CHAPTER 3

FORECASTING

Solutions to Problems

1. a&b. Plotting each data series (see below) reveals that blueberry muffin sales are stable, varying around
an average (constant). Therefore, the naive forecast for workday 16 should be the last value, 33
dozens. The demand for cinnamon buns has an increasing trend. The last change was from 31 to 33
(33-31 = 2). Using the last value and adding the last trend change, the forecast for workday 16 is 33
+ 2 = 35 dozens. Demand for cupcakes has seasonal variation with peaks every five workdays.
Workday 1 = 45, Workday 6 = 48, Workday 11 = 47 are peak workdays. There is no trend. Using
Workday 11’s sales, the forecast for Workday 16 is 47 dozens.

60

50

40
Blueberry Muffins
30 Cinnamon buns
Cupcakes
20

10

0
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15

c. For Blueberry Muffins, an averaging method such as moving average. For Cinnamon Buns, linear
trend (regression). For Cupcakes. Time series decomposition using seasonal relatives.

2. a.

3-1
Instructor’s Manual, Chapter 3
 25

20
f(x) = 0.5 x + 16.86

15
Sales

10 Linear (Sales)

0
Feb Mar Apr May Jun Jul Aug

b. (i) Instead of manually calculating a and b, we will use Excel’s Trendline function in the Layout
Menu: y = .5x + 16.857 (where x = index of month; x = 1 for Feb )
For Sept., x = 8, and y = .5(8) + 16.857 = 20.857 (000)
20+18+ 22+ 20
MA 4 = =20
(ii) 4

(iii) PROBLEM NO. 3 - 2 ES 0.1

Fore ca st Alpha =
Month Num be r Sold 0.1
Feb 19
Mar 18 19.0
A pr 15 18.9
May 20 18.5
Jun 18 18.7
Jul 22 18.6
A ug 20 18.9
S ep 19.0

(iv) 20
(v) .5 (20) + .3(22) + .2(18) = 20.2
c. Naïve, because data seems to fluctuate up and down (see the chart above).
d. That sales are reflective of demand (i.e., no stockouts or backorders occurred).

3-2 Operations Management, 5/C/e

3 a. 88 + .2(89.6 – 88) = 88.32
b. 88.32 + .2(92 – 88.32) = 89.056

4. a. 22
22+18+21+22
=20. 75
b. 4
c. F3 = 20 + .30(22 – 20) = 20.6
F4 = 20.6 + .30(18 – 20.6) = 19.82
F5 = 19.82 + .30(21 – 19.82) = 20.17
F6 = 20.17 + .30(22 – 20.17) = 20.72

3-3
Instructor’s Manual, Chapter 3
5. a. Using Excel, the equation of linear trend is Y = 18.996x + 208.48, where x is the index of week.

600

f(x) = 19 x + 208.48
500

400

Number
300
Linear (Number)

200

100

0
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18

b. Y19 = 18.996(19) + 208.48 = 569.40

Y20 = 18.996(20) + 208.48 = 588.40

700  208.48
 25.87
c. 19 (700 – 208.48) / 18.996 = 25.87 or 26th week

6. a.

12

b. There is no trend, therefore use an averaging method.

10
3-4 Operations Management, 5/C/e
7. a.

PROBLEM 3 - 7

Dozen
Day Deliveries
1 200
2 214
3 211
4 228 350
5 235
300
6 232 f(x) = 7 x + 195.466666666667
250
7 248
8 250 200
9 253 150
10 267
100 Deliveries
11 281 Linear
12 275 50 (Deliveries)

13 280 0
14 288

Y = 7x + 195.47

Forecasted demand for day 16 is: Y16 = 195.47 + (7)(16) = 307.47

PROBLEM 3 - 7

Dozen
Day Deliveries
1 200
2 214
3 211
4 228 1.2
5 235
1
6 232 f(x) = 1
7 248 0.8
8 250
0.6
9 253
10 267 0.4
11 281
0.2 Linear ()
12 275
13 280 0
14 288 1
15 310

3-5
Instructor’s Manual, Chapter 3
8. The head office demand forecast for seed #1234 in 2008 is .9(3.4) = 3.06 thousand tonnes. Plotting the
data shows declining sales. Various trend models available in Trendline of Excel were tried. The
following two are the closest:
Polynomial of order 2:
p3-8

#1234 sales 4.50

2000 0.10 4.00
2001 2.10 f(x) = − 0.1452380952 x² + 1.7071428571 x − 1.1535714286
3.50
2002 2.80 3.00
2003 3.10 2.50
2004 3.90 2.00
#1234 sales
2005 3.70 1.50 Polynomial
2006 3.50 1.00 (#1234 sales)
2007 3.40 0.50
2008 0.00
2000 2001 2002 2003 2004 2005 2006 2007 2008

With forecast for 2008 of 2.4, and linear trend only applied to period 2004-2007):
p3-8

#1234 sales 4.50

2000 0.10 4.00
2001 2.10 f(x) = − 0.1452380952 x² + 1.7071428571 x − 1.1535714286
3.50
2002 2.80 3.00
2003 3.10 2.50
2004 3.90 2.00
#1234 sales
2005 3.70 1.50 Polynomial
2006 3.50 1.00 (#1234 sales)
2007 3.40 0.50
2008 0.00
2000 2001 2002 2003 2004 2005 2006 2007 2008

With forecast for 2008 of 3.2. Therefore, the above two models provide a range of 2.5-3.2, with
average of 2.85. The head office’s forecast (3.06) is actually higher than 2.85. Therefore, it is not low
(in fact it is rather optimistic).

