Akita University

Combining Economic, Geological, and Technical Uncertainties in

Mining Projects Valuation Using Real Options Analysis

2022

Mohammad Rahman Ardhiansyah

Akita University
Akita University

Doctoral Thesis 2022

Combining Economic, Geological, and Technical Uncertainties in

Mining Projects Valuation Using Real Options Analysis

Department of Geosciences, Geotechnology, and Material Engineering for Resources

Graduate School of International Resource Sciences

D6519113

Mohammad Rahman Ardhiansyah

Supervisor Professor Tsuyoshi Adachi

September 2022
Akita University
Akita University

Acknowledgment

Professor Tsuyoshi Adachi, as supervisor, deserves special thanks for his patience,
support, and academic counsel during this project. The author also expresses gratitude to
Professor Youhei Kawamura, Professor Atsushi Shibayama, and Professor Tadao Imai for
advising to make this research better.

The author would like to thank the Ministry of Education, Culture, Sports, Science, and
Technology Japan for financial support during the study, namely MEXT scholarship program.
His sincere thanks also go out to Akita University's Academic Affairs Division and
International Affairs Division employees. The author also appreciates his comrades at the
Mineral Economics Laboratory for their warm smiles, assistance, support, and general
camaraderie.

The writer's thankfulness is also to the Faculty of Mining and Petroleum Engineering of
Institut Teknologi Bandung for recommending and being accepted at this study level and
Universitas Islam Bandung for allowing him to take higher education at Akita University. In
addition, the author is sincerely grateful to PT Timah, Tbk, who gave him a permit to do
research on their mining area and the exploration and related economic data.

This thesis is dedicated to author's wife, Alfi Rachmawati, for her undying support,
unwavering patience, and bountiful love, enough to last a lifetime and a bit more. Nevertheless,
the author's special gratitude extends to his parents for their everlasting love and support.

i
Akita University

Table of Contents
Acknowledgment .................................................................................................................. i
Table of Contents................................................................................................................. ii
List of Tables ...................................................................................................................... iv
List of Figures .......................................................................................................................v
Abstract............................................................................................................................. viii

Chapter 1 Introduction ................................................................................................... 1

1.1. Background .....................................................................................................................1
1.2. Problem Statement ...........................................................................................................4
1.3. Purpose of Study..............................................................................................................5
1.4. Significance of Study .......................................................................................................6
1.5. Outline of Study ..............................................................................................................7
1.6. References .......................................................................................................................9

Chapter 2 Project Evaluation ....................................................................................... 11

2.1. Exploration .................................................................................................................... 11
2.2. Mine Planning ............................................................................................................... 18
2.3. Mining Economics ......................................................................................................... 22
2.4. Sensitivity Analysis ....................................................................................................... 28
2.5. References ..................................................................................................................... 30

Chapter 3 Real Option Analysis ................................................................................... 31

3.1. Real Option Analysis Framework .................................................................................. 31
3.2. Black Scholes (BS) Model ............................................................................................. 36
3.3. Binomial Lattice (BL) Valuation .................................................................................... 41
3.4. Simulation Method ........................................................................................................ 43
3.5. References ..................................................................................................................... 46

Chapter 4 Multi-stage Stratified State Aggregation .................................................... 47

4.1. Stratified State Aggregation (SSA) ................................................................................ 47
4.2. Economic Uncertainty Model ........................................................................................ 51
4.3. Geological Uncertainty Model ....................................................................................... 56
4.4. Technical Uncertainty Model ......................................................................................... 61

ii
Akita University

4.5. Combined Uncertainty Assessment ................................................................................ 65

4.6. Economic and Geological Uncertainty Case ................................................................... 74
4.7. References ..................................................................................................................... 80

Chapter 5 Conclusions .................................................................................................. 83

APPENDIX ...................................................................................................................... 87
A. Price Modelling ........................................................................................................ 87
B. Resource Summary ................................................................................................... 90
C. Reserve Summary ..................................................................................................... 92
D. Cost Calculation........................................................................................................ 95

iii
Akita University

List of Tables
Table 2.1 Resource Summary ............................................................................................... 15

Table 2.2 Mine Planning Parameters..................................................................................... 19

Table 2.3 Ore and Waste Mine Production ...........................................................................19

Table 2.4 Cash Flow Model for Mid Reserve ........................................................................ 26

Table 2.5 Cash Flow Model for South Reserve ..................................................................... 27

Table 3.1 Empirical Example of the ROV and DCF for Commodities................................... 32

Table 3.2 Option Parameters................................................................................................. 35

Table 3.3 Project Volatility Estimation Based on Historical Tin Price ................................... 39

Table 3.4 Project Value ........................................................................................................ 43

Table 3.5 Project Value Minus Initial Investment ................................................................. 43

Table 3.6 Backward Calculation ........................................................................................... 43

Table 4.1 ADF test on Tin Price data (2016-2020) ................................................................ 51

Table 4.2 Tin Price Volatility .............................................................................................. 53

Table 4.3 Comparison of Real Option Approach ................................................................... 56

Table 4.4 ADF Test of Cost data........................................................................................... 62

Table 4.5 Present Value Regarding Price-Reserve Changing................................................. 70

Table 4.6 Option Value Regarding Price-Reserve Changing ................................................. 70

Table 4.7 Present Value of Reserve and Price Relative Changing ......................................... 76

Table 4.8 Option Value of Reserve and Price Relative Changing ..........................................76

iv
Akita University

List of Figures

Figure 1.1. Resource and Reserve Classification .....................................................................1

Figure 2.1 Basic Statistic Data .............................................................................................. 13

Figure 2.2 Variogram Model ................................................................................................ 14

Figure 2.3 Ordinary Kriging Model for South Reserve.......................................................... 16

Figure 2.4 Ordinary Kriging Model for Mid Reserve ............................................................ 16

Figure 2.5 Conditional Simulation Model for South Reserve ................................................ 17

Figure 2.6 Conditional Simulation Model for Mid Reserve ................................................... 17

Figure 2.7 Iterative Mine Planning Flow Chart ..................................................................... 18

Figure 2.8 Ultimate Pit Limit Design for Mid Reserve .......................................................... 20

Figure 2.9 Ultimate Pit Limit Design for South Reserve ....................................................... 21

Figure 2.10 Evaluation Metric at Feasibility Stage ................................................................ 22

Figure 2.11 Mid Monthly Cash Flow Model ......................................................................... 24

Figure 2.12 South Monthly Cash Flow Model ......................................................................24

Figure 2.13 Sensitivity Analysis ........................................................................................... 28

Figure 3.1 Result of option analysis ...................................................................................... 34

Figure 3.2 Price and Project Uncertainty ............................................................................... 40

Figure 3.3 Stock and Option Value in One Period ................................................................. 41

Figure 4.1 Multi-stage Stratified State Aggregation Methodology......................................... 48

Figure 4.2 Stratified State Aggregation (SSA) Methodology ................................................. 49

Figure 4.3 Price Simulation Model (Geometric Brownian Model) ........................................ 54

Figure 4.4 Project Value Considering Price Changing only ................................................... 55

Figure 4.5 Conditional Simulation for Modelling Reserve .................................................... 57

v
Akita University

Figure 4.6 Mid Reserve Grade-Tonnage ............................................................................... 57

Figure 4.7 Mid Grade Uncertainty ........................................................................................ 58

Figure 4.8 Probability Distribution (Mid Reserve) ................................................................ 58

Figure 4.9 South Reserve Grade-Tonnage .............................................................................59

Figure 4.10 Probability Distribution (South Reserve) ............................................................ 59

Figure 4.11 South Grade Uncertainty .................................................................................... 60

Figure 4.12 Project Value Considering Grade Changing only ............................................... 60

Figure 4.13 Historical Operating Cost Data ..........................................................................61

Figure 4.14 Modified Historical Unit Cost Data.................................................................... 62

Figure 4.15 Regression Analysis of Historical Cost .............................................................. 63

Figure 4.16 Cost Uncertainty Model (GBM vs MR) ............................................................. 64

Figure 4.17 Project Value Considering Cost Changing only .................................................. 65

Figure 4.18 Generated Cells with Three Uncertainties Driver (Random Cells Picked) ..........65

Figure 4.19 Exercise Value with Three Uncertainties Driver (Random Cells Picked) ............ 66

Figure 4.20 Waiting Value with Three Uncertainties Driver (Random Cells Picked).............66

Figure 4.21 Probability Distribution...................................................................................... 67

Figure 4.22 Project Value with Initial Price Changing .......................................................... 68

Figure 4.23 Project Value with Initial Total Reserve Changing .............................................69

Figure 4.24 Project Value with Initial Total Cost Changing .................................................. 69

Figure 4.25 Initial Price and Total Reserve Decision Map..................................................... 71

Figure 4.26 Initial Price and Unit Cost Decision Map ........................................................... 72

Figure 4.27 Reserve - Cost Uncertainty Scenario Analysis.................................................... 73

Figure 4.28 Project Value with Price Changing..................................................................... 74

Figure 4.29 Project Value with Reserve Changing ................................................................ 75

vi
Akita University

Figure 4.30 Decision Initial Price-Reserve Map .................................................................... 77

Figure 4.31 Real Option (RO) Value with Price and Reserve Uncertainties Alteration ..........78

vii
Akita University

Abstract
The selling of examined metal is the only mining company revenue generator. It implies
a certain amount of metal and selling price would grow their revenue. Before mining, the
company should estimate the reserves with tremendous uncertainty. In addition, fluctuation of
the metal price as economic uncertainty forces the company to deal with an advanced strategy
to gain optimum value because profit is very sensitive to the selling price. Nevertheless, the
cost of mining operations is unstable due to both local and global economic conditions. Cost
instability represents the technical uncertainty of the process that consists of mining and
processing-related expenditure.

Geology and mining operations need excessive capital expenditure to run the project.
In contrast, finance has to press the expense to guarantee profit. A balance between them is the
crucial success of mining. Exploration is necessary to estimate the resource. The estimation is
detailed with core drilling at a considerable cost and yielded a certain confidence level of
reserves with inherent uncertainty. Sales prices and expenses are commonly assumed in
constant or constant growth without compromising their fluctuation. However, project
evaluation cannot ignore those uncertainties to get the project value.

A standard method to evaluate a project value, Discounted Cash Flow (DCF), could not
adequately account for the future risk. Net Present Value (NPV) as the decision parameter of
the DCF method theoretically only generates two decision areas that are accepted and declined.
While in reality, management commonly takes no action to wait for a reasonable commodity
price, stable cost and ensure the number of reserves by collecting more exploration data. As a
result, recent study dedicated to accounting for uncertainty, real options (RO) valuation, adapt
financial option theory to be practiced in a real business. On the other hand, uncertainty in
reserves is modelled by geostatistics methodology, kriging, and conditional simulation, which
captures the spatial variability of the deposits.

There are three approaches in RO methodology: Black Scholes (BS) Valuation,

Binomial Lattice (BL) Valuation, and simulation. The complexity of RO in BS and BL
approach arises when considering multi uncertainties in project evaluation. On the other hand,
simulation approach in RO is not well developed. As a result, RO studies often only consider
price as an uncertainty driver. This research combined price, grade and cost uncertainty in a
viii
Akita University

mining project evaluation through the simulation approach, namely Multistage Stratified Stage
Aggregation (MSSA), which would be the study's originality. Conditional simulation methods
in geostatistics will be utilized to account for grade uncertainty. Thus, the expected reserve and
the deviation are incorporated with commodity price and cost uncertainties.

This study demonstrated a project evaluation method covering resource estimation,

mine planning, economic evaluation, and uncertainty assessment. The data was collected from
PT Timah, Tbk, the most significant world tin producer in 2020, located in Indonesia. The data
consisted of drill hole exploration and historical operation costs, while price data was recorded
from the S&P 500. Those data were followed by resource estimation using conditional
simulation, particularly Sequential Gaussian Simulation, which was run with the GEOVIA
Surpac mining program. Mine planning and project evaluation were conducted with the
GEOVIA Whittle mining program and converted to a monthly cash flow model. Finally, the
uncertainty assessment was done with the real options method, especially MSSA, which
utilized java programming language. The originality of the methodology was the development
of path generation through java which is an essential step in the MSSA method.

Our research was a pilot method that demonstrated a combination of advances in

resource estimation and economic evaluation. The conditional simulation method indicated that
each reserve location had its geological uncertainty. Furthermore, Geometric Brownian Model
(GBM) was used to make price simulations and get the price uncertainty. Lastly, we utilized
the Mean Reverting Process to model cost uncertainty. The three uncertainties represented
geological, economic, and technical uncertainties, respectively. MSSA is an alternative method
to get project value considering those three uncertainties. A benchmark comparison of the
MSSA result with BS and BL approaches resulted in a slight difference.

ix
Akita University

In summary, we developed a project evaluation methodology that considered

geological, economic, and technical uncertainties represented by grade, price and cost,
respectively. In our case study, the manager can run the project, but they must ensure the project
cost. In addition, the price and geological uncertainties will not be revealed until they decide to
mine its reserve; thus, controlling the production and price is essential to guarantee project
profitability.

Keywords: project evaluation, real options analysis, uncertainty modeling, combining

uncertainties, Multistage Stratified State Aggregation

x
Akita University

Chapter 1 Introduction

1.1. Background

Standard mining project valuation starts from exploration to economic evaluation

conducted by the company. The general purpose of exploration is to delineate valuable
resources and estimate the reserve. Moreover, resources and reserves classification in
exploration involves several branches of knowledge. Figure 1.1 depicts the framework for
classifying mineral deposit to reflect varying levels of geological confidence and different
degrees of technical and economic judgement, summarized as modified factors. Mineral
resources can be estimated based on geoscientific data. Ore reserves, which are an altered
selection of the indicated and measured mineral resources (shown within the dashed outline in
Figure 1.1), must consider the modifying factors that affect extraction. It should be assessed
with input from various disciplines. A competent person can transform measured mineral
resources into either proven or probable ore reserves (The JORC Code 2012 Edition, 2012).

Figure 1.1. Resource and Reserve Classification

Source: JORC Code 2012 Edition, 2012

1
Akita University

After exploration, a mining company obtains reserve estimation in a particular area with
inherited uncertainty. Then, the reserve is evaluated in a feasibility study, including mine
planning to extract the expected amount of metal. Once the reserve is extracted, the company
invests in equipment, processing plant, and other expenditure. The investment should be
assessed in detail, including the uncertainty. Practically, the evaluation method used by a
company is the traditional Discounted Cash Flow (DCF) methodology. However, the method
employed by common mining companies does not account for risk and uncertainty. Decision-
makers frequently modify the discount rate to account for risk and integrate technical risk
premium. Aside from that, they conducted a sensitivity analysis to forecast the impact of
changing cash flow parameters. In some circumstances, this method might be a significant
limitation in adoption; and an improper valuation technique (Nicholas, 2014).

Geology, mine plan, and finance departments forecast the DCF return of a typical
mining project based on significant assumptions about price, reserves, and cost. These profit
determinants come with risk and uncertainty. To date, the best method for modelling the
resource beneath the surface is to use a coring drill. A significant geological uncertainty, grade
uncertainty, will be assessed in this study through conditional simulation (CS) methodology.
CS resulted in resource estimation and inherited geological uncertainty (Dowd, 1994). The CS
method developed further kriging method, particularly generation of grade uncertainty. The
study considered geological uncertainty which is typically not identified in economic evaluation
instead of estimation.

With the confirmation of modifying factors, the resource was transformed into reserves.
Operating costs, including mining, processing, and metallurgical operations, represent technical
uncertainty. Furthermore, price uncertainty is a mining project's most significant profit
determinant. Failure to respond to those risks and uncertainties will cause a significant mistake
in evaluating the future worth of the project. Risk refers to the possibility of an undesirable
occurrence occurring. In contrast, uncertainty refers to the future outlook of the project being
unpredictable (Mun, 2006).

Researchers devised the Real Option (RO) to replace the DCF technique to account for
risk and uncertainty. An option is a financial derivative whose value depends on the price of a
stock. Option holders have the right but no obligation to execute their option by the maturity
date. These alternatives are quantified by option calculation to justify option value. The
adaptation of option calculation in the actual project is called real options (RO) (Hull, 2015).

2
Akita University

To date, three approaches are used to calculate the value of an option: Black-Scholes
Formulation, Lattice Valuation and Simulation Method. The Black Scholes Formula employed
a closed mathematical approach (Black & Scholes, 1973). On the other hand, lattice valuation
(Cox et al., 1979) used risk-neutral probability, which simplified the Black Scholes method's
mathematical equation.

The simulation technique has similarities with lattice valuation. However, based on the
Monte Carlo simulation methodology, the simulation method is designed to calculate high
dimensional risk and uncertainty (Barraquand & Martineau, 1995; Hull, 2015). By optimizing
the exercise and waiting for the value of the project, the lattice and simulation methods consider
management flexibility. A temporal stochastic model and theoretical computation for natural
resource project appraisal are introduced in the study of RO in natural resources (Brennan &
Schwartz, 1985). Furthermore, the RO technique considers the managerial flexibility to amend
a future choice if the ambiguity is exposed. The Black Scholes method (Black & Scholes, 1973)
and lattice valuation (Cox et al., 1979), two common RO approaches, become complicated
when accounting for more than one uncertainty. Nonetheless, the mathematical complexity
(Haque et al., 2014) and usability (Ajak & Topal, 2015) of RO in the industry are still limiting
factors. As a result, the simulation technique is vital to be developed further to solve the
limitation.

The limitation of mathematical complexity was solving higher-order and dimensional

partial differential equations. The shortage of empirical BS and BL approach was generally
addressed by the simulation approach, Stratified State Aggregation (SSA). It was founded by
Barraquand & Martineau (Barraquand & Martineau, 1995) for the American option and later
developed by Adachi et al. for an oil industry (Adachi et al., 2008). However, mining is a unique
industry that has complex uncertainties, and no detailed research has been developed.
Therefore, we proposed an originality on evaluation method on a mining project considering
grade, price and cost uncertainties through developed SSA approach.

A case study is conducted on tin mining projects at PT Timah Tbk, a mining company
in Indonesia, particularly on monthly mine planning for underwater mining and using dredge
as the primary equipment. Economic, geological, and technical factors of uncertainties are
represented by price, grade, and cost, respectively. The application of RO in the mining industry
could be improved by incorporating these uncertainties. However, modelling those
uncertainties is challenging. Therefore, we proposed a modelling method to be applied in the

3
Akita University

mining industry. The model can be applied to other mining with adaptation to assumptions
relating to mineral properties, mining methods and global economic conditions. In addition, CS
methodology should increase RO application, particularly in analyzing geological uncertainties
in a multi-stage of a short-term mining project. Multi-stage terminology refers to the total
amount of mining locations as more than one. Our model covered resource estimation through
the geostatistics method to mine plan on a monthly basis and economic valuation of two areas
simultaneously. The option type is the waiting to operate option based on the data availability.
The company has the right to operate, wait for the uncertainty to be resolved, or abandon the
mining projects.

