Question

The manager of the I-85 Carpet outlet needs to be able to forecast accurately the demand of Soft Shag carpet (its biggest seller). If the manager does not order enough carpet from the carpet mill, customers will buy their carpets from one of the outlets many competitors. The manager has collected the following demand data for the past eight months.

Month	Demand for Soft Shag Carpet
1	5000
2	10000
3	6000
4	8000
5	14000
6	10000
7	9000
8	12000

1. Compute a three-month moving average forecast for months 6 through 9.

2. Compute a weighted three-month moving average forecast for months 6 through 9. Assign weights of 0.55, 0.33 and 0.12 to the months in sequence, starting with the most recent month.   

3. Compute an exponential smoothed forecast using α=0.4 for months 6 through 9 (the forecasting of the fifth month is 7160 ; F(5)= 7160)                                                     

d. Indicate the most accurate method between a, b and c, and explain why?

in b & c it says to assign like you have to follow equation, so they won't give in the schedule.

Answer 1

For 3-month moving average,

Forecast for any month N = Average of the actual demand of the months (N-1, N-2, N-3)

For 3-month weighted moving average,

Forecast for any month N = 0.55 * Actual demand of month N-1 + 0.33 * Actual demand of month N-2 + 0.12 * Actual demand of month N-3

For Exponential Smoothing,

forecast for any month = forecast for previous month + alpha * (Actual sales previous month - forecast for the previous month)

We calculate the Mean Absolute Demand (MAD) for each of the 3 forecasting methods that give us an idea of the accuracy of the forecasting method.

MAD = Average of Absolute difference between the actual demand and forecast for all the months

please refer to image 1 and 2 for calculations

MAD for 3-month moving average = 2066.67

MAD for 3-month weighted moving average = 2486

MAD for Exponential Smoothing = 2579.76

MAD for a 3-month moving average is the least among all the 3 forecasting methods.

Hence, a 3-month moving average method of forecasting is the most accurate.