> L%!246F
H
J
,.0246
FH,.
&F:$d$
&F d$
&F.d asset (the underlying)BuySell4.5
4.8 4.9 Carrying t4.9.1 4.9.1 4.104.11
31
DALARNA UNIVERSITY BACHELORS THESIS
DEPARTMENT OF ECONOMICS SUMMER 2001
THE OPTION TO DEFER
MODELING THE OPPORTUNITY COST UNDER COMPETITION
AUTHOR: EXAMINATOR:
HKAN JANKENSGRD PROFESSOR LARS HULTKRANTZ
TOC \o "13" 1 Introduction PAGEREF _Toc521815160 \h 4
1.1 Background PAGEREF _Toc521815161 \h 4
1.2 Problem Statement PAGEREF _Toc521815162 \h 4
1.3 Purpose PAGEREF _Toc521815163 \h 6
2 Method PAGEREF _Toc521815164 \h 7
2.1 Game Theory for Modeling Competition PAGEREF _Toc521815165 \h 7
2.2 Options Pricing as Opposed to Options Analysis PAGEREF _Toc521815166 \h 8
2.3 The Choice of Literature in This Essay PAGEREF _Toc521815167 \h 8
2.4 On the Level of Mathematics PAGEREF _Toc521815168 \h 9
3 Game Theory PAGEREF _Toc521815169 \h 11
3.1 The Basics PAGEREF _Toc521815170 \h 11
3.2 Solution concepts PAGEREF _Toc521815171 \h 12
3.2.1 Dominant Strategy Solution and IEDS Solution PAGEREF _Toc521815172 \h 12
3.2.2 Nash Equilibrium PAGEREF _Toc521815173 \h 14
3.3 Information and Uncertainty in Game Theory PAGEREF _Toc521815174 \h 14
3.4 Games of Perfect Information PAGEREF _Toc521815175 \h 15
3.5 Games of Imperfect Information PAGEREF _Toc521815176 \h 16
3.6 Games of Incomplete Information PAGEREF _Toc521815177 \h 17
3.6.1 Introduction PAGEREF _Toc521815178 \h 17
3.6.2 Bayes Nash Equilibrium PAGEREF _Toc521815179 \h 18
4 Option Theory PAGEREF _Toc521815180 \h 19
4.1 Some Notation PAGEREF _Toc521815181 \h 19
4.2 The Arbitrage Principle PAGEREF _Toc521815182 \h 19
4.3 Options: Some Definitions PAGEREF _Toc521815183 \h 21
4.4 Option Pricing by Creating a Risk Free Hedge PAGEREF _Toc521815184 \h 21
4.4 Option Pricing by Assuming Risk Neutral Investors PAGEREF _Toc521815185 \h 23
4.5 The Stochastic Properties of the Underlying PAGEREF _Toc521815186 \h 24
4.6 Modeling Dividends in the Binomial Tree PAGEREF _Toc521815187 \h 25
4.7 Real Options: The Analogy PAGEREF _Toc521815188 \h 26
4.8 Carrying The Analogy Over PAGEREF _Toc521815189 \h 27
4.4.1 Using a Twin Security PAGEREF _Toc521815190 \h 27
4.4.2 The Marketed Asset Disclaimer PAGEREF _Toc521815191 \h 29
4.9 The Option to Defer PAGEREF _Toc521815192 \h 30
4.10 Problems with the Analogy PAGEREF _Toc521815193 \h 32
5 Real Option Valuation under Competition PAGEREF _Toc521815194 \h 33
5.1 On Economic Rents and Competition PAGEREF _Toc521815195 \h 33
5.2 Investment Tactics under Competition and Uncertainty PAGEREF _Toc521815196 \h 34
5.3 Analyzing First Mover Advantage PAGEREF _Toc521815197 \h 35
5.3.1 The General Principles PAGEREF _Toc521815198 \h 35
5.3.2 Sources of first mover advantage PAGEREF _Toc521815199 \h 37
5.3.3 An Example PAGEREF _Toc521815200 \h 38
5.3.4 Comments on the Example PAGEREF _Toc521815201 \h 42
5.4 The Opportunity Cost in Real Options Analysis PAGEREF _Toc521815202 \h 42
5.5 The Opportunity Cost as a Function of Incomplete Information PAGEREF _Toc521815203 \h 45
5.5.1 Introduction PAGEREF _Toc521815204 \h 45
5.5.2 A Game Theoretic Framework PAGEREF _Toc521815205 \h 45
5.5.3 The impact of imperfect information PAGEREF _Toc521815206 \h 46
5.6 Valuing the Option to Defer under Strategic Uncertainty PAGEREF _Toc521815207 \h 50
5.6.1 A description of the approach PAGEREF _Toc521815208 \h 50
5.6.2 A oneperiod binomial model PAGEREF _Toc521815209 \h 51
5.7 Some Thoughts on the Usefulness of the Approach PAGEREF _Toc521815210 \h 54
6 SUMMARY PAGEREF _Toc521815211 \h 56
1 Introduction
Background
We live in a world of changes. Things change, a statement that is as true as ever in the corporate world. It is a simple statement, yet its implications are truly profound for the reason that we cannot perfectly foresee change. Uncertainty about the future is an everpresent aspect of decisionmaking.
Some see the enormous growth of the derivatives markets in the past 20 years as a sign of our times. Uncertainty is what drives the value of derivatives, and, the argument goes, the thriving of derivatives markets is telling us that the world is a highly volatile and uncertain place, where often little is known about the future.
One implication of the fact that the future is uncertain is that information has value. Knowing more gives us a better shot at making the right decision. One way to gain more information, is to wait and let time reveal more about some variables that are important to our decision. There is also a strategic aspect: Having information that no one else has access to gives a decisionmaker an edge over less informed individuals.
Another implication is that flexibility has value. If we are, for some reason, able to change our course of action at some point down the road, a bad decision may not be a terrible one, and a good one may turn out to be a great one. Flexibility means that it is possible for us to make or change decisions in the future as new information becomes available, when the source of uncertainty reveals more about itself.
Problem Statement
The issues of uncertainty, information and flexibility are evidently important when we wish to evaluate the decision of whether to invest or not. We are interested in assessing the value of an investment opportunity in order to make the right decision. The value of the investment will be a function of the flexibility inherent in the project or asset we are about to invest in, the information we have at the time of the valuation, and the uncertainty about key variables.
One way to evaluate capital investments is the socalled NPVmethod, which has us making an estimate of the cash flows we expect to get were we to go ahead with the project, and discount them to adjust for the time value of money and risk. The basic rule says that if the NPV is a positive number, the investment should be given the goahead. For a number of reasons this approach has fallen from grace, the main one being that it fails to incorporate flexibility on part of the decisionmaker, therefore falling short of being an accurate model of reality.
To remedy this shortcoming, academics realized that projects could be seen as opportunities, but not obligations, to invest. Based on this analogy they turned to financial options theory and began incorporating the techniques for valuing financial options into corporate capital budgeting, which has led to the establishment of real options theory.
The analogy has been very successful and is currently attracting a huge amount of interest from researchers as well as practitioners. The consensus is growing, however, that the analogy is lacking in one vital respect, namely that competition affects the value of real options in a way that has no natural equivalence in the case of financial options.
The bias in real options theory has been towards situations where a) the firm has a monopoly over an investment situation and b) the product market is perfectly competitive. In a monopoly there is, by definition, no strategic considerations, and perfectly competitive product markets means that the investment affects neither price nor market structure. In the real world, these conditions seem to be the exception rather than the rule.
The attention is turning to game theory, which appears tailormade for analyzing competition after all, it is a theory about strategic interaction. In game theory, for its part, issues of preemption and strategic advantages through, for example, cost effectiveness, has been an intensively researched field of study for many years. But little research has been carried out with regard to how exogenous uncertainty impacts these strategic considerations.
Thus the search is on for ways to incorporate competition into real options analysis. But exactly how should this be done? At a first glance, these theories do not seem to have much in common. And they both seem to be overwhelmingly mathematical, so the mere thought of integrating them can clearly be intimidating. Yet, the integration should preferably be done without sacrificing too much of the underlying intuition that capital investment opportunities can be seen as options.
Purpose
In this essay I propose a way to model the opportunity cost of waiting using game theory that can be integrated into the binomial model for options pricing. My hypothesis is that the opportunity cost can be derived from seeing the capital investment as a game of incomplete information, and that this cost can in turn be modeled in much the same way dividends are modeled in options pricing.
Who is this essay intended for?
This essay is directed to economics students at Bachelors or Masters Level, especially those with an interest in the theory and application of real options theory.
All the necessary concepts and definitions from both theories are included in the essay to make it, in some respect, selfcontained. Nevertheless, it will be very helpful to have been exposed them prior to reading the essay because they are rather abstract in nature and one doesnt get a good grasp of their workings from simply reading through the definitions provided here.
Of course, providing the definitions and basic theory is of little value to those readers who are already involved in the subject, and what I suggest is that they read only chapter 5.
Method
Game Theory for Modeling Competition
It is recognized that what real option theory is lacking is a way to acknowledge the presence of competition and how this influences the value of a real option. The way I see it, there are three available options for modeling the competitive aspect of capital investment.
Corporate Strategy
Microeconomic Theory
Game Theory
It is my understanding that corporate strategyliterature does not have a reasonably objective framework for modeling an investment situation. There is a profusion of texts available on how to gain an edge over competitors, but they dont deal with predictions about how a specific investment situation might unfold. So obviously, this would be a dead end.
Traditional microeconomic theory does indeed offer a framework for modeling different types of markets and investment decisions. But historically microeconomic theory has focused its attention on two extreme types of markets, namely monopoly and perfect competition. Here, strategic interactions play no part of the analysis. In monopoly by definition, and in perfect competition because it is unrealistic to assume that a player should be able to keep track of all its competitors.
Game theory, on the other hand, it a theory that has been developed precisely to deal with strategic interactions. In a situation where strategically aware players interact, and the circumstances are well enough understood, game theory can be used to predict the outcome of the game in question. What is more, game theory has solutions concepts that are very appealing because they allow for a neutral, logicbased way to derive a prediction of how the game will be played.
Given these facts it is clear that the most promising route is offered by game theory, which is why this essay will try to analyze how it can be integrated with real options theory. It should be said, however, that the distinction between Microeconomic theory and game theory is probably not relevant today, because game theory is actually a subfield of Microeconomics.
Options Pricing as Opposed to Options Analysis
There is a distinction between options analysis and actual options pricing. Options pricing is a very exact science, usually highly mathematical, aimed at assessing a given investments true value by way of an argument that says there are no arbitrage opportunities in financial markets.
Options analysis is can also be carried out without the link to financial markets. Game theory, for example, deals with entry as well as exit options, and it is possible to analyze quite complex options within its framework. Decision tree analysis is another tool for analyzing options, but it fails to give an accurate valuation of the project due to incorrect assumptions about the discount rate. This has to do with the fact that it assumes a constant discount rate under all scenarios in the tree, even though the risk of the project changes as the value of the underlying project changes. The higher the levels of cash flows that the project is throwing off, the less risky it will be as an investment.
The main difference between options pricing and options analysis, in my interpretation, is that the former has a strong link to financial markets and the arbitrage argument. It is a neutral, marketbased way to value a derivative. In this essay I will try to keep as strictly as possible to option pricing. Although there may be very few real world situations where such a model would apply, I think it has the benefit of necessitating strict model building that will clarify the principles at work. Also, thanks to the Marketed Asset Disclaimer approach, it is possible to value project even if there is no apparent link between the project and a traded security.
The Choice of Literature in This Essay
This essay has a special interest in discussing games of incomplete information. The material in the chapter on game theory has been chosen accordingly, only presenting bits of the theory that are relevant to these particular applications. These include backward induction, games of perfect information and games of imperfect information.
Also, the game theoretic solution concepts  Nash equilibrium, Dominant Strategy, and IEDS  will be given some attention. This is because they are based on impeccable logic and selfinterested behavior. This is a plus since we are trying to connect game theory with the very precise science of options pricing. It is preferable not to get too subjective, because then the valuation arrived at will be more questionable.
As for options theory, I will focus entirely on the binomial pricing model rather than Black & Scholes. The reason is that the binomial model is much more intuitive, which is a stated goal of this essay. It also shares the backward induction principle with game theory and allows us to see very clearly what it is we are doing at every step.
Given the reliance on arbitrage arguments, some sections will be devoted to describing these principles. Also, since the hypothesis is to model competition as an opportunity cost, I will go to some length to describe how dividends are treated in the binomial model.
The option that will receive almost all attention is the option to defer. Therefore one section will be devoted entirely to describing the principles behind this type of investment situation.
The theorychapters are thus not intended as standalone introductions to either theory. It is assumed that the reader is already familiar with these theories and will therefore only describe the parts of them relevant to testing the hypothesis.
On the Level of Mathematics
Ito Calculus, a high level subfield of mathematics, dominates the field of options pricing. Game Theory, in its purest form, also makes use of highly abstract mathematics. An integration of the two would seem to suggest that there is going to be some pretty heavy mathematics involved.
Nonetheless, you will find very little of that in this essay. The main reason is that the essay is intended simply to map out an area that has not been fully explored. Not much has been written on the subject and even less, of course, in a way that stresses intuition and logical reasoning over hardcore mathematics and microeconomic models. (For a dose of the latter, consult the book Game Choices).
The benefit of mathematics is that it is a wonderful way to prove things. If a proposition can be proved mathematically, it will carry an awful lot more weight than just a simple claim that thisisso. I have tried to provide some type of validation for some of the propositions in the essay, but they are wordy as opposed to strict. This leaves one unsure if these are general principles or just something that seemed right given the lines of reasoning I had in my head at the time of writing.
Either way, the essay should be viewed as an attempt to explore a relatively new area of research and consolidate some of the principles at work. As of today, the integration of the two is probably long ways away from having any influence in industry. But for those interested, this essay could be an introduction to some of the insights that the merging of these theories hold the potential to produce.
Game Theory
The Basics
Game theory is the study of rational choices by players whose payoffs depend on each others actions and who recognize this fact. Strategic thinking has been described as the art of outdoing an adversary, knowing that the adversary is trying to do the same to you.
For a game to be analyzed we need to specify the rules of a game. The rules of a game answer the questions:
Who is playing?
What are they playing with?
When does each player get to move?
How much do they stand to gain (or loose)?
Here will follow a number of definitions that in closer detail describes the nature of the rules of the game.
Definition 3.1 Players
The players are the individuals who make decisions. Each players goal is to maximize his utility by choice of actions
Definition 3.2 Nature
Nature is a nonplayer who takes random actions at specified points in the game with specified probabilities.
Definition 3.3 Action
An action or move by player i is a choice he can make
Definition 3.4 Strategy
A strategy is a rule that tells a player which action to choose at each instant of the game
Definition 3.5 Payoff
The payoff is the utility the player receives after all players and Nature have picked their strategies and the game has been played out.
Three other essential elements of a game are information, outcomes and payoffs.
Definition 3.6 Information set
A players information set at any particular point of the game is the set of different nodes in the game tree that he knows might be the actual node, but between which he cannot distinguish by direct observation
Definition 3.7 Outcome
The outcome of the game is a set of interesting elements that the modeler picks from the values of actions, payoffs, and other variables after the game is played out
Definition 3.8 Equilibrium
An equilibrium S* = (s*1, , S*n) is a strategy combination consisting of a best strategy for each of the n players in the game
The following definition will also be useful
Definition 3.9 Node
A node is a point in the game at which some player or Nature takes an action, or the game ends
In game theory common knowledge of the rules is assumed. This means that Player A knows the rules, and he knows that Player B knows that he knows, and in turn Player B knows that player A knows that he knows and so on ad infinitum.
It is also assumed that players are rational decisionmakers trying to maximize their utility. The utility should incorporate everything the player cares about in the game. For example, a player may care not only about monetary amounts but also about the wellbeing of others.
The rules of the game are described pictorially in either the extensive form or strategic form. The extensive form has the same appearance as a decision tree, with the difference that it has different players making decisions at the nodes in the tree rather than just a single decisionmaker.
Solution concepts
3.2.1 Dominant Strategy Solution and IEDS Solution
When we look for the solution of a game, that is, how we expect the game to be played, one possible solution is to look for dominant strategies. If a player has a dominant strategy it means that this strategy earns him a better payoff than all his other strategies regardless of what strategies the other players in the game use. The formal definition is
Definition 3.9 Dominant strategy
Strategy Si strongly dominates all other strategies of player i if the payoff to S is strictly greater than any other strategy, regardless of which strategy is chosen by the other player.
Since we know that a player with a dominant strategy will choose this strategy, all other players will be expected to choose the strategy that maximizes their payoffs given that the dominant strategy is used. A player with a dominant strategy can ignore strategic complications such as What will the other players do? and How will that affect my payoff?
Dominance can also be weak, so that an alternative dominance definition is the following
Definition 3.10 Alternative dominance definition
A strategy S*i is dominated by another strategy Si if the latter does at least as well as S*i against every strategy of the other players, and against some it does strictly better, such that
(i (Si, Si) ( (i (S*i, Si) for all Si
(i (Si, Si) ( (i (S*i, Si) for some Si
Either of these two things has to be true
There may be a dominant strategy. All remaining strategies are then dominated.
There may not be a dominant strategy, but there has to be at least one undominated strategy.
A strategy is undominated if it is not dominated by any other strategy. In the case where there is no dominant strategy, we can eliminate dominated strategies. This logic can lead to a chain reaction because once a strategy is eliminated there is a new relevant game in which a player may find that certain of his strategies, that originally were not dominated, are now in fact dominated. This procedure is called Iterated elimination of dominated strategies (IEDS).
Definition 3.11 Iterated Dominant Strategy Equilibrium
An iterated dominant strategy equilibrium is a strategy combination found by deleting a weakly dominated strategy from the strategy set of one of the players, recalculating to find which remaining strategies are weakly dominated, deleting one of them, and continuing the process until only one strategy remains for each player
Dominant strategy and IEDS solutions are appealing because of their simplicity. However, many games do not have such solutions.
3.2.2 Nash Equilibrium
Now we take the step to assuming we have an idea about what the other players are going to do. What we are interested is in how our strategies perform against this known strategy.
A strategy is a best response if it does better than any of a players other strategies when played against his guess about the opponents strategy. If all players have guessed correctly and played their best strategies, none would have an incentive to switch strategy and we have what is known as Nash Equilibrium.
Definition 3.12 Best response strategy
A strategy S*i is a best response to a strategy vector S*i of the other players if
(i (S*i, S*i) ( (i (Si, S*i) for all Si
That is, S*i is a dominant strategy provided that the other players play S*i. So what we need at this point is a condition such that player is guess is right (and all other players guesses as well). This leads us to the definition of Nash Equilibrium.
Definition 3.13 Nash Equilibrium
The strategy vector S* = S*1, S*2, , S*N is a Nash Equilibrium if (i (S*i, S*i) ( (i (Si, S*i) for all Si and all i
This condition includes the two requirements of a Nash Equilibrium:
Each player must be playing a best response against a conjecture
The conjectures must be correct
A Nash strategy only has to be a best response to the other Nash strategies, not to all possible strategies. This is the equilibrium concept most commonly used in game theory, because although less obviously correct than dominant strategy equilibrium it is more often applicable.
Information and Uncertainty in Game Theory
Information and uncertainty are in many cases two vital components of a game. It is worth spending some time clarifying how these factors influences the game theoretic analysis, especially since uncertainty is the key driver of an options value.
There are several distinctions regarding information in game theory. They have been summarized below.
