Answers Forecasting

2.3 a) Time Series

b) Regression or Causal Model
c) Delphi Method

2.8 The manager should have been prepared for the consequences of forecast errors.

2.10 This type of criteria would be closest to MAPE, since the errors measured are relative
not absolute. It makes more sense to use a relative measure of error in golf. For example,
an error of 10 yards for a 200 yard shot would be fine for most golfers, but a similar
error for a 20 yard shot would not.

2.13
|E1/D| |E2/D|
F1 F2 D E1 E2 (E1)^2 (E2)^2 |E1| |E2| *100 *100
223 210 256 -33 -46 1089 2116 33 46 12.89 17.97
289 320 340 -51 -20 2601 400 51 20 15 5.88
430 390 375 55 15 3025 225 55 15 14.67 4
134 112 110 24 2 576 4 24 2 21.82 1.82
190 150 225 -35 -75 1225 5625 35 75 15.56 33.33
550 490 525 25 -35 625 1225 25 35 4.76 6.67
1523.5 1599.17 37.17 32.17 14.12 11.61
MSE 1 MSE 2 MAD 1 MAD 2 MAPE 1 MAPE 2

275+188+312
2.16 MA (3) forecast: = 258.33
3
223+286+212+275+188+312
MA (6) forecast: = 249.33
6
MA (12) forecast: 205.33

2.17-2.19

Demand
350

300

250

200

150 Demand

100

50

0
0 2 4 6 8 10 12 14

57+144+221+177
Example: two-step ahead forecast July = = 149.75
4
 Forecast 1 Forecast 2
Month (one-step (two-step Demand
ahead) ahead) E1 E2
July 205.50 149.75 223 -17.50 -73.25
August 225.25 205.50 286 -60.75 -80.50
September 241.50 225.25 212 29.50 13.25
October 250.25 241.50 275 -24.75 -33.50
November 249.00 250.25 188 61.00 62.25
December 240.25 249.00 312 -71.75 -63.00
MAD 44.2 54.3

The one step ahead forecasts gave better results (and should have according to the theory).

350

300

250

200 Demand
1-step-fcst
150
2-step-fcst
100

50

0
0 2 4 6 8 10 12 14

2.20

Month Demand MA(3) MA(6)

July 223 226 161.33
August 286 226.67 183.67
September 212 263 221.83
October 275 240.33 233.17
November 188 257.67 242.17
December 312 225 244

MA (6) Forecasts exhibit less variation from period to period.

2.22
𝜶=0.15
𝐹𝐹𝑒𝑏 = 0.15 ⋅ 23.3 + 0.85 ⋅ 25 = 24.745
𝐹𝑀𝑎𝑟𝑐ℎ = 0.15 ⋅ 72.3 + 0.85 ⋅ 24.745 = 31.88
𝐹𝐴𝑝𝑟 = 0.15 ⋅ 30.3 + 0.85 ⋅ 31.88 = 31.64
𝐹𝑀𝑎𝑦 = 0.15 ⋅ 15.5 + 0.85 ⋅ 31.63 = 29.22
𝜶=0.4
𝐹𝐹𝑒𝑏 = 0.40 ⋅ 23.3 + 0.60 ⋅ 25 = 24.32

Month Demand 𝜶=0.15 𝜶=0.4

January 23.3
February 72.3 24.745 24.32
March 30.3 31.88 43.47
April 15.5 31.64 38.20
May 29.22 29.12
MSE 838.04 993.74

α = .15 gave a better forecast

2.23 Small α implies little weight is given to the most recent demand observation and
virtually all weight is given to the most recent forecast. This means that the forecast
will be stable but not responsive.

Large α implies that a great deal of weight is applied to the most recent demand
observation. This means that the forecast will adjust quickly to changes in the demand
pattern but will vary considerably from period to period.

