ISSN 1392-8619 print/ISSN 1822-3613 online

http://www.tede.vgtu.lt
FUZZY REAL OPTIONS VALUATION FOR OIL INVESTMENTS
İrem Uçal
1
, Cengiz Kahraman
Istanbul Technical University, Department of Industrial Engineering,
Macka, Istanbul, 34367, Turkey
Received 26 March 2009; accepted 20 August 2009
Abstract. Traditional valuation methods are less viable under uncertainty. Hence, other methods
such as real options valuation models, which can minimize uncertainty, have become more im-
portant. In this study, the hybrid approach suggested by Carlsson and Fuller is examined for the
case of discrete compounding as this approach better models risky cash fows. A new real options
valuation model that will evaluate the investment in a more realistic way is suggested by postponing
the defuzzifcation of parameters in early stages. Te suggested model has been applied to the data
of an oil feld investment and in conclusion the loss of information caused by early-defuzzifcation
has been determined.
Keywords: Fuzzy Sets, Real Options Valuation, Fuzzy Real Options, Investment.
Reference to this paper should be made as follows: Uçal, İ.; Kahraman, C. 2009. Fuzzy real op-
tions valuation for oil investments, Technological and Economic Development of Economy 15(4):
646–669.
1. Introduction
Te investment valuation methods that are currently in use try to provide decision makers
with enough information for making investment decisions. In literature, there exist numerous
methods for investment evaluation including traditional valuation methods and real options
methods. Traditional valuation methodologies based on discounted cash fows (DCF) do
not consider some of the intrinsic attributes of the asset or investment opportunity (Mun
2002). In the DCF methods, expected value and priced risk characteristics of cash fows
summarize unknown future cash fows. Te DCF model values the equity of an investment
by discounting free cash fows from the operations of the investment to a present value.
1
Corresponding author, e-mail: ucal@itu.edu.tr Phone: +902122931300
TechNologIcal aNd ecoNomIc developmeNT oF ecoNomY
Baltic Journal on Sustainability
2009
15(4): 646–669
doi: 10.3846/1392-8619.2009.15.646-669
647 Technological and Economic Development of Economy, 2009, 15(4): 646–669
Te operating fexibility and strategic value aspects of various projects cannot be properly
captured by traditional DCF techniques because of their discretionary asymmetric nature
and their dependence on future events that are uncertain at the time of the initial decision
(Trigeorgis 1996). Brach (2003) identifed three fundamental diferences between discounted
cash fows methods and real options valuation methods. Te frst diference is that in the
discounted cash fows method, decisions cannot be changed in the future, while in the real
options valuation (ROV) method it is likely to perform directional changes propped up by
obtaining new information. Secondly, in DCF method estimated cash fows group is con-
sidered as a base, whereas in ROV method, cash fows depend on the ambiguous conditions
in the future. Tirdly, in the former method, sensitivity and scenario analyses are static;
while in the latter there is a managerial fexibility in order to provide adaptability to chang-
ing conditions. As a result, it can be claimed that the diferences arise in decision variation,
dependencies, and dynamism.
Real options analysis has appeared as a tool to give more distinct results than traditional
DCF analysis in investment projects by giving the option of postponing the project to a later
time or abandoning the project whenever it is necessary. Mun (2002) explains real options
as a systematic approach and integrated solution making use of fnancial theory, economic
analysis, management science, decision sciences, statistics, and econometric modeling in ap-
plying options theory to valuing real physical assets, as opposed to fnancial assets defned as
a dynamic and uncertain business environment where business decisions are fexible in the
context of strategic capital investment decision making, valuing investment opportunities
and project capital expenditures. Te general analysis of real options as a strategic tool rather
than a mere valuation tool, that is, proactive rather than just reactive fexibility represents
an advance on current thinking in this area (Leslie and Michaels 1997). Real options theory
provides a method to better valuate investment projects in the presence of managerial fex-
ibilities as information option, waiting option and abandonment option (Rocha et al. 2007).
ROV is practically the same as the valuation of fnancial options; yet, there exist a number of
diferences. A fnancial option gives the holder of the right to buy or sell a specifed quantity
of an underlying asset at a fxed price (it is also called a strike price or an exercise price) at or
before the expiration date of the option. Since it is a right and not an obligation, the holder can
prefer not to exercise the right and allow the option to expire. Tere are two types of options
- call options and put options. In a call option, the buyer of the option has the right, but not
the obligation, to buy an agreed quantity of a particular commodity or fnancial instrument
from the seller of the option at a certain time for a certain price, and the seller is obligated to
sell the commodity or fnancial instrument if the buyer exercises the option. Te put option
allows its buyer the right but not the obligation to sell a commodity or fnancial instrument to
the seller of the option at a certain time for a certain price, and with respect to that the seller
has the obligation to purchase the underlying asset at that strike price, if the buyer exercises
the option. Tere are two main types of options: American options and European options.
A primary distinction between American and European types of options is that American
options can be exercised at any time prior to their expiry date, while European options can
be exercised only at expiration. Te possibility of early exercise makes American options
more valuable than otherwise similar European options; it also makes them more difcult
648 İ. Uçal, C. Kahraman. Fuzzy real options valuation for oil investments
to value. Tere is one compensating factor that enables the former to be valued using models
designed for the latter. In most cases, the time premium associated with the remaining life of
an option and transactions costs makes early exercise sub-optimal. In other words, the hold-
ers of in the money options will generally get much more by selling the option to someone
else than by exercising the options.
Mun (2002) emphasized nine distinctive characteristics of fnancial and real options.
Te frst one is that the life span of fnancial options is shorter than real options. In fnancial
options the main variable determining the value is the price of the fnancial asset, while in
real options the main variables determining the value are the cash fows determined by the
demand, management and competition. Te values of the fnancial options are usually small,
while the values of the real options are usually extremely large. Te value of options cannot
be controlled by changing the stock prices in fnancial options, while in real options the value
of strategic options can be increased by managerial decisions and fexibility. In fnancial
options market or competition efects are irrelevant in determining the value and pricing
of options, whereas in real options, market or competition efects determine the strategic
value of strategic options. Financial options have been used for the past 30 years, while real
options are newly developed to be used in fnancial activities of the corporations for the last
few years. Financial options are usually solved by applying closed formed partial diferential
equations and simulation, decreasing the variance methods, whereas real options are usually
solved by applying closed formed equations and the simulation of the main variables with
two binomial letters. Financial options are commercially secured and they can be marketed
with price information and likewise, while real options are not commercially in use and
they are naturally private and they do not have market resemblance. Lastly, management
acceptance and movements are not infuential on the valuation of fnancial options, while
they determine the value of real options.
Option pricing theory has made vast strides since 1972, when Black and Scholes (1972)
published their pioneering paper suggesting a model for valuing dividend-protected European
options. Black and Scholes used a “replicating portfolio” to come up with their fnal formula-
tion. Although their derivation is mathematically complicated, there is a simpler binomial
model for valuing options that draws on the same logic.
In valuing real options, we ofen use the Black Scholes model which has fairly limited ap-
plicability. Most real options are not analogs of European options (Black and Scholes 1973).
However, the Black Scholes partial diferential equation itself has a wider applicability. Given
the wide appropriate boundary conditions, this partial diferential equation can be solved to
evaluate many types of options, such as American and compound options.
Te value of a real option is computed by (Leslie and Michaels 1997):

ROV S e N d Xe N d
T rT

( )

( )
− −
0 1 2
δ
, (1)
where
d
S X r T
T
1
0
2
2

+ − + ln( / ) ( / ) δ σ
σ
, d d T
2 1
− σ , (2)
where S: present value of expected cash fows, X: present value of fxed costs, δ: value lost
over duration of the option, r: risk-free interest rate, σ: uncertainty of expected cash fows,
649 Technological and Economic Development of Economy, 2009, 15(4): 646–669
t: time to expiry, and N(d): cumulative normal distribution function. Brach (2003) pointed
out the equivalence of the real options parameters in fnancial options parameters as follows:
the exercise price used in fnancial options indicates costs to acquire the asset, stock price ac-
counts for the present value of the future cash fows from the asset, time to expiration stands
for the length of time option viable, and the variance of stock returns is replaced by riskiness
of the asset or, in other words, variance of the best and worst case scenario.
Te aim of this study is to propose a new fuzzy real options valuation model which evalu-
ates investments in a more realistic way. Since fuzzy logic fts best to the uncertain nature of
the investment decisions, it has been utilized in the model. Te proposed model postpones
defuzzifcation to the very end stages of the process. Tis prevents the loss of information
at the beginning of the process. Another superiority of our model is that the conversion of
continuous compounding into discrete compounding provides more intelligibility, which
increases its usability in practice.
Te rest of the paper is organized as follows. Literature reviews on fuzzy real options
valuation and real options valuation applications in oil investment valuations are given in
Sections 2 and 3, respectively. In Section 4, we propose a model of fuzzy real options valua-
tion. In Section 5, there is an application of this new model. In the last section we conclude
the obtained results.
2. Fuzzy real options
In classical mathematics the binary valued logic and set theory are used. For an element that
belongs to a set of all possible elements and to any given specifc subset, it can be said exactly
whether that element is or is not a member of it. For example, a person belongs to the set of
all human beings and given a specifc subset, such as all males, one can say whether or not
each particular person belongs to this set. Unfortunately, not everything can be described
using binary valued sets. Te classifcations of persons into males and females is easy, but it
is problematic to classify them as being young or not. Te set of young people is far more
difcult to defne as there is no distinct cut-of point at which age youngness begins or ends.
Tis is not a measurement problem and measuring the age of all elements more precisely is
not helpful. Such a problem is ofen distorted so that it can be described using the well-known
existing methodology. Fuzzy logic gives a method to simulate the ability of human reasoning.
With fuzzy logic an element could partially belong to a set and this is represented by the set
membership. Unlike crisp theories, fuzzy logic enables vagueness and ambiguity as well as
it avoids clear distinctions and limits.
Zadeh (1965) frst founded the fuzzy set theory and he suggested that the notion of a fuzzy
set provides a convenient point of departure for the construction of a conceptual framework
which is parallel in many respects to the framework used in the case of ordinary sets; but
it is more general than the latter and, potentially, may prove to have a much wider scope of
applicability, particularly in the felds of pattern classifcation and information processing.
Fuzzy set theory is a marvelous tool for modeling the kind of uncertainty associated with
vagueness, with imprecision, and/or with a lack of information regarding a particular ele-
ment of the problem at hand (Ross 1995). A fuzzy set is a set containing elements that have
650 İ. Uçal, C. Kahraman. Fuzzy real options valuation for oil investments
varying degrees of membership between 0 and 1 in the set, where 0 means the element is not
a member of the set, 1 means the element is completely a member of the set which means
full membership, and the values between these two values gives a partial membership of
the element on the set. A fuzzy set is prescribed by vague or ambiguous properties hence its
boundaries are ambiguously specifed (Ross 1995).
A fuzzy number is a fuzzy set which is both normal and convex. Most common types of
fuzzy numbers are triangular and trapezoidal. Other types of fuzzy numbers are possible, such
as bell shaped or gaussian fuzzy numbers, yet these types of fuzzy numbers are rarely used
in literature. Triangular fuzzy numbers are defned by three parameters, while trapezoidal
require four parameters.
Fuzzy sets have been used for valuating real options in the literature. Carlsson et al. (2001)
applied fuzzy real options on the project selection by identifying the optimal path of the
dynamic decision trees with the biggest real option value in the end of the planning period.
Carlsson and Fuller (2003) improved a fuzzy approach for real options valuation, which is
one of the mostly used real options valuation approaches, by applying a heuristic real option
rule in a fuzzy setting, where the present values of expected cash fows and expected costs
are estimated by trapezoidal fuzzy numbers and they determined the optimal exercise time
with the assistance of possibilistic mean value and variance of fuzzy numbers. A = (a, b, α,
β) is a trapezoid fuzzy number which can be shown by
A a b
[ ]
− − + −
[ ]
∀ ∈
[ ]
γ
γ α γ β γ ( ) , ( ) , , 1 1 0 1 (3)
which means the most possible values of expected A lie in the interval [a, b] and (a–α)
is the downward potential and (b+β) is the upward potential for A. Tey suggested Eq. 4
for computing fuzzy real options values, where is the estimated present value of expected
cash fows,

X x x ′ ′
( )
1 2
, , , α β is estimated present value of expected costs, E S

0
( )
denotes
the possibilistic mean value of the present value of expected cash fows, E X

( )
denotes the
possibilistic mean value of expected costs, and σ σ
( )

S
0
is the possibilistic variance of the
present value of expected cash fows.

FROV S e N d Xe N d
T rT

( )

( )
− −
 
0 1 2
δ
,
(4)
where

d
E S
E X
r T
T
1
0
2
2

|
(
'
'
`
J
J
J
+ − +
|
(
'
'
`
J
J
J
ln
( )
( )
,


δ
σ
σ
(5)
d d T
2 1
− σ . (6)
Using Eq. 3 for arithmetic operations on trapezoidal fuzzy numbers they fnd the fuzzy
real options formula as below:

FROV s s e N d x x e N d
s e N d
T rT
T

( ) ( )
− ′ ′
( ) ( )

− −

1 2 1 1 2 2
1 1
, , , , , , α β α β
δ
δ
(( )

( ) ( )

