Abstract (eng)
Investing in the earths ecosystem, as it is the fundament of our existence, is an absolutely necessary undertaking, not at least because of the substantiated damage from which it already suffered as a result of human activities. As the future is uncertain, and environmental harmful economic actions are partly irreversible, a precise valuation of ecological investment becomes more and more important, and part of scientific economic discussion. The traditional Net-Present Value rule, as pointed out clearly in this thesis, is not suitable enough to account for uncertainty, irreversibility and flexibility. Thus, the real option approach is used to evaluate and determine the optimal timing as well as the optimal amount of an emission reducing policy carried out by a policy maker. This approach, which refers to an option value comparable to the financial call or put option, allows giving future possibilities of investment a value which is not considered under the NPV-rule.
The mathematical background used in the main part of this thesis, chapter 3, is presented in chapter 2, as dynamic programming and certain types of Brownian motion, especially the Ito-process. The latter allows to model stochastic trends which face uncertainties and irreversibilities. Dynamic programming is used to break the whole future which is regarded in two sequences, the present decision and a second one which encapsulates all future consequences and possibilities.
Chapter 3 describes the model by Pindyck (2000): “Irreversibilities And The Timing of Environmental Policy” in detail. First of all, the analytical framework is presented, which is then extended to analyse the implications of ecological and economic uncertainty, irreversibility and the possibility of delaying the investment decision, as well as the possibility of reducing emissions at once or gradually.
Economic uncertainty together with a once and for all reduction in emissions is modelled in chapter 3.1. Here, the greater the uncertainty over future social cost of pollution, the greater the incentive to wait rather than adopt the policy immediately. The same incentive to wait is given if the discount rate r increases. Under ceteris paribus conditions, a greater current cost of pollution makes immediate policy adoption favourable, which the traditional NPV-rule takes into account as well. The timing of an emission reducing policy is also affected by irreversibility. In this chapter, the higher the natural rate at which pollution depreciates, which means that pollution gets more reversible, the smaller the sunken benefit of adopting the policy now rather than waiting, and the policymaker tends to delay the adoption. Furthermore, the timing of policy adoption does not depend on the initial level of pollution, but the value of society’s option to adopt the policy increases linearly with the initial level of pollution. Numerical as well as graphical solution can be found at the end of chapter 3.2.
In chapter 3.2.1 economic uncertainty, a convex cost function, and the possibility of partial emission-reductions are assumed. As in chapter 3.2, the more uncertainty, the later an emission reducing policy is adopted. But now, uncertainty also determines the amount by which emissions are reduced. The higher uncertainty, the lower the emission reduction, whereas an increase in the convex cost function increases the amount by which emissions are reduced.
Next, a convex benefit function together with economic uncertainty is modelled in chapter 3.2.2, in contrast to before, where the benefit function was linearly correlated with the stock of pollutant. The higher this stock, the earlier a policy gets adopted, because a higher stock of pollution implies a higher marginal cost of additional emissions. Besides, a higher emission level, a higher cost of emission reduction and a higher decay rate – more reversibility, lead to later adoption.
Gradual emission reduction is then assumed in chapter 3.3. Here, a policy maker faces the possibility to reduce emissions gradually and continuously. Thus, the optimal timing and the optimal amount of emission reduction has to be determined. The cost function is assumed to be convex, and the benefit function is assumed to be linear. Uncertainty then affects not only the emission reduction over time, but also the initial reduction.
From chapter 3.4 onwards, ecological uncertainty is assumed but no economic uncertainty. After a short overview of the new analytical framework, in chapter 3.2.1 complete reversibility is assumed in order to find an analytical solution. As ecological uncertainty is assumed to control the stock of pollutant stochastically, the future emission rates are known but still the evolution of the ecosystem is uncertain. This uncertainty, analogous to the economic uncertainty, delays policy adoption. A numerical example shows that the greater uncertainty, the later an emission reducing policy is adopted.
A more general case with ecological uncertainty is modelled in chapter 3.4.2, where environmental damage is partly irreversible. A solution is found only numerically in this case. The higher uncertainty over the evolution of the stock of pollutant, the later a policy is adopted. The same delay is true for lower irreversibility. Thus, a higher pollution decay rate implies that the stock of pollution faces a lower drift rate, and as a consequence the present value of the flow of social cost for any current value of the pollution stock is smaller. Therefore, to compensate for the sunk cost of policy adoption, a higher stock of pollutant is needed to trigger policy implementation.
Summarizing chapter 3, there is a possibility to delay policy adoption, called flexibility, and if the costs for a policy adoption are assumed to be sunk, then immediate emission reduction imposes an opportunity cost on society. However, there is also an opportunity “benefit” of early adoption because the stock of pollutant gets reduced, which otherwise would impose a nearly irreversible cost on society. And the higher uncertainty of the future costs and benefits of reduced emissions, i.e. economic uncertainty, or of the evolution of the stock of pollutant, i.e. ecological uncertainty, is assumed, the more the policy adoption gets delayed in order to gain more information. This is true for once and for all reduction as well as for gradual emission reductions, and in both cases the delay gets reduced the greater irreversibility.
This basic framework is then extended to two directions in order to analyse more precisely the impacts of the discount rate on the one hand and of the possibility of extreme events on the other hand. This is done in chapter 4.
First, a model by Di Vita (2003) “Is the Discount Rate Relevant in Explaining the Environmental Kuznets Curve?” is presented and discussed. The matter of investigation is the question how the interest rate affects environmental policies. If the discount rate is high due to economic growth as found in developed countries, the income-pollution pattern states that environmental policies are more likely to be adopted than in developing countries which face high interest rates. Thus, the discount rate and income move in opposite directions, such as the discount rate and the willingness to adopt emission reducing policies. Furthermore, the coherence between economic growth and pollution depends on the discount rate in the following way: countries with a low level of income and high discount rates show a positive relationship between economic growth and pollution, whereas developed countries illustrate the opposite. Both, a once-and-for all reduction and a gradual emission reduction were modelled, although the main findings were the same.
In chapter 4.2 extreme variations in pollution stock levels and in the socioeconomic costs were assumed. Again, the optimal timing and the optimal amount of an emission reducing policy were determined, but now accounting for the possibility of sudden jumps in the emission level and in pollutant-related socio-economic costs. Economic and ecological uncertainties were not modelled together. First, in chapter 4.2.1 sudden jumps were assumed in the stock of pollution. Then, in chapter 4.2.2 large and unexpected changes in the future flow of social cost were regarded. In both cases, the possibility of sudden unpredictable changes lead to earlier policy adoption.
This thesis gives on overview of ecological investment theory, especially if uncertainty, irreversibility and flexibility are concerned. The real option approach is explained and two different extensions of Pindyck’s basic evaluation method were discussed. Needless to say, more research on this fields has to be done, and actually is done, in order to minimize the extremely high costs which may result from environmentally harmful production techniques.