SCAF Workshop Integrated Cost and Schedule Risk Analysis. Tuesday 15th November 2016 The BAWA Centre, Filton, Bristol

Transcription

1 The following presentation was given at: SCAF Workshop Integrated Cost and Schedule Risk Analysis Tuesday 15th November 2016 The BAWA Centre, Filton, Bristol Released for distribution by the Author

2 What are the chances of that? Quantification with Monte Carlo simulation Doug Oldfield Palisade Corporation

3 Agenda Introduction of myself & Palisade Introduction to Monte Carlo simulation Adding uncertainty to a cost estimate Incorporating a risk register into the estimate Applying uncertainty to a schedule Adding risk events to a schedule Combining cost and schedule uncertainty analysis Questions?

4 Doug Oldfield who am I? Look after sales & business development in the south of the UK & the Middle East Married with a four year old daughter & an eleven month old son Joined Palisade 8 years ago in 2008 and have worked with clients in industries such as oil and gas, defence, pharma, mining, finance, insurance, construction, engineering, retail, manufacturing, and many more. Degree in Politics Contact via doldfield@palisade.com

5 Palisade Corporation a brief overview Palisade was founded in 1984 and we provide risk and decision analysis solutions for multiple industries, Through techniques like Monte Carlo simulation, decision trees & optimisation we enable you to understand risks and make better decisions More than 200,000 users worldwide. Palisade tools are currently used by 93% of the Fortune 100, and 76% of the Fortune 500 companies worldwide Taught annually to more than 40,000 MBA students More than 1,500 training & consulting engagements

6 Some of our clients

7 The starting point

8 The issue

9 Bombs & Casinos: Embryonic Simulation

10 Monte Carlo simulation Uses probability distributions to capture and model does this within Microsoft Excel and Project. Generates thousands of scenarios to give probabilistic outputs instead of static (deterministic) ones. Uses confidence levels (P-values) to quantify the likelihood of events or outcomes.

11 Why is this important? Distributions everywhere!

12 Why is quantitative risk analysis important? Modeling - create a representation of a real-life situation to analyze it, usually mathematically. Traditional analyses combine single point estimate of a model s variables to predict a single estimate of results. Combined errors in each estimate often lead to significantly different real-life results. Decision you made on your expected results might have been different if you had a complete picture of all possible outcomes.

13 The starting point Microsoft Excel

14 How a regular Excel model works Volume Project Metrics Costs Point Estimates Price/Mix Cost Static (deterministic) Excel Model Start Date NPV IRR Revenue OOIP

15

16

17 Monte Carlo how does it work? Build a model in Excel, or use an existing one. Determine which components are uncertain in your model then convert them into ranges using probability then generates thousands of scenarios for each of the uncertain inputs The resulting output scenario (cost, NPV, risks, etc.) are computed over and over again.

18 Monte Carlo models Value Ranges Monte Carlo Simulation Outcomes in ranges: Costs, Start Date, NPV, IRR, Volume Price/Mix Cost Same Excel Model Project Metrics Key risk drivers Revenue Risk and opportunities

19 Benefits of Monte Carlo simulation #1 Makes models less wrong & more useful! You get to see everything that might happen: best case, worst case, most likely case, and everything in between You get to see the different probabilities of these outcomes occurring Fewer surprises and better prepared for contingencies Understand what could happen, and then knowing how likely it is, be able to plan accordingly Model, test & implement effective mitigation strategies

20 Benefits of Monte Carlo simulation #2 Capture correlations between related variables Identify drivers of risk Sensitivity analysis ranks variables according to impact on your output and enables you to understand which factors are most important Target specific variables which have the biggest impact Don t waste resources on low-impact events, or allocate too much to low-probability events

21 Monte Carlo simulation Let s take a look.