Our established valuation approach addressed limitations in RO applications,

particularly concerning assessing several sources of uncertainty in a multi-stage short-term
mining project. The application exhibited practical simplicity as well as applicability. We
enhanced previous studies (Adachi et al., 2008) by employing the adaptation of spatial
uncertainty through conditional simulation (CS) and the construct path generation algorithm to
incorporate the three uncertainties. Closely related studies demonstrated the same using
empirical models (Brennan & Schwartz, 1985; Cortazar et al., 2001). However, we utilized the
simulation method (Barraquand & Martineau, 1995) and the uncertainty drivers. The
methodology incorporated the economic impact of the reserve, pricing, and cost uncertainties.
However, other modifying factors such as political, legal, social, and environmental are
assumed to be stable.

1.2. Problem Statement

Metal prices have been fluctuating remarkably in recent years. Fluctuation of the price
certainly affects mining company value. On the other hand, total available reserves are always
uncertain until the deposit is depleted. The company should optimize the cost and obtain
exploration data to estimate the reserves. Moreover, the history of mine operational cost
indicated significant volatility that gives the most considerable risk to the tin mining company.
These three uncertainties are calculated in this research through real options methodology,
particularly Multi-stage SSA.

4
Akita University

In this study, there are five main problems regarding measuring project value. As a case
study, we demonstrated the case of the Indonesian tin company, PT Timah Tbk. They are as
follows:

1. How do a mining company quantify geological uncertainty and use it to calculate project
value?
2. How can a mining project value be calculated simultaneously considering economic,
technical, and geological uncertainties?
3. How to evaluate the multi-project mining using RO methodology?
4. What is the advantage and disadvantage of the simulation method compared with Black-
Scholes and lattice valuation?
5. What are the practical consequences of risk and uncertainty caused by price, cost, and
reserve in the mining industry?

1.3. Purpose of Study

The mining industry faced uncertainties, particularly price fluctuation, the geological
presence of the reserves, and technical-cost uncertainty. A developed evaluation method of a
project is required considering the three most significant problems in mining. In addition,
decision-making in a mining company practically needs coordination of geological, the finance
and the mining engineering department. The finance department urges to decide the cheapest
operating cost at the highest price. In contrast, the geological and mining engineering
department wants to minimize the uncertainty of the existence of the ore and easy operation
(Crundwell, 2009).

This research tried to meet their need using geostatistics and real options methodology.
It resolved the price fluctuation that directly affected the project value and assessed the
exploration step to account for grade volatility. Moreover, the fluctuation of the historical cost
was measured through statistical method. The proposed method calculated the consequence of
uncertainties to the project value. The assessment was conducted in a monthly time frame to
describe how the real option worked.

We proposed an enhanced evaluation SSA method of mining projects considering their

inherent uncertainties. The price and cost uncertainty was analyzed using its historical data with
Geometric Brownian Model and Mean Reverting Process. On the other hand, geological

5
Akita University

uncertainty was explained through conditional simulation. Furthermore, we combined those

uncertainties with the utilized SSA approach. The improvement of applicability of real options
valuation and the utilization of conditional simulation to capture geological uncertainty was
expected.

1.4. Significance of Study

To date, researchers developed methods to measure and mitigate the risk embedded in
a business. They made parameters to calculate the return of a project and the inherited risk.
Principally, a higher risk of a project should give the investor the best return. Furthermore, the
most popular method was discounted cash flow analysis, which decided to either accept or
decline to start. However, it had assumption that the risk would be constant during the project
life. The developed risk assessment methodology, namely real options analysis, was adapted
from financial option theory and practiced in real projects. It figured out the risk based on the
potential growth of the project and the probability embedded.

The originality of proposed SSA methodology exhibits practical simplicity and

enhances previous studies (Adachi et al., 2008) by employing the adaptation of spatial
uncertainty through conditional simulation (CS) and the construction path generation algorithm
to incorporate price and grade uncertainties. Closely related studies demonstrated using
empirical models (Brennan & Schwartz, 1985; Cortazar et al., 2001). In contrast, we utilized
the simulation method (Barraquand & Martineau, 1995) to assess the uncertainty drivers, which
are more straightforward and flexible. Researchers concluded that the methodology is growing
in metal mining due to increased uncertainty of markets and the complexity of new projects
(Savolainen, 2016).

The uncertainties are assessed in short-term mining design or monthly mine planning
incorporating economic, technology, and geological uncertainties. Additionally, the utilization
of conditional simulation, which is typically an estimation tool, was first performed to assess
grade uncertainty which would be the significance of this study. By examining the research, the
project's decision-making was resulted in optimizing the project value. The methodology was
a pilot demonstration to improve the applicability of RO methodology by capturing several
uncertainties in a single evaluation. We developed the path generation in the MSSA method
and applied specific model in uncertainties measurement. Nevertheless, technical issues such
as geotechnical, weather, recovery, and other modifying factors (JORC, 2012) could be

6
Akita University

calculated in the future using RO Methodology, particularly simulation method with MSSA
approach.

1.5. Outline of Study

The research on "Combining Economic, Geological, and Technical Uncertainties in

Mining Projects Valuation Using Real Options Analysis" consisted of 5 chapters, including
introduction (Chapter 1) and conclusion (Chapter 5).

Chapter 1 introduced the background of this research and set the purposes we intend to
achieve. A brief explanation of recent studies regarding project evaluation, notably RO
Analysis, and the gap of recent research is described in the chapter. In addition, we explained
the significance of the studies that would be the originality and contribution to current research.

Chapter 2 consisted of explaining the case study using the typical traditional evaluation
method. It covered exploration, mine planning, and project evaluation. Moreover, it contained
collected data and how the project valuation was carried out. This chapter also assessed the
positive and negative points of current practice. The drawback of the traditional method relied
on a lack of uncertainty consideration in the evaluation. Brief analysis, namely sensitivity
analysis, is conducted which has been standard practice in the industry.

The main focus of the research, Real Options Analysis, was described in Chapter 3.
After summarizing previous literature regarding real options analysis, we calculated the project
value of our case study. The recent three approaches are explained and demonstrated in this
chapter. Real options analysis was conducted to get the optimal decision of the monthly mine
planning whether go mine, get more data by doing next exploration, or waiting for more
information regarding the price change and cost prediction.

While in Chapter 4, we focused on the simulation method represented by the Multistage

Stratified State Aggregation approach. The chapter started with modeling uncertainties deal
with the mining industry in general. Furthermore, we detailed the economic, technical, and
geological uncertainties in our case study. This methodology could be adapted for another
mining location with adjustments based on the mineral, reserve property, mining method, and
the corporate historical data regarding the cost of mining. It was developed to combine those
three uncertainties and simulate them through java programming as a tool. Compared to the
other two methods, this new approach's drawbacks and benefits are analyzed. The chapter
7
Akita University

analyzed the result of project value through traditional DCF and real options using the MSSA
method. It contained a decision map and further analysis when the cost uncertainty was
eliminated. This assumption was based on the ability of the company to control the expense in
their project.

Chapter 5 covered the conclusion based on all analyses in our research.

8
Akita University

1.6. References

Adachi, T., Mogi, G., & Adachi, T. (2008). Real Option Analysis of Multi-Stage Investment
on Resources Development Project by Improved SSA Method. Mining and Materials
Processing Institute of Japan, 124(9), 576–582.

Ajak, A. D., & Topal, E. (2015). Real option in action: An example of flexible decision making
at a mine operational level. Resources Policy, 45, 109–120.
https://doi.org/10.1016/j.resourpol.2015.04.001

Barraquand, J., & Martineau, D. (1995). Numerical Valuation of High Dimensional

Multivariate American Securities. The Journal of Financial and Quantitative Analysis,
30(3), 383–405. https://doi.org/https://doi.org/10.2307/2331347

Black, F., & Scholes, M. (1973). The pricing of options and corporate. The Journal of Political
Economy, 81(3), 637–654.

Brennan, M. J., & Schwartz, E. S. (1985). Evaluating Natural Resource Investments. The
Journal of Business, 58(2), 135. https://doi.org/10.1086/296288

Cortazar, G., Schwartz, E. S., & Casassus, J. (2001). Optimal exploration investments under
price and geological-technical uncertainty: A real options model. R and D Management,
31(2), 181–189. https://doi.org/10.1111/1467-9310.00208

Cox, J. C., Ross, S. A., & Rubinstein, M. (1979). Option pricing: A simplified approach. Journal
of Financial Economics, 7(3), 229–263. https://doi.org/10.1016/0304-405X(79)90015-1

Dowd, P. A. (1994). Risk assessment in reserve estimation and open-pit planning. Transactions
- Institution of Mining & Metallurgy, Section A, 103. https://doi.org/10.1016/0148-
9062(95)97056-o

Haque, M. A., Topal, E., & Lilford, E. (2014). A numerical study for a mining project using
real options valuation under commodity price uncertainty. Resources Policy, 39(1), 115–
123. https://doi.org/10.1016/j.resourpol.2013.12.004

Mun, J. (2006). Real options analysis: tools and techniques for valuing strategic investments
and decisions. In John Wiley & Sons, Inc (2nd ed.). John Wiley & Sons, Inc.

Nicholas, G. (2014). An Integrated Risk Evaluation Model for Mineral Deposits [The
University of Adelaide]. https://hdl.handle.net/2440/86234

9
Akita University

Savolainen, J. (2016). Real options in metal mining project valuation: Review of literature.
Resources Policy, 50, 49–65. https://doi.org/10.1016/j.resourpol.2016.08.007

JORC. (2012). Australasian Code for Reporting of Exploration Results, Mineral Resources and
Ore Reserves (The JORC Code).

10
Akita University

Chapter 2 Project Evaluation

2.1. Exploration

Project evaluation is a process of establishing the project's economic feasibility that

requires a capital investment and making an investment decision. The industry whose revenue
relied on ore selling is termed the resources industry. Ore is a metalliferous mineral, or an
aggregate of metalliferous minerals, mixed with gangue which can be mined or processed at a
profit (Hustrulid et al., 2013). Mineral and mining projects commonly experience unique
evaluation hurdles. They are the difficulty of estimating reserves, catastrophic issues
forecasting commodity prices and production costs, lengthy evaluation durations during
production, unsure regulatory and environmental expenses, and, in most cases, the long project
lives. Economic and technological circumstances can significantly alter the project value. The
project's evaluation of technical, financial, social, environmental, and political components of
the ecosystem must be determined before executing. They overlap to some level, but they form
the economic foundation for an evaluation when taken together. In addition, every mine
resource is unique; and tough to quantify the industry economics. Nevertheless, information is
costly, especially in the exploration stage, which has no guarantee of success.

In general, the levels of feasibility progress in order from conceptual (pre-feasibility) to

preliminary, intermediate, and final. To determine stages of feasibility, various firms employ
different terminologies. Analysts determine whether to progress to the next stage and,
ultimately, construct after each stage based on economics, environmental factors, and market
timing. The first step of mining exploration is conceptual study, an initial assessment of a
mining project. Flowsheet development, cost calculation, and production scheduling rely on
limited test work and engineering design. Exploration data from drilling and sampling have to
be enough to describe a resource accurately. This level of work is important for determining
future engineering inputs and necessary investigations (Dyas, 2002).

A resource is a group of naturally existing materials in a concentrated form. It can be

profitably extracted now or in the future. Geologic evidence is employed to determine the
materials' location, grade, quality, and quantity. There are three types of resources which are

11
Akita University

measured, indicated, and inferred. Furthermore, the reserve is a fraction of a resource that meets
minimum product specifications and may be reasonably expected to be profitably and
technically produced at the time of determination. It is classified into two categories which are
proven and probable (Hustrulid et al., 2013). As a result, the exploration stage, namely the
feasibility study, has a critical role in transforming resources into reserves.

Predicting rock characteristics at unknown locations and anticipating the future flow
behavior of complicated geological and engineering systems are the main problematic in
resource exploration. Addressing this complex situation requires a variety of theories and
assumptions (Pyrcz & Deutsch, 2014). A presumed orebody is a starting point for examination
in exploration. Tonnage, grades, elevation, and physical address are used to define the potential
deposit. Additionally, exploring a known orebody is carried out to discover new reserves and
better understand current deposits to assure their dependability. Additional reserves almost
always result in increased value straightforwardly (Runge, 1998). The last exploration step,
feasibility study, provides technical, environmental, and commercial base for an investment
decision. Iterative processes to optimize all critical elements of the project will be a procedure
in feasibility study. The ultimate aim of the stage is identifying the production capacity,
technology, investment and production costs, sales revenues, and return on investment.

Two standards of importance in the exploration can be defined for mines, i.e. (1) a
minimum ore reserve equal to that required for all the years that the cash flows are projected in
the feasibility report must be known with accuracy and confidence and (2) an ultimate tonnage
potential, projected generously and optimistically, should be calculated to define the area
adversely affected by the mining and within which dumps and plant buildings must not
encroach (Hustrulid et al., 2013). Those expected exploration results can be generated after
conducting drill hole surveys, sampling, splitting, assaying, logging drill holes, plotting, cross-
sectioning, drill hole spacing, cutoff grade estimation, geostatistical analysis, and finally
making reserve categories.

This research performed a project evaluation from exploration to economic decision

recommendation. A case study was conducted at an underwater tin mine project in Indonesia
operated by PT Timah, Tbk, as a mining company. Exploration of the company had been
conducted since around 1970 and the available data was part of their concession. The
exploration data existed at two locations, which later we termed as mid and south reserves.
Exploration data consisted of drill collar and drill assay. The drill hole data amount was 55

12
Akita University

holes with a total depth of 1459.6 meters. They were ranged from 4.8 meters under sea level
(MUSL) to 47.3 MUSL. Additionally, tin content data were ranged from 0.008 kg/m3 to 1.599
kg/m3. The company have right to extract as well as abandon the resources. This research
evaluated further the resource with its inherited uncertainties. The evaluation was started with
modeling the resources using GEOVIA Surpac 6.8.1. After plotting the drill hole data, we
composited it with its statistic (Figure 2.1) and variogram modeling (Figure 2.2) using the
geostatistical methodology.

Statistics is the collection, organization, and interpretation of data and the conclusion
and decision-making process. While geostatistics is a subfield of applied statistics. It
encompasses the following concepts: spatial context (geological), spatial connections,
volumetric support/scale, and uncertainty. The branch of knowledge was first developed by
Krige (1951) and followed by Matheron (1963), which generated a statistical framework to
resolve the relationship between vein width, the efficiency of exploration data in terms of
sampling, measuring, and assaying. Geostatistics consists of three main concepts: regionalized
variables, random function, and random variables. The regionalized variable is a function that
takes a value at every point in the space of regionalization. However, the function varies
irregularly from point to point in the space of regionalization. The random function is the set of
random variables at all possible locations. The unique outcome that exists at every location
realizes the random function. The grade may be considered a random variable that assumes a
series of outcome values at unsampled locations. The series of possible outcome values for the
random variable at each location is determined by a probability distribution (Nicholas, 2014).

Figure 2.1 Basic Statistic Data

13
Akita University

Variogram model or geostatistical semi variance is a function of difference over distance

and is dissimilarity measurement. It is the half variance of grade differences for each drill hole
pair with a certain direction mathematically expressed by Equation 2.1. The result of variance,
𝛾 (ℎ), will be more significant as the increase of distance (h). Nugget variance exists, which
indicates the variance of the combined distribution of pairs of points separated by an infinitely
small distance (h value almost 0). Typically, variogram experiments produced an asymptotic
behavior termed sill. The distance when sill is reached is called range. This range is later used
to define the recommended minimal distance of core drilling to get optimum data with a certain
cost. Based on the model (Figure 2.2), the drill hole data has a range of influence of 30 m with
sill 1.2 and nugget effect 0.3.

𝑛(ℎ)
1 Equation 2.1
𝛾 (ℎ ) = ∑[𝑍(𝑥𝑖 ) − 𝑍(𝑥𝑖+ℎ )]2
2𝑛(ℎ)
𝑖=1

Where: n(h) = the number of couples separated by distance h,

h = distance between sample pairs,
Z(xi) = value of sample at xi,
Z(xi+h) = value of sample at xi+h.

Figure 2.2 Variogram Model

Estimation is process of determining the best single value for a spatial attribute in an
unsampled area. Kriging is a phrase coined by G. Matheron in 1963 in tribute to Danie Krige.
It produces an interpolation function based on a covariance or variogram model derived from
the data. Kriging method is defined as optimal regression of observed Z values of surrounding

14
Akita University

(real) data and weighted according to covariance values from semivariogram results. Local
scale of kriging model is accurate and it takes local priority over global geographical variability.
The value of estimation can be formulated as Equation 2.2 for simple kriging type.
𝑛

𝑍 ∗ = ∑ 𝜆𝑖 𝑍𝑖 Equation 2.2
𝑖=1

Where Z* = Estimated block value

λi = unknown weight for value at i,
Zi = measured value at i.
By contrast, simulation is the process of achieving one or more acceptable values for a
reservoir attribute at an unsampled location. The idea of conditional simulations is to build a
representation of the phenomenon that is consistent with the data observed sampling interval,
as kriging is, and yet reproduces the local fluctuations. The procedure of globally accurate and
consistent with global data are used. Conditional simulation generates value at sampled points
and produces the same dispersion characteristics of the original data set or mean, variance, and
covariance or semivariogram. The conditional simulation generates sets of realization that each
reproduces histogram, spatial variability, and known data of a variable. Each of them is
independent and equal to be drawn from referred set. The obtained block model will be
evaluated and compared with the results.

Inputting parameters of compositing data and variogram model resulted kriging model
(Figure 2.3, and 2.4) and conditional simulation model (Figure 2.5 and 2.6). The resource
estimation through Kriging and Conditional simulation method is described at Appendix A and
summarized at Table 2.1.

Table 2.1 Resource Summary

Location Method Tonnage Grade (%)

Mid Kriging 4,824,000 0.013
Mid Conditional Simulation 11,503,500 0.016
South Kriging 728,250 0.010
South Conditional Simulation 1,205,250 0.019

15
Akita University

Figure 2.3 Ordinary Kriging Model for South Reserve

Figure 2.4 Ordinary Kriging Model for Mid Reserve

16
Akita University

Figure 2.5 Conditional Simulation Model for South Reserve

Figure 2.6 Conditional Simulation Model for Mid Reserve

17
Akita University

2.2. Mine Planning

Mine planning is defined as a systematic process set out in a three-dimensional

flowchart in Figure 2.7. The process consists of a similar series of steps in each phase, each
undertaken in the same order. Three such phases are illustrated in Figure 2.7. However, in
practice, there may be any number of phases. The same series of steps are undertaken in varying
amounts of detail, depending on the precision, economic action, or decision being sought
(Runge, 1998). The ultimate aim of mine planning is to get the value of the project, which is
feasible to be extracted practically. Once the reserve geology and quality are well understood,
then mine planning can begin concerning the following: (1) orientation of mine works; (2)
access to the reserve; (3) determination of opening widths; (4) selection of mine method; (5)
selection of development and secondary mining heights; (6) appropriate inter-burden
thicknesses; and (7) examining the stability of the mine (Newman et al., 2020).