3-6 Operations Management, 5/C/e

PROBLEM 3 - 7

Dozen
Day Deliveries
1 200
350
2 214
3 211 300
4 228 f(x) = 7 x + 195.466666666667
5 235 250
6 232
7 248 200
8 250 Deliveries
150 Linear (Deliveries)
9 253

100

10. yt = 70 + 6t 50 t = 0 (June of last year)

t = 1 (July of last year)
0 t = 7 (January of this year)
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15
t = 8 (February of this year)
t = 9 (March of this year)
t = 19 (January of next year)
t = 20 (February of next year)
t = 21 (March of next year)
yJan = 70 + (6)(19) = 184
yFeb.= 70 + (6)(20) = 190
yMar. = 70+ (6)(21) = 196

Month Trend * Seasonal Relative = Forecast

January 184 * 1.10 = 202.4
February 190 * 1.02 =193.8
March 196 * .95 = 186.2

3-7
Instructor’s Manual, Chapter 3
P3-11
a
Week Day Sales CMA3 Sales/CMA 3
1 Fri 149
Sat 250 188.3 1.327
Sun 166 190.0 0.874
2 Fri 154 191.7 0.803
Sat 255 190.3 1.340
Sun 162 189.7 0.854
3 Fri 152 191.3 0.794
Sat 260 194.3 1.338
Sun 171 193.7 0.883
4 Fri 150 196.3 0.764
Sat 268 197.0 1.360
Sun 173 200.0 0.865
5 Fri 159 201.7 0.788
Sat 273 202.7 1.347
Sun 176 204.0 0.863
6 Fri 163 205.0 0.795
Sat 276 207.3 1.331
Sun 183

Fri Sat Sun

Week 1 1.327 0.874
Week 2 0.803 1.340 0.854
Week 3 0.794 1.338 0.883
Week 4 0.764 1.360 0.865
Week 5 0.788 1.347 0.863
Week 6 0.795 1.331 sum
avg 0.789 1.341 0.868 2.997

3-8 Operations Management, 5/C/e

P3-11
a
Week Day Sales CMA3 Sales/CMA 3
1 Fri 149
Sat 250 188.3 1.327
Sun 166 190.0 0.874
2 Fri 154 191.7 0.803
Sat 255 190.3 1.340
Sun 162 189.7 0.854
3 Fri 152 191.3 0.794
Sat 260 194.3 1.338
Sun 171 193.7 0.883
4 Fri 150 196.3 0.764
Sat 268 197.0 1.360
Sun 173 200.0 0.865
5 Fri 159 201.7 0.788
Sat 273 202.7 1.347
Sun 176 204.0 0.863
6 Fri 163 205.0 0.795
Sat 276 207.3 1.331
Sun 183

Fri Sat Sun

Week 1 1.327 0.874
Week 2 0.803 1.340 0.854
Week 3 0.794 1.338 0.883
Week 4 0.764 1.360 0.865
Week 5 0.788 1.347 0.863
Week 6 0.795 1.331 sum
avg 0.789 1.341 0.868 2.997
adjusted 0.790 1.342 0.868 3.000

b
Week Day Sales SR Sales/SR
1 Fri 149 0.790 188.662
Sat 250 1.342 186.320
Sun 166 0.868 191.145
2 Fri 154 0.790 194.993
Sat 255 1.342 190.047

3-9
Instructor’s Manual, Chapter 3
12. Wednesday = .15 x 4 = 0.60
Thursday = .20 x 4 = 0.80
Friday = .35 x 4 = 1.40
Saturday = .30 x 4 = 1.20

13. a. There is a long-term upward trend in the data (see the following chart). The averaging technique
forecasts will consistently underestimate the data.

b. Linear trend seems appropriate:

PROBLEM 3 - 13

W e ek Tons Shipped Forecast

1 405
2 410 y = 4.5417x + 397.3
3 420 480
4 415 470
5 412 460
6 420 450
7 424 440
8 433 430
9 438 420
10 440 410
11 446 400
12 451 1 3 5 7 9 11 13 15 17
13 455
14 464
15 466
16 474
17 476
18 479.05

3-10 Operations Management, 5/C/e

PROBLEM 3 - 14
a
Numbe r No. se rve d
Day Serve d CMA7 /CMA7
1 80
2 75
3 78
4 95 90.57 1.049
5 130 90.86 1.431
6 136 91.14 1.492
7 40 91.43 0.438
8 82 91.29 0.898
9 77 90.57 0.850
10 80 90.43 0.885
11 94 90.71 1.036
12 125 91.00 1.374
13 135 91.00 1.484
14 42 91.43 0.459
15 84 91.71 0.916
16 77 93.14 0.827
17 83 93.86 0.884
18 96 93.14 1.031
19 135 93.57 1.443
20 140 94.29 1.485
21 37 96.43 0.384
22 87 97.43 0.893
23 82 98.71 0.831
24 98 99.29 0.987
25 103 100.86 1.021
26 144
27 144
28 48

Day
1 2 3 4 5 6 7
1.049 1.431 1.492 0.438
0.898 0.850 0.885 1.036 1.374 1.484 0.459
0.916 0.827 0.884 1.031 1.443 1.485 0.384
0.893 0.831 0.987 1.021
Total
avg: 0.902 0.836 0.919 1.034 1.416 1.487 0.427 7.021
Adjusted: 0.900 0.833 0.916 1.031 1.412 1.482 0.426 7.000

3-11
Instructor’s Manual, Chapter 3
PROBLEM 3 - 14 b
Number No. served
Day Served SR / SR
1 80 0.900 88.916
2 75 0.833 89.994
3 78 0.916 85.154
4 95 1.031 92.123
5 130 1.412 92.096
6 136 1.482 91.738
7 40 0.426 93.983
8 82 0.900 91.139
9 77 0.833 92.394
10 80 0.916 87.337
11 94 1.031 91.154
12 125 1.412 88.553
13 135 1.482 91.064
14 42 0.426 98.682
15 84 0.900 93.361
16 77 0.833 92.394
17 83 0.916 90.613
18 96 1.031 93.093
19 135 1.412 95.638
20 140 1.482 94.436
21 37 0.426 86.935
22 87 0.900 96.696
23 82 0.833 98.393
24 98 0.916 106.988
25 103 1.031 99.881
26 144 1.412 102.013
27 144 1.482 97.134
28 48 0.426 112.780

No. served / SR
120.000

100.000
f(x) = 0.513240796766983 x + 86.653795611075
80.000

60.000
No. served / SR
40.000
Linear (No. served / SR)
20.000

0.000
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28

Day Trend frcst Frcst

29 101.537 91
30 102.050 85
31 102.563 94
32 103.076 106
15.
a.
3-12 Operations Management, 5/C/e
 a. Computing seasonal relatives
Calipe r
Quarter Sales CM A4 Sales/CM A4
1990 1 2370
2 2058
3 2778 2644.875 1.0503
4 3238 2696.250 1.2009
1991 1 2641 2681.250 0.9850
2 2198 2614.250 0.8408
3 2518 2534.750 0.9934
4 2962 2441.750 1.2131
1992 1 2281 2344.875 0.9728
2 1814 2207.625 0.8217
3 2127
4 2255

