SQQS3163 Forecasting Methods Matric No.: …………………..

QUESTION 1 (10 MARKS)

a) Describe the differences between qualitative forecasting techniques and quantitative

forecasting techniques.
(2 marks)

b) Complete the following sentences:

i. The overall upward or downward pattern of the data in annual time series will be
contained in the ____________ component.
(1 mark)

ii. The effect of an unpredictable, rare event will be contained in the ___________
component.
(1 mark)

iii. Based on the Mean Absolute Error (MAE) criterion in TABLE 1, the most appropriate
model is _____________.

TABLE 1

Model
Error measure
Linear trend Quadratic trend Exponential trend AR(2)
MAE 1.38 1.22 1.39 0.71
(1 mark)

c) Is the plot Y versus T in FIGURE 1 shows a stationary time series? Explain.

500

400

300

200

100

0
Y

1 5 9 13 17 21 25 29 33 37 41 45 49 53 57 61 65

FIGURE 1
(2 marks)

1
SQQS3163 Forecasting Methods Matric No.: …………………..

d) A study was conducted to forecast the number of visitors in BBC Garden for a few weeks.

number of visitors (Y)

20000
18000
16000
14000
12000
10000
8000
6000
4000
2000
0
1 3 5 7 9 11 13 15 17 19 21 23 25 27 29 31

week (T)

FIGURE 2

Based on the FIGURE 2, explain the pattern of data and state the TWO (2) suitable
techniques to analyze the data.
(3 marks)

QUESTION 2 (15 MARKS)

Quarterly profit of paper sales (MYR thousands) for the years 2013 to 2014 displayed in
TABLE 2. The OUTPUT 1 and OUTPUT 2 represents the result of time series regression for
deseasonalize data and trend respectively.

TABLE 2

Year Quarter Time (t) Profit (yt)

2013 1 1 58
2 2 63
3 3 65
4 4 77
2014 1 5 79
2 6 81
3 7 88
4 8 95

2
SQQS3163 Forecasting Methods Matric No.: …………………..

OUTPUT 1: Time series regression for profit of paper sales (yt) and time (t)

Coefficients Standard Error t Stat P-value

Intercept 52.5 1.808819 29.02447 1.11E-07
t 5.166667 0.358199 14.42399 6.96E-06

OUTPUT 2: Time series regression for deseasonalize data (Dt) and time (t)

Coefficients Standard Error t Stat P-value

Intercept 53.411 1.369 39.027 0.000
t 4.964 0.271 18.318 0.000

a) By choosing the suitable information given in OUTPUT 1 and OUTPUT 2, analyze the data
in TABLE 2 using the multiplicative decomposition method and produce the estimated
trend.
(13 marks)

b) Based on part (a), calculate the predicted value at the time period sixth. (Assume there is no
cycle component).
(2 marks)

QUESTION 3 (20 MARKS)

a) Give the appropriate situation to use Holt’s method.

(2 marks)

b) What is the range for smoothing constants in Holt’s method?

(1 mark)

c) Show that SSE   et2 using Holt’s method formula.

(4 marks)

3
SQQS3163 Forecasting Methods Matric No.: …………………..

d) ABC Airlines would like to know the number of passengers who will fly on ABC planes. By
using Holt’s method, the numbers of passengers who fly on ABC planes from year 1990 to
2013 have been analyzed. TABLE 3 display the results when smoothed series () and trend
estimate (β) values are respectively equal to 0.9999 and 0.9675. (Use formula:
St  yt  1   St 1  Tt 1  and Tt   St  St 1   1   Tt 1 ).

TABLE 3

Year Time Number of Smoothed Trend Forecast m Forecast

period passengers series estimate period into the error
(thousands) future
t yt S t  Tt  Ft m  et
1990 1 22.8000 A B C
1991 2 26.1000 D E 22.8000
1992 3 29.4000 29.4000 3.2974 29.3105

2012 23 147.9000 147.9005 3.2695 153.1104

2013 24 150.1000 150.1001 2.2344 F G

i. Find the value for A, B, C, D, E, F, G.