( )
(
( )
+ ′
− − −
− −
x e N d s e N d x e N d
e N d e N
rT T rT
T rT
2 2 2 1 1 2
1
,
δ
δ
α β dd e N d e N d
T rT
2 1 2
( ) ( )
+ ′
( )
)
− −
, . β α
δ
(7)
651 Technological and Economic Development of Economy, 2009, 15(4): 646–669
Garcia (2004) used the fuzzy real options valuation (FROV) model in a real investment
project from the energy sector. In Garcia’s paper, the model suggested by Carlsson and Fuller
(2003) has been applied to timing decision and selection of power station for various invest-
ment alternatives. Han and Zheng (2005) applied fuzzy options to fail to perform risk analysis
for municipal bonds in China. Zeng et al. (2007) compared the application of traditional
net present value (NPV) method with real options in investment evaluations, analyzed the
uncertainty of power grid investment project, and discussed how to make the investment
decision when investment cost and cash fow are both fuzzy numbers. Tey also introduced
fuzzy expectation and fuzzy variance concept to construct evaluation model and, fnally,
they performed a simulation model to show the validity. Tao et al. (2007) developed a com-
prehensive methodology based on fuzzy risk analysis and real options approach to evaluate
information technologies investments in a nuclear power station. Allenotor and Tulasiram
(2007) used a fuzzy trinomial real options model on pricing grid resources and proved the
feasibility of the model through experiments. Kahraman and Ucal (2008) used the certainty
equivalent approach for real options valuation in an oil investment with fuzzifed data. Tolga
and Kahraman (2008) considers the multidimensional and vague side of the R&D project
selection process. Te fuzzy analytic hierarchy process, which takes monetary (fuzzy real
option value) and nonmonetary (capability, success probability, trends, etc.) criteria into
account, was used to make this selection among alternative R&D projects. Based on fuzzy
real option valuation, Majlender (2008) proposed the application of a possibilistic decision
rule for optimal investment strategy. Te proposed decision rule had to be reapplied every
time when new information arrived during the deferral period to be able to analyze how the
optimal investment strategy should be changed in the light of the new information.
3. Energy investments
Energy is one of the necessary elements that is required for human life and industrial pro-
duction. Energy consumption increases sharply in parallel with the increase in population,
standard of living, industrial and technologic developments. Planning of the exploition of
energy resources has become a very crucial issue. Given insufcient existing resources, the
importance of investment in new resources increases. Te main energy resources are fossil
fuels, coal, oil, natural gas, nuclear, and renewable energy resources: wind, solar, hydropower,
biomass and geothermal.
Te survey carried out by the International Energy Outlook 2008 (US.DoE 2008) shows
that liquid fuels consumption increases at an average annual rate of 1.2 percent from 2005
to 2030. Renewable energy and coal are the fastest growing energy sources, with consump-
tion increasing by 2.1 percent and 2.0 percent respectively, and it is estimated that the World
energy consumption increments by 50% from 2005 to 2030.
According to the Worldwide Trends in Energy Use and Efciency 2008 report (IEA 2008),
oil products remained the most important fnal energy commodity with a global share of
37% in 2005, driven by their use in transport in which energy consumption has grown most
quickly by rising passenger travel and freight transport. Due to a relative importance of the
transport sector in the OECD countries, oil products accounted for 47% of total fnal energy
652 İ. Uçal, C. Kahraman. Fuzzy real options valuation for oil investments
use in 2005. Oil products also have the largest share of consumption in non-OECD countries,
accounting on average for 28% of total fnal energy use in 2005. In many of these countries
oil products are the most important natural resources. Tey are used as gasoline, jet fuel, and
diesel fuel to run cars, trucks, aircraf, ships, and other vehicles. Home heat sources include
oil, natural gas, and electricity, which in many areas are generated by burning natural gas.
Petroleum and petroleum-based chemicals are important in manufacturing plastic, wax,
fertilizers, lubricants, and many other goods. Tese data reveal how important and critical
oil production is throughout the whole world.
A typical oil investment project has three main phases which are exploration, development,
and extraction. Te exploration phase comprises three main activities: scouting, concession
and exploration. Development activity can be separated into two stages: appraisal and technical
development activities and the two aspects of the extraction phase are particularly relevant
to econometric modeling: operating expenditure and taxation (Favero, Pesaran 1991).
Investment in the power sector has three important characteristics; irreversibility, uncer-
tainty, and fexibility. Traditional appraisal methodologies for project investment can hardly
incorporate the above three characteristics. ROA is relatively new in assessing investments
in climate change projects which handles the investment problem and uncertainty in a par-
ticular way. For example, it focuses on the timing of the investment decision whether to do
the project or not (Yang and Blyth 2007).
As mentioned in the world energy outlook 2006 (Birol 2007), to increase geographic and
fuel-supply diversity and to mitigate climate-destabilizing emissions, the need to curb the
growth in fossil-energy demand is more urgent than ever. Global primary energy demand
in the Reference Scenario is projected to increase by just over one-half between 2006 and
2030 – an average annual rate of 1.6%. Oil investment – three-quarters of which goes to the
upstream – amounts to over $4 trillion in total over 2005–2030. With this huge demand oil
investment decisions become more important in global economy.
According to world energy outlook 2007 (IEA 2007), the developing countries whose
economies and populations are growing fastest contribute 74% of the increase in global
primary energy use where China and India alone account for 45% of this increase.
In the literature there are lots of models which are used to valuate energy investments.
Paddock et al. (1988) used option valuation theory to develop a new approach to value leases
for ofshore petroleum. Bergmann et al. (2006) estimated the magnitude of external costs and
benefts such as landscape quality, wildlife and air quality for the case of renewable technologies
in Scotland, a country which has set particularly ambitious targets for expanding renewable
energy. C. Locatelli (2006) explored the investment strategies of oil companies of Russia.
Chorn and Shokhor (2006) reported the union of a real options algorithm with the Bellman
equation by allowing the analyst and management to avoid estimating outcome probabilities
and computing expected values for investments with sequential stages using project experts’
technical insight. Menegaki (2008) used environmental cost–beneft analysis for the evalu-
ation of renewable energy projects. Dinca et al. (2007) aimed to select the optimal energetic
scenario applied to a consumer with 100 000 inhabitants from the residential–tertiary sector,
from the ecological, energetical and economic points of view. Kjærland (2007) presented a
valuation study of hydropower investment opportunities in the Norwegian context. Mohn and
653 Technological and Economic Development of Economy, 2009, 15(4): 646–669
Misund (2009) presented a micro-econometric study of corporate investment and uncertainty
in a period of market turbulence and restructuring in the international oil and gas industry.
Rodríguez (2008) applied a real option model for the valuation of destination fexibility in
long-term liquefed natural gas supplies. Abadie and Chamorro (2008) used Monte Carlo
approach to evaluate natural gas resources investments. Lin et al. (2009) developed a hybrid
interval-fuzzy two-stage stochastic energy systems planning model and applied the model
to a hypothetical regional energy system.
Tere are some factors on the evaluation of energy investments: longer time of return of
energy investments comparing to other investments, higher capital necessity and longevity
of duration of management increase in the risk and ambiguity. Te uncertainty and risks
of the energy investments are also increasing due to the changes in the energy policies and
escalating energy requirements of the world. Using fuzzy real options minimizes the efects
of these uncertainties and risks on the investment.
4. Proposed model for real options valuation
In this study we suggest a new model based on Carlsson & Fuller’s hybrid approach using
discrete compounding instead of continuous compounding by defuzzifying the costs and
revenues at a later stage than the based model and the based model has been improved by
fuzzifying interest rates and probabilities.
4.1. Defuzzifcation methods used in the model
Defuzzifcation is the process of producing a quantifable result in fuzzy logic. It is the con-
version of a fuzzy quantity into a precise quantity, just as justifcation is the conversion of
a precise quantity into a fuzzy quantity (Ross 1995). We use three defuzzifcation methods:
Total Integral Value Method, Centroid Method, and a defuzzifcation method for normal dis-
tribution. Tey are summarized as follows:
4.1.1. Total Integral Value Method
Liou and Wang (1992) developed the total integral value method based on the mean of the
integral value, and it ranks triangular fuzzy number F = (a, b, c). Te total integral value of
F is defned as

I A c b a
T
α
α α ( ) . + + −
( )

]
]
0 5 1
, (8)
where α is an index of optimism that represents the degree of optimism of the decision-maker
and has a value between [0,1]. Te optimism coefcient also refects the decision maker’s risk
taking trend. A lower optimism coefecient leads a risk-averse decision maker and likely a
higher optimism coefecient leads a risk-seeking decision maker. Tis method has been used
to defuzzify fuzzy triangular numbers.
654 İ. Uçal, C. Kahraman. Fuzzy real options valuation for oil investments
4.1.2. Centroid Method
Centroid method is the most prevalent and physically appealing of all the defuzzifcation
methods. Tis method determines the centre of the area of the combined membership func-
tions and mimics the center of gravity approach in physics. Eq. 9 gives the algebric expression
of this method (Ross 1995).

z
z zdz
z dz
C
C
*
( )
( )



µ
µ


. (9)
Although there are lots of other defuzzifcation methods available, the centroid method
has been chosen to use in this paper for its simplicity and fast computation.
4.1.3. A defuzzifcation method for normal distribution
Figure 1 demonstrates the probability values that have a normal distribution. If the weights
used to balance the probability values are taken respectively as t
1
, t
2
, t
3
, and t
4
as determined
by the experts, since a balancing must be applied towards the summit point of the bell curve
that is observed in normal distribution. Ten, for Fig. 1 these values have to provide the in-
equality of t
1
> t
2
> t
3
> t
4
. Under these conditions, Eq. 10 is suggested in order to defuzzify
the fuzzy numbers that have a normal distribution.
d
d d d d

+ + +
+ + +
τ τ τ τ
τ τ τ τ
1 1 2 2 3 3 4 4
1 2 3 4
. (10)
4.2. Discrete compounding
Discrete compounding has more widespread usage in engineering economy community
than continuous compounding. Carlsson & Fuller’s (2003) real options formula is modifed
by applying discrete compounding. To do this, the equation F Pe P i
rn n
+ ( ) 1 has been
considered. Te discrete interest rate is obtained by the below operations.
e i rn i r i i e
rn
r
n
r
n
r r
r
+ ÷ → ÷ + ÷ → ÷ + ÷ → ÷ − ( ) ln( ) ln( ) 1 1 1 1 , (11)
Fig. 1. Normal distribution
d
1
d
2
d
3
d
4
655 Technological and Economic Development of Economy, 2009, 15(4): 646–669
e i n i i i e
n n n δ
δ δ δ δ
δ
δ δ + ÷ → ÷ + ÷ → ÷ + ÷ → ÷ − ( ) ln( ) ln( ) 1 1 1 1, (12)
where i
r
is risk-free interest rate in dicrete cases and i
δ
is the percentage value lost over the
duration of option in discrete cases. When the discrete interest rate formulas acquired in
the fuzzy real options formula are used instead of continuous interest rate in the equations,
below formulas are reached:




d
E S
E X
i i T
T
r
1
2
1 1
2

|
(
'
`
J
J
+ + − + +
|
(
'
'
`
J
J
J

ln
( )
( )
(ln( ) ln( ) l
δ
σ
σ
nn
( )
( )
ln
E S
E X
i
i
T
T
r
T


|
(
'
`
J
J
+
+
+
|
(
'
`
J
J
+
1
1 2
2
δ
σ
σ
(13)
and
FROV S i N d X i N d
T
r
T
+
( ) ( )
− +
( ) ( )
− −
 
0 1 2
1 1
δ
, (14)
where

S
0
is the fuzzy present value of expected cash fows,

X is the fuzzy present value of
fxed costs, and other symbols are the same as the ones in the crisp formulas.
For operational convenience, the variable w has been designated for Eq. 15 and afer some
simplifcations Eq.s 17 and 16 have been reached.

w
i
i
r

+
+
1
1
δ
. (15)
If we substitute Eq. 15 in Eq. 13, Eqs. 16 and 17 are obtained.





d
E S
E X
i
i
T
T
E S
E
r
T
1
2
1
1 2

|
(
'
`
J
J
+
+
+
|
(
'
`
J
J
+

ln
( )
( )
ln
ln
( )
(
δ
σ
σ
XX
w e
T
E S
E X
w
T
T
T
)
ln
ln
( )
( )
|
(
'
`
J
J
|
(
'
'
`
J
J
J
+

|
(
'
`
J
J
|
(
'
'
`
J
J
J
σ
σ
2
2


11
2
σ
σ
T
T
e
|
(
'
'
`
J
J
J
|
(
'
'
'
`
J
J
J
J
(16)





d
E S
E X
i
i
T
T
E S
E
r
T
1
2
1
1 2

|
(
'
`
J
J
+
+
+
|
(
'
`
J
J
+

ln
( )
( )
ln
ln
( )
(
δ
σ
σ
XX
w e
T
E S
E X
w
T
T
T
)
ln
ln
( )
( )
|
(
'
`
J
J
|
(
'
'
`
J
J
J
+

|
(
'
`
J
J
|
(
'
'
`
J
J
J
σ
σ
2
2


11
2
σ
σ
T
T
e
|
(
'
'
`
J
J
J
|
(
'
'
'
`
J
J
J
J
.




d
E S
E X
w e
T
T
T
1
1
2

|
(
'
`
J
J
|
(
'
'
'
`
J
J
J
J
|
(
'
'
`
J
J
J
ln
( )
( )
σ
σ
σ
. (17)
If we substitute Eq. 17 in Eq. 6, Eq. 18 is obtained.




d
E S
E X
w e T
T
T
T
2
1
2

|
(
'
`
J
J
|
(
'
'
'
`
J
J
J
J

|
(
'
'
`
J
J
J
ln
( )
( )
σ
σ
σ
σ . (18)
4.3. Fuzzifying the discrete interest rates
In this subsection i
r
and i
δ
are accepted as triangular fuzzy numbers. We prefer using Buckley’s
notation. Buckley’s membership function for a future value

F is given by
656 İ. Uçal, C. Kahraman. Fuzzy real options valuation for oil investments

µ x M m
f y M
m
m
f y M
m



( )

( )
( )
|
(
'
'
`
J
J
J
1
1
2
2
2
3
, , , ,
(19)
which is determined by

f y M f y F f y r
ni i k
n
 

( )

( )
+
( ) ( )

1
(20)
for i = 1,2 where k = i for negative

F, k = 3 – i for positive

F(Buckley 1987).
Te equations that demonstrate the right and lef convergences towards the i
δ
and i
r

functions are given by Eqs. 21–24.

f y i i y i i
r r
L
r
M
r
L
1

( )
+ −
( )
, (21)
f y i i y i i
r r
R
r
M
r
R
2

( )
+ −
( )
, (22)

f y i i y i i
L M L
1

δ δ δ δ
( )
+ −
( )
, (23)
f y i i y i i
R M R
2

δ δ δ δ
( )
+ −
( )
. (24)
Fig. 2 demonstrates the membership functions of the interest rates i
δ
and i
r
.
Fig. 2. Membership functions of i
δ
and i
r
f y i i y i i
r r
L
r
M
r
L
1

( )
+ −
( )
f y i i y i i
r r
R
r
M
r
R
2

( )
+ −
( )
i
r
L
i
r
M
i
r
R
i
r
μ(i
r
)
f y i i y i i
L M L
1

δ δ δ δ
( )
+ −
( )
f y i i y i i
R M R
2

δ δ δ δ
( )
+ −
( )
i
δ
L
i
δ
M
i
δ
R
i
δ
μ(i
δ
)
With substituting Eqs. 21–24 into Eq. 15, we obtain Eq. 25.

w
i
i
f y i
f y i
r
k r
i

+
+

+
( )
+
( )
1
1
1
1
δ
δ
, (25)
where k = 3 – i and i = 1,2.
If it is substituted in Eq.s 17 and 18, Eqs. 26–27 are obtained.



d
E S
E X
f y i
f y i
e
T
k r
i
T
T
1
1
2
1
1

|
(
'
`
J
J
+
( )
+
( )
|
(
'
'
`
J
J
J
|
ln
( )
( )
σ
δ
σ
σ
((
'
'
`
J
J
J
|
(
'
'
'
'
`
J
J
J
J
J

|
(
'
`
J
J
+ + −
( )
ln
( )
( )
E S
E X
i y i i
T
r
R
r
M
r
R


1
1
σ
11
2
+ + −
( )
|
(
'
'
`
J
J
J
|
(
'
'
'
'
`
J
J
J
J
J
|
(
'
'
`
J
J
J
i y i i
e
L M L
T
T
δ δ δ
σ
σ

657 Technological and Economic Development of Economy, 2009, 15(4): 646–669




d
E S
E X
f y i
f y i
e
T
k r
i
T
T
1
1
2
1
1

|
(
'
`
J
J
+
( )
+
( )
|
(
'
'
`
J
J
J
|
ln
( )
( )
σ
δ
σ
σ
((
'
'
`
J
J
J
|
(
'
'
'
'
`
J
J
J
J
J

|
(
'
`
J
J
+ + −
( )
ln
( )
( )
E S
E X
i y i i
T
r
R
r
M
r
R


1
1
σ
11
2
+ + −
( )
|
(
'
'
`
J
J
J
|
(
'
'
'
'
`
J
J
J
J
J
|
(
'
'
`
J
J
J
i y i i
e
L M L
T
T
δ δ δ
σ
σ
. (26)




d
E S
E X
i
i
e
T r
R
L
T
T
1
1
2
1
1

|
(
'
`
J
J
+
+
|
(
'
'
`
J
J
J
|
|
(
'
'
`
J
J
J
ln
( )
( )
σ
δ
σ
σ
((
'
'
'
'
`
J
J
J
J
J
|
(
'
'
'
'
|
(
'
`
J
J
+
+
|
(
'
'
`
J
J
J
; ln
( )
( )
E S
E X
i
i
T r
M
M


1
1
1
σ
δ
TT
T
T r
L
e
E S
E X
i
i
σ
σ
σ
δ
2
1
1
1
|
(
'
'
`
J
J
J
|
(
'
'
'
'
`
J
J
J
J
J
|
(
'
`
J
J
+
+
;
ln
( )
( )


RR
T
T
e
|
(
'
'
`
J
J
J
|
(
'
'
'
'
`
J
J
J
J
J
`
J
J
J
J
J
|
(
'
'
`
J
J
J
σ
σ
2
.
(27)




d
E S
E X
f y i
f y i
e
T
k r
i
T
T
2
1
2
1
1

|
(
'
`
J
J
+
( )
+
( )
|
(
'
'
`
J
J
J
|
ln
( )
( )
σ
δ
σ
σ
((
'
'
`
J
J
J
|
(
'
'
'
'
`
J
J
J
J
J

|
(
'
`
J
J
+ + −
σ
σ
T
E S
E X
i y i i
T
r
R
r
M
r
ln
( )
( )


