Question

During the past 4 quarters, the Port of Baltimore has unloaded large quantities of grain from ships.  The ports operations manager wants to test the use of exponential smoothing to see how well the technique works in predicting tonnage unloaded.  He guesses that the forecast of grain unloaded in the first quarter was 175 tons.  Two values of α are examined: α = .10 and α =.50.  Calculate the forecast for the 5^th quarter. Make sure to fill the spaces left blank and show all your calculations

Quarter	Actual Tonnage Unloaded	Forecast  (0.10)	Error	Absolute Error	Forecast  (0.50)	Error	Absolute Error
1	180	175	5	5	175	5	5
2	168	176	-8	8	178	-10	10
3	159	175	-16	16	173	-14	14
4	175	173	2	2	166	9	9
5	190	173	17	17	170.5	9	9
		ME4 =	-4.25		ME4 =	-2.75
		MAD4 =	7.75		MAD4 =	9.5

a)Based on the data shown above, can you decide which α is best to be used? Explain the reason for your answer, and how would you proceed in order to make a decision if your answer for the previous question was in the negative.

b)Can you say which model tends to underestimate more the demand? What are the implications with respect to operations of a forecast that tends to underestimate demand? (Write down your assumptions about the operations.)

c)Can you say which model gives a higher error magnitude? What are the implications with respect to operations of a higher error magnitude? (Write down your assumptions about the operations.)

Answer 1

(i) Upto period 4 forecast values, MAD is lower with alpha =0.1 than that with alpha =0.5, hence it should be the choice. However, after including period 5, the MAD5 =9.6 for alpha =0.1 while MAD5=9.4 for alpha =0.5 which makes the second one more favourable.

(Ii) ME5 =0 for alpha =0.1 while ME5 is -1 for alpha =0.5 which means that the model with alpha =0.1 more underestimates the demand ( errors are more positive than negative). If the forecasts are lower than anticipated demand, there is a risk of shortage of products / services which might result in lost opportunities.

(iii) HIgher error magnitude is with alpha =0.1, which means that there is higher variability from the forecast ( both positive and negative) which makes the forecasting process less efficient and the company might end up either overstocking which increases inventory cost, or make run short of supplies which might result in lost sales.