QUARTER
Year 1 2 3 4
1990 1.0503 1.2009
1991 0.9850 0.8408 0.9934 1.2131
1992 0.9728 0.8217
Total
Average: 0.9789 0.8312 1.0219 1.2070 4.0390
Adjusted: 0.9694 0.8232 1.0120 1.1954 4.0000

b. Forecast of Q1 of 1993 = (6334 / 4)(.9694) = 1,535.0

Forecast of Q2 of 1993 = (6334 / 4)(.8232) = 1,303.5
Forecast of Q3 of 1993 = (6334 / 4)(1.0120) = 1,602.5
Forecast of Q4 of 1993 = (6334 / 4)(1.1954) = 1,892.9
c. The method in Part b is much simpler. It requires a forecast for the whole year which is easier
to calculate using a trend model. On the other hand, it loses some details of how the demand was
behaving within the years which may affect the trend.

16. a. Yes. This would cause concern because you wouldn’t know the actual demand for the pain
reliever.

3-13
Instructor’s Manual, Chapter 3
 PROBLEM 3 - 16

Day no. sold

1 36 α β
2 38 0.2 0.3
3 42 S T TAF = S + T
4 44 40.00 2.67 42.67
5 48 43.73 2.99 46.72
6 49 47.18 3.12 50.30
7 50 50.24 3.11 53.35
8 49 52.48 2.84 55.32
9 52 54.66 2.65 57.30
10 48 55.44 2.09 57.53
11 52 56.42 1.76 58.18
12 55 57.54 1.56 59.11
13 54 58.09 1.26 59.34
Linear trend is not appropriated for the first half of the data (it seems to be non-linear), but it
seems appropriate for the second half.

PROBLEM 3 - 16 c

Day no. sold

1 36 α β
2 38 0.3 0.3
3 42
4 44
5 48
6 49
7 50 S T TAF = S + T
8 49 44.50 1.86 46.36
9 52 48.05 2.37 50.42
10 48 49.69 2.15 51.84
11 52 51.89 2.16 54.05
12 55 54.33 2.25 56.58
13 54 55.81 2.02 57.82
14 56 57.28 1.85 59.13

3-14 Operations Management, 5/C/e

17. a. Yes, there is a slightly increasing trend.

900
800
700
600
500
400
300
PROBLEM 3 - 17 b
200
Units
Month
100
Sold Index De-seasonalized
Jan 640 00.80 800.0
Feb 648 0.80 810.0
Mar 630 Jan
0.70 900.0Feb Mar Apr Ma
Apr 761 0.94 809.6
May 735 0.89 825.8
Jun 850 1.00 850.0
Jul 765 0.90 850.0
Aug 805 1.15 700.0
Sep 840 1.20 700.0
Oct 828 1.20 690.0
Nov 840 1.21 694.2

c.
1000

900

800

700

As it can be observed in the following chart, the “trend” observed in data was due to seasonality,
as it does not appear in the de-seasonalized data.
600

3-15
Instructor’s Manual, Chapter
500 3
 PROBLEM 3 - 18
Sales Sales
Quarter ($000) SR / SR
1 88 1.100 80.000
2 99 0.990 100.000
3 108 0.900 120.000
4 141 1.010 139.604

160.000
140.000
f(x) = 19.8811881188119 x + 60.1980198019802
120.000
100.000
80.000
60.000 Sales / SR
40.000 Linear (Sales / SR)
20.000
0.000
1 2 3 4

5 19.881(5)+60.198= 159.6 159.6(1.1)= 175.6

3-16 Operations Management, 5/C/e

PROBLEM 3 - 19 a

a. Computing se as onal re lative s

Shipm ent
Q uarter Shipm ents CMA / CMA
Year 1 Q1 200
2 250
3 210 251.3 0.8358
4 340 252.8 1.3452
Year 2 Q1 210 253.3 0.8292
2 252 256.0 0.9844
3 212 259.1 0.8181
4 360 260.8 1.3806
Year 3 Q1 215 262.8 0.8183
2 260 263.5 0.9867
3 220 264.5 0.8318
4 358 267.3 1.3396
Year 4 Q1 225 270.4 0.8322
2 272 273.8 0.9936
3 233 276.4 0.8431
4 372 278.8 1.3345
Year 5 Q1 232 281.1 0.8253
2 284 283.1 1.0031
3 240
4 381

QUARTER
Ye ar 1 2 3 4
1 0.8358 1.3452
2 0.8292 0.9844 0.8181 1.3806
3 0.8183 0.9867 0.8318 1.3396
4 0.8322 0.9936 0.8431 1.3345
5 0.8253 1.0031

Ave rage : 0.8262 0.9919 0.8322 1.3500

Adjus te d: 0.8262 0.9919 0.8321 1.3499

3-17
Instructor’s Manual, Chapter 3
b. Seasonal Deseasonalized
Quarter Shipments Relatives Sales
Year 1 Q1 200 0.826 242.08
2 250 0.992 252.05
3 210 0.832 252.37
4 340 1.350 251.88
Year 2 Q1 210 0.826 254.19
2 252 0.992 254.07
3 212 0.832 254.77
4 360 1.350 266.69
Year 3 Q1 215 0.826 260.24
2 260 0.992 262.13
3 220 0.832 264.39
4 358 1.350 265.21
Year 4 Q1 225 0.826 272.35
2 272 0.992 274.23
3 233 0.832 280.01
4 372 1.350 275.58
Year 5 Q1 232 0.826 280.82
2 284 0.992 286.33
3 240 0.832 288.42
4 381 1.350 282.25

Fit a linear trend to the deseasonalized data (pink):

y = 2.1936x + 242.97
450.00

400.00
De se a sona liz e d
S a le s
350.00

Line a r
300.00 (De se a sona liz e d
S a le s)