1
1
RR
L M L
T
T
i y i i
e T
( )
+ + −
( )
|
(
'
'
`
J
J
J
|
(
'
'
'
'
`
J
J
J
J
J

|
(
'
'
`
J
J
J
1
2
δ δ δ
σ
σ
σ .
(28)




d
E S
E X
i
i
e
T r
R
L
T
T
2
1
2
1
1

|
(
'
`
J
J
+
+
|
(
'
'
`
J
J
J
|
|
(
'
'
`
J
J
J
ln
( )
( )
σ
δ
σ
σ
((
'
'
'
'
`
J
J
J
J
J
|
(
'
'
'
'

|
(
'
`
J
J
+
+
|
(
'
'
`
σ
σ
δ
T
E S
E X
i
i
T r
M
M
; ln
( )
( )


1
1
1
JJ
J
J
|
(
'
'
'
'
`
J
J
J
J
J

|
(
'
`
J
J
+
|
(
'
'
`
J
J
J
T
T
T
e T
E S
E X
i
σ
σ
σ
σ
2
1
1
;
ln
( )
( )


rr
L
R
T
T
i
e T
1
2
+
|
(
'
'
`
J
J
J
|
(
'
'
'
'
`
J
J
J
J
J

`
J
J
J
J
J
|
(
'
'
`
J
J
J
δ
σ
σ
σ .





d
E S
E X
i
i
e
T r
R
L
T
T
2
1
2
1
1

|
(
'
`
J
J
+
+
|
(
'
'
`
J
J
J
|
|
(
'
'
`
J
J
J
ln
( )
( )
σ
δ
σ
σ
((
'
'
'
'
`
J
J
J
J
J
|
(
'
'
'
'

|
(
'
`
J
J
+
+
|
(
'
'
`
σ
σ
δ
T
E S
E X
i
i
T r
M
M
; ln
( )
( )


1
1
1
JJ
J
J
|
(
'
'
'
'
`
J
J
J
J
J

|
(
'
`
J
J
+
|
(
'
'
`
J
J
J
T
T
T
e T
E S
E X
i
σ
σ
σ
σ
2
1
1
;
ln
( )
( )


rr
L
R
T
T
i
e T
1
2
+
|
(
'
'
`
J
J
J
|
(
'
'
'
'
`
J
J
J
J
J

`
J
J
J
J
J
|
(
'
'
`
J
J
J
δ
σ
σ
σ .
(29)




d
E S
E X
i
i
e
T r
R
L
T
T
2
1
2
1
1

|
(
'
`
J
J
+
+
|
(
'
'
`
J
J
J
|
|
(
'
'
`
J
J
J
ln
( )
( )
σ
δ
σ
σ
((
'
'
'
'
`
J
J
J
J
J
|
(
'
'
'
'

|
(
'
`
J
J
+
+
|
(
'
'
`
σ
σ
δ
T
E S
E X
i
i
T r
M
M
; ln
( )
( )


1
1
1
JJ
J
J
|
(
'
'
'
'
`
J
J
J
J
J

|
(
'
`
J
J
+
|
(
'
'
`
J
J
J
T
T
T
e T
E S
E X
i
σ
σ
σ
σ
2
1
1
;
ln
( )
( )


rr
L
R
T
T
i
e T
1
2
+
|
(
'
'
`
J
J
J
|
(
'
'
'
'
`
J
J
J
J
J

`
J
J
J
J
J
|
(
'
'
`
J
J
J
δ
σ
σ
σ .

FROV equation is adapted for fuzzy discrete interest rates as follows;

FROV S f y i N d X f y i N d
i
T
i r
T
+
( ) ( ) ( )
− +
( ) ( ) ( )
− −
0 1 2
1 1
δ
  
, (30)
658 İ. Uçal, C. Kahraman. Fuzzy real options valuation for oil investments

FROV S i N d X i N d S i
N d
L
T
r
R
T
M
T
+
( ) ( ( )
− +
( ) ( )
+
( )

− −
0 1 2 0
1
1 1 1
δ δ
  

; ,
(( )
− +
( ) ( )
+
( ) ( )
− +
( ) ( ))
− − −
    
X i N d S i N d X i N d
r
M
T
R
T
r
L
T
1 1 1
2 0 1 2
;
δ
..
(31)
Eq. 31 is developed for the situations that occur while dealing with fuzzy discrete parame-
ters. It is helpful when there is incomplete information under fuzzy decision environment.
4.4. Postponing the defuzzifcation of costs and revenues

S s s
0 1 2

( )
, , , α β and

X x x ′ ′
( )
1 2
, , , α β are trapezoidal fuzzy numbers. If ln
E S
E X


0

]
]

]
]
|
(
'
'
`
J
J
J

expression found in Eq. 5 is replaced by the fuzzy numbers

S
0
and

X, Eq. 32 is reached
subsequently.
ln ln , , , ln
E S
E X
s
x
s
x
s
x
s
x
s


0
1
4
2
3
3
2
4
1
1

]
]

]
]
|
(
'
'
`
J
J
J

]
]
]

xx
s
x
s
x
s
x
4
2
3
3
2
4
1
, ln , ln , ln

]
]
]
. (32)
If Eq.32 is placed in Eq. 6, Eq. 33 is reached.

d
s
x
w e
s
x
T
T
T
1
1
4
1
2
2
3

|
(
'
`
J
J
|
(
'
'
'
`
J
J
J
J
|
(
'
'
'
|
(
|
(
'
'
`
J
J
J
ln ,
σ
σ
σ
ln
''
`
J
J
|
(
'
'
'
`
J
J
J
J
|
(
'
`
J
J
|
(
'
'
`
J
J
J
|
(
1
2
3
2
1
2
σ
σ
σ
σ
σ
σ
T
T
T
T
T
T
w e
s
x
w e
,
ln
''
'
`
J
J
J
|
(
'
'
`
J
J
J
|
(
'
'
'
`
J
J
J
J
|
(
'
`
J
J
|
(
'
'
'
`
J
J
J
J
, ln
s
x
w e
T
T
T
4
1
1
2
σ
σ
σ ``
J
J
J
J
.
(33)
In d d T
2 1
− σ expression, σ σ σ σ σ T T T T T
~
( , , , )
( ) ( ) ( ) ( )
1 2 3 4
can be stated
and Eq. 34 is reached:

d
s
x
w e T
s
x
T
T
T
2
1
4
1
2
2

|
(
'
`
J
J
|
(
'
'
'
`
J
J
J
J
|
(
'
'
'

|
(
'
'
`
J
J
J
ln ,
σ
σ
σ
σ ln
33
1
2
3
2
1
|
(
'
`
J
J
|
(
'
'
'
`
J
J
J
J

|
(
'
`
J
J
|
(
'
'
`
J
J
J σ
σ
σ
σ
σ
σ
T
T
T
T
T
w e T
s
x
w
,
ln ee T
s
x
w e
T
T
T
T σ
σ
σ
σ
σ
2
4
1
1
2
|
(
'
'
`
J
J
J
|
(
'
'
`
J
J
J
|
(
'
'
'
`
J
J
J
J

|
(
'
`
J
J
|
, ln
((
'
'
'
`
J
J
J
J

`
J
J
J
J
σ T .
(34)
659 Technological and Economic Development of Economy, 2009, 15(4): 646–669
Carlsson and Fuller (2003) made a defuzzifcation in the very early stages of their model.
Eqs. 33 and 34 are developed for postponing the defuzzifcation to the very end stages of
the model.
4.5. Fuzzy probabilities
Afer d
1
and d
2
values obtained in Section 4.4 were defuzzifed by Eq. 10, probability values
could be obtained from the normal distribution table. By postponing the defuzzifcation,
operations could be carried out with fuzzy probabilities. Below, formulas have been developed
to be used in the case of carrying on with fuzzy probabilities.

FROV S i N
s
x
w e
T
T
T
T
+
( )
|
(
'
`
J
J
|
(
'
'
'
`
J
J
J
J
|

|
(
'
'
`
J
J
J

0
1
4
1
2
1
δ
σ
σ
σ
ln
((
'
'
'
|
(
'
`
J
J
|
(
'
'
'
`
J
J
J
J
|
(
'
`
|
(
'
'
`
J
J
J
, ,
ln
ln
s
x
w e
s
x
T
T
T
2
3
1
2
3
2
σ
σ
σ
JJ
J
|
(
'
'
'
`
J
J
J
J
|
(
'
`
J
J
|
(
'
'
`
J
J
J
|
(
'
'
1
2
4
1
1
2
σ
σ
σ
σ
σ
σ
T
T
T
T
T
T
w e
s
x
w e , ln
``
J
J
J

|
(
'
'
|
(
'
'
'
`
J
J
J
J
`
J
J
J
J

+
( )
|
(
'
`
J
J

X i N
s
x
w e
r
T
T
T
T
1
1
4
1
2
ln
σ
σ
σ
``
J
J
J
|
(
'
'
`
J
J
J
|
(
'
'
'
`
J
J
J
J
|
(
'
'
'

|
(
'
`
J
J
|
(
'
σ
σ
σ
σ
T
s
x
w e
T
T
T
, ln
2
3
1
2
''
'
`
J
J
J
J

|
(
'
`
J
J
|
(
'
'
'
`
J
J
J
J

|
(
'
'
`
J
J
J
σ
σ
σ
σ
σ
T
s
x
w e T
T
T
T
,
ln , ln
3
2
1
2
ss
x
w e T
T
T
T
4
1
1
2
|
(
'
`
J
J
|
(
'
'
'
`
J
J
J
J

`
J
J
J
J
|
(
'
'
`
J
J
J σ
σ
σ
σ .

(35)

(36)
FROV i s N
s
x
w e
T
T
T
T
+
( )
|
(
'
`
J
J
|
(
'
'
'
`
J
J
J
J
|
(

|
(
'
'
`
J
J
J
1
1
1
4
1
2
δ
σ
σ
σ
ln
''
'
'

`
J
J
J
J
J
|
(
'
`
J
J
|
(
'
'
'
`
J
J
J
|
(
'
'
`
J
J
J
, s N
s
x
w e
T
T
T
2
2
3
1
2
ln
σ
σ
σ
JJ
|
(
'
'
'
`
J
J
J
J
|
(
'
`
J
J
|
(
'
'
'
`
J
J
J
J
|
(
|
(
'
'
`
J
J
J
,
ln s N
s
x
w e
T
T
T
3
3
2
1
2
σ
σ
σ
''
'
'
`
J
J
J
J
|
(
'
`
J
J
|
(
'
'
'
`
J
J
J
J
`
J
J
J
J
|
(
'
'
`
J
J
J
, ln s N
s
x
w e
T
T
T
4
4
1
1
2
σ
σ
σ
||
(
'
'
'
'
]
]
]
]
]
]

+
( )
|
(
'
`
J
J
|
(
' −
|
(
'
'
`
J
J
J
1
1
1
4
1
2
i x N
s
x
w e
r
T
T
T
T
ln
σ
σ
σ
''
'
`
J
J
J
J

|
(
'
'
'
`
J
J
J
J

|
(
'
`
J
J
|
(
'
'
`
J
σ
σ
σ
σ
T x N
s
x
w e
T
T
T
,
2
2
3
1
2
ln
JJ
J
|
(
'
'
`
J
|
(
'
'
'
`
J
J
J
J

|
(
'
'
'
`
J
J
J
J
|
(
'
`
J
J
σ
σ
σ
σ
T
x N
s
x
w e
T
T
T
,
ln
3
3
2
1
2
JJ
J
|
(
'
'
`
J
|
(
'
'
'
`
J
J
J
J

|
(
'
'
'
`
J
J
J
J
|
(
'
`
J
J
σ
σ
σ
σ
T x N
s
x
w e
T
T
T
, ln
4
4
1
1
2
JJ
J
|
(
'
'
'
`
J
J
J
J

|
(
'
'
'
`
J
J
J
J
]
]
]
]
]
σ T .
660 İ. Uçal, C. Kahraman. Fuzzy real options valuation for oil investments
5. Application
Grafström and Lundquist (2002) examined whether the value of an undeveloped oilfeld dif-
fers depending on whether real options valuation or discounted cash fows analysis is used.
In that application the present value of the costs is found as
X 188777796$
and the present
value of the revenues is found as S
0
241188460 $. Table 1 reports fnancial parameters used
in ROV where δ is convenience yield, and r is instantaneous risk-free interest rate.
Table 1. Parameters used in Real Options Valuation
δ R σ
δ
σ
S
5.20% 5.00% 67.18% 33.66%
Te current values of the revenues and costs are fuzzifed in order to obtain the standart
deviation (33.66%) and the fuzzy real options value is calculated through their application
to Eq.s 5–7.
S
0
241188460 $,

S
0
229129037 253247883 176197913 176197913 ( ; ; ; ),
X 188777796$,

X ( ; ; ; ) 179338906 198216686 100000000 100000000 ,

E S

0
229129037 253247883
2
176197913 176197913
6
241188460
( )

+
+

$$
,
s s
2 1
253247883 229129037 24118846 − − ,
α β + + 176197913 176197913 352395826,
σ
α β α β


S
s s s s
E S
0
2 1 2 1
2
0
4 6 24
81184035
24118
( )


+
− +
+
+
( )

( ) ( )( ) ( )
88460
33 66 . %,
E X

( )

+
+
+

( ) 179338906 198216686
2
100000000 100000000
6
1887777966$,
d
1
2
241188460
188777796
0 05 0 052
0 3366
2
1

|
(
'
`
J
J
+ − +
|
(
'
'
`
J
J
J
ln . .
.
22
0 3366 12
0 7725
.
. ,
N d
1
0 7801
( )
. ,
d
2
0 7725 0 3366 12 0 3935 − − . . . ,
N d
2
0 3445
( )
. ,
661 Technological and Economic Development of Economy, 2009, 15(4): 646–669

FROV e
( )
⋅ ⋅
− ⋅
229129037 253247883 176197913 176197193
0 05212
; ; ;
.
00 7801
179338906 198216686 100000000 100000000
0 051
.
; ; ;
.

( )

− ⋅
e
22
0 3445
58294298 71944517 92552920 92552920

( )
.
; ; ; $.

