Identifying risk and recognizing the ability to respond to uncertainty is an important priority for mining and metals companies across Canada and beyond — especially given ongoing metal price volatility. Taking the necessary steps to recognize metal price volatility and the possibility of price declines in investment decision making has never been more important. Industry players cannot afford to ignore the investment effects of uncertainty.
Much of the industry relies on static cash flow (“Static CF”) models for investment analysis and portfolio reviews. These models estimate investment or portfolio cash flows by representing uncertain variables such as metal price or grade as a forecast. The risk exposure generated by these uncertainties may then be explored with a limited number of scenarios and other risk modelling techniques. Unfortunately, while making the cash flow modelling process easier, a forecast can be problematic as it is only a summary representation of an uncertain variable’s future possibilities.
This limitation associated with a forecast is particularly detrimental in today’s mining and metals environment. “All forecasts are ultimately wrong” is often heard when trying to explain away differences between actual metal prices and their corporate forecasts. However, it’s rare to hear this deviation considered to be a reflection of metal price behaviour rather than an embarrassing forecast error. A better approach, rather than relying on a single price forecast for investment decision-making, is to explicitly recognize uncertainty by pairing the forecast with a statistical description of possible price behaviour. Companies can better prepare for uncertainty by replacing a single forecast with a large number of simulated scenarios that describe possible movements around the original forecast
Figure 1 illustrates why simulated price scenarios can provide a better overall description of future price behaviour than a single forecast. In this graph, historic real gold prices are plotted from January 1975 to 31 December 2013. One method of projecting future gold prices from this date for investment analysis is to use a flat forecast based on the final known spot price (grey line). Alternatively, a large number of scenarios can be generated to represent future price possibilities through simulation and an updating uncertainty model that is statistically consistent with historic prices. These simulated price scenarios are exemplified by the five price scenarios (coloured lines) in Figure 1 describing possible gold price behaviour around the original forecast. These simulated scenarios each exhibit the erratic price movements that are consistent with historic gold price behaviour but, when considered in aggregate, have an expectation equal to the original forecast.
The choice between using a single price forecast and simulated price scenarios is important if we wish to continually improve the representation of the business environment in investment models. A Static CF model can be transformed to better reflect the business environment by using simulation to replace a single forecast scenario with a large number of price scenarios. This transformation is important for two reasons:
• The Static CF model and its reliance on a single forecast scenario may provide an erroneous cash flow estimate due to the presence of management flexibility or non-linear financing/taxation cash flow structures. Simulation and other numerical methods can reduce this estimation bias.
• The standard sensitivity or scenario analysis often performed with a Static CF model provides an incomplete description of cash flow uncertainty and investment risk exposures. Understanding potential risk exposures can be improved with numerical methods.
Explicitly recognizing metal price uncertainty with numerical methods such as simulation is not a needless complication for investment decision-making, though this extension should be used selectively. The purpose of techniques such as simulation is to improve the description of future cash flow variability in investment analysis and to ultimately provide management with insight into a proposed investment’s risk exposure that would not otherwise be gained with a Static CF model.
*Michael Samis, Ph.D., P.Eng., is an Associate Partner in EY’s Valuation and Business Modelling, Transaction Advisory Services practice. He is based in Toronto.