2.28
a) “Eyeball” estimates:

Demand
3000

2500

2000

1500

1000

500

0
0 1 2 3 4 5 6 7

2750
slope = 6 = 458.33
intercept = -500
 b)
Month 𝒊 Demand (𝑫𝒊 ) 𝒊𝑫𝒊
January 1 133 133
February 2 183 366
March 3 285 855
April 4 640 2,560
May 5 1876 9,380
June 6 2550 15,300
SUM 5,667 28,594
AVG 944.5

𝑛2 (𝑛+1)(2𝑛+1) 𝑛(𝑛+1) 2
𝑆𝑥𝑥 = 𝑛 ∑ 𝑥𝑖2 − (∑ 𝑥𝑖 )2 = −( ) = 105
6 2
𝑛(𝑛+1)
𝑆𝑥𝑦 = 𝑛 ∑ 𝑥𝑖 𝐷𝑖 − ∑ 𝑥𝑖 ∑ 𝐷𝑖 = 𝑛 ∑ 𝑖𝐷𝑖 − ∑ 𝐷𝑖 = 52,557
2

𝑆𝑥𝑦 52,557
𝑏=𝑆 = = 500.54
𝑥𝑥 105
1 1 1 𝑛+1
𝑎 = 𝑛 ∑ 𝐷𝑖 − 𝑏 ∙ 𝑛 ∑ 𝑖 = 𝑛 ∑ 𝐷𝑖 − 𝑏 ∙ = −807.39
2

c) Regression equation: 𝐹𝑡 = −807.4 + 500.54 ⋅ 𝑡

𝐹𝐽𝑢𝑙𝑦 = 𝐹7 = −807.39 + 500.54 ∙ 7 = 2,696.39
𝐹𝐷𝑒𝑐 = 𝐹12 = −807.39 + 500.54 ∙ 12 = 5,199.09

Month Forecast
July 2696
August 3197
September 3698
October 4198
November 4699
December 5199

d) One would think that peak usage would be in the summer months and the
increasing trend would not continue indefinitely. A warning sign came from the change-
in-trend from March to April, as observed in the graph; regression assumes a constant
trend.

2.30 a) 𝑆0 == −807.39 + 500.54 ∙ 6 = 2,196, 𝐺0 = 500.54

𝛼 = 0.15, 𝛽 = 0.10
July:
𝑆1 = 𝛼𝐷1 + (1 − 𝛼)(𝑆0 + 𝐺0 ) = 0.15 ∙ 2,150 + 0.85 ∙ (2,196 + 500.54) = 2,615
𝐺1 = 𝛽(𝑆1 − 𝑆0 ) + (1 − 𝛽)𝐺0 = 0.10 ∙ (2,615 − 2,196) + 0.90 ∙ 500.54 = 492.4

August:
𝑆2 = 0.15 ∙ 2,660 + 0.85 ∙ (2,615 + 492.4) = 3,040
𝐺2 = 0.10 ∙ (3,040 − 2,615) + 0.90 ∙ 492.4 = 485.7

b) One-step-ahead forecast in August for September:

𝐹𝑎𝑢𝑔,𝑠𝑒𝑝 = 𝑆2 + 𝐺2 = 3,525.7
 Two-step-ahead forecast in August for October:
𝐹𝑎𝑢𝑔,𝑜𝑐𝑡 = 𝑆2 + 2𝐺2 = 4,011.4

c) Forecast in July for December:

𝐹𝑗𝑢𝑙,𝑑𝑒𝑐 = 𝑆1 + 5𝐺1 = 2,615 + 5 ∙ 492.4 = 5,077

2.31 This observation would lower future forecasts. Since it is probably an "outlier" (non-
representative observation) one should not include it in forecast calculations.

2.32 Both regression and Holt's method are based on the assumption of constant linear trend.
It is likely in many cases that the trend will not continue indefinitely or that the observed
trend is just part of a cycle. If that were the case, significant forecast errors could result.

