5.4.5 Cost Risk Assessment
220.127.116.11 Description. This technique provides a
program-level cost estimate at completion (EAC) that is a function of
performance (technical), and schedule risks. It uses the results of previous
assessments of WBS elements and cost probability distributions developed for
each of the elements. These individual WBS elements are aggregated using a
Monte Carlo simulation to obtain a probability distribution of the
program-level cost EAC probability distribution function. These results are
then analyzed to determine the actual risk of cost overruns and to identify
the cost drivers.
The use of these cost probability distributions as the basis for
the program-level cost estimate results in a more realistic EAC than the
commonly used single point estimates for WBS elements, since they address both
the probability of occurrence and consequences/impacts of potential risk events.
Their use also eliminates a major cause of underestimating (use of point
estimates) and permits the evaluation of performance (technical) or schedule
causes of cost risk. Thus, this technique provides a basis for the
determination of an "acceptable" level of cost risk.
This technique can be used in any of the acquisition phases, preferably at
least once per phase beginning in the CTD Phase although suitable data may not
exist until the System Integration (SI) Part of the SDD Phase in some cases.
It should be used in conjunction with performance (technical) and schedule
risk assessments and may be performed by small Government-industry teams
consisting of risk analysts, cost analysts, schedule analysts and technical
experts who understand the significance of previous performance and schedule
risk assessments. They should report to the Program IPT. This technique
requires close and continuous cooperation among cost analysts and
knowledgeable technical personnel and the support of the prime contractor's
senior management to help get valid cost data.
18.104.22.168 Procedures. Figure 5-8 depicts the process used in
applying this technique. The first step is to identify the lowest WBS level
for which cost probability distribution will be constructed. The level
selected will depend on the program phase; e.g., during Phase 0, it may not be
possible to go beyond level 2 or 3, simply because the WBS has not yet been
developed to lower levels. As the program advances into subsequent phases and
the WBS is expanded, it will be possible and necessary to go to lower levels
(4, 5, or lower). Specific performance (technical) and schedule risks are then
identified for these WBS elements.
To develop the WBS elements cost probability distributions, the team,
working with the prime contractor's WBS element managers, determines the cost
range for each element being investigated. The cost range encompasses cost
estimating uncertainty, schedule risk, and technical risk. The validity of the
cost data used to construct the distribution is critical. In fact, collecting
good data is the largest part of the cost risk job. Consequently, PMOs
should place major emphasis on this effort.
The element cost probability distributions are aggregated and evaluated
using a Monte Carlo simulation program. All Monte Carlo processes contain
limitations, but they are more informative than point estimates. Any number of
these simulations are readily available to perform this aggregation, and one
that meets the specific needs of the program should be selected. The results
of this step will be a program-level cost EAC and a cost distribution that
shows the cumulative probability associated with different cost values. These
outputs are then analyzed to determine the level of cost risk and to identify
the specific cost drivers. Cost risk is determined by comparing the EAC with
the cost baseline developed as part of the acquisition program baseline. Since
the EAC and program cost distribution are developed from WBS element risk
assessments, it is possible to determine the cost risk drivers. The cost
drivers can also be related back to the appropriate performance and schedule
risks. The results of the analysis (cost risks and drivers) should be
documented in RIFs.