Flexible Design of Airport System Using Real Options Analysis - MIT

1.231 Planning and Design of Airport System 

Term Project 

Flexible Design of Airport System 

Using Real Options Analysis 

Case Study of New Runway Extension Project 

of Tokyo International Airport 

Dai Ohama 

Technology and Policy Program 

Engineering Systems Division 

Massachusetts Institute of Technology 

December 14, 2007

Table of Contents 

1. Introduction .................................................................................................................1 

2. Outline of the Case Project.........................................................................................1 

2.1 Tokyo International Airport....................................................................................1 

2.2 Runway Extension Project......................................................................................2 

2.3 Current Design is Optimal? ....................................................................................3 

2.4 Project Capacity of Airport.....................................................................................5 

2.5 Flexible Design.......................................................................................................5 

3. Real Options Analysis .................................................................................................6 

3.1 Financial Options Theory .......................................................................................6 

3.1.1 Black-Sholes Option Pricing Model ................................................................7 

3.1.2 Binomial Lattice Model ...................................................................................8 

3.1.3 Monte Calro Simulation...................................................................................9 

3.2 Real Options Analysis ............................................................................................9 

3.2.1 Real Options.....................................................................................................9 

3.2.2 Types of Real Options ...................................................................................10 

3.3 Application of Real Options Analysis to Case Study...........................................10 

3.3.1 Which Method should be used for the Case Study? ......................................10 

3.3.2 Real Options Analysis Using Monte Calro Simulation .................................11 

4. Analysis of Case .........................................................................................................11 

4.1 Analysis Condition ...............................................................................................11 

4.1.1 Capital Investment .........................................................................................12 

4.1.2 Revenues and Costs .......................................................................................13 

4.1.3 Capacity of Airport ........................................................................................14 

4.1.4 Discount Rate.................................................................................................14 

4.2 Uncertainty in System...........................................................................................14 

4.3 Demand Forecasting .............................................................................................15 

4.4 Summary of Analysis Condition ..........................................................................17 

4.5 Result of Analysis.................................................................................................18 

5. Conclusion..................................................................................................................21 

6. References ..................................................................................................................22

12/14/2007 

1.231 Planning and Design of Airport System Dai Ohama 

1. Introduction 

The Airport systems not only in the U.S. but also all over the world are now very 

complex. However, all of them are not always designed optimally. Some facilities are 

built excessively compared to their estimation during planning phases, others are required 

improvement since they are too large for the actual situations. [1] For example, the size of 

the New Denver Airport is too big, and as a matter of fact there is an unnecessary passenger 

building. The Newark airport is unsuited for the actual transfer and international traffic, 

and major changes should be required. The reason for them is that the master plan of these 

airport systems did not anticipate future risks and uncertainties of possible changes in market 

conditions, and did not consider the countermeasure of those real risks. Eventually, it leads 

to losses or extra costs, and even losses of opportunities happen. Thus, the master plan is 

often inflexible and inherently cannot respond to the risks. In order to respond to this 

situation, it is essential to plan strategically by considering future risks and uncertainties and 

by applying flexibility into design. 

This project is to consider one of the possible effective ways to incorporate flexibility 

into design by using real options analysis. The goal of this project is to demonstrate how 

the flexible design is conducted and how it works. The case of “Tokyo International 

Airport New Runway Extension Project” is applied as a case study. 

2. Outline of the Case Project 

2.1 Tokyo International Airport 

Tokyo International Airport or Haneda Airport (HND) is located in the bay area, near 

the center of Tokyo. Haneda Airport is the busiest and an important hub airport in Japan 

for domestic air traffic.[2] Haneda Airport is consistently ranked among the world's busiest 

passenger airports in terms of the number of passengers, and its ranking was fourth in 2006. 

Currently it has three runways (Runway A: 3,500m, Runway B: 2,500m, Runway C: 

3,000m), serving nearly 60 million passengers every year. The total capacity of Haneda 

Airport is 296,000 aircraft (a/c) per year. 

Term Project: Flexible Design of Airport System Page 1 of 22

12/14/2007 


Japan 

Tokyo 

Chiba 

Kyoto 

Osaka 

Tokyo 

Yokohama 

Tokyo Bay 

Haneda Airport 

Figure 2-1 Location 

Figure 2-2 Tokyo Int’l Airport 

Source: Ministry of Land, Infrastructure and Transport in Japan, Kanto 

Regional Development Bureau [2] 

2.2 Runway Extension Project 

However, its capacity has already reached a limit of airport capacity against the 

increasing demand, and it is necessary to respond to the demand as soon as possible. In 

order to solve this problem, Tokyo International Airport Extension Project was launched in 

2002 to build 4 th runway, which is called Runway D, to increase the total capacity of the 

airport. [2] This extension enables the airport to have the capacity from of 296,000 a/c per 

year to 407,000 a/c per year. 

Runway B 

Runway C 

Runway A 

Extension Project 

“Runway D” 

(Under Construction) 

Figure 2-3 Plan of New Runway Island in Haneda Airport 

Source: Ministry of Land, Infrastructure and Transport in Japan, Kanto Regional Development Bureau [2] 


12/14/2007 


The new runway will be constructed on the artificial island approximately 3,000 meters 

long and 500 meters wide, on the south side of the airport. As the new runway island is 

planned in the estuarine waters of the Tama River, the course of the river shall not be 

affected by reclaimed island. Therefore, pile elevated platform is applied for the part in the 

river mouth, and reclamation is applied for the part outside the alignments of the river mouth. 

Two taxiway bridges for aircraft connecting the existing airport to the new runway island, 

600 meters long and 60 meters wide, are planned as well. Pile elevated platform of the new 

runway island is 1,100 meters long and 500 meters wide in total. It is a jacket structure 

composed of approx. 200 steel jacket units 60 meters long and 45 meters wide. The water 

depth at the construction site is 15 to 19 meters, and the river has 30 to 40 meters thick 

sedimentary layers of soft clayey soil. Thus, long steel pipe piles are used as bearing piles 

for the pile elevated platform. 