250.00

200.00

150.00
Ye a r 2 3 4 Ye a r 2 3 4 Ye a r 2 3 4 Ye a r 2 3 4 Ye a r 2 3 4
1 Q1 2 Q1 3 Q1 4 Q1 5 Q1

Extend the line & reseasonalize:

Trend Reseasonalized
Quarter Forecast Forecast
Year 6 Q1 289.0 238.8
2 291.2 288.9
3 293.4 244.2
4 295.6 399.0

3-18 Operations Management, 5/C/e

PROBLEM 3-20

FCL - C om pu t in g Season al Relat iv es u sin g C en tered M o v in g A v erage

Demand/
Demand CMA 1 2 CMA 1 2
Jan-03 0
Feb 52
Mar 29
Apr 49
May 47
Jun 58
Jul 0 27.6 0.00
Aug 64 25.8 2.48
Sep 3 24.6 0.12
Oct 17 26.3 0.65
Nov 10 28.5 0.35
Dec 1 30.9 0.03
Jan-04 2 38.1 0.05
Feb 8 42.0 0.19
Mar 44 40.6 1.08
Apr 74 41.2 1.80
May 75 40.6 1.85
Jun 87 40.6 2.14
Jul 145 40.9 3.55
Aug 11 41.4 0.27
Sep 24 40.6 0.59
Oct 9 37.5 0.24
Nov 5 35.1 0.14
Dec 6 32.0 0.19
Jan-05 3 28.0 0.11
Feb 20 28.3 0.71
Mar 12 30.1 0.40
Apr 31 30.0 1.03
May 61 29.9 2.04
Jun 28 29.5 0.95
Jul 107
Aug 57
Sep 21
Oct 10
Nov 0
Dec 1
Month
Year J F M A M J J A S O N D
2003 0.00 2.48 0.12 0.65 0.35 0.03
2004 0.05 0.19 1.08 1.80 1.85 2.14 3.55 0.27 0.59 0.24 0.14 0.19
2005 0.11 0.71 0.40 1.03 2.04 0.95
To tal
Average: 0.08 0.45 0.74 1.41 1.94 1.55 1.77 1.37 0.36 0.44 0.25 0.11 10.48
Adjusted: 0.09 0.51 0.85 1.62 2.23 1.77 2.03 1.57 0.41 0.51 0.28 0.13 12.00

3-19
Instructor’s Manual, Chapter 3
FC L - Usin g th e Seaso n al Relativ es to De-seaso n alize dem an d

Seasonal Deseasonalized
Dem and relativ es Dem and
Jan-03 0 0.09 0.0
Feb 52 0.51 101.3
Mar 29 0.85 34.2
Apr 49 1.62 30.2
May 47 2.23 21.1
Jun 58 1.77 32.7
Jul 0 2.03 0.0
Aug 64 1.57 40.7
Sep 3 0.41 7.3
Oct 17 0.51 33.5
Nov 10 0.28 35.4
Dec 1 0.13 7.9
Jan-04 2 0.09 21.9
Feb 8 0.51 15.6
Mar 44 0.85 51.9
Apr 74 1.62 45.7
May 75 2.23 33.7
Jun 87 1.77 49.1
Jul 145 2.03 71.4
Aug 11 1.57 7.0
Sep 24 0.41 58.8
Oct 9 0.51 17.7
Nov 5 0.28 17.7
Dec 6 0.13 47.7
Jan-05 3 0.09 32.8
Feb 20 0.51 38.9
Mar 12 0.85 14.1
Apr 31 1.62 19.1
May 61 2.23 27.4
Jun 28 1.77 15.8
Jul 107 2.03 52.7
Aug 57 1.57 36.3
Sep 21 0.41 51.4
Oct 10 0.51 19.7
Nov 0 0.28 0.0
Dec 1 0.13 7.9

3-20 Operations Management, 5/C/e

FC L - fittin g a m odel to de-seaso n alized dem an d,
ex ten din g it 12 m o n th s, an d re-season alizin g it

120.0 Deseasonalized
Deseasonalized
y = -0.2013x + 34.241
100.0
Linear (Deseasonalized)
80.0
60.0
40.0
20.0
0.0

Trend Re-seasonalized
Forecast Forecast
Jan-06 26.79 2.45
Feb 26.59 13.66
Mar 26.39 22.39
Apr 26.19 42.44
May 25.99 57.87
Jun 25.79 45.67
Jul 25.59 51.99
Aug 25.38 39.88
Sep 25.18 10.29
Oct 24.98 12.69
Nov 24.78 7.00
Dec 24.58 3.09

3-21
Instructor’s Manual, Chapter 3
Problem 3-21 a
SaskEnergy Natural Gas Sales Forecasting -- Computing Seasonal Relatives
www.saskenergy.com/about_saskenergy

Sales Sales /
Quarter (peta joules) CMA 2 CMA2
2005 1 49
2 24
3 18 31.13 0.578
4 37 29.75 1.244
2006 1 42 29.50 1.424
2 20 30.50 0.656
3 20 32.00 0.625
4 43 33.25 1.293
2007 1 48 33.75 1.422
2 24 33.38 0.719
3 20 33.38 0.599
4 40 33.88 1.181
2008 1 51 33.88 1.506
2 25 34.13 0.733
3 19 34.50 0.551
4 43 34.38 1.251
2009 1 51 33.75 1.511
2 24
3 15

QUARTER
Year 1 2 3 4
2005 0.578 1.244
2006 1.424 0.656 0.625 1.293
2007 1.422 0.719 0.599 1.181
2008 1.506 0.733 0.551 1.251
2009 1.511

Total
Average: 1.466 0.702 0.588 1.242 3.999
Adjusted: 1.466 0.703 0.589 1.243 4.000

3-22 Operations Management, 5/C/e

Problem 3-21
SaskEnergy Natural Gas Sales Forecasting -- Computing Seasonal Relatives
www.saskenergy.com/about_saskenergy

Sales Sales /
Quarter (peta joules) CMA4 CMA 4
2005 1 49
2 24
3 18 31.13 0.578
4 37 29.75 1.244
2006 1 42 29.50 1.424
2 20 30.50 0.656
3 20 32.00 0.625
4 43 33.25 1.293
2007 1 48 33.75 1.422
2 24 33.38 0.719
3 20 33.38 0.599
4 40 33.88 1.181
2008 1 51 33.88 1.506
2 25 34.13 0.733
3 19 34.50 0.551
4 43 34.38 1.251
2009 1 51 33.75 1.511
2 24
3 15

