Using Dynamic Discounted Cash Flow

and Real Option Methods for Economic

Analysis in NI43-101 Technical Reports
M Samis1,5, L Martinez2, G A Davis3 and J B Whyte4,6

ABSTRACT
The introduction of dynamic discounted cash flow (‘dynamic DCF’) and real options (‘RO’) analysis
into a NI43-101 report is an intriguing new avenue for improving the communication of project
risk and the understanding of how a project’s risk profile influences project economics. However,
these techniques have not found wide-spread use in NI43-101 and other public technical reports
to date. This is beginning to change as in June 2010 Ivanhoe and Entree Gold both published
NI43-101 reports for the Oyu Tolgoi project containing dynamic DCF and RO analysis. In addition,
Canadian securities regulatory authorities have recently published amendments to their NI43-101
rules to make it clear companies can provide more sophisticated economic and risk analyses in
these reports.
This paper proposes an extended valuation and risk management framework that includes these
methods and discusses how they can be used to support the objectives of a NI43-101 technical
report. It highlights how these techniques can complement and extend conventional evaluation
analyses through a review of the dynamic DCF and RO analysis contained in the Ivanhoe NI43-
101 report. A set of best practice guidelines are then proposed to ensure that dynamic DCF and
RO analysis is presented in a professional and objective manner that benefits investors and is
consistent with a conventional DCF analysis.

INTRODUCTION
Public mining companies have the responsibility of informing public technical reports issued under NI43-101 rules (‘NI43-
their investors about significant new information affecting 101 Technical Reports’).
their exploration and development stage projects. The release An important focus of NI43-101 Technical Reports is the
of new information to investors related to the quality and technical information generated by exploration activities and
quantity of the mineral resource is controlled by securities how this information is processed to arrive at an estimate of
regulations. These regulations include the National Instrument the quality and quantity of mineral resource. Reporting on this
43-101 Standards of Disclosure for Mineral Projects (‘NI43- technical information includes discussing the characteristics
101’; Canadian Securities Administrators, 2005, 2011) rules of the exploration program, the structural interpretation
in Canadian financial markets, the Australasian Code for of geological samples, sampling and assaying techniques,
Reporting of Exploration Results, Mineral Resources and statistical analysis methods, and the approach to resource
Ore Reserves (JORC, 2004) prepared by the Australasian modelling. A key requirement for NI43-101 Technical
Joint Ore Reserves Committee and used in Australian Report technical sections is to ensure that investors have the
financial markets, and the South African Code for Reporting information available to judge the robustness of the resource
of Exploration Results, Mineral Resources and Mineral quality and quantity estimate.
Reserves (South African Mineral Resource Committee, 2009) A further requirement under NI43-101 guidelines is that the
used in South African financial markets. These codes all have a technical report must inform the investor about the economic
common goal of ensuring that new information about mineral viability of the resource given a suitable project design and
resources is released through public technical reports in a using reasonable assumptions about the current and future
manner that is transparent, professional, and of a consistent economic environment. This is an important requirement
standard for all public companies. In this paper, we focus on as it underlies the actual definition of a mineral resource: a

1. Vice President, Valuation and Business Modelling, Ernst & Young LLP, 222 Bay Street, PO Box 251, Toronto ON M5K 1J7, Canada. Email: [email protected]
2. MAusIMM, Principal and Manager, ROMPEV Pty Ltd, 15 Riverview Terrace, Indooroopilly Qld 4068. Email: [email protected]
3. FAusIMM, Professor, Division of Economics and Business, Colorado School of Mines, 1500 Illinois Street, Golden CO 80401-1887, USA. Email: [email protected]
4. Senior Geologist, Corporate Finance Branch, Ontario Securities Commission, Suite 1903, 20 Queen Street West, Toronto ON M5H 3S8, Canada. Email: [email protected]
5. The statements made and opinions expressed in this paper are those of the author and do not represent an official position of Ernst and Young LLP.
6. The statements and views contained within this paper are those of the author in his private capacity and should not be taken to represent an official view of the Ontario Securities Commission,
the Commission members or staff.

VALMIN SEMINAR SERIES / PERTH, WA, 18 OCTOBER 2011 / BRISBANE, QLD, 17 APRIL 2012 1
M SAMIS et al

mineral occurrence can only be declared a resource if it can be and sliding scale royalty rates may change with variation
demonstrated that there is a reasonable prospect of economic in metal prices, leading to changes in cash flow structure.
extraction. A static DCF model is unable to recognise the contingent
It is notable that the analysis supporting conclusions nature of cash flows and so may provide an erroneous
about the possibility of economic extraction is often not estimate of project expected net cash flow. A cash flow
performed to the same level of sophistication as the technical estimation method that recognises contingent cash flow
analysis supporting conclusions about the quantity and structures and corrects this error may be necessary to
quality of mineral resources. This is unfortunate because generate a reasonable cash flow estimate.
there are numerous techniques that may be used to extend 3. Static DCF risk adjustments do not recognise the
the conventional economic analysis provided in NI43-101 dynamic variation of cash flow risk through time: the
Technical Reports such that an investor is provided with a use of a single discount rate implies that project cash flow
better understanding of a project’s economic prospects. uncertainty increases through time in a regular manner.
However, most mine valuation professionals would agree
that the cash flow uncertainty changes in a dynamic and
CURRENT ECONOMIC ANALYSIS TECHNIQUES
erratic manner due to changes in metal grades and prices,
IN NI43-101 TECHNICAL REPORTS operating costs, mining method, exhaustion of tax shields,
The predominant economic analysis technique used in NI43- and tax and royalty rates among other things. A risk
101 technical reports is the static discounted cash flow (static adjustment method that responds to changes in cash flow
DCF) method. This technique estimates the future net cash uncertainty would be preferred.
flows from the project using annual single-point forecasts The economic analysis provided in a NI43-101 technical
of production and economic variables such as future metal report can be extended with better numerical methods and
prices, waste and ore production amounts, metal grades, concepts from finance theory to correct the problems with a
recoveries, prices and amounts of consumables, labour, and Static DCF analysis. This assertion does not mean that static
services. These forecasts are then used to construct an annual DCF models should not be used in NI43-101 technical reports.
expected project net cash flow equal to revenues less capital Indeed, it is most likely that static DCF analysis will continue
and operating costs, government and third party royalties, to be the predominant method employed in these reports,
corporate income taxes, cost of transporting output to with more involved numerical and financial analysis being
market, insurance, smelting and refining charges, and other only used when there is an important project characteristic
deductions. that would not be fully recognised by a static DCF model.
This expected net cash flow is then used to calculate a project However, it is important for mining professionals to realise
internal rate of return (IRR) or net present value (NPV) as an that NI43-101 rules do not prohibit the use of more advanced
indication of project viability. economic analysis methods in these technical reports. The
Project IRR represents the annualised effective compounded only restriction is that analytic techniques being used are
return rate on capital invested in the project. This is also the considered acceptable by industry professionals and that the
discount rate at which the NPV of expected project cash flows analysis is conducted by someone with professional experience
is zero. Generally, a project with an IRR that is higher than in the proposed analysis. In Canada, select advanced
its opportunity cost of capital is thought to be an attractive valuation techniques such as RO NPV have been recognised as
investment. professionally acceptable by the Canadian Institute of Mining,
Project NPV represents the value created or destroyed by Metallurgy, and Petroleum in their Standards and Guidelines
investing in a project. The calculation of a Static DCF NPV for the Valuation of Mineral Properties (‘CIMVal Standards’,
requires estimating net annual cash flows and then discounting Canadian Institute of Mining, Metallurgy, and Petroleum,
each annual cash flow for the value effects of uncertainty and Special Committee on Valuation of Mineral Properties, 2003)
time to determine a cash flow present value. NPV is the sum and so may be used when necessary for economic analysis in
of these present values. The value effect of uncertainty and NI43-101 Technical Reports. Indeed, the main challenge for
time is recognised by summarising their impact into a single the inclusion of these techniques in these reports is finding
constant risk-adjusted rate that is used in the discounting qualified professionals to conduct the analysis.
process. This discount rate is likely used for a broad class There have been recent changes to NI 43-101 rules in order
of investment projects regardless of the actual uncertainty to accommodate analysis that goes beyond a static DCF model
characteristics of the particular project. A positive static DCF in scope. The changes are largely in form F1, the form that
NPV suggests that a project is an attractive investment. specifies the content of the technical report. Formerly the
There are three potential shortcomings of a Static DCF NPV technical report form specified annual cash flow forecasts –
calculation: a requirement that in practice was met by a standard static
DCF calculation – and ‘sensitivity analyses with variants in
1. Static DCF ignores randomness in cash flow variables: metal prices, grade, capital and operating costs’ (Canadian
the reliance on only the expectations of uncertain cash Securities Administrators, 2005). The form also called for a
flow variables such as metal price excludes a more detailed discussion of the payback period and of expected mine life.
description of their randomness in the cash flow analysis. The requirement applied to properties in production and
Static DCF analysis may use sensitivity analysis that varies to ‘development properties’ – defined under the old rule as
a single variable at a time by a set percentage to gain properties being prepared for production whose economic
insight into how NPV changes with random outcomes for viability has been demonstrated by a feasibility study. While
that variable. However, this type of analysis is very limited it was possible to argue that a project at the pre-feasibility
compared to the insights into cash flow randomness that stage, or in earlier stages of economic assessment, was not
can be done with more complete methods of analysis. a ‘development property’, the rule’s requirement that a
2. Static DCF ignores the effects of contingent cash flows and technical report ‘not omit any material scientific and technical
flexibility: projects incorporate contingencies that cause information’ meant that any technical report supporting
the structure of cash flow to change with variations in the disclosure of economic information should include a similar
project environment. For example, both production policy discussion of the project’s economics.