Perfect information vs. imperfect information. The difference is that in the former, all moves in the game occur in a strict order of play whereas in the latter at one or more nodes in the game moves may occur simultaneously.
Symmetric information vs. asymmetric information. In a symmetric information game, no player has an informational advantage over the other players when he moves.
Complete information vs. incomplete information. In a game of complete information, nature does not move first, or her initial move is observable to all players.
Some of these concepts are very important to the ideas in this thesis, so they will be looked into more in detail in the sections that follow.
Another useful distinction in game theory is the following.
Games of certainty vs. games of uncertainty. A game of certainty has no moves by nature after any player moves.
Games of Perfect Information
We will begin this section with the definition.
Definition 3.15 Game of perfect information
A game of perfect information is an extensive form game with the property that there is exactly one node in every information set
In such a game, each time a player has to move she knows exactly the entire history of choices that were made by all previous players. We note that if moves are made simultaneously at any node, it is not a game of perfect information. Also, all players observe Natures moves before making any decisions.
In a game of perfect information, what is a reasonable prediction about play? We rely on what is called sequential rationality and backward induction. The principle of backward induction means that we start at the end of the tree and fold it one step at a time using utilityoptimizing as prediction for play. That is, each player at each instant is supposed to take the action that maximizes his benefit.
Backward induction is a general solution concept for games of perfect information. Kuhns Theorem tells us that every finite game of perfect information has a backward induction solution and that if for every player it is the case that no two payoffs are the same there is a unique solution.
Games of Imperfect Information
A game in extensive form that is not a game of perfect information is called a game of imperfect information. The distinguishing feature is that the extensive form contains one or more subgames that involve simultaneous moves.
A subgame is a collection of nodes and branches with the properties that:
It starts at a single node
It contains every successor to this node
If it contains any part of an information set, it contains all nodes in that information set
In trying to find a solution to a game of imperfect information we are seeking to establish what future behaviors are credible. In a game of perfect information the credibility of a strategy can be examined at any node by asking if it is in that players best interest to do what the strategy states. In a game of imperfect information it is the credibility of a groups decision that has to be judged and this is done by checking whether the strategy constitutes a Nash Equilibrium in that subgame.
The subgame perfect (Nash) equilibriumsolution to a game is defined as follow
Definition 3.16 Subgame perfect Nash Equilibrium
A pair of strategies is a subgame perfect (Nash) equilibrium if the strategies, when confined to any subgame of the original game, have the players playing Nash Equilibrium within that subgame.
The procedure is to start with a final subgame (a subgame that ends only in terminal nodes) and determine the Nash equilibrium within that subgame. Doing the same with all final subgames and working backwards to find Nash Equilibrium in every subgame of the tree all the way back to the initial node.
Games of Incomplete Information
3.6.1 Introduction
The definition of a game with incomplete information is
Definition 3.17 Game of Incomplete Information
A game of incomplete information is one in which players do not know some relevant characteristics of their opponents, which may include their payoffs, their available options and even their beliefs.
Restated, games with incomplete information are what we are dealing with when a player is unsure about the rules of a game.
Who are the other players?
What are their strategies?
What are their preferences (payoffs)?
This type of uncertainty is often expressed in terms of not knowing the other players type.
There are four levels of complexity.
I dont know the other players type but it doesnt matter because the other player has a dominant strategy regardless of his type.
I dont know the other players type but this time it matters because my payoffs depend on which type he is and he will have different dominant strategies depending on his type.
I dont know the other players type and it matters which type he is, but this time around he only has a dominant strategy in one of the cases.
I dont know the other players type and it matters which type he is, and this time around he doesnt have a dominant strategy in any of the cases.
In terms of solution, the first is simply a case of choosing my best response. The second can be solved by assigning probabilities to each possible type, that is, with probability q he is type 1 and with probability 1q he is type 2. I choose the strategy that gives the highest expected payoff.
The third is even more complicated because even if he is, say, type B, I cannot predict his strategy in that case because the likelihood of each action depends on his guess about my intentions. The fourth case is the most complicated because regardless of his type I need to predict his actions, which in turn will depend on what he is predicting about my actions.
3.6.2 Bayes Nash Equilibrium
There are three basic assumptions.
Player B knows her preferences, meaning that she knows whether she is type 1 or 2.
Player A doesnt know which type player B is and attaches a probability q that she is type 1 and probability 1q that she is type 2.
Player B knows Player As estimate of these probabilities.
The third point is the somewhat controversial assumption of a common prior. It is irrelevant whether Nature actually makes the moves with probability q and 1q as long as they are common knowledge. To solve games with incomplete information Harsayani proposed a twostep procedure.
Step 1 Turn the game of incomplete information into a game of imperfect information.
Step 2 Use Nash Equilibrium as the solutions concept
Step 1 means that we make the following reinterpretation of the game: Nature makes the first move and decides the type of Player B. Now all the players have the same exante knowledge about the rules of the game: it is therefore a game of complete information. In incomplete information games, the resulting Nash Equilibrium is given a special name: BayesNash Equilibrium.
Option Theory
Some Notation
In the following chapters, some notation will be useful.
S = Stock price
F = Forward price
r = Risk free annual interest rate, continuously compounded
T = Time to maturity
u = up parameter
d = down parameter
c = the price of a European call
C = the price of a American call
p = the riskneutral probability of an up move in the underlying
( = the value of a given portfolio
The Arbitrage Principle
Since the arbitrage principle is so central to option pricing, in this section I will briefly illustrate its workings. Although the essay is about options, I have for simplicity chosen a forward contract to show how its price can be determined by arbitrage arguments.
Definition 4.1
Arbitrage is a trading strategy with no initial cash flow and with positive probability to give a positive cash flow at settlement but without risk of negative cash flow
In efficient markets, such arbitrage doesnt exist for anything but very small amounts of time. This is because as traders realize the potential for risk free profit, they start very rapidly to bid for the asset causing its price to swing back into parity with other market variables.
To continue, investment assets can be divided into two categories.
Investment assets. These are assets held by investors for investment purposes only.
Consumption assets. These are assets that bring some physical benefit to the holder of the asset, such as commodities that have industrial uses.
Since the latter is somewhat more complex to analyze, consider an investment asset such as a stock, whose current price is 50$. The interest rate is 10%. What is a fair price of a forward contract to be settled four months hence?
Suppose it is 51. Suppose further that an investor who owns the stock sets up the following portfolio:
Sell the stock in the market and collect 50
Invests the 50$ in risk free government bonds yielding 10%
Go long on the forward contract
In four months, his risk free investment will have grown to 50 * e 0.1*(4/12) = 51.7. Under the forward contract, he can repurchase the stock for 51. But since his original 50 have grown to 51.7, he can close out the contract and still come out 0.7$ ahead no matter how the underlying stock moves. The reason is that the forward contract is underpriced relative to the other variables.
Suppose now that the price is 53 instead but that otherwise the situation is the same. An investor could then set up the following portfolio.
Go short on the forward contract
Borrow 50$
Buy the stock
In four months, the forward contract enables the investor to sell the stock for 53 regardless of how the stock moves. He must then repay the loan, and this amount will have grown to 50 * e 0.1*(4/12) = 51.7 (assuming he can borrow at 10%). But this leaves the investor with a profit of 53 51.7 = 1.3 in any possible scenario of the stock price. This arbitrage strategy was possible because the forward price was overpriced relative to the other variables.
In can be shown in this way that the only forward price that will prevent such arbitrage strategies is 51.7. Any other price would present investors with arbitrage opportunities. The forward price is therefore given by
F = Se r * t
Options: Some Definitions
We now move on to consider options. They differ from forward contracts in that they provide the holder with a beneficial asymmetry: they can choose to exercise the option only if it is to their advantage to do so.
The most common types of options are calls and puts, defined below.
Definition 4.2 European call option
The holder has a right but no obligation to buy the underlying asset at a predetermined time at a predetermined price. The writer has an obligation to sell
Definition 4.3 European put option
The holder has a right, but no obligation to sell the underlying asset at a predetermined time at a predetermined price. The writer has an obligation to buy.
There are also American options, which are the same as Europeantype except that the holder can choose to exercise it at any time during the life of the option. Since it is the same in all other regards but also provides this flexibility we have the general result C ( c (see definition section 4.1).
4.4 Option Pricing by Creating a Risk Free Hedge
In this and the following chapter I will try to give an intuitive derivation of the risk neutral pricing methodology. An equivalent approach is to value an option is by constructing a replicating portfolio, whose value is the same as the options in every state of the world and by necessity has to have the same value today. I will not review it here because it is generally considered that the riskneutral methodology is computationally easier.
The idea underlying option pricing is to create a portfolio about whose value there is no uncertainty at the settlement date of the option. Then the argument can be made that, since there is no risk concerning its value at the end date, its return must equal the risk free interest rate.
To see how this works, consider a stock selling at S0. It is known that at time T is value will be either S0u or S0d. The option to be valued is a European call with strike price X, so that its value at time T will be either Cu or Cd.
To create a risk free portfolio, we consider this portfolio
A long position in the share
A short position in the option.
We now wish to know the size of the position in the share, called the hedge ratio and denoted (, that will make the portfolio risk free over the next short time interval. This means solving for ( in the following identity:
(S0u  Cu = (S0d  Cd
Our risk free portfolio is therefore made up of ( shares and a short position in the option. The portfolio is risk free because if the value of the underlying decreases so will the call option that we have shorted. When the hedge ratio is chosen correctly, the gain/loss from the long position in the stock is completely offset by a corresponding gain/loss on the short position in the call option.
Using formal notation, the hedge ratio is defined as
( = (ud)S0 / Cu Cd
But we still do not know the value of the option today. To proceed we check the value of the portfolio in any of the two scenarios, which in the up case would be
(S0u  Cu = (T
Since the portfolio is risk free, it can be discounted using the risk free interest rate. We then know the present value of the portfolio. Since we now know both the portfolios composition and its present value, we can solve for the sole unknown C0, which is the price of the option today. That is,
(S0 C0 = (TerT
Inserting the definition of the hedge ratio and solving for C0 we obtain
C0 = [pCu + (1 p) Cd ]erT
where p is defined as
p = (er * t d) / (u d)
Option Pricing by Assuming Risk Neutral Investors
The risk neutral methodology assumes that the same value of the option will be obtained in a riskneutral world as in our own riskaverse world as shown in the previous section. A risk neutral world is one in which investors do not care about risk, meaning that we can assume the two following things.
The expected return from all traded securities is the risk free interest rate
Future cash flows can be valued by discounting their expected values at the risk free interest rate
As it turns out, to value the option assuming riskneutrality is much simpler. Since it says that all assets are expected to earn the risk free rate, we set the probabilities of an upmove and downmove respectively such that the discounted value of the underlying equals its spot price.
In the example from the previous section, we would set
S0 = [pS0u + (1p)S0d]e r* t
and solve for p, which gives
[4.1] p = (er * t d) / (u d)
Armed with these probabilities, we can now calculate the value of the option as the expected payoff at t = T discounted back to present, or
[4.2] C0 = [pC0u + (1 p)C0d]er* T
This is exactly the same option value we obtained by creating a risk free portfolio in the previous section. Very simply put, we get the same result because as we move from a riskneutral world the expected return on a risky asset increases, but this is exactly offset by the increase in the discount factor that comes about because of investors aversion to risk.
That the expected return actually decreases in a risk neutral world can understood from the fact that the riskneutral probability of an upward move in the underlying is lower than the subjective probability of an upward move in a risk averse world. This of course lowers the expected payoff, but this is compensated for by a lower discount rate (the risk free).
It is important to note that the risk neutral probabilities are not the actual probabilities of an up or down move in the underlying. Rather, they are the probabilities that would prevail in a riskneutral world to prevent arbitrage opportunities. The actual probabilities of an up or downward move does not enter the option valuation formula because they have already been accounted for when valuing the underlying and are thus baked into its current spot price.
4.5 The Stochastic Properties of the Underlying
The question we need to ask at this point is; how do we know the values of u and d? That is, what can be assumed about the future movements in the underlying asset?
What we need is a mathematical model that describes the development over time of the underlying asset from which the option derives its value. We have two basic possibilities:
The underlying takes on values according to a log normal distribution
The underlying takes on values according to a binomial distribution
The former is closely connected to the Black & Scholes formula and option valuation in continuous time (a continuous variable can take any value within a given range). The latter is associated with the binomial model and discrete time option valuation (a discrete variable can take only certain discrete values).
If we assume a multiplicative stochastic process (a Geometric Brownian process) the discrete distribution will converge with the log normal distribution, as the time intervals in the process become infinitesimal. They thus converge in the limit, meaning also that the discrete approximation becomes increasingly imperfect the longer the subintervals in the process.
It is clear that we are interested in a stochastic process meaning that we dont know which one of a certain known values that will be realized. The range of values that we assume it can take is associated with the volatility of the asset.
Consequently, the u and d parameters are chosen so as to match the volatility of the underlying as well as its expected growth rate (. In a risk neutral world the latter changes, but the volatility is the same as when the world is risk averse. So when we model the evolution of the underlying in a riskneutral world, the lattice will look the same as if we were assuming a riskaverse world (with respect to the possible values of the underlying).
The values of u and d, the multiplicative factors that determine the value of the underlying at the next node in the tree, are not uniquely determined, but one possible choice is:
[4.3] u = e(((t
[4.4] d = e(((t
The most common lattice used to visualize the stochastic development over time of the underlying risky variable is the binomial tree. It shows the values the underlying can take on at every node in the tree (corresponding to a given time period) having assumed up and down parameters. In the binomial tree the probabilities (usually riskneutral) of an up and downward move are also indicated.
Such a lattice has the nice property of being recombining, which means that the value of the underlying after one downmove and one upmove is the same as after one upmove followed by a downmove. Because of this property, the nodes in the middle of the tree will always contain the same value throughout the whole tree.
4.6 Modeling Dividends in the Binomial Tree
Since I later in this essay will attempt to model the alternative cost of the (real) option to defer, I will devote this section to describe how the financial options equivalent, namely dividends, are treated.
If the underlying asset on which the option is written pays out dividends, the value of the option will be affected. A call option negatively because it lowers the probability of the stock price being inthemoney at the exercise date. A put option positively because it increases the probability of the stock price being inthemoney at the exercise date.
It is assumed that the value of the stock prices goes down by the exact amount of the dividend at the dividenddate (in the real world, because of taxes, this need not always be so). This has the effect that the binomial tree describing the stock price will no longer be recombining.
To handle this problem, we note that if the size and time of payment of the dividends are known at the time of valuation, which they are usually assumed to be, this is a nonstochastic part of the underlyings value because their present value is certain. Therefore the stock price can be divided into two components.
A stochastic component, denoted S*
A known component, the present value of future dividends, denoted I
We can therefore write the price of a dividendpaying stock as
S = S* + I
which of course also means that S* = S I. After assuming that the volatility of S* is equal to the volatility of S we can proceed in three steps:
Construct a binomial tree for S* with the same up and down parameters as for S
Add the present value of dividends I to each node in S* ( we get an alternative tree for the underlying stock price
Value the option using this alternative tree.
The intuition is of course that, in the case of a call option, the dividend lowers the value of the option. We add back I at each node preceding the dividend date, which means that up to that point the value of the option is unaffected, whereas beyond it the options value suffer a loss from the fact that the dividend has occurred.
We can also model the dividend yield, which is when the dividend is a known proportion of the stock price. If the total dividend yield (there is one dividend yield per actual dividend and there can be many dividend dates during the life of the option) is denoted (, the underlying will correspond to
S0ujdij j = 0, 1, , i
before the dividend date occurs. If it has occurred, the value of the underlying will correspond to
S0(1()ujdij j = 0, 1, , i
4.7 Real Options: The Analogy
As discussed in section 1.1, we live in an uncertain world. That the NPVmethods dominance came to an end is because it doesnt deal well with this fact. Implicitly it makes some assumptions that rarely are good descriptions of reality, namely that the future is certain and that management must remain passive throughout the investments life.
Also, how it handles uncertainty is to discount heavily for it. Due to this practice, many potentially very promising projects may never have seen the light of day. Real options theory, on the other hand, offers a completely different take on uncertainty. It realizes that there may be limited downside risk in a project because of flexibility inherent in the project, but that the upside potential remains intact.
In the view of real options theory, uncertainty may actually raise the value of a project because with volatility the potential upside grows larger while the downside is still protected against through the projects own flexibility. For example, the option to defer investment means that we are free to proceed in the good state of the world and also free to not undertake the project should the bad state of the world come to pass.
The resemblance with financial options should be clear. The same beneficial asymmetry appears here: we have the right, but not the obligation to invest. A possible definition is the following:
Definition 4.4 Real option
A real option is the right, but not the obligation, to take an action (e.g. deferring, expanding, contracting or abandoning) at a predetermined cost called the exercise price, for a predetermined period of time the life of the option.
The option to defer is therefore analogous to a call option on a dividend paying stock. This means that the capital budgeting decision can be modeled in much the same way as financial options, thereby offering a superior way to incorporate uncertainty and flexibility into the valuation of the project.
And the analogy holds not just for call options. If we are in a situation where we can invest in a project and still be able to dispose of it should conditions deteriorate, what we have is in fact a put option on the gross value of the underlying project. There are other types of real options, all differing in the flexibility they provide.
To be a good description of reality, many real option models need to incorporate several types of options simultaneously, such as switching options present in gold mines. They may also involve several options that appear at different points in time and where the succeeding option is contingent on the exercise of the preceding option. These are called compound options and are often present when the projects consists of a series of phased investments.
4.8 Carrying The Analogy Over
Using a Twin Security
While it is not too difficult to see that capital budgeting decisions can be seen as options on a conceptual level, it is not as easy to see how the valuation principles can be implemented in practice.
In previous chapters I presented the principles underlying options pricing, and as could be seen options pricing relies heavily on arbitrage arguments. For us to correctly price the claims from a project, we must somehow establish a link to the financial markets where the arbitrage argument can be used.
The arbitrage argument cannot be used in real markets because assets are not traded frequently enough for that kind of market efficiency ever to come about. So what is needed is to find a security in the financial markets that has the same risk/return characteristics as the project we are contemplating investing in.
If such a security can be found, with it we can create a portfolio, called the cashequivalent portfolio, which yields the same cash flows as the proposed investment in every scenario. Alternatively, we could use the riskneutral methodology to calculate the probabilities that would equate the current price of the twin asset with its discounted future values one period hence.
Establishing a relationship between the price of a publicly traded asset and the cash flow from the proposed investment is the chief difficulty of this approach. Concerning the twin security, we make the following observations:
If the size of a cash flow is unrelated to any marketed asset, it will not add significant risk to a welldiversified portfolio and we can therefore treat the investment as essentially risk free.
A natural candidate for the cashequivalent portfolio is the firms own stock, but it introduces the problem that the stock price may already reflect the cash flow from the investment. An industry stock average or market portfolio could be used.
Lacking a body of theory that could associate traded assets with cash flows, a purely empirical relationship must suffice. That such an empirical relationship is enough for the arbitrage argument to hold is one of its chief virtues.
However, the consensus is growing that this is not a very fruitful approach for dealing with valuation of real options. There are two main reasons
It is nearly impossible to find marketpriced underlying risky assets
The volatility of the underlying project is not the same as a traded assets even if they are clearly correlated.
To understand the last point, consider a gold mine. The volatility of the price of gold is a feasible proxy for the volatility of the project, but not a perfect one, because there is operational and financial leverage in a gold mine increasing the risk (i.e. the volatility) of the underlying project as compared to the price of gold.