Figure 2.7 Iterative Mine Planning Flow Chart

18
Akita University

In this research, monthly mine planning is conducted for underwater mining with dredge
as main equipment. We developed a mine planning simulation using GEOVIA Surpac 6.8.1
and utilized GEOVIA Whittle 4.6 mining software to optimize the mine design. Optimizer
software can reduce the iterative activity of mine planning. It employed Lerch - Grossman
algorithm to get the maximum NPV. The algorithm pit limit based on fixed slope angles
governed by block dimensions. The parameter we used to develop the mine plan is described
in Table 2.2. The cost parameter is obtained from averaging historical cost with similar mining
method and equipment. We simulated for two reserve locations using the Kriging model or
Conditional Simulation model (Table 2.3 and Appendix B). The design result of mine planning
is shown at Figure 2.8 and Figure 2.9.

Table 2.2 Mine Planning Parameters

Parameters Unit Value

Natural Slope Degree 45
Tin Selling Price USD/Ton 23,199
Mining Operation Cost USD/ Ton Ore 1.46
Mining Recovery 0.9
Mining Dilution 1.1
Processing Cost USD/kg/m3 1.15
Processing Recovery 0.9
Selling Cost USD/kg/m3 0.5

Table 2.3 Ore and Waste Mine Production

Location Volume Tones Grade (Kg/ M3) Tin (Ton) NPV (USD)
Ordinary Kriging Model
mid 748,000 2,244,000 0.36 269.28 1,147,183.00
south 172,000 516,000 0.17 28.896 -279,926.00
Conditional Simulation Model
mid 3,680,500 11,041,500 0.39 1,428.03 7,294,826.00
south 580,875 1,742,625 0.36 208.53 882,590.00

19
Akita University

Figure 2.8 Ultimate Pit Limit Design for Mid Reserve

20
Akita University

Figure 2.9 Ultimate Pit Limit Design for South Reserve

21
Akita University

2.3. Mining Economics

The Discounted Cash Flow (DCF) approach is one of the most frequent methods for
evaluating a mining project. For most mining prospects, DCF methodologies are used to make
investment decisions (Fig. 2.10). This strategy has been a popular tool for numerous decades
for completing appraisals and allocating scarce funds. It is based on cash flow and is simple to
comprehend. The DCF method assesses the entire project by modifying or discounting the
project net cash flow to account for risk and time. The bigger the investment risk, the higher
the discount rate. DCF approaches do not allow for management flexibility and it assures
parameters throughout the project. Decisions must be taken "now or never," and the use of an
appropriate discount rate is critical.

Figure 2.10 Evaluation Metric at Feasibility Stage

Source: Whittle D., et al., 2005

Evaluating a project is a multidisciplinary task and is not solely the domain of any
profession or management. It involves engineering economics, capital budgeting, financial
management, and strategic planning (Crundwell, 2008). The key concept of the resource

22
Akita University

industry is extraction leading to a profit. Therefore, several parameters in the economic

modeling of mining projects: capital expenditure, operating cost, and commodity price have to
be defined. The capital expenditure is the sum of money required to develop and install a
manufacturing facility. At the same time, operating costs are those costs that are incurred in the
direct manufacture of the items. The commodity price is from the world trade price represented
in London Metal Exchange. The cash flow model summarizes the money stream in and out of
a planned project. While the cost is resumed at Appendix D.

Well-known economic theory, the time value of money, explained that the current value
of money is bigger than the future and less than the past (Equation 2.3). There are three reasons
for this: inflation, risk, and liquidity (Crundwell, 2008). The Net Present Value (NPV), as
expressed in Equation 2.4. is the total of all cash flows discounted at present using the time
value of the money. The investor's value is increased when the net present value is positive. If
it is less than zero, the investor's interest is expected to be lost.

FV = PV(1 + 𝑟)𝑡 Equation 2.3

𝑛
𝐶𝐹𝑡
𝑁𝑃𝑉 = ∑ −𝐼 Equation 2.4
(1 + 𝑟)𝑡
𝑡=0

Where FV = future value

PV = present value
r = discount rate
t = number of period
NPV = Net Present Value
CF = cash flow
I = initial investment
The evaluation of cash flow in the traditional DCF method is simply the subtractions
revenues and costs. The revenues stream only comes from selling material with a certain selling
price. Therefore, revenue is governed by grade, throughput, recovery, and metal or product
price. Of these, price is: (a) by far the most difficult to estimate and (b) the one quantity largely
outside the estimator's control. Even ignoring inflation, selling prices are widely variable with
time. While for the costs can be divided into two big groups: capital investment and operating
cost. Generally, accuracy in capital or operating cost estimating goes back to accuracy in
quantities, reliable unit prices, and adequate provision for indirect or overhead items. Based on
the evolution of commodity prices, the mine operation can be altered (Rimélé et al., 2020).
23
Akita University

Traditional DCF techniques do not account for decisional flexibility in project capital
planning. DCF analysis fails to account for the value of genuine alternatives inherent in capital
budgeting, which explains why it does not consider the true worth of an investment (Trigeorgis,
1993). Misuses of DCF methods include discounting real flows with nominal interest rates,
failing to implement inflation correctly; applying unnecessary risk-adjustment factors, failing
125
120
115
Present Value (K USD)

110
105
100
95
90
85
80
1
3
5
7
9
11
13
15
17
19
21
23
25
27
29
31
33
35
37
39
41
43
45
47
49
51
53
55
57
59
Production Month

Net Cash FLow Discounted Cash Flow

Figure 2.11 Mid Monthly Cash Flow Model

17

16

15
Present Value (K USD)

14

13

12

11

10

8
1
3
5
7
9
11
13
15
17
19
21
23
25
27
29
31
33
35
37
39
41
43
45
47
49
51
53
55
57
59

Production Month

Net Cash Flow Discounted Cash Flow

Figure 2.12 South Monthly Cash Flow Model

24
Akita University

to recognize that risk diversification should not be considered; using unsuitable criteria in
measuring profitability, such as internal rate of return, which frequently yield erroneous
conclusions; and frequently gauging cash flows using standardized accounting principles.

In the case study, the evaluation was conducted in February 2021. Therefore, the used
price was flat on January 2021. While for the operating cost was resumed from the company
with a similar operation method and similar location. The overall operation did not cause initial
capital expenditure. The research evaluated the reserve that had been explored before. The
company had the right to execute or abandon the project. As we assumed the production was
flat in 5 years of operation, the cash flow and cash flow models are shown in Table 2.4, Table
2.5, Figure 2.11, and Figure 2.12. It resulted in Net Present Value of 6.6 MUSD and 0.9 MUSD
for mid and south reserve, respectively.

25
Akita University

Table 2.4 Cash Flow Model for Mid Reserve

Unit Total
Component Unit 1 2 3 … 58 59 60
value value
Production Ore 1000 ton 7,204 120 120 120 … 120 120 120
Grade kg/m3 0.594 0.594 0.594 0.594 … 0.594 0.594 0.594
Tin Got ton 1,426 23.77 23.77 23.77 … 23.77 23.77 23.77
Selling Price usd/ton 23,199 23,199 23,199 23,199 … 23,199 23,199 23,199
Gross Revenue 1000 usd 551.51 551.51 551.51 … 551.51 551.51 551.51
Mining Cost 12,785 213.08 213.08 213.08 … 213.08 213.08 213.08
Processing Cost 6,738 112.30 112.30 112.30 … 112.30 112.30 112.30
Operating Cost 1000 usd 19,523 325.38 325.38 325.38 … 325.38 325.38 325.38
Net Income Before
1000 usd 226.13 226.13 226.13 … 226.13 226.13 226.13
Tax
Tax 1000 usd 28% 63.32 63.32 63.32 … 63.32 63.32 63.32
Cash Flow 1000 usd 118.69 118.69 118.69 … 118.69 118.69 118.69
Discounted value 1000 usd 0.25% $118.40 $118.10 $117.81 … $102.69 $102.43 $102.18
Net Present Value = 6.605 Million USD

26
Akita University

Table 2.5 Cash Flow Model for South Reserve

Unit Total
Component Unit 1 2 3 … 58 59 60
value value
Production Ore 1000 ton 1,123 18.72 18.72 18.72 … 18.72 18.72 18.72
Grade kg/m3 0.558 0.558 0.558 0.558 … 0.558 0.558 0.558
Tin Got ton 208.88 3.48 3.48 3.48 … 3.48 3.48 3.48
Selling Price usd/ton 23,199 23,199 23,199 23,199 … 23,199 23,199 23,199
Gross Revenue 1000 usd 80.76 80.76 80.76 … 80.76 80.76 80.76
Mining Cost 1.77 1,993.08 33.22 33.22 33.22 … 33.22 33.22 33.22
Processing Cost 0.87 986.80 16.45 16.45 16.45 … 16.45 16.45 16.45
Operating Cost 1000 usd 3.35 2,979.87 49.66 49.66 49.66 … 49.66 49.66 49.66
Net Income Before
1000 usd 31.10 31.10 31.10 … 31.10 31.10 31.10
Tax
Tax 1000 usd 28% 8.71 8.71 8.71 … 8.71 8.71 8.71
Cash Flow 1000 usd 16.32 16.32 16.32 … 16.32 16.32 16.32
Discounted value 1000 usd 0.25% $16.28 $16.24 $16.20 … $14.12 $14.09 $14.05
Net Present Value = 0.908 Million USD

27
Akita University

2.4. Sensitivity Analysis

The risks related to mining are diverse and complicated, with the orebody constituting
the primary source of risk. Mining is distinct from most other industries in that product
knowledge relies primarily on estimations that carry a degree of uncertainty. World commodity
prices heavily influence possible income fluctuations and, hence, the amount of the commercial
mineral inventory (Dominy, 2016).

Sensitivity analysis is the process of changing one or more elements to examine how
the variation affects the project's value. While sensitivity analysis helps to understand the
implications of uncertainty, it does not provide a project value adjusted for the uncertainty. One
of the most important benefits of sensitivity analysis is that it identifies the elements that have
the greatest impact on the economics of a project. The analysis enables assessors to collect more
information more effectively.

12,000

11,000

10,000
Net Present Value (1000 USD)

9,000

8,000

7,000

6,000

5,000

4,000

3,000

2,000
-20 -10 0 10 20
Relative Changing (%)

production price discount rate mining cost

processing cost total cost tax

Figure 2.13 Sensitivity Analysis

We conducted a sensitivity analysis on price, reserve, and cost variables on our case
study (Figure 2.13). Price and production were the most sensitive parameters, followed by total
cost, mining cost, processing cost, tax, and discount rate. However, because sensitivity analyses
are usually conducted on a single ‘best estimate’ of the resource model, it should reflect that
level of local inaccuracy and could be misleading decision-making. Therefore, combining more
28
Akita University

than one uncertainties evaluation is necessary and analysis based on simulation should be
conducted to anticipate an unpredictable situation in the future.
Sensitivity analysis can understand project risk straight forward. In addition, all
stochastic variables need to be considered in a single evaluation model with all possible
permutations (Nicholas, 2014). Practically, general mining companies conduct sensitivity
analysis to measure the effect of parameter changing. It runs analyses that reflect various
commodity prices, reserve amount, and operating and capital costs to determine the effect of
such variations. In very simplistic terms, sensitivity analysis is 'what if' analysis, which is an
important notion at the core of any application of decision tools and may be applied to a wide
range of uses. One-way sensitivity analysis is the simplest form of sensitivity analysis where a
given amount varies one value in the model to examine the impact of the change on the model's
results (Taylor, 2009).

29
Akita University

2.5. References

Runge, I. C. (1998). Mining Economics and Strategy. Society for Mining, Metallurgy, and
Exploration, Inc.

Dyas. K. E. (2002). Project Evaluation. In Mining Reference Handbook (pp. 1–9). Society for
Mining, Metallurgy, and Exploration, Inc.

Hustrulid, W., Kuchta, M., & Martin, R. K. (2013). Open Pit Mine Planning & Design (3rd
ed.). CRC Press.

Stermole, F. J., & Stermole, J. M. (2000). Economic Evaluation and Investment Decision
Methods (10th ed.). Investment Evaluations Corporation.

Krige, D. G. (1951). A Statistical Approach to some Basic Mine Valuation Problems on the
Withwatersrand. J. South. Afr. Inst. Min. Metall., 52(6), 119–139.

Matheron, G. (1963). Principles of Geostatistics. Economic Geology, 58, 1246–1266.

Crundwell, F. (2008). Finance for Engineers: Evaluation and Funding of Capital Projects.
Springer-Verlag London Limited.

Rimélé, A., Dimitrakopoulos, R., & Gamache, M. (2020). A dynamic stochastic programming
approach for open-pit mine planning with geological and commodity price uncertainty.
Resources Policy, 65(December 2019). https://doi.org/10.1016/j.resourpol.2019.101570

Newman, C., Newman, D., & Dupuy, R. (2020). Development of a multiple level underground
limestone mine from geology through mine planning. International Journal of Mining
Science and Technology, 30(1), 63–67. https://doi.org/10.1016/j.ijmst.2019.12.007

Trigeorgis, L. (1993). The Nature of Option Interactions and the Valuation of Investments with
Multiple Real Options Author. Journal of Financial and Quantitative Analysis, 28(1), 1–
20.

Whittle, D., Whittle, J., Wharton, C., & Hall, G. (2005). Strategic Mine Planning (Issue
January). Gemcom Software International Inc.

30
Akita University

Chapter 3 Real Option Analysis

3.1. Real Option Analysis Framework

An option is a financial derivative whose value depends on the price of a stock. Option
holders have rights but no obligation to execute their option in the maturity date. These
alternatives are quantified by option calculation to justify option value. The adaptation of option
calculation in the actual project is called real options (RO). For instance, if a company decides
to invest in a mine, they exercise right. They lose the right, but they earn the asset. Discounted
Cash Flow (DCF) method theoretically assumed static decision over project life time without
ability to change decision. In contrast, RO assumed dynamic future decisions when the
uncertainties resolved. That managerial flexibility is not accommodated in the traditional DCF
methodology. For instance, the operation of a project can be contracted, delayed, or abandoned
and real options discount uncertain cash flows at the correct rate (Crundwell, 2008).

The analysis of options was introduced by Fisher Black and Myron Scholes to analyze
a derivative market product (Black and Scholes, 1973). It had been popular with the lattice
method, which was developed further. The real options theory was largely accepted in the
financial literature, and many scholars sought to alter it (Table 4.1). Cox, Ross, and Rubinstein
developed a basic economic concept of option pricing under the non-arbitrage assumption in
1979 and a simple and efficient numerical approach for valuing options (Cox et al., 1979).
Myers (1977) stated real option first time in the study of corporate debt policies, which account
for the flexibility of the company management. Mun (2006) added that a real option is a
systematic approach and integrated solution using financial theory, economic analysis,
management, decision science, statistic, and econometrics to evaluate real asset dynamic and
uncertain.

Brennan and Schwartz (1985) used an option pricing model to analyze a natural resource
project for the first time. Natural resources, they said, contain much unpredictability in terms
of resources and cost. According to the Brennan and Schwartz model, continuous-time non-
arbitrage methods and stochastic control theory may be used to assess such enterprises and find
31
Akita University

the best strategies for creating, managing, and abandoning them. Trigeorgis offered examples
of applied actual options valuation (Trigeorgis, 1993). They demonstrated how to postpone,
enlarge, and shut a variety of investment project decisions, particularly in the natural resource
sector. They indicated that commodity price changes have a major impact on natural resource
investment decisions. All prior empirical research on ROV and DCF in mining operations is
included in Table 3.1. This thesis employed ROV in a mining project with a combination of
uncertainties which will be the originality of the research.