(11 marks)

ii. Forecast the number of passengers who will fly on ABC planes in year 2017.
(2 marks)

QUESTION 4 (15 MARKS)

a) State THREE (3) assumptions used when estimating a regression model.

(3 marks)

b) What is the effect of the model if the error terms for a time series model are autocorrelated?
(1 mark)

c) The computer output in APPENDIX A represents the result of time series regression analysis
on PSC printer sales. Based on the output, answer the following questions.

i. Write the equation of the estimate model.

(1 mark)

ii. Comment on the model obtained. Your answer should be based on the given information
in APPENDIX A. (Use α = 0.05)
(10 marks)

4
SQQS3163 Forecasting Methods Matric No.: …………………..

QUESTION 5 (20 MARKS)

a) Build a flow diagram or pseudocode to display the Box-Jenkins model-building strategy.

(5 marks)

b) Suppose the following time series model has been fit to historical data and found to be an
adequate model.
Yˆt  35  0.25 t 1  0.3 t  2

i. Name the model in part 5(b) and state the stationary condition of the model.
(2 marks)

ii. The first four observations are Y1=32.5, Y2=36.6, Y3=33.3 and Y4=31.9. Assume Yˆ1  35
and  0  0 , calculate forecast value for periods three if period two is the forecast origin.
(4 marks)

c) APPENDIX B presents the estimation, diagnostic checking and forecasting of shampoo sales
using model zt  1 zt 1  at where zt  yt  yt 1. Determine if this model is a reasonable
model for forecasting shampoo sales at α = 0.05. Write a brief report to support your
argument.
(9 marks)

END OF QUESTIONS

5
SQQS3163 Forecasting Methods Matric No.: …………………..

APPENDIX A

Notes: time = time = t

timesq = (time)2 = t2

Plot of y*time. Legend: A = 1 obs, B = 2 obs, etc.

y |
1500 +
|
|
|
|
| A
1000 + A A A A A
| A A
| A A A
|
| A
| A A A A
500 + A A
| A A
| A
| A
| A A
|
0 +
--+-------+-------+-------+-------+-------+-------+-------+--
0 4 8 12 16 20 24 28
time
NOTE: 4 obs had missing values.

The REG Procedure

Model: MODEL2
Dependent Variable: y

Analysis of Variance

Sum of Mean
Source DF Squares Square F Value Pr > F
Model 2 1872232 936116 266.69 <.0001
Error 21 73713 3510.14608
Corrected Total 23 1945945

Root MSE 59.24649 R-Square 0.9621

Dependent Mean 671.66667 Adj R-Sq 0.9585
Coeff Var 8.82082

Parameter Estimates

Parameter Standard
Variable DF Estimate Error t Value Pr > |t|
Intercept 1 100.25000 39.52936 2.54 0.0192
time 1 56.19070 7.28554 7.71 <.0001
timesq 1 -0.64147 0.28292 -2.27 0.0340

Durbin-Watson D 1.682
Number of Observations 24
1st Order Autocorrelation 0.145

6
SQQS3163 Forecasting Methods Matric No.: …………………..

APPENDIX B

Notes: SAS output of estimation, diagnostic checking and forecasting of shampoo

sales using model zt  1 zt 1  at where zt  yt  yt 1

Plot of y*time. Legend: A = 1 obs, B = 2 obs, etc.

y |
|
500 +
|
| A
|
450 + AAAA A A
|
| A A A
| A A A
400 + A AA A
| AA AA AA A A
| A
|
350 + A
| A
| A A
|
300 +
|
--+---------+---------+---------+----------------------------
0 10 20 30
time

Plot of z*time. Legend: A = 1 obs, B = 2 obs, etc.

z |
|
100 +
|
|
| A A
50 + A
| A A A
| A
| A A A A A
0 + AA A A A A A
| A A
| A A A A A
| A
-50 + A A
|
|
|
-100 +
|
--+---------+---------+---------+----------------------------
0 10 20 30
time

NOTE: 1 obs had missing values.