2 VALMIN SEMINAR SERIES / PERTH, WA, 18 OCTOBER 2011 / BRISBANE, QLD, 17 APRIL 2012
 USING DYNAMIC DISCOUNTED CASH FLOW AND REAL OPTION METHODS FOR ECONOMIC ANALYSIS

Under the changes to form F1 (now grouped under the AN EXTENDED EVALUATION AND RISK
form’s item 22), the technical report must include annual
MANAGEMENT FRAMEWORK
cash flow forecasts and an annual production schedule,
using either mineral reserves or, in the case of preliminary This section provides a brief overview of an extended
economic assessments, mineral resources. There must also evaluation and risk management framework (‘Extended
be a discussion of NPV, IRR, and the payback period. The Evaluation Framework’) for mine project evaluation that
uncertainties in the Static DCF economic analysis are covered might be included in a technical report. This framework
under the form’s item 22 (e), which requires the technical divides the cash flow modelling process into two parts. The
report include: first part translates the project into a qualitative cash flow
model description which highlights important characteristics
...sensitivity or other analyses using variants in of the business environment, the project, and the participation
commodity price, grade, capital and operating costs,
terms of project stakeholders. The framework’s second part
or other significant parameters, as appropriate, and
comprises the analytical processes used to estimate cash
discuss the impact of the results.
flows, calculate economic benefits, and assess risk.
This change to the rules recognises that single-variable It is important to remember that any project cash flow
sensitivity analyses superimposed on static DCF calculations model built within the Extended Evaluation Framework is
are a useful, but limited, way of describing the uncertainties still just an approximate model estimating project cash flows
surrounding a mineral project. ‘Other analyses’ could include generated within a highly complex project and economic
dynamic DCF or RO NPV calculations that incorporate environment. Cash flow modelling within this framework can
uncertainty analysis techniques such as Monte Carlo produce substantial amounts of information about a project’s
simulation or risk adjustments tuned to unique project
economic and risk characteristics that are both informative
uncertainty characteristics. The requirement in the revised
and helpful when making an investment decision. However,
Form F1 applies explicitly to all ‘advanced properties’ – defined
as with any economic analysis, these results should still
in the rule as a property with mineral reserves or a property
be treated with some caution and tempered with industry
with mineral resources whose potential economic viability
experience.
is supported by at least a preliminary economic assessment.
With that change, the old definition of ‘development property’
no longer had a function in the rule and was removed.
Describing the project with a cash flow model
Translating a project’s development and production plan
The parts of the technical report that deal with the
into a cash flow model consists of two components. The first
uncertainty of the economic model are still subject to the
component is a description of the most important uncertainties
general requirement (in s. 2.1 of NI 43-101) that all scientific
associated with the project and the business environment.
and technical information disclosed by an issuer on a material
The second component is an outline of project structure which
mineral project be the responsibility of a qualified person
encompasses not only the project’s cash flow consequences
(QP). While the 2011 revisions to NI 43-101 did provide
but also possible sources of management flexibility and the
some additional flexibility for the QP to rely on other experts
division of the underlying project cash flow between equity,
for matters that are not within their realm of scientific or
creditors, and government. Figure 1 provides a representation
engineering practice, that does not extend to the economic
of these two components.
analysis or to studies of uncertainties in it.
Given the complexity of mining projects, a cash flow
It should be made clear that demonstrating the economic
model will by necessity be only a crude representation of the
viability of a resource in a NI 43-101 Technical Report is not
project’s full cash flow generating capabilities since it is not
considered a formal business valuation. Formal valuations are
only required under Multilateral Instrument 61-101, Protection possible or even desirable to model every project uncertainty,
of Minority Security Holders in Special Transactions, which is management flexibility, or financing and taxation term. An
a rule applying to takeover bids, adopted only in the provinces important function of the valuation professional is to identify
of Ontario and Quebec. The Companion Policy for this rule the most important project characteristics that influence
refers to general business valuation standards as a ‘reasonable value and risk while ignoring secondary influences so that the
approach’ to disclosure. MI 61-101 is silent on the specifics of interpretation of results is transparent and manageable.
mineral property valuations, but an approach incorporating
the CIMVal Standards will normally meet similar business Description of uncertainty
valuation standards. The CIMVal Standards recognise RO A mining project is exposed to many uncertainties, including
analysis as a primary valuation method for properties with economic and financial uncertainties (eg metal prices, foreign
mineral reserves and in some cases for properties with exchange rates), technical or physical uncertainties (eg
resources. metal grades, resource tonnage, processing efficiencies), or

Components of a cash flow model

Description of uncertainty Project structure
1) Type 1) Cash flow
Economic / technical / political Timing / leverage / non-linearities
2) Resolution / updating 2) Flexibility
Continuous / discrete / Bayesian Operations / development / other
3) Interaction 3) Financing / taxation
Systemic / unsystemic / correlation Equity / creditors / government

FIG 1 - Components of a project cash flow model.