QUARTER
Year 1 2 3 4
2005 0.578 1.244
2006 1.424 0.656 0.625 1.293
2007 1.422 0.719 0.599 1.181
2008 1.506 0.733 0.551 1.251
2009 1.511

Total
Average: 1.466 0.702 0.588 1.242 3.999
Adjusted: 1.466 0.703 0.589 1.243 4.000

3-23
Instructor’s Manual, Chapter 3
Problem 3-22 a
Ridgeline Sales Forecasting -- Computing Seasonal Relatives
CANSIM II Series V29850230

Sales Sales / 25.0

Quarter (units) CMA 2 CMA2
2005 1 11.5 20.0
2 19.2
3 17.4 15.250 1.141 15.0

4 12.7 14.788 0.859

10.0
2006 1 11.9 14.275 0.834
2 15.1 14.200 1.063
5.0
3 17.4 13.963 1.246
4 12.1 14.300 0.846 0.0
2007 1 10.6 14.363 0.738

2
3
4

2
3
4

2
3
4

2
3
1

1
2 19.1 13.850 1.379
2005

2006

2007

2008
3 13.9 13.725 1.013
4 11.5 12.063 0.953
2008 1 10.2 9.413 1.084
2 6.2
3 5.6

QUARTER
Year 1 2 3 4
2005 1.141 0.859
2006 0.834 1.063 1.246 0.846
2007 0.738 1.379 1.013 0.953
2008 1.084

Total
Average: 0.885 1.221 1.133 0.886 4.126

3-24 Operations Management, 5/C/e

Problem 3-22 a
Ridgeline Sales Forecasting -- Computing Seasonal Relatives
CANSIM II Series V29850230

Sales Sales / 25.0

Quarter (units) CMA4 CMA 4
2005 1 11.5 20.0
2 19.2
3 17.4 15.250 1.141 15.0

4 12.7 14.788 0.859

10.0
2006 1 11.9 14.275 0.834
2 15.1 14.200 1.063
5.0
3 17.4 13.963 1.246
4 12.1 14.300 0.846 0.0
2007 1 10.6 14.363 0.738

2
3
4

2
3
4

2
3
4

2
3
1

1
2 19.1 13.850 1.379

2007

2008
2005

2006
3 13.9 13.725 1.013
4 11.5 12.063 0.953
2008 1 10.2 9.413 1.084
2 6.2
3 5.6

QUARTER
Year 1 2 3 4
2005 1.141 0.859
2006 0.834 1.063 1.246 0.846
2007 0.738 1.379 1.013 0.953
2008 1.084

Total
Average: 0.885 1.221 1.133 0.886 4.126
Adjusted: 0.858 1.184 1.099 0.859 4.000

Problem 3-22 b
Ridgeline Sales Forecasting -- Computing Forecasts

Sales Seasonal Deseasonalized

Quarter (units) Relatives Sales
2005 1 11.5 0.858 13.40

3-25
Instructor’s Manual, Chapter 3
Problem 3-23 a
MA & PA
Boned-Head Removed Trout
Computing seasonal relatives

Sales
Quarter (pounds) CMA Sales/CMA
1997 1 664
2 1338
3 1170 1030.0 1.14
4 1069 969.8 1.10
1998 1 422 1035.9 0.41
2 1098 1103.8 0.99
3 1939 1123.1 1.73
4 843 1212.3 0.70
1999 1 803 1162.1 0.69
2 1430 1055.6 1.35
3 1206 1027.6 1.17
4 724 970.3 0.75
2000 1 698 918.9 0.76
2 1076
3 1149

QUARTER
Year 1 2 3 4
1997 1.14 1.10
1998 0.41 0.99 1.73 0.70
1999 0.69 1.35 1.17 0.75
2000 0.76

Average: 0.62 1.17 1.35 0.85

3-26 Operations Management, 5/C/e

Problem 3-23 a
MA & PA
Boned-Head Removed Trout
Computing seasonal relatives

Sales
Quarter (pounds) CMA Sales/CMA
1997 1 664
2 1338
3 1170 1030.0 1.14
4 1069 969.8 1.10
1998 1 422 1035.9 0.41
2 1098 1103.8 0.99
3 1939 1123.1 1.73
4 843 1212.3 0.70
1999 1 803 1162.1 0.69
2 1430 1055.6 1.35
3 1206 1027.6 1.17
4 724 970.3 0.75
2000 1 698 918.9 0.76
2 1076
3 1149

QUARTER
Year 1 2 3 4
1997 1.14 1.10
1998 0.41 0.99 1.73 0.70
1999 0.69 1.35 1.17 0.75
2000 0.76
Total
Average: 0.62 1.17 1.35 0.85 3.99
Adjusted: 0.62 1.18 1.35 0.85 4.00

Problem 3-23 b
Computing forecasts

Sales Seasonal Deseasonalized

3-27
Instructor’s Manual, Chapter 3
PROBLEM 3 - 24 a

Average Number
Sold Per Day, y Price, x
200 $ 6.00
190 6.50
188 6.75
180 7.00
170 7.25
162 7.50
160 8.00
155 8.25
156 8.50
148 8.75
140 9.00
133 9.25

slope = -19.5126
Intercept = 315.9829

220
No. sold, y

200
f(x) = − 19.5126122274476 x + 315.982898674647
180

160

140

120

100
5 5.5 6 6.5 7 7.5 8 8.5 9 9.5 10
Price, x

Note: Trendline's equation shows rounded values for the intercept and slope.