Table 3.1 Empirical Example of the ROV and DCF for Commodities

No Year Author (s) Method Commodity Mine/project name Project location

1 1985 M J Brennan, E S Schwartz ROV Copper Hypothetical Not available
2 1985 M J Brennan, E S Schwartz ROV Gold Hypothetical Not available
3 1986 S K Palm, N D Pearson DCF.ROV Copper Not available Not available
J L Paddock, D R Siegel, F L
4 1988 DCF.ROV Oil Gulf of Mexico USA
Smith
5 1992 N Kulatilaka, A J Marcus DCF.ROV Oil, gas Not available Not available
6 1992 B Cavender DCF.ROV Gold Hypothetical USA
7 1993 D G Laughton, H D Jacoby DCF.ROV Oil Not available Not available
8 1993 J L Mardones DCF.ROV Copper Not available Chile
9 1993 E Pickles, J L Smith DCF.ROV Oil Not available USA
10 1994 N Kulatilaka, L Trigeorgis DCF.ROV Oil Hypothetical Not available
11 1996 D G Laughton DCF.ROV Copper Not available Not available
12 1996 M Samis, R Poulin DCF.ROV Gold Not available Not available
13 1997 S Frimpong J Whiting DCF.ROV Copper Confidential Not available
14 1998 G Salahor DCF.ROV Gas Not available Not available
15 1998 M Samis, R Poulin DCF.ROV Copper Not available Not available
16 1998 G Cortazar, J Whiting DCF.ROV Copper Not available Not available
Underground
17 1998 W S Dunbar, S Dessureault ROV Not available Not available
mine
18 1998 W S Dunbar, S D, M Scoble ROV Gold Not available Not available
19 1998 F S Sagi. EE Hiob. S Jones ROV Copper Not available Not available
Papua New
20 1998 Skelly FE Smith, K F McCardle DCF.ROV Gold Lihir Island
Guinea
21 1999 J E Smith, K F McCarkle DCF.ROV Oil, gas Hypothetical Not available
22 1999 F P Camus, C W Pelley ROV Copper Not available Not available
23 2000 R T McKnight DCF.ROV Copper Not available Not available
24 2001 M E Slade DCF.ROV Copper 21 mines Canada
25 2001 S Faiz DCF.ROV Oil Chevron Texaco USA
Olkusz
26 2002 J A Drieza, J Kicki, P Saluga DCF.ROV Zinc, lead Poland
Pomorzany
27 2002 A Moel, P Tufano ROV Gold 285 mines North America
D Colwell, T Henker, J Ho, K
28 2003 DCF.ROV Gold 27 companies Australia
Fong
W bailey, A B, S Faiz, Srinivasan,
29 2004 ROV Gas Elba Island Georgia
H Weeks
30 2004 S Kelly DCF.ROV Gold 41 mines Australia

32
Akita University

31 2005 V Blais, R Poulin, M Samis DCF.ROV Copper, gold Not available Canada
32 2006 M Samis, G A Davis, D Laughton DCF.ROV Copper Not available Not available
33 2007 J Hall, S Nicholls DCF.ROV Coal Hypothetical Not available
S Dessureault, V N Kazakidis, Z Sudbury, Western
34 2007 DCF.ROV Nickel, copper Canada, USA
Mayer Arizona
35 2007 P Guj, R Garzon DCF.ROV Nickle Not available Not available
G Dogbe, S Frimpong, J
36 2007 DCF.ROV Copper Hypothetical Not available
Szymanski
37 2008 S Shafiee, E Topal DCF.ROV Gold Hypothetical Australia
38 2009 Li Shu-xing, Knights Peter DCF ROV Not available Hypothetical Not available
39 2009 S.A. Abdel Sabour and G. Wood DCF ROV Copper, Gold Not available Canada
Akbari Afshin Dehkharghani,
40 2009 Osanloo Morteza, Shirazi Mohsen DCF ROV Copper Not available Not available
Akbarpour
S. A. Abdel Sabour and R.
41 2011 DCF ROV Copper Not available Canada
Dimitrakopoulos
Hesam Dehghani, Majid Ataee-
42 2014 DCF ROV Gold, Copper Grasberg Indonesia
pour, Akbar Esfahanipour
Snehamoy Chatterjee, Manas
43 2015 Ranjan Sethi, Mohammad Waqar DCF ROV Iron Not available India
Ali Asad
44 2015 Ajak Duany Ajak, Erkan Topal ROV Iron Not available Australia
45 2017 Jyrki Savolainen ROV Not available 170 mines Not available
Oscar Miranda, Luiz E. Brandao,
46 2017 DCF ROV Silver Not available Peru
Juan Lazo Lazo
Renewable
47 2018 Xiaoran Liu and Ehud I. Ronn ROV Not available China
Energy
48 2019 Aldin Ardian and Mustafa Kumral DCF ROV Gold Not available Canada, Indonesia
Shiwei Yu, Zhenxi Li, Yi-Ming
49 2019 DCF ROV Geothermal Xiong New Area China
Wei, Lancui Liu
Giorgio Locatelli, Mauro Mancini,
50 2020 DCF ROV Energy Not available Not available
Giovanni Lotti

Result of a project or underlying asset will remain uncertain until the end of the project
period. Therefore, the owner of the right does not know whether their choice will be profitable
at the option's maturity. The strike cost for real option is making the production facility to start
the project. Real option valuation parameter is resumed at Table 3.2. Option value can be
measured in various ways, such as path-dependent simulation, closed-form formulas, partial
differential equations, and lattice methods. Real option concept based on the manager's
flexibility to take action when information is obtained. The method implies that the manager
should defer or abandon the mine project when the project will not profitable respecting
uncertainty model.

The option evaluation in the mining project is divided into each mining stage. It
contained exploration, development, extraction, and abandonment. For instance, delay option
type is put option with the time maturity from acquisition to closing. The other option type is
the call option with the mature period, which is acquisition to the next stages. In the extraction
stage, the partition came to mine design, mine planning, temporary close, expansion, defer

33
Akita University

building, and switch option. Every type of options analysis has its particular framework, as
calculating option value is various. In this thesis, the calculation of option is focus on the wait
to operate option which the manager has right to delay operation considering uncertainty of the
project.

Figure 3.1 Result of option analysis

(Crundwell, 2008)

34
Akita University

Table 3.2 Option Parameters

Financial Option
Real Option Parameter Definition
Parameter
Strike Price Capital Cost The price to get the option
Maturity Project period Length of the option
Underlying Asset Present Value The asset value
Call Option Invest option Right to buy an asset
Put Option Abandon option Right to sell an asset
Long Position Investor Owner of an option
Short Position - The writer who buy the option
The option can only be exercised at
European Option -
maturity
The option can be exercised at any time
American Option American Option
before the option expires
Cortazar has provided a case study of real options in mining exploration which
accounted for geological and economical uncertainties (Cortazar et al., 2001). In line with this
research, there were three sources of uncertainties: geological, technical and price uncertainties.
Recently, the evidence of uncertainty in mining project valuation has been highlighted. Their
sources are recovery, grade, commodity price, discount rate, and mining cost. They drive the
successes of mining projects when the company is accurate in predicting. However, simulating
multiple uncertainty parameters can overcome the difficulty of prediction (Ardian & Kumral,
2020). Theoretically, the ultimate result of the option value can be described in Figure 3.1. It
consist of option value, strike price, Net Present Value and the option premium. The asset value
will be varied as uncertainty parameter is changed.

Inputs of cash flow are sources of uncertainty. Technical parameters refer to production
uncertainty which consists of the reserve, mining, and metallurgical uncertainty. While selling
price, capital cost, and variable cost lead to economic uncertainty. Real option foundation relies
on the flexibility of management to take action when uncertainty is resolved in the future time.

35
Akita University

3.2. Black Scholes (BS) Model

Black and Scholes model employed the capital asset pricing model to derive a link
between the market's necessary return on the option and the required return on the stock. Black
Scholes method is based on the potential underlying asset, and the cost expensed during the
project (Hull, 2015).

The assumption of the approach is:

1. The stock price follows the Geometric Brownian Model process with growth (µ) and
volatility (σ) constant (Equation 3.1),
2. The short-selling of securities with full use of proceeds is permitted,
3. There are no transaction costs or taxes. All securities are perfectly divisible,
4. No dividends during the life of the product,
5. There are no riskless arbitrage opportunities,
6. Security trading is continuous, and
7. The risk-free rate, r, is constant and identical for all periods.

Calculating Black-Scholes Formula needs a comprehensive understanding of the

volatility of an asset. The volatility, σ, is a measure of the uncertainty regarding an asset's
returns. It described as the standard deviation of the return generated by the stock in 1 year
when the return is stated using continuous compounding.
𝛿𝑆
= 𝜇 (𝛿𝑡) + 𝜎𝜀√𝛿𝑡 Equation 3.1
𝑆

where:
𝛿𝑆
= Percent change of variable
𝑆

𝜇(𝛿𝑡) = deterministic part of the change

𝜎𝜀√𝛿𝑡 = stochastic part of the movement
𝜇 = drift term or growth factor
𝜎 = volatility parameter
𝛿𝑡 = time-steps
𝜀 = simulated parameter
Black-Scholes formula combined underlying asset with normal distribution to calculate
the option value (Equation 3.2).

36
Akita University

𝑐 = 𝑆0 𝑁(𝑑1 ) − 𝐾𝑒 −𝑟𝑇 𝑁(𝑑2 ) Equation 3.2

Where:

𝑆 𝜎2
ln ( 𝐾0 ) + (𝑟 + 2 ) 𝑇 Equation 3.3
𝑑1 =
𝜎 √𝑇
𝑆 𝜎2 Equation 3.4
ln ( 𝐾0 ) + (𝑟 − 2 ) 𝑇
𝑑2 = = 𝑑1 − 𝜎√𝑇
𝜎 √𝑇
c : call option value
𝑆0 : underlying asset
K : strike price
𝜎 : asset volatility
r : continuously compounded risk free rate
T : time to maturity of the option
The normal distribution N(x) is a likelihood of a call option being executed in a risk-
free environment. It is the cumulative probability distribution function for a variable or the
probability of a variable will be less than a constant (x). In a risk-neutral universe, where stock
prices less than the strike price are treated as zero. Furthermore, the expression S0N(d1)ert is the
predicted stock price at time T. The strike price is only paid if the stock price is greater than K,
which has a chance of N(d2).
In the case study the option is defined as waiting to operate option. The company has
right to operate, delay and abandon the project when the project is unprofitable. Respecting the
multi uncertainty of cash flow, Black Scholes formula could not explain it. Under no investment
assumption (K=0), Equation 3.2 will be error. However, the uncertainty is existing over project
period. Therefore, the important of simulation method is to identify and assess the project value
and the uncertainty when empirical approach cannot resolve it.

We employed BS model to the project with adaptation on the assumption. We add

assumption of investment and focus only on the price uncertainty. The practice is done to be a
benchmark to our proposed model later. We assumed an expenditure 1 Million USD to be strike
price. Other assumption parameters are underlying asset of 7.514 MUSD, risk free rate of 0.25%
and maturity time of 60 months.

Then, we model the asset uncertainty by measuring the volatility of the asset. We
calculated the asset volatility as indicated by its standard deviation (Equation 3.5).

37
Akita University

1 𝑛
𝜎=√ ∑ (𝑢𝑖 − 𝑢̅)2 Equation 3.5
𝑛 − 1 𝑖=1

Where:
𝜎 : standard deviation
𝑃𝑖
𝑢𝑖 ∶ present value growth = ln ( )
𝑃𝑖−1
𝑢̅ : mean of 𝑢𝑖
n+1 : number of observation
The calculation conducted further by applying price into cash flow model periodically
(Figure 3.2). In Table 3.3, the calculated volatility of the project relating to price uncertainty
was 29% in a monthly time frame.

1
𝜎𝑝𝑟𝑜𝑗𝑒𝑐𝑡 = √ (5.11) = 0.29
60 − 1

38
Akita University

Table 3.3 Project Volatility Estimation Based on Historical Tin Price

Tin, LME US$/t Project NPV 2

Period Growth rate (𝑢𝑖) (𝑢𝑖 − 𝑢̅)2 ∑(𝑢𝑖 − 𝑢
̅)
(Pi) (MUS$) (Ci)

1/1/2016 14,906 $743.15 𝑢𝑖 = Ln(Pi/Pi-1) 5.11

2/1/2016 15,976 $1,395.45 0.63 0.35
3/1/2016 16,729 $1,854.50 0.28 0.06
4/1/2016 17,260 $2,178.21 0.16 0.02
5/1/2016 16,324 $1,607.60 (0.30) 0.12
6/1/2016 17,075 $2,065.43 0.25 0.05
7/1/2016 17,840 $2,531.79 0.20 0.03
8/1/2016 18,882 $3,167.03 0.22 0.04
9/1/2016 20,164 $3,948.57 0.50 0.21
10/1/2016 20,885 $4,388.11 0.11 0.00
11/1/2016 21,320 $4,653.30 0.06 0.00
12/1/2016 21,205 $4,583.19 (0.02) 0.00
… … … … …
… … … … …
… … … … …
1/1/2020 16,425 1,669.17 (0.24) 0.08
2/1/2020 16,267 1,572.85 (0.06) 0.01
3/1/2020 14,667 597.45 (0.97) 1.01
4/1/2020 15,274 967.49 0.48 0.20
5/1/2020 15,502 1,107.00 0.13 0.01
6/1/2020 16,819 1,909.36 0.55 0.26
7/1/2020 17,909 2,573.86 0.30 0.07
8/1/2020 17,550 2,355.00 (0.09) 0.02
9/1/2020 17,448 2,293.28 (0.03) 0.00
10/1/2020 17,724 2,461.08 0.07 0.00
11/1/2020 18,642 3,020.72 0.20 0.03
12/1/2020 20,544 4,180.53 0.49 0.21

39
Akita University

25,000 8,000
7,000

Project Value (K USD)

20,000
Tin Price ( USD/ Ton)

6,000

15,000 5,000
4,000
10,000 3,000
2,000
5,000
1,000
0 -
Jan-16 Jul-16 Feb-17 Aug-17 Mar-18 Sep-18 Apr-19 Nov-19 May-20 Dec-20
Period

Price Present Value

Figure 3.2 Price and Project Uncertainty

Thus, we input the parameter into formulation:

𝑐 = 𝑆0 𝑁(𝑑1 ) − 𝐾𝑒 −𝑟𝑇 𝑁(𝑑2 )

𝑐 = 7.514 · 𝑁(2.022) − 1 𝑒 −0.25% ·60 𝑁(−0.224)

c = 7.514 · 0.98 - 1 𝑒 −0.25% ·60 0.411
c = 6.997 MUSD
Where:

𝑆 𝜎2
ln ( 𝐾0 ) + (𝑟 + 2 ) 𝑇
𝑑1 =
𝜎 √𝑇
7.514 0.292
ln( )+(0.25%+ )·60
1 2
=
0.29√60

= 2.022
𝑆 𝜎2
ln ( 𝐾0 ) + (𝑟 − 2 ) 𝑇
𝑑2 =
𝜎 √𝑇
= 𝑑1 − 𝜎√𝑇
= 2.022 − 0.29√60
= -0.224

40
Akita University

3.3. Binomial Lattice (BL) Valuation

Lattice valuation was founded by Cox et al., (1979) which simplified Black Scholes
methodology. It started from the assumption that an asset follows a random walk. In each step,
the asset has a certain probability of increasing and decreasing by a certain percentage amount.
The approach followed a very important principle, namely risk-neutral valuation. When the
time of lattice simulation is limited to zero, the value is equal to the result of the Black Scholes
method. Recently, the lattice method has been popular in the academic research of real options
analysis (Crundwell, 2008). Johnathan Mun described that the initial lattice approach of real
option is the Brownian motion (Mun, 2006). The uncertainty in the future is explained by
simulation. A likelihood distribution may be built at an increasing period based on all the virtual
paths. The simulation pathways were developed using a Geometric Brownian Motion with set
volatility defined at Equation 3.1.

In an assumption of no-arbitrage, an asset with S0 price has an option with f value.

During the T date, the stock can either move up to S0u, where u >1, or down to S0d, where d <1.
In Figure 3.3, the option value will follow stock movement with fu and fd, respectively (Hull,
2015).

Figure 3.3 Stock and Option Value in One Period

The approach begins with measuring the growth of the project value using natural
logarithm (ln) to address asset’s volatility. The Brownian motion theory (Equation 3.1) can be
derived to the exponential continue form (Equation 3.6) to account for the continuity of the
option.

𝛿𝑆
= 𝑒 𝜇(𝛿𝑡)+𝜎𝜀√𝛿𝑡 Equation 3.6
𝑆
The upside factor (u) is when the simulated parameter generates a positive value or

𝑒 𝜎√𝛿𝑡 . At the same time, down factor (d) is obtained when a negative value is produced or

41
Akita University

𝑒 −𝜎√𝛿𝑡 . A manager should decide which maximizes the project value whether execution, delay
or abandonment, to get the option value. Executing the project resulted company should pay
the investing expenditure. In contrast, delaying and abandoning projects resulted the company
expense or producing nothing. Furthermore, the probability of each way is calculated using
risk-neutral probability. For instance, if the probability of upside is denoted as P, the expected
value of a project is expressed at Equation 3.7. However, the project is conducted over time.
Thus, the discount factor will be added in the exponential formula and generate Equation 3.8.

𝑓0 = 𝑃(𝑓𝑢𝑝 ) + (1 − 𝑃)(𝑓𝑑𝑜𝑤𝑛 ) Equation 3.7

𝑓0 = [𝑃(𝑓𝑢𝑝 ) + (1 − 𝑃)(𝑓𝑑𝑜𝑤𝑛 )]𝑒 −𝑟(𝑡) Equation 3.8

When assuming that the project is accepted when the initial value of asset (S0) is minimal
one. The probability can be defined as Equation 3.10.

𝑒 𝑟 (𝑡) − 𝑑 Equation 3.9

𝑃=
𝑢−𝑑
Finally, after generating Probability (P), up factor (u), and down factor (d), the option
value at present (f0) can be obtained with Equation 3.8 (Crundwell, 2008).

The Lattice method is a simplification of Black Scholes formulation. Similar with Black
Scholes formulation, the three uncertainty drivers cannot be accounted for by lattice valuation
methodology. They adapted from financial option which written on an asset with certain
exercise price with single uncertainty driver. The underlying asset for real option is the present
value of cash flow while strike price is the initial investment. However, when the methodology
applied to ongoing project which has no investment needed, they cannot calculate the
uncertainty although uncertainty is always existing.

We added the assumption as we did at Black Scholes approach to compare Binomial

Lattice (BL) method with other approach later. Then we applied the approach started calculated
the up and down factor of 1.33 and 0.74. Following stage is calculating the up (P) and down (1-
P) probability of 0.43 and 0.57. Finally, we made the tree model of the project value at Table
3.4.

42
Akita University

Table 3.4 Project Value

Period (month) 0 1 2 3 … 60
project value (KUSD) 7,514 10,042 13,420 17,935 … 270,762,143,780
5,622 7,514 10,042 … 151,599,282,030
4,207 5,622 … 84,880,190,381
3,148 … 47,524,279,948

Then, we subtract with the initial investment of 1 Million USD and we obtain Table 3.5.

Table 3.5 Project Value Minus Initial Investment

Period (month) 0 1 2 3 … 60
NPV (KUSD) 6,514 9,042 12,420 16,935 … 270,762,142,780
4,622 6,514 9,042 … 151,599,281,030
3,207 4,622 … 84,880,189,381
2,148 … 47,524,278,948

Last step is backward calculation by applying Equation 3.8 and we got Table 3.6.

Table 3.6 Backward Calculation

Period (month) 0 1 2 3 … 60
Option Value (KUSD) 6,961 9,479 12,811 17,282 … 270,762,142,780
5,074 6,984 9,464 … 151,599,281,030
3,641 5,127 … 84,880,189,381
2,526 … 47,524,278,948

3.4. Simulation Method

The simulation or numerical method is used when the analytic approach does not exist.
It follows a stochastic process to predict the future. A stochastic process is defined as a variable
whose value fluctuates unpredictably over time. Theoretically, stochastic modeling should
satisfy Markov properties, Wiener, and Ito process. Markov properties is a form of a stochastic
process explain that the present value is all that matters when forecasting the future. Since they
are uncertain, future predictions must be described in terms of probability distributions. The
Markov property states that the probability distribution of the price at any given future time is
independent of the past price's path. At the same time, Wiener Process is a particular type of
Markov stochastic process with a mean change of zero and a constant variance rate. The last
property, the Ito process, is a generalized Wiener process. The parameters a and b are functions
of the value of the underlying variable x and time t (Equation 3.10). Incorporating the three
processes result in Equation 3.11.
43
Akita University

𝑑𝑥 = 𝑎(𝑥, 𝑡)𝑑𝑡 + 𝑏(𝑥, 𝑡)𝑑𝑧 Equation 3.10

∆𝑥 = 𝑎(𝑥, 𝑡)∆𝑡 + 𝑏(𝑥, 𝑡)𝜀√∆𝑡 Equation 3.11

This equation applies a small approximation. It presumes that the drift and variance rate
of x remain consistent, equal to the values at time t, during the time interval between t and t+∆t.
Note that the process in Equation 3.10 is Markov because the change in x at time t depends only
on the value of x at time t, not on its history.