These difficulties have constituted a major stumbleblock for implementing real options analysis, but fortunately for practitioners a new methodology has been proposed and this will be the subject of the next section
The Marketed Asset Disclaimer
First, we make the proposal that the twin security to be used in valuing the project should be the present value of the project itself (assuming no flexibility). To defend this choice, it can be noticed that no asset could be better correlated with the project that the project itself and that its present value without flexibility is the best, unbiased estimate of the market value of a project.
In making the last assumption we arrive at the Marketed Asset Disclaimer (MAD), which allows us to value a real option and get the same result as if we had available a perfectly correlated financial asset. The great advantage, of course, is that the present value of the cash flows is easier to estimate.
To make the approach implementable we need a second assumption: that properly anticipated prices fluctuate randomly. This is a result that has been proven theoretically by Paul Samuelsson and it says that regardless of the pattern (cyclicality) of the cash flows produced by the project/company, the rate of return of any security that derives its value from these cash flows will nevertheless follow a random walk.
To see why, we note that it assumes that investors have complete information about future cash flows and that this information is incorporated in the current price of the security. So because of complete information, no matter the cyclicality of the cash flows, only deviations from those predicted cycles will cause volatility in the rate of return on this particular asset.
Making use of these two assumptions, valuing the option is a fourstep procedure:
Compute base case present value of the project assuming no flexibility
Model the uncertainty in a so called eventtree to understand how the present value of the underlying develops over time
Identify managerial flexibility in the project and turn the event tree into a decision tree
Value the project with flexibility by starting at the end node of the tree and find the value maximizing decision at each node in the tree rolling backwards
The question remains as to what the projects volatility is. The overall volatility in the rate of return on a project may have several sources: there may be uncertainty about future price levels, future costs, future quantity of output and so forth.
The volatility of the risky variables that drive the rate of return on the project can be estimated in two ways: either by using historical data or subjective estimates by management. They can be correlated with each other or itself through time (i.e. autocorrelation or mean reversion). If we use Monte Carlo Simulation to simulate the projects present value in different states of the world, however, these uncertainties can be combined into one the volatility in the rate of return of the project.
4.9 The Option to Defer
This essay takes aim primarily at valuing the option to defer investment in a project. This type of option has been described as in the definition below.
Definition 4.5 The option to defer
A deferral option is an American call option found in most projects where one has the right to delay the start of the project. Its exercise price is the money invested in getting the project started
If we let V denote the value of the cash flows it is expected to generate and X the cost of getting the project started, at the end of the options life t = T, its value is mathematically written as
Max [VT X, 0]
That is, if the value of the underlying project exceeds the exercise price, we will exercise the option (meaning that the project will be carried out). If the value of the underlying isnt high enough to justify paying the costs to get the project started, we will not exercise and our payoff from the option will be zero (rather than having to suffer the losses).
If we value this option, we will arrive at an expanded NPVvalue of the project, that is the sum of the NPV without flexibility and the value of the option.
Definition 4.6 Value of a project with the option to defer
Value of project with flexibility = NPV of project assuming no flexibility + Value of the option to defer.
The value of the project at t = 0 will be the maximum of
Vt = 0 (Investing immediately)
0 (Not investing)
The option value of deferring
There will be a trigger value V* , at which the value of the missed net cash inflows equals the flexibility value of deferring. Any flexibility value above this point will induce deferment, whereas a flexibility value below it will induce immediate investment.
This rest of section will further discuss the nature of and intuition behind the value of a project in which the owner has the option to defer it (infinitely). The standard application of the NPVrule misses the point that any investment opportunity competes not just with other projects but also with itself delayed in time.
The value of a project stems from three sources:
Its NPV if it were undertaken today
Embedded options built into the project itself
An option on the movement on capital costs and prices
The third factor refers to the observation that if investment is undertaken today, there is an opportunity cost in the fact that we lose the option to undertake the project at some later date when financing terms are more favorable (i.e. capital costs are lower).
This is best understood through an example. Suppose there is an investment with an up front cost and a positive cash flow at some point in the future. The cash flow is sure to be realized, so we need only discount with the risk free interest rate. At current interest rates, the NPVformula yields a slightly negative value.
The decisionmaking in this project has three aspects to it.
It is a good decision to not undertake the project at the current interest rate, because the NPV is negative
It is a bad decision to sell the project for a trivial amount because the project includes all the rights to undertake it in the future. With interest rate uncertainty, the project is equivalent to a call option on a bond with equal length as the project. Just because the NPV is negative today doesnt mean the project is worthless.
It is an ugly decision to undertake the project if the interest is such that the NPV is positive by a trivial amount. When interest rates are uncertain, there is a trade off between taking on the project today and the cost of foregoing the option to undertake the project at some later date when interest rates are more favorable.
The options approach thus means moving away from valuing an investment under the assumption that it must be undertaken today to valuing the opportunity to invest.
4.10 Problems with the Analogy
Although the analogy has made a major contribution to our ability to evaluate capital decisions, it remains imperfect. Here, some of the limitations of the analogy will be discussed. This will lead us to the problem this essay is about, namely the impact of competition on the value of real options.
In both financial and real options theory risk is assumed to be exogenous that is, the uncertainty about future rates of return is beyond the control of the holder of the option. This assumption is reasonable in the case of financial options but obviously less so with real options, because the holder of the option is the company itself, and managements actions can affect the value and risk of the underlying.
And then there is the issue about estimating the volatility of the underlying risky asset. With financial options, we have available time series of historical data regarding the returns on the underlying. Using these data, the standard deviation can be estimated. With real options, the underlying is rarely a traded asset, so the estimation of the volatility is often trickier, although possible, as we have seen, even if no traded twinsecurity can be found.
Also, a financial option gives the holder an exclusive right to the underlying asset on the settlement date. In many cases, this exclusiveness does not exist for real options. This problem is particularly acute in the case of the option to defer investment, because there may be competition about the opportunity to invest.
The value of the option to defer is likely to decrease under competition, because what the other aspiring players do will affect our own payoff from the investment opportunity and their interest is often diametrically opposed to our own. An important concept is that in a competitive market, such as a duopoly, the state value of the investment opportunity will be the Nash Equilibrium outcome of an investment game rather than MAX[V I, 0] and the option value.
If we look solely at the option value, there will be a bias toward waiting whereas if we look only at the combative aspect we will have a bias toward investing straight away to preempt. What the next chapter will try to do is finding an analytical framework for striking the right balance.
Real Option Valuation under Competition
5.1 On Economic Rents and Competition
At this point it will be useful to define the meaning of economic rents, as this concept will resurface many times in this essay.
Definition 5.1 Economic rents
Economic rents are excess profits above the opportunity cost of capital
In order to analyze how options theory interacts with strategy it will be useful to briefly go through the different forms of market. A market is usually classified into one of three categories.
Monopoly
Oligopoly/Duopoly (Imperfect competition)
Perfect competition
Under monopoly, the firm is the sole player in the market and there will not be any competition about investment opportunities. From a monopolists point of view the option to defer is equivalent to a call option on a dividend paying stock that has a constant dividend payout ratio.
Under perfect competition, there is a tendency for the economic rents of a project to quickly converge to the cost of capital because competitors are free to enter. In the long run competitive equilibrium, all firms earn their cost of capital.
Real markets, however, are often imperfectly competitive. In such a market it is possible to consistently earn economic rents. Under competition, these economic rents should be thought of in terms of sustainable competitive advantage. In other words, a firm can be expected to earn economic rents if it has a competitive advantage in realizing the project.
Depending on the strength of the competitive advantage, the speed of the convergence of economic rents towards the cost of capital will be different. Sources of competitive advantage can be, for example, economies of scale/scope, absolute cost advantages, product differentiation and other barriers to entry.
5.2 Investment Tactics under Competition and Uncertainty
In the previous section it was reviewed how different forms of market means different expectations of the evolvement of economic rents over the life of the project. Most interesting of these market forms is oligopoly because here strategic interactions are often a central part of the analysis. In addition to the development of market demand, firms will have to factor in the behavior of competitors in shaping their investment decision.
So strategic considerations have already been introduced. In this section I will discuss what happens when the situation is complicated further by uncertainty about some exogenous risky variable that will have an impact on the value of the project. This discussion refers primarily to the option to defer investment assuming no exit option.
What variables could logically affect the value of the investment project? In what follows, some investment tactics will be developed based on two factors.
The relative market power of the firms in the industry
The NPV of the investment opportunity in relation to the degree of exogenous uncertainty
Obviously, the degree of uncertainty will be important. More uncertainty will make the investment more risky, so that we can make the following proposal (for proof of propositions 5.15.3, see Smit H & Ankum LA, A real options and game theoretic approach to corporate investment under competition (1993).
Proposition 5.1
A higher degree of uncertainty about the future distribution of the exogenous risky variable will increase the risk of investing, thereby inducing companies to defer investment
Next, the Net Present Value as of t = 0 will be important. If the NPV is high, there is less to be gained from waiting, because the probability of the project ever ending up outofthe money is smaller. Also, competitors are also likely to move in quicker if the base case NPV looks high. Therefore, we have proposition 5.2.
Proposition 5.2
A high base case NPV of the project will induce companies to invest sooner rather than later because the option to defer is worth less when the probability of negative NPV is small. Another reason is that it is likely to make competitors move in early to maximize their share of the market.
Finally, we consider the market power of a firm. The principle at work here is that firms with strong market power has less to worry about competition, and they can therefore attach more weight to the consequences of committing in the presence of market uncertainty. This is because the effect on NPV from letting competitors with neglecteable market power enter the market can be afforded in light of the uncertainty about market demand.
Proposition 5.4
A firm with high market power will have more incentive than one with weak market power to defer investment when there is uncertainty about market demand because less is to be lost to competition effects than from giving up flexibility.
The principles in propositions 13 can be summarized into a set of investment tactics in an oligopoly situation with uncertainty about future demand.
If the NPV of the project is low and markets are uncertain, a strong firm will take advantage of the relatively higher flexibility value and defer investment
If the NPV is high and markets are stable, the flexibility value is relatively small so the dominant firm will invest early instead.
If the NPV is low, a firm with a weak market position will defer investment to benefit from the resolution of uncertainty.
If the NPV is high, a firm with a weak market position will invest early hoping to preempt competitors.
5.3 Analyzing First Mover Advantage
The General Principles
In section 5.2 it was suggested that the market power of a firm is a determinant in choosing the optimal investment strategy. The subject of preempting the competitors was also touched upon, and in this section I will probe further into the subject of being first on the market.
In many instances there may be a benefit from being first on the market. This could result from name recognition, being able to build the best network, setting the standard of the product and so forth. It would seem that the more symmetric the market power, the more valuable the first mover advantage would tend to be. The reason is that when firms are fundamentally similar, anything that makes them stand out from the crowd will be beneficial, sort of like a focal point (supposing, of course, that the thing is not something intrinsically negative). But naturally a distinct first mover advantage will change the dynamics of the game even when symmetric market power is not the case.
If symmetric market power is at hand the implications of Proposition 5.3 for our investment strategy fall out, and the things we need to focus on are instead the degree of uncertainty and the size of the NPV of the investment opportunity. What I would like to bring to the table is the idea that if the NPV is strongly influenced by who is able to get the first mover advantage, we can reformulate the tradeoff as being between moving first and the flexibility value provided by the option to defer.
Conjecture 5.1
In the case of an investment situation in an oligopoly market with a first mover advantage, the trade off between the flexibility value of the option to defer and the first mover advantage of investing first is crucial in determining the optimal investment strategy, but not relative market power
According to conjecture 5.1 relative market power no longer makes a difference in making the tradeoff. Although the general rules presented in section 5.2 are still valid, the relative market power itself would not enter the strategic considerations in finding the optimal decision.
Table 5.1 Attempt at validating conjecture 5.1
Assume two players, Player A and Player B, involved in an investment game.
Player A has a market power giving it a share of 70% of the market. Player B has the remaining 30%. This is common knowledge.
The game has no first mover advantage. Future market demand is uncertain. Investing means committing.
In choosing to invest, both players know that whenever both players have invested the market will settle and they will get their predetermined share of the market. As long as only one player has invested, it will have 100% of the market
As a consequence, the game is perfectly solvable because both players know the division of market power and that being first doesnt, at any stage in the game, affect this division.
Both players will choose whether to invest according to their own ability to bear the uncertainty and their expected profitability, given the market power that the rules specify.
The relative market power will still be important, but it will only affect the individual players opportunity cost but not any strategic considerations such as preempting.
The purpose of bringing this up is that it focuses the problem on the fact that for there to be an competitioninduced opportunity cost involved in waiting, there must be a risk of something being lost which cannot be regained after the postponement period. If the same division of market power holds no matter what there is nothing lost by waiting other than the net cash inflows that this market power dictates. And the important point is that this opportunity cost is not competitioninduced in the sense that it has to do with competitors actions in the future, only their predetermined market power.
In order to decompose this analytical problem further, we note that the investment decision will be a function of two things; the size of the first mover advantage and the value of the flexibility provided by the deferraloption. Knowing that the value of an option is primarily driven by the uncertainty (volatility) about the underlying, we have the following
Conjecture 5.2
The trade off between the first mover advantage and the deferral option will be a function of the size of the first mover advantage and the degree of uncertainty about the exogenous risky variable
This trade off is of course at the very heart of the whole problem of valuing options under competition and will be exemplified in section 5.3.3.
Sources of first mover advantage
In game theory a first mover advantage is analyzed in terms of setting the stage in a way that favorably affects the outcome of the game. This results from the fact that the first mover is able to commit to a particular action, so that the other players choice of action is reduced to finding the best response given the particular action of the first mover. Any value resulting from this principle could be termed commitment value and a possible definition could be
Definition 5.2 Commitment value
Commitment value is a positive effect on NPV resulting from credibly committing and thereby changing the outcome in a favorable way by removing some strategy combinations, forcing the other players to act contingent on the first movers action
The preemptive effects of such commitments have been widely analyzed by game theorists, and the conditions for commitments to be credible are also well understood. Due to time constraints neither will be reviewed here though.
In a market setting the first mover advantage may result not only from committing. It could also come about because there may be a genuine advantage to be had from investing first because of some characteristic of the market situation. It is a rule of the game, so to speak, and it could be about one or several of the following (the list is naturally not definitive).
Name recognition
Creating the standard of the product
Achieving the best distribution net
Getting a lead in the learning curve
The acquisition of future growth opportunities relative to competitors
These could be termed intrinsic first mover advantages, for lack of a better word, and it will be defined as
Definition 5.3 Intrinsic first mover advantage
An intrinsic first mover advantage is one due to some inherent characteristic of the game resulting in a positive NPV effect that cannot be recouped by later entrants in the game
A perhapsuseful way to think of intrinsic first mover advantages could be that they alter the predetermined market power in a favorable way.
We can now define the total first mover advantage as the sum of the commitment value and the intrinsic first mover advantage.
Definition 5.4 Total first mover advantage
Total first mover advantage = Commitment value + Intrinsic first mover advantage
Committing has a positive effect on NPV through its effect on other players choice of action. An intrinsic first mover advantage on the other hand has a positive effect from the sheer fact of being first independent of what actions the other players are going to do at a later stage in the game.
We can summarize the decision rules in an oligopoly situation as follows
Table 5.2 Decision rules for firms in oligopoly with symmetric market power
First mover advantageMarket uncertaintyDecision ruleYesNoInvest immediately (if positive NPV)NoYesDefer investmentNo NoInvest (if positive NPV)Yes YesTradeoff
An Example
As follows from looking at table 5.2 the interesting case is when there is a first mover advantage but also market uncertainty. To analyze this further, I will proceed by means of an example. The game describes a situation of irreversible investment under imperfect competition where players have the option to defer investment, and where the simplifying assumption is made that the market price is independent of the number of producers.
Analyzing such a game provides some interesting insights into how competition and options interact. There is no commitment value in this example (although it might first appear that it is), but this is not meant to suggest that this should be the prevailing case in the real world. Nor is the example meant to be innovative in any way, it is only a simplistic attempt at looking at the principles at work when there is both an option to wait and, at least seemingly, value from commitment. Far more sophisticated models describing this type of situations are available elsewhere.
This game will be reintroduced in section 5.3 to illustrate how strategic uncertainty can reduce the value of an investment opportunity.
The rules of the game are as follows
Table 5.3 Capital investment game with deferral options and exogenous uncertainty
Players
Player A and Player B
Information sets
There is symmetric market power and complete information. At t = 0 both players knows nature makes a move at t = 1 to determine market demand, and they also know the probability distribution of the value of market demand.
Actions and events
Both players choose at t = 0 whether to invest or defer investment
At t = 0.5 any player can, having observed at t = 0 the other players decision, choose to invest
At t = 1 nature makes a move and determines market demand. With a probability of 0.5 demand is high, and with probability 0.5 demand is low.
We note the following: Invest at t = 0, t = 0.5 and at t = 1 are mutually exclusive. Investing means committing, so there is no flexibility to exit the game at t = 1 should demand be low. At t = 0.5 no new information has arrived.
Payoffs
If nature picks high, present value (PV) of demand is 400. If nature picks low, demand is 50. Player As costs have a PV of 200. Player Bs costs have a PV of 150. The symmetric market power means that the players share the market equally if both invest.
Outcomes
The outcome includes a description of who invests and when as well as the profit to each player.
Let us first analyze the game assuming there is no flexibility, meaning that the investment must be undertaken today. This is thus a detour from the rules of the game as outlined above, but it better illustrates how the option to defer works when it is reintroduced, illustrating the philosophy of opportunity to invest rather than have to invest today. Since both players must make the decision at t = 0, the game is one of simultaneous moves.
With a 0.5 probability the game will have the following strategic form.
Table 5.4 Game matrix assuming no deferral option, high demand
Player A/Player BInvestNot investInvest0/50200/0Not invest0/2500/0
With a 0.5 probability, the game will be:
Table 5.5 Game matrix assuming no deferral option, low demand
Player A/Player BInvestNot investInvest175/125150/0Not invest0/1000/0
We note that the game in table 5.4 is dominance solvable, because Player B will invest irrespective of what Player A does. The equilibrium outcome is (0,250), because not investing is Player As best response. It follows that in the game in table 5.5 player A would not choose to invest either because the expected payoff is even worse in that state of nature. Player B will choose to invest because the expected payoff of investing is 0.5 * 250 + 0.5 * (100) = 75. This is the value of the investment, to Player B if it had to be undertaken today.
Now we return to the rules of the game as specified in table 5.3. Both players have the option to defer investment to t = 1 right after nature makes a move. Conceptually, this means we are moving away from valuing the investment under the assumption that it must be undertaken today and instead viewing it as an opportunity to invest at either t = 0 or t = 1.
Player B might think along the following lines. Instead of investing right now, which renders an expected payoff of 0.5 * 250 + 0.5 * (100) = 75, it is possible to wait and let uncertainty resolve itself. Since Player B would only invest in the good state of the world, we are now contemplating the expected payoff 0.5 * 250 + 0.5 * 0 = 125. Since 125 ( 75 Player B would draw the conclusion that it is better to defer, indicating an option value of 50.
But this is forgetting that there is still competition from Player A about the opportunity to invest. The option to defer changes the dynamics of the game. If Player B decides to wait, player A can choose to turn the game into a sequential move game by investing at t = 0.5 and make the analysis that it will be in player Bs best interest not to invest given that Player A has already invested. Logically, since no new information has arrived, the decision should be a repeat of the decision at t = 0, but lets see what happens
Note that Player A would be assuming that Player B, knowing that Player A has already invested, will be comparing the following two payoffs:
If I invest, my expected payoff is 0.5 * (125) + 0.5 * 50 = 37.5. The first term represents low demand (50/2 150 = 125), the second high demand (400/2 150 = 50).
