In: Operations Management
A company has the following history of sales for an item.
Month Sales Volume
Jan. 120
Feb. 90
Mar. 100
Apr. 75
May 110
Jun. 50
Jul. 75
Aug. 130
Sep. 110
Oct. 90
Compute the monthly demand forecast for April through November using a 3-month moving average. Compute the monthly demand forecast for June through November using a 5-month moving average. Compute the monthly demand forecast for April through November using a 3-month weighted moving average. Use weights of 0.5, 0.33, and 0.17, with the heavier weights on the more recent months. Compute the mean absolute deviation (MAD) for June through October for each of the methods used. Which method would you use to forecast demand for November?
3-month moving average: It is calculated by the average of the sales volumes of the previous 3 months
So, the 3-month moving average of April = (Sales Volume of Jan + Sales Volume of Feb + Sales Volume of March) / 3
The 3-month moving average from April to November is calculated below:
Month | Sales Volume | 3-month moving average |
Jan | 120 | |
Feb | 90 | |
Mar | 100 | |
Apr | 75 | 103.33 |
May | 110 | 88.33 |
Jun | 50 | 95 |
Jul | 75 | 78.33 |
Aug | 130 | 78.33 |
Sep | 110 | 85 |
Oct | 90 | 105 |
Nov | 110 |
5-month moving average: It is calculated by the average of the sales volumes of the previous 5 months
So, the 5-month moving average of June = (Sales Volume of Jan + Sales Volume of Feb + Sales Volume of March Sales Volume of April + Sales Volume of May) / 5
The 5-month moving average from June to November is calculated below:
Month | Sales Volume | 5-month moving average |
Jan | 120 | |
Feb | 90 | |
Mar | 100 | |
Apr | 75 | |
May | 110 | |
Jun | 50 | 99 |
Jul | 75 | 85 |
Aug | 130 | 82 |
Sep | 110 | 88 |
Oct | 90 | 95 |
Nov | 91 |
3-month weighted moving average: It is calculated by below formula
The 3-month weighted moving average of April = ((Sales Volume of Jan x Weight of Jan)+ (Sales Volume of Feb x Weight of Feb) + (Sales Volume of March x Weight of March) / (Total Weight)
The weights are 0.5, 0.33 and 0.17, and the highest weight shall be allocated to the most recent months.
So in the above formula, the Weight of March shall be 0.5, Weight of Feb shall be 0.33 and Weight of Jan shall be 0.17.
Also, Total weight = 0.5 + 0.33 + 0.17 = 1
The 3-month weighted moving average from April to November is calculated below:
Month | Sales Volume | 3-month Weighted moving average |
Jan | 120 | |
Feb | 90 | |
Mar | 100 | |
Apr | 75 | 100.1 |
May | 110 | 85.8 |
Jun | 50 | 96.75 |
Jul | 75 | 74.05 |
Aug | 130 | 72.7 |
Sep | 110 | 98.25 |
Oct | 90 | 110.65 |
Nov | 103.4 |
Mean Absolute Deviation:
Mean Absolute Deviation is calculated as the average value of absolute errors(Sales - Forecast) for all months.
So, for 3-month moving average:
Month | Sales Volume | 3-month moving average | Error | Absolute Error |
Jan | 120 | |||
Feb | 90 | |||
Mar | 100 | |||
Apr | 75 | 103.33 | -28.33 | 28.33 |
May | 110 | 88.33 | 21.67 | 21.67 |
Jun | 50 | 95 | -45.00 | 45.00 |
Jul | 75 | 78.33 | -3.33 | 3.33 |
Aug | 130 | 78.33 | 51.67 | 51.67 |
Sep | 110 | 85 | 25.00 | 25.00 |
Oct | 90 | 105 | -15.00 | 15.00 |
Nov | 110 |
MAD = (28.33 + 21.67 + 45 + 3.33 + 51.67 + 25 + 15) / 7 = 190 / 7 = 27.17
For 5-month moving average:
Month | Sales Volume | 5-month moving average | Error | Absolute Error |
Jan | 120 | |||
Feb | 90 | |||
Mar | 100 | |||
Apr | 75 | |||
May | 110 | |||
Jun | 50 | 99 | -49 | 49 |
Jul | 75 | 85 | -10 | 10 |
Aug | 130 | 82 | 48 | 48 |
Sep | 110 | 88 | 22 | 22 |
Oct | 90 | 95 | -5 | 5 |
Nov | 91 |
MAD = ( 49 + 10 + 48 + 22 + 5 ) / 5 = 134 / 5 = 26.8
For 3-month weighted moving average:
Month | Sales Volume | 3-month Weighted moving average | Error | Absolute Error |
Jan | 120 | |||
Feb | 90 | |||
Mar | 100 | |||
Apr | 75 | 100.1 | -25.1 | 25.1 |
May | 110 | 85.8 | 24.2 | 24.2 |
Jun | 50 | 96.75 | -46.75 | 46.75 |
Jul | 75 | 74.05 | 0.95 | 0.95 |
Aug | 130 | 72.7 | 57.3 | 57.3 |
Sep | 110 | 98.25 | 11.75 | 11.75 |
Oct | 90 | 110.65 | -20.65 | 20.65 |
Nov | 103.4 |
MAD = ( 25.1 + 24.2 + 46.75 + 0.95 + 57.3 + 11.75 + 20.65) / 7 = 186.7 / 7 = 26.67
For forecasting the demand for November, I would use the 3-month weighted average method. This method considers the sales volume for 3 months and also gives more weightage to recent months. The 3-month moving average just considers the average of 3 months and doesn't give any weightage to months. The 5-month moving average considers a broader range ie 5 months, which may not be very accurate to forecast and also doesn't give any weightage to months.
Due to the above reasons, I would use the 3-month weighted average.