7
SQQS3163 Forecasting Methods Matric No.: …………………..

The ARIMA Procedure

Conditional Least Squares Estimation

Iteration SSE AR1,1 Lambda R Crit

0 14482 -0.64181 0.00001 1

1 14482 -0.64031 1E-6 0.001924
2 14482 -0.64031 1E-7 1.941E-8

ARIMA Estimation Optimization Summary

Estimation Method Conditional Least Squares

Parameters Estimated 1
Termination Criteria Maximum Relative Change in Estimates
Iteration Stopping Value 0.001
Criteria Value 2.347E-8
Alternate Criteria Relative Change in Objective Function
Alternate Criteria Value 3.77E-16
Maximum Absolute Value of Gradient 0.000357
R-Square Change from Last Iteration 1.941E-8
Objective Function Sum of Squared Residuals
Objective Function Value 14482.41
Marquardt's Lambda Coefficient 1E-7
Numerical Derivative Perturbation Delta 0.001
Iterations 2

Conditional Least Squares Estimation

Standard Approx
Parameter Estimate Error t Value Pr > |t| Lag

AR1,1 -0.64031 0.14765 -4.34 0.0002 1

Variance Estimate 517.2288

Std Error Estimate 22.74267
AIC 264.4869
SBC 265.8542
Number of Residuals 29
* AIC and SBC do not include log determinant.

Autocorrelation Check of Residuals

To Chi- Pr >
Lag Square DF ChiSq ------------------Autocorrelations-----------------

6 2.02 5 0.8464 0.029 0.107 0.056 -0.020 0.030 -0.193

12 7.47 11 0.7600 0.012 0.141 -0.225 0.072 -0.154 0.130
18 13.30 17 0.7156 0.169 0.045 0.236 -0.041 0.042 -0.061
24 21.56 23 0.5468 -0.144 -0.014 -0.129 -0.081 0.043 -0.137

8
SQQS3163 Forecasting Methods Matric No.: …………………..

Autocorrelation Plot of Residuals

Lag Covariance Correlation -1 9 8 7 6 5 4 3 2 1 0 1 2 3 4 5 6 7 8 9 1 Std Error

0 517.229 1.00000 | |********************| 0

1 14.885433 0.02878 | . |* . | 0.185695
2 55.111720 0.10655 | . |** . | 0.185849
3 28.961053 0.05599 | . |* . | 0.187944
4 -10.442400 -.02019 | . | . | 0.188518
5 15.494169 0.02996 | . |* . | 0.188593
6 -99.923700 -.19319 | . ****| . | 0.188757
7 6.038524 0.01167 | . | . | 0.195456
8 73.117856 0.14136 | . |*** . | 0.195480
9 -116.251 -.22476 | . ****| . | 0.198974
10 37.411342 0.07233 | . |* . | 0.207544
11 -79.399418 -.15351 | . ***| . | 0.208411
12 67.155623 0.12984 | . |*** . | 0.212275
13 87.224109 0.16864 | . |*** . | 0.214996
14 23.027297 0.04452 | . |* . | 0.219509

"." marks two standard errors

Inverse Autocorrelations

Lag Correlation -1 9 8 7 6 5 4 3 2 1 0 1 2 3 4 5 6 7 8 9 1

1 -0.06092 | . *| . |
2 -0.12068 | . **| . |
3 -0.07793 | . **| . |
4 -0.05594 | . *| . |
5 0.08716 | . |** . |
6 0.18770 | . |**** . |
7 -0.10322 | . **| . |
8 -0.19902 | . ****| . |
9 0.16755 | . |*** . |
10 -0.01667 | . | . |
11 0.15690 | . |*** . |
12 -0.06574 | . *| . |
13 -0.15822 | . ***| . |
14 -0.06732 | . *| . |