If I dont invest, my expected payoff is 0.5 * 0 + 0.5 * 0 = 0.
Since under this assumption Player B would not invest (0 ( (37.5)), player A would be thinking that investing is the best action. Her strategy Invest at t = 0.5 yields a positive expected payoff (assuming a decision not to invest from Player B at t = 1) because 0.5 * 200 + 0.5 * (150) = 25. The first term is the high demand scenario (400 200 = 200) and the second is the low demand scenario (50 200 = (150))
So if Player A is making the right analysis, the conclusion must be that there exists a first mover advantage in this game from committing, because by waiting to invest Player B lets Player A seize this advantage. This leads to a worse outcome for Player B. If Player B recognizes this line of reasoning, which he will because game theory assumes rational players, he will undertake the investment immediately receiving an expected payoff of 0.5 * 250 + 0.5 * (100) = 75.
This is of course the same payoff as in the game where none of the players had the option to defer. The interpretation seems to be that the option to defer is worth zero, which is equivalent to saying that the value of the opportunity to invest is equal to the investment if undertaken at t = 0 and that the reason for this being so is the existence of competition.
The only problem is that Player As analysis is wrong in the first place. There is another layer of analysis in this game. To see why it is wrong, we note that the rules of the game in table 5.3 say that both players have the option to defer investment until t = 1 upon seeing natures move. This means that even if Player A invests in an attempt to seize the apparent first mover advantage, Player B can choose to invest at t = 1 if demand turns out to be high.
So in reality, Player A will have an expected payoff from investing at t = 0 of 0.5 * 0 + 0.5 * (150) =  75. The first term is 400/2 200 = 0 because when nature picks high, Player B will exercise the option to profit from high demand.
So the fact that Player B has the option to defer means that the equilibrium outcome as wrongly calculated by Player A is destroyed. This brings us back to the game analysis where Player B can defer investment to take advantage of letting uncertainty resolve itself without risk of Player A preempting.
In conclusion, then, the true equilibrium outcome is that both players defer at t = 0 and that Player B invests at t = 1 if nature picks high demand and chooses not to invest if demand is low. In equilibrium, Player A will defer at t = 0 and not undertake the investment at t = 1 in either state of the world because the payoff from doing so is zero.
Comments on the Example
The example in the previous chapter showed the principles underlying the interaction between competition and flexibility. In some circumstances when the first mover consist only of the ability to set the stage, it will be neutralized because the other player will always take advantage of the flexibility provided by the option to defer. The commitment didnt work in this example because the volatility was high enough to render it worthless.
It can easily be shown that there is commitment value in the game in the previous section if we change the numbers a little. Lets say instead that demand is 300 if high and 200 if low and that A commits. If demand is high B will receive 300/2 150 = 0 from investing at t = 1 and therefore will not bother. A therefore gets an expected payoff from committing at t = 0.5 of 0.5 * 300 + 0.5 * 0 = 150.
The difference is that here the volatility of the risky variable (demand) is much lower. This has served to illustrate the quite intuitive point that moving first and committing means bearing uncertainty and the higher the variance the less advantageous this will be (when investing means committing I should add).
5.4 The Opportunity Cost in Real Options Analysis
In this section the analogy between financial and real options is extended to include dividends. In section 4.6 we looked at how dividends are adjusted for in the binomial model assuming a Europeantype option. It has already been established that the option to defer is analogous to a call option. But we have also noted that it is an Americantype option because the holder of the option to defer is usually free at every point during the options life to choose whether to exercise or not.
A general result in options theory is that early exercise of an American call option on an underlying that pays no dividends is never optimal. The intuition is that when exercised, the owner of the option gets the underlying asset whereas if unexercised the holder has the rights to the underlying, but also the protection against a downward move during the remaining life of the option. So in such cases we would have the equality C = c (see notation in 4.1).
This is an important issue because if we apply valuation techniques for European call options on real project opportunities, we will get the unrealistic decision rule that it is always optimal to wait as long as possible before exercising the option.
Introducing dividends changes the picture because a dividend is only due to be paid to the owner of the underlying. So the holder of the option must trade off the value of exercising and immediately receiving the dividend and keeping the option and therefore the protection against downside risk. If the option is European we clearly dont have the choice of early exercise so with dividends we have C ( c.
To value an Americantype option on a dividendpaying underlying asset we have the following steps.
Construct a binomial tree for S* with the same up and down parameters as for S
Add the present value of dividends I to each node in S* ( we get an opportunity tree for the underlying stock price
Work backwards through the three and compare at every node the value of the option if left unexercised (as calculated with the riskneutral methodology) with the value of exercising to get the underlying (S X). At every node choose the highest of the two
We make the observation that it is only optimal to exercise immediately preceding the dividend payment, and that it can also be showed that in most circumstances it is only the final dividend date that need to be considered.
If we now look at the deferral option on a nontraded project, what can we say about the opportunity cost of waiting? What principles apply?
With a financial option, we have the right to acquire an underlying that will drop in value at some future date because of a dividend cash payment.
With a real option (to defer) we have the right to acquire an underlying asset (the project) which decreases in value starting a t = 0 because of foregone cash inflow
There is a fundamental difference here. The opportunity cost of in the case of a financial option occurs at a known date in the future, and, given that we can estimate the volatility, we can make a tradeoff between the value of the dividend and the value of the protection against uncertainty.
In the case of a real option, the opportunity cost consists of missed operating net cash inflows that occurs as of t = 0 and going forward, not a known date in the future, say, t = 1. Instead of at a known date in the future, we can think of the missed cash flows as occurring up to a known date (the decision nodes in the binomial tree).
To the extent theses missed net cash inflows exceed the firms cost of capital, it constitutes an economic rent and therefore an opportunity cost. The reason that a firm, under some circumstances, is willing to sacrifice this economic rent is of course that it is valuable to wait and see under market uncertainty. As always, just because the project is inthemoney today doesnt mean we should invest right away because giving up the downside protection of the deferral option counts as a real economic cost. Anyway, lets denote this economic rent ER
The next step in evaluating the dividendanalogy is to introduce competition because as we discussed in section 4.10, competition has no natural equivalent in the financial markets. If we assume the existence of a first mover advantage, we have to add to the opportunity cost the probability that waiting will cause it to be lost to competitors. We could call this E(LFMA), where LFMA stands for Loss in First Mover Advantage.
I think it would a helpful way to frame the whole concept if we think of E(LFMA) as a present value concept, that is, the discounted value of the cash flows the first mover advantage would pay off. That way the dividend analogy holds up very well: the opportunity cost is the discounted value of cash payments occurring between now and some future date, or in the case of a financial option, on some future date. The main difference is that with a real option, the missed cash flows could be anything between a steady stream and a few large cash inflows, depending on the type of the project, whereas dividends are simply known amounts to be paid at discrete points in time.
A possible definition of the total opportunity cost in real options analysis is
Definition 5.5 Total opportunity cost of deferring
The total opportunity cost of deferring is equal to I(ER)+ E(LFMA)
Where
I = The invested capital
ER = the amount by which the expected rate of return on the project exceeds the projects cost of capital
E(LFMA) = The expected loss in first mover advantage, that is the NPV of the first mover advantage * the probability that waiting will cause loss of first mover advantage.
We see that one part of the opportunity cost is determined by the expected return on the underlying, as determined, partly, by the division of market power between the players in the game, while the other part is purely competitioninduced. The expected loss first mover advantage is still a very vague concept, and needs to be better understood. This, and how to actually estimate the opportunity cost, will be the subject of the next chapter.
The Opportunity Cost as a Function of Incomplete Information
Introduction
In the previous section I speculated that the total opportunity cost of deferring consists of two parts:
The foregone economic rent as given by the predetermined market power
A competitioninduced risk of losing the first mover advantage
So at this point what we need to ask is of course how these can be estimated.
About the foregone economic rent we can note that it concerns the value of the project without the option to defer, which follows from the fact that we are talking about returns that would accrue to the project were it already running (under some assumption about competition). As a sidenote: if the rate is negative, it means that without firstmover effects we would definitely defer and enjoy maximum optionvalue because we have no incentive to invest today.
To estimate the first mover advantage, as already hinted at, one possibility is to view it as a present value concept. We could ask the question: what additional cash flows would we expect as a result from getting the first mover advantage? That is, we need to know what is to be gained from being the first mover and how much it is worth in terms of present value. As managers, our guess is as good as any. Valuation is about finding objective estimates wherever possible, but using best guesses in circumstances where subjectivity is a must. This is evidently one of those instances.
Armed with estimates of both the foregone rate of return and the value of the first mover advantage we are one step closer to being able to make the valuemaximizing tradeoff between moving first and keeping the option to defer. In the next subsection these factors will be put in a game theory context with some perhaps unexpected results.
A Game Theoretic Framework
Having assumed that we know the present value of the first mover advantage and the foregone rate of return the next step is to evaluate how these factors change the strategic dynamic of the game.
The answer should be that they dont. With the standard assumption of common knowledge, the sizes of these variables should be known to all players and factored into the payoffs of the game. If the probability distribution of the risky exogenous factor is also known, the players will be able to apply the usual solution concepts to calculate their best move in such a way that a equilibrium outcome can be found.
Conjecture 5.3
In a game of perfect information with a stochastic exogenous variable affecting the payoffs and with a first mover advantage, an equilibrium outcome can be found (given that one exists) by applying standard game theory solution concepts. This equilibrium outcome will encompass the value maximizing tradeoff between moving first and deferring for all players and produce the value of the opportunity to invest at t = 0 for any given player.
The size of the opportunity cost is known and will be incorporated by the players like any other cash flow. For example in the game presented in section 5.3.3, we could just add the precondition that there is a first mover advantage with a PV of 50 and rework it using the same analytical tools to find the equilibrium. According to this, then, there will never exist an unexpected loss of first mover advantage because it is known from the beginning and traded off at t = 0, resulting in the equilibrium outcome.
We can think about it like this. If there is competition about an opportunity to invest, the only time we have to worry about this competition is when there is a first mover advantage (see Table 5.1). If there isnt one we can safely choose to defer, provided that the option value is larger than the value of the foregone economic rent, because demand will be divided up according to the predetermined market power of the players. Only a present value in the form of a first mover advantage could in such a case possibly induce a player to give up the option defer, so this is what we should zero in on.
So long as the PV of moving first is common knowledge, the game can be solved as usual because players will use this knowledge to trade it off against the option value and we would have equilibrium. Now heres the catch: we would only get this result in a world with complete information, in which there is no uncertainty whatsoever about the rules of the game. This would rarely be the case in an actual investment situation, which is the issue of the following subsection.
The impact of imperfect information
To repeat the previous conjecture: if all players know PV of the first mover advantage, this is part of the payoff in the game and it is possible to calculate the value maximizing action of each player. When all players are choosing their value maximizing decision given the rules of the game, what we have is the equilibrium outcome (this is assuming that the game is such that equilibrium exists).
The obvious problem is of course that such equilibrium would not be very robust. A robust equilibrium is one that remains intact even when the values of the model parameters are altered (within a reasonable range). If even a small change in one of the parameters results in a different equilibrium, the models prediction about play is not reliable.
And if we think about it, a model that requires player to accurately predict a stochastic, profitabilityaffecting variable as well as a vaguely defined first mover advantage (in terms of present value!) is bound to be fraught with uncertainty. Furthermore, players are required to have perfect information about factors like relative differences in efficiency and capacity. What this means is that we are faced with a game of incomplete information rather than complete information.
To derive the E(LFMA), how could we proceed? For starters, it would be necessary to make an estimate of the relevant rules of the game. They would include, among other things:
The probability distribution of the stochastic variable
Present value of demand in each scenario
The present value of the costs of each player
The present value of the first mover advantage
To keep things simple, in the rest of the essay I will focus on uncertainty about the size of the first mover advantage.
If we assume that the size of the first mover advantage, as estimated by us, is the correct one and that the other players are reaching the same conclusion, the equilibrium behavior of each player can be deduced. Then its time to allow for the inevitable uncertainty in the model. We realize that the other players could be making different estimates. For example, if they believe that the present value of the first mover advantage is in fact X and not Y, what happens to the equilibrium outcome?
There are four levels of complexity in games of incomplete information (see section 3.6) depending on what the uncertainty means for the strategy choice of the other player. What they have in common, though, is the need to estimate the probability of the other player being a certain type. That is, in this case we need the probability that our best guess about the other players estimate of the size of the first mover advantage is wrong.
The equilibrium will dictate a certain best strategy. If there were no uncertainty about the rules, this exercise would also indicate the value of the investment opportunity. But the fact that there is uncertainty means a probability that the game will be played out differently because competition might act in a way we thought they wouldnt. Stated differently, there is a risk we have guessed wrong about the conjecture competition has made, and this might lead us to make a decision at t = 0 that turns out to be suboptimal.
The only relevant case is when competitions estimate of the size of the first mover advantage (or any other relevant variable) differs from the equilibrium scenario to such a degree that it would induce competition to act differently from what equilibrium dictates. And some further thought tells us that the only time this would signify a economic cost is when there is a probability that we will loose out on the first mover advantage when equilibrium tells us that it is safe to defer.
Conjecture 5.4
The only case where there is an opportunity cost of deferring is when the optimality of the decision to defer hinges on competition also deferring. The opportunity cost will then be given as the probability that competition will act in this nonequilibrium way times the size of the first mover advantage.
Once one starts to think about it, this is a trivial insight. But to clarify things, consider the three mutually exclusive scenarios in Table 5.6.
Table 5.6 Attempt at validating conjecture 5.4
SCENARIOACTIONS IN EQUILIBRIUMScenario 1We invest at t = 0. The option game is overScenario 2The game analysis tells us it is in our best interest to defer regardless of what competition does (dominant strategy)Scenario 3The game analysis tells us that it is in the best interest of competition to defer at t = 0 and that it is in our best interest to defer given this fact
The idea is that there is always a probability that competition will not use their equilibrium strategy (because we have specified the game incorrectly), and if this means a probability of an undesirable outcome this should count as an economic cost and reduce the value of the opportunity to invest.
The only scenario where an opportunity cost might arise is in scenario 3 because in it the decision to defer is value maximizing if and only if competition also defers, leaving the first mover advantage intact until the subsequent stage of the game.
Now, to repeat, equilibrium tells us that we are safe in deferring. However, we know that the rules of the game in our analysis are only guesses and there is a probability that competition invests in spite of it not being in their best interest to do so (according to our game analysis). The opportunity cost of waiting will depend on our estimate of the probability that this undesired scenario does occur.
If the uncertainty in the game analysis regards a certain variable, for example competitions estimate of the size of the first mover advantage, there will be a given value of the variable at which competition will choose some other action than the one specified in equilibrium. Such a threshold value is usually called the trigger value.
To exemplify, denote the uncertain variable X. According to our analysis variable X will take on the value XH and this is common knowledge. The equilibrium we get using X predicts that both players will defer. But what if there is a probability that competition doesnt realize it is in his best interest to defer? And is there a probability he might be attaching some other value to variable X than XH? Let us say XQ is the value where it becomes optimal for competition to invest at t = 0 and that XQ(XH, that is, XQ is the trigger value. The question now is this: what is the total probability, in our view, of competition attaching the value XQ or higher to variable X? We know he shouldnt (according to our own estimates) but we know he could be. This probability is therefore equal to the probability of losing out on the first mover advantage, and we know we need to at least share in on this first mover advantage for our investment to be profitable.
At this point I think it could be useful to find a simple term for expressing this type of uncertainty, instead of the complexities associated with the terminology of incomplete information. Since what we are dealing with doesnt involve uncertainty about all aspects of the game, only competition, a narrowing down seems plausible. More specifically, the uncertainty is about factors like:
Are our competitors actually pursuing the goals that we think they ought to be? Can we count on them acting rationally? What is their true value system anyway?
Do they have the cost structure we think they have? Or do they really have the production capacity that our analysis suggests they have?
Can we be sure that they are expecting demand to develop at the rate we believe them to?
Does competition take it for granted that we know the rules of the game or could it be attaching uncertainty to how we perceive the game?
And so on. If we are unsure about some aspect ourselves, such as Who are the players in this game? we do not know exactly what the game we are in looks like. This is uncertainty about the rules of the game. But to be strategic uncertainty it must involve competition, and the relevant question with strategic uncertainty would be We know the list of players, but do our competitors really know who the players are?
Unable to find a definition of strategic uncertainty in the literature, I will use the following definition.
Definition 3.16 Strategic Uncertainty
Strategic uncertainty is uncertainty on our part regarding competitions perception about the rules of the game
Strategic uncertainty is worth defining because it addresses the important issue of how much faith we can have in our equilibrium analysis of an investment situation that has a combative aspect to it. Later in the essay I will argue that the opportunity cost of deferring can be derived from analyzing the strategic uncertainty in a game.
5.6 Valuing the Option to Defer under Strategic Uncertainty
5.6.1 A description of the approach
I have put forward the (notsorevolutionizing) thought that the opportunity cost under strategic uncertainty is intimately connected to the probability that waiting leads to a loss of first mover advantage. The risk of the predicted equilibrium outcome being wrong is, in some circumstances, equivalent to an economic cost that reduces the value of the investment. In this section I will develop one approach to deal with the problem, keeping it on a conceptual level. In section 5.6.1 the approach will be somewhat formalized.
But the analysis so far has been concentrated to a game theoretic framework and ignored the factor that makes option valuation superior (in the sense of estimating market value) to other option analysis tools, namely the correct discounting of uncertainty. To actually value the opportunity to invest, including the option feature as well as competition, we need to fit the data into an option valuation model.
What an opportunity cost means in practical terms is of course that the value of the investment is lower. The perhaps most logical way to adjust for strategic uncertainty would be to adjust the value of the underlying, that is, to create an alternative binomial tree with which to value the option, just as we do with dividends when we value a financial option.
The fact that the opportunity cost can be seen as a present value means that an adjustment could be made to the value of the opportunity to invest as calculated using the equilibrium scenario but without the option to defer. In what follows, I will assume that
There exists a first mover advantage
The situation corresponds to scenario 3 in Table 5.4
To proceed, I assume no flexibility (meaning that all players invest at t = 0) and make a game theoretic analysis. Use the equilibrium outcome to value the investment in line with the first of the fourstepprocess proposed by Copeland et al. It is never stated explicitly in their book, but a basecase NPV must be made under some assumption about competition. We then have
Base case NPV = NPV without flexibility under some assumption about competition
Normal Real Options theory would now have us constructing a binomial tree with this NPV as underlying to get the expanded NPV, calculated using the standard option pricing methodology of working backwards the tree. This renders us
Expanded NPV = NPV without flexibility + Value of option to defer.
But this is ignoring the effects of strategic uncertainty. The opportunity cost is, as argued previously, the probability that deferring will induce a loss in first mover advantage times the size of the first mover advantage.
If a second analysis is performed to allow for this uncertainty, we can investigate the effects on competitive equilibrium under various scenarios. It would necessitate that the PV of the first mover advantage is estimated, as well as the total probability of some variable reaching a trigger value that would cause this advantage to be lost. And since it will come out as a present value, we can simply subtract it from the valuation as indicated using the analysis.