In the defuzzifcation process, the centroid method based on the centers of gravity of the
possibility distribution is used.

z
z dz zdz
z
zdz
*
+ + −

|
(
'
`
J
J
2
1
71944517
92552920
71944517
164497437
∫∫ ∫ ∫



+ +
58294298
71944517
34258622
58294298
1
71944517
92
zdz dz
z
5552920
71944517
164497437
58294298
71944517
3425862
|
(
'
`
J
J ∫ ∫

dz
22
58294298
71415953 66

. $
.
Te fuzzy real options value of the investment calculated through Carlsson and Fuller
approach is found as 71415953 66 . $ afer being defuzzifed through the centroid method.
Below, the interest rates for the case of discrete compounding are obtained by applying
the data of oil feld investment example to Eq.s 11 and 12.
i e e
r
r
− − 1 1 0 0513
0 05 .
. ,
i e e
δ
δ
− − 1 1 0 0534
0 052 .
. ,
w
i
i
r

+
+

1
1
0 998
δ
. .
Fuzzy revenues and costs are:


S
0
59931124 229129037 253247883 429445796 ( ; ; ; ),


X ( ; ; ; ) 79338906 179338906 198216686 298216686
.
Te expected values of the revenues and costs are given below.
E S

0
229129037 253247883
2
229129037 52931124 42944579
( )

+
+
− − ( ) ( 66 253247883
6
− )


E S

0
241188460
( )
$


E X

( )

+
+
− − 179338906 198216686
2
179338906 79338906 298216686 ( ) ( −−198216686
6
)


E X

( )
188777796$

When the values are placed in Eqs. 17 and 18, the fuzzy real options value is calculated
as below:
d e
1
1
0 3366 12
12
0 3366
0 3366
241188460
188777796
0 998
|
(
'
`
J
J
ln .
.
.
. 112
2
0 7726
|
(
'
'
'
`
J
J
J
J
. ,
,
,
,
.
662 İ. Uçal, C. Kahraman. Fuzzy real options valuation for oil investments
d
2
0 7726 0 3366 12 0 3934 − − . . . ,
N d
1
0 78012
( )
. ,

N d
2
0 34701
( )
.
,

FROV
( )
⋅ ⋅

52931124 229129037 253247883 429445796 1 0534 0
12
; ; ; . .7 78012
79338906 179338906 198216686 298216686 1 0513
12

( )
⋅ ⋅

; ; ; . 00 34701
34660540 58012568 71686162 164359270
.
; ; ;


( )

FROV −
( )
34660540 58012568 71686162 164359270 ; ; ; .
Te fuzzy real options value of the investment is defuzzifed through the centroid method
and found to be 72971941$.
i
δ
and i
r
are fuzzifed at the rate of ±%5 to calculate the real options value of the invest-
ment in the condition of fuzzifying discrete interest rates.
i
r

( )
0 0487 0 0513 0 0539 . ; . ; . ,

i
δ

( )
0 0507 0 0534 0 0561 . ; . ; .
.
When the data are applied to Eq.s 21 and 22, the below membership functions for i
r
are
obtained.
f y i i y i i y
r r
L
r
M
r
L
1
0 0487 0 0513 0 0487 0 0487 0 0

( )
+ −
( )
+ −
( )
+ . . . . . 0026y,

f y i i y i i y
r r r r
R M R
2
0 0539 0 0513 0 0539 0 0539 0 0

( )
+ −
( )
+ −
( )
− . . . . . 0026y
.
When approached from the lef: for y = 0 f y i
r 1
0 0487

( )
. , for y = 1 f y i
r 1
0 0513

( )
. , and
when approached from the right: for y = 0 f y i
r 2
0 0539

( )
. and for y = 1 f y i
r 2
0 0513

( )
.
values are obtained.
When the data are applied to Eq.s 23 and 24, the below membership functions for i
δ
are
obtained.

f y i i y i i y
L M L
1
0 0507 0 0534 0 0507 0 0507 0 0

δ δ δ δ
( )
+ −
( )
+ −
( )
+ . . . . . 0027y
,
f y i i y i i y
R M R
2
0 0561 0 0534 0 0561 0 0561 0 0

δ δ δ δ
( )
+ −
( )
+ −
( )
− . . . . . 0027y.
When approached from the lef: for y = 0 f y i
1
0 0507

δ
( )
. , for y = 1 f y i
1
0 0534

δ
( )
. ,
and when approached from right: for y = 0
f y i
2
0 0561

δ
( )
.
and for y = 1 f y i
2
0 0534

δ
( )
.
values are obtained,
Te fuzzy number

d
1
is calculated below by the help of Eq. 27.
.
663 Technological and Economic Development of Economy, 2009, 15(4): 646–669

d
1
1
0 3366 12 241188460
188777796
1 0 0487
1 0 0561

|
(
'
`
J
J
+
+
|
(
'
`
ln
.
.
.
JJ
J
|
(
'
'
'
`
J
J
J
J
|
(
'
'
'
|
(
'
'
`
J
J
J
12
0 3366
0 3366 12
2
241188460
18
.
.
;
ln
e
88777796
1 0 0513
1 0 0534
1
0 3366 12
12
0 3366
0 33
|
(
'
`
J
J
+
+
|
(
'
`
J
J
. .
.
.
.
e
666 12
2
1
0 3366 241188460
188777796
|
(
'
'
`
J
J
J
|
(
'
'
'
`
J
J
J
J
|
(
'
`
J
J
;
ln
. 112
12
0 3366
0 3366 12
2 1 0 0539
1 0 0507
+
+
|
(
'
`
J
J
|
(
'
'
'
`
J
|
(
'
'
`
J
J
J
.
.
.
.
e
JJ
J
J
`
J
J
J
J
,



d
1
0 720767 0 772595 0 824428
( )
. ; . ; . .
Using the defuzzifcation method suggested for the fuzzy numbers which have normal
distribution,

d
1
0 720767 0 772595 0 824428
( )
. ; . ; . is defuzzified and d
2
is calculated. If
τ τ τ
11 12 13
0 50 0 30 0 20 . ; . ; . is accepted;
d
1
0 50 0 720767 0 30 0 772595 0 20 0 824428
0 50 0 30 0 20

⋅ + ⋅ + ⋅
+ +
. . . . . .
. . .
0 757048 . ,
d
2
0 757048 0 3366 12 0 408968 − − . . . .
Te areas under the standard normal distribution curve are given below for d
1
and d
2
:
N d ( ) .
1
0 775489 ,
N d ( ) .
2
0 341281 .
Te fuzzy real options value is calculated with the help of Eq. 31.

FROV FROV FROV FROV S i N d X i N d
L
T
r
R
T
+
( ) ( ( )
− +
( )


( ; ; )
1 2 3 0 1
1 1
δ
  
22
0 1 2 0 1
1 1 1
( )
+
( ) ( )
− +
( ) ( )
+
( ) ( )
− − −
;
; S i N d X i N d S i N d
M
T
r
M
T
R
T
δ δ
   
−− +
( ) ( ))

 
X i N d
r
L
T
1
2
;


FROV
1
1
52931124 229129037 253247883 429445796 1 0 0507
( )
+
( )

; ; ; .
22
0 775489
79338906 179338906 198216686 298216686 1 0 05
⋅ −
( )
+
.
; ; ; . 339 0 341281
31531751 62125072 75888533 169545362
12
( )


( )

.
; ; ; .

FROV
1
is found by defuzzying −
( )
31531751 62125072 75888533 169545362 ; ; ; through
the centroid method and found to be $ 63082694.
664 İ. Uçal, C. Kahraman. Fuzzy real options valuation for oil investments

FROV
2
1
52931124 229129037 253247883 429445796 1 0 0534
( )
+
( )

; ; ; .
22
0 775489
79338906 179338906 198216686 298216686 1 0 05
⋅ −
( )
+
.
; ; ; . 1 13 0 341281
33850206 58064338 71617684 163532227
12
( )


( )

.
; ; ; .

FROV
2
is found by defuzzying −
( )
33850206 58064338 71617684 163532227 ; ; ; through
the centroid method and found to be $ 70256880.

FROV
3
1
52931124 229129037 253247883 429445796 1 0 0561
( )
+
( )

; ; ; .
22
0 775489
79338906 179338906 198216686 298216686 1 0 04
⋅ −
( )
+
.
; ; ; . 8 87 0 341281
36199402 54065743 67422616 157687761
12
( )


( )

.
; ; ; .

FROV
3
is found by defuzzying −
( )
36199402 54065743 67422616 157687761 ; ; ; through
the centroid method and found to be $ 84940120.
FROV
( )
63082694 70256880 84940120 ; ; $.
FROV is found by defuzzying 63082694 70256880 84940120 ; ; $
( )
through the centroid
method and found to be $ 66734120.
Te data in the example of oil feld investment are applied to Eq.s 33 and 34, and d
1
and
d
2
values are found by postponing the defuzzifcation.

d
1
1
0 3366 12
12
0 52931124
298216686
1 0513
1 0534

|
(
'
`
J
J
|
(
'
`
J
J
ln
.
.
. ..
.
,
3366
0 3366 12
2
229129037
198216
e
|
(
'
'
`
J
J
J
|
(
'
'
'
`
J
J
J
J
|
(
'
'
'

ln
6686
1 0513
1 0534
1 12
0 3366
0 3366 12
2 |
(
'
`
J
J
|
(
'
`
J
J
|
|
(
'
'
`
J
J
J
σ T
e
.
.
.
.
((
'
'
'
`
J
J
J
J
|
(
'
`
J
J
|
(
'
`
J
J
,
ln
.
.
253247883
179338906
1 0513
1 0534
1
σ T
112
0 3366
0 3366 12
2
429445796
79338906
.
.
,
ln
e
|
(
'
'
`
J
J
J
|
(
'
'
'
`
J
J
J
J
|
((
'
`
J
J
|
(
'
`
J
J
|
(
'
'
'
|
(
'
'
`
J
J
J
1 12
0 3366
0 3366 12
2 1 0513
1 0534
σ T
e
.
.
.
. ``
J
J
J
J
`
J
J
J
J
;

d
1
1 041530 0 565438 0 737106 1 889470 − ( . ; . ; . ; . );
d
2
1 041530 0 3366 12 0 565438 0 3366 12 0 737106 0 3366 12 − − − − ( . . ; . . ; . . ;; . . ) 1 889470 0 3366 12 −
d
2
1 041530 0 3366 12 0 565438 0 3366 12 0 737106 0 3366 12 − − − − ( . . ; . . ; . . ;; . . ) 1 889470 0 3366 12 − ;
d
2
2 207547 0 598500 0 428911 0 723453 − − − ( . ; . ; . ; . ).
665 Technological and Economic Development of Economy, 2009, 15(4): 646–669
In the model suggested for the defuzzifcation of the membership functions show-
ing normal distribution: using the weights τ
11
0 11 . ; τ
12
0 41 . ; τ
13
0 39 . ; τ
14
0 09 . ;
the fuzzy number d
1
1 041530 0 565438 0 737106 1 889470 − ( . ; . ; . ; . ) is defuzzyfied, and
using the weights τ
21
0 08 . ; τ
22
0 35 . ; τ
23
0 40 . ; τ
24
0 17 , ; the fuzzy number
d
2
2 207547 0 598500 0 428911 0 723453 − − − ( . ; . ; . ; . ) is defuzzifed.
d
1
0 11 1 041530 0 41 0 565438 0 39 0 737106 0 09 1 889

⋅ − + ⋅ + ⋅ + ⋅ . ( . ) . . . . . . 4470
0 11 0 41 0 39 0 09
0 5748
. . . .
.
+ + +
,
d
2
0 08 2 207547 0 35 0 598500 0 40 0 428911 0 17

⋅ − + ⋅ − + ⋅ − + . ( . ) . ( . ) . ( . ) . ⋅⋅
+ + +

0 723453
0 08 0 17 0 35 0 40
0 4346
.
. . . .
.
N d
1
0 7173
( )
. ,
N d
2
0 3319
( )
. .
When the values are placed in Eq. 14, the fuzzy real options value in the case of postpon-
ing the defuzzifcation of costs and revenues with defuzzifying the probabilities is calculated
as below:
FROV S X
( )

( )
− −
 
0
12 12
1 0534 0 7173 1 0513 0 3319 . . . . ,

FROV
( )
⋅ − 52931124 229129037 253247883 429445796 0 3842
79338
; ; ; .
9906 179338906 198216686 298216686 0 1821
33965177 5194
; ; ; .
;
( )

− 22806 64647237 150555219 ; ; $.
( )

Te real options value of the investment is found by defuzzifying the calculated value
through the centroid method and found to be $ 77408863.
Using Eq. 36, the fuzzy real options value is calculated without defuzzifying the pos-
sibilities.

FROV N N
( )
⋅ ⋅ − ⋅

1 0534 52931124 1 04153 229129037 0 5654
12
. ( ( . ); ( . );;
( . ); ( . )
. (
253247883 0 7371 429445796 1 8895
1 0513 7933
12
⋅ ⋅ −
( )

N N
88906 2 207547 179338906 0 598500
198216686 0
⋅ − ⋅ ≤ −
⋅ −
N N
N
( . ); ( . );
( .4428911 298216686 0 723453 ); ( . )) ⋅ N

FROV − ( ; ; ; ) 120991904 51323816 77347074 222672378 .
FROV is found by defuzzfying ( ; ; ; ) −120991904 51323816 77347074 222672378 through
the centroid method and found to be $ 105867783.
,
;
666 İ. Uçal, C. Kahraman. Fuzzy real options valuation for oil investments
Table 2. Te results of the application with suggested model
Fuzzy Result Defuzzifed Result
Discrete Compounding
(–34660540; 58012568;
71686162; 164359270)$
72971941$
Fuzzifying the discrete interest rates
(63082694; 70256880;
84940120)$
66834120$
Postponing the defuzzifcation of costs and
revenues with defuzzifying the probabilities
(–33965177; 51942806;
64647237; 150555219)$
77408863$
Postponing the defuzzifcation of costs
and revenues without defuzzifying the
probabilities
(–120991904; 51323816;
77347074; 222672278)$
105867783$
Table 2 represents the application results of the suggested model together. Te last achieved
value is the most sensitive value which means that the case of postponing the defuzzifcation
of costs and revenues without defuzzifying the probabilites shows more information about
the investment. Consequently, the result of the case of postponing the defuzzifcation of
costs and revenues without defuzzifng the probabilites is the fuzzy number which has the
widest range.
6. Conclusion
Real options valuation method, which is vastly diferent from the traditional investment valu-
ation methods, makes more exact assessments since it considers future uncertainties as well
as dependencies and dynamism. By using the real options valuation method particularly to
analyse the risky investments, wrong decisions could be easily avoided. When examined by
real options, an investment, that is rejected because its current value is found to be negative
by using the discounted cash fow analysis, may deliver positive values in case of postponing
the investment and, therefore, a method that does not consider the postponing option could
miss the investment opportunity.
Due to the fact that estimations made by people have multilateral structures that do not
have sharp certainties, fuzzy logic is more expressive than classical mathematics. Terefore,
on consideration of a fuzzy manner, the parameters used in the real options valuation method
will lead to more realistic results which concern human reasoning. Tus, the results obtained
can be more trusted.
Using real options valuation methods to analyse an investment decision reduces the un-
certainty to minimum and it ensures that the investment assessment is made in as the most
realistic way as possible. Te model suggested by Carlsson and Fuller (2003) has been found
to be the most frequently used method amongst the fuzzy real options assessment methods
during literature researches. In this model, however, it is observed that the fuzzy revenue
and expenditure values were defuzzifed at a relatively early stage. Early defuzzifcation of the
fuzzy parameters causes information loss. In this study, a new model has been suggested; it
667 Technological and Economic Development of Economy, 2009, 15(4): 646–669
postpones the defuzzifcation of fuzzy parameters in the real options valuation in order to
avoid this information loss. Carlsson and Fuller’s model has been re-arranged for discrete
compounding and then the defuzzifcation of revenue and expenditures has been postponed
and relavant equations have been formed for the cases of early defuzzifying probabilities and
defuzzifying them at the last stage. On the other hand, the diference between the applications
of this new model and of the early defuzzifying model has been found to be the information
loss due to the early defuzzifcation.
For further research, it is suggested that the information loss caused by uncertainty should
be measured by fuzzifying the other real options valuation methods that have a broad area
of usage such as game theory, binomial, and trinomial valuation methods.
References
Abadie, L. M.; Chamorro, J. M. 2008. Monte Carlo valuation of natural gas investments, Review of Financial
Economics 18(1): 10–22. doi:10.1016/j.rfe.2008.10.002.
Allenotor, D.; Tulasiram, R. K. 2007. A grid resources valuation model using fuzzy real option, Lecture
Notes in Computer Science, Parallel and Distributed Processing and Applications, Springer, Vol. 4742:
622–632. doi:10.1007/978-3-540-74742-0_56.
Bergmann, A.; Hanley, N.; Wright, R. 2006. Valuing the attributes of renewable energy investments,
Energy Policy 34(9): 1004–1014. doi:10.1016/j.enpol.2004.08.035.
Birol, F. 2007. Energy economics: a place for energy poverty in the Agenda?, Te Energy Journal 28(3):
1–6.
Black, F.; Scholes, M. 1972. Te valuation of option contracts and a test of market efciency, Te Journal
of Finance 27(2): 399–417. doi:10.2307/2978484.
Black, F.; Scholes, M. 1973. Te pricing of options and corporate liabilities, Journal of Political Economy
81(3): 637–654. doi:10.1086/260062.
Brach, M. A. 2003. Real Options in Practice. John Wiley&Sons, USA.
Buckley, J. J. 1987. Te fuzzy mathematics of fnance, Fuzzy Sets and Systems 21: 257–273.
doi:10.1016/0165-0114(87)90128-X.
Carlsson, C.; Fuller, R. 2003. A fuzzy approach to real option valuation, Fuzzy Sets and Systems 139:
297–312. doi:10.1016/S0165-0114(02)00591-2.
Carlsson, C.; Fullér, R.; Majlender, P. 2001. Project selection with fuzzy real options, in Proceedings of the
Second International Symposium of Hungarian Researchers on Computational Intelligence, Budapest,
81–88.
Chorn, L. G.; Shokhor, S. 2006. Real options for risk management in petroleum development investments,
Energy Economics 28(4): 489–505. doi:10.1016/j.eneco.2006.03.002.
Dinca, C.; Badea, A.; Rousseaux, P.; Apostol, T. 2007. A multi-criteria approach to evaluate the natural
gas energy systems, Energy Policy 35(11): 5754–5765. doi:10.1016/j.enpol.2007.06.024.
Favero, C. A.; Pesaran, M. H. 1991. Oil investment in the North sea, Oxford Institute for Energy Studies
EE15: 24.
Garcia, F. A. A. 2004. Fuzzy real option valuation in a power station reengineering project, in Sof Com-
puting with Industrial Applications – Proceedings of the Sixth Biannual World Automation Congress,
Spain: 281–287.
Grafström, C. and Lundquist, L. 2002. Real Option Valuation vs. DCF Valuation: An application to a North
Sea Oilfeld. Tesis in Financial Economics, Stockholm University.
668 İ. Uçal, C. Kahraman. Fuzzy real options valuation for oil investments
Han, L.; Zheng, C. 2005. Fuzzy options with application to default risk analysis for municipal bonds in
China, Nonlinear Analysis 63(5–7): e2353–e2365. doi:10.1016/j.na.2005.02.019.
International Energy Agency (IEA). 2007. World Energy Outlook 2007: China and India insights, Inter-
national Energy Agency, Paris.
International Energy Agency (IEA). 2008. Worldwide Trends in Energy Use and Efciency, International
Energy Agency, Paris.
Kahraman, C.; Ucal, İ. 2008. Fuzzy real options valuation for oil investments, in Proceedings of the 8
th

