By Aldin Ardian – Chief Strategy Officer (CSO) at PT Studio Mineral Batubara (SMB) and lectures at UPN “Veteran” Yogyakarta
The mining industry has long been recognized as a capital-intensive and high-risk industry. According to several studies, many mining projects end in failure, or at least around 80% of projects experience budget overruns of up to 43%. This indicates that uncertainty in the industry is considerably high. Exploration activities may fail to discover economically mineable resources, mining sites may be located within protected forest areas, operational costs may exceed initial plans, or commodity prices may decline, causing projects that were once considered feasible to become unviable for further development. The analysis conducted to determine whether a mining project is feasible is commonly known as a feasibility study (FS).
Parameters in Preparing a Feasibility Study
In an FS, exploration resources are engineered and modeled according to mining systems/methods, technologies, economic parameters, and other required variables into a systematic mining plan covering the entire mining lifecycle, from development to mine closure and reclamation. Currently, FS documents generally assess project feasibility using net present value (NPV), internal rate of return (IRR), and payback period as the main indicators for determining whether a project is feasible.
Theoretically, a project is considered feasible if the NPV is greater than 0, the IRR is higher than the risk-free investment return (such as deposit interest rates or bond yields), and the payback period is shorter than the mine life. However, upon closer examination, these three parameters can be manipulated simply by changing the assumptions used in the calculations. Commodity price assumptions, mining/processing recovery factors, cost assumptions, and discount rate values are highly sensitive variables that can significantly and rapidly alter NPV, IRR, and payback period results.
Variables in FS Are Static in Nature
Furthermore, the variables used in FS calculations are generally static and assumption-based. For example, if the latest benchmark coal price is USD 100 per ton, that value is often applied statically for future years until the end of the mine life. Another example is the use of recovery assumptions. Recovery values are generally considered normal within a range of 80% to 95%. In reality, even a small increase in recovery can have a disproportionately large impact on NPV. In other words, increasing recovery by 5% may increase NPV by more than 5%, since recovery directly affects revenue without necessarily increasing mining costs. Similarly, IRR will increase and payback periods will become shorter simply by applying higher recovery assumptions.
The Influence of Discount Rates on NPV
Apart from cash flow variables, NPV is also highly dependent on the discount rate, which represents project risk and serves as a crucial variable in present value calculations.
There is no universally accepted standard for determining discount rates. Each analyst or expert may use different discount rates depending on their methodology and experience. Analysts may argue that discount rates should range between 5% and 12%. Consequently, to present a more attractive NPV, lower discount rates are often applied. Conversely, higher discount rates are commonly used to portray investments as relatively safer. According to research by Smith (2002), differences in discount rate assumptions in mining projects can alter NPV values by up to 50%, which is highly significant for investment decisions.
Probability Concepts as a Method for Determining Mining Project Feasibility
Therefore, this article discusses a more dynamic approach to determining mining project feasibility using probability concepts. This model addresses the shortcomings of conventional feasibility studies, which are static and heavily assumption-based. The results of this model are presented in the form of project scenarios, where the number of scenarios with NPV > 0 is divided by the total number of scenarios. Ultimately, project feasibility is expressed as a percentage confidence level.
This probability-based feasibility study model utilizes simulation models and/or forecasting techniques for each uncertain variable. We recommend simulating five major variables: commodity prices, average grade, recovery, mining costs, and interest rates or discount rates. Forecasting methods such as regression, exponential smoothing, or autoregressive models can also be applied to determine time-series variables such as commodity prices, mining costs, and interest rates.
Moreover, the mining industry is naturally cyclical, where commodity prices rise and fall over macroeconomic cycles. According to several studies, these five variables are considered critical in mining feasibility analyses (Ardian and Kumral, 2020).
Variable Simulation
Each combination of simulated or forecasted variable values represents a scenario, resulting in varying NPV, IRR, and payback period outcomes. In principle, the more scenarios generated, the better the analysis. Some studies utilize between 1,000 and 1,000,000 scenarios. However, even with fewer scenarios, there will still be a range of NPV values that appear most frequently, which can be considered the expected value.
The simulation results may produce NPV values ranging from approximately negative USD 200 million to positive USD 400 million with varying frequencies. Such distributions also illustrate the project’s confidence level of feasibility. For instance, if 87% of all scenarios produce NPV > 0, the project can be considered feasible with an 87% confidence level.
In addition, the simulation model may indicate an expected NPV value of around USD 100 million, based on the most frequently occurring scenario.
Compared to traditional static assumption-based methods, which only generate a single NPV value and are vulnerable to manipulation, the probability-based approach provides a far more realistic representation. For example, a project may appear to have an NPV of USD 300 million under certain assumptions, but simulation results may reveal that such an outcome only occurs in approximately 1% of scenarios.
Considering the high level of uncertainty in the mining industry, such as commodity price declines or lower-than-expected ore grades, there is always a possibility that projects become economically unfeasible.
For example, simulation results may indicate a 13% probability that a project will generate NPV < 0, meaning the investment could result in economic losses. This provides a much more realistic picture of investment risk.
Recommendations
From a practical standpoint, considerable time is required between the preparation of a feasibility study, government approval, and actual project implementation. During this period, time-sensitive variables such as prices, costs, and interest rates may change significantly. The feasibility study model discussed in this article captures these changes through dynamic scenario analysis.
Currently, risks in FS documents are typically represented by discount rates, while risk profiles are illustrated using sensitivity analysis. However, discount rates cannot adequately capture rare or extreme events such as global financial crises or pandemics, which may cause sudden fluctuations in NPV values.
Likewise, sensitivity analysis only measures changes in NPV, IRR, or payback period by altering one variable at a time while assuming all other variables remain constant. In reality, changes in one variable are usually accompanied by changes in others. For example, high commodity prices are often followed by increased mining costs, while higher ore grades may also improve recovery rates.
Conclusion
In conclusion, this feasibility study model applies probability concepts and confidence percentages to determine mining project feasibility. To generate varying NPV, IRR, and payback period outcomes, the model can utilize simulation methods such as Monte Carlo simulations or stochastic processes, as well as forecasting techniques such as regression or exponential smoothing.
The ultimate goal is to produce more realistic feasibility study analyses through the use of dynamic variables and reduced reliance on assumptions.
The expectation is that feasibility study results will become more realistic by using dynamic variable values and fewer assumptions.
Furthermore, mining project evaluations should no longer rely solely on a single NPV, IRR, or payback period value, but rather on the percentage confidence level that the project will generate NPV > 0. For example, a mining project may be considered feasible if the confidence level for NPV > 0 exceeds 50% or 75%.
Editor: Chaesary Husna R.