2.33
1. Compute sample mean entire data set
2. Divide each observation by sample mean
3. Average factors for like-periods in a season

𝑌̅ = 58.96 sample mean over Year 1 and Year 2

January:
12 16
+
58.96 58.96 = 0.24
2

Month Year 1 Year 2 𝒀𝟏 𝒀𝟐 Seasonal

(𝒀𝟏 ) (𝒀𝟐 ) ̅
𝒀 ̅
𝒀 factor
January 12 16 0.20 0.27 0.24
February 18 14 0.31 0.24 0.27
March 36 46 0.61 0.78 0.70
April 53 48 0.90 0.81 0.86
May 79 88 1.34 1.49 1.42
June 134 160 2.27 2.71 2.49
July 112 130 1.90 2.20 2.05
August 90 83 1.53 1.41 1.47
September 66 52 1.12 0.88 1.00
October 45 49 0.76 0.83 0.80
November 23 14 0.39 0.24 0.31
December 21 26 0.36 0.44 0.40
TOTAL 689 726

2.51

Summary:
MAD MSE MAPE
MA(3) 86.62 10547.47 38.31
MA(6) 101.07 16663.00 49.73
a) and b)

Month Sales MA(3) Error |Error| Error^2 |Error|%

1 238
2 220
3 195
4 245 217.67 -27.33 27.33 747.11 11.16
5 345 220.00 -125.00 125.00 15625.00 36.23
6 380 261.67 -118.33 118.33 14002.78 31.14
7 270 323.33 53.33 53.33 2844.44 19.75
8 220 331.67 111.67 111.67 12469.44 50.76
9 280 290.00 10.00 10.00 100.00 3.57
10 120 256.67 136.67 136.67 18677.78 113.89
11 110 206.67 96.67 96.67 9344.44 87.88
12 85 170.00 85.00 85.00 7225.00 100.00
13 135 105.00 -30.00 30.00 900.00 22.22
14 145 110.00 -35.00 35.00 1225.00 24.14
15 185 121.67 -63.33 63.33 4011.11 34.23
16 219 155.00 -64.00 64.00 4096.00 29.22
17 240 183.00 -57.00 57.00 3249.00 23.75
18 420 214.67 -205.33 205.33 42161.78 48.89
19 520 293.00 -227.00 227.00 51529.00 43.65
20 410 393.33 -16.67 16.67 277.78 4.07
21 380 450.00 70.00 70.00 4900.00 18.42
22 320 436.67 116.67 116.67 13611.11 36.46
23 290 370.00 80.00 80.00 6400.00 27.59
24 240 330.00 90.00 90.00 8100.00 37.50

MAD MSE MAPE

86.62 10547.47 38.31
c)

Month Sales MA(6) Error |Error| Error^2 |Error|%

1 238
2 220
3 195
4 245
5 345
6 380
7 270 270.50 0.50 0.50 0.25 0.19
8 220 275.83 55.83 55.83 3117.36 25.38
9 280 275.83 -4.17 4.17 17.36 1.49
10 120 290.00 170.00 170.00 28900.00 141.67
11 110 269.17 159.17 159.17 25334.03 144.70
12 85 230.00 145.00 145.00 21025.00 170.59
13 135 180.83 45.83 45.83 2100.69 33.95
14 145 158.33 13.33 13.33 177.78 9.20
15 185 145.83 -39.17 39.17 1534.03 21.17
 16 219 130.00 -89.00 89.00 7921.00 40.64
17 240 146.50 -93.50 93.50 8742.25 38.96
18 420 168.17 -251.83 251.83 63420.03 59.96
19 520 224.00 -296.00 296.00 87616.00 56.92
20 410 288.17 -121.83 121.83 14843.36 29.72
21 380 332.33 -47.67 47.67 2272.11 12.54
22 320 364.83 44.83 44.83 2010.03 14.01
23 290 381.67 91.67 91.67 8402.78 31.61
24 240 390.00 150.00 150.00 22500.00 62.50

MAD MSE MAPE

101.07 16663.00 49.73

MA(6) has higher MAD, MSE and MAPE .

