Quantitative Techniques in Management

Chapter 3
Forecasting Techniques
Contents
Forecasting Techniques .............................................................................................................................. 62
3.1 Estimation Using the Regression Line ............................................................................................... 62
3.2 Scatter Diagram ................................................................................................................................ 69
3.3 Karl Pearson’s Coefficient of Correlation .......................................................................................... 70
3.4 Time Series Analysis .......................................................................................................................... 72
3.5 Forecasting ........................................................................................................................................ 82
 Quantitative Techniques in Management

Forecasting Techniques
So, far we have studied problems relating to one variable only. In practice we come across large
number of problems involving the use of two or more than two variables. If two quantities vary
in such a way that movements in one are accompanied by movements in other, these quantities
are correlated.

Managers make personal and professional decisions based on prediction of future events. For this
they rely on relationship between known and what is to be estimated. If decision makers can
determine how the known is related to the unknown, they can aid the decision-making process
considerably.

Regression and correlation analyses show us how to determine both the nature and the strength
of relationship between two variables. In regression analysis, we develop an estimating equation
i.e. a mathematical formula that relates the known variables to the unknown variable. Afterwards
we can apply correlation analysis to determine the degree to which the variables are related.

Relationship Type

Regression and correlation analyses are based on the relationship, or association, between two
(or more) variables. The known variable is called the independent variable. The variable which is
need to predict is dependent variable.

Scatter Diagrams

First step in determining whether there is a relationship between two variables is to examine the
graph of the observed data. This graph or chart is called the scatter diagram. Scatter diagram
gives two types of information. First, we can look for patterns that indicate that the variables are
related. Then, if variables are related, then what kind of estimating equation, describes this
relationship.

3.1 Estimation Using the Regression Line

In the preceding chapter, it was mentioned that the first step in the correlation analysis is to plot
the values of the two series, X and Y, on the graph and to have a visual idea of the scatter
diagram that emerges. In that chapter, we had also given four different types of scatter diagrams
to indicate different types of relationships between two variables. These were: positive, negative,
spurious and non-linear, here we shall show how to calculate the regression line by using an
equation that relates the two variables. It may be noted that we are concerned here with the linear
relationship between two variables only. The equation for a straight line is:

Y = a + bX
 Quantitative Techniques in Management
where Y is the dependent variable, X is the independent variable, a is the Y-intercept, which is
the point at which the regression line crosses the Y-axis (the vertical axis) and b is the slope of
the regression line. It should be noted that the values of both a and b will remain constant for any
given straight line. On the basis of this equation, we can find the value of Y for any value of X if
values of a and b are known to us.

Suppose that a is 6 and b is 3 and we have to find the value of Y if X is 10.

Y = a + bX
= 6 + (3 *
Now how to determine the values of two constants a and b? The slope of the straight line can be
numerically calculated by the following formula:
b = (Y2 – Y1) / (X2 – X1)
On the basis of this equation, we can now find the values of the dependent variable Y for varying
values of X.

Least Squares Method

In a scatter diagram, if a straight line is drawn, it obviously does not touch several points. In fact,
it may not pass through any of the scattered points. Some points may be above the straight line
while the remaining points may be below it. Surely, there must be a technique to draw a line that
is best fit‘. Such a straight line will minimize the error between the estimated point that lies on
the line and the actual point based on the data given. The technique used for this purpose is
known as the method of least squares.
We have used Y so far to represent the individual value of the observed points on the Y-axis. A
new symbol Yc (computed or estimated value of Y) is used to represent individual values of the
estimated points, that is, those points that actually lie on the estimating line. In view of this, the
equation for the estimating line becomes Yc = a + bX. The difference between the actual point
and the estimated point is shown by a vertical line in respect of each of the seven points either
above or below it.
We may raise a pertinent issue here: which would give a better fit-one or two points which are
far away from the observed points in an estimating equation or a few points showing small
differences from the original or observed points? We can visualize that the second alternative-
smaller differences between the estimated points and the observed points would give a better
fit‘. This leads us to square the individual differences or errors before summing them up. This
process of squaring serves two purposes. First, it magnifies the larger differences or errors.
Second, the effects of positive and negative differences get cancelled. This is because the square
of negative differences or errors is a positive figure. This process is known as the least square
method as it minimizes the sum of the squares of the error in the estimating line.
With the help of this method of least squares, we can find out whether one estimating line is
better than the other. But, as a very large number of estimating lines can be drawn on the basis of
the same set of data, it is extremely difficult for us to say that the estimating line that we have
obtained is the best fitting line. To overcome this problem, a set of two equations known as the
normal equations is used to determine both the Y-intercept and the slope of the best-fitting
regression line. The two normal equations are:

