Moving Average (3 months)

2023
Month Demand (actual) Demand Forecast
January 100 -
February 105 -
March 100 -
April 110 -
May 107 -
June 103 -
July 115 -
August 112 -
September 100 -
October 106 -
November 110 106
December 120 105
January 112

Weighted Moving Average (3 months) - 0.2, 0.3, 0.5

2023
Month Demand (actual) Weights Demand Forecast
January 100 - -
February 105 - -
March 100 - -
April 110 - -
May 107 - -
June 103 - -
July 115 - -
August 112 0.2 -
September 100 0.3 -
October 106 0.5 -
November 100 105.40
December 110 101.80
January - 106.20

Exponential Smoothing Example

Month Forecasted Demand Actuals
𝑭_𝒕=𝑭_𝒕−𝟏+𝜶(𝑨_𝒕−𝟏−𝑭
January 100 101
February 105 104
March 100 99
April 110 110
May 107 108
June 103 103
July 115 110
August 112 115
September 100 95
October 99 102
November 99.30 110
December 100.37 120
January 102.33 100
February 102.10
𝑭_𝒕=𝑭_𝒕−𝟏+𝜶(𝑨_𝒕−𝟏−𝑭_𝒕−𝟏)
𝛼 = 0.10
LINEAR TREND

Week Actual Sales

1 700
2 724
3 720
4 728
5 740
6 742
7 758
8 750
9 770
10 775

1. Graph the data to determine if there is indeed an linear trend.

Sales
800

780

760

740

720

700

680

660

640

620

600
1 2 3 4 5 6 7

Actual Sales
2. Use "FORECAST.LINEAR" function to forecast sales in the succeeding weeks.

"=FORECAST.LINEAR(Period Forecasted,Range of actual sales or Y,Range of actual time

Week (X) Actual Sales (Y) Forecast

1 700
850
2 724
3 720
4 728 800

5 740
6 742 750
7 758
8 750
700
9 770
10 775
11 Err:502 650

12 790
13 797 600
1 2 3 4
14 805
15 812
16 820

Alternative Way (long method using MS Excel)

SLOPE (b) 7.51
INTERCEPT (a) 699.40

Week 11 forecast 782

Week 12 forecast 789.51
Week 13 forecast 797.02
7 8 9 10
ng weeks.

nge of actual time periods given or X)"

Sales

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

Actual Sales Forecast

15
SUMMARY OUTPUT

Regression Statistics
Multiple R 0.9166657
R Square 0.840276
Adjusted R S 0.8243036
Standard Erro 2.34403615
Observations 12

ANOVA
df SS MS F Significance F
Regression 1 289.054945 289.054945 52.608 2.74837E-05
Residual 10 54.9450549 5.49450549
Total 11 344

CoefficientsStandard Error t Stat P-value Lower 95% Upper 95% Lower 95.0%
Intercept -0.91208791 1.77627215 -0.51348433 0.61876583 -4.86986891 3.04569309 -4.86986891
X Variable 1 52.7472527 7.27233622 7.25313725 2.74837E-05 36.5434779 68.9510276 36.5434779
Upper 95.0%
3.04569309
68.9510276
Y variable - Predicted or dependent variable
X variable - Predictor or independent variable
*Start by plotting the data and see if there is a linear rel
Unit Sales (X) Profits (Y)
in $ millions in $ millions
7 0.15
2 0.1
6 0.13
4 0.15
14 0.25
15 0.27
16 0.24
12 0.2
14 0.27
20 0.44
15 0.34
7 0.17

Run the Regression function of Excel (Data > Data Analysis > Regression)
SUMMARY OUTPUT

Regression Statistics
Multiple R 0.916665698641932
R Square 0.840276003066701
Adjusted R Square 0.824303603373371
Standard Error 0.040735724184138
Observations 12

ANOVA
df SS MS F
Regression 1 0.087297674419 0.0872976744 52.608
Residual 10 0.016593992248 0.0016593992
Total 11 0.103891666667

Coefficients Standard Error t Stat P-value

Intercept 0.050600775193798 0.026869449164 1.8832085051 0.0890522754
Unit Sales (X)
in $ millions 0.01593023255814 0.002196323054 7.2531372523 2.748368E-05

1. To do your forecast using simple linear regression, consider the least square line formula Y = a + bx

2. Use the coefficients derived from the Regression run of Excel and plug them in the least square line formula.

3. Example, if sales (X) is $10 million, what is the forecasted profit?

X 10
a 0.050600775193798
b 0.01593023255814

Y = a + bx

Y 0.209903100775193 or $209,900
the data and see if there is a linear relationship between the 2 variables

Significance F
2.7483678E-05

Lower 95% Upper 95% Lower 95.0% Upper 95.0%

-0.0092680884 0.1104696388 -0.0092680884 0.1104696388

0.01103651983 0.0208239453 0.0110365198 0.0208239453

line formula Y = a + bx

n the least square line formula.

 Actual Forecast
Period
(A) (F)
1 107 110
2 125 121
3 115 112
4 118 120
5 108 109
(A − F) Error |Error| Error2
-3 3 9
4 4 16
3 3 9
-2 2 4
-1 1 1
SUM 13 39
MAD MSE
n=5 n-1=4
2.6 9.75

Average difference of the

actual values versus
forecasted values.

Control Chart Setup

Square Root of the MSE 3.1224989991992
UCL (0 + 2 times the sq rt of MSE) 6.2449979983984
LCL (0 - 2 times the sq rt of MSE) -6.2449979983984
[|Error|/Actual] × 100
2.80%
3.20%
2.61%
1.69%
0.93%
11.23%
MAPE
n=5
2.25% 97.7500%

Average percentage
difference of forecast
against actual values. This is
the average % error of the
forecast. Deduct this from
100% to get the forecast
accuracy.