This step means that we have corrected the investments value for the existence of competition, and we can use it as the underlying in the binomial lattice. We would then have an alternative event tree and if we value the option to defer using this alternative we end up with the value of the investment opportunity when both the flexibility and competition have been taken into account. This gives the value as
Expanded NPV adjusted for strategic uncertainty = NPV without flexibility + Value of the option to defer Expected loss in first mover advantage
A oneperiod binomial model
Some notation:
Z = Present value of first mover advantage
z = Differential in estimate of PV first mover advantage
q = Probability that waiting will cause Z to be lost
V0 = Base case present value of project under some assumptions about competition
( = dividend yield (expected loss in first mover advantage as a proportion of V0)
W = Value of option to defer
ER = Foregone economic rent
S#i = Strategy Defer for player i
Si = Strategy Invest for player i
S#i = Strategy Defer for all other players
Si = Strategy Invest for all other players
p = risk neutral probability of an upmove in the underlying
u = up parameter
d = down parameter
The up and down parameters, as well as p, are calculated using the MADmethodology presented in Copeland (2001) with V0 as underlying (see equations 4.1, 4.3 and 4.4).
For us to be, and for there to be an competitioninduced opportunity cost, in the situation characterized by scenario 3 in Table 5.6 we need the following conditions.
[1] W ( 0
[2] Z ( 0
[3] q ( 0
[4] S#i and S#i are equilibrium strategies
[5] W ( ER provided [4]
Assuming away q the investment project could be valued using Figure 5.1
Figure 5.1 Binomial Tree when q = 0
V0u
p
V0
1p
V0d
The value of the project is then the maximum of [V0, 0] and [p * V0u + (1p) * V0d]er* T
When q ( 0 we can create an alternative binomial tree that has a lower value of V, reflecting the fact that there is a possible loss in first mover advantage.
Figure 5.2 Binomial Tree when q (0
V(1()0u
p
V(1()0
1p
V(1()0d
But figure 5.2 specifies the trade off incorrectly, because if we invest at t = 0 q is irrelevant. So we should add back ( to the initial node in the tree. This is equivalent to assuming a dividend date occurring between t = 0 and t = 1, as a result of competitive uncertainty.
Figure 5.3 Binomial Tree with dividend date between t = 0 and t = 1
V(1()0u
p
V0
1p
V(1()0d
The value of the project with the option to defer is then [p * V(1()0u + (1p) * V(1()0d]er* T. If scenario 3 in Table 5.6 is at hand, condition 4 tells us that deferring is the optimal choice conditional on competitors also deferring. What Figure 5.3 asks is if deferring is still the optimal choice after allowing for the probability q that doing so will lead to Z being lost.
Why could this happen? For many reasons obviously, but one could be that competition is in fact attaching the value Z + z to the first mover advantage, enough to tip the equilibrium strategy from S#i to Si.In this case, q would be the probability that competition is estimating the size of the first mover advantage at Z + z.
5.7 Some Thoughts on the Usefulness of the Approach
The basic idea of this approach is to use game theoretic equilibrium as a prediction of play and let it guide the decision of investing or not investing by help of stating the first mover advantage in terms of present value. Then we analyze the strategic uncertainty in the model to check for possible factors that could induce competition to act in a way that will be costly to us, and in this way we will arrive at the opportunity cost of deferring.
If the model is to be useful, the three factors listed below must come together (and probably some more).
Nash Equilibrium must be a reasonably reliable tool in the first place for predicting what will happen (it must be reasonably robust)
It must be possible to make a reasonable estimate of the present value of the first mover advantage
We must be in a situation where the optimality of the decision to defer in fact depends on competition also deferring (see Table 5.6) because otherwise there can be no opportunity cost (at least according to the line of reasoning presented in this essay)
The first point is the subject of a lengthy debate among game theorists, and I will not present the arguments for or against here. But there are circumstances more conducive to Nash equilibrium being reliable, and they are presented in Table 5.7.
Table 5.7 Circumstances contributory to Nash Equilibrium being reliable
INDICATOREXPLANATIONConcentrated competitionAmplifies the marginal effects of one rivals choices on anothersMutual familiarityIncreases the likelihood and speed of convergence of adaptive, dynamic learning processes to an equilibrium point Repeated interactionIncreases the likelihood that competitors will learn to achieve a particular outcome as a result of an adaptive, dynamic learning processConsistent strategic rolesCompetitors pursue their given strategic roles within the industry, for example leader and follower, and this reflects their behaviorStrategic complementarityThere can be local complementarity among competitors, for example the optimal response to a price increase is often a price increase
From what I can tell, the consensus seems to be that Nash Equilibrium can be used as a starting point but that the analyst must be cautious in applying it and well aware of its limitations. And the approach presented here asks nothing more of the Nash Equilibrium than that: that it be used as a starting point for prediction of play.
The second point is very interesting. It undoubtedly puts some strain on the analysts creativity, but I think it could be a useful framework for thinking about the whole situation because it begs questions like: Why do we really want to rush off and invest now? What is to be gained from it that cannot be recouped should someone else beat us to the punch? Is the risk of losing this gain enough to make us want to bear the uncertainty it means to go now instead of later when we know more?
What it would mean in practical terms is conjecturing a scenario that answers how big the cash flow differential over the relevant time horizon will be, should we happen to be the first player on the field and get all the benefits this means. It could also be about the cash flow differential from at least not being the last one to enter the game (sharing in on the first mover advantage). It would also involve discounting these cash flow differentials at some appropriate rate. Maybe this could be the subject of a new, more practically oriented essay?
As for the point three, my limited experience stops me from hypothesizing about the value of first mover advantages in real world situations (as it should have stopped me from writing this thesis). But isnt it a general wisdom that the big returns goes to the original, whereas the copycats that move in later only make marginal returns, because by now everyone knows what a great idea it is so everyone is doing it?
SUMMARY
The problem that motivated this essay was that Real Options Valuation currently lacks a way of factoring in competitive aspects in valuing investment opportunities. Game Theory has been recognized as the most potent analytical tool for achieving the searchedfor integration of correct valuation techniques based on noarbitrage arguments and strategic considerations resulting from the existence of competition.
Game theory offers the concept of equilibriumbased predictions of play, and it relies on the logic of players acting in their best interest. This kind of equilibrium is a desirable because it allows an objective approach to analyzing competitive interaction, which is more compatible with the neutral arbitrage based principles of option pricing.
The option to defer is valuable because it gives the investor protection against downside risk. The investor will only choose to invest if uncertainty at some future point in time resolves in a way that gives the investment a positive Net Present Value. There are two factors counterbalancing the incentive to defer: 1) The investment may be inthemoney today, so that deferring means foregoing an economic rent and 2) There may be a first mover advantage in the game that cannot be recouped by later entrants.
It was suggested that there are two possible sources of this first mover advantage. The first is the commitment value that comes about as a consequence of being able to commit to a certain action. The second is an intrinsic part that results from the sheer fact of being first, for example brand recognition, being able to set the product standard or getting the best distribution channels.
It was then argued that the combined value of the first mover advantage can be estimated by management and expressed in terms of present value. This trick then allows us to add the present value of the first mover advantage as a rule of the game. Assuming perfect information, equilibrium can be found in which each player makes the valuemaximizing trade off between the value of the option to defer and the cash flow benefits of investing immediately
In a world of no uncertainty this equilibrium analysis would indicate the optimal course of action and the value of the investment of under this scenario. There would be no opportunity cost as such in such a world because the perfect information makes it possible to factor in all aspects of competition and optimally trade them off in order to guide our decision.
BIBLIOGRAPHY
Literature
Bernstein P, Against the gods (1998), Wiley & Sons
Brandenburger A & Dixit J, Coopetition (1996), HarperCollins
Copeland Tom, Real Options A Practitioners Guide, (2001), Texere
Dixit A & Skeath S, Games of strategy (1999), WW Norton
Dixit A & Nalebuff B, Thinking strategically (1993), WW Norton
Dutta, Strategies and Games, (1999) MIT Press
Ghemawhat P, Games businesses play (1997) MIT Press
Grenadier S (ed), Game choices The intersection of Real Options and Game Theory (2001), Risk Books
Hull J, Options, futures and other derivatives, 4th edition (2000), Prentice Hall
Rasmussen E, Games and information: An introduction to Game theory (1994) Blackwell Publishers
Ross S et al, Corporate finance, 4th edition (1999), Richard D Irwin
Trigeorgis L, Real Options: Managerial flexibility and strategy in resource allocation (1996), Praeger Publishers
Articles
Trigeorgis (ed) Real Options An Overview, Real options in capital investment (1995)
Kutilaka N & Perotti E, Strategic growth options, (1998) Management Science 44/8
Smit H & Ankum LA, A real options and game theoretic approach to corporate investment under competition (1993) Financial Management, Autumn Issue
Gehr A, Riskadjusted capital budgeting using arbitrage (1981) Financial Management, Winter issue
Ross S, Uses, abuses and alternatives to the NetPresentValue rule (1995) Financial Management, Autumn issue
Ingersoll J & Ross S, Waiting to invest: investment and uncertainty (1992) Journal of Business, vol.65
Olmsted E, Methods for Evaluating Capital Investment Decisions under Uncertainty, Real options in capital investment (1995) Trigeorgis (Ed)
Bernstein P, Against the gods (1998)
Dutta, Strategies and Games, (1999)
Dutta, Strategies and Games, (1999)
Copeland Tom, Real Options A Practitioners Guide, (2001)
Ross S et al, Corporate finance, 4th edition (1999), Richard D Irwin
See Copeland Tom, Real Options A Practitioners Guide, (2001), Texere
See introduction in Grenadier S (ed), Game choices The intersection of Real Options and Game Theory
Kutilaka N & Perotti E, Strategic growth options, (1998)
Ibid
Ibid
Smit H & Ankum LA, A real options and game theoretic approach to corporate investment under competition (1993)
Dutta, Strategies and Games, (1999)
Smit H & Ankum LA, A real options and game theoretic approach to corporate investment under competition (1993)
Jerker Holm, Lund University
Trigeorgis L, Real Options: Managerial flexibility and strategy in resource allocation (1996)
Copeland Tom, Real Options A Practitioners Guide
See section 4.4.2
This section is based on chapter 2 in Dutta, Strategies and Games, (1999)
Dixit A & Nalebuff B, Thinking strategically (1993)
These definitions are from Rasmussen E, Games and information: An introduction to Game theory (1994)
This section is based on chapters 3,4 and 5 in Dutta, Strategies and Games, (1999)
This section is based on chapter 2 in Rasmussen E, Games and information: An introduction to Game theory (1994)
This section is based on chapter 11 in Dutta, Strategies and Games, (1999)
This section is based on chapter 13 in Dutta, Strategies and Games, (1999)
This section is based on chapter 20 in Dutta, Strategies and Games, (1999) and chapter 2 in Rasmussen E, Games and information: An introduction to Game theory (1994)
This section is based on chapter 3 in Hull J, Options, futures and other derivatives, 4th edition (2000)
From lecture notes, Theory of Options, fall 2000, Lund University
From lecture notes, Theory of Options, fall 2000, Lund University
This section is based on section 9.1 in Hull J, Options, futures and other derivatives, 4th edition (2000)
This section is based on sections 9.2 and 11.6 in Hull J, Options, futures and other derivatives, 4th edition (2000) as well as on chapter 5 in Copeland Tom, Real Options A Practitioners Guide, (2001)
This section is based on chapter 9 in Copeland Tom, Real Options A Practitioners Guide, (2001) and chapter 10 in Hull J, Options, futures and other derivatives, 4th edition (2000)
This section is based on section 16.3 in Hull J, Options, futures and other derivatives, 4th edition (2000) as well as on lecture notes, Theory of Options, fall 2000, Lund University
This section is based on chapter 1 in Copeland Tom, Real Options A Practitioners Guide, (2001),
This section is based on chapter 4 in Copeland Tom, Real Options A Practitioners Guide, (2001) and on Gehr A, Riskadjusted capital budgeting using arbitrage (1981)
This section is based on chapters 4, 8 and 9 in Copeland Tom, Real Options A Practitioners Guide, (2001)
Copeland Tom, Real Options A Practitioners Guide, (2001)
Trigeorgis (ed) Real Options An Overview, Real options in capital investment (1995), Praeger Publishers
Smit H & Ankum LA, A real options and game theoretic approach to corporate investment under competition (1993)
Ibid
Ross S, Uses, abuses and alternatives to the NetPresentValue rule (1995)
Ibid
Ibid
This section draws on Grenadier S (ed), Game choices The intersection of Real Options and Game Theory (2001) and chapter 4 in Copeland Tom, Real Options A Practitioners Guide, (2001)
Smit H & Ankum LA, A real options and game theoretic approach to corporate investment under competition (1993)
This section is based on Smit H & Ankum LA, A real options and game theoretic approach to corporate investment under competition (1993)
This section is based on Smit H & Ankum LA, A real options and game theoretic approach to corporate investment under competition (1993)
See for example chapter 7 in Dixit A & Nalebuff B, Thinking strategically (1993)
This is a very interesting case, see Kutilaka N & Perotti E, Strategic growth options, (1998)
Hull J, Options, futures and other derivatives, 4th edition (2000)
Olmsted E, Methods for Evaluating Capital Investment Decisions under Uncertainty
Hull J, Options, futures and other derivatives, 4th edition (2000)
Hull J, Options, futures and other derivatives, 4th edition (2000)
Ghemawhat P, Games businesses play (1997)
Ghemawhat P, Games businesses play (1997)
Dixit A & Skeath S, Games of strategy (1999)
PAGE 4
PAGE 11
,R_`suv
56789IJdefgh89:;<պհզ՜ՒjqUmHjUmHjwUmHjUmHj}UmHjUmH
jUmHmHj56UCJmHCJ$CJ,5CJ0CJ0mHCJmHmH569,RSTUVWXYZ[\]^_`uv$
f#,RSTUVWXYZ[\]^_`uv:i=Jz= X .
k
S0o*
j
V!SG{eCnpr_:i=Jz=
f#
Xf#
f#`<pq)*DEFHIYZtuvxy789;<lmjUmHjYUmHjUmHj_UmHjUmHjeUmHjUmHjkUmHjUmH
jUmHmH=
7 8 R S T V W k l
(
)
*
,

J
K
e
f
g
i
j
j
UmHjA
UmHj UmHjG UmHjUmHjMUmHjUmHjSUmH
jUmHmH== X .
k
S0o*
j
V!S
f#
f#
Xf#
23MNOQR*+,./NOijkmn
j#UmHjUmHj)UmHj
UmHj/
UmHjUmHj5UmHjUmHmH
jUmHj;UmH=
$
%
&
(
)
I
J
d
e
f
h
i
56PQRTUz{ 23MNOQjUmHjUmHjUmHjUmHjUmHjUmHjUmHjUmHjUmHmH
jUmH=QRqr&'ABCEFZ[uvwyzDE_`acdjUmHjpUmHjUmHjvUmHjUmHjUmHjUmHjUmHjUmHmH
jUmH=SG{eCnpr$d
$
&F7d
$d
f#
Xf#
f#"#=>?ABMNhijlmnoqdk
()`aABDE ""$$%%...33<4=45588EEEFFKGLGGG4H5HNHHHJIaIIIJKH
j0JU656j56UjUmHjjUmHmH
jUmHObcdvwQRGH!!""$$$$½}xsnid[V7
`a7
!"tu7 bcdvwQRGH!!""$$$$E&F&f&
&F7
$
&F7d$d$E&F&f&g&
'(@)A)B)D)K)L)M)N)s)t)****** ,!,. .//Z1[1\1½{vqlgb]XSijEF.
7 7
97
of&g&
'(@)A)B)D)K)L)M)N)s)t)****** ,!,. .$
&Fd
$
&F7d
$
&F7d$d .//Z1[1\11122v5w5888<8=8996;7;<<==Z>[>Q?
$
&F7d$d\11122v5w5888<8=8996;7;<<==Z>[>Q?R?S?o?p?@@%B&BDznid_ZUP.7
{7
/HI7
Q?R?S?o?p?@@%B&BDDEEEEEEEEFFKGLG\G$
&F d
$
&F7d
$
&F7d$dDDEEEEEEEEFFKGLG\GxGGGG6H7HNHHHHJIKIaIIIIJ~ytoje`[V,cyzv)
L
h
x y7 7
4
5
\GxGGGG6H7HNHHHHJIKIaIIIIJ
J#JJJJJKKKd$d$
&F dJ
J#JJJJJKKK LLLLKMLMyMzMMMMNNOOQQQ'QZQ[QRRĿ~ymgb]Xi
7
6JKxy J#JJKK LLLLLLLLMMMNN%Q&QRRRSUSVTVaVbVcVdVVVVVVVVVVYZ]]]]]]]]]]]]]]]]^^^ ^^^^"^#^;^D^b^d^8_S_T_X_Y_`_c__}_~___________H* j>6 j6 jP6
j0JU6H*6YK LLLLKMLMyMzMMMMNNOOQQQ'QZQ[QRRRSS
$
&F7d$ddRRSSUU`UaUURVSVVVVVV3WWWdYeYYZZe[f[}[~[4\5\p]q]]Ŀvqlgb]X12
CD
C Dno !Fh SUU`UaUURVSVVVVVV3WWWdYeYYZZe[f[}[~[4\5\$
&Fdd$d5\p]q]]]^^__8_____7`W`X`paqaraaabbbb
$
&F7d$
&F
dd$d]]^^__8_____7`W`X`paqaraaabbbbccdd(e)eeeeeǽzupkfa\WRIJst7
C/
p
q.OPOPz _______W`X`aabc.cccddd(ee$fgffffggukvkImJmopqqar)stt[tdtvvwx{yefjklς#5хۅ]ijՑ֑hH* jH*j0J6U jP56
j0JU66H*Ubccdd(e)eeeeegfhfiffffffgghghh0j1jUkVkd
$
&F7d$degfhfiffffffgghghh0j1jUkVkWkwkxk\l]lllllJmKm?o@oo½xnd_ZUP
j
7
^897
L
{VkWkwkxk\l]lllllJmKm?o@ooooppqqqqqqqd$
&Fd
$
&F7d$doooppqqqqqqqqqqqq1r2rar)s*sssssttbtctdttĸ~tje`[VQ
#
> ?
7
qqqqqq1r2rar)s*sssssttbtctdtttuu$
&Fd$
&Fdd
$
&F7d$dttuuPvvvxx}y~yyyyyzzzz{{Dyz}}~~~½~ytoje]XS7
:
k
Eh
%
[ \uPvvvxx}y~yyyyyzzzz{{Dyz}}~
$
&F7d$
&Fd$d$
&Fd~~~~~J~K~[~m~~~~~$elmd$d
&F7~~~J~K~[~m~~~~~$elmς#Ŀwrmf\WRg
7
hyz 3DY7 ς#1ṁڇۇ$
&Fd$
&Fd$
&Fd$d1ṁڇۇً½zupd_ZUP
7
P
yz
Y
ًڋ]^d
&F7$dًڋ]^ʒ˒
ƕǕߕ~ytoje`[V
DE!ʒ˒
ƕǕߕ×ėٗڗ#$$d$
&Fd$dhi͓̓12ɕʕӕՕٕؕݕޕƗǗȗɗ͗Ηїҗӗԗח').0;=?Bch 7<%+ߡ jsH* jm6
j0JUH* jPH* jDW×ėٗڗ#$CDZ[xyzܙݙ+)*LĿzupkfa\WRPQnoO
7
x12HIhi $CDZ[xyzܙݙ+)*LMٜڜ$
&Fd
&F7$dLMٜڜopܟݟ9:ģţopcd%&ک۩~ytoje`[VLM@A0
v w'(
yz.!ڜopܟݟ9:ģţopcd%&ک$
&Fd$dک۩ͬάϬЭѭ78UV>?OP$
&Fd$dmnٳڳ12Z[\]^a߶4G}~MOYZef
FGKR6734j0J6UH* jd j6
j0JU jsH*H* jDH* jH*SͬάϬЭѭ78UV>?OP5ٳڳ)*Vrhc^YT%
8 9yz%&5ٳڳ)*VWtuٶڶ z{ͻλd$d$
&FdVWtuٶڶ z{ͻλȾɾ ! !{vqe`[7 q78ghUpq23Z[xy!Ⱦɾ ! !\]{$
&Fd
$
&F7d$d\]{ef@ab½qlgb]XSN{ RS7
.VW


K L01{ef@ab.d
$
&F7d$
&Fd$
&Fd$db.uJ~FGYZef
½~ytoje`[V>y`a)LM
g
z .uJ~FGYZef$dd$d$
&Fdf
IJk}Bq89$
&Fd$
&F.d$ddIJk}Bq89B4½xnd_ZWRM
9
,67n
=>Db.
t.