∑Y = na +b ∑X
∑XY = a∑X + ∑X2
 Quantitative Techniques in Management
where
∑Y = the total of Y series
n = number of observations
∑X = the total of X series
∑XY = the sum of XY column
∑X2 = the total of squares of individual items in X series
a and b are the Y-intercept and the slope of the regression line, respectively. We now take up an
example to illustrate the use of the two normal equations. Given the following data, find the
regression equation of Y and X.

Example:
X 2 3 4 5 6
Y 7 9 10 14 15

Solution
We have now to set up a worksheet to get the values of the terms shown earlier. Worksheet for
Computing Correlation
X Y XY X2
2 7 14 4
3 9 27 9
4 10 40 16
5 14 70 25
6 15 90 36
∑X = 20 ∑Y = 55 ∑XY = 241 ∑X2 =90
Substituting these values in the normal equations given above
55 = 5a + 20b (i)
241 = 20a + 90b (ii)
Solving these we get,
a = 2.6
b = 2.1
Therefore, the regression equation of Y on X is
Y = 2.6 + 2.IX

Alternative Approach
We can use an alternative approach, which involves the use of two formulae-one to calculate the
Y-intercept and the other to calculate the slope. The formula for calculating the slope is

b = ∑XY- nXY )/( ∑X 2- nX2 )

In order to apply the above formula, we should know the values of X and Y in addition to those
of ∑XY and ∑X2 .
X =∑X /n =20/5 =4
Y =∑Y /n =55/5 =

Calculating the value of b‘ from above equation, we get

b = 2.1
 Quantitative Techniques in Management
The formula for calculating the Y-intercept of the line is
a =Y bX
where a is Y-intercept. Applying this formula to calculate the Y-intercept, we get
a = 2.6
Hence, the regression equation is Y = 2.6 + 2.1X (same as was obtained earlier by applying the
normal equations). On the basis of this regression, we can find the value of Y for any value of X.
Suppose that we have to ascertain the value of Y where X is 14. Applying this value of X = 14 in
the above equation,

Y = 2.6 + (2.1 *

We are now clear as to how the regression line is obtained. The question is how to check the
accuracy of our results. One method is to draw a scatter diagram with original data pertaining to
X and Y series and then to fit a straight line. This graph will give a visual idea about the
suitability of the straight line fitted. A more refined and, therefore, better approach is based on
the mathematical properties of a line fitted by the method of least squares. This means that the
positive and the negative errors (i.e. differences between the original data points and the
calculated points) must be equal so that when all individual errors are added together, the result
is zero.

X Y Yc Y-Yc
2 7 2.6 + (2.1 *
3 9 2.6 + (2.1 *
4 10 2.6 + (2.1 * -1.0
5 14 2.6 + (2.1 *
6 15 2.6 + (2.1 * -0.2
Total 0

Here, the calculated value of Y is shown as Yc. We find that the sum of positive errors Y - Yc, is
equal to 1.2. The same is true for negative errors. Thus, the sum of the column Y- Yc, comes to
zero. This means problem solved is correct.

Regression Coefficients
So far our discussion on regression analysis related to finding the regression of Y on X. It is just
possible that we may think of X as dependent variable and Y as an independent one. In that case,
we may have to use X = a + bY as an estimating equation. Then, the normal equations will be
∑X = na +b ∑Y
∑XY = a∑Y + ∑Y2
Here, we will have to get the values of ∑X, ∑Y, ∑XY, ∑Y2 and n. Once these values are known,
we may enter them in the two normal equations. The equations then can be solved in the same
manner as in the case of regression of Y on X.
i. Regression equation of Y on X
Y Y =r(s y /s x )(X X )
The term r(s y /s x ) can be written as ∑XY / ∑X2
ii. Regression equation of X on Y
X X =r(s x /s y )(Y Y )
 Quantitative Techniques in Management
The term r(s y /s x ) can be written as ∑XY / ∑Y2
It may be noted that the square root of the product of two regression coefficients is the value of
the coefficient of correlation.
We may write
bxy or r(s x /s y ) =∑xy/∑y2
byx or r(s x /s y ) =∑xy/∑x 2
r =(bxy *b yx )^0.5
Another point to note is that x and y are the deviations in X and Y series from their arithmetic
means.

