Question

3. The following time series represent the annual demand for 50 lb bags of fertilizer for a garden supply company between 1991 and 2002.

Year | Bags

1991 | 4000

1992 | 6000

1993 | 4000

1994 | 5000

1995 | 10,000

1996 | 8000

1997 | 7000

1998 | 9000

1999 | 12,000

2000 | 14,000

2001 | 15,000

2002 | 14,000

a. Use a 2-period moving average to forecast the demand for the bags of fertilizers in 2003.

b. Use a 3-period moving average to forecast the demand for the bags of fertilizers 2003

c. Which averaging period provides a better historical fit in your opinion and why?

d. Use a 3-period weighted moving average to forecast the demand for the bags of fertilizers in 2003. Use solver to determine the optimal weights based on minimizing the MAPE criterion.

e. Use exponential smoothing with a smoothing constant of 0.8 (remember the difference between the smoothing constant and the damping factor on Excel) to forecast the demand for the bags of fertilizers in 2003.

f. Use exponential smoothing with a smoothing constant of 0.5 to forecast the demand for the bags of fertilizers in 2003.

g. Which one of the above (part e or part f) is a more accurate forecast. Explain in detail.

Answer 1

a. 2-period moving average to forecast the demand for the bags of fertilizers in 2003.

2001	15,000	2-period moving average
2002	14,000	= 15000+14000 ) /2 =
2003		= 14500

b. 3-period moving average to forecast the demand for the bags of fertilizers 2003

2000	14,000
2001	15,000	2-period moving average
2002	14,000	= 14000+15000+14000 +) /3 =
2003		= 14333.33

c. period 3 years of moving average is better because greater the number of periods in the moving average, smoother will be average demand forecasing .