Figure 2-4 Structure of Runway D 


By the end of the fiscal year of 2006, the design has been completed, the construction 

started in March 2007, and it is supposed to be in service in December 2010. 

2.3 Current Design is Optimal? 

As discussed above, the new runway is expected not only to improve the airport 

capacity but also to bring about a major impact on the overall economy. However, is this 


12/14/2007 


new project really optimal? Can the design of this project respond to the future 

uncertainty? 

The current design of Runway D consists of the Runway (R/W) and the Parallel 

Taxiway (PT/W). At the Runway D, airplanes depart from south to north and land from 

north to south, and the projected capacity is based on this operation. A parallel taxiway is 

usually necessary for a runway since airplanes have to move to a runway in order to take off 

and to go to terminal after landing. But at this Runway D, a parallel taxiway will be used 

only when those airplanes which are running for takeoff stop and return to the terminal. 

This possibility is very low. Because there is turning pad at the end of the runway, it is 

possible that those airplanes which are running for takeoff return to the terminal by using 

this area. Therefore, the operation frequency of the PT/W is extremely low. In addition, 

the area at the south edge of the runway (S/E), is also not used since airplanes do not depart 

and land this way. Therefore, PT/W & S/E are not necessary for the currently planned 

operation, and it can be said that this runway island is not designed optimally. This project 

was based on taxpayers’ money (¥570 billion) and it is excessively invested, so it is 

indispensable to evaluate if the PT/W and S/E is worthwhile investing or not. 

North 

Departure 

Runway 

(R/W) 

South 

Arrival 

Parallel Taxiway 

(PT/W) 

South Edge 

(S/E) 

In the currently designed 

operation, these areas 

are not necessary. 

Existing 

Airport 

Figure 2-5 Operation Concepts of Runway D 


12/14/2007 


These areas such as PT/W and S/E would be used if airplanes depart from north to south 

and land from south to north. Thus, if they need this operation in the future, they should be 

constructed, otherwise they should not be constructed. 

2.4 Projected Capacity of Airport 

So when are such departure and landing necessary? In the current plan, when north 

wind blows, runway B and D are used for departure and runway A and C are used for 

landing. When the south wind blows, runway A and C are used for departure and runway 

B and D are used for landing. In this planning, during south wind, only 12 a/c land at the 

runway D per hour, while during north wind 28 a/c depart from the runway D per hour. 

This is because runway usages are decided so that the entire capacity of the airport could be 

maximized. However, at the runway D, if airplanes depart from N to S and land from S to 

N, it is possible for the entire capacity of the airport to be increased. 

North Wind 

Departure: B-12 ac/hr 

D-28 

Landing : A-28 ac/hr 

C-12 

Total : 40 ac/hr 

South Wind 

Departure: A-22 ac/hr 

C-18 

Landing : B-28 ac/hr 

D-12 

Total : 40 ac/hr 

Figure 2-6 Currently Designed Capacity 


2.5 Flexible Design 

As a matter of fact, estimating this capacity is very complicated because all of four 


12/14/2007 


runways’ capacities are related complexly to one another. So, in this project, estimating 

this capacity is outside of the scope of the goal. I assume that the opposite departure (from 

north to south) and landing (from south to north) could increase the total capacity of the 

airport. In other words, PT/W and S/E could increase the capacity. If the demand is more 

than the capacity, PT/W and S/E should be constructed. If the demand is less than the 

capacity, it’s not necessary to construct them. Therefore, PT/W and R/E should be 

constructed when they become necessary. In other words, they don’t have to be 

constructed until the demand of passengers exceeds the projected capacity. If such 

flexibility is incorporated into design, managers are able to respond to future risks and 

uncertainties such as demand of passengers. Therefore, flexible design can minimize the 

initial investment, reduce risks, respond to uncertainties, and maximize the value of the 

project. By doing so, this runway extension project would be optimized. 

3. Real Options Analysis 

In order to evaluate those projects, one of the best possible ways is real options analysis, 

which is an evaluation method for those projects which involve decision opportunities. 

Real options analysis is the application of financial options theory to the actual projects such 

as infrastructure developments, real asset investments, research and development, and so on. 

In these projects, those who make decision generally have options such as the option to defer, 

the option to defer, and the option to abandon. In this case study, the option to expand 

should be applied. 

3.1 Financial Options Theory 

Financial options theory is a basis for the real options analysis. In finance, an option 

contract is an agreement where the owner has the right, but not the obligation, to buy or sell 

an “underlying asset,” such as a stock, at a pre-determined price on or before the expiration 

date. [3] 

There are two types of options in terms of the right. A call option refers to the right to 

buy a stock at the strike price, the pre-determined price. A put option refers to the right to 

sell a stock at the strike price. [3] Options are also divided into two categories in terms of 


12/14/2007 


expiration date. An American option is exercised when the owners are allowed to exercise 

them on or before the expiration date. A European option is exercised only on the 

expiration date. [3] 

Option is not the obligation but the right to buy or sell the underlying asset. So the 

owner would exercise the option only when it is favorable to do so. [3] For example, those 

who hold option would exercise a call option only if the price of the underlying asset (S), is 

higher than the strike price (K), and would earn a profit of S-K; otherwise, they would buy 

the asset not from the option, but from the market. A European put option is the right to 

“sell” the underlying asset at a certain price. So it can be exercised only when the market 

price of the asset (S) is less than the strike price (K), where those who hold the option can 

make the profit of K-S. Figure 3-1 shows the general payoff of a European call option and 

put option. Although American options can be exercised at any time on or before the 

expiration date, they have the same payoffs as European options. 

Payoff ($) 

Underlying Assets 

S 

Payoff ($) 

Underlying Assets 

S-K 

K-S 

Option 

0 

S=K 

Asset Price ($) 

Payoff of European call = Max (S-K, 0) 

0 

S=K 

Asset Price ($) 

Payoff of European put = Max (K-S, 0) 

Figure 3-1 Payoff Diagram for a European Call Option (Left) and Put Option (Right) 

Source: R de Neufville Lecture notes of the MIT course of ESD.71 (Fall 2006)[4] 

3.1.1 Black-Scholes Option Pricing Model 

The Black-Scholes option pricing model is the most well-known method. It applies to 

those European call and put options that do not provide dividends. It can produce the 

theoretical value of the option by applying five factors such as the price of the underlying 

asset (S), the strike price (K), the time until expiration (T), the risk-free rate of interest (rf), 

and the volatility (i.e., standard deviation) of returns on the stock. [3] The following 

formula shows this model for a European call option. [3] 


12/14/2007 


C = S ⋅ N 

−r T 

( d ) − K ⋅ e 

f 

⋅ N( ) 

1 

d 2 

Where C: Theoretical value of a European call option with no dividends 

d 

d 

1 

2 

ln 

= 

2 

( S / K ) + ( r / 2) 

= d 1 

− σ 

T 

σ 

f 

+ σ ⋅T 

T 

N(x): Cumulative probability function for a standardized normal distribution 

The following formula shows the theoretical value of European put options. 

P = K ⋅ e 

−r f T 

⋅ N 

( − d ) − S ⋅ N( − ) 

2 

d 1 

Where P: Theoretical value of a European put option with no dividends 

3.1.2 Binomial Lattice Model 

Binomial Lattice Model, which is also widely used, is a more simplified discrete-time 

approach to valuation of options compared to the Black-Scholes Option Pricing Model. [4] 

It assumes that the price of the underlying asset will change to one of the only two possible 

values during the next period of time. Figure 3-2 shows a three-stage binomial example. 

Figure 3-2 Binomial Tree Representations of Changes in Underlying Asset 

Source: R de Neufville Lecture notes of the MIT course of ESD.71[4] 

In the Binomial Lattice Model, the price of an underlying asset at a certain time can 

change to one of the only two possibilities, which are an upward movement with multiplier u 

with the probability of p, and a downward movement with multiplier d with the probability 

of 1-q. [4] 


12/14/2007 


u = e 

d = e 

σ ΔT 

−σ 

ΔT 

p = 0.5 + 0.5 

= 1/ u 

( v / d ) ΔT 

3.1.3 Monte Calro Simulation 

Monte Calro Simulation is an analytical method that generates the stochastic 

distribution of possible outcome that correspond to probability-distributed sampled inputs. 

[5] Because of the development of computer technology, large computer simulation such 

as Monte Calro Simulation can be constructed easily. Spread sheet software such as 

Microsoft Excel is the tool for conducting Monte Calro Simulation. 

3.2 Real Options Analysis 

In the previous section, financial options theory including Black-Scholes option pricing 

model, Binomial Lattice Model, and Monte Calro Simulation were introduced. In this 

section, real options analysis, which was the application of financial options theory to the 

actual projects, is introduced. 

3.2.1 Real Options 

Real option is the option to undertake some business decision under high uncertainty. 

In real options analysis, they are associated with flexibilities in designing systems or the 

evolution of projects. [4] Option - like flexibilities are included in systems and projects, 

represent opportunities to increase the value of the project through design or through 

management actions. [4] It is not obligation, and only when future asset price is preferable 

for you, you can exercise it, otherwise you don’t have to do it. This allows 

decision-makers to avoid downside losses as well as to obtain upside opportunities. These 

flexibilities are called “real” options. [4] Managers have applied real options analysis to 

business strategies including large-scale engineering systems. Although real options are 

very similar to financial options, they are different from financial options in terms of what 

should be assessed and the time span. [4] In financial options, the actual price such as 


12/14/2007 


stock price is evaluated, while in real options, the value of the project itself is evaluated. 

Financial options are usually very short span such as two years, while the time span of real 

options is very long-term. 

3.2.2 Types of Real Options 

There are several types of real options, such as deferral options, abandonment options, 

expansion options, growth options, and compound options. [6] Applying appropriate type 

of options enables managers to actively manage risks and uncertainties. For example, in 

this case study, expansion options is appropriate to be applied. In the current design, which 

is inflexible design, the whole facilities such as a runway, a parallel taxiway, and a south 

edge would be constructed at the initial construction. On the other hand, in flexible design 

that should be proposed in this study, only a runway would be constructed at the initial 

construction and then sometime after 10 years of operation, if the demand of passengers 

exceeds the capacity, a parallel taxiway and a south edge should be constructed; otherwise 

they would not be constructed. Thus, if uncertain demand is unfavorable, it is not 

necessary to exercise the option. It is possible to reduce upfront capital investment and 

therefore reduce losses. The Flexibility to expand them takes advantage of demand that is 

higher than expected in the deterministic projections. 

Payoff 

Expansion 

No Expansion 

Value of Project 

Figure 3-3 Payoff of Expansion Option 

3.3 Application of Real Options Analysis to Case Study 

3.3.1 Which Method should be used for the Case Study 


12/14/2007 


As introduced in section 3.1, there are three types of evaluation method in financial 

options, and these are also applied to real options analysis. Black-Scholes option pricing 

model and Binomial Lattice Model are widely used in financial options, but because they are 

very complex and needs financial skills, they are not so widely used in real options in 

engineering practice as in financial field. [7] Even if those methods are used appropriately, 

it is very difficult to explain to those managers who are not familiar with financial skills. [7] 

On the other hand, Monte Calro Simulation can be used more easily by using spread sheet 

than those two methods. Compared to conventional methods using financial mathematics, 

real options analysis using Monte Calro Simulation has advantages such as user friendly 

procedure, being based on data availability in practice, and being easy to explain graphically. 

[7] Thus real options analysis based on Monte Calro Simulation can be the appropriate as a 

tool for the valuation method for this case study. 

3.3.2 Real Options Analysis Using Monte Calro Simulation 

In this case study, real options analysis using Monte Calro Simulation method is used 

and its procedure is: 1) to estimate cash flows pro forma including capital investment, future 

costs and revenues of the project, and calculate the economic value of currently designed 

case, 2) to explore the effects of uncertainty by simulating possible scenarios, each of which 

leads to a various NPVs, and the collection of each scenario generate both an “expected net 

present value” (ENPV) and the distribution of possible outcomes for a project by 

demonstrating cumulative distribution functions, and 3) to explore ways to avoid the 

downside risk and take advantage of upside potential by exercising options. [7] 

4. Analysis of Case - Tokyo Int’l Airport New Runway Extension Project 

4.1 Analysis Condition 

In order to conduct real options analysis, it is necessary to set up analysis conditions 

such as cash flow pro forma including capital investment, revenue scheme, operating and 

maintenance costs. 

In this case study, three types of design should be considered. Case A refers to the 


12/14/2007 


current design that government actually planned and that would construct the whole runway 

island, including a parallel taxiway and a south edge. This design does not include 

flexibility. 

Case B refers to the design that constructs the whole runway island in the same way as 

Case A. But this design recognizes future uncertainty which is demand of passengers, 

while Case A considers deterministic projection of the demand of passengers. 

Case C refers to the design that constructs only runway area without parallel taxiway 

and south edge of the runway at the initial construction, and expands the parallel taxiway 

and south edge after 10 years operation if the demand exceeds the capacity of the airport for 

two consecutive years. 

Table 4-1 shows the summary of case of analysis. 

Table 4-1 Cases of Analysis 

Case 

Initial Future 

Investment Uncertainty 

Flexibility 

Case A 

R/W, 

PT/W & SE 

N/A 

N/A 

Case B 

R/W, 

PT/W & SE 

Recognizing N/A 

Case C R/W Recognizing 

Future Expansion 

PT/W & SE 

Source: Applied R de Neufville,et al [7] 

4.1.1 Capital Investment 

Capital investment here means the initial construction fee and the expansion fee of the 

parallel taxiway and the south edge. In the “New Runway Extension Project”, the 

construction fee is ¥570 billion, [2] so the initial investment in Case A and Case B is ¥570 

billion. In Case C, I assume that the initial investment can be set by prorating the area 

where it needs. In the area of the taxiway (827,625m 2 ), the structure of the runway island 

is landfill (¥64,700/m 2 ), [2] so the construction fee is ¥53.55 billion. In the area of the 

south edge (45,120m 2 ), the structure of the runway island is piled pier (¥786,500/m 2 ), [2] so 

the construction fee is ¥35.48 billion. Therefore, considering the bank revetment area 

(assumed 5% extra) and the joint area for the both areas so that the expansion could be 

possible, the initial construction fee of Case C is ¥505 billion. ([570-(53.55+35.48)]*1.05) 


12/14/2007 


The expansion cost of Case C is only the area of taxiway and the south edge, which is the 

same as the difference between the initial cost of Case A, B and that of Case C. 

Considering the extra fee of the joint area (assumed 10% extra), the expansion fee is ¥97.9 

billion. ((53.55+35.48)*1.05) The initial investment is assumed to be paid equally every 

year through the construction for 4 years, while the expansion investment is supposed to be 

paid equally every year for 2 years. 

Runway 

(R/W) 

Parallel Taxiway 

South Edge 

(Landfill) 

(Piled Pier) 

827,625m 2 45,120m 2 

Figure 4-1 Expansion Area of Runway D 

4.1.2 Revenues and Costs 

There are several types of revenues and costs in the airport system such as from 

aeronautical charges, non-aeronautical charges, and off-airport or non-operate charges. [8] 

In the current operation of Tokyo Int’l Airport, these charges applies, but in terms of a 

runway, aeronautical charges, which is the charges for services or facilities directly related to 

the processing of aircraft and their passengers, is the main charge. [8] The aeronautical 

charges have several categories such as landing charge, terminal-area air navigation charge, 

passenger service charge in terminals, security charge, charges for airport noise, and so on. 

[8] In terms of a runway, landing charge is the main source of the revenue. Thus, I 

assume that the revenue is only from the landing charge. The current landing charge in 

Tokyo Int’l Airport as of 2006 is ¥490,000 for B747-400D, ¥350,000 for B777-200, and 

¥230,000 for B777-724, and so on. [9] I set the average landing fee of this airport as 

¥332,000 / aircraft by assuming the probabilities for each type of aircraft of 23% for 

B747-400D, 35% for B777-200D, and 41% for B767-300. I also assume that this landing 

charge is fixed for the next 20 years. For the simplicity, I also set the revenue for each year 

is landing charge multiplied by the minimum of the number of passenger or the capacity. 


12/14/2007 


Costs for the operation and maintenance are categorized into the ones above. From the 

past disclosed information by the government, operating and maintenance cost in all airports 

in Japan in 2005 was ¥147.4 billion. [10] I assume that these costs depend on the scale of 

air traffic. The whole air traffic in Japan in 2005 was 1,431,000 and 300,000 in Tokyo Int’l 

Airport, which is about 20% of the whole traffic, so the operating and maintenance costs for 

Tokyo Int’l Airport in each year is assumed ¥29.48 billion, which is also assumed equally for 

the current three runways. So the operating cost and maintenance costs for the one runway 

is one-third of ¥29.48, which is equal to ¥9.84 billion. In addition, in the current 

construction plan indicates that the specific maintenance cost for the runway D is ¥100 

billion for the next 30 years. Thus the maintenance cost of ¥3.33 billion should also be 

included in the operating and maintenance cost for the new runway. Therefore, the 

operating and maintenance costs for the runway D are set as ¥13.17 billion. 