Example:
X Y x =X- X x2 y =Y- Y y2 xy
2 7 -2 4 -4 16 8
3 9 -1 1 -2 4 2
4 10 0 0 -1 1 0
5 14 1 1 3 9 3
6 15 2 4 4 16 8
20 55 ∑x = 0 ∑x 2 = 10 ∑y = 0 ∑y2 = 46 ∑xy = 21

X =20/5 = 4
Y =55/5 = 11
Regression equation of X on Y:
X X =r(s x /s y )(Y Y )
On putting the values, we get
Y = 2.6 + 2.1X
Also,
r =(bxy *b yx)^0.5
= ∑xy/∑y 2 )*( ∑xy/∑x2 ))^0.5
= ((21/46)*(21/10))^0.5
= (0.9587)^0.5
= 0.98
Regression equation of X on Y is
X X =r(s x /s y )(Y Y )
or X = 40 + 0.5 (10/9) (Y - 45)
or X = 40 + 0.556 (Y - 45)
or X = 40 + 0.556Y- 25.02
or X = 14.98 + 056Y
In order to estimate the value of Y for X = 48, we have to use the regression equation of Y on X
Y = 27 + 0.45X
when X= 48
Y = 27 + (0.45 *
or Y= 27 + 21.6
or Y = 48.6

Properties of Regression Coefficients

 Quantitative Techniques in Management
At this stage, we should know the properties of the regression coefficients, which are given
below.
1. The coefficient of correlation is the geometric mean of the two regression coefficients.
2. As the coefficient of correlation cannot exceed 1, in case one of the regression coefficients is
greater than 1, then the other must be less than 1.
3. Both the regression coefficients will have the same sign, either positive or negative. If one
regression coefficient is positive, then the other will also be positive.
4. The coefficient of correlation and the regression coefficients will have the same sign. If the
regression coefficients are positive, then the correlation coefficient will also be positive and vice
versa.
5. The average of the two regression coefficients will always be greater than the correlation
coefficient. Symbolically,

This should be obvious as r happens to be the square root of the two regression coefficients.
6. Finally, regression coefficients are not affected by change of origin. But this is not the case in
respect of scale.

The Standard Error of Estimate

We now turn to the concept of the standard error of estimate. It is the measure of the spread of
observed values from the estimated ones, expressed by regression equation. This concept is
similar to the standard deviation, which measures the variation of individual items about the
arithmetic mean. As in the case of standard deviation, we can find the standard error of an
estimate by calculating the mean of squares of deviations between the actual or observed values
and the estimated values based on the regression equation. It can be written symbolically

where Yc is the calculated or estimated value of Y. Y is the actual or observed value of variable
n is the total number of observations Syx is the standard deviation of regression of Y values on X
values It may be noted that the sum of the squared deviations is divided by n-2 and not by n This
is because we have lost 2 degrees of freedom in estimating the regression line. One can reason
out that while obtaining values of a and b from the sample data, 2 degrees of freedom have been
lost while estimating the regression line from these points. There is a short-cut method for
finding the standard error of estimate. The formula is

where
X = values of the independent variable
Y = values of dependent variable
a = Y-intercept
b = slope of the estimating equation
n = number of observations
It should be obvious that this formula gives a short-cut method. When we estimate the regression
equation, all the values, which we need, are already determined.
 Quantitative Techniques in Management
Example
Suppose that we have been given the following data pertaining to two series X and Y

X 40 30 20 50 60 40 20 60
Y 100 80 60 120 150 90 70 130
X series indicates advertising expenditure in thousand rupees and Y series relates to sales in
units. We are told the regression equation is Yc =24.444 + 1.889 X. We are asked to calculate
the standard error of estimate.

Solution
X Y Yc (Yi-Yc) (Yi-Yc)2
40 100 100 0
30 80 81 -1 1
20 60 62 -2 4
50 120 119 1 1
60 150 138 12 144
40 90 100 -10 100
20 70 62 8 64
60 130 138 -8 64
Total: 378

= (378/6)^0.5
= (63)^0.5
= 7.9 (standard error)
Higher the magnitude of the standard error of estimate, the greater is the dispersion or variability
of points around the regression line. In contrast, if the standard error of estimate is zero, then we
may take it that the estimate in equation is the best estimator of the dependent variable. In such a
case, all the points would lie on the regression line. As such, there would be no point scattered
around the regression line.