One way of gaining an intuitive understanding of a stochastic process for a variable is

to simulate the behavior of the variable. The simulation involves dividing a time interval into
many small time steps and randomly sampling possible variables' possible paths. The future
probability distribution for the variable can then be calculated. A technique for sampling
random outcomes for a stochastic process is named Monte Carlo simulation. Monte Carlo
simulation has the benefit of being able to be employed when the payment is conditional on the
path taken by the underlying variable S and when it is just dependent on the final value of S.
However, Monte Carlo simulation has the disadvantage of being computationally intensive and
unable to handle scenarios when early exercise opportunities exist (Hull, 2015).

Monte Carlo simulation employs the risk-neutral valuation outcome when valuing an
option. The steps are:

1. Sample a random path for S in a risk-neutral simulation.

2. Calculate the payoff from the derivative.
3. Reproduce steps 1 and 2 to get multiple sample values of the payoff from the derivative
in a risk-neutral simulation.
4. Compute the mean of the representative payoffs to estimate the expected payoff in a
risk-neutral simulation.
5. Discount this expected payoff at the risk-free rate to get an estimate of the value of the
derivative.

This thesis developed real option methodology particularly simulation method in

mining industry. The mining industry has several uncertain variables during production. In this
study, an evaluation method based on Monte Carlo Simulation was developed for economic
assessment through real option (RO) analysis, namely Stratified State Aggregation (SSA). The
methodology benefits the simplicity and flexibility to incorporate all uncertainties in one
assessment. In the oil industry, SSA assessment based on simulation has been performed

44
Akita University

(Adachi et al., 2008). However, none of them is applied in the mining project, which has
different calculation methods on uncertainties calculation. The considerable risk and
uncertainties in this study are economic, technical, and geological factors represented by price,
cost, and reserve, respectively. However, modeling the uncertainty is unique, and we developed
it through this research. Nevertheless, integration of those uncertainties will improve the
applicability of RO in the mining industry. Detail work of simulation approach is explained at
Chapter 4.

45
Akita University

3.5. References

Adachi, T., Mogi, G., & Adachi, T. (2008). Real Option Analysis of Multi-Stage Investment
on Resources Development Project by Improved SSA Method. Mining and Materials
Processing Institute of Japan, 124(9), 576–582.

Black, F., & Scholes, M. (1973). The pricing of options and corporate. The Journal of Political
Economy, 81(3), 637–654.

Brennan, M. J., & Schwartz, E. S. (1985). Evaluating Natural Resource Investments. The
Journal of Business, 58(2), 135. https://doi.org/10.1086/296288

Crundwell, F. (2008). Finance for Engineers: Evaluation and Funding of Capital Projects.
Springer-Verlag London Limited. https://doi.org/10.1007/978-1-84800-033-9

Cox, J. C., Ross, S. A., & Rubinstein, M. (1979). Option pricing: A simplified approach. Journal
of Financial Economics, 7(3), 229–263. https://doi.org/10.1016/0304-405X(79)90015-1

Hull, J. C. (2015). Options, futures, and other derivatives. In Asset Pricing (Ninth, Vol. 59,
Issue 2). Pearson Education, Inc.

Myers, S. C. (1977). Determinants of corporate borrowing. Journal of Financial Economics,

5(2), 147–175. https://doi.org/10.1016/0304-405X(77)90015-0

Mun, J. (2006). Real options analysis: tools and techniques for valuing strategic investments
and decisions. In John Wiley & Sons, Inc (2nd ed.). John Wiley & Sons, Inc.

Nicholas, G. (2014). An Integrated Risk Evaluation Model for Mineral Deposits [The
University of Adelaide]. https://doi.org/https://hdl.handle.net/2440/86234

Shafiee, S., Topal, E., & Nehring, M. (2009). Adjusted real option valuation to maximise
mining project value - A case study using century mine. Australasian Institute of Mining
and Metallurgy Publication Series, 3, 125–134.

Trigeorgis, L. (1993). The Nature of Option Interactions and the Valuation of Investments with
Multiple Real Options Author. Journal of Financial and Quantitative Analysis, 28(1), 1–
20.

46
Akita University

Chapter 4 Multi-stage Stratified State Aggregation

4.1. Stratified State Aggregation (SSA)

Stratified State Aggregation (SSA) algorithm was a systematic numerical technique to

value American options. It was able to compute the prices of complex American instruments in
seconds on a workstation. The algorithm was aimed to resolve the problem of the classical
method to overcome high dimensional valuation. The methodology applied Monte Carlo
Simulation for modeling uncertainty and performed a backward calculation to get the value at
present (Barraquand & Martineau, 1995).

Multi-stage SSA (MSSA) is an RO simulation method which run through Java

programming. Developed MSSA algorithm was used for the oil industry with price uncertainty
driver (Adachi et al., 2008). It was developed to analyze oil development projects, including
early stages of exploration, development, and production. Every stage had its capital
expenditure, and the success probability increased with the expenditure. They calculated the
value of a crude oil development project using real options analysis in a multi-stage investment
with geological and pricing uncertainties that decreased as the project progressed. The research
approached investment initiatives as a multi-staged compound or rainbow option investment
akin to a research and development project. Pricing uncertainty and diminishing geological
uncertainty were accounted for as project values vary overtime to find the best moment to go
from exploration to development and eventually extraction.

This research performed identical Java scripts in mining areas which has different
properties with the oil industry, especially uncertainty modeling. We practiced on developed
mining site, especially underwater tin mine planning. The company has the right to extract the
mineral considering the uncertainty of selling price, operating cost, and the amount of mineral
itself. Regarding the limited number of dredges as main equipment, every mining location is

47
Akita University

treated as stage extraction. Therefore, MSSA is an alternative method to compute the real value
of multi project with multi uncertainties driver.

MSSA method started with building stratification map, which represents a partition or
state-space path traveled by the asset's value. It was separated into thin layers along with the
maps like the optimal stochastic control path shown in Figure 4.1. The controlling path was
price, grade, and cost uncertainty.

(MUSD)

(Month)

Figure 4.1 Multi-stage Stratified State Aggregation Methodology

The algorithm simulates project value 100,000 times considering three uncertainties.
The method is an alternative to combine three uncertainties and more than one project. In the
case study, the option type calculation is delay option of two mining project. The company has
the right to mine, wait or abandon the project considering the three uncertainties.

48
Akita University

Figure 4.2 Stratified State Aggregation (SSA) Methodology

Figure 4.2. detailed the steps of the MSSA algorithm. The algorithm was started by
building 400 cells as a state-space of the project value. The multi-stage development was
defined according to the parameters S.

𝑎
𝑆ℎ,𝑗 (i) : the value of the asset a (a=1,….A) at time 𝑡𝑗 to maturity in the ith path of stage h

1 2 𝐴
𝑆ℎ,𝑗 = ( 𝑠ℎ,𝑗 (i), 𝑆ℎ,𝑗 (i),….𝑆ℎ,𝑗 (i) ): the ith mapped path of multiple assets in stage h

The state-space path was divided to run through several cells (cell division stage)
corresponding to the average expected payout of the total number of pathways.

𝑘
Then, whole range of 𝑆ℎ,𝑗 was divided into intervals 𝐶ℎ,𝑗

𝑘 𝑘 𝑘+1
𝐶ℎ,𝑗 = [𝑐ℎ,𝑗 ,𝑐ℎ,𝑗 ] (k = 1,…K)

𝑘
𝑆ℎ,𝑗 = in the state k when 𝑆𝑗 ∈ 𝐶ℎ,𝑗

𝑃ℎ,𝑗 (k) = the exercise value of the investment option at 𝑡𝑗 in stage h

𝑉ℎ,𝑗 (k) = the continuation value when the option is not exercised at 𝑡𝑗 in stage h

49
Akita University

Furthermore, the division path of project value was generated using Monte Carlo
simulation to model uncertain parameters. Generated the mapped paths of the asset 𝑆ℎ,𝑗 (i) and
𝑘
counted the number of paths 𝑎ℎ,𝑗 (k) such that 𝑆ℎ,𝑗 ∈ 𝐶ℎ,𝑗 (Equation 4.1).

𝑘
𝑆ℎ,𝑗 (i) ∈ 𝐶ℎ,𝑗 ∩ 𝑆ℎ,𝑗+1
𝑓ℎ.𝑗 (k) = ∑𝑆 𝑘
ℎ.𝑗 ∈ 𝐶ℎ,𝑗
(𝑒 −𝑟𝑈ℎ 𝑃ℎ+1,𝑗+𝑈ℎ (m) -𝐼𝑜 ) Equation 4.1
∑𝑆 (𝑒 −𝑟𝑈ℎ 𝑆ℎ,𝑗+𝑈𝐻 -𝐼𝑜 )
ℎ,𝑗 ∈𝐶𝑘
ℎ,𝑗

Where

r : Risk free interest rate

𝐼0 : Exercise price (investment) in stage h.

𝑚
𝑆ℎ,𝑗+ 𝑈ℎ = 𝑆ℎ+1,𝑗+𝑈ℎ ∈ 𝐶ℎ+1,𝑗+𝑈ℎ
when h = 1,……,H-1

𝑘 𝑚
Next step was calculating transition probabilities from 𝐶ℎ,𝑗 to 𝐶ℎ,𝑗+1 and forward
𝑓ℎ,𝑗 (𝑘)
exercise value 𝑒ℎ,𝑗 (𝑘) = (Equation 4.2). Finally, we maximized exercise value and
𝑎ℎ,𝑗 (𝑘)

waiting value (Equation 4.3).

k k+1
Number of paths from Ch,j to Ch,j+1
∏h,j (k. k + 1) = Equation 4.2
k
Number of paths through Ch,j
Ph,j (k) = max[Eh,j (k), Vh,j (k)]
Where: 𝑉ℎ,𝑗 (𝑘) = 𝑒 −𝑟(𝑡𝑗+1−𝑡𝑗 ) ∑𝐾
𝐼=1 ∏ℎ,𝑗 (𝑘, 𝑙 )𝑃ℎ,𝑗+1 (𝑙) Equation 4.3
𝑓ℎ,𝑗 (𝑘)
𝐸ℎ,𝑗 (𝑘 ) =
𝑎ℎ,𝑗 (𝑘)
As long as the value of waiting to invest is higher than the value of investing
immediately, the company should defer the investment decision (Mun J., 2006). Otherwise,
they do not need to wait to decide to invest or abandon the project.

50
Akita University

4.2. Economic Uncertainty Model

Modeling uncertain parameters must follow several principles of stochastic calculus. A

stochastic process is a variable that changes over time in a way that is at least in part random.
Formally, it is defined by probability law for evolution variable (x) through time (t). The model
must satisfy Markov processes which are described as later variable estimation (xt+1) depending
only on previous data (xt) and not additionally on condition before time t (Dixit & Pindyck,
1993).

Price is a time series variable in the cash flow parameter. Therefore, forecasting price
and modelling the uncertainty should be started by statistical analysis of historical price. One
of the analyses is unit root test to check the stationary data. The most popular one, and the one
that we used, is the Augmented Dickey–Fuller (ADF) test. This test is based on the process yt
= ϼyt-1+vt is stationary when ϼ < 1, but when ϼ=1, it becomes the nonstationary random walk
process yt = ϼyt-1+vt. Hence, one way to test for stationarity is to examine the value of ϼ. In other
words, we test whether ϼ is equal to one or significantly less than one. where the vt are
independent random errors with zero mean and constant variance δ2. Non stationary data can be
modelled using random walk model or Geometric Brownian Motion (GBM) model (Hill et al.,
2011).
The tin price is collected from January 2016 to December 2020 to model the price in 5
years’ production. The ADF test resulted the data was not stationary (Table 4.1). The null
hypothesis that tin price has a unit root can be rejected because the error probability was 27%
or more than 5% threshold. Therefore, we used Geometric Brownian Motion (GBM) model to
simulate SSA. Brownian motion model is a continuous-time stochastic process with three
important properties: Markov process, Wiener process (probability distribution is independent
of any other time interval), and normally distributed with increasing variance over time.
Table 4.1 ADF test on Tin Price data (2016-2020)

Null Hypothesis: TINPRICE has a unit root

Exogenous: Constant
Lag Length: 1 (Automatic - based on SIC, maxlag=10)

t-Statistic Prob.*

Augmented Dickey-Fuller test statistic -2.032564 0.2725

Test critical values: 1% level -3.546099
5% level -2.911730
10% level -2.593551

*MacKinnon (1996) one-sided p-values.

51
Akita University

Further uncertainty modelling stage is measuring the volatility of the asset. Thus, we
analyze the price volatility as indicated by its standard deviation (Equation 4.4). The price is
frequently watched at defined periods to assess the volatility of a price experimentally.

1 𝑛
𝑠=√ ∑ (𝑢𝑖 − 𝑢̅)2 Equation 4.4
𝑛 − 1 𝑖=1

Where:
s : standard deviation
𝑃𝑖
𝑢𝑖 ∶ price growth = ln ( )
𝑃𝑖−1
𝑢̅ : mean of 𝑢𝑖
n+1 : number of observation
The standard deviation of the 𝑢𝑖 is 𝜎 √𝜏, where τ is the length of time, thus, σ=s/√τ. It is
not easy to pick a fair value for n. More data typically leads to greater accuracy, although
accuracy varies with time, and too old data for predicting future volatility may be irrelevant.
Alternatively, n can be specified as the number of days to which volatility will be applied as a
rule of thumb (Hull, 2015).
Mining project revenue depends on the price of sold material. In the case study,
historical price sourced from S&P data from January 2016 to December 2020. The unit root
analysis through Augmented Dickey Fuller test resulted a nonstationary data with probability
27%. Therefore, we used GBM assumption to model the price in the project life. Application
of Equation 4.4 to tin price (Table 4.2) resulted standard deviation of 4.7%. The fluctuation was
based on 60 monthly data. The volatility will be used to predict 5 years later.

1
𝑠𝑝𝑟𝑖𝑐𝑒 = √ (0.1373)
60 − 1

𝑠𝑝𝑟𝑖𝑐𝑒 = 0.047

52
Akita University

Table 4.2 Tin Price Volatility

Tin, LME US$/t 2

Period Growth rate (𝑢𝑖) (𝑢𝑖 − 𝑢̅)2 ∑(𝑢𝑖 − 𝑢
̅)
(Pi)

1/1/2016 14,906 𝑢𝑖 = Ln(Pi/Pi-1) 0.1373

2/1/2016 15,976 0.07 0.0043
3/1/2016 16,729 0.05 0.0018
4/1/2016 17,260 0.03 0.0007
5/1/2016 16,324 (0.06) 0.0036
6/1/2016 17,075 0.04 0.0017
7/1/2016 17,840 0.04 0.0016
8/1/2016 18,882 0.06 0.0028
9/1/2016 20,164 0.07 0.0038
10/1/2016 20,885 0.04 0.0010
11/1/2016 21,320 0.02 0.0003
12/1/2016 21,205 (0.01) 0.0001
… … … …
… … … …
… … … …
1/1/2020 16,425 (0.04) 0.0027
2/1/2020 16,267 (0.01) 0.0003
3/1/2020 14,667 (0.10) 0.0123
4/1/2020 15,274 0.04 0.0011
5/1/2020 15,502 0.01 0.0001
6/1/2020 16,819 0.08 0.0055
7/1/2020 17,909 0.06 0.0031
8/1/2020 17,550 (0.02) 0.0008
9/1/2020 17,448 (0.01) 0.0002
10/1/2020 17,724 0.02 0.0001
11/1/2020 18,642 0.05 0.0019
12/1/2020 20,544 0.10 0.0081
Logarithmic natural growth analysis based on historical data resulted in the price has an
average monthly growth rate (α) of 0.74% at Table 4.2 and detailed ata Appendix A. However,
to be applied at our evaluation method, we assumed the price growth would follow the
stochastic differential equation model, particularly the one-factor constant convenience yield
model for risk-neutral prices stated in Equation 4.5 (Cortazar et al., 2001). Assumed parameters
were risk-free rate (r) of 0.25%, convenience yield (δ) 0.17% and calculated price volatility ( 𝜎𝑃 )
of 4.7%. We simulated the GBM price forecasting with the assumed parameter at Figure 4.3.

53
Akita University

𝑑𝑃
= (r-δ)𝑑𝑡 + 𝜎𝑃 𝑑𝑧𝑃 Equation 4.5
𝑃

Where P = price,
r = risk free rate,
δ = convenience yield,
𝜎𝑃 = price volatility,
𝑑𝑧𝑃 = Wiener Process.

Figure 4.3 Price Simulation Model (Geometric Brownian Model)

Economic uncertainty is represented by price fluctuation. MSSA approach is an

alternative to evaluate that risk. With the assumption of no added investment and price
uncertainty, the project should continue to operate in January 2021. The price of tin per ton was
23,199 USD in January 2021. It had a net present value (NPV) of 7.5 MUSD, real option value
(ROV) of 7.68 MUSD, and a 1.99% option premium below our 5% threshold. The critical
present value of 3 MUSD was reached when the tin price dropped 25% to 17,399 USD/ton
(Figure 4.4). The condition altered the option premium to 8%, and the company should delay
the project. Moreover, the decreasing price of 45% into 12,759 USD/ton made option value
reach zero, and consequently, the project is completely unprofitable.

The variability of real option value can only be captured when we simulated varies of
the initial investment in either Black Scholes or lattice valuation. An ongoing project can only

54
Akita University

be evaluated through the simulation method. The underlying asset represented by present value
could not capture permutation between cash flow variable which MSSA approach could do.

10
8 -10%
-20%
6
Project Value (MUSD)

-30%
4
-40%
2 -75% -50%

0
Initial
-2 Assumption
-4 Critical Value

-6
-8
-10
5 10 15 17.4 20 23 25
Price ($/Ton) Thousands

Present Value Option Value

Figure 4.4 Project Value Considering Price Changing only

Conducting RO analysis through common approach is needed to benchmark proposed

method. As a result, we assumed a capital cost of 1 MUSD and considering price uncertainty
only. We conducted on the 3 approaches of Black Scholes (BS) Formulation, Binomial Lattice
(BL) valuation, and Stratified State Aggregation (SSA) methodology. Asset or project
uncertainty is used to calculate real option value through Black-Scholes and Lattice method.
The comparison result is resumed at Table 4.3. It concluded that the differences between
approaches are not significant. However, the SSA has advantages on the calculation on multi
uncertainty and multi stage of project. The only disadvantage of SSA method is calculation
time of the number simulation. In the case study, we used 100,000 simulations to get optimum
number of simulation.