VALMIN SEMINAR SERIES / PERTH, WA, 18 OCTOBER 2011 / BRISBANE, QLD, 17 APRIL 2012 3
M SAMIS et al

uncertainties linked to the political or regulatory environment introduce complexity because they alter stakeholder cash
(eg environmental regulations, cancelling of mining licenses flow distributions in some situations (eg high metal price
due to social pressure). scenarios) but not in others (eg low metal price scenarios). For
The prime focus of modelling a particular uncertainty is example, project financing may have payouts that are linked
finding a set of conditional probability distributions that to commodity prices through hedging arrangements (eg
describe the possible future random outcomes for this ‘costless’ collars) while tax regimes have contingent features
variable. A further consideration is how each uncertainty is such as restrictions on tax loss carry forwards, depreciation
resolved during the project, as this may affect the manner in and net profit royalties.
which the project is managed and valued. Uncertainty can be A complete model of project structure needs to fully articulate
resolved at one distinct point in time through an event such as the dynamic interaction of equity, creditor, and government
the awarding of a mining license when an exploration license interests in order to correctly estimate each stakeholder’s cash
expires or partially resolved in discrete steps such as through flow under various possible random project outcomes.
a multi-year exploration or commercial research program.
There are also uncertainties that are never fully resolved such Project cash flow calculation, value estimation
as future commodity prices that have their forecasts revised
and risk assessment
as new information is received on a continual basis from
financial markets. Once the details of a cash flow model have been decided, a
three-part analytical process is followed to estimate cash
A final aspect of modelling project uncertainty involves the flow, analyse the project’s economic benefits, and describe
relationship between the various uncertainties and also their its uncertainty and risk characteristics. Figure 2 outlines this
interaction with the project and business environment. Some three-part analytical process. The actual details of this process
of these uncertainties may be independent of each other, will be determined by the objective of the evaluation exercise,
such as metal concentration and metal price. Other project the availability of information, and resources available to
uncertainties may display some correlation, such as copper perform the evaluation.
and gold prices. For valuation and risk discounting purposes,
it is also important to understand the relationship between Numerical method
an individual uncertainty and the overall financial markets as The standard approach for estimating mine project cash flows
this can influence how cash flow discounting is performed. is the Static DCF method, where expected or median outcomes
for uncertain project variables such as metal price, ore
Project structure grades, and processing are used in a simple and well-known
Project structure refers to the design and operational details, set of multiplicative and additive calculations to arrive at an
management’s ability to change development and operation, expected net cash flow estimate. It is important to recognise
and the terms of finance and taxation that combine to influence that a static cash flow estimation technique will generate
the generation of project cash flow. Static DCF models tend erroneous cash flow expectations when there is uncertainty,
to be concerned with only the timing of cash flows and their flexibility, and contingent tax and financing terms that create
composition. However, when modelling project uncertainty expectations over a non-linearity. In some situations, this
there may be reason to consider a more detailed description of error may be minor and can be ignored. However, static cash
structure that incorporates flexibility and contingent tax and flow error can be large in other situations which may result
financing payoffs. in incorrect conclusions about a project’s economic viability.
At some projects, managers may be able to change project There are several advanced numerical methods that can be
structure and cash flow when the arrival of new information used to avoid static cash flow error. Among these techniques,
causes them to update their expectations about the future. binomial trees (or lattices) are popular as cash flow estimation
This ability to modify project cash flows is called management tools since these methods are more visual and can be
flexibility. It increases value as it allows managers to mitigate implemented relatively easily in a spreadsheet. However,
possible losses in adverse conditions, enhance economic binomial trees have difficulty correctly estimating the effects
gains in advantageous situations and manage risk through of path-dependent cash flow structures such as tax-loss carry
operational decisions as new information is received. The forwards or net profit royalties that allow the recovery of an
actual value benefit can be large or small depending on the initial investment. For this reason, Monte Carlo simulation
flexibility available, the costs of that flexibility, and the may be preferred as a method of modelling uncertainty.
characteristics of underlying project uncertainty. A more powerful numerical approach finding increased
The presence of contingent financing and taxation terms use in the mining industry is based on the Longstaff-
may also require an extended modelling of cash flow structure Schwartz algorithm (Longstaff and Schwartz, 2001). This
as these terms can complicate the division of cash flow flexible simulation technique uses Monte Carlo simulation to
between equity, creditors, and government. Contingent terms generate a large number of random project cash flows based

An extended economic
and risk assessment toolkit

Numerical techniques
Numerical method of Estimating
Value and return Risk
Risk assessment
estimating cash flow value and return
estimation assessment
Static DCF NPV Sensitivity analysis
Binomial trees IRR CFAR / tails analysis
Flexible simulation Real option NPV Event probabilities

FIG 2 - An extended project evaluation and risk management analytical process.

4 VALMIN SEMINAR SERIES / PERTH, WA, 18 OCTOBER 2011 / BRISBANE, QLD, 17 APRIL 2012
 USING DYNAMIC DISCOUNTED CASH FLOW AND REAL OPTION METHODS FOR ECONOMIC ANALYSIS