SIMPLE LINEAR CORRELATION

As the summer heat rises, hill stations, are crowded with more and more visitors. Ice-cream sales
become more brisk. Thus, the temperature is related to number of visitors and sale of ice-creams.
Correlation analysis is a means for examining such relationships systematically. It deals with
questions such as:
• Is there any relationship between two variables? If the value of one variable changes, does the
value of the other also change? Do both the variables move in the same direction? How strong is
the relationship?
The purpose of correlation analysis is to measure and interpret the strength of a linear or
nonlinear (eg, exponential, polynomial, and logistic) relationship between two continuous
variables. When conducting correlation analysis, we use the term association to mean linear
association
 Quantitative Techniques in Management

Simple linear correlation is a measure of the degree to which two variables vary together, or a
measure of the intensity of the association between two variables. The parameter being measure
is r (rho) and is estimated by the statistic r, the correlation coefficient. r can range from -1 to 1,
and is independent of units of measurement. The strength of the association increases as r
approaches the absolute value of 1.0. A value of 0 indicates there is no association between the
two variables tested.
A better estimate of r usually can be obtained by calculating r on treatment means averaged
across replicates. Correlation does not have to be performed only between independent and
dependent variables. Correlation can be done on two dependent variables. The X and Y in the
equation to determine r do not necessarily correspond between an independent and dependent
variable, respectively.
Methods measuring Correlation-
Widely used techniques for the study of correlation are scatter diagrams, Karl Pearson‘s
coefficient of correlation and Spearman‘s rank correlation. A scatter diagram visually presents
the nature of association without giving any specific numerical value. A numerical measure of
linear relationship between two variables is given by Karl Pearson‘s coefficient of correlation. A
relationship is said to be linear if it can be represented by a straight line. Another measure is
Spearman‘s coefficient of correlation, which measures the linear association between ranks
assigned to individual items according to their attributes. Attributes are those variables which
cannot be numerically measured such as intelligence of people, physical appearance, honesty etc.

3.2 Scatter Diagram

A scatter diagram is a useful technique for visually examining the form of relationship, without
calculating any numerical value. In this technique, the values of the two variables are plotted as
points on a graph paper. The cluster of points, so plotted, is referred to as a scatter diagram. From
a scatter diagram, one can get a fairly good idea of the nature of relationship. In a scatter diagram
the degree of closeness of the scatter points and their overall direction enable us to examine the
relationship. If all the points lie on a line, the correlation is perfect and is said to be unity. If the
scatter points are widely dispersed around the line, the correlation is low. The correlation is said
to be linear if the scatter points lie near a line or on a line.

Scatter plots are a useful means of getting a better understanding of your data.
 Quantitative Techniques in Management

3.3 Karl Pearson’s Coefficient of Correlation

This is also known as product moment correlation and simple correlation coefficient. It gives a
precise numerical value of the degree of linear relationship between two variables X and Y. The
linear relationship may be given by Y = a + bX This type of relation may be described by a
straight line. The intercept that the line makes on the Y-axis is given by a and the slope of the
line is given by b. It gives the change in the value of Y for very small change in the value of X.
On the other hand, if the relation cannot be represented by a straight line as in Y = X2 the value
of the coefficient will be zero. It clearly shows that zero correlation need not mean absence of
any type of relation between the two variables. Let X1, X2, ..., XN be N values of X and Y1, Y2
,..., YN be the corresponding values of Y. In the subsequent presentations the subscripts
indicating the unit are dropped for the sake of simplicity. The arithmetic means of X and Y are
defined as

and their variances are as follows

The standard deviations of X and Y respectively are the positive square roots of their variances.
Covariance of
X and Y is defined as

Where x = X-X and y = X – Y are the deviations of the ith value of X and Y from their mean
values respectively.
The sign of covariance between X and Y determines the sign of the correlation coefficient. The
standard deviations are always positive. If the covariance is zero, the correlation coefficient is
always zero. The product moment correlation or the Karl Pearson‘s measure of correlation is
given by
 Quantitative Techniques in Management

Strong and weak are words used to describe correlation. If there is strong correlation, then the
points are all close together. If there is weak correlation, then the points are all spread apart.
There are ways of making numbers show how strong the correlation is. These measurements are
called correlation coefficients. The best known is the Pearson product-moment correlation
coefficient. If the answer is 1, then there is strong correlation. If the answer is -1, then there is
weak correlation. Another kind of correlation coefficient is Spearman's rank correlation
coefficient.