55
Akita University

Table 4.3 Comparison of Real Option Approach

Parameters SSA Value BL Value BS Value Unit

Initial Project Value 7.54 7.54 7.54 MUSD
Exercise Price 1.0 1.0 1.0 MUSD
Price uncertainty value 4.7%/month 4.7%/month 4.7%/month
Option Value 6.921 6.960 6.997 MUSD
Option premium 6.8% 7.6% 8.2%
Decision Delay Delay Delay

4.3. Geological Uncertainty Model

Reserves have varying levels of geoscientific confidence and uncertainty. A mining

company must spend much capital to get confidence in geological uncertainty. Geological
uncertainty is a spatial uncertainty that revealed after the reserve is extracted. Typically, kriging
methodology is used to estimate the total amount of reserves. However, it could not be enough
to capture the uncertainty. Therefore, we used Conditional Simulation (CS) method to measure
the geological uncertainty. It is a developed methodology based on the kriging method (Dowd,
1994).

The valuation of mining projects is based on conditional simulation (CS) of ore body
parameters. CS is a geostatistical tool that can generate punctually or block ‘realizations’ of
mineral grades. Each realization is intended to honor the histogram and semivariogram of the
true grade distribution and honor known data points (Nicholas, 2014). It is a technique to
measure spatial uncertainty based on spatial Monte Carlo Analysis (MCS). CS reproduces the
properties of an ore body and quantifies the variability, which we call “spatial uncertainty.” The
realization of CS methodology on our case study is figured at Figure 4.5. The utilization of
MCS is for the NPV probability distribution function and the variability of the projected cash
flow (Jurdziak L and Wiktorowicz, 2008).

56
Akita University

Figure a Conditional Simulation Figure b Conditional Simulation Figure c Expected Conditional

1 1000 Simulation

Figure 4.5 Conditional Simulation for Modelling Reserve

Figure 4.6 Mid Reserve Grade-Tonnage

We found that each location of resource has its expected grade and tonnage as well as
volatility (Figure 4.6 and Figure 4.9). In our 1000 simulation, we get reserves amount 7.2 Mton,
grade 0.594 volatility 9% (Figure 4.6, and Figure 4.7) and grade amount 1.1 Mton, grade 0.558,
volatility 15.8% (Figure 4.9, and Figure 4.10) for mid and south reserve, respectively. The
stability of the result is reached after more than 400 simulations. The probability distribution of
Conditional Simulation result followed normal distribution (Figure 4.8 and Figure 4.11). Thus,
we concluded the result is reliable to be the base of economic evaluation using discounted cash
flow analysis and real option valuation methodology.

57
Akita University

Figure 4.7 Mid Grade Uncertainty

Figure 4.8 Probability Distribution (Mid Reserve)

In the research, one originality is the utilization of conditional simulation results, which
are typically only for estimating reserve. While, herein, we used to calculate the grade
uncertainty and calculate further the project value through real options analysis. The common
Kriging method could not provide variability of the reserve. Therefore, conditional simulation
was alternative method to calculate grade uncertainty and anticipate it in a project lifetime.

58
Akita University

Geological uncertainty was represented by the grade uncertainty, explained in the

previous chapter. The originality of this research was the utilization of conditional simulation
to be combined with real options methodology. It can only be measured using MSSA
methodology and resulted in Figure 4.10. The result of geological uncertainty measurement
showed a similar outcome with economic risk. The reason was with a similar modeling method,
which belongs to the Geometric Brownian Model and the modeling at income sector.

Figure 4.9 South Reserve Grade-Tonnage

Figure 4.10 Probability Distribution (South Reserve)

Economic (price) uncertainty had positive growth assumptions while geological risk
modeling assumpted only followed normal distribution assumption. The calculation through
59
Akita University

Figure 4.11 South Grade Uncertainty

SSA approach had a net present value (NPV) of 7.5 MUSD and real option value (ROV) of
7.63 MUSD (Figure 4.12). SSA method replicated the distribution of conditional simulation
and multiply the probability and the value of each simulation. It indicated an option premium
of 1.29% and recommended continuing the project. When the reserve is dropped 25%, the
option premium reached 9%. It surpassed our 5% threshold and recommended delaying the
project. Furthermore, when the grade is dropping 40%, the decision recommendation shifted to
postpone the project.

10
8 -10%
Project Value (MUSD)

6
4 -30%
-50% -20%
2 -75% -40%
0
-2 Current
Critical Assumption
-4 Value
-6
-8
-10
2 4 6 6.2 8 8.3 10
Total Reserve (M Ton) Thousands
Present Value Option Value

Figure 4.12 Project Value Considering Grade Changing only

60
Akita University

4.4. Technical Uncertainty Model

We measure the technical uncertainty represented by cost volatility. The cost consisted
of mining, processing and metallurgical cost. Unit cost of production model construction was
based on historical cost given by the company with a similar location and production technique
(Figure 4.13). Similar ADF methodology was applied to the summarized historical unit cost
data (Figure 4.14). The unit root test concluded that the cost data is stationer (Table 4.4).
Therefore, the construction of the technical uncertainty model was used the Mean Reverting
(MR) model (Dixit A.K. and Pindyck R.S, 1993) instead of the GBM model. We used the
simplest mean-reverting process-also known as the Ornstein-Uhlenbeck process (Equation 4.8).
The process follows Markov property and does not have independent increments.

800
Thousands

700
Operating Cost (USD)

600
500
400
300
200
100
-
Jul-17
Jul-15

Oct-18
Oct-15

Jul-16
Oct-16

Oct-17

Jul-18

Jul-19
Oct-19

Jul-20
Oct-20
Jan-15

Apr-16
Apr-15

Jan-16

Jan-17
Apr-17

Jan-18
Apr-18

Jan-19

Jan-20
Apr-19

Apr-20
Period

Fuel Cost Employee Cost Material Cost Depreciation Others Total (USD)

Figure 4.13 Historical Operating Cost Data

61
Akita University

4.00

Cost Per Ton Ore (USD/Ton)

3.50
3.00
2.50
2.00
1.50
1.00
0.50
-
Jan-15

Sep-15
Jan-16

Sep-16
Jan-17

Sep-17
Jan-18

Sep-18
Jan-19

Sep-19
Jan-20

Sep-20
May-15

May-16

May-17

May-18

May-19

May-20
Period

cost

Figure 4.14 Modified Historical Unit Cost Data

Table 4.4 ADF Test of Cost data

Null Hypothesis: COST has a unit root

Exogenous: Constant
Lag Length: 0 (Automatic - based on SIC, maxlag=10)

t-Statistic Prob.*

Augmented Dickey-Fuller test statistic -6.633826 0.0000

Test critical values: 1% level -3.544063
5% level -2.910860
10% level -2.593090

*MacKinnon (1996) one-sided p-values.

Parameters calculation for the MR process started by running discrete-time data

regression (Equation 4.6).

𝑥𝑡 − 𝑥𝑡−1 = 𝑎 + 𝑏 𝑥𝑡−1 + ε𝑡 Equation 4.6

Then, we calculated x = -a/b, and η = −log(1 + b), and finally, we calculated the
standard error of the regression using equation 4.7.

log(1 + 𝑏)
𝜎 = 𝜎𝜀 √ Equation 4.7
(1 + 𝑏)2 − 1

62
Akita University

2.5

1.5

0.5
X(t)-X(t-1)

0 Y
0 0.5 1 1.5 2 2.5 3 3.5 4 Predicted Y
-0.5

-1

-1.5

-2

-2.5
X(t-1)

Figure 4.15 Regression Analysis of Historical Cost

The regression model in Figure 4.15 resulted in intercept (b) 1,33 and variable (a) (-
0.88). In addition, it has an R squared error of 44%. Although the accuracy is quite low, the
simulation shows similarity with the historical data. It has better results than the GBM model
(Figure 4.16). In the Figure 4.16, the data was real historical value before January 2021and
followed by the prediction. In summary, the linear regression cost model tends to be back to its
mean in 92% rate (η). Nevertheless, the volatility σc rate is very high, reaching 39%.

dC
= η(x̄ − x)dt + σc dz𝑐 Equation 4.8
C

63
Akita University

In our research, technical uncertainty is related to the cost of operation, which consists
of mining cost, processing cost, and metallurgical cost. The purpose is to compare with other
uncertainty and the strategy to hedge the uncertainty. Historical data analysis showed that the
uncertainty is very high, reaching 39%. Additionally, the model had a different approach with
other uncertainties, which used the Mean Reverting Process and affected the expense side of
the project. Increasing cost decreased project value and vice versa. With the cost uncertainty
being the only driver as assumptions, the project is recommended to be delayed with Net Present
Value (NPV) 7.5 MUSD and Real Option Value (ROV) 15 MUSD (Figure 4.17). It had an
option premium of 100% far beyond our assumed threshold. The option premium can be
pressed below our threshold when either the cost dropped 75% or the volatility decreased to
5%.

Figure 4.16 Cost Uncertainty Model (GBM vs MR)

64
Akita University

20
-75% -50%
-25%
+25%
15 +50% +75%

Project Value (MUSD)

10

Current assumption
0
0 100 200 300 400 500 600 700

-5
Total monthly cost (*1000 USD)

Present Value Option Value

Figure 4.17 Project Value Considering Cost Changing only

4.5. Combined Uncertainty Assessment

Price, grade, and cost uncertainties were the driver of project risk assessment. The
mining project resulted in net present value output with SSA method in Java was 7.5 MUSD
and real option value 17.6 MUSD. Generated, exercise, and waiting cells values as illustrated
in Figure 4.18, Figure 4.19, and Figure 4.20., respectively. They were randomly picked
simulations from both locations of mining. The exercise values were the overall payoff if one
decides to invest. The waiting values were the expected overall payoff if one decides not to
100

80
Project Value (MUSD)

60

40

20

0
0 10 20 30 40 50 60
-20

-40
Month

1 34 79 138 163 209

233 293 313 347 382 400

Figure 4.18 Generated Cells with Three Uncertainties Driver (Random

Cells Picked)

65
Akita University

120

100

80
Project Value (MUSD)

60

40

20

0
0 10 20 30 40 50 60
-20

-40

-60
Month

1 34 79 138 163 209

233 293 313 347 382 400

Figure 4.19 Exercise Value with Three Uncertainties Driver (Random Cells
Picked)

60

50
Project Value (MUSD)

40

30

20

10

0
0 10 20 30 40 50 60
-10
Month

1 34 79 138 163 209

233 293 313 347 382 400

Figure 4.20 Waiting Value with Three Uncertainties Driver (Random Cells
Picked)

invest and wait to invest in future time. Furthermore, the project value will be the maximum
values between the exercise option computation and wait-to-operate value computations in each
cell at each time.

66
Akita University

Combination of the three uncertainties is aimed to evaluate the project with the
uncertainties. The exercise and waiting values are compared at each period to decide whether
to proceed with expansion or wait for more favorable economic conditions. These are computed
in the backward calculation. Each cell's exercise and waiting value will always be compared to
obtain the project value for each cell.

The exercise, waiting, and project value computations indicated that the project values
can take any value when deciding to exercise now or in the future. These results showed the
impact uncertainties effects on the mining project. However, the decision-making from the java
results can be made from the frequency distribution, probability distribution, and most
importantly, on the output net present value and real option value with its sensitivity analysis.
A total of 100,000 paths were simulated to calculate the expansion option payoff, and 84,433
were output from the payoff frequency distribution in Figure 4.21. It showed that the largest
number (16,000) of paths remained at payoff range 0, implying the computation recommended
to wait for the project. However, other simulations yielded positive values with most
probabilities at 7 MUSD and normally distributed to 23 MUSD.

18000

16000

14000
Number of simulation

12000

10000

8000

6000

4000

2000

0
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23
Payoff Value

Figure 4.21 Probability Distribution

67
Akita University

Our assumed MSSA simulation with price, grade, and cost uncertainty consideration
resulted in a project real option value of 17.6 MUSD compared to 7.5 MUSD when using the
traditional DCF Method. We simulated for changes in the critical parameter, which are initial
price (Figure 4.22), initial total reserve (Figure 4.23), and initial total cost (Figure 4.24). With
the determined uncertainty parameter, changing price and reserve gave a similar project value
(Figure 4.22 and 4.23). These results showed a delaying decision recommendation with option
premium reached, 135% which is far beyond our 5% threshold. Moreover, the decision
recommendation shifted into abandoning the project when the project value decreased 90% due
to the option value approaching zero.

40

30

Current Assumption
Project Value (MUSD)

20 +90%
+75%
10 +50%
+25%
0
-15 -10 -5 0 -25% 5 10 15 20 25 30
-50%
-10
-75%
-90%
-20
Present Value (MUSD)

Present Value Option Value

Figure 4.22 Project Value with Initial Price Changing

68
Akita University

40

30

Current Assumption +90%

Project Value (MUSD)

20
+75%
+50%
10
+25%

0 -25%
-15 -10 -5 0 5 10 15 20 25 30
-50%
-75% -10

-90%
-20
Present Value (MUSD)

Present Value Option Value

Figure 4.23 Project Value with Initial Total Reserve Changing

20

15 -90%
-75%
10 -50%
-25%
Project Value (MUSD)

5
Current Assumption
+25%
0
-25 -20 -15 -10 -5 0 +50% 5 10 15 20
-5
+75%
-10

-15

-20
+200%
-25
Present Value (MUSD)

Present Value Option Value

Figure 4.24 Project Value with Initial Total Cost Changing

Cost changing impacts wait-to-operate value to execution when it drops more than 75%.
Otherwise, the value has remained at 17 MUSD (Figure 4.24). The option value was peaked at
the project value without cost. The result explained that cost uncertainty has huge impact on
69
Akita University

option value. Therefore, we studied further about the contribution of the uncertainty. The
dominant uncertainty was cost uncertainty (39%), followed by geological (9% and 16%) and
price uncertainty (5%). The simulation recommended waiting for the project to be stable
because of the huge option premium which is the differences between present value and option
value.

Table 4.5 Present Value Regarding Price-Reserve Changing

Present Value Price

(MUSD) -90% -75% -50% -25% 0 25% 50% 75% 90%
-90% (15.00) (14.61) (13.97) (13.33) (12.68) (12.04) (11.40) (10.76) (10.37)
-75% (14.61) (13.65) (12.04) (10.43) (8.82) (7.21) (5.61) (4.00) (3.04)
-50% (13.97) (12.04) (8.83) (5.61) (2.40) 0.51 2.90 5.22 6.60
Reserve

-25% (13.33) (10.43) (5.62) (0.86) 2.90 6.38 9.85 13.31 15.41
0 (12.68) (8.83) (2.40) 2.90 7.52 12.16 16.78 21.43 24.22
25% (12.04) (7.21) 0.50 6.37 12.16 17.92 23.72 29.52 32.98
50% (11.39) (5.61) 2.90 9.85 16.79 23.76 30.67 37.62 41.79
75% (10.76) (4.00) 5.22 13.32 21.42 29.52 37.63 45.73 50.54
90% (10.37) (3.04) 6.62 15.40 24.19 32.99 41.81 50.60 55.84

Table 4.6 Option Value Regarding Price-Reserve Changing

Option Value Price

(MUSD) -90% -75% -50% -25% 0 25% 50% 75% 90%
-90% - - - - - - - 0.19 0.53
-75% - - - 0.49 1.81 3.09 4.44 5.77 6.64
-50% - - 1.82 4.41 7.15 9.97 12.68 15.17 16.70
Reserve

-25% - 0.48 4.39 8.62 12.77 16.49 20.05 23.31 25.26

0 - 1.82 7.18 12.64 17.66 22.24 26.46 30.62 33.09
25% - 3.12 10.04 16.49 22.31 27.51 32.67 37.73 40.73
50% - 4.42 12.69 20.01 26.55 32.69 38.72 44.71 48.41
75% 0.21 5.77 15.26 23.32 30.68 37.74 44.85 51.87 56.21
90% 0.54 6.60 16.72 25.27 33.02 40.80 48.43 56.23 60.96
The superiority of the real options methodology compared to the traditional method
enables the manager to plan activities when uncertainty is revealed. The option value will be
taken from the possibility of the improved economic parameter (Crundwell, 2008). Therefore,
we made a decision map when initial price, and reserve change in the future as shown at Table
4.5, Table 4.6, Figure 4.25. The way to recommend a decision was by comparing present value
and option value. With a threshold of 5% and determined price, grade and cost uncertainty
parameters, MSSA methodology will always recommend holding the project. It turned to
abandon recommendation when the price or reserve drops 90%. The other scenario between
price and reserve value was described in Table 4.5, Table 4.6, Figure 4.25. The results of the

70
Akita University

MSSA methodology (Table 4.5 and Figure 4.25) opposed the results of the traditional method,
which recommends immediate exercising of the project. The traditional method recommended
exercising the project when the present value exceeds zero. The present value becomes negative
when the price or reserve drops by 50%. The overall result in both traditional and real option
method is symmetrical because both affect the income value of the cash flow. Discounted Cash
Flow methodology only views the projected income to determine the decision. The assumed
project cash flow is sensitive to income and outcome parameters (Table 4.5). Price and reserve
parameters are the income side of the project. Both parameters will be revealed when the
company decides to exercise the project.

15,821

14,572
INITIAL TOTAL RESERVE (K TON)

12,491

10,409 Assumption

8,327

6,245

4,164
Wait to Operate
2,082
Abandon
833

INITIAL PRICE (USD/TON)

Figure 4.25 Initial Price and Total Reserve Decision Map

71
Akita University

4.73
Abandon

Initial unit cost (USD/ton ore)

4.05

3.38 Assumption

2.70

2.03

Execute
1.35 Wait

0.68
5,800 11,600 17,399 23,199 28,999 34,799 40,598
Initial Price (USD/ton)

Figure 4.26 Initial Price and Unit Cost Decision Map

Figure 4.25 and Figure 4.26 showed a condition of decision recommendation
considering the initial price, total reserve, and monthly cost shifting. It is divided into two areas
which are waiting to operate and abandoning the project. When the mine planning is done, the
initial condition was a tin price of 23,199 USD/ton and obtained expected total reserve from
conditional simulation of 8.3 Mton. On the other hand, the unit cost to operate the two mining
areas was 2.7 USD/ton ore. Those conditions resulted in wait-to-operate decision
recommendations when considering price, grade, and cost uncertainties. The decision can shift
in a certain situation as shown in the maps (Figure 4.25 and Figure 4.26).

Price and reserve uncertainties had similar properties related to the shift in present and
option values. However, cost shifting had different outcome patern. Increasing cost definitely
decreasing value of the project. The option premium as determinant of decision
recommendation can be decreased by increasing the profit or decreasing the uncertainties driver.
We have simulated further for the uncertainties driver at Figure 4.27.

72
Akita University

Figure 4.27 Reserve - Cost Uncertainty Scenario Analysis

Figure 4.27 showed that project uncertainties accounted for option premium in the real
options methodology. Price and reserve volatility is combined regarding their similar properties
of affecting income side of the project, while cost uncertainty separated to assess the result. The
figure explained that cost uncertainty was the most dominant factor of the option premium
compared to other uncertainties. For instance, if the company decides to have a 10% maximum
risk, they should decrease the cost volatility below 5%, and so on. Otherwise, the project is
delayed.