PROBLEM 3 - 24 b

r = -.985 implies a high, negative relationship between price and number sold.
3-28 Operations Management, 5/C/e
PROBLEM NO. 3 - 25

Observation x y
1 15 74
2 25 80
3 40 84
4 32 81
5 51 96
6 47 95
7 30 83
8 18 78
9 14 70
10 15 72
11 22 85
12 24 88

a
120

100

80

60
y
40

20

0
10 15 20 25 30 35 40 45 50 55
x

b
slope = 0.570571
Intercept = 66.33333 y = 66.33 + .57 x

Approximately 76.5% of the variation in the dependent variable is explained by the independent
variable.

d y = 66.33 + .57 (41) = 89.70

3-29
Instructor’s Manual, Chapter 3
PROBLEM 3 - 26

x y
Fertilizer No. of Mowers
Period Sales (tons) Sold
1 1.6 10
2 1.3 8
3 1.8 11
4 2.0 12
5 2.2 12
6 1.6 9
7 1.5 8
8 1.3 7
9 1.7 10
10 1.2 6
11 1.9 11
12 1.4 8
13 1.7 10

a
Correlation coeff. = 0.960
R^2 = 0.921
Since the value of R^2 is close to 1, there is a strong relationship between these variables.
Yes, good predictions.

b
Slope = 6.1527
c Y = 6.1527(2) -.649 = 11.66. Therefore, we expect 12 mowers to be sold.

3-30 Operations Management, 5/C/e

Problem 3-27

Day Avg. Temp (F) Demand (thousand dekatherm)

-10 33 1005
-9 46 501
-8 41 612
-7 47 499
-6 37 692
-5 36 709
-4 32 951
-3 34 1053
-2 38 746
-1 43 458

a
1200

1000

800

600
Demand

400

200

0
30 32 34 36 38 40 42 44 46 48
Temp

Yes, there seems to be a negative strong linear reltionship.

Note: R^2 = 0.83

b
Intercept = 2156.09
Slope= -37.04 y = 2156.09 -37.04 x

3-31
Instructor’s Manual, Chapter 3
28.

Units Linear
T sold Naive e |e| e2 | e |/Actual*100 Trend e | e | e2 | e |/Actual*100
11 147  146 1 1 1 .68
12 148 147 1 1 1 .68 148 0 0 0 0
13 151 148 3 3 9 1.99 150 1 1 1 .66
14 145 151 –6 6 36 4.14 152 –7 7 49 4.83
15 155 145 10 10 100 6.45 154 1 1 1 .65
16 152 155 –3 3 9 1.97 156 –4 4 16 2.63
17 155 152 3 3 9 1.94 158 –3 3 9 1.94
18 157 155 2 2 4 1.27 160 –3 3 9 1.91
19 160 157 3 3 9 1.88 162 –2 2 4 1.25
13 31 177 20.32 –16 22 90 14.55%

MAD = 31/8 = 3.875 MAD = 22/9 = 2.44

MSE = 177/8 = 22.125

MSE = 90/9 = 10

MAPE = 20.32/8 = 2.54% MAPE = 14.55/9 = 1.62%

Linear trend provides more accurate forecasts because it has smaller values for any measure of forecast error:
MAD, MSE, and MAPE.

3-32 Operations Management, 5/C/e

Problem 3-29

Period Demand F1 e ½e½ e2 | e |/Actual*100 F2 e ½e½ e2 | e |/Actual*100

1 68 66 2 2 4 2.94% 66 2 2 4 2.94%
2 75 68 7 7 49 9.33% 68 7 7 49 9.33%
3 70 72 -2 2 4 2.86% 70 0 0 0 0.00%
4 74 71 3 3 9 4.05% 72 2 2 4 2.70%
5 69 72 -3 3 9 4.35% 74 -5 5 25 7.25%
6 72 70 2 2 4 2.78% 76 -4 4 16 5.56%
7 80 71 9 9 81 11.25% 78 2 2 4 2.50%
28 160 37.56% 22 102 30.28%
a, b, c 4 22.86 5.37% 3.14 14.57 4.33%
MAD MSE MAPE MAD MSE MAPE
F2 has a lower value of MAD, MSE, and MAPE.

d Yes, all three provide the same conclusion, but this may not be the case. E.g, if obs 7 for F1 was 76:

Period Demand F1 e ½e½ e2 | e |/Actual*100 F2 e ½e½ e2 | e |/Actual*100

1 68 66 2 2 4 2.94% 66 2 2 4 2.94%
2 75 68 7 7 49 9.33% 68 7 7 49 9.33%
3 70 72 -2 2 4 2.86% 70 0 0 0 0.00%
4 74 71 3 3 9 4.05% 72 2 2 4 2.70%
5 69 72 -3 3 9 4.35% 74 -5 5 25 7.25%
6 72 70 2 2 4 2.78% 76 -4 4 16 5.56%
7 80 76 4 4 16 5.00% 78 2 2 4 2.50%
23 95 31.31% 22 102 30.28%
3.286 13.57 4.47% 3.14 14.57 4.33%
MAD MSE MAPE MAD MSE MAPE

F1 will have a lower MSE than F2.

3-33
Instructor’s Manual, Chapter 3
Problem 3-30

Period Demand F1 e ½e½ e2 | e |/Actual*100 F2 e ½e½ e2 | e |/Actual*100

1 770 771 -1 1 1 0.13% 769 1 1 1 0.13%
2 789 785 4 4 16 0.51% 787 2 2 4 0.25%
3 794 790 4 4 16 0.50% 792 2 2 4 0.25%
4 780 784 -4 4 16 0.51% 798 -18 18 324 2.31%
5 768 770 -2 2 4 0.26% 774 -6 6 36 0.78%
6 772 768 4 4 16 0.52% 770 2 2 4 0.26%
7 760 761 -1 1 1 0.13% 759 1 1 1 0.13%
8 775 771 4 4 16 0.52% 775 0 0 0 0.00%
9 786 784 2 2 4 0.25% 788 -2 2 4 0.25%
10 790 788 2 2 4 0.25% 788 2 2 4 0.25%
28 94 3.59% 36 382 4.62%
a 2.8 9.4 0.36% 3.6 38.2 0.46%
MAD MSE MAPE MAD MSE MAPE
F1 has a lower value of MAD, MSE, and MAPE.

b Yes, all three provide the same conclusion, but this may not be the case. E.g, if obs 4 for F2 was 789:

