Many supply managers use a monthly reported survey result known as the purchasing managers' index (PMI) as a leading indicator to forecast future sales for their businesses. Suppose that the PMI and your business sales data for the last 10 months are the following:
|
Month:
|
1
|
2
|
3
|
4
|
5
|
6
|
7
|
8
|
9
|
10
|
|
PMI:
|
42.1
|
43.0
|
41.0
|
38.2
|
40.2
|
44.1
|
45.8
|
49.0
|
48.7
|
52.0
|
|
Sales (1000s):
|
121
|
123
|
125
|
120
|
118
|
118
|
122
|
127
|
135
|
136
|
a. Construct a causal regression model using PMI as the causal variable. How well does your model fit the data?
b. Suppose that the PMI is truly a leading indicator. That is, the PMI value in one period influences sales in the following period. Construct a new regression model using this information. Is the new model better or worse than the model you made for part a?
c. Pick the best model from parts a and b, and create a forecast for sales given PMI 5 47.3.
Text Book: Managing Operation Across the Supply Chain 2nd Edition.