2.52

Month Sales Forecast Error |Error| Error^2 |Error|% Alpha

1 238 225.00 -13.00 13.00 169.00 5.46 0.1
2 220 226.30 6.30 6.30 39.69 2.86
3 195 225.67 30.67 30.67 940.65 15.73
4 245 222.60 -22.40 22.40 501.63 9.14
5 345 224.84 -120.16 120.16 14437.78 34.83
6 380 236.86 -143.14 143.14 20489.51 37.67
7 270 251.17 -18.83 18.83 354.47 6.97
8 220 253.06 33.06 33.06 1092.65 15.03
9 280 249.75 -30.25 30.25 915.07 10.80
10 120 252.77 132.77 132.77 17629.15 110.65
11 110 239.50 129.50 129.50 16769.56 117.72
12 85 226.55 141.55 141.55 20035.72 166.53
13 135 212.39 77.39 77.39 5989.65 57.33
14 145 204.65 59.65 59.65 3558.55 41.14
15 185 198.69 13.69 13.69 187.37 7.40
16 219 197.32 -21.68 21.68 470.05 9.90
17 240 199.49 -40.51 40.51 1641.27 16.88
18 420 203.54 -216.46 216.46 46855.50 51.54
19 520 225.18 -294.82 294.82 86915.99 56.70
20 410 254.67 -155.33 155.33 24128.54 37.89
21 380 270.20 -109.80 109.80 12056.10 28.89
22 320 281.18 -38.82 38.82 1507.01 12.13
23 290 285.06 -4.94 4.94 24.39 1.70
24 240 285.56 45.56 45.56 2075.31 18.98

MAD MSE MAPE

79.18 11616.03 36.41
 MAD MSE MAPE
0.1 79,18 11616.03 36.41
0.5 68,37 7563.55 30.39
0.8 62,80 5824.36 26.75

The error turns out to be a declining function of α for this data. Hence, α = 1 gives the lowest
error.

2.53 a)

Year (𝑖) 1 2 3 4 5 6 7 8 Sums Averages

Sales (𝑌𝑖 ) 6.4 8.3 8.8 5.1 9.2 7.3 12.5 57.6 8.23
Births prec. year (𝑋𝑖 ) 2.9 3.4 3.5 3.1 3.8 2.8 4.2 23.7 3.39
𝑋𝑌 18.56 28.22 30.8 15.81 34.96 20.44 52.5 201.29 28.76
𝑋2 8.41 11.56 12.25 9.61 14.44 7.84 17.64 81.75 11.68
𝑌2 40.96 68.89 77.44 26.01 84.64 53.29 156.25 507.48 72.50

𝑆𝑥𝑦 = 𝑛 ∑ 𝑋𝑖 𝑌𝑖 − ∑ 𝑋𝑖 ∑ 𝑌𝑖 = 43.91, 𝑆𝑥𝑥 = 𝑛 ∑ 𝑋𝑖2 − (∑ 𝑋𝑖 )2 = 10.56

𝑛2 (𝑛+1)(2𝑛+1) 𝑛(𝑛+1) 2
Note: cannot use 𝑆𝑥𝑥 = 𝑛 ∑ 𝑥𝑖2 − (∑ 𝑥𝑖 )2 = −( ) because
6 2
𝑋 is not 1,2,3,…
𝑆𝑥𝑦 1 1
𝑏=𝑆 ≈ 4.16, 𝑎 = 𝑛 ∑ 𝑌𝑖 − 𝑏 ∙ 𝑛 ∑ 𝑋𝑖 ≈ −5.85
𝑥𝑥

Hence 𝑌𝑡 = −5.85 + 4.16𝑋𝑡−1 is the resulting regression equation.

b) 𝑌10 = − 5.8 + 4.16 ⋅ 3.3 = 7.9214
c)

Year (𝑖) 1 2 3 4 5 6 7 8 9
Births (𝑋𝑖 ) 2.9 3.4 3.5 3.1 3.8 2.8 4.2 3.7
Forecast ES (0.15) 3.2 3.3 3.2 3.4 3.4

Forecasted births for years 9 and 10 is 3.4 million.

d) 𝑌𝑡 = −5.85 + 4.16𝑋𝑡−1 with 𝑋𝑡−1 = 3.4 in years 8 and 9.

𝑌𝑡 = −5.85 + 4.16 ⋅ 3.4 = 8.3372 sales in each of years 9 and 10. Hence, the forecast of
total aggregate sales in these years is 8.3372 ⋅ 2 = 16.6744 or $1,667,440.

2.55 The method assumes that the "best"  based on a past sequence of specific demands
will be the "best"  for future demands, which may not be true. Furthermore, the best
value of the smoothing constant based on a retrospective fit of the data may be either
larger or smaller than is desirable on the basis of stability and responsiveness of
forecasts.