on the underlying project uncertainty model. Then, it uses and output prices. Measures of the estimated risk associated
dynamic programming combined with regression techniques with these uncertainties are developed based on capital market
to estimate cash flow consequences of various project designs information and its supporting finance theory. If the prices of
and operating decisions. When matched with either the uncertain inputs and outputs are observable in well-developed
dynamic DCF or RO valuation methods it can identify those financial markets, the forward price curve for the particular
designs and decisions that maximise expected project value input or output may be used as risk-adjusted expected prices in
within the valuation framework. Natural resource project the cash flow calculation because a commodity’s forward price
examples of using this approach include Blais, Poulin and may be interpreted as a risk-adjusted expected price (Geman,
Samis (2005), Sabour and Poulin (2006), and Laughton, 2005; Markert and Zimmermann, 2008). Alternatively, a
Guerro and Lessard (2008). corporate forecast price may be risk adjusted using a model
such as the CAPM to estimate risk-adjusted expected prices
Estimate of project value and return if corporate guidelines require the use of a forecast price or
The project cash flow estimate generated by the chosen there is a lack of market information.
numerical method is used to determine whether participating Risk-adjusted cash flows for the project are then estimated
in a project is expected to have a positive net benefit for a using these risk-adjusted forecasts. Under an assumption that
potential stakeholder. There are two primary measures of this process has dealt with all uncertainties, the risk-adjusted
whether project participation is advantageous. These two cash flow is valued by discounting it for the time value of
measures are IRR and NPV. IRR was briefly discussed in the money (ie at the risk-free rate) based on the long-term yield of
introduction and will not be considered further. a publicly-traded government bond.
NPV is the expected net benefit or loss to a participant from A real option model can be further supplemented with a
making an investment in a project after accounting for the residual project risk factor applied to the risk-adjusted cash
risk-adjusted opportunity cost of capital. The natural resource flow to account for aspects of technical and commercial
industries primarily use the DCF method to calculate project project uncertainty that have not been explicitly recognised
NPV. A distinguishing feature of a static and dynamic DCF under the RO methodology (Smith and McCardle, 1998,
NPV is that adjustments for the value effects of uncertainty 1999). Support for the use of a residual risk premium is also
and time are applied on an aggregate basis to the project found in recent accounting guidelines requiring the use of a
expected net cash flow stream using a constant risk-adjusted counter-party risk premium when estimating the fair value of
discount rate (RADR). That RADR is applied via the bond a forward contract and other derivative financial instruments
pricing formula 1/(1+RADR)t. The choice of RADR may be (CICA Emerging Issues Committee, 2009). These guidelines
based on a financial market model of asset returns such as the recognise that not all the uncertainty and risk attached to the
capital asset pricing model (CAPM) or the weighted average payout of a forward contract or other derivative is recognised
cost of capital of the investor. One discount rate is often used by a standard derivative valuation model. One method of
across a broad class of investment projects regardless of the recognising additional contract risks such as counterparty
actual uncertainty characteristics of each particular project. default is to supplement the discounting process in the
The primary difficulty with DCF NPV is reconciling the risk standard model with an additional risk premium.
adjustment applied to a cash flow with the cash flow’s actual The difference between DCF and RO risk-adjustment may
uncertainty and risk characteristics. appear small but it has important implications for project
The RO method is an alternative approach to calculating evaluation. The key benefit of the RO approach is its ability
project NPV that is being investigated by some participants to recognise the complex variation of cash flow risk over the
in the natural resource industries. The RO NPV method is project life due to the characteristics of individual project
derived from the same finance theory as the DCF method, uncertainties, project structure and management flexibility,
with additional influences from the development of financial and the terms of financing and taxation.
option pricing theory by Black, Scholes and Merton during In contrast, the DCF method’s approach to risk adjustment
the early 1970s (Black and Scholes, 1973; Merton, 1973). The implies that cash flow uncertainty increases in a regular
term ‘real option’ was suggested as a means of identifying manner over the life of the project. This may be an appropriate
the application of option theory to real assets as opposed to simplification for some project evaluations but it may also
financial assets. The RO method described in the next few produce misleading results for other evaluations. Situations
paragraphs may seem novel to some mining professionals where a DCF risk adjustment may cause problems include:
but it can be considered a certainty equivalent approach
• long-life projects where the short-term price of inputs or
to calculating an Expected Net Present Value which is an outputs fluctuates around a long-term equilibrium level;
accepted method for calculating fair value under IFRS 13
• projects where input or output prices do not exhibit
by the International Accounting Standards Board (IFRS
reversion (ie they fluctuate in a manner similar to stock
Foundation, 2011). prices) and management has the flexibility to respond to
The distinguishing feature of the RO method is its risk this uncertainty; and
adjustment approach. This may be surprising to some as • projects in which the structure of royalty, taxation, and
many RO publications state that the explicit recognition of financing cash flows change significantly with output price
flexibility is the main difference between the DCF and RO levels or are dependent on the path of past prices.
NPV methods. However, RO’s recognition of flexibility is This list does not imply the DCF NPV method is
not unique, as modifications to the static DCF method for inappropriate in these cases. It suggests evaluation situations
flexibility (ie dynamic DCF) have been proposed since at where a valuation professional should consider whether the
least the mid-1960s (Magee, 1964), although they are only DCF risk-adjustment approach is suitable.
infrequently used.
The RO method takes the underlying static or dynamic Project risk assessment
cash flow model and applies risk adjustments to the primary Traditionally, quantitative risk assessment within a static DCF
commercial sources of cash-flow uncertainty, such as input model has been limited to scenario and sensitivity analyses.

VALMIN SEMINAR SERIES / PERTH, WA, 18 OCTOBER 2011 / BRISBANE, QLD, 17 APRIL 2012 5
M SAMIS et al

Scenario analysis re-calculates project cash flows and NPVs and other factors. Over the life of the project, average annual
for a number of alternative project scenarios while sensitivity after-tax operating cash flow is estimated to be US$978 M in
analysis changes a particular input variable over the life of real 2010 dollar terms with an average profit margin of 43 per
the project. Both techniques provide an indication of the cent given a forecast long-term copper price of US$2.00/lb
variability in project NPV to changes in input assumptions. and gold price of US$850/oz. Capital costs in real terms for
However, they do not provide any guidance on project risk the initial development and expansion phases were estimated
since they do not define nor provide a summary measure of to be US$7.2 B while sustaining capital was projected at
project risk. US$11.4 billion over the project’s production horizon.
The use of Monte Carlo simulation combined with risk The primary problem confronting a valuation professional
management concepts from the finance industry has extended when dealing with a project like the Oyu Tolgoi Project is how
the ability of valuation professionals to perform project risk to recognise the benefits of forecasted strong positive cash
analysis and communicate the results in a more concise manner. flows over a very long production horizon against the cost of
Monte Carlo simulation allows the reporting of not only large amounts of capital initially incurred to bring the project
expected cash flow but also cash flow confidence boundaries and into production. Static DCF models have difficulty fully
measures of cash flow uncertainty. Further analysis can then communicating the uncertainty and risk characteristics of
be performed on histograms of cumulative project cash flow to this cash flow pattern generated by a base metal mine. Using
highlight potential cash flow losses or gains from investing in dynamic DCF and RO methods within an extended evaluation
the project as well as how management flexibility or financial framework provides a more complete set of tools with which
strategies may limit the losses and enhance the gains (Samis to perform an economic analysis.
and Davis, 2009). Lastly, the information from simulation can The dynamic DCF and RO analyses in the OT NI43-101
be used to estimate the probability of certain events occurring, Report are broken into four sections. These include:
such as closing early due to low metal prices or the development
of subeconomic resources if metal prices increase given the 1. a description of metal price uncertainty,
uncertainty assumptions used in the cash flow model. This is 2. an overview of cash flow uncertainty characteristics when
particularly useful as a means of probabilistically estimating there is no management flexibility,
reserves and resources. 3. a statement of the dynamic DCF and RO evaluation results
The use of quantitative financial risk management concepts and sensitivity to uncertainty model parameters, and
during project evaluation are still a novel application in 4. an explanation for the differences between the DCF and
the mining industry. Although it is not clear at this time RO NPV results.
how financial risk management concepts can be adapted
for analysing mining projects, the range and depth of risk Description of metal price uncertainty
management applications in the financial industry provide The dynamic DCF and RO analyses in the OT NI43-101 report
useful tools (and warnings) for mining project risk assessment incorporate explicit models of copper, gold, and molybdenum
and decision making. price uncertainty, while uncertainties related to capital and
operating costs and technical characteristics are not modelled.
THE OYU TOLGOI PROJECT – AN EXAMPLE OF In this paper, we discuss only the copper price model because
copper revenues comprise 80 per cent of overall project
DYNAMIC DISCOUNTED CASH FLOW/ revenues. Copper price uncertainty is modelled with a one-
REAL OPTION ANALYSIS factor reverting lognormal stochastic process. Reversion is the
Two NI43-101 technical reports have been published during tendency for a metal spot price to fluctuate randomly around
the last two years that include dynamic DCF and RO analyses a long-term equilibrium level. When a metal price exhibits
as a complement to static DCF analysis and as a means of reversion, prices that are much higher or lower than the long-
recognising specific value-influencing project characteristics. term equilibrium price are hard to sustain as it is assumed
The first was Ivanhoe Mines’ NI43-101 technical report for that supply/demand forces in the economy move metal prices
the Oyu Tolgoi Project (‘OT NI43-101 Report’; Ivanhoe Mines back to the equilibrium level.
Limited, 2010), which performed these analyses on a ‘no- This type of price model also recognises that market
flexibility’ basis (ie the effects of management flexibility were participants can update their price forecasts in response to
ignored) for the Life-of-Mine Sensitivity Case. This report was new price information. This characteristic is illustrated in
followed shortly afterwards by a similar report published by Figure 3, where a single simulated copper price scenario in
Entree Gold for their interest in the same project. We provide constant monetary terms is displayed as a solid black line.
a short review of the analyses in the Ivanhoe report to illustrate Note that this price scenario is just one of many generated
how these techniques can be used to communicate the impact during the Monte Carlo simulation process. Figure 3 also
of specific project characteristics on economic viability. provides a snapshot of forecast prices and the associated
level of price uncertainty at various project times in constant
The Oyu Tolgoi Project is one of the world’s largest
monetary terms. At the start of the project, price expectations
undeveloped copper-gold projects. It is located in Mongolia,
are represented by the dashed purple line. These expectations
approximately 550 km south of the capital, Ulaanbaatar. It is
decline from the market spot price on 31 December 2009
being jointly developed by Ivanhoe Mines, Rio Tinto, and the
towards the long-term Ivanhoe forecast of $2.00/lb.
Government of Mongolia. The OT NI43-101 Report predicts
the initial project development phase will last to December The uncertainty associated with this forecast is illustrated
2013 followed by an expansion development phase that by 90 per cent/ten per cent confidence boundaries which are
continues to 2020. The project’s production horizon lasts from demarcated by the dotted light purple lines. These boundaries
2014 to 2071 during which an estimated three billion tonnes provide a price range within which 80 per cent of metal
of run-of-mine (ROM) ore, assuming resources are converted prices are expected to fall at a particular time. The confidence
to reserves, will be processed. This ore will produce 50 billion boundaries for copper price stabilise at $3.00/$1.16 per lb due
pounds (lb) of copper, 25 million troy ounces (oz) of gold, to the effect of reversion.
161 million oz of silver, and 136 million lb of molybdenum The updating of forecasts and uncertainty are shown by
after adjustments for recoveries, smelter/refinery deductions, the blue and green lines in Figure 3. These lines depict the

