Risk AMP
Risk AMP
Risk AMP
www.riskamp.com
Page 1 of 4
www.riskamp.com
%n a Monte Carlo simulation, a random value is selected for each of the tasks, based on the range of estimates. The model is calculated based on this random value. The result of the model is recorded, and the process is repeated. ' typical Monte Carlo simulation calculates the model hundreds or thousands of times, each time using different randomly(selected values. When the simulation is complete, e have a large number of results from the model, each based on random input values. These results are used to describe the likelihood, or probability, of reaching various results in the model. For E#am$le )or e"ample, consider the model described above* e are estimating the total time it ill take to complete a particular project. %n this case, it$s a construction project, ith three parts. The parts have to be done one after the other, so the total time for the project ill be the sum of the three parts. 'll the times are in months. Task ,ob ,ob / ,ob 1 Time +stimate . Months 0 Months . Months
Total -0 Months a!le 1" #asic $orecasting Model %n the simplest case, e create a single estimate for each of the three parts of the project. This model gives us a result for the total time* -0 months. !ut this value is based on three estimates, each of hich is an unkno n value. %t might be a good estimate, but this model can$t tell us anything about risk. 2o likely is it that the project ill be completed on time3 To create a model e can use in a Monte Carlo simulation, e create three estimates for each part of the project. )or each task, e estimate the minimum and ma"imum e"pected time 4based on our e"perience, or e"pertise, or historical information5. We use these ith the 6most likely7 estimate, the one that e used above* Task ,ob ,ob / ,ob 1 Minimum 0 Months 1 Months 0 Months Most 8ikely . Months 0 Months . Months Ma"imum 9 Months : Months : Months -; Months
Total -- Months -0 Months a!le %" $orecasting Model &sing 'ange (stimates
Page % of 4
www.riskamp.com
This model contains a bit more information. <o there is a range of possible outcomes. The project might be completed in as little as -- months, or as long as -; months. %n the Monte Carlo simulation, e ill randomly generate values for each of the tasks, then calculate the total time to completion-. The simulation ill be run .== times. !ased on the results of the simulation, e ill be able to describe some of the characteristics of the risk in the model. To test the likelihood of a particular result, e count ho many times the model returned that result in the simulation. %n this case, e ant to kno ho many times the result as less than or equal to a particular number of months. Time -/ Months -1 Months -0 Months -. Months -: Months -9 Months <umber of Times 4>ut of .==5 1-91;0 0B/ 0;; ?ercent of Total 4@ounded5 =A :A 10A 9;A ;:A -==A -==A
The original estimate for the 6most likely7, or e"pected case, as -0 months. )rom the Monte Carlo simulation, ho ever, e can see that out of .== trials using random values, the total time as -0 months or less in only 10A of the cases. ?ut another ay, in the simulation there is only a 10A chance about - out of 1 that any individual trial ill result in a total time of -0 months or less. >n the other hand, there is a 9;A chance that the project ill be completed ithin -. months. )urther, the model demonstrates that it is e"tremely unlikely, in the simulation, that e ill ever fall at the absolute minimum or ma"imum total values. This demonstrates the risk in the model. !ased on this information, e might make different choices hen planning the project. %n construction, for e"ample, this information might have an impact on our financing, insurance, permits, and hiring needs. 2aving more information about risk at the beginning means e can make a better plan for going for ard.
- %n this e"ample, e use the beta(?+@T distribution to generate random values based on a minimum, most likely, and ma"imum value. The ?+@T distribution is often used to model estimates of e"pert data. )or more information on this and other probability distributions, see the documentation on our ebsite.
Page ) of 4
www.riskamp.com
100%
$igure 1" Pro!a!ility of Completion Within Specified ime *Months+ Ho Relia%le !s !t?
8ike any forecasting model, the simulation ill only be as good as the estimates you make. %t$s important to remember that the simulation only represents probabilities and not certainty. <evertheless, Monte Carlo simulation can be a valuable tool hen forecasting an unkno n future.
CCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCC &%out Ris"&M' @isk'M? is a Monte Carlo simulation engine that orks ith Microsoft +"celD. The @isk'M? 'dd( in adds comprehensive probability simulation to spreadsheet models and +"celD applications. The 'dd(in includes // random distributions, -9 statistical analysis functions, a iEard for creating charts and graphs, and F!'D support all for a fraction of the price of competing packages. )or more information, visit our ebsite at http*GG .risk'M?.com.
Page 4 of 4