4.1.3 Capacity of Airport 

According to the Ministry of Land, Infrastructure and Transport in Japan, the improved 

maximum capacity of the airport by the runway extension in Tokyo Int’l Airport is 40 

aircrafts per hour, and it estimated this capacity can accommodate 87 million passengers per 

year. [2] So the capacity in terms of the number of passenger in this airport is 87 million. 

4.1.4 Discount Rate 

Discount rate is the opportunity cost of capital, and it can be calculated by the capital 

asset pricing model (CAPM). CAPM requires the risk-free rate of interest, the sensitivity 

of specific project or company to stock price, and the expected rate of return of the market. 

However, the government used the discount rate of 4% when it assessed not only this project 

but also any airport related project. [11] Thus, the discount rate of 4% is also used in this 

case study. 

4.2 Uncertainty in System 

There are usually a lot of uncertainties in large-scale engineering systems which include 

technical change, economic change, regulatory change, industrial change, political change, 

and so on. [4] This case also includes a lot of uncertainties such as demand of the number 


12/14/2007 


of passengers, capital investment in the future, the operating and maintenance costs, and 

unexpected events in the future. For the sake of the simplicity, I set only demand of the 

number of passengers as uncertainty. 

When forecasting the number of passenger in airports, 

the time span should be considered at most 20 years span and the volatility, which is the 

range of the chance the demand is higher or lower, should be considered by plus or minus 

50%. [12] 

4.3 Demand Forecasting 

According to the government estimation, it forecasted the demand of the number of 

passengers as 73.2 million in 2012, 80.3 million in 2017, and 85.5 million in 2022. [2] 

90.00 

Past and Predicted Number of Passengers at Tokyo Int'l Airport 

80.00 

Number of Passengers 

70.00 

60.00 

50.00 

40.00 

30.00 

20.00 

10.00 

0.00 

1983 

1985 

1987 

1989 

1991 

1993 

1995 

1997 

1999 

2001 

2003 

2012 

2022 

Year 

Figure 4-2 Demand Forecasting 

Source: Ministry of Land, Infrastructure and Transport in Japan [2] 

This estimation was based on the estimate of the total population in Japan, the estimate 

of GDP and other several factors. [13] However, forecast is always wrong. The table 

below is comparison of 10 years forecasts of international passenger to Japan with actual 

results. [12] These results clearly show that forecast has been always wrong. In addition 

the forecast of the total population in Japan and GDP are also very difficult to assess. Thus, 

the demand forecast above could be wrong. 


12/14/2007 


Table 4-2 Comparison of 10-year forecasts of international 

passengers to Japan with actual results 

Forecast Passengers (millions) Percent error 

For Done in Actual Forecast Over actual 

1980 1970 12.1 20.0 65 

1985 1975 17.6 27.0 53 

1990 1980 31.0 39.5 27 

1995 1985 43.6 37.9 (13) 

Source: R. de Neufville, A. Odoni, Airport Systems: Planning, Design, and Management[12] 

However, when assessing the real options analysis, it is necessary to rely on some 

forecasts, although they are nothing but estimation. Thus, in this case study, the demand 

curve that is based on the estimation of the total population and the estimation of GDP in 

Japan should be considered. The important thing is to recognize that this forecast is just an 

assumption and to evaluate this by using various demand scenarios with volatility. 

In this case study, the demand forecast is estimated based on the estimation of 

population [14] and the estimation of GDP [15] in Japan by conducting regression analysis. 

[16] The table and the graph below show the estimation of the number of passenger for the 

next 20 years in Tokyo Int’l Airport. 

Year 

Population 

(Thousands) 

GDP 

(in trillion) 

Passengers 

(million) 

1985 121.05 ¥350.3 24.27 

1986 121.66 ¥360.8 25.88 

1987 122.24 ¥374.5 28.19 

1988 122.75 ¥400.0 30.28 

1989 123.21 ¥421.2 34.96 

1990 123.61 ¥443.1 38.09 

1991 124.10 ¥458.2 40.06 

1992 124.57 ¥462.7 39.98 

1993 124.94 ¥463.7 39.12 

1994 125.27 ¥468.8 41.44 

1995 125.57 ¥477.7 43.02 

1996 125.86 ¥489.9 45.08 

1997 126.16 ¥496.8 47.43 

1998 126.47 ¥488.0 49.91 

1999 126.67 ¥487.0 52.27 

2000 126.93 ¥501.3 54.77 

2001 127.32 ¥503.2 57.01 

2002 127.49 ¥503.9 59.49 

2003 127.69 ¥512.8 59.41 

2004 127.79 ¥524.6 59.05 

2005 127.77 ¥538.9 59.47 

2006 127.76 ¥544.8 61.08 

2007 127.69 ¥550.8 62.23 

2008 127.57 ¥556.9 63.37 

2009 127.40 ¥563.0 64.49 

2010 127.18 ¥569.2 65.59 

2011 126.91 ¥578.9 67.35 

2012 126.60 ¥588.7 69.11 

2013 126.25 ¥598.7 70.87 

2014 125.86 ¥608.9 72.63 

2015 125.43 ¥619.3 74.39 

2016 124.96 ¥629.8 76.14 

2017 124.46 ¥640.5 77.89 

2018 123.92 ¥651.4 79.64 

2019 123.34 ¥662.4 81.38 

2020 122.73 ¥673.7 83.12 

2021 122.10 ¥683.1 84.46 

2022 121.43 ¥692.7 85.80 

2023 120.74 ¥702.4 87.11 

2024 120.01 ¥712.2 88.42 

2025 119.27 ¥722.2 89.70 

2026 118.50 ¥732.3 90.98 

2027 117.71 ¥742.6 92.24 

2028 116.90 ¥753.0 93.48 

2029 116.07 ¥763.5 94.71 

2030 115.22 ¥774.2 95.92 

Demand (PAX in million) 

120 

100 

80 

60 

40 

20 

0 

Demand Forecasting 

Projected Demand (Government) 

Projected Demand (This Project) 