6 VALMIN SEMINAR SERIES / PERTH, WA, 18 OCTOBER 2011 / BRISBANE, QLD, 17 APRIL 2012
 USING DYNAMIC DISCOUNTED CASH FLOW AND REAL OPTION METHODS FOR ECONOMIC ANALYSIS

5.00

4.50

Constant dollar copper price ($/lb)

4.00

3.50

3.00

2.50

2.00

1.50

1.00
Long-term expected price = $2.00/lb
0.50

0.00
01-Jan-2010 01-Jan-2020 01-Jan-2030 01-Jan-2040 01-Jan-2050 01-Jan-2060

Project time (year)

One simulated stochastic copper price path.
Expected CU price at 31/12/09 CU price of $3.33/lb. 10%/90% confidence bdy for 31/12/09 forecast CU price.
Expected CU price at 31/12/19 CU price of $3.19/lb. 10%/90% confidence bdy for 31/12/19 forecast CU price.
Expected CU price at 31/12/29 CU price of $1.37/lb. 10%/90% confidence bdy for 31/12/29 forecast CU price.

FIG 3 - A simulated price path with expectations and confidence boundaries for a one-factor stochastic copper price model.

model’s revision of copper price expectations and associated and miscellaneous charges but does not include capital
confidence boundaries based on the simulated spot price expenditures or working capital. Based on the information in
at a select point in time. They can be re-drawn at any point the OT NI43-101 Report, annual average after-tax operating
along the simulated path. The dashed blue line and its cash flow is US$978 M, though annual expected amounts vary
accompanying dotted light blue lines outline the revised metal greatly. The cash flow confidence boundaries demarcate a
price expectations and confidence boundaries for 31 December range in which 80 per cent of the cash flows are estimated to
2019 when simulated copper price is $3.19/lb as of that date. fall for a specific time and provide some indication of cash flow
The dashed green and dotted light green lines present revised uncertainty. The level trend in the uncertainty of cash flow
price expectations and confidence boundaries based on that amounts between 2025 and 2050 is due to the combination
simulated outcome. For both lines, copper price expectations of similar annual production amounts and copper price
revert back towards their long-term equilibrium levels while reversion.
long-term confidence boundaries after 2040 are unaffected by
short-term price changes because of price reversion. It is important to realise that the project cash flow
expectations and uncertainty characteristics in Figure 4 along
Price model inputs such as price volatility and reversion with the addition of capital expenditure and working capital
strength where estimated by econometric statistical techniques
underlie both of the dynamic DCF and RO NPV calculations.
from historic price data. The assumption of historic copper
These two NPV techniques start with the same project cash
price reversion was tested and supported by a statistical test.
flow and apply different risk adjustment methods to arrive at
Price medians in the model were set so that the metal price
what are most likely different NPVs.
models had long-term expectations equal to Ivanhoe’s long-
range price forecasts.
Dynamic discounted cash flow and real options
Cash flow expectations and uncertainty evaluation results
characteristics The primary economic analysis section in the OT NI43-101
The static cash flow model underlying Ivanhoe’s economic Report used the Static DCF method to estimate the project’s
analysis of the Oyu Tolgoi Project was rebuilt to facilitate the NPV at US$5.6 billion. This calculation was done in constant
simulation of metal price uncertainty and the effect of this dollars with a real eight per cent risk-adjusted discount rate.
uncertainty on project cash flows. Metal prices were simulated The dynamic DCF and RO analysis conducted in the report
in constant dollar terms and then inflated at a two per cent rate differed from the conventional Static DCF calculation as
into nominal terms to calculate cash flow. Price-dependent follows:
calculations such as government taxes and royalties and the • A nominal 10.2 per cent risk-adjusted discount rate was
Entrée cash flow distribution were adapted to allow for the used in the dynamic DCF NPV calculation which is a
wide range of metal prices generated by simulation while still discrete discounting adjustment applied to the real eight
reflecting the original terms. Cash flows were then deflated per cent DCF discount rate to recognise a two per cent
into real terms for various reports and graphs. inflation rate.
Figure 4 presents the expected real after-tax operating • The RO NPV calculation used a residual risk premium of
cash flow for the Oyu Tolgoi project and the associated 2.2 per cent to risk-adjust cash flows for non-metal price
ten per cent and 90 per cent confidence boundaries. After- uncertainty. This premium was estimated by splitting the
tax operating cash flow is defined as revenue less direct risk rate in the real DCF discount rate into revenue and
and indirect costs, taxes, royalties, the Entrée distribution, non-revenue components. Risk-adjusted RO cash flows

VALMIN SEMINAR SERIES / PERTH, WA, 18 OCTOBER 2011 / BRISBANE, QLD, 17 APRIL 2012 7
M SAMIS et al

4500

4000
After-tax operating cash flow ($ million)

3500

3000

2500

2000

1500

1000

500

0
01-Jan-2010 01-Jan-2020 01-Jan-2030 01-Jan-2040 01-Jan-2050 01-Jan-2060 01-Jan-2070

-500
Date

Expected after-tax operating cash flow 90% confidence bdy 10% confidence bdy

FIG 4 - Real after-tax operating cash flow expectations with 90 per cent/ten per cent confidence boundaries.
were discounted using a nominal time discount (risk-free) professionals an understanding of how well each NPV
rate of 4.7 per cent. method adapts discounting and risk adjustments to changes
• The capital asset pricing model and financial market data in cash flow uncertainty during the project. Insight into these
was used to estimate individual risk adjustments for gold, adaptations may assist a valuation professional to choose
copper, and molybdenum prices in the RO calculation as a between DCF and RO NPV results if it is important to have
proxy for how markets would evaluate these risks. a detailed correspondence between risk adjustment and the
• Initial prices for copper and molybdenum for the cash flow level of cash flow uncertainty.
analysis were set to market prices as of 31 December 2009 The difference between dynamic DCF and RO NPV results is
and allowed to revert back to the Ivanhoe long-term metal attributed in the OT NI43-101 Report to an overall difference
price forecast instead of starting at the long-term forecast in effective discounting. The report relies on annual RO and
price. Dynamic DCF cash flow discount factors (CFDF) for the after-
• Cash flow estimates were generated with simulation in tax operating cash flow to explain the NPV difference.
order to remove possible cash flow estimation errors
A CFDF is defined as the ratio of a cash flow’s present value
linked to static cash flow calculations and to generate cash
to its expected value and reflects adjustments for both cash
flow uncertainty information.
flow uncertainty through a risk adjustment and the time value
Over the project life, the dynamic DCF and RO analyses
of money. This ratio estimates the amount the market would
estimate cumulative net cash flow at US$38.4 billion. The
static DCF calculation estimated this figure to be $37.7 billion. be willing to pay now for a dollar of uncertain project cash
The difference in the two estimates is due to the dynamic flow at a specific future time. Dynamic DCF and a RO CFDFs
DCF and RO analyses starting from the actual current market will likely differ because each NPV method adjusts for risk in
prices for copper and molybdenum and small non-linear cash a different manner.
flow effects related to expectations over taxes and royalties. Figure 5 presents the dynamic DCF and RO CFDFs for the
The NPVs from dynamic DCF and RO calculations in the Oyu Tolgoi Project and reveals why the RO NPV is higher.
supplemental analysis were US$6.0 billion and US$7.6 billion In this figure, RO CFDFs, demarcated by the dashed green
respectively. The difference between the two estimates is line, indicate that the RO method tends to value cash flow
the result of risk adjustment differences. The particular risk more highly than the dynamic DCF method whose CFDFs
associated with price reversion is recognised by RO through an are delineated by a solid grey line. For example, for after-
explicit risk adjustment factor to the copper and molybdenum tax operating cash flow occurring in 2039, the RO method
revenue streams. The RO NPV is higher than the DCF NPV in estimates that each $1 of expected cash flow has a present
this particular instance due to what turns out to be a lower risk value of $0.103 while the dynamic DCF method estimates that
adjustment. This outcome may not occur for other projects. each $1 of expected cash flow in the same year has a present
value of $0.058. This graph shows that the RO CFDFs vary
Discounting and risk adjustment differences dynamically from year-to-year due to changes in cash flow
between dynamic discounted cash flow and uncertainty caused by changes in operating leverage (change
real options net present value results in unit cost) while the DCF CFDFs decrease in a steady manner
The OT NI43-101 Report provides a comparison of the of related to the functional form of the discounting formula.
the discounting and risk adjustments applied to the project The analysis of dynamic DCF and RO discounting in the OT
after-tax operating cash flow by the dynamic DCF and RO NI43-101 report goes further by considering how the price risk
NPV methods. This comparison is made to give valuation adjustment effectively applied by each NPV method responds

8 VALMIN SEMINAR SERIES / PERTH, WA, 18 OCTOBER 2011 / BRISBANE, QLD, 17 APRIL 2012
 USING DYNAMIC DISCOUNTED CASH FLOW AND REAL OPTION METHODS FOR ECONOMIC ANALYSIS

to changes in cash flow uncertainty. This is done by removing is sourced increasingly from the lower-grade Heruga Deposit.
the adjustments for the time value of money and residual This graph shows that the variation of cash flow uncertainty
risk from the CFDFs to generate cash flow price risk discount is dynamic and non-constant over the life of the Project.
factors (CF-PRDF) and comparing the modified dynamic DCF Uncertainty increases or decreases and displays sudden large
and RO factors to changes in cash flow uncertainty. variations as year-to-year unit costs change due to variations
Figure 6 presents this detailed risk adjustment analysis. in ore grade or changes in mining methods. The gradual trend
In this figure, after-tax operating cash flow uncertainty is of increasing cash flow uncertainty between the years 2020 to
tracked by the dash-dot blue line against the left-hand y-axis 2060 is due to the interaction between copper price reversion
using the cash flow coefficient of variation (CoV). Cash flow and a gradual decline in ore grades.
CoV is defined as the standard deviation of the uncertain cash The right-hand y-axis in Figure 6 plots the pattern of RO
flow divided by its expected amount. A larger CoV indicates a and dynamic DCF cash flow-price risk discount factors (CF-
relatively larger amount of uncertainty. The cash flow CoV line PRDF). These CF-PRDFs represent an estimate of the amount
shows that annual after-tax operating cash flow uncertainty the market would be willing to pay now on a metal-price risk
varies within a range of just under 40 per cent to 100 per cent adjusted but not time discounted or residual risk-adjusted
before 2063 and then increasing markedly after production basis for a dollar of a future cash flow.

1.0

0.8
Net cash flow discount factor

0.6

0.4

0.2

0.0
01-Jan-2010 01-Jan-2020 01-Jan-2030 01-Jan-2040 01-Jan-2050 01-Jan-2060 01-Jan-2070
Date
DCF Net cash flow discount factor RO Net cash flow discount factor

FIG 5 - Dynamic discounted cash flow and real option net cash flow discount factors.
500% 1.0

400% 0.8
Cash flow Price Risk-Discount Factor
Cash flow coefficient of variation (%)

300% 0.6

200% 0.4

100% 0.2

0% 0.0
01-Jan-2010 01-Jan-2020 01-Jan-2030 01-Jan-2040 01-Jan-2050 01-Jan-2060 01-Jan-2070

-100% -0.2

-200% -0.4
Date
After-tax operating cash flow coefficient of variation DCF CF Price Risk-Discount Factor
RO CF Price Risk-discount factor

FIG 6 - Dynamic discounted cash flow and real option cash flow coefficient of variation and cash flow price risk discount factors.

