Determine the risk (probability) of meeting forecasted demand

I'm doing a project for a car company for my final year studying Industrial Engineering.
The company forecast demand 4 months ahead but, according to history data, the company rarely meets the forecast demand.

The Supply Chain Department asked me to develop a model that would, with the help of actual historic demand, assign a risk factor or probability that the forecast demand will be met. The SPD wil use this to challenge the Volume Control Department's plan.

My question is which methods/techniques would you suggest to solve this problem with the best accuracy. I thought of Monte Carlo simulation. Is this sufficient or is there better ways. Please help me!!:confused:


TS Contributor
I would first plot the forecasted and the realized value in time order just to see whether the forecast has a tendency to overestimate he demand (or under). Then I would compare the MSE of the demand to the MSE of the forecast error - if the latter is bigger then the forecast is basically useless. Of course, if your demand has a trend you would need to de-trend the data first.

I hope this helps