Nevertheless, the uncertainty alteration does not change the decision using a traditional
methodology represented by the present value. The simulation has proven that cost uncertainty
assurance is the key to project value using MSSA methodology. Additionally, the changing
price and reserve are related to the present and option value shift. However, it made a pattern
with the constant of option premium. Therefore, we simulated further for the no-cost
uncertainty case.

73
Akita University

4.6. Economic and Geological Uncertainty Case

Notating the distinction of cost parameters, we investigated the project value with no
cost uncertainty (Figure 4.28, and Figure 4.29). The assumption is the company can control the
cost to secure the profit. Our multi-stage SSA simulation with no cost uncertainty case yielded
an RO value of 7.8 MUSD, compared to the corresponding net present value (NPV) of 7.5
MUSD obtained from the standard DCF method. The simulation resulted in a significant
waiting value decreasing from 17.6 MUSD. The magnitude of the cost uncertainty parameter
affected the decision recommendation significantly. In addition, we modeled the changes using
important parameters: initial price and initial reserves (Figure 4.28, and Figure 4.29). They were
treated as if their positive and negative values fluctuated. Nonetheless, the critical value of
changing decision recommendation was specified when the option value approached the present
value and was close to zero.

10
-10%
8
-20%
6
-30%
Project Value (MUSD)

4
-40%
-75% -50%
2

0
Initial
-2 Critical Value Assumption
-4

-6

-8

-10
20.8
5 7 9 11 13 15 17 19 21 23 25
Price ($/Ton) Thousands

Present Value Option Value

Figure 4.28 Project Value with Price Changing

74
Akita University

10

8
-20%
6
-30%
Project Value (MUSD)
-10%
4
-40%
-75% -50%
2

-2 Critical Current
Value Assumption
-4

-6

-8

-10
2 3 4 5 6 7 7.5 8 8.3 9 10
Total Reserve (Ton) Thousands

Present Value Option Value

Figure 4.29 Project Value with Reserve Changing

When the price dropped by more than 10% from our assumption, the option premium
value surpassed the assumed 5% threshold (Figure 4.28). Because the trend of option premium
was increasing, the wait to execute decision was recommended to delay the execution of the
project. When the price dropped by 45% and the RO value was almost zero, the decision was
changed to "abandon the project."

The changing reserves parameter (Figure 4.29) had a similar result to the price alteration
(Figure 4.28). A decrease of 10% reserve from our expected CS shifted the decision from
execution to wait until the uncertainty was resolved. Moreover, when the reserves dropped by
45% from the estimated value in CS, the recommendation was changed from waiting to execute
the project to abandon it as the threshold exceeded 5%.

Scenario analysis was performed to map the decision recommendation and spread from
the original assumption shift in price and reserve grade. Table 4.7 and Table 4.8 represent the
project's present value and option value, respectively. Recommending a decision through RO
was performed by comparing the present and option values. In contrast, the traditional method
produced the present value as the only decision parameter.

75
Akita University

Table 4.7 Present Value of Reserve and Price Relative Changing

Present Value Price
(MUSD) -90% -75% -50% -25% 0 25% 50% 75% 90%
-25% (13.32) (10.43) (5.61) (0.87) 2.90 6.37 9.86 13.31 15.39
-20% (13.20) (10.12) (4.97) (0.02) 3.83 7.54 11.24 14.93 17.16
-15% (13.07) (9.79) (4.32) 0.76 4.75 8.68 12.63 16.55 18.92
-10% (12.94) (9.47) (3.68) 1.48 5.69 9.84 14.02 18.18 20.67
(12.81) (9.15) (3.04) 2.20 6.62 11.00 15.41 19.78 22.42
Reserve

-5%
0 (12.68) (8.83) (2.40) 2.90 7.54 12.18 16.80 21.38 24.20
5% (12.55) (8.51) (1.77) 3.59 8.47 13.31 18.18 23.04 25.95
10% (12.43) (8.18) (1.16) 4.30 9.39 14.49 19.57 24.67 27.69
15% (12.30) (7.86) (0.58) 4.99 10.31 15.64 20.95 26.30 29.48
20% (12.17) (7.54) (0.02) 5.68 11.26 16.79 22.32 27.89 31.24
25% (12.04) (7.22) 0.50 6.38 12.17 17.96 23.73 29.52 32.99

Table 4.8 Option Value of Reserve and Price Relative Changing

Option Value Price
(MUSD) -90% -75% -50% -25% 0 25% 50% 75% 90%
-25% - - - - 3.29 6.67 10.14 13.55 15.60
-20% - - - 0.52 4.20 7.85 11.47 15.17 17.35
-15% - - - 1.25 5.10 8.96 12.87 16.77 19.10
-10% - - - 1.95 6.01 10.11 14.25 18.39 20.90
-5% - - - 2.60 6.92 11.26 15.60 20.03 22.59
Reserve

0 - - - 3.29 7.82 12.39 16.97 21.57 24.35

5% - - - 3.95 8.72 13.53 18.40 23.23 26.15
10% - - - 4.67 9.66 14.70 19.78 24.89 27.90
15% - - 0.05 5.35 10.52 15.83 21.13 26.47 29.65
20% - - 0.62 6.03 11.51 17.01 22.51 28.11 31.42
25% - - 1.03 6.70 12.39 18.09 23.88 29.70 33.18

The colored blocks in Table 4.7 represent the decision recommendation through the
traditional DCF method. A positive present value would be reached if the price exceeded 12,000
dollars per ton ore or drops 50% from our initial assumption. With option premium due to the
parameter's wait-to-operate value resulting from uncertainty, the RO methodology and the SSA
approach improved the decision to neither execute (green color) nor reject (red color) the mine
project. Instead, it provided waiting recommendations for specific cases, as shown by the
yellow color in Table 4.8 and Figure 4.30.

76
Akita University

12490.5
Execute
11657.8

10825.1 Assumption Execute

Initial Total Reserve (K Ton)

9992.4

9159.7

8327 Wait
Wait
7494.3

6661.6
Abandon
5828.9

4996.2 Abandon

4163.5

3330.8

Initial Price ($/ton)

Figure 4.30 Decision Initial Price-Reserve Map

The DCF methodology only views projected income to determine the decision.
However, the assumed project cash flow was sensitive to income and outcome parameters. The
calculation of option premium resulted a risk that should be paid by company if they decided
to continue the project. Therefore, the risk apatite of a manager will be determinant of execution
decision. Theoretically, higher risk should give higher return and vice versa. In this research,
5% threshold was applied to recommend a decision. Shifting initial price and total reserve as
uncertainty driver was assessed. Current assumption stood at the borderline of execution and
wait to operate decision. The recommended decision can be shifted with changing price, reserve
or increasing the risk threshold.

Uncertainties were not accounted for in the traditional method; therefore, they did not
affect the present value and remained constant at 7.5 MUSD either using manual-based DCF
calculations or SSA through the JAVA program language. These calculations summarized the
cash flow model with the additional time value of money. Figure 4.31 displays the option value
alteration where the uncertainty was changed in increments of 5%. The value of waiting implied
that the project had to be delayed when the option value was far from the calculated present

77
Akita University

value. A shift in the price uncertainty had an almost symmetrical effect with the reserves
uncertainty to the option value. When the parameter was altered below the critical value of 8.3
MUSD, the decision recommendation was changed from execution to wait-to-operate or to
abandon the project. The price and reserves parameters presented the income side of the project;
therefore, both parameters were incorporated when the company decided to exercise the project.
The two uncertainties had similar properties, implying future decisions based on future
conditions.

Figure 4.31 Real Option (RO) Value with Price and Reserve Uncertainties Alteration

A mining company does not have control over price and reserve. Price will always
fluctuate in the future, which depends on the supply and demand of the commodity. On the
other hand, the reserve is a spatial base parameter. It will be known when mining the reserve.
Otherwise, the company should increase the exploration drilling to assure the reserve. The
spacing between drill holes determines the volatility. It can be additional data to confirm the
reserve amount and decrease the geological uncertainty. However, the geological uncertainty
will never be eliminated caused by the nugget effect. The term "nugget" is borrowed from
geostatistics, referring to the unexpected nugget of minerals founding a mining process (Yin J
et al., 2011). Cressie (1993) concluded that the nugget effect in geostatistics is caused by two
factors: micro-scale variation and measurement error.

The critical value was calculated to secure the project profitability. Dixit and Pindyck
emphasized that this critical value is reached when the waiting value is identical to or less than

78
Akita University

the executing value (Dixit & Pindyck, 1993). However, the uncertainties created distance
between the two values. Mun stated that the company's behavior determined the risk of a project
(Mun, 2006). Therefore, the critical value may vary among decision-makers. Although both
methods recommended immediate execution of the project, SSA improved the results of the
traditional method. The traditional method recommended executing the project when the
present value exceeded zero. This improvement caused the RO methodology to recommend the
delay option under several other conditions.

This study further demonstrates the application of RO in the mining industry. It

successfully combines the advancement of geostatistics in terms of conditional simulation and
real option valuation through the SSA approach. The current limitation concerning real options,
which include mathematical complexity and applicability (Haque et al., 2014), is solved using
our methodology. Monte Carlo simulation solves the mathematical problem while
incorporating grade uncertainty to determine the application, particularly mining. Capturing
additional uncertainty increases the future applicability of the RO. Each modifying factor in the
conversion resource to be reserved and from geological confidence (The JORC Code 2012
Edition, 2012) is supposed to be modeled further to obtain proper project value and ultimately
guarantee a successful project with calculated risk.

79
Akita University

4.7. References

Adachi, T., Mogi, G., & Adachi, T. (2008). Real Option Analysis of Multi-Stage Investment
on Resources Development Project by Improved SSA Method. Mining and Materials
Processing Institute of Japan, 124(9), 576–582.

Taylor, M. (2009). What is sensitivity analysis. Consortium YHE: University of York, April
2009, 1–8.

Dixit, A. K., & Pindyck, R. S. (1993). Investment under uncertainty. Princeton University Press.

Jurdziak, L., & Wiktorowicz, J. (2008). Conditional and Monte Carlo Simulation - the Tools
for Risk Identification in Mining Projects. In E. J. Sobczyk & J. Kicki (Eds.), Economic
Evaluation and Risk Analysis of Mineral Projects (pp. 61–72). Taylor & Francis.

Dowd, P. A. (1994). Risk assessment in reserve estimation and open-pit planning. Transactions
- Institution of Mining & Metallurgy, Section A, 103. https://doi.org/10.1016/0148-
9062(95)97056-o

Cortazar, G., Schwartz, E. S., & Casassus, J. (2001). Optimal exploration investments under
price and geological-technical uncertainty: A real options model. R and D Management,
31(2), 181–189. https://doi.org/10.1111/1467-9310.00208

Deutsch, C. V., & Journel, A. G. (1997). GSLIB: geostatistical software library and user’s
guide. Second edition. In GSLIB: geostatistical software library and user’s guide. Second
edition (Second Edi). Oxford University Press.

Dominy, S. C. (2016). Errors and Uncertainty in Mineral Resource and Ore Reserve Estimation
: The Importance of Getting it Right Errors and Uncertainty in Mineral Resource and Ore
Reserve Estimation : The Importance of Getting it Right. January 2002.
https://doi.org/10.2113/11.1-4.77

Hill, R. C., Griffiths, W. E., & Lim, G. C. (2011). Principles of Econometrics (4th ed.). John
Wiley & Sons, Inc.

Nicholas, G. (2014). An Integrated Risk Evaluation Model for Mineral Deposits. The
University of Adelaide.

80
Akita University

Adachi, T., Mogi, G., & Adachi, T. (2008). Real Option Analysis of Multi-Stage Investment
on Resources Development Project by Improved SSA Method. Mining and Materials
Processing Institute of Japan, 124(9), 576–582.

Barraquand, J., & Martineau, D. (1995). Numerical Valuation of High Dimensional

Multivariate American Securities. The Journal of Financial and Quantitative Analysis,
30(3), 383–405. https://doi.org/https://doi.org/10.2307/2331347

Mun, J. (2006). Real options analysis: tools and techniques for valuing strategic investments
and decisions. In John Wiley & Sons, Inc (2nd ed.). John Wiley & Sons, Inc.

Sefemo, D. F., & Adachi, T. (2015). Management Flexibility and Its Real Option Pricing in
Expansion Project of a Metal Mine: Actual Case Study in Selibe Phikwe Mine , Botswana ,
Africa. April, 922069.

Crundwell, F. (2008). Finance for Engineers: Evaluation and Funding of Capital Projects.
Springer-Verlag London Limited. https://doi.org/10.1007/978-1-84800-033-9

Yin, J., Ng, S. H., & Ng, K. M. (2011). Kriging metamodel with modified nugget-effect: The
heteroscedastic variance case. Computers and Industrial Engineering, 61(3), 760–777.
https://doi.org/10.1016/j.cie.2011.05.008

Cressie, N. A. C. (1993). Statistics for spatial data revised edition. In John Wiley & Sons, Inc.
John Wiley & Sons, Inc.

Mun, J. (2006). Real options analysis: tools and techniques for valuing strategic investments
and decisions. In John Wiley & Sons, Inc (2nd ed.). John Wiley & Sons, Inc.

Dixit, A. K., & Pindyck, R. S. (1993). Investment under uncertainty. Princeton University Press.

Haque, M. A., Topal, E., & Lilford, E. (2014). A numerical study for a mining project using
real options valuation under commodity price uncertainty. Resources Policy, 39(1), 115–
123. https://doi.org/10.1016/j.resourpol.2013.12.004

The JORC code 2012 Edition. (2012). https://www.jorc.org/docs/JORC_code_2012.pdf

81
Akita University

82
Akita University

Chapter 5 Conclusions

In summary, we had constructed a valuation method using the developed MSSA

methodology. The method is adapted from well-known financial theory, namely option
analysis, which have similarities in analyzing risk in the real business. An option is a financial
derivative whose value depends on the price of a stock. Option holders have rights but no
obligation to execute their option in the maturity date. These alternatives are quantified by
option calculation to justify option value. In this thesis, we applied it at a mining project in
Indonesia, particularly production stage of underwater tin mine planning. The objective of the
valuation is giving recommendation to mine company whether execute the project, delay or
abandon it when the condition is not favorable. We addressed the risk of price, reserve, and cost
which represented economical, geological and technical uncertainties. Our established
valuation approach addressed limitations in RO applications, particularly concerning assessing
several sources of uncertainty in a multi-stage short-term mining project. The application
exhibited practical simplicity as well as applicability. We enhanced previous studies by
employing the adaptation of spatial uncertainty through conditional simulation (CS) and the
construct path generation algorithm to incorporate the three uncertainties.

The mining project had been planned using GEOVIA Surpac and Whittle Mining
Software based on real tin exploration data. The assumed price was at January 2020 while cost
is calculated at the average historical data in 5 years with similar mining method and equipment
when the project was planned. The obtained present value for twos mining areas was 7.5 million
USD. Uncertainty in mining could be modeled using the real option and our result was 17.6
million USD with price, grade and cost uncertainties driver. Thus, the MSSA approach in real
option methodology evaluated multi-uncertainty and multi-stage mining (rainbow option). Our
developed algorithm could produce real option value using JAVA Programming as a simulation
tool.

83
Akita University

Generally, a mining company could not control the price and reserve uncertainties. We
used the Geometric Brownian model to account for price and reserve uncertainties. While for
cost uncertainty, the Mean Reverting model gave a better result. Price would be revealed when
the company decided to mine. At the same time, the reserve will be proven by the production.
The increasing of exploration drill data theoretically could decrease the uncertainty. However,
a huge exploration expenditure would be the consequence.

Logarithmic natural growth analysis based on historical data resulted in the price has an
average monthly growth rate (α) of 0.74%. However, to be applied at our evaluation method,
we assumed the price growth would follow the stochastic differential equation model,
particularly the one-factor constant convenience yield model for risk-neutral prices. Assumed
parameters were risk-free rate of 0.25%, convenience yield 0.17% and calculated price
volatility of 4.7%.

This research started to utilize the CS method to quantify the reserve uncertainty in real
options valuation. CS is a geostatistical tool that can generate punctually or block ‘realizations’
of mineral grades. Each realization is intended to honor the histogram and semivariogram of
the true grade distribution and honor known data points. It is a technique to measure spatial
uncertainty based on spatial Monte Carlo Analysis (MCS). CS reproduces the properties of an
ore body and quantifies the variability, which we call “spatial uncertainty.” We incorporated
the advanced geostatistical application as well as advanced economic valuation methodology.

We measure the technical uncertainty represented by cost volatility. The cost consisted
of mining, processing and metallurgical cost. Unit cost of production model construction was
based on historical cost given by the company with a similar location and production technique.
Similar ADF methodology was applied to the summarized historical unit cost data. The unit
root test concluded that the cost data is stationer. Therefore, the construction of the technical
uncertainty model was used the Mean Reverting (MR) model. We used the simplest mean-
reverting process-also known as the Ornstein-Uhlenbeck process.

A mining company does not have control over price and reserve. Price will always
fluctuate in the future, which depends on the supply and demand of the commodity. On the
other hand, the reserve is a spatial base parameter. It will be known when mining the reserve.
We simulated the combination relative shifting and volatility of price, reserve and cost. The
real option result opposed traditional Discounted Cash Flow in several condition particularly in
the waiting decision condition. In real option method, option premium is the decision
84
Akita University

determinant. Our 5% threshold assumption recommended to execute the project when the
company can eliminate cost volatility. The condition resulted option premium 3.8% which
dropped from 135% when cost uncertainty is existing. The simulation revealed that cost
uncertainty of 39% was the dominant factor of project valuation using real options methodology
comparing to price uncertainty (4.7%) and grade uncertainty (9% and 16%). The company
should ensure cost planning before executing the project to make the project profitable.

There are many uncertainties in the mining industry. The most significant uncertainties
are economical, geological, and technical, represented by price, grade, and cost. The traditional
method could not explain those uncertainties, and the previous real options method did not
incorporate the uncertainties. This research is a pilot project for combining uncertainties
simultaneously through the MSSA approach. Practically, a mining company can execute at
calculated risk or wait to resolve uncertainty or abandon the project to save from loss.
Additionally, we cannot perform standard real options approaches Black-Scholes and lattice
method to evaluate multi uncertainties. These methodologies resulted only in price risk
measurement by project. They cannot assess an ongoing project and multi uncertainties. In
contrast, the MSSA could capture more permutation of the cash flow parameters. Nonetheless,
the MSSA needed adaptation to model each uncertainty faced by the mining company and high
computation power to simulate.

We simulated further for changing parameter of uncertainty. Price, grade and cost is
assumed as control variable to get critical value. Moreover, those uncertainty is shifted to obtain
profitable project. The uncertainty alteration does not change the decision using a traditional
methodology represented by the present value. The simulation has proven that cost uncertainty
assurance is the key to project value using MSSA methodology. Additionally, the changing
price and reserve are related to the present and option value shift identically. However, it made
a pattern with the constant of option premium.