VALMIN SEMINAR SERIES / PERTH, WA, 18 OCTOBER 2011 / BRISBANE, QLD, 17 APRIL 2012 9
M SAMIS et al

The dynamic DCF CF-PRDF and pattern of DCF metal price This paper proposes some analytical elements of good
risk discounting is demarcated by the solid dark grey line. professional practice when working with the Extended
This line reflects a pattern of metal price risk adjustments Evaluation Framework. These elements include:
associated with cash flow uncertainty that is continually • a description of project uncertainty with statistical support
increasing in a regular manner whereas the profile of cash flow for model assumptions,
CoVs suggests a cash flow uncertainty pattern that increases • analysis of cash flow uncertainty characteristics when
in a more erratic fashion. The dynamic DCF CF-PRDFs do not there is both no flexibility and management flexibility (if
appear particularly consistent with the pattern of cash flow considered),
uncertainty at the Oyu Tolgoi Project.
• details of the links between static DCF analysis and
The dashed green line in Figure 6 delineates the RO CF- the analysis conducted with the Extended Evaluation
PRDFs and shows the RO metal price risk adjustments Framework, and
respond in a manner that is more consistent with the erratic • a sensitivity analysis of the economic input parameters for
annual changes in after-tax operating cash flow uncertainty. the dynamic DCF and RO models.
Specifically, increases in cash flow uncertainty result in an
increase in risk adjustment (ie a decline in the CF-PRDF and Project uncertainty
an increase in risk compensation) and a decline in uncertainty The introduction of explicit uncertainty models will require
results in a corresponding decline in risk adjustment (ie an that each individual model be supported by some statistical
increase in CF-PRDF and a decline in risk compensation). justification. There is a tendency in the mining industry to
Choosing between the dynamic DCF and RO NPV results is a use inappropriate models (eg modelling metal price density
matter of professional preference. However, Figure 6 suggests functions as triangular distributions) that are based on
that this choice could be made on how the risk adjustments of personal judgement and experiences in unrelated areas.
each NPV method respond to changes in cash flow uncertainty. For commodity prices used in estimating revenues and
A fundamental principle of finance theory is that investors are costs, there are several useful models of price movements
risk averse and require more compensation for investing in a ranging from fairly simple one-factor stochastic models to
project with more uncertain cash flows. Risk compensation is more sophisticated three-factor models, in which information
determined by a cash flow risk discount whereby an investor about price standard deviation, skewness, kurtosis, and
would be willing to pay an amount that is smaller than the intertemporal correlation can be included. Statistical
expected amount of an uncertain cash flow. This concept analysis of historic price data must be used to estimate model
of risk/return implies that annual variations in cash flow parameters and to demonstrate whether a particular type
uncertainty should be accompanied by changes in the amount of model is appropriate for the commodity being modelled.
an investor is willing to pay on a risk-adjusted basis per cash Long-term metal price expectations in the uncertainty models
flow dollar. In the case of the Oyu Tolgoi Project, the pattern should be consistent with those used in the base case static
of RO CF-PRDFs is more consistent with the pattern of cash DCF model, and any differences between this expectation
flow uncertainty than the dynamic DCF CF-PRDFs, which and financial market trends should be discussed. Lastly, a
suggests that RO NPV may be more representative of the qualitative description of each model, supported by graphs,
project cash flow risk profile caused by reversion in copper should be included to illustrate the uncertainty characteristics
price and changes in unit costs over time. in non-mathematical terms.
Technical uncertainty will be more difficult to integrate
into the Extended Evaluation Framework given that there
SOME THOUGHTS ON BEST PRACTICE FOR may be insufficient project information to develop a model
DYNAMIC DISCOUNTED CASH FLOW/REAL (eg a grade uncertainty model at an early exploration stage).
OPTION ANALYSIS IN NI43-101 REPORTS However, some late-stage projects may have large amounts of
information from earlier exploration and engineering work
The use of dynamic DCF and RO methods in a NI43-101
that allows a qualified professional to develop a technical
technical report to demonstrate the prospects for the economic
uncertainty model that is justified in a manner conceptually
extraction of a proposed resource requires some thought as to
similar to models of economic uncertainty.
what is considered professionally acceptable analysis. A static
DCF model is a straightforward analytic framework which
Cash flow uncertainty and flexibility
is broadly applied in a consistent manner across the mining
industry. The use of the Extended Evaluation Framework Advanced numerical methods such as Monte Carlo simulation
with its advanced valuation theory and numerical methods or lattice methods are able to assess the economic sensitivity of
requires a specialist skill set that is less common in industry. a resource to changes in its business and project environment.
This increases the prospect of inconsistent and unprofessional This information generated with these methods provides
dynamic DCF and RO analysis being included in NI43-101 a richer description of cash flow uncertainty than the more
technical reports. Examples of poor professional practice commonly used sensitivity analysis involving spider plots or
using variations of the Extended Evaluation Framework in tornado diagrams.
general mining valuation situations include the use of the Analysis of cash flow uncertainty must start with the base
Black-Scholes model based on superficial similarities between case development and production policy under the condition
an undeveloped mining project and a financial call option, and of no management flexibility as this is directly comparable
the use of real option models where uncertainty is modelled to the static DCF model used in earlier sections of the
in aggregate at the level of operating profit, rather than at NI43-101 technical report. Analysis may include graphical
individual cash flow elements, such as is discussed in Copeland representations of expected annual cash flows and confidence
and Antikorov (2001) and other standard RO textbooks. The boundaries, relative measures of annual cash flow uncertainty
worry here is that the use of these discredited approaches such as CoV plots, histograms of cumulative cash flow,
will lead to the production of economic information that is and other risk management tools such as conditional tails
misleading and confusing (Borison, 2005; Smith, 2005). expectations. Cash flow uncertainty characteristics should

10 VALMIN SEMINAR SERIES / PERTH, WA, 18 OCTOBER 2011 / BRISBANE, QLD, 17 APRIL 2012
 USING DYNAMIC DISCOUNTED CASH FLOW AND REAL OPTION METHODS FOR ECONOMIC ANALYSIS

