Correlated Valuation Methodology (CVM)

Correlated Valuation Methodology (CVM)

Relationships can be seen to exist between parameters in mining valuations, consider the following graphs.

CVM Set of 4 Graphs

These relationships have all been analysed by Quantified Strategies, and all of these are statistically significant. As such each of these parameter pairs can be seen to be correlated, meaning that the two variables move in relation to each other. Note that this does not imply causality.

If these correlations are not considered, the resulting valuations will have an unquantified error. The CVM covers this and more.

The purpose of the CVM is to:

  • Increase the valuation model accuracy.
  • Recognise and incorporate the relationships between parameters.
  • Reduce and quantify the risk resulting from systemic market forces.
  • Reduce the variation in the residual uncertainty.
  • Increase the overall confidence in the valuation outcomes.

The concept has a pragmatic framework which selects a central parameter (e.g.: for a copper mine this would typically be the copper price), and then seeks to assess and model the correlations between the remaining key correlated parameters (typically including diesel, steel-based consumables, labour, power, tyres, explosives, chemicals, etc., etc.).

Whilst sections of the CVM analysis are complicated, the concept overview and the presentation of the outcomes are designed to be simple to interpret and easy to understand.

Once the CVM has been incorporated in the financial valuation framework it becomes possible to calculate the internal cashflow probability (ICP) of both the preceding valuation and the CVM valuation. The ICP is a measure of the risk consequent from systemic market forces. This risk is considered to be comprised primarily of variations in the expected outcomes from the correlated parameters.

Resultant CVM internal cashflow probabilities are typically +P80, whereas standard models which do not incorporate the correlations between parameters are often sub-P50. Thus, the accuracy of an otherwise high-quality financial model can be significantly lower if the relationships between key parameters are not recognised and modelled.

Subsequent to calculating the internal cashflow probability it is then possible to calculate a value at risk (VaR) as a measure of the value of the resultant risk-based exposure to systemic market forces.

The main tasks required to be completed when incorporating the CVM into a valuation model includes the following.

  1. Select a central parameter. This would typically be the main revenue driver, i.e., for a copper mine this would be the copper price.
  2. Assess and analyse the historical relationships between the central parameter and each of the other relevant input parameters.
  3. Where correlations are identified, develop a regression analysis equation including the central parameter as a variable in the equation.
  4. Utilise a combination of the predictive time series for the central parameter (typically sourced from the corporate assumptions) and the CVM equations developed to generate the cashflow statement.
  5. Measure the resulting internal probability of the cashflow.
  6. Simulate and assess the residual uncertainty.

Almost all projects and operations have the main revenue price assumption (the central parameter) as the parameter to which the value is the most sensitive.

If the uncertainty in this variable is modelled stochastically inclusive of the CVM, the resultant distribution changes. Consider the following two graphs.

Note that the second distribution which is inclusive of the CVM has a narrower distribution and a reduced range. The range has been reduced from –US$806M – US$4,299M in the first graph (excluding the CVM), to –US$89M – US$3,645M for the second graph (inclusive of the CVM). Subsequent to including the CVM this project could say with a level of confidence based on sound analyses that it should not expect to lose circa a billion dollars.

The inclusion of the CVM in the project valuation has resulted in the uncertainty-based variance being reduced by 19%, and the standard deviation by 10%.

The coefficient of uncertainty (coefficient of variation) has been reduced from 0.52 to 0.38, a reduction of 27%, and therefore a significant improvement.

The final valuation outcome, inclusive of the uncertainty caused by market volatility, is now more constrained. An intuitive outcome which allows greater confidence in investment decisions.