The current limitation for RO, including mathematical complexity and applicability, is
solved using our methodology. SSA simulation solves the mathematical problem while
incorporating grade uncertainty in the simulation to determine the project value and risk.
However, the SSA methodology needs optimization on number of simulation and modeling
methods. Balancing calculation time and consistency of result is the key of method limitation.
Furthermore, comparing to Black-Scholes and Binomial Lattice method, only SSA can be
performed on the case study. Adding capital assumption and eliminating grade uncertainty, they

85
Akita University

have similar result with difference less than 1%. Nonetheless, capturing additional uncertainty
increases the future applicability of the RO. Each modifying factor in the conversion resource
to be reserved and from geological confidence is supposed to be modeled further to obtain
proper project value and ultimately guarantee a successful project with calculated risk.

This study presented a project that combined uncertainties and multi projects in mining
ventures. We considered pricing and geological factors in two mining locations, whereas other
modifying factors were assumed to be stable. Owing to its ability to capture real-world
parameters, the utilization of the proposed RO methodology in the mining industry is expected
to increase in future. In terms of future development, the valuation method could be further
enhanced to combine additional uncertainties, project valuation, and stages of investment. The
outcome parameter in cash flow in terms of mining expenditure may present an additional
application of the proposed methodology. The method can be developed further to combine
more uncertainties, project valuation, and more stages of investment. The utilization of real
options methodology in the mining industry is expected to increase by capturing the real
uncertainty as stated at modifying factor to convert resource to be reserve. Finally, uncertainties
could be overcome by management's flexibility to either hold or abandon the project.

86
Akita University

APPENDIX
A. Price Modelling

Price Log Price

Date Cash Change ƩUi Ui2 ƩUi2 n s Δt ơ Notes
($/tonne) (Ui)
Monthly
1/1/2016 14,906.00 0.0693 0.4423 0.0048 0.1406 61 0.047448 1 4.74%
volatility
2/1/2016 15,976.00 0.0461 0.0021
3/1/2016 16,729.00 0.0312 0.0010
4/1/2016 17,260.00 -0.0558 0.0031
5/1/2016 16,324.00 0.0450 0.0020
6/1/2016 17,075.00 0.0438 0.0019
7/1/2016 17,840.00 0.0568 0.0032
8/1/2016 18,882.00 0.0657 0.0043
9/1/2016 20,164.00 0.0351 0.0012
10/1/2016 20,885.00 0.0206 0.0004
11/1/2016 21,320.00 -0.0054 0.0000
12/1/2016 21,205.00 -0.0694 0.0048
1/1/2017 19,783.00 -0.0297 0.0009
2/1/2017 19,205.00 0.0520 0.0027
3/1/2017 20,230.00 -0.0119 0.0001
4/1/2017 19,990.00 0.0220 0.0005
5/1/2017 20,435.00 -0.0101 0.0001
6/1/2017 20,230.00 0.0275 0.0008
7/1/2017 20,795.00 0.0037 0.0000
8/1/2017 20,873.00 -0.0013 0.0000
9/1/2017 20,845.00 -0.0664 0.0044
10/1/2017 19,505.00 0.0094 0.0001
11/1/2017 19,690.00 0.0204 0.0004
12/1/2017 20,096.00 0.0839 0.0070
87
Akita University

Price Log Price

Date Cash Change ƩUi Ui2 ƩUi2 n s Δt ơ Notes
($/tonne) (Ui)
1/1/2018 21,855.00 -0.0099 0.0001
2/1/2018 21,640.00 -0.0217 0.0005
3/1/2018 21,175.00 0.0066 0.0000
4/1/2018 21,315.00 -0.0327 0.0011
5/1/2018 20,630.00 -0.0396 0.0016
6/1/2018 19,830.00 0.0172 0.0003
7/1/2018 20,175.00 -0.0572 0.0033
8/1/2018 19,052.50 -0.0102 0.0001
9/1/2018 18,860.00 0.0146 0.0002
10/1/2018 19,138.00 -0.0394 0.0016
11/1/2018 18,398.00 0.0592 0.0035
12/1/2018 19,520.00 0.0697 0.0049
1/1/2019 20,930.00 0.0383 0.0015
2/1/2019 21,747.00 -0.0139 0.0002
3/1/2019 21,447.00 -0.0829 0.0069
4/1/2019 19,741.00 -0.0475 0.0023
5/1/2019 18,825.00 0.0004 0.0000
6/1/2019 18,833.00 -0.0840 0.0071
7/1/2019 17,315.00 -0.0573 0.0033
8/1/2019 16,350.00 -0.0273 0.0007
9/1/2019 15,910.00 0.0358 0.0013
10/1/2019 16,490.00 0.0008 0.0000
11/1/2019 16,504.00 0.0400 0.0016
12/1/2019 17,178.00 -0.0448 0.0020
1/1/2020 16,425.00 -0.0097 0.0001
2/1/2020 16,267.00 -0.1035 0.0107
3/1/2020 14,667.00 0.0406 0.0016
4/1/2020 15,274.00 0.0149 0.0002
5/1/2020 15,502.85 0.0815 0.0066
6/1/2020 16,819.00 0.0628 0.0039

88
Akita University

Price Log Price

Date Cash Change ƩUi Ui2 ƩUi2 n s Δt ơ Notes
($/tonne) (Ui)
7/1/2020 17,909.00 -0.0202 0.0004
8/1/2020 17,550.00 -0.0058 0.0000
9/1/2020 17,448.75 0.0157 0.0002
10/1/2020 17,724.00 0.0505 0.0025
11/1/2020 18,642.00 0.0972 0.0094
12/1/2020 20,544.50 0.1215 0.0148
1/1/2021 23,199.00 0.0000
Average 18,870.06 0.00737242
Data Standard
Notes Drift rate Period Variance
Amount Deviation

89
Akita University

B. Resource Summary

South Resource Summary

Grade Range Minimum Grade Maximum Grade Expected Grade Std dev
Tonnes Cummulative Volume (m3) Tonnes
(kg/ ton) (kg/ ton) (kg/ ton) (kg/ ton) (kg/ ton)
0.0 -> 0.1 0.01 0.138 0.03 0.01 1,205,250 128,375 385,125
0.1 -> 0.2 0.01 0.627 0.143 0.08 820,125 35,000 105,000
0.2 -> 0.3 0.01 1.084 0.253 0.16 715,125 8,000 24,000
0.3 -> 0.4 0.06 1.825 0.357 0.19 691,125 11,625 34,875
0.4 -> 0.5 0.09 1.144 0.449 0.17 656,250 14,625 43,875
0.5 -> 0.6 0.15 1.239 0.555 0.17 612,375 29,000 87,000
0.6 -> 0.7 0.22 1.436 0.645 0.20 525,375 33,625 100,875
0.7 -> 0.8 0.23 1.597 0.749 0.24 424,500 18,625 55,875
0.8 -> 0.9 0.23 2.244 0.846 0.34 368,625 13,750 41,250
0.9 -> 1.0 0.23 2.614 0.949 0.36 327,375 10,750 32,250
1.0 -> 1.1 0.28 2.518 1.052 0.34 295,125 11,750 35,250
1.1 -> 1.2 0.36 2.423 1.155 0.34 259,875 13,375 40,125
1.2 -> 1.3 0.36 2.605 1.246 0.37 219,750 13,875 41,625
1.3 -> 1.4 0.26 2.845 1.351 0.39 178,125 15,625 46,875
1.4 -> 1.5 0.48 2.952 1.444 0.42 131,250 12,625 37,875
1.5 -> 1.6 0.35 3.15 1.549 0.45 93,375 9,625 28,875
1.6 -> 1.7 0.39 3.642 1.646 0.50 64,500 9,000 27,000
1.7 -> 1.8 0.20 4.308 1.747 0.67 37,500 4,500 13,500
1.8 -> 1.9 0.11 5.757 1.84 0.86 24,000 4,625 13,875
1.9 -> 2.0 0.04 5.908 1.947 1.11 10,125 2,875 8,625
2.0 -> 2.1 0.02 7 2.031 1.28 1,500 500 1,500
Grand Total 0.328 0.905 0.578 0.09 401,750 1,205,250

90
Akita University

Mid Resource Summary

Grade Range Minimum Grade Maximum Grade Expected Grade Std dev
Tonnes Cummulative Volume (m3) Tonnes
(kg/ ton) (kg/ ton) (kg/ ton) (kg/ ton) (kg/ ton)
0.0 -> 0.1 0.02 0.08 0.04 0.01 11,503,500 964,500 2,893,500
0.1 -> 0.2 0.06 0.28 0.15 0.03 8,610,000 484,500 1,453,500
0.2 -> 0.3 0.10 0.46 0.25 0.05 7,156,500 332,000 996,000
0.3 -> 0.4 0.18 0.60 0.35 0.07 6,160,500 308,500 925,500
0.4 -> 0.5 0.23 0.71 0.45 0.08 5,235,000 258,000 774,000
0.5 -> 0.6 0.30 0.88 0.55 0.10 4,461,000 227,500 682,500
0.6 -> 0.7 0.33 1.10 0.65 0.12 3,778,500 151,000 453,000
0.7 -> 0.8 0.38 1.30 0.75 0.13 3,325,500 160,000 480,000
0.8 -> 0.9 0.49 1.38 0.85 0.14 2,845,500 169,500 508,500
0.9 -> 1.0 0.54 1.51 0.95 0.15 2,337,000 177,000 531,000
1.0 -> 1.1 0.53 1.73 1.05 0.18 1,806,000 138,500 415,500
1.1 -> 1.2 0.57 1.84 1.15 0.19 1,390,500 105,000 315,000
1.2 -> 1.3 0.61 1.92 1.25 0.22 1,075,500 90,000 270,000
1.3 -> 1.4 0.76 2.10 1.35 0.24 805,500 97,000 291,000
1.4 -> 1.5 0.72 2.39 1.45 0.27 514,500 84,500 253,500
1.5 -> 1.6 0.59 2.71 1.54 0.35 261,000 44,500 133,500
1.6 -> 1.7 0.33 3.69 1.64 0.53 127,500 16,500 49,500
1.7 -> 1.8 0.13 4.14 1.74 0.66 78,000 8,500 25,500
1.8 -> 1.9 0.07 5.61 1.85 0.91 52,500 6,000 18,000
1.9 -> 2.0 0.03 5.24 1.96 1.03 34,500 4,500 13,500
2.0 -> 2.1 0.02 6.42 2.04 1.18 21,000 3,500 10,500
2.1 -> 2.2 0.00 7.00 2.12 1.88 10,500 1,000 3,000
2.2 -> 2.3 - 7.00 2.27 2.71 7,500 500 1,500
2.4 -> 2.5 - 7.00 2.43 2.77 6,000 500 1,500

91
Akita University

Grade Range Minimum Grade Maximum Grade Expected Grade Std dev
Tonnes Cummulative Volume (m3) Tonnes
(kg/ ton) (kg/ ton) (kg/ ton) (kg/ ton) (kg/ ton)
2.6 -> 2.7 - 7.00 2.66 2.79 4,500 500 1,500
2.7 -> 2.8 - 7.00 2.74 2.89 3,000 500 1,500
3.0 -> 3.1 - 7.00 3.08 2.84 1,500 500 1,500
Grand Total 0.37 0.60 0.48 0.04 3,834,500 11,503,500

C. Reserve Summary

South Reserve Summary

Grade Range Minimum Grade Maximum Grade Expected Grade Std dev
Tonnes Cummulative Volume (m3) Tonnes
(kg/ ton) (kg/ ton) (kg/ ton) (kg/ ton) (kg/ ton)
0.0 -> 0.1 0.01 0.14 0.03 0.01 1,123,125 128,375 385,125
0.1 -> 0.2 0.01 0.63 0.14 0.08 738,000 35,000 105,000
0.2 -> 0.3 0.01 1.06 0.25 0.16 633,000 7,875 23,625
0.3 -> 0.4 0.06 1.83 0.36 0.19 609,375 11,625 34,875
0.4 -> 0.5 0.09 1.16 0.45 0.18 574,500 14,125 42,375
0.5 -> 0.6 0.13 1.27 0.55 0.18 532,125 25,375 76,125
0.6 -> 0.7 0.18 1.69 0.65 0.21 456,000 24,250 72,750
0.7 -> 0.8 0.20 1.82 0.75 0.26 383,250 13,375 40,125
0.8 -> 0.9 0.20 2.30 0.84 0.35 343,125 10,500 31,500
0.9 -> 1.0 0.22 2.63 0.95 0.37 311,625 10,375 31,125
1.0 -> 1.1 0.26 2.54 1.05 0.34 280,500 11,625 34,875
1.1 -> 1.2 0.36 2.42 1.16 0.34 245,625 13,375 40,125
1.2 -> 1.3 0.36 2.61 1.25 0.37 205,500 13,875 41,625
1.3 -> 1.4 0.26 2.85 1.35 0.39 163,875 15,625 46,875
1.4 -> 1.5 0.48 2.94 1.44 0.42 117,000 11,875 35,625
92
Akita University

Grade Range Minimum Grade Maximum Grade Expected Grade Std dev
Tonnes Cummulative Volume (m3) Tonnes
(kg/ ton) (kg/ ton) (kg/ ton) (kg/ ton) (kg/ ton)
1.5 -> 1.6 0.37 3.47 1.55 0.49 81,375 8,250 24,750
1.6 -> 1.7 0.27 3.75 1.65 0.53 56,625 7,875 23,625
1.7 -> 1.8 0.21 4.27 1.75 0.68 33,000 4,125 12,375
1.8 -> 1.9 0.11 6.20 1.84 0.90 20,625 4,250 12,750
1.9 -> 2.0 0.04 6.00 1.95 1.17 7,875 2,500 7,500
2.0 -> 2.1 - 7.00 2.03 2.50 375 125 375
Grand Total 0.31 0.88 0.56 0.09 374,375 1,123,125

Mid Reserve Summary

Grade Range Minimum Grade Maximum Grade Expected Grade Std dev
Volume (m3) Tonnes Tonnes Cummulative
(kg/ ton) (kg/ ton) (kg/ ton) (kg/ ton) (kg/ ton)
0.0 -> 0.1 0.01 0.10 0.05 0.02 472,500 1,417,500 7,204,500
0.1 -> 0.2 0.05 0.37 0.14 0.05 220,500 661,500 5,787,000
0.2 -> 0.3 0.09 0.54 0.25 0.07 184,500 553,500 5,125,500
0.3 -> 0.4 0.14 0.64 0.35 0.08 193,000 579,000 4,572,000
0.4 -> 0.5 0.18 0.83 0.45 0.11 141,500 424,500 3,993,000
0.5 -> 0.6 0.25 1.03 0.55 0.12 149,000 447,000 3,568,500
0.6 -> 0.7 0.22 1.17 0.65 0.15 89,500 268,500 3,121,500
0.7 -> 0.8 0.34 1.37 0.75 0.15 116,500 349,500 2,853,000
0.8 -> 0.9 0.47 1.45 0.86 0.15 131,500 394,500 2,503,500
0.9 -> 1.0 0.52 1.56 0.95 0.16 156,500 469,500 2,109,000
1.0 -> 1.1 0.54 1.73 1.05 0.19 125,500 376,500 1,639,500
1.1 -> 1.2 0.59 1.93 1.15 0.21 92,000 276,000 1,263,000
1.2 -> 1.3 0.61 1.95 1.25 0.23 85,000 255,000 987,000
1.3 -> 1.4 0.67 2.16 1.35 0.25 87,000 261,000 732,000
93
Akita University

Grade Range Minimum Grade Maximum Grade Expected Grade Std dev
Volume (m3) Tonnes Tonnes Cummulative
(kg/ ton) (kg/ ton) (kg/ ton) (kg/ ton) (kg/ ton)
1.4 -> 1.5 0.69 2.34 1.45 0.27 75,500 226,500 471,000
1.5 -> 1.6 0.59 2.74 1.54 0.35 41,000 123,000 244,500
1.6 -> 1.7 0.34 3.67 1.64 0.54 16,000 48,000 121,500
1.7 -> 1.8 0.09 4.23 1.74 0.71 7,500 22,500 73,500
1.8 -> 1.9 0.07 5.61 1.85 0.91 6,000 18,000 51,000
1.9 -> 2.0 0.03 5.24 1.96 1.03 4,500 13,500 33,000
2.0 -> 2.1 0.02 6.58 2.04 1.28 3,000 9,000 19,500
2.1 -> 2.2 0.00 7.00 2.12 1.88 1,000 3,000 10,500
2.2 -> 2.3 - 7.00 2.27 2.71 500 1,500 7,500
2.4 -> 2.5 - 7.00 2.43 2.77 500 1,500 6,000
2.6 -> 2.7 - 7.00 2.66 2.79 500 1,500 4,500
2.7 -> 2.8 - 7.00 2.74 2.89 500 1,500 3,000
3.0 -> 3.1 - 7.00 3.08 2.84 500 1,500 1,500
Grand Total 0.44 0.77 0.05 2,401,500 7,204,500

94
Akita University

D. Cost Calculation

Operating Cost

Standard
Unit Year 2015 2016 2017 2018 2019 2020 Average
Deviation
USD Safety Cost 1,869,484 1,143,824 286,234 811,732 1,145,695 1,039,192 1,049,360 470,769
USD Labor cost 1,000,608 1,023,419 1,179,021 583,316 1,095,567 850,748 955,446 193,960
USD Materials 529,372 349,770 84,216 135,668 441,541 198,217 289,797 162,494
USD Depreciation 1,491,945 1,408,482 1,681,748 1,724,791 3,101,246 2,824,272 2,038,748 666,858
USD Others 95,564 64,943 47,940 37,105 661,762 230,894 189,701 220,736
USD Total 4,986,973 3,990,438 3,279,161 3,292,613 6,445,812 5,143,323 4,523,053 1,128,560
Ton Sn Production 199 253 75 40 274 383 204 118
USD/ton Sn COGS 25,116 15,772 98,734 82,315 23,525 13,429 43,149 34,075
USD/ton ore COGS 1.84 1.48 0.92 2.74 2.21 1.46 1.78 0.59
General Cost

Standard
Unit Year 2015 2016 2017 2018 2019 2020 Average
Deviation
Ore Tin
Ton Sn 26,361 24,347 31,035 44,514 82,522 38,235 41,169 19,727
Production
Sales
MT 30,087 26,677 29,914 33,818 64,801 50,235 39,255 13,737
Volume
Overhead
Usd/Ton 2,565 2,877 2,461 1,936 1,294 2,440 2,262 514
Cost
Smelting
USD/MT and Refining 731 701 668 1,084 857 1,000 840 156
Cost
USD/MT Sales cost 1,587 1,424 1,738 1,689 1,790 1,503 1,622 130

95