Period Demand F1 e ½e½ e2 | e |/Actual*100 F2 e ½e½ e2 | e |/Actual*100

1 770 771 -1 1 1 0.13% 769 1 1 1 0.13%
2 789 785 4 4 16 0.51% 787 2 2 4 0.25%
3 794 790 4 4 16 0.50% 792 2 2 4 0.25%
4 780 784 -4 4 16 0.51% 789 -9 9 81 1.15%
5 768 770 -2 2 4 0.26% 774 -6 6 36 0.78%
6 772 768 4 4 16 0.52% 770 2 2 4 0.26%
7 760 761 -1 1 1 0.13% 759 1 1 1 0.13%
8 775 771 4 4 16 0.52% 775 0 0 0 0.00%
9 786 784 2 2 4 0.25% 788 -2 2 4 0.25%
10 790 788 2 2 4 0.25% 788 2 2 4 0.25%
28 94 3.59% 27 139 3.47%
2.8 9.4 0.36% 2.7 13.9 0.35%
MAD MSE MAPE MAD MSE MAPE

c 2s control limits are 0  2 √ MSE

F1 02 9.4  0  6.13 Bec. all errors are within these limits, F1 is in control.
F2 02 38.2  0  12.36 Bec. the error for month 4 is outside these limits, F2 is not in
control.

3-34 Operations Management, 5/C/e

31. a.
Month Error Error2
1 -8 64
2 -2 4
3 4 16
4 7 49
5 9 81
6 5 25
7 0 0
8 -3 9
9 -4 16
264

MSE = 264 / 9 = 29.33

2s control limits are 0  2 √ MSE = 0  2 29.33 =  10.83
All errors are within these limits.

b.
12

0 Err
10 11 12 13 14 15 16 17 18 or
-4

-8

-12

Although all the last 9 forecast errors are within the control limits, the errors are not random around
0, showing a possible seasonality. This suggests that possible seasonality in demand is being
overlooked in forecasting.

3-35
Instructor’s Manual, Chapter 3
32. a.
PROBLEM 3 - 32
y = 3.8321x + 36.418
75.0
Year Sa les (000) Forecast error error^2 70.0
1 40.2 40.3 -0.1 0.00 65.0
2 44.5 44.1 0.4 0.17 60.0
3 48.0 47.9 0.1 0.01 55.0
50.0
4 52.3 51.7 0.6 0.31
45.0
5 55.8 55.6 0.2 0.05 40.0 Sales (000)
6 57.1 59.4 -2.3 5.34 35.0 Linear (Sales (000))

7 62.4 63.2 -0.8 0.71 30.0

8 69.0 67.1 1.9 3.71 1 2 3 4 5 6 7 8
1.29

2s control limits are 0  2 √ MSE = 02 √ 1.29 =  2.27

b.

Forecast error 6.0

9 73.7 70.9 2.8 4.0
10 77.2 74.7 2.5 2.0
0.0
11 82.1 78.6 3.5 1 2 3 4 5
-2.0
12 87.8 82.4 5.4
-4.0
13 90.6 86.2 4.4 -6.0
error

c. No the forecasting process is not performing adequately, because the forecast errors for years 9
to 13 fall outline the control limits.

3-36 Operations Management, 5/C/e

33. a.
PROBLEM 3- 33

Period Demand Method 1 e1 e1^2 Method 2 e2 e2^2

1 37 36 1 1 36 1 1
2 39 38 1 1 37 2 4
3 37 40 -3 9 38 -1 1
4 39 42 -3 9 38 1 1
5 45 46 -1 1 41 4 16
6 49 46 3 9 52 -3 9
7 47 46 1 1 47 0 0
8 49 48 1 1 48 1 1
9 51 52 -1 1 52 -1 1

Both forecasting methods have approximately the same MSE.

b. The 2s control limits for Method 1 are 0  2 √ MSE = 02 √ 3.7 =  3.8. The 2s control limits
for Method 2 are 0  2 √ 3.8 =  3.9. The errors for Method 1 are all within its control limits.
However, the error plot (see below) seems to be non-random (wavy). The period 5 error for Method
2 is just outside its control limits, but the error plot seems random. Based on the control charts, both
methods have problems.

control limits: 3.8 3.9

4 4
2 2

0 e1 0 e2

1 2 3 4 5 6 7 8 9 -2 1 2 3 4 5 6 7 8 9
-2
-4 -4

3-37
Instructor’s Manual, Chapter 3
 PROBLEM 3-34

No. used MA3 F MA5 F ES0.1 F ES0.3 F Trend F

Jan. 10 25.27
Feb. 10 10.00 10.00 25.55
Mar. 66 10.00 10.00 25.82
Apr. 32 28.67 15.60 26.80 26.09
May 34 36.00 17.24 28.36 26.36
June 18 44.00 30.40 18.92 30.05 26.64
July 24 28.00 32.00 18.82 26.44 26.91
Aug. 9 25.33 34.80 19.34 25.71 27.18
Sep. 14 17.00 23.40 18.31 20.69 27.45
Oct. 48 15.67 19.80 17.88 18.69 27.73
Nov. 23.67 22.60 20.89 27.48 28.00

70
No. used
60
MA3 F
50
MA5 F
40

30

20

10

Jan. Feb. Mar. Apr. May June July Aug. Sep. O ct. Nov.

70
ES0.1 F

60
ES0.3 F

50
No. used

40

30

20

10

Jan. Feb. Mar. Apr. May June July Aug. Sep. O ct. Nov.

y = 0.2727x + 25
70

60

50

40

30

20

10

1 2 3 4 5 6 7 8 9 10 11

Best fit to data is MA3 or ES0.3.