1982 1987 1992 1997 2002 2007 2012 2017 2022 2027 

Time (Year) 

PAX = 2 * A * { - 0.1818 + 0.0077 * B } 

(A: Estimated Population, B: Estimated GDP) 

Figure 4-3 Demand Forecast of Number of Passengers in Tokyo Int’l Airport 


12/14/2007 


The demand curve shown in Figure 4-3 is still deterministic projection, and this is 

usually used for the static analysis. However, it is essential to recognize and consider the 

uncertainty in this demand forecasting. Thus, uncertainty is recognized in the model by 

simulating possible scenarios. It indicates how fluctuations can be incorporated around 

deterministic projections based on the relevant probability distribution. [17] In this case 

study, 2,000 Monte Calro Simulations are generated where all of those simulations create 

each demand scenarios over the 20 years span. Figure 4-4 shows some of the examples of 

simulations of the uncertain demand. All of these scenarios can be considered and 

incorporated into the calculation of the expected value of the plans statistically. 

120 


120 



100 

80 

60 

40 

20 

0 



Demand scenario 

1982 1987 1992 1997 2002 2007 2012 2017 2022 2027 

Time (Year) 


100 

80 

60 

40 

20 

0 




1982 1987 1992 1997 2002 2007 2012 2017 2022 2027 

Time (Year) 

120 


120 



100 

80 

60 

40 

20 





100 

80 

60 

40 

20 




0 

1982 1987 1992 1997 2002 2007 2012 2017 2022 2027 

Time (Year) 

0 

1982 1987 1992 1997 2002 2007 2012 2017 2022 2027 

Time (Year) 

Figure 4-4 Examples of Simulation of the Uncertain Demand 

4.4 Summary of Analysis Condition 

The Table 4-3 shows the summary of analysis condition. 


12/14/2007 


Table 4-3 Summary of Analysis Condition 

Initial Future 

Investment Expansion 

Perspective Simulation Option 

Design A 

R/W, 

PT/W & SE 

N/A Deterministic No No 

Design B 

R/W, 

Recognizing 

N/A 

PT/W & SE 

Uncertainty 

Yes No 

Design C R/W PT/W & SE Incorporating 

Flexibility 

Yes Yes 

Source: Applied R de Neufville, S. Scholtes, T. Wang [7] 

4.5 Result of Analysis 

Table 4-4, 4-5, 4-6 shows the cash flow pro forma for three cases. Table 4-4 shows the 

cash flow pro forma and calculating the NPV of the project assuming the demand of the 

number of passengers grows as projected. (Case A) In this case, the expected NPV 

(ENPV) of the project is ¥678.8 billion. However, this value is not realistic since the actual 

demand of the number of passengers can change from this deterministic value. 

Table 4-5 shows the cash flow pro forma and calculating the NPV of the project 

recognizing uncertainty. (Case B) In this case, Monte Calro Simulation is conducted and it 

produces 2,000 possible demand scenarios. The ENPV of this case is ¥638.3 billion. But 

this ENPV is just one of the 2,000 scenarios. This simulation can generate 2,000 ENPV in 

each scenario, and can also generate distribution for each scenario. The actual ENPV is 

distributed as shown in Figure 4-5. The average of ENPV is ¥569.3 billion, which is less 

than that of the Case A. Although this design assumes that there are equal chances that 

demand changes to higher and to lower, this design limits the higher value of the project 

since the capacity is fixed, while there are still lower chances to generate losses. [7] 

Table 4-6 shows the cash flow pro forma and calculating the NPV of the project 

recognizing uncertainty and holding the option to expand. (Case C) This case also 

recognizes uncertainty and Monte Calro Simulation is conducted, which produces the ENPV 

of ¥686.7 billion. But this ENPV is also just one of the 2,000 scenarios. In this design, 

the actual ENPV is distributed as shown in Figure 4-5, and the average of ENPV is ¥621.6 

billion, which is higher than that of Case B. Because the option to expand is exercised 

when the demand is higher than the current capacity, the ENPV is improved higher. The 


12/14/2007 


estimated value of the option exercised in order to incorporate the flexibility into design can 

be calculated by the difference between the ENPV of Case C and that of Case B, which is 

¥52.3 billion. Table 4-7 shows the summary of this analysis. Case C, which holds 

flexibility in design has advantages over Case B in every aspect. Table 4-6 shows the 

summary of the analysis. [7] The initial investment in Case C is less than that of Case B, 

which means that the flexibility can reduce the initial investment. The ENPV, the 

maximum and minimum NPV in Case C is higher than that of Case B, so the flexibility can 

enhance the overall value of the project. 

Probability 

100.0% 

90.0% 

80.0% 

70.0% 

60.0% 

50.0% 

40.0% 

30.0% 

20.0% 

10.0% 

0.0% 

Cumulative distribution function 

Case B 

Average ENPV(B) 

¥569.3 B 

200,000 300,000 400,000 500,000 600,000 700,000 800,000 900,000 

Target value \ million 

Case C 

Case A 

(¥678.7 B) 

Option Value 

¥52.3 B 

Average ENPV(C) 

¥621.6 B 

Figure 4-5 Cumulative Distribution Function (Value at Risk) 

Table 4-7 Summary of the Analysis 

Case B 

(No Flexibility) 

Case C 

(with Flexibility) 

Initial Investment (¥ in billion) 570.0 505.0 

ENPV (¥ in billion) 569.3 621.6 

Minimum NPV (¥ in billion) 339.3 338.1 

Maximum NPV (¥ in billion) 773.4 899.1 


12/14/2007 


Table 4-4 Analysis Result (Case A) 

NPV Static Model (Design A) 

Construction Phase Operation Phase 

Year 

2010 2007 2008 2009 2011 2012 2013 2014 2015 2016 2017 2018 2019 2020 2021 2022 2023 2024 2025 2026 2027 

Demand (Deterministic Projection) 62,233,839 63,366,512 64,486,139 65,592,131 67,353,149 69,114,281 70,874,568 72,633,340 74,389,968 76,143,790 77,894,119 79,640,235 81,381,152 83,116,111 84,463,179 85,795,633 87,113,387 88,416,454 89,704,914 90,978,530 92,237,068 

Capacity (Constant) 87,000,000 87,000,000 87,000,000 87,000,000 87,000,000 87,000,000 87,000,000 87,000,000 87,000,000 87,000,000 87,000,000 87,000,000 87,000,000 87,000,000 87,000,000 87,000,000 87,000,000 87,000,000 87,000,000 87,000,000 87,000,000 

Revenue (Yen in M) 107,696 110,512 113,326 116,139 118,948 121,752 124,551 127,343 130,126 132,900 135,054 137,185 139,111 139,111 139,111 139,111 139,111 

Operating & Maintenance Cost (Yen in M) 13,170 13,170 13,170 13,170 13,170 13,170 13,170 13,170 13,170 13,170 13,170 13,170 13,170 13,170 13,170 13,170 13,170 

Initial Capital Investment (Yen in M) 142,500 142,500 142,500 142,500 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 

Cash Flow (Yen in M) -142,500 -142,500 -142,500 -142,500 94,526 97,342 100,156 102,969 105,778 108,582 111,381 114,173 116,956 119,730 121,884 124,015 125,941 125,941 125,941 125,941 125,941 

Present Value Cash Flow (Yen in M) -142,500 -137,019 -131,749 -126,682 80,801 80,008 79,155 78,248 77,291 76,288 75,245 74,164 73,051 71,907 70,385 68,861 67,241 64,655 62,168 59,777 57,478 

Net Present Value (Yen in M) 678,770 

Table 4-5 Analysis Result (Case B) 

NPV Model with Flexibility (Design B) 

Construction Phase Operation Phase 

Year 

2010 2007 2008 2009 2011 2012 2013 2014 2015 2016 2017 2018 2019 2020 2021 2022 2023 2024 2025 2026 2027 

Demand (Deterministic ,131 Projection) 62,233,839 63,366,512 64,486,139 65,592 67,353,149 69,114,281 70,874,568 72,633,340 74,389,968 76,143,790 77,894,119 79,640,235 81,381,152 83,116,111 84,463,179 85,795,633 87,113,387 88,416,454 89,704,914 90,978,530 92,237,068 

Random Demand 62 ,233,839 57,868,139 65,908,629 38,319,447 62,649,153 96,258,574 105,906,007 107,792,690 40,892,745 70,852,191 107,971,158 67,571,602 102,961,943 115,202,209 103,423,185 54,083,607 108,556,015 89,944,244 51,261,467 122,025,498 82,989,813 

Capacity (Constant) 87,000,000 87,000,000 87,000,000 87,000,000 87,000,000 87,000,000 87,000,000 87,000,000 87,000,000 87,000,000 87,000,000 87,000,000 87,000,000 87,000,000 87,000,000 87,000,000 87,000,000 87,000,000 87,000,000 87,000,000 87,000,000 

Revenue (YenM) 100,174 139,111 139,111 139,111 65,386 113,291 139,111 108,045 139,111 139,111 139,111 86,478 139,111 139,111 81,966 139,111 132,698 

Operating & Maintenance Cost (Yen in M) 13,170 13,170 13,170 13,170 13,170 13,170 13,170 13,170 13,170 13,170 13,170 13,170 13,170 13,170 13,170 13,170 13,170 

Initial Capital Investment (Yen in M) 142,500 142,500 142,500 142,500 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 

Cash Flow (Yen in M) -142,500 -142,500 -142,500 -142,500 87,004 125,941 125,941 125,941 52,216 100,121 125,941 94,875 125,941 125,941 125,941 73,308 125,941 125,941 68,796 125,941 119,528 

Present Value Cash Flow (Yen in M) -142,500 -137,019 -131,749 -126,682 74,372 103,514 99,533 95,705 38,154 70,343 85,081 61,629 78,662 75,637 72,728 40,705 67,241 64,655 33,959 59,777 54,551 


Table 4-6 Analysis Result (Case C) 

NPV Model with Expansion Option (Design C) 

Ope ration Phase (No Expansion) 

Construction Phase Operation Phase with Expansion Optiton 

Year 

2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 2020 2021 2022 2023 2024 2025 2026 2027 2028 2029 2030 

Demand (Deterministic Projection) 62,233,839 63,366,512 64,486,139 65,592 ,131 67,353,149 69,114,281 70,874,568 72,633,340 74,389,968 76,143,790 77,894,119 79,640,235 81,381,152 83,116,111 84,463,179 85,795,633 87,113,387 88,416,454 89,704,914 90,978,530 92,237,068 93,480,029 94,706,438 95,915,430 

Random Demand 62,233,839 39,894,335 54,103,487 51,525,702 76,708,558 90,411,186 78,516,141 85,421,842 58,639,900 53,141,813 101,463,792 81,211,366 116,851,429 99,558,999 80,045,196 46,217,658 76,364,557 94,004,581 71,889,072 85,106,617 118,002,960 108,274,523 116,423,514 134,211,221 

Ca pacity (Flexible) 87,000,000 87,000,000 87,000,000 87,000,000 87,000,000 87,000,000 87,000,000 87,000,000 87,000,000 87,000,000 87,000,000 87,000,000 87,000,000 87,000,000 121,800,000 121,800,000 121,800,000 121,800,000 121,800,000 121,800,000 121,800,000 121,800,000 121,800,000 121,800,000 

Expansion Expand 

Build Extra Capacity 0 0 0 0 0 0 0 0 0 0 0 0 0 34,800,000 0 0 0 0 0 0 0 0 0 0 

Revenue (YenM) 122,655 139,111 125,545 136,587 93,764 84,972 139,111 129,855 139,111 139,111 127,990 73,901 122,105 150,311 114,949 136,083 188,683 173,128 186,158 194,755 

