Demand Forecasting

Demand Forecasting refers to the process of predicting the future demand for the firm.
Methods of forecasting demand can be qualitative or quantitative.

Regression is a quantitative method of forecasting demand and is considered as the most

popular method. It uses past and present data of the firm in determining the relationship of the
dependent and independent variables.

Problem

A firm has the following data

Year Sale
1 600
2 1550
3 1500
4 1500
5 2400
6 3100
7 2600
8 2900
9 3800
10 4500
11 4000
12 4900

Use the least square method to determine the sales for the next four years. Also, find out the
standard error of estimate.

In order to get the sales for the 13 th,14th , 15th and 16th year of a firm, linear regression analysis
must be used.

In linear regression analysis, there are two types of data – the Dependent and Independent
variable. In such case, sales will be the dependent variable (Y) and years will be the independent
variable (X).

In linear equation, the equation is written as y=mx+c where y is the dependent variable, x is the
independent variable, m is the slope of x and c is the constant or y-intercept.

In linear regression analysis, the equation is similarly written as Y=a+bx where y is the
dependent variable, x is the independent variable, b is the slope of x and a is the constant or y-
intercept.

Graphically

Sales Regression line

Y’

Year
The distance from the Y’ of the regression line to the sales data Y is known as a deviation. The
sum of the squares of the vertical distances between each data point and its corresponding data
point on the line is minimized. In such case we can say that the regression line is good fit. Note:
to determine the goodness of fit of the regression line, standard error of estimation must be
computed.

Y =a+bX
a=Ý −b X́

b=
∑ XY −n X́ Ý
∑ X 2 −n X́ 2

Where a is the y intercept

b is the slope of X

Ý is the mean of all Ys

X́ is the mean of all Xs
Y is each value of the dependent variable

X is each value of the independent variable

n is the sample size

X Y XY X2
1 600 600 1
2 1550 3100 4
3 1500 4500 9
4 1500 6000 16
5 2400 12000 25
6 3100 18000 36
7 2600 18200 49
8 2900 23200 64
9 3800 34200 81
10 4500 45000 100
11 4000 44000 121
12 4900 58800 144
78 33350 268200 650

X́ =
∑ X = 78 =6.5
n 12

Ý =
∑ Y = 33350 =2779.17
n 12

268200−(12)(6.5)(279.17)
b=
650−(12)(6.5)2
b=359. 6
a=2779.17−359.6 ( 6.5 )
a=441.7 7
Y =a+bX
Y =441.77+359.6 X regression model

Since the regression model is Y =441.77+359.6 X , substitute 13, 14,15, 16 to X to get the
predicted sales.

Y 13=441.77+359.6(13)
Y 13=5116.57

Y 14=441.77+359.6(14 )
Y 14=5476.17

Y 15=441.77+359.6(15)
Y 15=5835.77

Y 16=441.77+359.6(16)
Y 16=6195.37

Standard error of estimate

- is similar to standard deviation

- it determine how widely the errors are dispersed along regression of line and
determine the goodness of fit of the regression model.

s yx =√ ∑ ¿ ¿ ¿¿

Where yi is the given data of the dependent variable

yi is the value of the dependent variable computed by the equation

n is the sample size
X Y XY X2 Y’ (Yi-Yi’)2
1 600 600 1 801.37 40549.87
2 1550 3100 4 1160.97 151344.36
3 1500 4500 9 1520.57 423.12
4 1500 6000 16 1880.17 144529.22
5 2400 12000 25 2239.12 25882.37
6 3100 18000 36 2599.37 250630.39
7 2600 18200 49 2958.97 128859.46
8 2900 23200 64 3318.57 175200.84
9 3800 34200 81 3678.17 14842.54
10 4500 45000 100 4037.77 213656.57
11 4000 44000 121 4397.37 157902.91
12 4900 58800 144 4756.97 20957.58
78 33350 268200 650 1324279.21

Y 1=441.77+359.6 ( 1 )=801.37 Y 2=441.77+359.6 ( 2 )=1160.97

 (Yi-Yi’)2= (600-801.37)2=40549.87

1324279.21
s yx =
√ 12−2
s yx =363.9 Note: the lower the value of the estimate the better

Other method of computing the standard error of estimate

∑ Y 2−a ∑ Y −b ∑ XY
s yx =
√ n−2

X Y XY X2 Y2
1 600 600 1 360000
2 1550 3100 4 2402500
3 1500 4500 9 2402500
4 1500 6000 16 2250000
5 2400 12000 25 5760000
6 3100 18000 36 9610000
7 2600 18200 49 6760000
8 2900 23200 64 8410000
9 3800 34200 81 14440000
10 4500 45000 100 20250000
11 4000 44000 121 16000000
12 4900 58800 144 24010000
78 33350 268200 650 112502500

∑ Y 2−a ∑ Y −b ∑ XY
s yx =
√ n−2

112502500−441.77 ( 33350 ) −359.6(268200)

s yx =
√ 12−2
s yx =363.9

By using the two methods of computing the standard error of estimation, you will arrive with
the same answer.