Importance if Correlation Analysis-

The study if correlation is of immense use in impractical life in view of following:
1. Most of the variables in economics and business area show relationship.
2. Once the correlation is established between two variables, regression analysis helps us to
estimate value of dependent variable for the given value of independent variable.
3. Correlation analysis together with regression analysis helps us to understand the behavior
of various social and economic variables.
4. The effect of correlation is to reduce the range of uncertainty in our predictions.

There are following types of Correlations-

1. Positive or Negative
2. Linear or Non-linear

Positive and Negative Correlation-

Positive or direct correlation refers to the movement of variables in the same direction. The
correlation is said to be positive or direct when the increase (decrease) in the value of one
variable is accompanied by an increase (decrease) in the value of the other variable.

Linear and Non-Linear Correlation

In perfect linear correlation the amount of the change in one variable bears a constant ratio to the
amount of change in one variable bears a constant ratio to the amount of change in the other. The
graph of variables having such a relation will be a straight line.

On the other hand, in non-linear correlation, the amount in one variable does not bear a constant
ratio to the amount of change in other variable.

Properties of Correlation Coefficient

• r has no unit. It is a pure number. It means units of measurement are not part of r.
• A negative value of r indicates an inverse relation. A change in one variable is associated with
change in the other variable in the opposite direction.
 Quantitative Techniques in Management
• If r is positive the two variables move in the same direction.
• If r = 0 the two variables are uncorrelated. There is no linear relation between them. However
other types of relation may be there.
• If r = 1 or r = –1 the correlation is perfect. The relation between them is exact.
• A high value of r indicates strong linear relationship. Its value is said to be high when it is close
to +1 or –1.
• A low value of r indicates a weak linear relation. Its value is said to be low when it is close to
zero.
• The value of the correlation coefficient lies between minus one and plus one, –1 <= r<= 1. If,
in any exercise, the value of r is outside this range it indicates error in calculation.
• The value of r is unaffected by the change of origin and change of scale.

3.4 Time Series Analysis

Arrangement of statistical data in chronological order ie., in accordance with occurrence of time,
is known as Time Series . Such series have a unique important place in the field of Economic
Business statistics. A business man is interested in finding out his likely sales in the near future,
so that the businessman could adjust his production accordingly and avoid the possibility of
inadequate production to meet the demand. In this connection one usually deal with statistical
data, which are collected, observed or recorded at successive intervals of time. Such data are
generally referred to as time series‘.
According to Mooris Hamburg A time series is a set of statistical observations arranged in
chronological order Ya-Lun- chou defining the time series as A time series may be defined as
a collection of readings belonging to different time periods, of some economic variable or
composite of variables. A time series is a set of observations of a variable usually at equal
intervals of time. Here time may be yearly, monthly, weekly, daily or even hourly usually at
equal intervals of time.
The Primary purpose of the analysis of time series is to discover and measure all types of
variations which characterise a time series. The central objective is to decompose the various
elements present in a time series and to use them in business decision making.
Components of Time series:
The components of a time series are the various elements which can be segregated from the
observed data. The following are the broad classification of these components.

In time series analysis, it is assumed that there is a multiplicative relationship between these four
components.
Symbolically,
 Quantitative Techniques in Management
Y=T S C I
Where Y denotes the result of the four elements; T = Trend; S = Seasonal component; C =
Cyclical components; I = Irregular component
In the multiplicative model it is assumed that the four components are due to different causes but
they are not necessarily independent and they can affect one another. Another approach is to treat
each observation of a time series as the sum of these four components. Symbolically
Y = T + S+ C+ I
The additive model assumes that all the components of the time series are independent of one
another.
1) Secular Trend or Long - Term movement or simply Trend
2) Seasonal Variation
3) Cyclical Variations
4) Irregular or erratic or random movements (fluctuations)

Secular Trend:
It is a long term movement in Time series. The general tendency of the time series is to increase
or decrease or stagnate during a long period of time is called the secular trend or simply trend.
Methods of Measuring Trend:
Trend is measured by the following mathematical methods.
1. Graphical method
2. Method of Semi-averages
3. Method of moving averages
4. Method of Least Squares
Graphical Method:
This is the easiest and simplest method of measuring trend. In this method, given data must be
plotted on the graph, taking time on the horizontal axis and values on the vertical axis. Draw a
smooth curve which will show the direction of the trend. While fitting a trend line the following
important points should be noted to get a perfect trend line.
(i) The curve should be smooth.
(ii) As far as possible there must be equal number of points above and below the trend line.
(iii) The sum of the squares of the vertical deviations from the trend should be as small as
possible.
(iv)If there are cycles, equal number of cycles should be above or below the trend line.
(v) In case of cyclical data, the area of the cycles above and below should be nearly equal.