Operating & Maintenance Cost (Yen in M) 12,580 12,580 12,580 12,580 12,580 12,580 12,580 12,580 12,580 12,580 13,170 13,170 13,170 13,170 13,170 13,170 13,170 13,170 13,170 13,170 

Initial Capital Investment (Yen in M) 126,255 126,255 126,255 126,255 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 

Expansion Expenditure (Yen in M) 0 0 0 0 0 0 0 0 0 0 0 48,967 48,967 0 0 0 0 0 0 0 0 0 0 0 

Cash Flow (Yen in M) -126,255 -126,255 -126,255 -126,255 110,075 126,531 112,965 124,007 81,184 72,392 126,531 68,308 77,564 126,531 114,820 60,731 108,935 137,141 101,779 122,913 175,513 159,958 172,988 181,585 

Present Value Cash Flow (Yen in M) -126,255 -121,399 -116,729 -112,240 94,092 103,999 89,278 94,235 59,320 50,862 85,480 44,372 48,446 75,991 66,306 33,722 58,161 70,404 50,241 58,340 80,102 70,195 72,993 73,674 

Present Value Expansion Cost (Yen in M) 0 0 0 0 0 0 0 0 0 0 0 31,808 30,584 0 0 0 0 0 0 0 0 0 0 0 



12/14/2007 


5. Conclusion 

There are a lot of facilities that are not designed optimally in airport systems. The 

master plan of those designs does not anticipate and consider future risks and uncertainties. 

Thus, inflexible design cannot manage risks and uncertainties, and it leads to losses. In 

order to solve this problem, it is essential to incorporate flexibility into design. As 

demonstrated in the case study in this paper, the real option analysis is the useful way to 

recognize uncertainty and incorporate flexibility into design. The case study demonstrated 

that flexibility can enhance the expected value of the project, and reduce the possible losses 

by using the option to expand. Also it can demonstrate that flexibility can keep the initial 

cost lower than that of the current design. Therefore, at the initial construction phase, only 

runway should be constructed and if the demand of the number of passenger exceeds the 

capacity after 10 years operation, a parallel taxiway and a south edge should be constructed. 

Furthermore, real options analysis using Monte Calro Simulation is very useful in that it just 

focuses on the ENPV of options and it does not require any financial skills unlike typical 

financial options approaches such as Black-Sholes option pricing model and binomial lattice 

model. Thus, those who are in charge of design can relatively easily use this method. 

Flexible design thus enables projects to be optimized and reduce excessiveness and losses in 

airport systems. 


12/14/2007 


6. Reference 

[1] R. de Neufville, Lecture note of the MIT course of 1.231: Planning and Design of Airport 

System (Fall 2007). 

[2] Ministry of Land, Infrastructure and Transport in Japan, Outline of Tokyo Int’l Airport 

New Runway Extension Project, 2007. 

http://www.mlit.go.jp/koku/04_outline/01_kuko/02_haneda/index.html 

[3] R. A. Brealey, S.C. Myers, and F. Allen, Principles of Corporate Finance, 8th ed., (New 

York: McGraw-Hill Publishing Company, 2005). 

[4] R. de Neufville, Lecture notes of the MIT course of ESD.71: Engineering Systems 

Analysis for Design (Fall 2006). 

[5] K. Hodota, “R&D and Deployment Valuation of Intelligent Transportation Systems: A 

Case Example of the Intersection Collision Avoidance Systems,” M.S. Thesis in Master 

of Science in Transportation, MIT, Cambridge, MA, 2006. 

[6] T. Copeland, T. Koller, J. Murrin, Valuation: Measuring and Managing the Value of 

Companies, Fourth Edition (McKinsey & Company Inc., 2002). 

[7] R. de Neufville, S. Scholtes, T. Wang, Valuing Real Options by Spread Sheet : Parking 

Garage Case Example, January 2005. 

[8] A. Odoni, Lecture notes of the MIT course of 1.231: Planning and Design of Airport 

System (Fall 2007). 

[9] Ministry of Land, Infrastructure and Transport in Japan, Civil Aviation Breau. 

http://www.mlit.go.jp/singikai/koutusin/koku/seibi/14/images/shiryou1_22.pdf 

[10] Ministry of Land, Infrastructure and Transport in Japan, Survey of the Air Traffic 

Condition in Japan, 2003, http://www.mlit.go.jp/kisha/kisha03/12/120523_3/05.pdf 

[11] Katsuya Hihara, Research for New Operation System of Transportation Policy 

Considering Uncertainty, 2004, Policy Research Institute for Land Infrastructure and 

Transport, https://www.mlit.go.jp/pri/houkoku/gaiyou/pdf/kkk9.pdf 

[12] R. de Neufville, A. Odoni, Airport Systems: Planning, Design, and Management, 2nd 

ed., (New York: McGraw-Hill Publishing Company, 2003) 

[13] Ministry of Land, Infrastructure and Transport in Japan, Forecast of air traffic, 

http://www.mlit.go.jp/singikai/koutusin/koku/07_9/01.pdf 

[14] National Institute of Population and Social Security Research, Forecast of Population 

http://www.ipss.go.jp/syoushika/tohkei/suikei07/suikei.html#chapt1-1 

[15] Ministry of Land, Infrastructure and Transport in Japan, Transition and Forecast of GDP, 

http://www.mlit.go.jp/road/kanren/suikei/7-1.pdf 

[16] R. de Neufville, Forecasting Assignment of the MIT course of 1.231: Planning and 

Design of Airport System (Fall 2007) 

[17] Michel-Alexandre Cardin, “Facing Reality: Designing and Management of Flexibility 

Engineering Systems,” M.S. Thesis in Master of Science in Technology and Policy, 

Massachusetts Institute of Technology, Cambridge, MA, 2007.

Flexible Design of Airport System Using Real Options Analysis - MIT

You also want an ePaper? Increase the reach of your titles

Delete template?

Save as template?