be discussed in the context of project features such as linking appropriate for future economic and project conditions by
increases or decreases in cash flow uncertainty to changes some investors. Some of these parameters may include the
in metal mix, grades, mining method, depth of operation, or price of risk in financial markets (a measure of investor risk
mining area. aversion), commodity price volatility, correlation between
Management flexibility has yet to be introduced into the commodity prices and financial markets, strength of reversion
economic analysis section of a NI43-101 technical report. or even the presence of reversion. A sensitivity analysis
However, flexibility analysis is a powerful method for should be provided that shows how project NPV varies with
describing the economic conditions in which subeconomic uncertainty model parameters. Providing this analysis will
resources may be developed at a profit or the how a project will allow investors to assess the project’s economic prospects
be managed in adverse conditions such as low metal prices. under alternative market and technical circumstances.
Management flexibility should be described where possible
with a decision tree and details of each development and CONCLUSION
production alternative represented by the individual decision Techniques for evaluating mining project economics are
tree branches. The cash flow uncertainty characteristics of continually improving and becoming more sophisticated. The
each operating alternative and the conditions under which analysis associated with the traditional Static DCF method
each alternative would be selected should be discussed. Cash is little changed from when mining engineers first used this
flow uncertainty would be articulated in a manner similar method in the early 1900s. This paper provides an overview of
to that of no-flexibility analysis while selection of operating how the Static DCF method can be adapted and extended by
alternatives can be described in terms of policy plots that link incorporating some of the advances in finance and economics
price boundaries to changes in operating policy. Ultimately, that have been made in the past 40 years. The two specific
a dynamic programming application such as a binomial approaches to evaluation that we describe are called dynamic
lattice, partial differential equations, or Longstaff-Schwartz DCF and real option analyses. Both of these methods have the
simulation will be needed to solve for the project value given potential to estimate project cash flows given explicit models
the optimal operating policy under various random price of project uncertainty and dynamic project structure such as
outcomes. management flexibility. However, the two analyses techniques
Finally, the impact of flexibility should be considered on a vary in the manner in which a risk adjustment is applied.
project-wide basis to explain how flexibility affects value and Dynamic DCF risk adjusts cash flows through discounting the
risk through changing the histogram of cumulative net cash net cash flow stream with a constant discount rate while the
flows. The dynamic programming solution will also provide real option method risk adjusts individual sources of cash flow
information as to how management is predicted to react uncertainty (eg metal price) using a tailored risk adjustment
under various random outcomes, and this can be checked for before discounting the resulting risk-adjusted net cash flow for
realism and constraint violations. the time value of money. Risk adjustment differences between
these techniques will produce different valuation results which
Linking static discounted cash flow and will require explanation that will ultimately provide valuable
dynamic discounted cash flow/real option information to investors interested in the project.
analysis Recent amendments to the guidelines for NI43-101 Technical
There will inevitably be differences between conventional Reports allow economic analysis using a Static DCF model
Static DCF NPV and NPVs calculated using the dynamic DCF to be supplemented with analysis considering the economic
and RO methods. These differences need to be explained to impact of project uncertainty. We discuss an example of this
support the validity of results produced with the Extended supplemental analysis in which Ivanhoe Mines and Entrée
Evaluation Framework, particularly given the mining Gold used the dynamic DCF and real option methods in
industry’s familiarity with the Static DCF method. their respective NI43-101 reports for the Oyu Tolgoi Project
to provide additional information about the characteristics
A starting point for explaining the differences in results is of project risk and its effect on project value. This analysis
a component NPV build up highlighting how project NPV showed that the real option method tends to vary its effective
changes with the introduction of metal price expectations risk discounting commensurately with the variation of the net
linked to a stochastic model, the effect of recognising cash cash flow risk, in contrast to the dynamic DCF approach which
flow non-linearities with simulation, the use of RO risk- employs a risk adjustment that does not vary with changes in
adjustments in place of DCF discounting, switching between project risk over time.
constant dollar and nominal dollar cash flow calculations, and
the value impact of various types of flexibility. The purpose There is now considerable latitude for qualified persons to
of the build-up is to indicate the primary factors impacting use more sophisticated economic analyses in technical reports.
the economic extraction of a resource which may be able to We would even argue that best practice requires that the effort
include among other considerations the presence of non-linear spent on economic analysis in technical reports be bolstered to
windfall taxes or sliding-scale royalties, the flexibility value of match the effort and sophistication on the technical side. Since
modular capital development or the contingent development these economic techniques require skills that are new to the
of subeconomic resources, or the effect of price reversion on industry, the main difficulty in their implementation is finding
long-term cash flows. An effort should be made to explain how a professional with the appropriate training to undertake
the risk adjustments profiles differ between Dynamic DCF and the analysis and finding senior managers comfortable with
RO NPV calculations to assist with understanding the value interpreting and explaining the results. It is becoming
change produced by switching between the two NPV methods. increasingly evident that project evaluation requires mining
professionals and senior managers that are familiar with
Sensitivity to changes in dynamic discounted both the technical aspects of mining and the financial aspects
of commodity markets. However, the expense and effort
cash flow/real option uncertainty model required to gain this familiarity will be more than offset with an
parameters improved understanding of how the characteristics of project
Dynamic DCF and RO models rely on the estimation of and market uncertainty combine with financing arrangements
parameters from historic data that may not be considered and management flexibility to affect project value and risk.

VALMIN SEMINAR SERIES / PERTH, WA, 18 OCTOBER 2011 / BRISBANE, QLD, 17 APRIL 2012 11
M SAMIS et al

REFERENCES JORC, 2004. Australasian Code for Reporting of Exploration

Results, Mineral Resources and Ore Reserves (The JORC Code)
Black, F and Scholes, M, 1973. The pricing of options and corporate
[online]. Available from: <http://www.jorc.org> (The Joint Ore
liabilities, Journal of Political Economy, 81:637-54. Reserves Committee of The Australasian Institute of Mining and
Blais, V, Poulin, R and Samis, M, 2005. Using real options to Metallurgy, Australian Institute of Geoscientists and Minerals
incorporate price risk into the valuation of a multi-mineral mine, Council of Australia).
in Orebody Modelling and Strategic Mine Planning (ed: R Laughton, D, Guerrero, R and Lessard, D, 2008. Real asset
Dimitrakopoulos), pp 9 -15 (The Australasian Institute of Mining valuation: A back-to-basics approach, Journal of Applied
and Metallurgy: Melbourne). Corporate Finance, 20(2):46-65.
Borison, A, 2005. Real options analysis: Where are the emperor’s Longstaff, F A and Schwartz, E S, 2001. Valuing American options by
clothes? Journal of Applied Corporate Finance, 17(2):17-31. simulation: A simple least-squares approach, Review of Financial
Studies, 14:113-147.
Canadian Institute of Mining, Metallurgy, and Petroleum,
Special Committee on Valuation of Mineral Properties, 2003. Magee, J, 1964. Decision trees for decision making, Harvard
Standards and Guidelines for Valuation of Mineral Properties. Business Review, 42(4):126-138.

Canadian Securities Administrators, 2005. National Markert, V and Zimmermann, H, 2008. The relationship between
Instrument 43-101 Standards of Disclosure for Mineral Projects risk premium and convenience yield models, in The Handbook
of Commodity Investing (eds: F Fabozzi, R Fϋss and D Kaiser),
(‘old’ NI 43-101).
Chapter 5 (John Wiley and Sons: New Jersey).
Canadian Securities Administrators, 2011. National Instrument
Merton, R, 1973. The theory of rational option pricing, Bell Journal
43-101 Standards of Disclosure for Mineral Projects.
of Economic and Management Science, 4:373-413.
CICA Emerging Issues Committee 173, 2009. Credit risk and Sabour, S A A and Poulin, R, 2006. Valuing real capital investments
the fair value of financial assets and financial liabilities, CICA using the least-squares Monte Carlo method, Engineering
Standards and Guidance Collection, 20 January. Economist, 51(2):141-60.
Copeland, T E and Antikarov, V, 2001. Real Options: A Practitioner’s Samis, M and Davis, G A, 2009. Using dynamic DCF and real
Guide, 372 p (Texere Publishing Limited: New York). options methods to value and assess flexible mine project design,
Entrée Gold Incorporated, 2010. Lookout Hill Property – in Mineral Resources/Reserves and Valuation Standards, CIM
Technical Report June 2010, 565 p [online]. Available from: Special Volume 56, pp 632-650 (Canadian Institute of Mining,
Metallurgy and Petroleum: Montreal).
<http://www.sedar.com>.
Smith, J, 2005. Alternative approaches for solving real-options
Geman, H, 2005. Commodities and Commodity Derivatives –
problems (comment on Borison et al, 2005), Decision Analysis,
Modeling and Pricing for Agricultural, Metals, and Energy, 2.2:89-102.
396 p (John Wiley and Sons: United Kingdom).
Smith, J and McCardle, K, 1998. Valuing oil properties: Integrating
International Accounting Standards Board, 2011. 2012 option pricing and decision analysis approaches, Operations
International Financial Reporting Standards IFRS – Research, 46(2):198-217.
Consolidated Without Early Applications, 3008 p (IFRS
Smith, J and McCardle, K, 1999. Options in the real world: Lessons
Foundation Publications: London).
learned in evaluating oil and gas investments, Operations
Ivanhoe Mines Limited, 2010. Oyu Tolgoi Project – Technical Research, 47(1):1-15.
Report June 2010, 629 p [online]. Available from: <http://www. South African Mineral Resource Committee, 2009. South
sedar.com>. African Code for the Reporting of Exploration Results, Mineral
Resources and Mineral Reserves.

12 VALMIN SEMINAR SERIES / PERTH, WA, 18 OCTOBER 2011 / BRISBANE, QLD, 17 APRIL 2012