereby the value of the project. Some examples of news that affects the option value are:
- The pilot phase requires additional investments. - A competitor has developed a better product.
- The demand for the product has significantly increased/decreased.
- The exercise price of the option (investment costs 2nd phase) has increased
or decreased.
regular evaluation (e.g. yearly or at the middle of the pilot phase) could already be helpful.
A Real Options Analysis is primarily useful when traditional DCF calculations do not provide convincing results and consequently do not provide a clear advice for the investment decision. Considering optionality in such cases might convince management to adopt the project or not. In cases where DCF calculations provide very convincing results, a ROA would presumably not change the valuation such that the advice of the finance department would be different.
6. Conclusion and recommendations
In this paper, I have outlined a set of valuation techniques that can be applied for valuing R&D renewables projects. I explained that traditional Discounted Cash Flow models are sufficient when uncertainty is low. However, with significant uncertainty, the value of having managerial flexibility increases and DCF models provide no longer accurate valuations due to inaccurate implied assumptions. In such cases, using a Real Options Analysis complementary to the Discounted Cash Flow model is preferred. Including real option value for simple options can be done using the Black-Scholes option pricing model. When using this model, practitioners should be aware to accurately determine the required inputs. For compound options, the Black-Scholes option pricing model cannot be used. In such cases, more complex option pricing models are required, such as the Geske model. I would however advice Gasunie to initially limit Real Options Analyses to simple options, as compound options are far more complex and generally less understandable for people with a non-finance background. Finally, using a Real Options Analysis requires to regard the determined real options as not obligatory investments, but depending on managerial choices resulting from better (more recent) information.
static way most firms look at projects by recognizing the concept of learning and adaptive behavior. For Gasunie specifically, I would recommend to request further research on the applicability of post-completion auditing during the pilot phase of R&D renewable projects.
7. Reference list
Allessandri, T., Ford, D., Lander, D., Leggio, K., Taylor, M., 2004. Managing risk and uncertainty in complex capital projects. The Quarterly Review of Economics and Finance 44, 751-767.
Amram, M., Kulatilaka, N., 2000. Stategy and shareholder value creation: the real options frontier. Journal of Applied Corporate Finance 18(2), 40-47.
Black, F., Scholes, M., 1973. The pricing of options and corporate liabilities. The Journal of Political Economy 81(3), 637-654.
Brantjes, M., von Eije, J., Lemkert, J., Pennink, B., 1999. Post-completion auditing within Heineken. Management Accounting 4, 20-22.
Copeland, T., Antikarov, V., 2001. Real Options: A Practitioner’s Guide. Texere Publishing, New York.
Dixit, A., Pindyck, R., 1994. Investment under Uncertainty. Princeton University Press, New Jersey.
Energy technology RD&D budgets: Overview, 2017. International Energy Agency. Retrieved from http://www.iea.org/statistics/rdd/.
Falkner, R., 2016. The Paris Agreement and the new logic of international climate politics. International Affairs 92, 1107-1125.
Fama, E., French, K., 2004. The capital asset pricing model: theory and evidence. Journal of Economic Perspectives 18(3), 25-46.
Holthausen, R., Zmijewski, M., 2012. Valuation with market multiples: how to avoid pitfalls when identifying and using comparable companies. Journal of Applied Corporate Finance 24(3), 26-38.
Koller, T., Goedhart, M., Wessels, D., 2015. Valuation: measuring and managing the value of firms. Wiley, New Jersey.
Lefley, F., 1996. The payback period of investment appraisal: a review and synthesis. Int J. Production Economics 44, 207-224.
Luehrman, T., 1998. Investment opportunities as real Options: getting started on the numbers. Harvard Business Review 76(4), 51-66.
Makropoulou, V., 2011. Decision tree analysis and real options: a reconciliation. Managerial and Decision Economics 32(4), 261-264.
Menegaki, A., 2008. Valuation for renewable energy: a comparative review. Renewable and Sustainable Energy Reviews 12, 2422-2437.
Merton, R., 1973. Theory of rational option pricing. The Bell Journal of Economics and Management Science 4(1), 141-183.
Mun, J., 2006. Real options and monte carlo simulation versus traditional DCF valuation in layman’s terms. In: Leggio, K., Bodde, D., Taylor, M. (Ed.), Managing enterprise risk: what the electric industry experience implies for contemporary business. North Holland, Amsterdam, pp. 75-100.
Myers, S., 1977. Determinants of corporate borrowing. Journal of Financial Economics 5, 147-175.
Newton, P., Paxson, D., Widdicks, M., 2004. Real R&D options. International Journal of Management Reviews 5/6, 113-130.
Ross, S., 1976. The arbitrage theory of capital asset pricing. Journal of Economic Theory 13, 341-360.
Schneider, M., Tejeda, M., Dondi, G., Herzog, F., Keel, S., Geering, H., 2008. Making real options work for practitioners: a generic model for valuing R&D projects. R&D Management 38(1), 85-106.