International FLINS Conference, 1027–1032.
Kjærland, F. 2007. A real option analysis of investments in hydropower – the case of Norway, Energy
Policy 35(11): 5901–5908. doi:10.1016/j.enpol.2007.07.021.
Leslie, K. J.; Michaels, M. P. 1997. Te real power of real options, Te McKinsey Quart. 3: 5–22.
Lin, Q. G.; Huang, G. H.; Bass, B.; Qin, X. S. 2009. IFTEM: An interval-fuzzy two-stage stochastic optimi-
zation model for regional energy systems planning under uncertainty, Energy Policy 37(3): 868–878.
doi:10.1016/j.enpol.2008.10.038.
Liou T. S.; Wang, M. J. 1992. Ranking fuzzy numbers with integral value, Fuzzy Sets and Systems 50:
247–255. doi:10.1016/0165-0114(92)90223-Q.
Locatelli, C. 2006. Te Russian oil industry between public and private governance: obstacles to inter-
national oil companies’ investment strategies, Energy Policy 34(9): 1075–1085.
doi:10.1016/j.enpol.2004.09.007.
Majlender, P. 2008. Sof decision support systems for evaluating real and fnancial investments, in Fuzzy
Engineering Economics with Applications (Ed. C. Kahraman), Studies in Fuzziness and Sof Computing
233, Springer Verlag, 307–338. doi:10.1007/978-3-540-70810-0.
Menegaki, A. 2008. Valuation for renewable energy: a comparative review, Renewable and Sustainable
Energy Reviews 12(9): 2422–2437. doi:10.1016/j.rser.2007.06.003.
Mohn, K. and Misund, B. 2009. Investment and uncertainty in the international oil and gas industry,
Energy Economics 31(2): 240–248. doi:10.1016/j.eneco.2008.10.001.
Mun, J. 2002. Real Option Analysis Tools and Techniques for Valuing Strategic Investments and Decisions.
John Wiley & Sons, USA.
Paddock, J. L.; Siegel, D. R.; Smith, J. L. 1988. Option valuation of claims on real assets: the case of
ofshore petroleum leases, Te Quarterly Journal of Economics 103(3): 479–508.
doi: 10.2307/1885541.
Rocha, K.; Salles, L.; Garcia, F. A. A.; Sardinha, J. A.; Teixeira, J. P. 2007. Real estate and real options – a
case study, Emerging Markets Review 8(1): 67–79. doi: 10.1016/j.ememar.2006.09.008.
Rodríguez, R. Y. 2008. Real option valuation of free destination in long-term liquefed natural gas sup-
plies, Energy Economics 30(4): 1909–1932. doi: 10.1016/j.eneco.2007.03.006.
Ross, T. J. 1995. Fuzzy Logic with Engineering Applications. Mc-Graw Hill, USA.
Tao, C.; Jinlong, Z.; Shan, L.; Benhai, Y. 2007. Fuzzy real option analysis for IT investment in nuclear power
station, Lecture Notes in Computer Science, ICCS 2007, Springer, Vol. 4489: 953–959. doi:10.1007/978-
3-540-72588-6_152.
Tolga, A. Ç.; Kahraman, C. 2008. Fuzzy multiattribute evaluation of R&D projects using a real options
valuation model, International Journal of Intelligent Systems 23(11): 1153–1176.
doi:10.1002/int.20312.
Trigeorgis, L. 1996. Real Options Managerial Flexibility and Strategic in Resource Allocation. Te MIT
Press Cambridge, London.
669 Technological and Economic Development of Economy, 2009, 15(4): 646–669
US.DoE. 2008. International energy outlook, US department of energy – energy information adminis-
tration, valuation for renewable energy: A comparative review, Renewable and Sustainable Energy
Reviews 12(9): 2422–2437.
Yang, M.; Blyth, W. 2007. Modeling investment risks and uncertainties with Real Options approach, IEA
Working Paper, 27.
Zadeh, L. 1965. Fuzzy sets, Inform and Control 8: 338–353.
Zeng, M.; Wang, H.; Zhang, T.; Li, B.; Huang, S. 2007. Research and application of power network invest-
ment decision-making model based on fuzzy real options, in Service Systems and Service Management,
International Conference on, 9–11 June, 1–5.
REALIŲ ĮVERČIŲ METODO TAIKYMAS INVESTICIJOMS
Į NAFTOS VERSLĄ VERTINTI
İ. Uçal, C. Kahraman
Santrauka
Tradiciniai vertinimo metodai yra mažiau patikimi esant neapibrėžtumams. Vadinasi, kiti metodai, tokie
kaip realių pasirinkčių vertinimo modeliai, kurie gali minimizuoti neapibrėžtumus, tampa svarbesni.
Šiame straipsnyje nagrinėjamas hibridinis Carlsson ir Fuller metodas, kuris buvo panaudotas diskrečia-
jam rizikingų pinigų srautų modeliavimui. Pasiūlytas naujas realių pasirinkčių vertinimo modelis, kuris
realistiškiau įvertins investicijas, rodiklius apibūdinančią neapibrėžtą informaciją apdorojant ankstyvojoje
stadijoje. Pasiūlytas modelis buvo pritaikytas investicijoms į nafos verslą modeliuoti, nustatytas infor-
macijos nuostolis, kuris atsiranda dėl ankstyvo neapibrėžtų duomenų apdorojimo.
Reikšminiai žodžiai: neapibrėžtos aibės, realių pasirinkčių vertinimas, neapibrėžtos pasirinktys, in-
vestavimas.
İrem UÇAL. MSc, Research Assistant. Department of Industrial Engineering. Istanbul Technical Uni-
versity. First degree in environmental engineering, Istanbul Technical University (2004). Master of Sci-
ence in engineering management, Istanbul Technical University (2008). PhD in industrial engineering
(continued), Istanbul Technical University. Research interests: engineering economics, multiple criteria
decision making and energy resources.
Cengiz KAHRAMAN. Prof. Department of Industrial Engineering. Istanbul Technical University. First
degree in Industrial Engineering Istanbul Technical University (1988). Master of Science (1990). Doctor
(1996). Assoc. Prof. (1997). Professor (2003). Author of about 70 scientifc articles. Research interests:
engineering economics, multiple criteria decision making, quality management and control, statistical
decision making, fuzzy sets applications to decision making.
Copyright of Technological & Economic Development of Economy is the property of Technological &
Economic Development of Economy and its content may not be copied or emailed to multiple sites or posted to
a listserv without the copyright holder's express written permission. However, users may print, download, or
email articles for individual use.

Technological and Economic Development of Economy, 2009, 15(4): 646–669

647

The operating flexibility and strategic value aspects of various projects cannot be properly captured by traditional DCF techniques because of their discretionary asymmetric nature and their dependence on future events that are uncertain at the time of the initial decision (Trigeorgis 1996). Brach (2003) identified three fundamental differences between discounted cash flows methods and real options valuation methods. The first difference is that in the discounted cash flows method, decisions cannot be changed in the future, while in the real options valuation (ROV) method it is likely to perform directional changes propped up by obtaining new information. Secondly, in DCF method estimated cash flows group is considered as a base, whereas in ROV method, cash flows depend on the ambiguous conditions in the future. Thirdly, in the former method, sensitivity and scenario analyses are static; while in the latter there is a managerial flexibility in order to provide adaptability to changing conditions. As a result, it can be claimed that the differences arise in decision variation, dependencies, and dynamism. Real options analysis has appeared as a tool to give more distinct results than traditional DCF analysis in investment projects by giving the option of postponing the project to a later time or abandoning the project whenever it is necessary. Mun (2002) explains real options as a systematic approach and integrated solution making use of financial theory, economic analysis, management science, decision sciences, statistics, and econometric modeling in applying options theory to valuing real physical assets, as opposed to financial assets defined as a dynamic and uncertain business environment where business decisions are flexible in the context of strategic capital investment decision making, valuing investment opportunities and project capital expenditures. The general analysis of real options as a strategic tool rather than a mere valuation tool, that is, proactive rather than just reactive flexibility represents an advance on current thinking in this area (Leslie and Michaels 1997). Real options theory provides a method to better valuate investment projects in the presence of managerial flexibilities as information option, waiting option and abandonment option (Rocha et al. 2007). ROV is practically the same as the valuation of financial options; yet, there exist a number of differences. A financial option gives the holder of the right to buy or sell a specified quantity of an underlying asset at a fixed price (it is also called a strike price or an exercise price) at or before the expiration date of the option. Since it is a right and not an obligation, the holder can prefer not to exercise the right and allow the option to expire. There are two types of options - call options and put options. In a call option, the buyer of the option has the right, but not the obligation, to buy an agreed quantity of a particular commodity or financial instrument from the seller of the option at a certain time for a certain price, and the seller is obligated to sell the commodity or financial instrument if the buyer exercises the option. The put option allows its buyer the right but not the obligation to sell a commodity or financial instrument to the seller of the option at a certain time for a certain price, and with respect to that the seller has the obligation to purchase the underlying asset at that strike price, if the buyer exercises the option. There are two main types of options: American options and European options. A primary distinction between American and European types of options is that American options can be exercised at any time prior to their expiry date, while European options can be exercised only at expiration. The possibility of early exercise makes American options more valuable than otherwise similar European options; it also makes them more difficult

The value of options cannot be controlled by changing the stock prices in financial options. In other words. management acceptance and movements are not influential on the valuation of financial options. when Black and Scholes (1972) published their pioneering paper suggesting a model for valuing dividend-protected European options. The first one is that the life span of financial options is shorter than real options. this partial differential equation can be solved to evaluate many types of options. In most cases. Although their derivation is mathematically complicated. while they determine the value of real options. σ: uncertainty of expected cash flows. decreasing the variance methods. In valuing real options. X: present value of fixed costs. Financial options are commercially secured and they can be marketed with price information and likewise. we often use the Black Scholes model which has fairly limited applicability. whereas in real options. There is one compensating factor that enables the former to be valued using models designed for the latter. The values of the financial options are usually small. market or competition effects determine the strategic value of strategic options. Given the wide appropriate boundary conditions. such as American and compound options. Fuzzy real options valuation for oil investments to value. . Mun (2002) emphasized nine distinctive characteristics of financial and real options. r: risk-free interest rate. C. δ: value lost over duration of the option. Most real options are not analogs of European options (Black and Scholes 1973). whereas real options are usually solved by applying closed formed equations and the simulation of the main variables with two binomial letters. However. while the values of the real options are usually extremely large. there is a simpler binomial model for valuing options that draws on the same logic. while real options are not commercially in use and they are naturally private and they do not have market resemblance. Option pricing theory has made vast strides since 1972. where S: present value of expected cash flows.648 İ. Financial options are usually solved by applying closed formed partial differential equations and simulation. Lastly. Kahraman. while real options are newly developed to be used in financial activities of the corporations for the last few years. (1) (2) where d1 = ln( S0 / X ) + (r − δ + σ2 / 2)T σ T . the holders of in the money options will generally get much more by selling the option to someone else than by exercising the options. Financial options have been used for the past 30 years. In financial options the main variable determining the value is the price of the financial asset. Black and Scholes used a “replicating portfolio” to come up with their final formulation. management and competition. d 2 = d1 − σ T . In financial options market or competition effects are irrelevant in determining the value and pricing of options. the Black Scholes partial differential equation itself has a wider applicability. the time premium associated with the remaining life of an option and transactions costs makes early exercise sub-optimal. while in real options the main variables determining the value are the cash flows determined by the demand. Uçal. The value of a real option is computed by (Leslie and Michaels 1997): ROV = S0 e − δT N ( d1 ) − Xe − rT N ( d 2 ) . while in real options the value of strategic options can be increased by managerial decisions and flexibility.

it has been utilized in the model. there is an application of this new model. not everything can be described using binary valued sets. with imprecision. For an element that belongs to a set of all possible elements and to any given specific subset. 2. 2009. Zadeh (1965) first founded the fuzzy set theory and he suggested that the notion of a fuzzy set provides a convenient point of departure for the construction of a conceptual framework which is parallel in many respects to the framework used in the case of ordinary sets. but it is more general than the latter and. Brach (2003) pointed out the equivalence of the real options parameters in financial options parameters as follows: the exercise price used in financial options indicates costs to acquire the asset. Unlike crisp theories. The rest of the paper is organized as follows. In Section 5. and the variance of stock returns is replaced by riskiness of the asset or. but it is problematic to classify them as being young or not. Fuzzy set theory is a marvelous tool for modeling the kind of uncertainty associated with vagueness. variance of the best and worst case scenario. Literature reviews on fuzzy real options valuation and real options valuation applications in oil investment valuations are given in Sections 2 and 3. respectively. and N(d): cumulative normal distribution function. a person belongs to the set of all human beings and given a specific subset. one can say whether or not each particular person belongs to this set. For example. which increases its usability in practice. Such a problem is often distorted so that it can be described using the well-known existing methodology. The proposed model postpones defuzzification to the very end stages of the process. Unfortunately. fuzzy logic enables vagueness and ambiguity as well as it avoids clear distinctions and limits. A fuzzy set is a set containing elements that have .Technological and Economic Development of Economy. The aim of this study is to propose a new fuzzy real options valuation model which evaluates investments in a more realistic way. potentially. particularly in the fields of pattern classification and information processing. in other words. 15(4): 646–669 649 t: time to expiry. Since fuzzy logic fits best to the uncertain nature of the investment decisions. and/or with a lack of information regarding a particular element of the problem at hand (Ross 1995). With fuzzy logic an element could partially belong to a set and this is represented by the set membership. Another superiority of our model is that the conversion of continuous compounding into discrete compounding provides more intelligibility. In the last section we conclude the obtained results. time to expiration stands for the length of time option viable. such as all males. stock price accounts for the present value of the future cash flows from the asset. may prove to have a much wider scope of applicability. we propose a model of fuzzy real options valuation. Fuzzy logic gives a method to simulate the ability of human reasoning. The classifications of persons into males and females is easy. The set of young people is far more difficult to define as there is no distinct cut-off point at which age youngness begins or ends. This is not a measurement problem and measuring the age of all elements more precisely is not helpful. In Section 4. it can be said exactly whether that element is or is not a member of it. Fuzzy real options In classical mathematics the binary valued logic and set theory are used. This prevents the loss of information at the beginning of the process.