Partial Autocorrelations

Lag Correlation -1 9 8 7 6 5 4 3 2 1 0 1 2 3 4 5 6 7 8 9 1

1 0.02878 | . |* . |
2 0.10581 | . |** . |
3 0.05082 | . |* . |
4 -0.03443 | . *| . |
5 0.02017 | . | . |
6 -0.19448 | . ****| . |
7 0.01984 | . | . |
8 0.18720 | . |**** . |
9 -0.23420 | . *****| . |
10 0.05306 | . |* . |
11 -0.12655 | . ***| . |
12 0.13546 | . |*** . |
13 0.20818 | . |**** . |
14 0.08359 | . |** . |

9
SQQS3163 Forecasting Methods Matric No.: …………………..

APPENDIX C

Durbin-Watson Statistic

Critical values at 5% level of significance

n = number of observations
k' = number of explanatory variables excluding the intercept

k’ = 1 k’ = 2 k’ = 3 k’ = 4 k’ = 5
n dL dU dL dU dL dU dL dU dL dU
6 0.610 1.400 - - - - - - - -
7 0.700 1.356 0.467 1.896 - - - - - -
8 0.763 1.332 0.559 1.777 0.368 2.287 - - - -
9 0.824 1.320 0.629 1.699 0.455 2.128 0.296 2.588 - -
10 0.879 1.320 0.697 1.641 0.525 2.016 0.376 2.414 0.243 2.822
11 0.927 1.324 0.658 1.604 0.595 1.928 0.444 2.283 0.316 2.645
12 0.971 1.331 0.812 1.579 0.658 1.864 0.512 2.177 0.379 2.506
13 1.010 1.340 0.861 1.562 0.715 1.816 0.574 2.094 0.445 2.390
14 1.045 1.350 0.905 1.551 0.767 1.779 0.632 2.030 0.505 2.296
15 1.077 1.361 0.946 1.543 0.814 1.750 0.685 1.977 0.562 2.220
16 1.106 1.371 0.982 1.539 0.857 1.728 0.734 1.935 0.615 2.157
17 1.133 1.381 1.015 1.536 0.897 1.710 0.779 1.900 0.664 2.104
18 1.158 1.391 1.046 1.535 0.933 1.696 0.820 1.872 0.710 2.060
19 1.180 1.401 1.074 1.536 0.967 1.685 0.859 1.848 0.752 2.023
20 1.201 1.411 1.100 1.537 0.998 1.676 0.894 1.828 0.792 1.991
21 1.221 1.420 1.125 1.538 1.026 1.669 0.927 1.812 0.829 1.964
22 1.239 1.429 1.147 1.541 1.053 1.664 0.958 1.797 0.863 1.940
23 1.257 1.437 1.168 1.543 1.078 1.660 0.986 1.785 0.895 1.920
24 1.273 1.446 1.188 1.546 1.101 1.656 1.013 1.775 0.925 1.902
25 1.288 1.454 1.206 1.550 1.123 1.654 1.038 1.767 0.953 1.886
26 1.302 1.461 1.224 1.553 1.143 1.652 1.062 1.759 0.979 1.873
27 1.316 1.469 1.240 1.556 1.162 1.651 1.084 1.753 1.004 1.861
28 1.328 1.476 1.255 1.560 1.181 1.650 1.104 1.747 1.028 1.850
29 1.341 1.483 1.270 1.563 1.198 1.650 1.124 1.743 1.050 1.841
30 1.352 1.489 1.284 1.567 1.214 1.650 1.143 1.739 1.071 1.833
31 1.360 1.500 1.300 1.570 1.230 1.650 1.160 1.740 1.090 1.830
32 1.370 1.500 1.310 1.570 1.240 1.650 1.180 1.730 1.110 1.820
33 1.380 1.510 1.320 1.580 1.260 1.650 1.190 1.730 1.130 1.810
34 1.390 1.510 1.330 1.580 1.270 1.650 1.210 1.730 1.150 1.810
35 1.400 1.520 1.340 1.580 1.280 1.650 1.220 1.730 1.160 1.800

10