. 9B45
5
$
&Fd$d$
&F d45
56 +,GH<=Ŀxnid_ZU!
!!
*! +7UV
)*DE!56 +,GH<=!"$
&F!dd$d4$xCvwu%O
)nopdu+,DM~
!"##1#2#J#K#k#l##(((())++,,@/////E0W0q000558899KKVMWMN j>CJ5CJ5CJ5566CJCJ
j0JU6V=!"OPCDvwuv
ƿ~ytoje`[cdRbc?OPGWX/
/
6kl "OPCDvwuv
1$
&F"dd$
&F/d$d
1:;RSjk
%N
O
lmeo"/ȿ~ytoje`[VQjkpT%&2 "
"
"
" :;RSjk
%N
O
lme$d%d&d'dD d
$
&F2d$d$
&F"do"/0op$%ef
&F2d$d$d%d&d'dD /0op$%efw.~½ukfa\WRM8gh0
C0
g0
0
0
2
9fw.~ !!6"7"""""" $d$d$d$
&F0d !!6"7""""""## ##1#2#5#9#J#K#O#R#k#l#q#u#Ľ}yrkd`YR
6Itu"## ##1#2#5#9#J#K#O#R#k#l#q#u#######ִdքP
$
&F2d$d$$F
# $d$u#######E%F%''((=(>(((((())))W***+++,,,{vqlgb]X+DEjk<=2
#E%F%''((=(>(((((())))W***+++,,$d$d%d&d'dD d$d,,@BD/M/N///////////////LX$$F
# $d$d$d$d$d%d&d'dD ,@BD/M/N///////////////00E0W0^0i0j0q0{00Ľ~wpie^WP
$
+
={
45}>@B/00E0W0^0i0j0q0{00000000022+4,455d\$$F
# $d$d$d0000000022+4,45588888A999$;%;<=o>p>?@@@0A1AaBĿwrmhc^YTQRBC]^A#
# z{VW
588888A999$;%;<=o>p>?@@@0A1AaBbBCCCEE
$
&F2d$
&F#d$daBbBCCCEEG.GiHjHkHHHJJ[L\LYMZMNNXOYOOPQ QRRRR¿wmhc^YThiKLM$
$
$ xyYZ)*S2
!EG.GiHjHkHHHJJ[L\LYMZMNNXOYOOPQ QRRRR$
&F$d$dNNNNVOWOXOYOOOQ QSSUUVVO]]]]aajxj{jktt{{Zlւׂ6CJIJVT=I~آj2312TVg jd56 j>CJ5CJ5mHH* j6
j0JU jTR&S'SSSTTMVNVxXyX%Z&Z\\]]O]]]]]]^^^``````{vqle`T3 3 OOPEFrsD%
E
% R&S'SSSTTMVNVxXyX%Z&Z\\]]O]]]]]]^^^`d$d$
&F%d```````@aAaaaabbcc3f4fgggglhmhj$
&F&d
$
&F3d
&F3$d``@aAaaaabbcc3f4fgggglhmhjjjkkmm8p9prr:r½}xsnid_Qu3
uwSTp 3
>B&
& jjjkkmm8p9prr:r;rss+u,uwwwww%xSxxx$
&F'd
$
&F3dd$d:r;rss+u,uwwwww%xSxxxxxzz~~¸wrmhc^YR
`$%WfgQRgh#$'
'
'
R' SJKxxxzz~~ւׂ4$$F
t0# $d$d$dւׂ678hideR<=ƏǏ !ÿwmh^YO
O
P
z {
y
4
?@
W678hideR$d $$F
t# 4$$F
t0# $d$<=ƏǏ !PQUVd$
&Fd$dPQUV"#*+01VTU<=}xsnid_ZU^_5
5 *+
jk1WXjk!"#*+01VTU<=ef:;עآjk$
&F5d$def:;עآjkã1Ԥ
/R~
`alzupkfa\W
]^q<^z:oB6 hi!kã1Ԥ
/R~
`almwx$d
&F6ghst~=>GHRSvw~ !
#$%&nopqqsy{=>"#ĵŵf&'l~CJ5CJ
j0JU6 jdH*H* j>ZlmwxʧͧΧ;<@ABDEFIJKOPQUVWXRĿzsoje`[V
!"
%&'
)*+
/0TU!ʧͧΧ;<@ABDEFI$$FO$$Fhif $d$$dIJKOPQUVWXRSTUyz$$FJKr$$FZ $d$$d$$FRSTUyz~wsngc^W
WXY
[\]
fg
z$$FZ$$F$$FO$d$$Fhif $d$
$$FO$$Fhif $d$$d
f$d$$FJKr()*+,\]GFG?@½wrmhc\U
f
p,
,
, tuvw@ABC
LMN
RST()*+,$$FJKr$$FZ $d$$d$$F,\]GFG?@t$$F# $d$$
&F,d$d"#>ĵŵߵfghiԻջyz{~ž~ytoje`[Vyz}~#$PQ=>
34
`
uv
@
YZ"#>ĵŵߵfghiԻջyz{~؈؈$d$$F# $d$()CQRS^_PQZ[
S~{xurolif56NO+,pq
?rsYZ3
x(()CQRS^_PQ$d
$
&F3d,>g~Gc/02:;bB\Gz%GHU&'/EMN]
D6>?G]dey
j0JUCJH*6_QZ[
ST)*{qrST)*{qrHI&M>dT N@y3BZ&\p#hFr¿
bc45pqz{KHI&M>dT N$dyMTUd !JGNO~.S@A.VWYab>fgiqryz3H*
j0JU6`@y3BZ&\p#hFr34i;BCZ[o &'.WT\]hpq#$UVX`ahiv346>?FGUkrsH*6
j0JU` "24"& " " "
lnfhRVZlnfhRVZ0J#mH0J#
j0J#U
j0JU6DpzPRXZFH,.p$
&F d$
&F.d
&F:$d"
&
00P. A!"#$% asset (the underlying)BuySell4.5
4.8 4.9 Carrying t4.9.1 4.9.1
31
}DyK_Toc521815160}DyK_Toc521815161}DyK_Toc521815162}DyK_Toc521815163}DyK_Toc521815164}DyK_Toc521815165}DyK_Toc521815166}DyK_Toc521815167}DyK_Toc521815168}DyK_Toc521815169}DyK_Toc521815170}DyK_Toc521815171}DyK_Toc521815172}DyK_Toc521815173}DyK_Toc521815174}DyK_Toc521815175}DyK_Toc521815176}DyK_Toc521815177}DyK_Toc521815178}DyK_Toc521815179}DyK_Toc521815180}DyK_Toc521815181}DyK_Toc521815182}DyK_Toc521815183}DyK_Toc521815184}DyK_Toc521815185}DyK_Toc521815186}DyK_Toc521815187}DyK_Toc521815188}DyK_Toc521815189}DyK_Toc521815190}DyK_Toc521815191}DyK_Toc521815192}DyK_Toc521815193}DyK_Toc521815194}DyK_Toc521815195}DyK_Toc521815196}DyK_Toc521815197}DyK_Toc521815198}DyK_Toc521815199}DyK_Toc521815200}DyK_Toc521815201}DyK_Toc521815202}DyK_Toc521815203}DyK_Toc521815204}DyK_Toc521815205}DyK_Toc521815206}DyK_Toc521815207}DyK_Toc521815208}DyK_Toc521815209}DyK_Toc521815210}DyK_Toc521815211 asset (the underlying)BuySell4.5
4.8 n uncertain world. That the NPV method has come under intense criticism is4.9 Carrying t4.9.1 4.9.1 . I4.104.11 Samuelsson s result and its relevance for real option valuation is fully explained in chapter 8 in Copeland (2001)
49
J$&FHJ,.0246FHJ,.02$
&F d$
&F.d
&F:$d asset (the underlying)BuySell4.5
4.8 n uncertain world. That the NPV method has come under intense criticism is4.9 Carrying t4.9.1 4.9.1 . I4.10
4.11Proposition 5.3 Source: The authorCorollary Source: The authorin Conjecture 5.2Conjecture 5. asset (the underlying)BuySell4.5
In Figure 4.1 a one period binomial tree is depicted, although without the lines that usually connect the nodes.
Figure 4.1 Binomial tree Source: The author
SSS
4.8 n uncertain world. That the NPV method has come u3 Source: The authorDefinition 5.6nceptual level. In section 5.6.26 (p.48) Source: The author Source: The author Source: The author Samuelsson s result and its relevance for real option valuation is fully explained in chapter 8 in Copeland (2001)
53
asset (the underlying)BuySell4.5
In Figure 4.1 a one period binomial tree is depicted, although without the lines that usually connect the nodes.
Figure 4.1 Binomial tree Source: The author
SSS
4.8 n uncertain world. That the NPV method has come uOh+'0x
(4@
LX`hp1 INTRODUCTION INPerssonersNormalLasse Svensson14sMicrosoft Word 8.0@@Հ@NXU@lHBycrosoft Office0Mallar5Ȕ@tCq
@<@՜.+,D՜.+,@hp
Privat&
1 INTRODUCTIONTitle 6>
_PID_GUIDAN{C28F01C486CC11D59018B59801341654}tCq
@<@ "24"& " " "
lnfhRVZlnfhRVZ0J#mH0J#
j0J#U
j0JU6GpzPRXZ$
&F d$
&F.d$d
possible to those new to these theories to understand the topics discussed could only
7
4dT1 INTRODUCTIONPerssonLasse Svensson$
[$@$NormalmH F@FRubrik 1$<@&5CJKHOJQJD@DRubrik 2$<@&56CJOJQJ>@>Rubrik 3$<@&CJOJQJBA@BStandardstycketeckensnitt2P2
Brdtext 2dx@Y@DokumentversiktD OJQJ,B@,BrdtextxCJD>@"DRubrik$<@&56CJ$KHOJQJ.+2.
Slutkommentar2&@A2FotnotsreferensH**@R*Fotnotstext6@6
Innehll 1x
56CJ&/r&Lista
*
*Index 1
8.!@.Indexrubrik588
Innehll 2
x5CJ,@,
Innehll 3,,
Innehll 4X,,
Innehll 5 ,,
Innehll 6,,
Innehll 7,,
Innehll 8 x,,
Innehll 9!@, @",Sidfot
"
p#")@1" Sidnummer
(`AD R(+).a/%063@BYCJL\afmzݓ
ިRt>418'GbIJbKp
!"#$%&'()*+,./012345667(OvEg}Iw$i \
kNi
p
@TL*V,RSTUVWXYZ[\]^_`uv:i=Jz=X.kS0o* j V
!SG{
e
CnprbcdvwQRGH E"F"f"g"
#e$f$g$i$p$q$r$s$$$%%%%%%E'F').) +
+,,,,,008393:3a3b344[6\6778899v:w:x:::;;J=K=C?D?@@@@@@@@BBpBqBBBBBB[C\CsCCCDoDpDDDDD1E2EHEEE#F$FCFGG.GGGGpHqHHHHIIIIKK7L8L9LLLLLMMNNN+P,PPPPwQxQQQQR RXRRRTTTV VVVVVYWZWXXXY;Y[\~23gh,ST ,)*VW[\]_`adefjklpqrs9yz!"23ǮȮ
9:WXfg{
ԽսֽNOtumnWX2XY$%lAvw>?QR]^ABcu@o67@A34
34)*EF:; MNABtust
/89PQhi#L M j
k
w
4AB}12&rs:;CD>?BFWX\_xy~R!S!##$$%$J$K$$$$$$%%%&{&' ''''(()d)f)h))+q+r++++++++++++++,*,+,i,{,,,,,,,,,,,,,,..70801244444M555071799{::K<L<<===m>n>???AA9C:CuDvDwDDDFFgHhHeIfIJJdKeKK*L+M,MNNNN2O3OOOQQYRZRTT1V2VXX'Y(Y[YYYYYY(ZZZ\\\\\\\L]M]]]]"^#^__?b@bccccxdydff(fggiiDlEl!n"nFnGnoo7q8qsssst1t_tttu uvvxxzz}}~~~~~~./:VWXr34\]@Aˌ̌pqގߎtuʐː̐ ./ABޔߔIJOPu)*z{bc`a"W3Uxѡ 34ޢߢףأ<=tuyz{}~ťƥϥХѥӥԥեݥޥߥcdmnoqrsvwx}~
STUVWPQBZ[n#$>ŲƲǲȲ34عٹڹ۹ݹ߹no4512vwST
lm&'(12mn?@56^1{%,4SI;GZQ.Gv=`rrr7 7 ddddddddddddddd7 7 F"F"F"F"F"F"7 i$i$i$7 i$N%N%N% N% N% N%N%N%N%N%N%N%N%N%N%N%7 D%\\\\\\\\7 D%444444444444447 D%S;S;S;S;S;S;S;S;S;7 AAA7 AAAAAA A A A AAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAA7 A(M'M'M'M'M'M'M'M'M'M'M'M'M'M'M'M'M'M'M 'M 'M'M'M'M'M'M'M'M'M(MfWfWfWfWfWfWfWfWfWfWfWfWfWfWfWfW
fW
fWfWfWfWfW7 Ar]r]r]r]r]r]r]r]r]r]r]r]r]r]r]r]r]r]7 Aibibibibibibibibibibibibib7 AWgWgWgWgWg Wg Wg WgWgWgWgWgWgWgWgWgWgWgWgWgWgWgWgWg7 Am(mmmmmmmmm m m mmmmmmm m m m mmmmmm(m~u~u~u ~u ~u ~u~u~u~u~u~u~u~u~u7 yy7 yzzzzzzzzzzzzzzz7 y{{{{{{{{{{ { {{{{{{ { { {{{{{{ { { {{{{{{{{{{{7 y asset (the underlying)BuySell4.5
4.8 n uncertain world. That the NPV method has come under intense criticism is4.9 Carrying t4.9.1 4.9.1 . I4.10
4.11Proposition 5.3 Source: The authorCorollary Source: The author Samuelsson s result and its relevance for real option valuation is fully explained in chapter 8 in Copeland (2001)
39
asset (the underlying)BuySell4.5
4.8 n uncertain world. That the NPV method has come under intense criticism is4.9 Carrying t4.9.1 4.9.1 . I4.10
4.11Proposition 5.3 Source: The authorCorollary Source: The authorin Conjecture 5.2Conjecture 5.3 Source: The authorDefinition 5.6nceptual level. In section 5.6.26 (p.48) Samuelsson s result and its relevance for real option valuation is fully explained in chapter 8 in Copeland (2001)
49
nder intense criticism is4.9 Carrying t4.9.1 4.9.1 . I
4.10
flexibility
4.11Proposition 5.3 Source: The authorCorollary Source: The authorin Conjecture 5.2Conjecture 5.3 Source: The authorDefinition 5.6nceptual level. In section 5.6.26 (p.48) Sourc nder intense criticism is4.9 Carrying t4.9.1 4.9.1 . I
4.10
flexibility
4.11Proposition 5.3 Source: The authorCorollary Source: The authorin Conjecture 5.2Conjecture 5.3 Source: The authorDefinition 5.6nceptual level. In section 5.6.26 (p.48) Source: The author Source: The author Source: The author Samuelsson s result and its relevance for real option valuation is fully explained in chapter 8 in Copeland (2001)
33
pzPRXZ$
&F d$
&F.d$d$
[$@$NormalmH F@FRubrik 1$<@&5CJKHOJQJD@DRubrik 2$<@&56CJOJQJ>@>Rubrik 3$<@&CJOJQJBA@BStandardstycketeckensnitt2P@2
Brdtext 2dx@Y@@DokumentversiktD OJQJ,B@,BrdtextxCJD>@"DRubrik$<@&56CJ$KHOJQJ.+@2.
Slutkommentar2&@A2FotnotsreferensH**@R*Fotnotstext6@6
Innehll 1x
56CJ&/@r&Lista
*
@*Index 1
8.!@.Indexrubrik58@8
Innehll 2
x5CJ,@,
Innehll 3,@,
Innehll 4X,@,
Innehll 5 ,@,
Innehll 6,@,
Innehll 7,@,
Innehll 8 x,@,
Innehll 9!@, @",Sidfot
"
p#")@1" Sidnummer
(`AD )*/<014AB4D%M]bugm{}ĆŊOxH=JK=Lqw`
!"#$%&'()*+,./012345667(OvEg}Iw$i \
kNi
p
@TL*V,RSTUVWXYZ[\]^_`uv:i=Jz=X.kS0o* j V
!SG{
e
CnprbcdvwQRGH E"F"f"g"
#$@%A%B%D%K%L%M%N%s%t%&&&&&& (!(* *++Z[\..v1w1444<4=45567778899Z:[:Q;R;S;o;p;<<%>&>@@AAAAAAAABBKCLC\CxCCCC6D7DNDDDDJEKEaEEEEF
F#FFFFFGGG HHHHKILIyIzIIIIJJKKMMM'MZM[MNNNOOQQ`QaQQRRSRRRRRR3SSSdUeUUVVeWfW}W~W4X5XpYqYYYZZ[[8[[[[[7\W\X\p]q]r]]]^^^^__``(a)aaaaagbhbibbbbbbgchcdd0f1fUgVgWgwgxg\h]hhhhhJiKi?k@kkkkllmmmmmmmmmmmm1n2nan)o*oooooppbpcpdpppqqPrrrtt}u~uuuuuvvvvwwDxyxzxyyzzzzzJzKz[zmzzzzz{${{{{{{{}}~~~~:H !ʃՃƆǆȈfgǊȊӎԎ
ϑБ͓̓,LMcd867YZ
}BC͟Οxylm./12678:;<?@AEFGKLMNTUx
h23ABVWr_`Ͼо)*OPHI23
\]34GefQR,89t>Pstz{J !{()Z[kbcsOP`NO
+,CD'
(
EFR
\WX
MN_fgvwhi ! 2 3 7 : S T Y ] g h i j u v "."t$u$$%%%&%%%%%%&&&&V'''((()))?*A*C*,L,M,,,,,,,,,,,,,,,DV]hipz//11224445}5(6h6i68899V;W;&='=>>H?I?@@@BBDDPEQEREEEsGtGBICI@JAJKK?L@LLMNNNNxOyO
PPPPQQ4S5S_U`UW
WYYZZ6ZyZzZZZZ[[[p]q]r]]]]]'^(^n^^^^^``ccrdsdddSeTeffghhjjm mnn!o"opprrsstttu:uiujuuuwwyy{{}}}~~ln
123cd_`M78KLߏOP
$%*+PUV=>fg;<؟ٟklĠ2ա0SabmnxyˤΤϤOPTUVXYZ]^_cdeijklƥǥfghiɦʦͦΦ>?HIJLMNQRSWXYbcde./012fg+,yz°̰ذٰ56Iұ\]xĺźbcIJ}
QR./GH
cdHI9`Vx.$Z"5z, : @@@@@@@@: e: The author Source: The author Source: The author Samuelsson s result and its relevance for real option valuation is fully explained in chapter 8 in Copeland (2001)
30
(оооооооооо оо оо оо$
[$@$NormalmH F@FRubrik 1$<@&5CJKHOJQJD@DRubrik 2$<@&56CJOJQJ>@>Rubrik 3$<@&CJOJQJBA@BStandardstycketeckensnitt2P@2
Brdtext 2dx@Y@@DokumentversiktD OJQJ,B@,BrdtextxCJD>@"DRubrik$<@&56CJ$KHOJQJ.+@2.