b] and (a–α) is the downward potential and (b+β) is the upward potential for A. x2 . Carlsson et al. b. A fuzzy number is a fuzzy set which is both normal and convex. where is the estimated present value of expected   cash flows. β′ ) is estimated present value of expected costs. such as bell shaped or gaussian fuzzy numbers.  T   . Triangular fuzzy numbers are defined by three parameters. α. (4) where   E ( S0 )   σ2 ln  +r −δ+  E( X )    2   d1 =  σ T d 2 = d1 − σ T . A = (a. while trapezoidal require four parameters. A fuzzy set is prescribed by vague or ambiguous properties hence its boundaries are ambiguously specified (Ross 1995). They suggested Eq. α′. which is one of the mostly used real options valuation approaches. Carlsson and Fuller (2003) improved a fuzzy approach for real options valuation. βe − δT N ( d1 ) + α′e − rT N ( d 2 ) . where 0 means the element is not a member of the set. s2 e − δT N ( d1 ) − x1e − rT N ( d 2 ) (7) αe − δT N ( d1 ) + β′e − rT N ( d 2 ) . Fuzzy sets have been used for valuating real options in the literature. 4 for computing fuzzy real options values. Fuzzy real options valuation for oil investments varying degrees of membership between 0 and 1 in the set. β ) e − δT N ( d1 ) − ( x1 . and σ = σ S0 is the possibilistic variance of the present value of expected cash flows.1] (3) which means the most possible values of expected A lie in the interval [a. 3 for arithmetic operations on trapezoidal fuzzy numbers they find the fuzzy real options formula as below: FROV = ( s1 . where the present values of expected cash flows and expected costs are estimated by trapezoidal fuzzy numbers and they determined the optimal exercise time with the assistance of possibilistic mean value and variance of fuzzy numbers. Uçal. and the values between these two values gives a partial membership of the element on the set. x2 . b + (1 − γ)β] . E X  possibilistic mean value of expected costs. Kahraman. X = ( x1 . s2 . α′. Most common types of fuzzy numbers are triangular and trapezoidal. β) is a trapezoid fuzzy number which can be shown by [ A]γ = [ a − (1 − γ)α. yet these types of fuzzy numbers are rarely used in literature. α. β′ ) e − rT N ( d 2 ) = (s e 1 − δT N ( d1 ) − x2 e − rT N ( d 2 ) . C.650 İ. E S0 denotes  denotes the the possibilistic mean value of the present value of expected cash flows. ( ) ( ) ( )   FROV = S0 e − δT N ( d1 ) − Xe − rT N ( d 2 ) . (2001) applied fuzzy real options on the project selection by identifying the optimal path of the dynamic decision trees with the biggest real option value in the end of the planning period. (5) (6) Using Eq. ∀γ ∈ [ 0. ) . 1 means the element is completely a member of the set which means full membership. Other types of fuzzy numbers are possible. by applying a heuristic real option rule in a fuzzy setting.

Energy consumption increases sharply in parallel with the increase in population. success probability. (2007) compared the application of traditional net present value (NPV) method with real options in investment evaluations. Tolga and Kahraman (2008) considers the multidimensional and vague side of the R&D project selection process. Majlender (2008) proposed the application of a possibilistic decision rule for optimal investment strategy. The fuzzy analytic hierarchy process. Han and Zheng (2005) applied fuzzy options to fail to perform risk analysis for municipal bonds in China. with consumption increasing by 2. According to the Worldwide Trends in Energy Use and Efficiency 2008 report (IEA 2008). The main energy resources are fossil fuels. driven by their use in transport in which energy consumption has grown most quickly by rising passenger travel and freight transport. 2009. etc. finally. nuclear. oil. the model suggested by Carlsson and Fuller (2003) has been applied to timing decision and selection of power station for various investment alternatives. Energy investments Energy is one of the necessary elements that is required for human life and industrial production. (2007) developed a comprehensive methodology based on fuzzy risk analysis and real options approach to evaluate information technologies investments in a nuclear power station.1 percent and 2. natural gas.Technological and Economic Development of Economy. trends. oil products remained the most important final energy commodity with a global share of 37% in 2005. solar. the importance of investment in new resources increases.2 percent from 2005 to 2030. The survey carried out by the International Energy Outlook 2008 (US. Tao et al. Kahraman and Ucal (2008) used the certainty equivalent approach for real options valuation in an oil investment with fuzzified data. hydropower. Zeng et al. Renewable energy and coal are the fastest growing energy sources. Due to a relative importance of the transport sector in the OECD countries. Based on fuzzy real option valuation. Planning of the exploition of energy resources has become a very crucial issue. which takes monetary (fuzzy real option value) and nonmonetary (capability. biomass and geothermal. 15(4): 646–669 651 Garcia (2004) used the fuzzy real options valuation (FROV) model in a real investment project from the energy sector.0 percent respectively. industrial and technologic developments. The proposed decision rule had to be reapplied every time when new information arrived during the deferral period to be able to analyze how the optimal investment strategy should be changed in the light of the new information. standard of living. and renewable energy resources: wind. Given insufficient existing resources. In Garcia’s paper. and discussed how to make the investment decision when investment cost and cash flow are both fuzzy numbers. 3.) criteria into account. oil products accounted for 47% of total final energy . Allenotor and Thulasiram (2007) used a fuzzy trinomial real options model on pricing grid resources and proved the feasibility of the model through experiments. coal. analyzed the uncertainty of power grid investment project. was used to make this selection among alternative R&D projects.DoE 2008) shows that liquid fuels consumption increases at an average annual rate of 1. and it is estimated that the World energy consumption increments by 50% from 2005 to 2030. they performed a simulation model to show the validity. They also introduced fuzzy expectation and fuzzy variance concept to construct evaluation model and.

trucks. uncertainty. and other vehicles. A typical oil investment project has three main phases which are exploration. Mohn and . accounting on average for 28% of total final energy use in 2005. These data reveal how important and critical oil production is throughout the whole world. Kjærland (2007) presented a valuation study of hydropower investment opportunities in the Norwegian context. In the literature there are lots of models which are used to valuate energy investments. a country which has set particularly ambitious targets for expanding renewable energy. ROA is relatively new in assessing investments in climate change projects which handles the investment problem and uncertainty in a particular way. and extraction. Uçal. Locatelli (2006) explored the investment strategies of oil companies of Russia. In many of these countries oil products are the most important natural resources. Kahraman. Paddock et al. As mentioned in the world energy outlook 2006 (Birol 2007). (1988) used option valuation theory to develop a new approach to value leases for offshore petroleum. development. Petroleum and petroleum-based chemicals are important in manufacturing plastic. Traditional appraisal methodologies for project investment can hardly incorporate the above three characteristics. (2006) estimated the magnitude of external costs and benefits such as landscape quality. concession and exploration. the developing countries whose economies and populations are growing fastest contribute 74% of the increase in global primary energy use where China and India alone account for 45% of this increase. aircraft. wildlife and air quality for the case of renewable technologies in Scotland. Fuzzy real options valuation for oil investments use in 2005. energetical and economic points of view. The exploration phase comprises three main activities: scouting. Chorn and Shokhor (2006) reported the union of a real options algorithm with the Bellman equation by allowing the analyst and management to avoid estimating outcome probabilities and computing expected values for investments with sequential stages using project experts’ technical insight. ships. Oil investment – three-quarters of which goes to the upstream – amounts to over $4 trillion in total over 2005–2030. which in many areas are generated by burning natural gas. and diesel fuel to run cars. fertilizers. and many other goods. and electricity. Menegaki (2008) used environmental cost–benefit analysis for the evaluation of renewable energy projects. C.6%. wax. Global primary energy demand in the Reference Scenario is projected to increase by just over one-half between 2006 and 2030 – an average annual rate of 1.652 İ. Pesaran 1991). They are used as gasoline. Oil products also have the largest share of consumption in non-OECD countries. jet fuel. the need to curb the growth in fossil-energy demand is more urgent than ever. (2007) aimed to select the optimal energetic scenario applied to a consumer with 100 000 inhabitants from the residential–tertiary sector. C. irreversibility. According to world energy outlook 2007 (IEA 2007). For example. from the ecological. Home heat sources include oil. Investment in the power sector has three important characteristics. natural gas. Dinca et al. With this huge demand oil investment decisions become more important in global economy. Development activity can be separated into two stages: appraisal and technical development activities and the two aspects of the extraction phase are particularly relevant to econometric modeling: operating expenditure and taxation (Favero. Bergmann et al. lubricants. it focuses on the timing of the investment decision whether to do the project or not (Yang and Blyth 2007). to increase geographic and fuel-supply diversity and to mitigate climate-destabilizing emissions. and flexibility.

This method has been used to defuzzify fuzzy triangular numbers. There are some factors on the evaluation of energy investments: longer time of return of energy investments comparing to other investments. They are summarized as follows: 4. Defuzzification methods used in the model Defuzzification is the process of producing a quantifiable result in fuzzy logic.1. A lower optimism coeffiecient leads a risk-averse decision maker and likely a higher optimism coeffiecient leads a risk-seeking decision maker. and it ranks triangular fuzzy number F = (a. just as justification is the conversion of a precise quantity into a fuzzy quantity (Ross 1995).1. It is the conversion of a fuzzy quantity into a precise quantity. Total Integral Value Method Liou and Wang (1992) developed the total integral value method based on the mean of the integral value. 4. and a defuzzification method for normal distribution. We use three defuzzification methods: Total Integral Value Method. .Technological and Economic Development of Economy. c).5 αc + b + (1 − α ) a  . b. The uncertainty and risks of the energy investments are also increasing due to the changes in the energy policies and escalating energy requirements of the world. higher capital necessity and longevity of duration of management increase in the risk and ambiguity. Centroid Method. 4. 15(4): 646–669 653 Misund (2009) presented a micro-econometric study of corporate investment and uncertainty in a period of market turbulence and restructuring in the international oil and gas industry. Proposed model for real options valuation In this study we suggest a new model based on Carlsson & Fuller’s hybrid approach using discrete compounding instead of continuous compounding by defuzzifying the costs and revenues at a later stage than the based model and the based model has been improved by fuzzifying interest rates and probabilities. Rodríguez (2008) applied a real option model for the valuation of destination flexibility in long-term liquefied natural gas supplies.1]. Abadie and Chamorro (2008) used Monte Carlo approach to evaluate natural gas resources investments. (2009) developed a hybrid interval-fuzzy two-stage stochastic energy systems planning model and applied the model to a hypothetical regional energy system.   (8) where α is an index of optimism that represents the degree of optimism of the decision-maker and has a value between [0. 2009. The total integral value of F is defined as α IT ( A) = 0. The optimism coefficient also reflects the decision maker’s risk taking trend. Using fuzzy real options minimizes the effects of these uncertainties and risks on the investment. Lin et al.1.

2. 10 is suggested in order to defuzzify the fuzzy numbers that have a normal distribution. e rn = (1 + ir ) n  rn = ln(1 + ir ) n  r = ln(1 + ir )  ir = e r − 1 . Then. 4. τ1 + τ2 + τ3 + τ4 (10) Discrete compounding has more widespread usage in engineering economy community than continuous compounding. Uçal. 9 gives the algebric expression of this method (Ross 1995). Carlsson & Fuller’s (2003) real options formula is modified rn n by applying discrete compounding.654 4. t2. Under these conditions. Centroid Method İ. Discrete compounding τ1d1 + τ2 d 2 + τ3 d3 + τ4 d 4 . 1 these values have to provide the inequality of t1 > t2 > t3 > t4. Kahraman.1. since a balancing must be applied towards the summit point of the bell curve that is observed in normal distribution.3. To do this. The discrete interest rate is obtained by the below operations. the equation F = Pe = P(1 + i ) has been considered. Eq. t3. Fuzzy real options valuation for oil investments Centroid method is the most prevalent and physically appealing of all the defuzzification methods. for Fig. ∫ µ  ( z ) zdz . Eq. (9) z* = C  ∫ µC ( z )dz Although there are lots of other defuzzification methods available. 1. This method determines the centre of the area of the combined membership functions and mimics the center of gravity approach in physics.1. the centroid method has been chosen to use in this paper for its simplicity and fast computation. d= 4. → → → (11) d1 d2 d3 d4 Fig. C.2. Normal distribution . A defuzzification method for normal distribution Figure 1 demonstrates the probability values that have a normal distribution. If the weights used to balance the probability values are taken respectively as t1. and t4 as determined by the experts.

17 in Eq. the variable w has been designated for Eq. 15(4): 646–669 655 (12) eδn = (1 + iδ ) n  δn = ln(1 + iδ ) n  δ = ln(1 + iδ )  iδ = eδ − 1 . (14)   where S0 is the fuzzy present value of expected cash flows. and other symbols are the same as the ones in the crisp formulas.s 17 and 16 have been reached. 2009. d1     E( X )      If we substitute Eq. 1 + ir .Technological and Economic Development of Economy. Eqs. 15 in Eq. 1     E (S )  σ T  d 2 = ln   w     E( X )   4. Eq. Buckley’s membership function for a future value F is given by . When the discrete interest rate formulas acquired in the fuzzy real options formula are used instead of continuous interest rate in the equations. (15) w= 1 + iδ If we substitute Eq. T Tσ    E (S )  T    E (S )   1 + ir  T σ2 ln   w  + ln e 2  ln  + ln  +    E( X )      2    E (S )  T    1 + iδ   E( X )    (16)w d1 = = = ln       σ T σ T    E( X )   T σ2    E (S )  T  T σ2 1 2 σ T   ln   w + ln e  +   E( X )     2      E (S )  T  σ T  2   . We prefer using Buckley’s  notation. X is the fuzzy present value of fixed costs. Fuzzifying the discrete interest rates T σ σ T   2 e     2 T σ    1 T σ T   2 e (17)   −σ T . → → → where ir is risk-free interest rate in dicrete cases and iδ is the percentage value lost over the duration of option in discrete cases. 6. 13.   (18) In this subsection ir and iδ are accepted as triangular fuzzy numbers. 18 is obtained.3.     = = ln    w  e    E( X )   σ T        1    T σ T       = ln   E ( S )  σ T w σ e 2   . below formulas are reached:   E (S )   σ2 ln   +  (ln(1 + ir ) − ln(1 + iδ ) +   2  E( X )    d1 = σ T T    E (S )   1 + ir  T σ2  T ln   + ln   +   2  1 + iδ   =  E( X )  σ T (13) and −T −T   FROV = S0 (1 + iδ ) N ( d1 ) − X (1 + ir ) N ( d 2 ) . 16 and 17 are obtained. For operational convenience. 15 and after some simplifications Eq.

   m2 f2 y M   which is determined by ( ) ( ) ( ) (19)    f ni y M = fi y F 1 + f k ( y r ) ( ) ( )( ) −n (20)   for i = 1. 2 demonstrates the membership functions of the interest rates iδ and ir . . we obtain Eq. ir f 2 ( y  ) = irR ir f1 ( y  ) = iδ L iδ f 2 ( y  ) = iδ R iδ rM ( + y (i + y (i + y (i − irR − iδ L δM δM − iδ R ) ).2 where k = i for negative F . Membership functions of iδ and ir With substituting Eqs. 2. Kahraman. If it is substituted in Eq.s 17 and 18. 25. = 1 + iδ 1 + fi ( y iδ ) (25) where k = 3 – i and i = 1. . f1 ( y  ) = irL + y irM − irL . C. Uçal. w= 1 + ir 1 + f k ( y ir ) . ).656 İ. μ(ir) μ(iδ) f1 ( y  ) = irL + y irM − irL ir ( ) ) f1 ( y  ) = iδ L + y iδM − iδ L iδ ( ) ) f 2 ( y  ) = irR + y irM − irR ir ( f 2 ( y  ) = iδ R + y iδM − iδ R iδ ( irL irM irR ir iδL iδM iδR iδ Fig.2. ). 21–24. 21–24 into Eq. m3  .  T  1 1      E ( S )  σ T  1 + f k ( y ir )  σ  σ 2T     σ T      = ln  E ( S )  = ln     e d1        E( X )   1 + fi ( y iδ )    E( X )            1 + ir + y ir − ir R M R   1+ i + y i − i δL δM δL  ( ( ) . Fuzzy real options valuation for oil investments    f1 y M m2  µ x M =  m1 . k = 3 – i for positive F (Buckley 1987). The equations that demonstrate the right and left convergences towards the iδ and ir functions are given by Eqs. 26–27 are obtained. (21) (22) (23) (24) Fig. 15. Eqs.