3-38 Operations Management, 5/C/e

 PROBLEM 3-34 g

Jan.
Feb.
3-39
Instructor’s Manual, Chapter 3
PROBLEM 3-35

Ice Cream avering & trend

Forecast Forecast
Year Sales MA2 MA4
1995 341 Sales MA2 MA4
1996 331 350
1997 317 336.0
330
1998 315 324.0
1999 321 316.0 326.0 310
2000 278 318.0 321.0 290
2001 298 299.5 307.8
270
2002 311 288.0 303.0
2003 302 304.5 302.0 250
1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008
2004 302 306.5 297.3
2005 335 302.0 303.3
2006 320 318.5 312.5
2007 285 327.5 314.8
2008 302.5 310.5

Exponential Smoothing
ES 0.2 ES 0.4
Sales ES 0.2 ES 0.4
Year Sales Forecast Forecast
1995 341 350
340
1996 331 341.00 341.00 330
1997 317 339.00 337.00 320
1998 315 334.60 329.00 310
300
1999 321 330.68 323.40 290
2000 278 328.74 322.44 280
270
2001 298 318.60 304.66 260
2002 311 314.48 302.00 250
1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008
2003 302 313.78 305.60
2004 302 311.42 304.16
2005 335 309.54 303.30
2006 320 314.63 315.98
2007 285 315.71 317.59
2008 309.56 304.55

Trend
Trend
Year Sales Forecast
1995 341 330.09 1
1996 331 325.98 2 350
1997 317 321.87 3 340
330
1998 315 317.76 4 320
310 f(x) = − 1.9945054945055 x
1999 321 313.65 5 300 + 325.96153846154
2000 278 309.55 6 290
280
2001 298 305.44 7 270
2002 311 301.33 8 260
250
2003 302 297.22 9 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008

2004 302 293.11 10

2005 335 289.00 11
2006 320 284.89 12
2007 285 280.78 13
2008 334.20

3-40 Operations Management, 5/C/e

PROBLEM 3-35

Ice Cream avering & trend

Forecast Forecast
Year Sales MA2 MA4
1995 341 Sales MA2 MA4
1996 331 350
1997 317 336.0
330
1998 315 324.0
1999 321 316.0 326.0 310
2000 278 318.0 321.0 290
2001 298 299.5 307.8
270
2002 311 288.0 303.0
2003 302 304.5 302.0 250
1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008
2004 302 306.5 297.3
2005 335 302.0 303.3
2006 320 318.5 312.5
2007 309 327.5 314.8
2008 314.5 316.5

Exponential Smoothing
ES 0.2 ES 0.4
Sales ES 0.2 ES 0.4
Year Sales Forecast Forecast
1995 341 350
340
1996 331 341.00 341.00 330
1997 317 339.00 337.00 320
1998 315 334.60 329.00 310
300
1999 321 330.68 323.40 290
2000 278 328.74 322.44 280
270
2001 298 318.60 304.66 260
2002 311 314.48 302.00 250
1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008
2003 302 313.78 305.60
2004 302 311.42 304.16
2005 335 309.54 303.30
2006 320 314.63 315.98
2007 309 315.71 317.59
2008 314.36 314.15

Trend
Trend
Year Sales Forecast
1995 341 330.09 1
1996 331 325.98 2 350
1997 317 321.87 3 340
330
1998 315 317.76 4 320
1999 321 313.65 5 310 f(x) = − 1.2032967032967 x
300 + 322.26923076923
2000 278 309.55 6 290
280
2001 298 305.44 7 270
2002 311 301.33 8 260
250
2003 302 297.22 9 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008

2004 302 293.11 10

2005 335 289.00 11
2006 320 284.89 12

3-41
Instructor’s Manual, Chapter 3
3-42 Operations Management, 5/C/e
Problem 3-36

v42190663; Toyota - Corolla

Alpha Linear
No. 0.97 No. Trend
Month produced ES Forecast |Fcst Error| produced Forecast |Fcst Error|
2008-01 4959 1 4959 8778.82 3819.8
2008-02 8463 4959.0 3504.0 2 8463 8967.74 504.7
2008-03 11706 8357.9 3348.1 3 11706 9156.66 2549.3
2008-04 9907 11605.6 1698.6 4 9907 9345.58 561.4
2008-05 9354 9958.0 604.0 5 9354 9534.5 180.5
2008-06 11297 9372.1 1924.9 6 11297 9723.42 1573.6
2008-07 8905 11239.3 2334.3 7 8905 9912.34 1007.3
2008-08 12249 8975.0 3274.0 8 12249 10101.26 2147.7
2008-09 19160 12150.8 7009.2 9 19160 10290.18 8869.8
2008-10 14264 18949.7 4685.7 10 14264 10479.1 3784.9
2008-11 13687 14404.6 717.6 11 13687 10668.02 3019.0
2008-12 8657 13708.5 5051.5 12 8657 10856.94 2199.9
2009-01 6733 8808.5 2075.5 13 6733 11045.86 4312.9
2009-02 1006 6795.3 5789.3 14 1006 11234.78 10228.8
2009-03 9328 1179.7 8148.3 15 9328 11423.7 2095.7
2009-04 9072 9083.6 11.6 16 9072 11612.62 2540.6
2009-05 8738 9072.3 334.3 17 8738 11801.54 3063.5
2009-06 12259 8748.0 3511.0 18 12259 11990.46 268.5
2009-07 9519 12153.7 2634.7 19 9519 12179.38 2660.4
2009-08 12113 9598.0 2515.0 20 12113 12368.3 255.3
2009-09 14044 12037.6 2006.4 21 14044 12557.22 1486.8
2009-10 16650 13983.8 2666.2 22 16650 12746.14 3903.9
2009-11 16912 16570.0 342.0 23 16912 12935.06 3976.9
2009-12 13851 16901.7 3050.7 24 13851 13123.98 727.0
2010-01 2923.3 2739.1

Plotting the data, it appears that there is a slight increasing trend,

but an averaging technique may also fit.
It is easier to use Exponential Smoothing (rather than Moving Average).
Alpha can be changed in order to find the minimum MAD.
This Turned out a very high alpha (.97), implying that the Naïve model is best.
Linear Trend model was tried. It led to be lower MAD.
Therefore, the best model is the linear trend model.
Bec. we only have 2 years of data and the data pattern seems to differ between the 2 years,
it is difficult to conclude that seasonality might be present.

25000

20000

15000 produced
ES Forecast
10000

5000

25000
No. produced
Linear (No. produced)
20000

15000
f(x) = 188.918695652174 x + 8589.89130434783
10000

5000

3-43
Instructor’s Manual, Chapter 3
 37. a. Annual sales are increasing by 15,000 bottles per year (the slope of the line)
b. Forecast for Year 6:
t = 6, Yt = 80 + 15(6) = 170, which is 170,000 bottles.
38. a.
850

Insurance
needed ($000)

010203040
Current Age of Head of Household (years)

b. y = 850 – 0.1(30) = 849.4. Thus, a 30-year old will need $849,000 of life insurance.

3-44 Operations Management, 5/C/e