Risk Quantification: Key to Investment Decision Making, and Project Control

To deliver an effective project plan, you need quality, targeted base cost and duration estimates. Cost estimating and scheduling use many techniques to translate project scope deliverable information into solid, integrated “base” estimates of time, costs and resources. However, all plans are uncertain and, at early scope definition phases, so too is the scope. So how are you going to deal with all the uncertainty? This is what Risk Quantification helps you with. 

Base estimates are just the starting point to understanding the range of possible cost and schedule outcomes. You need to make risk-aware investment decisions and establish a full control base including specific provisions for uncertainty and risk. AACE International® (AACE) refers to those provisions as contingency, escalation and currency, and management reserves, and the techniques to quantify them are called quantitative risk analysis or QRA.

Effective Quantitative Risk Analysis is About Modelling and Statistics, Grounded in Reality

While base estimating is about deterministically translating defined scope into time, cost and resources, Quantitative Risk Analysis is about stochastically predicting a range of possible outcomes recognizing what is not well defined or known now, and lesser known in the future. All quality QRA methods depend on models of the state of uncertainty, how risks will unfold, and how the project will be impacted by and react to them. And all QRA models must be probabilistic; i.e., result of the analysis is a distribution of possible outcomes. Some models are based on historical data which ground the analysis in objective reality, while others are based more on subjective assumptions.

Uncertainties and Risks; Quantitative Risk Analysis Methods Depend on their Nature

Uncertainties are a state of being; we are uncertain about our quantities, our pricing, our assumptions, our base estimate biases and so on. We are also uncertain as to the capability of our project system to perform effectively.  Risks are events or conditions that may occur or be discovered that cause or result in deviations from the base plan. Uncertainties (e.g., weak system) and risks (e.g., a permit delay) can interact and compound. For quantification, all of this must be represented in a model that either has inherent probabilistic properties (i.e., regression based) or as a platform for simulation (e.g., Monte-Carlo Simulation or MCS).

QRA is an evolving area of practice, but in general, the AACE community recognizes there are different types of risks for which different methods are best suited for quantifying them. However, for decision making we need unified, integrated outputs (e.g., total capex and in-service date distributions for NPV), not a hodge-podge; i.e., we need “hybrid” methods.

The Risk Types and Methods in a Nutshell

The uncertainty of the project system, including biases, is called “systemic” risk. Fortunately, we have 50+ years of empirical research that provides reliable, cross-industry models that realistically quantify the cost and duration impact of systemic risks. Being regression based from historical data, these “parametric” models are inherently probabilistic (no MCS required) and are grounded in reality. For risk events and conditions, which are specific to a given project (i.e., “project-specific”), the methods of choice are both MCS simulations of an integrated cost and schedule model. The two most common models are risk-driven, cost-loaded critical path method schedules (CPM) and risk-driven Expected Value (EV) (with EV being most generically applicable to projects of all sizes and phases). The parametric model of systemic risks, and simulation of project-specific risks can be put together in a hybrid.

Lastly, there is external economic-driven price risks which we call “escalation”. We include escalation in our base estimates, but QRA adds the probabilistic element for uncertainty in price trends, expenditure rates, and so on. By using the outputs of our cost and schedule QRA models as inputs to the escalation model, the outcome is integrated for NPV modelling.

Benchmarking and Quantitative Risk Analysis

As mentioned, the parametric method for systemic risks is based on the analysis of historical data. Whether we create our own parametric models (e.g., using regression) or use an industry model and calibrate it for our specific situation, we need to capture historical data in a benchmarking system. Also, for project-specific risks, we benefit from historical experience and lessons learned, also captured in a benchmarking system.

Putting it All Together

In the end, an integrated suite of QRA methods support investment decision making, and through various techniques, the QRA outcomes (considering business objectives and risk appetite) are the basis for determining the contingency, escalation, and reserve accounts for control purposes. Until that is done, no estimate or schedule is complete; the base estimate is just the beginning. However, realistic QRA depends on closing the loop using historical data and benchmarking.

Risk Quantification Seminar

If you want to know more about Quantitative Risk Analysis, join the 2-Day Risk Quantification Seminar in October 2018. During the seminar, you will discover practical methods supported by research and the AACEI, explore method variations by risk type and project phase. build decision and risk management skills, knowledge and competency. You can learn more and register for the seminar following this link: