To continue manufacturing rather than outsourcing, the difference (Delta) between manufacturing and outsourcing costs must be greater than zero. The delta (difference) was calculated as $50,000 + 125 * 1500 units - 158 * 1500 units, yielding $500. In this scenario, the decision is to manufacture. The mean of demand is 1500, with a standard deviation of 300. After 500 trials of simulation, the mean of the differences (Delta) is 745.8509, and the standard deviation is 9695.5468. The data from the 500 experiments was classified as either outsource or manufacturing. Outsourcing and manufacturing had 235 and 265 people, respectively. From this, it is evident that the proportion to outsource (Delta0) is 0.53. Since the proportion to outsource is smaller than the proportion to manufacture, the decision is to manufacture.
Using the F9 key in excel to recalculate this simulation ten times yielded the following values of the mean difference, Delta:
Time
Mean Difference
Decision
1
563.45
Manufacture
2
220.20
Manufacture
3
-116.61
Outsource
4
-386.82
Outsource
5
108.00
Manufacture
6
557.39
Manufacture
7
-129.29
Outsource
8
1157.37
Manufacture
9
-7.94
Outsource
10
440.36
Manufacture
Running the simulation yields different values every time. This explains the reason in the different mean differences Delta. From the table above, it is evident that of the ten times, the decision was to manufacture on six occasions while the decision was to outsource on 4 occasions. If more simulations are conducted, different values will be obtained. Since the decision to manufacture occurs more often than the decision to outsource, the overall decision in this case is to manufacture.
The simulation is the repeated again for 10,000 trials. After running the simulation once, the mean difference of Delta is 370.7816 and the standard deviation is 9935.1423. Next, the counts of the occasions on when the decision is to outsource or manufacture were obtained. The counts were 4854 for outsourcing and 5146 for manufacturing. The proportions are 0.4854 and 0.5146 respectively. It is evident that the proportion of the decision to manufacture is greater than the proportion of the decision to outsource and hence the overall decision is to manufacture. If more simulations are carried out, the proportions of the decision on whether to outsource or manufacture will change. At 0.05 level of significance, the confidence interval for the mean was obtained with a lower limit of 176.0564 and an upper limit of 565.0568. This means that one is 95% confident that the mean difference (Delta) lies within 176.0564 and 565.0568 and 5% confident that the mean difference will be either less than 176.0564 or greater than 565.0568.
If the mean of demand increases to 10,000 units and have a standard deviation of 1,000 units and a simulation of 10,000 trials is carried out, the mean differences, Delta were obtained. The mean difference was -279772.74 which is less than zero and hence the decision to outsource. The counts of when the loss will be more than $100,000 and less than $100,000 were obtained. The counts were 10,000 and 0 respectively. The probability that the loss will be more than $100,000
if they choose to keep the internal line operating instead of outsourcing the part was 1, that is, 10,000/10,000.