ln    d 2 =  ln   e       1 + iδ     E( X )     E( X )  L           T    1 1  σ T  T T   1 1 σ T    1+ i  σ Tσ   T      2  ( 1 + ir  σ  2   d =  ln  ES )) σ σT T 1 + ir rR σ σ 2T   − σ T. −T −T    FROV = S 1 + f ( y i ) N d − X 1+ f ( y i ) N d .    (27)  1 + ir L   1 + iδ R  T σ  σ T     2      .Technological and Economic Development of Economy. e    (28) T    1 1      E ( S )  σ T  1 + ir  σ  σ T     E (S )  σ T    2  R    − σ T . ln     e   1 + iδ    E( X )    1 + iδ  R          R          FROV equation is adapted for fuzzy discrete interest rates as follows.ln        E( X )       1 + ir M   1 + iδ M    .   E( X )     E X     1 + iδ L  e  (  1 + i L   ln  E ( X ) ) −  1 + iδ   δM  )       E( X )  L     1 + iδ R                      T  1 σ T   T   1 σ T      E ( S )  σ T  1 + ir  σ  2       L    − σ T . 0  1 + ir M   1 + iδ M   1 + ir  (29) M  1 + iδ M          T σ σ T   2 e T σ σ T   2 e ( i δ ) ( ) 1 ( i r ) ( ) 2 (30) . 2009. 15(4): 646–669 657  1 + f k ( y ir )     1 + fi ( y iδ )    T σ σ T    2    e   1    E (S )  σ T  = ln       E( X )        1 + ir R   1 + iδ L           1 + ir + y ir − ir R M R   1+ i + y i − i δL δM δL  T σ σ T   2 e     ( ( )  )  T σ σ T    2    e   .ln   E ( S )  σ T   E ( S   1      ) σ T   e    R     1 M    − σ 2T .        T σ σ T   2 e     (26)   1    E (S )  σ T   d1 =  ln        E( X )       1   E (S )  σ T  ln       E( X )      1    E (S )  σ T   .      e     σ T .ln   E ( S )  σ T   + ir  σ e 2   − σ T .    σ T  1 + irL  σ  2    e    − σ T  . e    T σ σ T   2 e      1   E (S )  σ T   d 2 = ln       E( X )     1   E (S )  σ T ln       E( X )     1 + f k ( y ir )     1 + fi ( y iδ )     1 + ir + y ir − ir R M R   1+ i + y i − i δL δM δL  ( ( )  )    −σ T =    T    σ σ T    2   −σ T.

4.    E X x3 x2 x1   x4 x3 x2 x1   x4    If Eq. If ln   0   S0 1 2   E X        expression found in Eq. Eq. 4. ( σ T ) ) can be stated 1 2 3 T σ σ T    2    e 1     s2  σ T  w  − σ T . ( σ T ) . 2 . 3 . ln 4  . S (1 + i ) N ( d ) − X (1 + i ) −T −T 2 0 δM −T 0 δR 1 rL −T  N d2 . s . 5 is replaced by the fuzzy numbers S0 and X . 34 is reached: 1      s1  σ T d 2 =  ln    w    x4    1    s3  σ T ln    w   x2   T σ σ T    2    e ~ 1      s4  σ T w  . σ T = ( σ T and Eq. Eq.32 is placed in Eq. C. x2 .   4 (33) ( ) . Fuzzy real options valuation for oil investments ( ( ( )   N ( d ) − X (1 + i ) ) −T rM   N d1 − X 1 + irR −T 2  ( ) N ( d ) . ln 3 . α′. ln 2 . 4  = ln 1 . 33 is reached. 1      s1  σ T d1 =  ln    w    x4    1    s3  σ T ln    w   x2   T σ σ T    2    e 1     s2  σ T  w  . α.    T σ T  σ e 2     In d 2 = d1 − σ T expression.    N ( d ) . ( )) (31) Eq.   E  S0   s s s s   s s s s  ln     = ln  1 . ln   x    1     T σ T  σ e 2        . ln   x    1     T σ T  σ e 2          − σ T . β ) and X = ( x1 . Kahraman.    T σ T  σ e 2     1      s4  σ T w  − σ T . ln   x   3       −σ T .658 FROV = S0 1 + iδ L 1 İ.     (34) . Uçal. 6. ln   x    3    T σ σ T    2    e (32)   . Postponing the defuzzification of costs and revenues   E S     = ( s . β′ ) are trapezoidal fuzzy numbers. 32 is reached subsequently. ( σ T ) . 31 is developed for the situations that occur while dealing with fuzzy discrete parameters. It is helpful when there is incomplete information under fuzzy decision environment. S (1 + i ) .

1      s1  σ T N  ln    w    x4    T σ σ T   2 e     1     s2  σ T  w  .5.Technological and Economic Development of Economy. Below. ln   x    1    T σ σ T   2 e     T σ σ T    2    e    −   T σ σ T   2 e      X (1 + ir ) −T 1      s1  σ T N  ln    w    x4    T σ σ T   2 e     1     s2  σ T  w  − σ T . s4 N  ln   x    −   1            T σ σ T   2 e     T σ σ T   2 e     1        s2  σ T   w  − σ T  . s2 N  ln   x     3      1         T σ T   T σ T      s4  σ T  2    2        wσ e wσ e   . 4. (36)     .     T σ σ T   2 e     (35) FROV = (1 + iδ ) −T 1        s1  σ T w  s1 N  ln        x4     1      s3  σ T s3 N  ln       x2    1       s2  σ T  w   . x4 N  ln   x     1            − σ T   . By postponing the defuzzification. 10.4 were defuzzified by Eq. ln   x    3    T σ σ T   2 e       −σ T . 33 and 34 are developed for postponing the defuzzification to the very end stages of the model. ln   x    1        −σ T  . probability values could be obtained from the normal distribution table. Fuzzy probabilities After d1 and d2 values obtained in Section 4.   T σ σ T    2    e 1      s4  σ T w  .   (1 + ir ) −T 1      s1  σ T w  x1 N ln       x4    T σ σ T   2 e     σ T   2 e          − σ T . x2 N  ln   x    3        T σ σ T   2 e     T σ   . Eqs.   1    s3  σ T ln    w   x2   1      s4  σ T w  − σ T . operations could be carried out with fuzzy probabilities.     1      s3  σ T x3 N  ln    w    x2    1         s4  σ T  w  − σ T  . ln   x    3    T σ σ T   2 e      FROV = S0 (1 + iδ ) 1    s3  σ T ln    w   x2   −T   . 15(4): 646–669 659 Carlsson and Fuller (2003) made a defuzzification in the very early stages of their model. formulas have been developed to be used in the case of carrying on with fuzzy probabilities. 2009.

.33662  241188460   ln   +  0.s 5–7.198216686.66%) and the fuzzy real options value is calculated through their application to Eq.052 + 2  188777796    0.05 − 0.20% R 5. X = 188777796$.176197913) .100000000) . α + β = 176197913 + 176197913 = 352395826 .  S0 = (229129037.18% σS 33. S0 = 241188460$. Table 1 reports financial parameters used in ROV where δ is convenience yield. 2 6 ( ) s2 − s1 = 253247883 − 229129037 = 24118846 . Parameters used in Real Options Valuation δ 5.  σ S0 = ( ) ( s2 − s1 ) ( s2 − s1 )(α + β) (α + β) 2 + + 81184035 4 6 24 = = 33. Uçal. Kahraman. Fuzzy real options valuation for oil investments 5. Table 1.66% .3366 12 = −0. N ( d 2 ) = 0.  241188460 E S0 ( ) (179338906 + 198216686) 100000000 + 100000000  E X = + = 188777796$ .7725 . Application Grafström and Lundquist (2002) examined whether the value of an undeveloped oilfield differs depending on whether real options valuation or discounted cash flows analysis is used.7801. 6 2 6 ( ) d1 = 0.  12   = 0.7725 − 0. 229129037 + 253247883 176197913 − 176197913  E S0 = + = 241188460$ . and r is instantaneous risk-free interest rate.66% The current values of the revenues and costs are fuzzified in order to obtain the standart deviation (33. In that application the present value of the costs is found as X = 188777796$ and the present value of the revenues is found as S0 = 241188460$.3366 12 N ( d1 ) = 0.176197913.3935 .00% σδ 67.100000000. 253247883.  X = (179338906. C. d 2 = 0.3445 .660 İ.

92552920.7801 − (179338906.176197193) ⋅ e −0.176197913.66$ after being defuzzified through the centroid method. 71944517. 229129037 + 253247883 (229129037 − 52931124) − (429445796 − 253247883) .Technological and Economic Development of Economy.0534 .05 − 1 = 0. 1 + iδ Fuzzy revenues and costs are:  S = (59931124. 179338906 + 198216686 (179338906 − 79338906) − (298216686 − 198216686) .66$ .998   = 0.  E S0 = + 2 6 ( ) ( )  E S0 = 241188460$ . 253247883.0513 .100000000. w= 1 + ir = 0.198216686. ( ) ( ) When the values are placed in Eqs. 298216686) .998 . In the defuzzification process. iδ = eδ − 1 = e0. 0  X = (79338906. The fuzzy real options value of the investment calculated through Carlsson and Fuller approach is found as 71415953.052⋅12 ⋅ 0.179338906. 229129037.05⋅12 ⋅ 0. the fuzzy real options value is calculated as below: 1  12 0. The expected values of the revenues and costs are given below. the interest rates for the case of discrete compounding are obtained by applying the data of oil field investment example to Eq. 253247883. 58294298 2 71944517 164497437  ∫−34258622 z dz + ∫58294298 zdz + ∫71944517  1 −  58294298 71944517 164497437  ∫−34258622 zdz + ∫58294298 dz + ∫71944517  1 −  92552920  z − 71944517  z − 71944517  z* =  zdz  dz 92552920  = 71415953.   188777796     . the centroid method based on the centers of gravity of the possibility distribution is used. 2009. 92552920 ) $. 429445796) .s 11 and 12.052 − 1 = 0. 15(4): 646–669 661 FROV = ( 229129037.3366 e 2 d1 = ln   0. Below.3366 12    241188460  0.3445 = ( 58294298.3366 12  0. ir = e r − 1 = e0.7726 .100000000 ) ⋅ e−0.  E X = + 2 6  E X = 188777796$ .198216686. 17 and 18.

f1 ( y  ) = iδ L + y iδM − iδ L = 0.164359270 ) .0487 ) = 0. iδ f 2 ( y  ) = iδ R iδ δM ( + y (i − iδ R ) ) = 0. 229129037.0534−12 ⋅ 0.0487.198216686. ir ir When approached from the left: for y = 0 f1 ( y  ) = 0.78012 .  The fuzzy number d1 is calculated below by the help of Eq. ir f 2 ( y  ) = irR ir M R ( + y ( ir ) − ir ) = 0.0027 y . 429445796 ) ⋅1. C.0561) = 0.0026 y . Fuzzy real options valuation for oil investments d 2 = 0. Kahraman. for y = 1 f1 ( y  ) = 0.s 23 and 24. .0561 + y ( 0.0487 + 0. f1 ( y  ) = irL + y irM − irL = 0.179338906.0513.0507 .0534 .0507 + y ( 0. FROV = ( 52931124.662 İ. ir = ( 0. 0.0534 − 0.0534 − 0.34701 = ( −34660540. 0.0507.0539 ) . 253247883.0513.0561 − 0.s 21 and 22.0539 − 0. iδ iδ When approached from the left: for y = 0 f1 ( y  ) = 0. the below membership functions for iδ are obtained.0513−12 ⋅ 0.7726 − 0. 58012568.0487.78012 − 7 ( 79338906.0561) .0487 + y ( 0.  ) = 0.164359270 ) .34701 . N ( d 2 ) = 0. When the data are applied to Eq.0539 ) = 0. 0. When the data are applied to Eq. iδ = ( 0. the below membership functions for ir are obtained.0027 y . iδ and ir are fuzzified at the rate of ±%5 to calculate the real options value of the investment in the condition of fuzzifying discrete interest rates. 27.0539 and for y = 1 f 2 ( y  ) = 0. 71686162. 58012568.0513 − 0. 298216686 ) ⋅1.3366 12 = −0.0539 + y ( 0. 71686162. and ir ir when approached from the right: for y = 0 f 2 ( y  ) = 0. Uçal.0534 and when approached from right: for y = 0 f 2 ( y iδ iδ values are obtained.0534. The fuzzy real options value of the investment is defuzzified through the centroid method and found to be 72971941$ . FROV = ( −34660540. 0.0513 values are obtained.0507 + 0.0513 − 0.0026 y . for y = 1 f1 ( y  ) = 0.0561 and for y = 1 f 2 ( y  ) = 0.0507 ) = 0.3934 . N ( d1 ) = 0.

30 ⋅ 0.50 ⋅ 0.824428 = 0.0513  0.   12         .772595. 62125072. 75888533.757048 − 0.408968 .824428 ) is defuzzified and d2 is calculated. 0.3366  e  1 + 0. −T 1 rL  N d2 . FROV1 is found by defuzzying ( −31531751. 429445796 ) (1 + 0.30 + 0.198216686.0539 )−12 ⋅ 0. 62125072. 298216686 ) (1 + 0.30.179338906. Using the defuzzification method suggested for the fuzzy numbers which have normal  distribution. 0.720767 + 0.0487  0.3366  e  1 + 0. 253247883. d1 = ( 0.20 is accepted.   12   1 + 0.3366 d1 =  ln       188777796    1    241188460  0.824428 ) .20 d 2 = 0. 0.772595 + 0.    d1 = ( 0. The fuzzy real options value is calculated with the help of Eq. S (1 + i ) ) −T −T 2 0 δR   N d1 − X 1 + irR −T ( )   N ( d ) − X (1 + i ) ) −T  N d2 .720767.3366  e  1 + 0.0539  0. FROV2 .3366 12    2    .0561    12 12  0. 75888533.169545362 ) through the centroid method and found to be $ 63082694. FROV = ( FROV1 .20 ⋅ 0. . ( )) FROV1 = ( 52931124. 0. FROV3 ) = S0 1 + iδ L S0 1 + iδM ( ( ( ) −T   N d1 − X 1 + irM ( ) ( ( ) (  ) N ( d ) . d1 = 0.3366 2   .0507     0.720767. τ13 = 0.772595.3366 12    2  12   1 + 0.3366 12 = −0. 31. 15(4): 646–669 663 1      241188460  0.3366 ln      188777796   1    241188460  0.341281 = ( −31531751. τ12 = 0.50 + 0.341281 .0507 ) −12 ⋅ 0.Technological and Economic Development of Economy.169545362 ) .50. 2009. N (d 2 ) = 0. If τ11 = 0.0534    12  0.775489 . 229129037.775489 − ( 79338906. The areas under the standard normal distribution curve are given below for d1 and d2: N (d1 ) = 0.3366 ln      188777796   12   1 + 0. 0.757048 .

341281 = ( −36199402. d1 =  ln   e   1. 70256880. 429445796 ) (1 + 0. C.3366 12 .775489 − ( 79338906. FROV2 is found by defuzzying ( −33850206.0513  0.157687761) . 0. 67422616.889470) .0534  .3366 12           52931124  0.3366  2  .3366  2  ln   e   1. FROV3 = ( 52931124.3366 12 ) .737106 − 0. .565438.3366 12 . FROV is found by defuzzying ( 63082694.1. 229129037. 70256880. 0.341281 = ( −33850206. 71617684. 229129037.163532227 ) .0534   198216686        1 12  0.3366 12         229129037  σ T  1.0534   179338906         1 12  0. 1 12  0.041530. 84940120 ) $ .3366 12 .775489 − ( 79338906. ln   e   1.3366 12           429445796  σ T  1. FROV3 is found by defuzzying ( −36199402.428911.3366 12  1. d 2 = (−1.163532227 ) through the centroid method and found to be $ 70256880.198216686.3366  2  .723453) .0513  0.0513)−12 ⋅ 0. 58064338. 0. 54065743. 54065743.3366  2  . The data in the example of oil field investment are applied to Eq.157687761) through the centroid method and found to be $ 84940120.179338906. ln   e   1. 253247883.207547. FROV = ( 63082694.889470 − 0.664 İ.0534 ) ⋅ 0. 298216686 ) (1 + 0. 71617684.3366 12         253247883  σ T  1.1.3366 12 .0561) −12 ⋅ 0. Kahraman.737106.  79338906        d1 = (−1. d 2 = (−2.737106 − 0. 253247883. 0.0487 )−12 ⋅ 0. Fuzzy real options valuation for oil investments −12 FROV2 = ( 52931124. −0. 429445796 ) (1 + 0.0534    298216686          1 12  0.565438 − 0. 0.s 33 and 34.889470 − 0.3366 12 . Uçal. 84940120 ) $ through the centroid method and found to be $ 66734120.179338906.565438 − 0. −0.598500. 0. 58064338. 0. 366 12 . 67422616.1.041530 − 0. and d1 and d2 values are found by postponing the defuzzification.198216686.0513  0.0513  0. 298216686 ) (1 + 0.

723453)) .17 . the fuzzy real options value in the case of postponing the defuzzification of costs and revenues with defuzzifying the probabilities is calculated as below: −12 −12   FROV = S (1.198216686.39.737106 + 0. d1 = d2 = 0.39 ⋅ 0.0534 ) 0. 51323816. τ14 = 0. 429445796 ) ⋅ 0.598500. 229129037.7173 .41 ⋅ 0.7173 − X (1.598500).5748 .04153).041530) + 0.0534 ) −12 ⋅ (52931124 ⋅ N (−1.35 + 0. τ12 = 0. 0.0513)−12 (79338906 ⋅ N (−2. the fuzzy number d1 = (−1.41.179338906. 14.39 + 0. 0.565438 + 0.889470) is defuzzyfied. Using Eq.08 .09 ⋅1.08 ⋅ (−2. τ24 = 0.207547.1.17 ⋅ 0.17 + 0. 0. 15(4): 646–669 665 In the model suggested for the defuzzification of the membership functions showing normal distribution: using the weights τ11 = 0.723453 = −0.207547) + 0.7371). 2009.3319 .428911) + 0.40 . 229129037 ⋅ N (0. 222672378) through the centroid method and found to be $ 105867783. FROV = (1. the fuzzy real options value is calculated without defuzzifying the possibilities.565438.3319 . 4 FROV = (−120991904.40 N ( d1 ) = 0.41 + 0. 77347074. 298216686 ) ⋅ 0.179338906⋅ ≤ N (−0. 198216686 ⋅ N (−0. 64647237. the fuzzy number d 2 = (−2. τ23 = 0. 0. −0. 51323816. τ22 = 0.40 ⋅ (−0. τ13 = 0.723453) is defuzzified. 222672378) .11 + 0.08 + 0.150555219 ) $.5654). 51942806.4346 .35.11.041530.428911. .0513) 0.889470 = 0. 253247883 ⋅ N (0. FROV is found by defuzzfying (−120991904.Technological and Economic Development of Economy. 36.598500) + 0.35 ⋅ (−0.737106.09 . When the values are placed in Eq. The real options value of the investment is found by defuzzifying the calculated value through the centroid method and found to be $ 77408863.428911). 253247883.11 ⋅ (−1. 429445796 ⋅ N (1.09 0. −0. 0 FROV = ( 52931124.1821 = ( −33965177.3842 − ( 79338906. N ( d 2 ) = 0. . 77347074. 298216686 ⋅ N (0. 0. and using the weights τ21 = 0.8895) − (1.207547).

a method that does not consider the postponing option could miss the investment opportunity. the result of the case of postponing the defuzzification of costs and revenues without defuzzifing the probabilites is the fuzzy number which has the widest range. 64647237. 222672278)$ Defuzzified Result 72971941$ 66834120$ 77408863$ 105867783$ Table 2 represents the application results of the suggested model together. Fuzzy real options valuation for oil investments Table 2. Early defuzzification of the fuzzy parameters causes information loss. Consequently. an investment. fuzzy logic is more expressive than classical mathematics. The last achieved value is the most sensitive value which means that the case of postponing the defuzzification of costs and revenues without defuzzifying the probabilites shows more information about the investment. therefore. The results of the application with suggested model Fuzzy Result Discrete Compounding Fuzzifying the discrete interest rates Postponing the defuzzification of costs and revenues with defuzzifying the probabilities Postponing the defuzzification of costs and revenues without defuzzifying the probabilities (–34660540. Kahraman. When examined by real options. 51323816. Thus. the parameters used in the real options valuation method will lead to more realistic results which concern human reasoning. 77347074. however. 70256880. that is rejected because its current value is found to be negative by using the discounted cash flow analysis. on consideration of a fuzzy manner. the results obtained can be more trusted. 164359270)$ (63082694. 51942806. a new model has been suggested. Uçal. which is vastly different from the traditional investment valuation methods. Using real options valuation methods to analyse an investment decision reduces the uncertainty to minimum and it ensures that the investment assessment is made in as the most realistic way as possible.666 İ. may deliver positive values in case of postponing the investment and. 150555219)$ (–120991904. it is observed that the fuzzy revenue and expenditure values were defuzzified at a relatively early stage. The model suggested by Carlsson and Fuller (2003) has been found to be the most frequently used method amongst the fuzzy real options assessment methods during literature researches. Therefore. it . 6. Conclusion Real options valuation method. 84940120)$ (–33965177. In this study. C. Due to the fact that estimations made by people have multilateral structures that do not have sharp certainties. 58012568. By using the real options valuation method particularly to analyse the risky investments. 71686162. makes more exact assessments since it considers future uncertainties as well as dependencies and dynamism. wrong decisions could be easily avoided. In this model.

06. doi:10. Allenotor. The Energy Journal 28(3): 1–6.035. 2008. Scholes. Fuzzy Sets and Systems 139: 297–312.. Wright. Pesaran.03. L. the difference between the applications of this new model and of the early defuzzifying model has been found to be the information loss due to the early defuzzification. A. in Proceedings of the Second International Symposium of Hungarian Researchers on Computational Intelligence.10..enpol. M.eneco. 2006. P. 2004.. in Soft Computing with Industrial Applications – Proceedings of the Sixth Biannual World Automation Congress. 2003. 1987. A. Fuzzy Sets and Systems 21: 257–273..2004. F. Energy economics: a place for energy poverty in the Agenda?. R. J.Technological and Economic Development of Economy. The Journal of Finance 27(2): 399–417. Brach... Bergmann. Carlsson and Fuller’s model has been re-arranged for discrete compounding and then the defuzzification of revenue and expenditures has been postponed and relavant equations have been formed for the cases of early defuzzifying probabilities and defuzzifying them at the last stage.08. Carlsson.1016/j. P. Thesis in Financial Economics. T. C. doi:10. Oil investment in the North sea.. Fuller. Journal of Political Economy 81(3): 637–654.1016/j. 2006..002.002.1007/978-3-540-74742-0_56. The fuzzy mathematics of finance.enpol. J. C. M. USA. Chamorro. M. Black. . C. Grafström.1086/260062. Real options for risk management in petroleum development investments. Review of Financial Economics 18(1): 10–22. Scholes. 2002.1016/S0165-0114(02)00591-2. C.. Budapest. F. Energy Policy 35(11): 5754–5765. 2001. Fullér.024. Hanley. John Wiley&Sons. 2009.. 2007. The valuation of option contracts and a test of market efficiency. A. D. L. Chorn. Thulasiram. Valuing the attributes of renewable energy investments. Energy Economics 28(4): 489–505.1016/0165-0114(87)90128-X. A.. Oxford Institute for Energy Studies EE15: 24. Apostol. A grid resources valuation model using fuzzy real option.. N. binomial. H. Dinca. 4742: 622–632. Parallel and Distributed Processing and Applications. Buckley. References Abadie. Badea. M. C. and Lundquist.2006. Majlender. Garcia. Real Option Valuation vs. doi:10. Stockholm University. and trinomial valuation methods.1016/j.2008.. Carlsson. Energy Policy 34(9): 1004–1014. doi:10. A. doi:10. Rousseaux. R. For further research. L. it is suggested that the information loss caused by uncertainty should be measured by fuzzifying the other real options valuation methods that have a broad area of usage such as game theory. 2007. 2003. A multi-criteria approach to evaluate the natural gas energy systems. Monte Carlo valuation of natural gas investments. Black. doi:10. The pricing of options and corporate liabilities. Birol. A fuzzy approach to real option valuation. 81–88.2307/2978484. Spain: 281–287. M. F. Project selection with fuzzy real options. Shokhor. 2007. doi:10. 15(4): 646–669 667 postpones the defuzzification of fuzzy parameters in the real options valuation in order to avoid this information loss. J. Favero. Real Options in Practice. F.1016/j. Fuzzy real option valuation in a power station reengineering project. Springer. G. On the other hand. 1972. A. R. 1973. S. Lecture Notes in Computer Science. M. Vol.rfe. DCF Valuation: An application to a North Sea Oilfield.2007. doi:10. R. doi:10. 1991.. K.

L.. Bass. A. Fuzzy options with application to default risk analysis for municipal bonds in China. 2008. in Fuzzy Engineering Economics with Applications (Ed. A. C. C.. Sardinha. Real Options Managerial Flexibility and Strategic in Resource Allocation. L.668 İ. ICCS 2007. Worldwide Trends in Energy Use and Efficiency.. Smith. Kahraman. J. Real Option Analysis Tools and Techniques for Valuing Strategic Investments and Decisions. Renewable and Sustainable Energy Reviews 12(9): 2422–2437. L. R.1016/j.eneco. doi:10.. 2008.1016/j. doi:10. doi:10. Mohn.1016/j.038. doi:10. International Journal of Intelligent Systems 23(11): 1153–1176. P. J.06. 2007. 2007. Jinlong.. Tolga. M. Springer..2005. doi:10. B. Paris. Salles. in Proceedings of the 8th International FLINS Conference. P.. 1992. Energy Policy 37(3): 868–878. Lecture Notes in Computer Science. T. Z.enpol. doi:10.. Nonlinear Analysis 63(5–7): e2353–e2365. Liou T.1007/9783-540-72588-6_152. Energy Economics 30(4): 1909–1932. The Quarterly Journal of Economics 103(3): 479–508. Rodríguez. S. 1027–1032.10. K.2008. Option valuation of claims on real assets: the case of offshore petroleum leases.09. Energy Policy 34(9): 1075–1085. H. X. Ç. Majlender. Vol. Valuation for renewable energy: a comparative review.019. International Energy Agency (IEA).07. Tao.1016/j. Ucal. 2002. Y. Zheng. doi:10. Siegel..eneco. L. A. A. 2005. 307–338. C. Y.na. Lin. 1988. doi: 10. Kahraman). S. 3: 5–22. Kjærland.enpol.003. 2007. and Misund. Shan. Garcia. P. Paris. J. The MIT Press Cambridge. 2006. World Energy Outlook 2007: China and India insights. Mun. G. Fuzzy multiattribute evaluation of R&D projects using a real options valuation model.enpol. Mc-Graw Hill. Wang. İ. Teixeira.. 2008. Michaels. 1996. C.1002/int. D.2004.1016/j. Benhai. J. Studies in Fuzziness and Soft Computing 233. C. B. J. 1997. Leslie. Q. K. G. Ross. Springer Verlag. doi:10. The Russian oil industry between public and private governance: obstacles to international oil companies’ investment strategies. International Energy Agency. doi: 10.. Uçal.007. Energy Economics 31(2): 240–248.001.1007/978-3-540-70810-0. 2009. L. Kahraman.2006.2008. .2007. John Wiley & Sons. 1995. A.1016/j. doi:10. R.. 2007.03.1016/j. F.1016/0165-0114(92)90223-Q.008. doi: 10. USA.ememar. 2008. Fuzzy Logic with Engineering Applications.021.. Fuzzy real option analysis for IT investment in nuclear power station. C..10. C. International Energy Agency (IEA). M..2007. Fuzzy Sets and Systems 50: 247–255. Ranking fuzzy numbers with integral value. Locatelli. London. F. K.rser. Emerging Markets Review 8(1): 67–79. J. The real power of real options.006. Menegaki. Kahraman. USA.02. Real option valuation of free destination in long-term liquefied natural gas supplies. 2009. 4489: 953–959. doi:10.2007. Soft decision support systems for evaluating real and financial investments..20312. Trigeorgis. J.. Fuzzy real options valuation for oil investments Han. L. Energy Policy 35(11): 5901–5908. A real option analysis of investments in hydropower – the case of Norway. IFTEM: An interval-fuzzy two-stage stochastic optimization model for regional energy systems planning under uncertainty. Investment and uncertainty in the international oil and gas industry. 2008. The McKinsey Quart. 2008. J. Qin. Rocha. Real estate and real options – a case study.1016/j. Fuzzy real options valuation for oil investments. Paddock. Huang. International Energy Agency.2307/1885541.09.

US department of energy – energy information administration. Kahraman Santrauka Tradiciniai vertinimo metodai yra mažiau patikimi esant neapibrėžtumams. kurie gali minimizuoti neapibrėžtumus. quality management and control. Istanbul Technical University.Technological and Economic Development of Economy. L. 1965. 15(4): 646–669 669 US. Research interests: engineering economics. Istanbul Technical University. kuris realistiškiau įvertins investicijas. Research interests: engineering economics. H. S. Department of Industrial Engineering. 2007. rodiklius apibūdinančią neapibrėžtą informaciją apdorojant ankstyvojoje stadijoje. 2008.. tampa svarbesni. 9–11 June. neapibrėžtos pasirinktys. International energy outlook. REALIŲ ĮVERČIŲ METODO TAIKYMAS INVESTICIJOMS Į NAFTOS VERSLĄ VERTINTI İ. Renewable and Sustainable Energy Reviews 12(9): 2422–2437. M. fuzzy sets applications to decision making. Pasiūlytas naujas realių pasirinkčių vertinimo modelis. Modeling investment risks and uncertainties with Real Options approach. Pasiūlytas modelis buvo pritaikytas investicijoms į naftos verslą modeliuoti. realių pasirinkčių vertinimas. Yang. Department of Industrial Engineering. Cengiz KAHRAMAN. Istanbul Technical University. M. Li. W.. Author of about 70 scientific articles. Prof. multiple criteria decision making and energy resources. IEA Working Paper.DoE. Professor (2003). Zadeh. MSc. valuation for renewable energy: A comparative review. Master of Science in engineering management. in Service Systems and Service Management. International Conference on. Zhang. Uçal. Master of Science (1990). Istanbul Technical University (2004). First degree in Industrial Engineering Istanbul Technical University (1988). C. Istanbul Technical University (2008). First degree in environmental engineering. .. Zeng. multiple criteria decision making. nustatytas informacijos nuostolis. Šiame straipsnyje nagrinėjamas hibridinis Carlsson ir Fuller metodas. kiti metodai. Research Assistant. Vadinasi. 2007. kuris buvo panaudotas diskrečiajam rizikingų pinigų srautų modeliavimui. Fuzzy sets. Blyth. Doctor (1996). investavimas.. B. PhD in industrial engineering (continued). Research and application of power network investment decision-making model based on fuzzy real options.. kuris atsiranda dėl ankstyvo neapibrėžtų duomenų apdorojimo. 1–5. Inform and Control 8: 338–353. Assoc. 2009. (1997). 27. statistical decision making. Reikšminiai žodžiai: neapibrėžtos aibės. Wang. Prof. tokie kaip realių pasirinkčių vertinimo modeliai. Huang. İrem UÇAL. T.

download. . However. users may print.Copyright of Technological & Economic Development of Economy is the property of Technological & Economic Development of Economy and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. or email articles for individual use.

Sign up to vote on this title
UsefulNot useful