Example:
Fit a trend line to the following data by graphical method.
Year 1996 1997 1998 1999 2000 2001 2002
Sales (in Rs 000) 60 72 75 65 80 85 95

Solution:
 Quantitative Techniques in Management

The dotted lines refers trend line

Merits:
1. It is the simplest and easiest method. It saves time and labour.
2. It can be used to describe all kinds of trends.
3. This can be used widely in application.
4. It helps to understand the character of time series and to select appropriate trend.

Demerits:
1. It is highly subjective. Different trend curves will be obtained by different persons for the
same set of data.
2. It is dangerous to use freehand trend for forecasting purposes.
3. It does not enable us to measure trend in precise quantitative terms.
Method of semi averages:
In this method, the given data is divided into two parts, preferably with the same number of
years. For example, if we are given data from 1981 to 1998 i.e., over a period of 18 years, the
two equal parts will be first nine years, i.e., 1981 to 1989 and from 1990 to 1998. In case of odd
number of years like 5,7,9,11 etc, two equal parts can be made simply by omitting the middle
year. For example, if the data are given for 7 years from 1991 to 1997, the two equal parts would
be from 1991 to 1993 and from 1995 to 1997, the middle year 1994 will be omitted.
After the data have been divided into two parts, an average of each part is obtained. Thus we get
two points. Each point is plotted at the mid-point of the class interval covered by respective part
and then the two points are joined by a straight line which gives us the required trend line. The
line can be extended downwards and upwards to get intermediate values or to predict future
values.

Example :
Draw a trend line by the method of semi-averages.
Year 1991 1992 1993 1994 1995 1996
Sales Rs in (1000) 60 75 81 110 106 117
Solution:
Divide the two parts by taking 3 values in each part.
 Quantitative Techniques in Management

Difference in middle periods = 1995 –1992 = 3 years

Difference in semi averages = 111 –72 = 39
Annual increase in trend = 39/3 = 13
Trend of 1991 = Trend of 1992 -13
= 72-13 = 59
Trend of 1993 = Trend of 1992 +13
= 72 + 13 = 85
Similarly, we can find all the values

The following graph will show clearly the trend line.

Merits:
1. It is simple and easy to calculate
2. By this method every one getting same trend line.
3. Since the line can be extended in both ways, we can find the later and earlier estimates.
Demerits:
1. This method assumes the presence of linear trend to the values of time series which may not
exist.
2. The trend values and the predicted values obtained by this method are not very reliable.

Method of Moving Averages:

This method is very simple. It is based on Arithmetic mean. Theses means are calculated from
overlapping groups of successive time series data. Each moving average is based on value
covering a fixed time interval, called period of moving average and is shown against the center
of the interval. The method of odd period of moving average is as follows. The moving averages
 Quantitative Techniques in Management
for three years is (a+b+c)/3, (b+c+d)/3, (c+d+e)/3, etc. Formula for five yearly moving average is
(a+b+c+d+e)/5, (b+c+d+e+f)/5, (c+d+e+f+g)/5, etc.

Steps for calculating odd number of years are following:

1. Find the value of three years total, place the value against the second year.
2. Leave the first value and add the next three years value (ie 2nd, 3rd and 4th years value) and=
put it against 3rd year.
3. Continue this process until the last year‘s value taken.
4. Each total is divided by three and placed in the next column.

