Mining Valuation: Three steps beyond a static DCF model

(As originally appeared in Canadian Mining Journal, December 2010)

  • Share

«Back to list

By Richard Crosson, Senior Partner, Transaction Advisory, Ernst & Young LLP and Mike Michael Samis, Vice President, Valuation and Business Modelling, Ernst & Young LLP

The primary valuation methodology for development properties and producing mines is discounted cash flow (“DCF”). The underlying valuation premise is that value reflects the current net economic benefit of the net cash flows that are expected to be generated over the life of the project. Calculating this benefit involves estimating expected after-tax cash flows and converting these cash flows into a present value or net asset value (“NAV”) through a process called “discounting”. The discounting process incorporates a “discount factor” to provide a rate of investment return that accounts for both the time value of money and risk factors.

DCF NAV calculations have many uses in the mining industry, including justifying project design choices or investment decisions, supporting NI43-101 resource declarations and providing value measurements for accounting tests. While the DCF method is widely accepted, it is subject to limitations and is at times unreliable. A gold sector example is the calculation of a gold project’s NAV by valuators or analysts who then multiply the result by a factor of (say) 1.8 in order to reconcile the DCF NAV with market valuations. This discrepancy is usually attributed to the DCF method not recognizing optionality or exploration potential of the project.

Valuators at the forefront of mine valuation are seeking to improve DCF calculations through three significant enhancements: the dynamic modelling of uncertainty, the recognition of the cash flow effects and contingent payoffs and the use of market-based risk discounting methods such as real options.

Dynamic modelling of uncertainty

Standard DCF models are static; they incorporate a single net cash flow stream with most-likely or expected values of project and business variables such as metal grade. Analysts recognize that these variables are uncertain and may assess how uncertainty impacts NAV with a sensitivity analysis where variables such as metal price are discretely varied over a range. More involved uncertainty analysis may include Monte Carlo simulation, in which a set probability distribution describes possible changes in a specific variable during the project.

Neither of these analytic techniques considers how forecasts are updated and uncertainty is resolved during a project. Introducing stochastic processes into Monte Carlo simulation can provide a richer description of the dynamics of uncertainty and its effect on cash flow. For example, a log-normal stochastic process can be used to model reversion in base metal prices. Reversion is the tendency of a variable such as metal price to revert over time to a long-term equilibrium level and may act to restrain long-term cash flow uncertainty. These models can be extended to reflect other characteristics such as uncertainty in long-term equilibrium price levels and the structure of forward curves.

Flexibility and Contingent Payoffs

A static DCF model relies on a cash flow scenario built on a single development / production policy, combined with predetermined financing and taxation payouts. Unfortunately, this scenario may provide a flawed estimate of cash flow, since operating policy may be altered and the structure of financing and taxation payoffs transformed in response to changes in the project and business environment. This limitation of static DCF is particularly problematic when attempting to estimate the value of sub-economic resources at a gold or copper-gold mine, the economic impact of windfall taxes, or the true cost of a financing arrangement with embedded commodity derivatives.

It is standard industry practice to offset this DCF defect by building a series of models for different designs and project environments. However, this approach results in a scattering of NAVs that often provide little insight into how flexibility and contingent payoffs interact with uncertainty to affect project NAV.

Market-based valuation methods

The structure of risk discounting in a standard DCF NAV calculation implies an uncertainty profile in which net cash flow uncertainty grows over time in a well-behaved manner. However, for most projects, net cash flow uncertainty is not well behaved as it changes annually in an erratic manner during the project due to factors such as changes in metal price and grades and the exhaustion of tax shields. The inconsistency between a project’s actual risk profile and the profile implied by the DCF risk discounting formula is particularly problematic when valuing long-life base metal projects where large upfront capital expenditures are required before cash flows are generated over periods of more than 30 years.

Finance has developed market-based methods such as option pricing to value financial assets which have dynamic risk characteristics. These methods, when adapted for the mining industry, use the same information and have a similar approach to building a cash flow model as the standard DCF method. However, the DCF and market-based NAV methods are differentiated by how they apply a risk adjustment. Whereas the DCF approach applies an aggregate risk adjustment to the net cash flow, market-based methods apply a tailored risk adjustment to individual uncertainty sources (e.g. copper price) based on their systemic risk component before building up a risk-adjusted net cash flow.

Not just theory

The discussion in this article by The EY Team may appear to be more theoretical than practical. However, the authors have observed that many leading mining companies are actively undertaking internal investigations of the three enhancements to improve their valuation and risk management processes. While most of the work being performed is currently for internal use, publicly disclosed examples involving these enhancements are starting to appear.