Slutkommentar2&@A2FotnotsreferensH**@R*Fotnotstext6@6
Innehll 1x
56CJ&/@r&Lista
*
@*Index 1
8.!@.Indexrubrik58@8
Innehll 2
x5CJ,@,
Innehll 3,@,
Innehll 4X,@,
Innehll 5 ,@,
Innehll 6,@,
Innehll 7,@,
Innehll 8 x,@,
Innehll 9!@, @",Sidfot
"
p#")@1" Sidnummer
(`AD )*/<014AB4D%M]bugm{}ĆŊOx
H>JK>Lqxa
!"#$%&'()*+,./012345667(OvEg}Iw$i \
kNi
p
aZ<kn,RSTUVWXYZ[\]^_`uv:i=Jz=X.kS0o* j V
!SG{
e
CnprbcdvwQRGH E"F"f"g"
#$@%A%B%D%K%L%M%N%s%t%&&&&&& (!(* *++Z[\..v1w1444<4=45567778899Z:[:Q;R;S;o;p;<<%>&>@@AAAAAAAABBKCLC\CxCCCC6D7DNDDDDJEKEaEEEEF
F#FFFFFGGG HHHHKILIyIzIIIIJJKKMMM'MZM[MNNNOOQQ`QaQQRRSRRRRRR3SSSdUeUUVVeWfW}W~W4X5XpYqYYYZZ[[8[[[[[7\W\X\p]q]r]]]^^^^__``(a)aaaaagbhbibbbbbbgchcdd0f1fUgVgWgwgxg\h]hhhhhJiKi?k@kkkkllmmmmmmmmmmmm1n2nan)o*oooooppbpcpdpppqqPrrrtt}u~uuuuuvvvvwwDxyxzxyyzzzzzJzKz[zmzzzzz{${{{{{{{}}~~~~:H !ʃՃƆǆȈfgǊȊӎԎ
ϑБ͓̓,LMcd867YZ
}BC͟Οxylm./12678:;<?@AEFGKLMNTUx
h23ABVWr_`Ͼо)*OPHI23
\]34GefQR,89t>Pstz{J}!"})*[\lcdtPQaOP,DE(
)
FGS
]XY
NO`ghwx ij " 3 4 8 ; T U Z ^ h i j k v w ."/"u$v$%%&%'%%%%%%&&&&W'''((()))@*B*D*,M,N,,,,,,,,,,,,,,,EW^ijq{//1122444}5~5)6i6j68
899W;X;'=(=>>I?J?@@@BBDDQERESEEEtGuGCIDIAJBJKK@LALLMNNNNyOzOPPPPQQ5S6S`UaU
WWYYZZ7ZzZ{ZZZZ[[[q]r]s]]]]](^)^o^^^^^``ccsdtdddTeUeffghhjj m!mnn"o#opprrssttt
u;ujukuuuwwyy{{}}}~~mo
234de`aN89ÌLMPQ
%&+,QVW>?gh<=ٟڟlmŠ3֡1Tbcnoyz̤ϤФPQUVWYZ[^_`defjklmǥȥghijʦ˦ΦϦ?@IJKMNORSTXYZcdef/0123gh,z{ðͰٰڰ67Jӱ]^yźƺcd¿JK~
RS/0HI
deIJ:a
Wy/%[#6{n}`
{ sl"N}~rrr7 7 ddddddddddddddd7 7 F"F"F"F"F"F"F"7 D%D%D%7 D%N%N%N% N% N% N%N%N%N%N%N%N%N%N%N%N%7 D%\\\\\\\\7 D%444444444444447 D%S;S;S;S;S;S;S;S;S;7 AAA7 AAAAAA A A A AAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAA7 A(M'M'M'M'M'M'M'M'M'M'M'M'M'M'M'M'M'M'M 'M 'M'M'M'M'M'M'M'M'M(MfWfWfWfWfWfWfWfWfWfWfWfWfWfWfWfW
fW
fWfWfWfWfW7 Ar]r]r]r]r]r]r]r]r]r]r]r]r]r]r]r]r]r]7 Aibibibibibibibibibibibibib7 AWgWgWgWgWg Wg Wg WgWgWgWgWgWgWgWgWgWgWgWgWgWgWgWgWg7 Am(mmmmmmmmm m m mmmmmmm m m m mmmmmm(m~u~u~u ~u ~u ~u~u~u~u~u~u~u~u~u7 yy7 yzzzzzzzzzzzzzzz7 y{{{{{{{{{{ { {{{{{{ { { {{{{{{ { { {{{{{{{{{{{7 y : @@@@@@@@: (оооооооооо оо оо оооо о ооооо( . . . ! ! ! / / " " " " (2
(2 0 0 0 0 0 (2 ` ` ` ` ` ` ` ` ` ` ` ` ` ` ` ` ` ` ` ` ` ` ` ` ` ` ` ` ` ` ` ` ` ` ` ` ` ` ` ` ` ` ` ` ` ` ` ` ` ` ` ` ` ` ` ` ` ` ` ` ` ` ` ` ` ` ` ` ` ` ` ` ` ` ` # ` # ` ` ` ` ` ` ` ` ` ` ` ` ` ` ` (2 @@@@@@@@HEHEHEHEHEHEHEHEHEHEHE$ HE$ HE$ HEHEHEHEHEHE% HEHE% HEHEHEHEHEHEHEHEHEHEHEHEHEHEHEHEHEHEHEHEHEHEHEHEHEHE3 h](3 h]h]]]]& ]& ]]]]]]]]]](3 h]h]idididididididididididididid(3 h]h]nnnnnnnnn' n' n' n' nnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnn nn nn nn nnnnnnnnnnnn(ڑڑڑڑڑڑڑڑڑ5 ڑ5 ڑڑڑڑڑڑڑڑڑڑڑڑڑڑڑڑڑڑ(6 bbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbb(((((, (, (, (((((((((((((((((((((((((((((((((((((((3 sssssssssssssssssssssssssssssssssssssssssssss@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@
@"@"@@"@"@
<
QJ_h4Ngy3Z
3ATcquwx=Sf& .Q?\GKS5\bVkqu~$ڜک{.f95"f"#,/5ER`jxkIz,QrJ2Z!#%&(*,/12579;<>@CEFHJLMOQSVWY[]^`befhiklnprtvy$\1DJR]eot~ًLVb4=
/u#,0aBR`:rlRS "$')+.0468:=?BDGIKNPRUXZ\_adgjmoszUnknownLasse Svensson
68Ieg9;p)EHYux8;l7SVk
),Jfi2NQ+.Njm % ( I e h 5
Q
T
z
2NQq&BEZvy
D
`
c
">AMiln
t%t%t%t%t%t%t%t%t%t%t%t%t%t%t%t%t%t%t%t%t%t%t%t%t%t%t%t%t%t%t%t%t%t%t%t%t%t%t%t%t%t%t%t%t%t%t%t%t%t%t%t%t
!!8@0(
B
S ?4
_Toc521815160
_Toc521815161
_Toc521815162
_Toc521815163
_Toc521815164
_Toc521815165
_Toc521815166
_Toc521815167
_Toc521815168
_Toc521815169
_Toc521815170
_Toc521815171
_Toc521815172
_Toc521815173
_Toc521815174
_Toc521815175
_Toc521815176
_Toc521815177
_Toc521815178
_Toc521815179
_Toc521815180
_Toc521815181
_Toc521815182
_Toc521815183
_Toc521815184
_Toc521815185
_Toc521815186
_Toc521815187
_Toc521815188
_Toc521815189
_Toc521815190
_Toc521815191
_Toc521815192
_Toc521815193
_Toc521815194
_Toc521815195
_Toc521815196
_Toc521815197
_Toc521815198
_Toc521815199
_Toc521815200
_Toc521815201
_Toc521815202
_Toc521815203
_Toc521815204
_Toc521815205
_Toc521815206
_Toc521815207
_Toc521815208
_Toc521815209
_Toc521815210
_Toc521815211rd D%N%\4S;AAM'MfWr]ibWgmm~uyz{оk @SEs]]tdnm3
!"#$%&'()*+,./0123u J%r%;4n;AA&MYMW]bvgmmuyz{ņƊ߳ξ+u @E]]d!o fĺ77}E~EHHNN O!OQQQQQQTRURWRZR^R_RdReRgRjRnRoRRRRRRRRRRRRRRRRRzSSSSYYYYYYYYYYYYZZZ ZZZ Z#Z`ZbZZZ}[~[[[[[[[[[[[[[[[uuwwyyۓ58CEIKkmCE%JX٭01&)AKv&T_/014RSWX
.=KQSX]c=?emryJN%<AchPVz
{1;',19x?HLQ>B}'+05?DIQ!$PUZ`~,QlmEEHHTRURRRYYY[\[[[_`eeooqqrrmznzzzzzzz${%{{{DFבڑFHMRfg>?$39ӴԴAMPr\crvabgqm%%NJWJVVWWYWkZmZ[[t
tњ՚ŠƠ~=?~Persson,C:\Mina dokument\The Option to Defer NYA.docPerssonAC:\WINDOWS\TEMP\terskapningsinfo fr The Option to Defer NYA.asdPersson,C:\Mina dokument\The Option to Defer NYA.docPersson,C:\Mina dokument\The Option to Defer NYA.docPerssonAC:\WINDOWS\TEMP\terskapningsinfo fr The Option to Defer NYA.asdPersson,C:\Mina dokument\The Option to Defer NYA.docPersson,C:\Mina dokument\The Option to Defer NYA.docLasse Svensson+C:\Mina dokument\TheOptiontoDeferNYA[1].docLasse Svensson@C:\WINDOWS\TEMP\terskapningsinfo fr TheOptiontoDeferNYA[1].asdLasse Svensson@C:\WINDOWS\TEMP\terskapningsinfo fr TheOptiontoDeferNYA[1].asd=*dkS2md qpGt
?9kdv7]rFd0PFH0(60HFM6!}!d`#$#d+d4,OX,,dO.d:0dz1d31=D1Wz;23V'4d@5tb6dp7dk;?0?d _BOiCdFhFHHdR4R&dYAgZ7^.6 bdx?bd=@b^lHiZ4Ind+rk0sC@[s.TltA+v)fwgwK.xvyd]%{d ThhOJQJo(o(o(.o(..o(... o( .... o(.....
`o(
......
`o(.......
o(........hhOJQJo(77.(.hho(o(.o(..o(... o( .... o(.....
o(
......
`o(.......
`o(........hhOJQJo((.hh.hhOJQJo( o(0o(.0o(..88o(... `o( .... o(.....
o(
......
ppo(.......
( o(........
5CJo(
05CJo(.
05CJo(..
885CJo(...
885CJo( ....
`5CJo(.....
5CJo(
......
5CJo(.......
pp5CJo(........(.77.(.77.hhOJQJo(ko(o(.o(..o(... o( .... o(.....
o(
......
`o(.......
`o(........hhOJQJo(hhOJQJo(77.77.hhOJQJo(hhOJQJo(hhOJQJo(hhOJQJo(77.XXo( 0o(.0o(..88o(... 88o( .... `o(.....
o(
......
o(.......
ppo(........hhCJOJQJo(q o(0o(.0o(..88o(... `o( .... o(.....
o(
......
ppo(.......
( o(........hhOJQJo((.hhOJQJo(hhOJQJo(hho(o(.o(..o(... o( .... o(.....
o(
......
`o(.......
`o(........hhOJQJo((.hhOJQJo(o(o(.m_ z"Qe];go(..o(... o( .... o(.....
`o(
......
`o(.......
o(........hhOJQJo((.77.hho(o(.o(..o(... o( .... o(.....
o(
......
`o(.......
`o(........(.(.hhOJQJo(hhOJQJo( o(0o(.0o(..88o(... `o( .... o(.....
o(
......
ppo(.......
( o(........ko(0o(.0o(..88o(... 88o( .... `o(.....
o(
......
o(.......
ppo(........hhOJQJo(o(o(.o(..o(... o( .... o(.....
`o(
......
`o(.......
o(........77.XXo( 0o(.0o(..88o(... 88o( .... `o(.....
o(
......
o(.......
ppo(........(.XXo( 0оо о ооооо( rrr7 . . . ! ! ! / / " " " " (2 (2 0 0 0 0 0
(2 j j j j j j j j j j j j j j j j j j j j j j j j j j j j j j j j j j j j j j j j j j j j j o(.0o(..88o(... 88o( .... `o(.....
o(
......
o(.......
ppo(........j j j j j j j j j j j j j j j j j j j j j j j j j j j j j 7 ddddddddddddddd7 7 F"F"F"F"F"F"F"7 D%D%D%7 D%N%N%N% N% N% N%N%N%N%N%N%N%N%N%N%N%7 D%\\\\\\\\7 D%444444444444447 D%S;S;S;S;S;S;S;S;S;7 AAA7 AAAAAA A A A AAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAA7 A(M'M'M'M'M'M'M'M'M'M'M'M'M'M'M'M'M'M'M 'M 'M'M'M'M'M'M'M'M'M(MfWfWfWfWfWfWfWfWfWfWfWfWfWfWfWfW
fW
fWfWfWfWfW7 Ar]r]r]r]r]r]r]r]r]r]r]r]r]r]r]r]r]r]7 Aibibibibibibibibibibibibib7 AWgWgWgWgWg Wg Wg WgWgWgWgWgWgWgWgWgWgWgWgWgWgWgWgWg7 Am(mmmmmmmmm m m mmmmmmm m m m mmmmmm(m~u~u~u ~u ~u ~u~u~u~u~u~u~u~u~u7 yy7 yzzzzzzzzzzzzzzz7 y{{{{{{{{{{ { {{{{{{ { { {{{{{{ { { {{{{{{{{{{{7 y : @@@@@@@@: (оооооооооо оо оо оооо о ооооо( . . . ! ! ! / / " " " " (2
(2 0 0 0 0 0 (2 ` ` ` ` ` ` ` ` ` ` ` ` ` ` ` ` ` ` ` ` ` ` ` ` ` ` ` ` ` ` ` ` ` ` ` ` ` ` ` ` ` ` ` ` ` ` ` ` ` ` ` ` ` ` ` ` ` ` ` ` ` ` ` ` ` ` ` ` ` ` ` ` ` ` ` # ` # ` ` ` ` ` ` ` ` ` ` ` ` ` ` ` (2 @@@@@@@@HEHEHEHEHEHEHEHEHEHEHE$ HE$ HE$ HEHEHEHEHEHE% HEHE% HEHEHEHEHEHEHEHEHEHEHEHEHEHEHEHEHEHEHEHEHEHEHEHEHEHE3 h](3 h]h]]]]& ]& ]]]]]]]]]](3 h]h]idididididididididididididid(3 h]h]nnnnnnnnn' n' n' n' nnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnn nn nn nn nnnnnnnnnnnn(ڑڑڑڑڑڑڑڑڑ5 ڑ5 ڑڑڑڑڑڑڑڑڑڑڑڑڑڑڑڑڑڑ(6 bbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbb(((((, (, (, (((((((((((((((((((((((((((((((((((((((3 sssssssssssssssssssssssssssssssssssssssssssss@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@
@"@"@@"@"@
<
QJ_h4Ngy3Z
3ATcquwx=Sf& .Q?\GKS5\bVkqu~$ڜک{.f95"f"#,/5ER`jxkIz,QrJ2pZ!#%&(*,/12579;<>@CEFHJLMOQSVWY[]^`befhiklnprtvy$\1DJR]eot~ًLVb4=
/u#,0aBR`:rlRS "$')+.0468:=?BDGIKNPRUXZ\_adgjmoszUnknownLasse Svensson
68Ieg9;p)EHYux8;l7SVk
),Jfi2NQ+.Njm % ( I e h 5
Q
T
z
2NQq&BEZvy
D
`
c
">AMiln
t%t%t%t%t%t%t%t%t%t%t%t%t%t%t%t%t%t%t%t%t%t%t%t%t%t%t%t%t%t%t%t%t%t%t%t%t%t%t%t%t%t%t%t%t%t%t%t%t%t%t%t%t
!!8@0(
B
S ?4
_Toc521815160
_Toc521815161
_Toc521815162
_Toc521815163
_Toc521815164
_Toc521815165
_Toc521815166
_Toc521815167
_Toc521815168
_Toc521815169
_Toc521815170
_Toc521815171
_Toc521815172
_Toc521815173
_Toc521815174
_Toc521815175
_Toc521815176
_Toc521815177
_Toc521815178
_Toc521815179
_Toc521815180
_Toc521815181
_Toc521815182
_Toc521815183
_Toc521815184
_Toc521815185
_Toc521815186
_Toc521815187
_Toc521815188
_Toc521815189
_Toc521815190
_Toc521815191
_Toc521815192
_Toc521815193
_Toc521815194
_Toc521815195
_Toc521815196
_Toc521815197
_Toc521815198
_Toc521815199
_Toc521815200
_Toc521815201
_Toc521815202
_Toc521815203
_Toc521815204
_Toc521815205
_Toc521815206
_Toc521815207
_Toc521815208
_Toc521815209
_Toc521815210
_Toc521815211rd D%N%\4S;AAM'MfWr]ibWgmm~uyz{оj @REr]]sdnl2
!"#$%&'()*+,./0123u J%r%;4n;AA&MYMW]bvgmmuyz{ņƊ߳ξ
*t @E]]d oeú77}E~EHHNN O!OQQQQQQTRURWRZR^R_RdReRgRjRnRoRRRRRRRRRRRRRRRRRzSSSSYYYYYYYYYYYYZZZ ZZZ Z#Z`ZbZZZ}[~[[[[[[[[[[[[[[[uuwwyyۓ58CEIKkmCE%JX٭01&)AKv&./03QRVW <JPRW\b<>
dlqxIM$
;@bgOUy{
z0:&+08w>GKP=AGIlpuz_f=Fir,QlmEEHHTRURRRYYY[\[[[_`eeooqqrrmznzzzzzzz${%{{{DFבڑFHMRfg>?$39ӴԴAM\cqu`afpl%%MJVJVVVWXWjZlZ[[t tКԚĠŠ}~<>
Persson,C:\Mina dokument\The Option to Defer NYA.docPerssonAC:\WINDOWS\TEMP\terskapningsinfo fr The Option to Defer NYA.asdPersson,C:\Mina dokument\The Option to Defer NYA.docPersson,C:\Mina dokument\The Option to Defer NYA.docPerssonAC:\WINDOWS\TEMP\terskapningsinfo fr The Option to Defer NYA.asdPersson,C:\Mina dokument\The Option to Defer NYA.docPersson,C:\Mina dokument\The Option to Defer NYA.docLasse Svensson+C:\Mina dokument\TheOptiontoDeferNYA[1].docLasse Svensson@C:\WINDOWS\TEMP\terskapningsinfo fr TheOptiontoDeferNYA[1].asdLasse Svensson@C:\WINDOWS\TEMP\terskapningsinfo fr TheOptiontoDeferNYA[1].asd=*dkS2md qpGt
?9kdv7]rFd0PFH0(60HFM6!}!d`#$#d+d4,OX,,dO.d:0dz1d31=D1Wz;23V'4d@5tb6dp7dk;?0?d _BOiCdFhFHHdR4R&dYAgZ7^.6 bdx?bd=@b^lHiZ4Ind+rk0sC@[s.TltA+v)fwgwK.xvyd]%{d ThhOJQJo(o(o(.o(..o(... o( .... o(.....
