The quantification of risks is an iterative process that relies upon reliable information and sound judgement. Some risks are readily quantifiable due to data available from prior experience in the management of these types of risk, while other risks are not easily quantifiable and an accurate quantification will depend upon the judgement and experience of the project team, as well as expert opinion.
Statistical risk analysis assists the project management team to make informed judgements of the impact of variables that are either controllable or subject to exogenous influences, but that at best can only be anticipated within a reasonable boundary of probability. These are the principal risk drivers, including input prices, technological innovation or obsolescence, vandalism, fire, theft and potentially the total destruction of the asset.
The aim of statistical analysis is to eliminate optimism bias that occurs in many forecasts by introducing an expected outcome as a weighted average of all probable outcomes. In its simplest form, discrete probabilities are assigned to each outcome to derive the expected outcome, as illustrated in Table 1 above. The assignment of probabilities can be assisted by software packages that provide multiple probability distributions. The use of probability distributions also addresses estimation problems when historical data on specific risk items is not available. The probability distribution, if correctly described, effectively replicates the historical data that would otherwise be used in the analysis.
For highly complex projects with correlated risks, consideration should be given to the use of advanced techniques such as Monte Carlo Simulation, which is used to estimate the probability distribution of a model's output by a random sampling of the probability distribution of each variable. Discrete probabilities can be assigned to each outcome, and for projects that may be implemented in a highly complex environment; Monte Carlo analysis can also be used to construct scenario outcomes relating to distant future periods.
Risk adjusted costs are therefore derived primarily from statistical risk analysis as discussed in the following section. However, when estimating the probability and consequence of risk, it is important to keep in mind that not all risks need to be assessed by means of advanced risk valuation techniques. A risk that has a low operational or financial impact can be evaluated using subjective probabilities that are based upon judgement and experience. In these cases, the value if the risk is simply:
Value of risk = probability x consequence + contingency (if necessary)
Even though an individual risk may have a low impact, the accumulated value of these types of risks can be significant. It is therefore preferable to use a subjective estimate that may not be completely accurate rather than no estimate at all. In practice, the experience and judgement of the project team should produce reasonable values for most risks.