Level:
TCM Section(s):
Venue: 2023 AACE International Conference & Expo
Abstract: Effective estimators develop risk-adjusted models that account for statistical uncertainty to calculate a range of possible cost outcomes for a project. The coefficient of variation (CV) of a risk-adjusted model is a statistic that normalizes uncertainty to determine if appropriate bounds have been captured in the estimate. The bottom-up application of cost uncertainty requires the use of Monte Carlo simulations. Despite many benefits, this approach is susceptible to underestimation of uncertainty for top-level work breakdown structure (WBS) elements. Failure to understand the causes of this underestimation could prevent realistic generation of bounding estimates. Unreasonable upper bounds for potential costs may present false flags while assessing performance as part of the PDCA cycle of TCM, as reasonable cost growth exceeds variance from the underestimated plan.
This topic illustrates how elements such as WBS size, correlation, and uncertainty of lower-level elements can impact top-level spread in a WBS using a derived equation and randomly generated WBSs. The discussion will identify root causes of risk underestimation in Monte Carlo based models along with mitigation steps for addressing the underlying issues. Program management professionals and cost estimators can apply these insights and guidelines to manage cost-growth risks more effectively by more accurately quantifying cost uncertainty to enable data-driven decision making.
TCM Section(s):
Venue: 2023 AACE International Conference & Expo
Abstract: Effective estimators develop risk-adjusted models that account for statistical uncertainty to calculate a range of possible cost outcomes for a project. The coefficient of variation (CV) of a risk-adjusted model is a statistic that normalizes uncertainty to determine if appropriate bounds have been captured in the estimate. The bottom-up application of cost uncertainty requires the use of Monte Carlo simulations. Despite many benefits, this approach is susceptible to underestimation of uncertainty for top-level work breakdown structure (WBS) elements. Failure to understand the causes of this underestimation could prevent realistic generation of bounding estimates. Unreasonable upper bounds for potential costs may present false flags while assessing performance as part of the PDCA cycle of TCM, as reasonable cost growth exceeds variance from the underestimated plan.
This topic illustrates how elements such as WBS size, correlation, and uncertainty of lower-level elements can impact top-level spread in a WBS using a derived equation and randomly generated WBSs. The discussion will identify root causes of risk underestimation in Monte Carlo based models along with mitigation steps for addressing the underlying issues. Program management professionals and cost estimators can apply these insights and guidelines to manage cost-growth risks more effectively by more accurately quantifying cost uncertainty to enable data-driven decision making.