`o(
......
`o(.......
o(........hhOJQJo(77.(.hho(o(.o(..o(... o( .... o(.....
o(
......
`o(.......
`o(........hhOJQJo((.hh.hhOJQJo( o(0o(.0o(..88o(... `o( .... o(.....
o(
......
ppo(.......
( o(........
5CJo(
05CJo(.
05CJo(..
885CJo(...
885CJo( ....
`5CJo(.....
5CJo(
......
5CJo(.......
pp5CJo(........(.77.(.77.hhOJQJo(ko(o(.o(..o(... o( .... o(.....
o(
......
`o(.......
`o(........hhOJQJo(hhOJQJo(77.77.hhOJQJo(hhOJQJo(hhOJQJo(hhOJQJo(77.XXo( 0o(.0o(..88o(... 88o( .... `o(.....
o(
......
o(.......
ppo(........hhCJOJQJo(q o(0o(.0o(..88o(... `o( .... o(.....
o(
......
ppo(.......
( o(........hhOJQJo((.hhOJQJo(hhOJQJo(hho(o(.o(..o(... o( .... o(.....
o(
......
`o(.......
`o(........hhOJQJo((.hhOJQJo(o(o(.o(..o(... o( .... o(.....
`o(
......
`o(.......
o(........hhOJQJo((.77.hho(o(.o(..o(... o( .... o(.....
o(
......
`o(.......
`o(........(.(.hhOJQJo(hhOJQJo( o(0o(.0o(..88o(... `o( .... o(.....
o(
......
ppo(.......
( o(........ko(0o(.0o(..88o(... 88o( .... `o(.....
o(
......
o(.......
ppo(........hhOJQJo(o(o(.o(..o(... o( .... o(.....
j # j # j j j j j j j j j j j j j j j (2 @@@@@@@@RERERERERERERERERERERE$ RE$ RE$ RERERERERERE% RERE% RERERERERERERERERE`o(
......
`o(.......
o(........77.XXo( 0o(.0o(..88o(... 88o( .... `o(.....
o(
......
o(.......
ppo(........(.XXo( 0o(.0o(..88o(... 88o( .... `o(.....
o(
......
o(.......
ppo(........(.(.77.hhOJQJo(hhOJQJo(hhCJOJQJo(q=`# Tz;2]%{+RHiGt
rFk;FMv7@5m:0OiCqpgw0?}!H,7^ _Bp7fwTlt6gZ]0H4, tb64R6!z1OX,#.6 bO.9kK.xs*431x?bF34Invy+r&dY=@b0PkS=D1C@[sA+vRERERERERERERERERERERERERERERERERE3 r](3 r]r]]]]& ]& ]]]]]]]]]](3 r]r]sdsdsdsdsdsdsdsdsdsdsdsdsdsd(3 r]r]nnnnnnnnn' n' n' n' nnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnn nn nn nn nnnnnnnnnnnn(5 5 (6 llllllllllllllllllllllllllllllllllllllllllllllllllllllllllllllllllllllllllllllllllllllllllllllllllllllll22222, 2, 2, 2222222222222222222222222222222222222223 ssssssssssssssssssssssssssssssssssssssssssss@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@
@"@"@@"@"@
<
QJ_h4Ngy3
3ATcquwx=Sf& .Q?\GKS5\bVkqu~$ڜک{.f95"f"#,/5ER`jxkIz,QrJ2p!#%&(*,/12579;<>@CEFHJLMOQSVWY[]^`befhiklnprtvy$\1DJR]eot~ًLVb4=
/u#,0aBR`:rlRS "$')+.0468:=?BDGIKNPRUXZ\_adgjmoszUnknownLasse Svensson
68Ieg9;p)EHYux8;l7SVk
),Jfi2NQ+.Njm % ( I e h 5
Q
T
z
2NQq&BEZvy
D
`
c
">AMiln
t%t%t%t%t%t%t%t%t%t%t%t%t%t%t%t%t%t%t%t%t%t%t%t%t%t%t%t%t%t%t%t%t%t%t%t%t%t%t%t%t%t%t%t%t%t%t%t%t%t%t%t%t
!!8@0(
B
S ?4
_Toc521815160
_Toc521815161
_Toc521815162
_Toc521815163
_Toc521815164
_Toc521815165
_Toc521815166
_Toc521815167
_Toc521815168
_Toc521815169
_Toc521815170
_Toc521815171
_Toc521815172
_Toc521815173
_Toc521815174
_Toc521815175
_Toc521815176
_Toc521815177
_Toc521815178
_Toc521815179
_Toc521815180
_Toc521815181
_Toc521815182
_Toc521815183
_Toc521815184
_Toc521815185
_Toc521815186
_Toc521815187
_Toc521815188
_Toc521815189
_Toc521815190
_Toc521815191
_Toc521815192
_Toc521815193
_Toc521815194
_Toc521815195
_Toc521815196
_Toc521815197
_Toc521815198
_Toc521815199
_Toc521815200
_Toc521815201
_Toc521815202
_Toc521815203
_Toc521815204
_Toc521815205
_Toc521815206
_Toc521815207
_Toc521815208
_Toc521815209
_Toc521815210
_Toc521815211rd i$s$,:3x:@@9LLLV\afl
mty'yzЅֽ9?wD\\c"n̐ W
!"#$%&'()*+,./0123u o$$,`3:@@KL~LV\afmmt$y4yzޓߨ 27O0?D\\cEn66DD
HHNNENFNPPPPQQyQzQQQQQQQQQQQQQQQQQQQQQQQQQQQRRRRXXXXYYYYYY#Y$Y&Y(Y*Y.Y8Y:YEYHYYYYYZZZZZZZZZZZZZZZZttvvxy#+
Z]hjnphj")<Jo}
UVǲ˲߲KNfp@KBD7=36STUXvw{٩ݩ.=Ra
ouw(ac
2<>@nr@I%/6`etz $'+.27<6<=AU_%*KPU]cl#(pubf"049>ln$)bk,Qlm
##DDHHyQzQQQYY~ZZ[[_9_dd:n>nqqqqyyyyyyzzIzJzzzikʃkmrw04>Acd&*!"*738CIX^òfr?B:?"0
$$rI{IUU{V}VYY(Z*Z,s.s~~"#ѡҡ !ɧ˧کݩ¸ĸac>@286Persson,C:\Mina dokument\The Option to Defer NYA.docPerssonAC:\WINDOWS\TEMP\terskapningsinfo fr The Option to Defer NYA.asdPersson,C:\Mina dokument\The Option to Defer NYA.docPersson,C:\Mina dokument\The Option to Defer NYA.docPerssonAC:\WINDOWS\TEMP\terskapningsinfo fr The Option to Defer NYA.asdPersson,C:\Mina dokument\The Option to Defer NYA.docPersson,C:\Mina dokument\The Option to Defer NYA.docLasse Svensson+C:\Mina dokument\TheOptiontoDeferNYA[1].docLasse Svensson@C:\WINDOWS\TEMP\terskapningsinfo fr TheOptiontoDeferNYA[1].asdLasse Svensson@C:\WINDOWS\TEMP\terskapningsinfo fr TheOptiontoDeferNYA[1].asd=*dkS2md qpGt
?9kdv7]rFd0PFH0(60HFM6!}!d`#$#d+d4,OX,,dO.d:0dz1d31=D1Wz;23V'4d@5tb6dp7dk;?0?d _BOiCdFhFHHdR4R&dYAgZ7^.6 bdx?bd=@b^lHiZ4Ind+rk0sC@[s.TltA+v)fwgwK.xvyd]%{d ThhOJQJo(o(o(.o(..o(... o( .... o(.....
`o(
......
`o(.......
o(........hhOJQJo(77.(.hho(o(.o(..o(... o( .... o(.....
o(
......
`o(.......
`o(........hhOJQJo((.hh.hhOJQJo( o(0o(.0o(..88o(... `o( .... o(.....
o(
......
ppo(.......
( o(........
5CJo(
05CJo(.
05CJo(..
885CJo(...
885CJo( ....
`5CJo(.....
5CJo(
......
5CJo(.......
pp5CJo(........(.77.(.77.hhOJQJo(ko(o(.o(..o(... o( .... o(.....
o(
......
`o(.......
`o(........hhOJQJo(hhOJQJo(77.77.hhOJQJo(hhOJQJo(hhOJQJo(hhOJQJo(77.XXo( 0o(.0o(..88o(... 88o( .... `o(.....
o(
......
o(.......
ppo(........hhCJOJQJo(q o(0o(.0o(..88o(... `o( .... o(.....
o(
......
ppo(.......
( o(........hhOJQJo((.hhOJQJo(hhOJQJo(hho(o(.o(..o(... o( .... o(.....
o(
......
`o(.......
`o(........hhOJQJo((.hhOJQJo(o(o(.o(..o(... o( .... o(.....
`o(
......
`o(.......
o(........hhOJQJo((.77.hho(o(.o(..o(... o( .... o(.....
o(
......
`o(.......
`o(........(.(.hhOJQJo(hhOJQJo( o(0o(.0o(..88o(... `o( .... o(.....
o(
......
ppo(.......
( o(........ko(0o(.0o(..88o(... 88o( .... `o(.....
o(
......
o(.......
ppo(........hhOJQJo(o(o(.o(..o(... o( .... o(.....
`o(
......
`o(.......
o(........77.XXo( 0o(.0o(..88o(... 88o( .... `o(.....
o(
......
o(.......
ppo(........(.XXo( 0o(.0o(..88o(... 88o( .... `o(.....
o(
......
o(.......
ppo(........(.(.77.hhOJQJo(hhOJQJo(hhCJOJQJo(q=`# Tz;2]%{+RHiGt
rFk;FMv7@5m:0OiCqpgw0?}!H,7^ _Bp7fwTlt6gZ]0H4, tb64R6!z1OX,#.6 bO.9kK.xs*431x?bF34Invy+r&dY=@b0PkS=D1C@[sA+v=@e$e$se$e$mHnH
&FÈìĈÈÈìĈ
&F:?C$Eƀb'
&F??$
&F?<<ĈÈ<<ĈÈ'
&F?<<ĈÈ<<ĈÈC$
&F ??78fg:;FHvx9;TV*,giOQ,.km& ( f h R
T
OQCEwy
a
c
?Ajlo#####C$G$N$S$\$a$zzՂقޓߓ)*UVWXablmqߨ9/\yֽٽqstRSy?@Au
$$$,f&f~~ߎ`asĥObc56QIvF0@1n
@0p
@1
@0
@18@0:@1@0@1@0@1t@0v@1
@0
@1
@0
@10@02@1@0@1@0@1Z@0^@1@0@1r@0v@1@0@1@0@1@0"@1@0@1@0@1@0@1@0@1T@0X@1@0@1L@0P@1@0@1@0@1@@0D@1@0@1X@0\@1@0@1P@0T@1@0@1L@0P@1@0@1b@0f@1@0@1@0@1.@02@1@0@1:@0>@1@0@1@0 @1@0@1 @0 @1 @0 @1r!@0v!@1"@0"@1"@0"@1N#(.(.77.hhOJQJo(hhOJQJo(hhCJOJQJo(q=`# Tz;2]%{+RHiGt
rFk;FMv7@5m:0OiCqpgw0?}!H,7^ _Bp7fwTlt6gZ]0H4, tb64R6!z1OX,#.6 bO.9kK.xs*431x?bF34Invy+r&dY=@b0PkS=D1C@[sA+v=@HsK{mHnH
&FÈìĈÈÈìĈ
&F:?C$Eƀb'
&F??$
&F?<<ĈÈ<<ĈÈ'
&F?<<ĈÈ<<ĈÈC$
&F rnprFHJFHJ
&F:$d"h"&`#$??,78fg:;FHvx9;TV*,giOQ,.km& ( f h R
T
OQCEwy
a
c
?Ajld{{{0123<=GHLų
7TоѾ־LNO.T_rstP_tw%%%q.t.fg}}ǏȏГ;<N*=>,$Q]^_!0@1n
@0p
@1
@0
@18@0:@1@0@1@0@1t@0v@1
@0
@1
@0
@10@02@1@0@1@0@1Z@0^@1@0@1r@0v@1@0@1@0@1@0"@1@0@1@0@1@0@1@0@1T@0X@1@0@1L@0P@1@0@1@0@1@@0D@1@0@1X@0\@1@0@1P@0T@1@0@1L@0P@1@0@1b@0f@1@0@1@0@1.@02@1@0@1:@0>@1@0@1@0 @1@0@1 @0 @1 @0 @1r!@0v!@1"@0"@1"@0"@1N#@0R#@1#@0#@1~$@0$@1$@0$@10@1.0
@140V@1<10@0D0Z1@1hC@0F0C@1H100,1.0010V@10*V@10JV@0W@1Y@00Y@0`@0000m@10n@110o@1214180@1N1P1@ 0>@ 0@@1Z1@0@
0@1f1j1@1l0@0n0ƭ@0ʭ@1p0ı@0x0@1z1@004@ 06@00@0@10@0F@10@106@10R@110Q@1ja@1"0c@1(04@1D0@1`0j@110%@10@1107@110tP@10R@1@0V@0V@0Կ@0@0@0@00@0@1f1j1~11141L0N0`@1@0X@0P0@1R0ʿ@0V0Կ@GTimes New Roman5Symbol3&ArialK,Bookman Old Style5&Tahoma;Wingdings"q""x,X
<hBz'24d1 INTRODUCTIONPerssonLasse SvenssonTimes New Roman5Symbol3&ArialK,Bookman Old StyIbjbj]ddT8RFtz8(```;y{{{{{{$pi;;ii``F4i``yiy1y`\*Mf}5=@__Hs_TK{mHnH
&FÈìĈÈÈìĈ
&F:?C$Eƀb'
&F??$
&F?<<ĈÈ<<ĈÈ'
&F?<<ĈÈ<<ĈÈC$
&F ??78fg:;FHvx9;TV*,giOQ,.km& ( f h R
T
OQCEwy
a
c
?Ajld{{{0123<=GHLų
7TоѾ־LNO.T_rstQ`ux%}%%r.u.fg}}ȏɏѓ<=O+>?}%R^_`}~0@1n
@0p
@1
@0
@18@0:@1@0@1@0@1t@0v@1
@0
@1
@0
@10@02@1@0@1@0@1Z@0^@1@0@1r@0v@1@0@1@0@1@0"@1@0@1@0@1@0@1@0@1T@0X@1@0@1L@0P@1@0@1@0@1@@0D@1@0@1X@0\@1@0@1P@0T@1@0@1L@0P@1@0@1b@0f@1@0@1@0@1.@02@1@0@1:@0>@1@0@1@0 @1@0@1 @0 @1 @0 @1r!@0v!@1"@0"@1"@0"@1N#@0R#@1#@0#@1~$@0$@1$@0$@10@1.0
@140V@1<10@0D0Z1@1hC@0F0C@1H100,1.0010V@10*V@10JV@0W@1Y@00Y@0`@0000m@10n@110o@1214180@1N1P1@ 0>@ 0@@1Z1@0@
0@1f1j1@1l0@0n0ƭ@0ʭ@1p0ı@0x0@1z1@004@ 06@00@0@10@10@106@10R@110Q@1ja@1"0c@1(04@1D0@1`0j@110%@10@1107@110tP@10R@1@0V@0V@0Կ@0@0@0@00@0@1f1j1~11141L0N0`@0P0@1R0ʿ@0V0Կ@GTimes New Roman5Symbol3&ArialK,Root Entry FlzU*M@Data
d1TableWordDocument9yBookman Old Style5&Tahoma;Wingdings"1""x,X;hBz'4dT1 INTRODUCTIONPerssonLasse Svensson
Privat%
1 INTRODUCTIONTitle 6>
_PID_GUIDA@0R#@1#@0#@1~$@0$@1$@0$@11Z11 O@1Q@1HR@11RR@1fR@10xR@10@1.0
@140V@1<10@0D0Z1@1hC@0F0C@1H100,1.0010V@10*V@10JV@0W@1Y@00Y@0`@0000m@10n@110o@1214180@1N1P1@ 0>@ 0@@1Z1@0@
0@1f1j1@1l0@0n0ƭ@0ʭ@1p0ı@0x0@1z1@004@ 06@00@0@10@0F@10@106@10R@110Q@1ja@1"0c@1(04@1D0@1`0j@110%@10@1107@110tP@10R@1@0V@0V@0Կ@0@0@0@00@0@1f1j1~11141L0N0`@1@0X@00@10ʿ@00Կ@GTimes New Roman5Symbol3&ArialK,Bookman Old Style5&Tahoma;Wingdings"q"#x,X@HBy&4d
!"#$%&'()*+,./012345678;:C<=>?@ABPERSTUVWXYZ[\]^_`abcdefghijklmnopqrstuvwxyz{}~z$1 INTRODUCTIONPerssonLasse SvenssonzIbjbj 6]dxTvvv8Pvvtxz6~(^~^~^~9kmmmmmm$0$p+
g99gg^~^~Dz4g^~^~kgk#k^~x\`dlP}v'Root Entry FlzU l"@Data
d1TableWordDocument #$SummaryInformation(
DocumentSummaryInformation8CompObj`0Table9yD
!"#$%&'()*+,./012345678;:C<=>?@ABPEFRGHIJKMNOQSTUVWXYZ[\]^_`abcdefghijklmnopqrstuvwxyz{}~
!"#$%&'()*+,./0123456789:;<=>?@ABCDEFGHIJKLMNOPQRSTUVWXYZ[\]^_`abcdefghijklmnopqrstuvwxyz{}~SummaryInformation(
DocumentSummaryInformation8CompObj`0Table
FWorddokument
MSWordDocWord.Document.89q՜.+,D՜.+,@hp
Privat'
1 INTRODUCTIONTitle 6>
_PID_GUIDAN{C28F01C486CC11D59018B59801341654}Oh+'0x
(4@
LX`hp1 INTRODUCTION INPerssonersNormalLasse Svensson13sMicrosoft Word 8.0@ha@Հ@NXU@MMhBz
!"#$%&'()*+,./0123456789:;<=>?@ABCDEFGHIJKLMNOPQRSTUVWXYZ[\]^_`abcdefghijklmnopqrstuvwx{}~
z
FWorddokument
MSWordDocWord.Document.89q՜.+,D՜.+,@hp
Privat&
1 INTRODUCTIONTitle 6>
_PID_GUIDAN{C28F01C486CC11D59018B59801341654}Oh+'0x
(4@
LX`hp1 INTRODUCTION INPerssonersNormalLasse Svensson14sMicrosoft Word 8.0@@Հ@NXU@lHBy
!"#$%&'()*+,./0123456789:;<=>?@ABCDEFGHIJKLMNOPQRSTUVWXYZ[\]^_`abcdefghijklmnopqrstuvw{}~