Question

Question 3. Monthly demand at A&D Electronics for flat-screen TVs are as follows:

Month Demand (units)

1 1,000

2 1,113

3 1,271

4 1,445

5 1,558

6 1,648

7 1,724

8 1,850

9 1,864

10 2,076

11 2,167

12 2,191

Estimate demand for the next two weeks using simple exponential smoothing with a = 0.3 and Holt’s model with a = 0.05 and b = 0.1. For the simple exponential smoothing model, use the level at Period 0 to be L0 =1,659 (the average demand over the 12 months). For Holt’s model, use level at Period 0 to be L0= 948 and the trend in Period 0 to be T0 = 109 (both are obtained through regression). Evaluate the MAD, MAPE, MSE, bias, and TS in each case. Which of the two methods do you prefer? Why?

Note: Please, solve the problems by using MS Excel.

Answer 1

Using the simple exponential smoothing with a = 0.3:

Alpha	0.3
Data			Forecasts and Error Analysis
Period	Demand		Forecast	Error	Absolute	Squared	Abs Pct Err
Period 1	1,000		1000	0	0	0	00.00%
Period 2	1,113		1000	113	113	12769	10.15%
Period 3	1,271		1033.9	237.1	237.1	56216.41	18.65%
Period 4	1,445		1105.03	339.97	339.97	115579.6	23.53%
Period 5	1,558		1207.021	350.979	350.979	123186.3	22.53%
Period 6	1,648		1312.315	335.6853	335.6853	112684.6	20.37%
Period 7	1,724		1413.02	310.9797	310.9797	96708.38	18.04%
Period 8	1,850		1506.314	343.6858	343.6858	118119.9	18.58%
Period 9	1,864		1609.42	254.5801	254.5801	64811.01	13.66%
Period 10	2,076		1685.794	390.206	390.206	152260.8	18.80%
Period 11	2,167		1802.856	364.1442	364.1442	132601	16.80%
Period 12	2,191		1912.099	278.901	278.901	77785.75	0.1272939
			Total	3319.231	3319.231	1062723	193.83%
			Average	276.6026	276.6026	88560.23	16.15%
				Bias	MAD	MSE	MAPE
					SE	325.9943
Next period	1995.76933

Using the Holt’s model with a = 0.05 and b = 0.1:

Alpha	0.05
Beta	0.1
Data			Forecasts and Error Analysis
Period	Demand		Smoothed Forecast, Ft	Smoothed Trend, Tt	Forecast Including Trend, FITt	Error	Absolute	Squared	Abs Pct Err
Period 1	1,000		1000		1000	0	0	0	00.00%
Period 2	1,113		1000	0	1000	113	113	12769	10.15%
Period 3	1,271		1005.65			265.35	265.35	70410.62	20.88%
Period 4	1,445		1018.918			426.0825	426.0825	181546.3	29.49%
Period 5	1,558		1040.222			517.7784	517.7784	268094.4	33.23%
Period 6	1,648		1066.111			581.8895	581.8895	338595.3	35.31%
Period 7	1,724		1095.205			628.795	628.795	395383.1	36.47%
Period 8	1,850		1126.645			723.3552	723.3552	523242.8	39.10%
Period 9	1,864		1162.813			701.1875	701.1875	491663.9	37.62%
Period 10	2,076		1197.872			878.1281	878.1281	771109	42.30%
Period 11	2,167		1241.778			925.2217	925.2217	856035.2	42.70%
Period 12	2,191		1288.039			902.9606	902.9606	815337.9	0.412123
			Total			6663.748	6663.748	4724188	368.46%
			Average			555.3124	555.3124	393682.3	30.70%
						Bias	MAD	MSE	MAPE
							SE	687.3273

A simple exponential smoothing method is better because it has lower MSE, MAD And MAPE values.