Example:
Calculate the three yearly averages of the following data.
Year 1975 1976 1977 1978 1979 1980 1981 1982 1983 1984
Production in (tones) 50 36 43 45 39 38 33 42 41 34

Solution:

Even Period of Moving Averages:

The middle period of each set of values lies between the two time points in case of even moving
period. So we must center the moving averages.
The steps are
1. Find the total for first 4 years and place it against the middle of the 2nd and 3rd year in the
third column.
2. Leave the first year value, and find the total of next four-year and place it between the 3rd and
4th year.
3. Continue this process until the last value is taken.
4. Next, compute the total of the first two four year totals and place it against the 3rd year in the
fourth column.
5. Leave the first four years total and find the total of the next two four years‘ totals and place it
against the fourth year.
6. This process is continued till the last two four years‘ total is taken into account.
 Quantitative Techniques in Management
7. Divide this total by 8 (Since it is the total of 8 years) and put it in the fifth column.
These are the trend values.

Example: The production of Tea in India is given as follows. Calculate the Four-yearly moving
averages
Year 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002
Production (tones) 464 515 518 467 502 540 557 571 586 612

Solution:

Merits:
1. The method is simple to understand and easy to adopt as compared to other methods.
2. Method is flexible as mere addition of more figures to the data will not change the entire
calculation. That will produce only some more trend values.
3. Regular cyclical variations can be completely eliminated by a period of moving average equal
to the period of cycles.
4. It is particularly effective if the trend of a series is very irregular.
Demerits:
1. It cannot be used for forecasting or predicting future trend, which is the main objective of
trend analysis.
2. The choice of the period of moving average is sometimes subjective.
3. Moving averages are generally affected by extreme values of items.
4. It cannot eliminate irregular variations completely.
 Quantitative Techniques in Management
Method of Least Square:
This method is widely used. It plays an important role in finding the trend values of economic
and business time series. It helps for forecasting and predicting the future values. The trend line
by this method is called the line of best fit.
The equation of the trend line is y = a + bx, where the constants a and b are to be estimated so as
to minimize the sum of the squares of the difference between the given values of y and the
estimate values of y by using the equation. The constants can be obtained by solving two normal
equations.
∑y = na + b∑x ………. (1)
∑xy = a∑x + b∑x2 ……… (2)
Here x represent time point and y are observed values. n‘ is the number of pair- values.
When odd numbers of years are given
Step 1: Writing given years in column 1 and the corresponding sales or production etc in column
2.
Step 2: Write in column 3 start with 0, 1, 2 .. against column 1 and denote it as X
Step 3: Take the middle value of X as A
Step 4: Find the deviations u = X - A and write in column 4
Step 5: Find u2 values and write in column 5.
Step 6: Column 6 gives the product uy
Now the normal equations become
∑y = na + b∑u where u = X-A
∑uy = a∑u + b∑u2
Since ∑u = 0, From above equation
a = y/n
∑uy = b∑u2
b = y/∑u2
The fitted straight line is
y = a + bu = a + b (X - A)

Example:
Fit a straight line trend by the method of least squares for the following data.
Year 1983 1984 1985 1986 1987 1988
Sales (Rs. in lakhs) 3 8 7 9 11 14
Also estimate the sales for the year 1991

Solution:
 Quantitative Techniques in Management

u = (X-A)/(1/ 2)
= 2 (X 2.5) = 2X 5
The straight line equation is
y = a + bX = a + bu

The normal equations are

y = na …….(1)
uy = b u2 ……(2)
From (1) 52 = 6a

a = 52/6
= 8.67
From (2) 66 = 70 b
b = 66/70
= 0.94
The fitted straight line equation is
y = a+bu
y = 8.67+0.94(2X-5)
y = 8.67 + 1.88X - 4.7
y = 3.97 + 1.88X
The trend values are
Put X = 0, y = 3.97 X = 1, y = 5.85
X = 2, y = 7.73 X = 3, y = 9.61
X = 4, y = 11.49 X = 5, y = 13.37
The estimated sale for the year 1991 is; put X = x –1983
= 1991 – 1983 = 8
y = 3.97 + 1.88 × 8
= 19.01 lakhs
The following graph will show clearly the trend line.
 Quantitative Techniques in Management

Merits:
1. Since it is a mathematical method, it is not subjective so it eliminates personal bias of the
investigator.
2. By this method we can estimate the future values. As well as intermediate values of the time
series.
3. By this method we can find all the trend values.
Demerits:
1. It is a difficult method. Addition of new observations makes recalculations.
2. Assumption of straight line may sometimes be misleading since economics and business time
series are not linear.
3. It ignores cyclical, seasonal and irregular fluctuations.
4. The trend can estimate only for immediate future and not for distant future.

Seasonal Variations:
Seasonal Variations are fluctuations within a year during the season. The factors that cause
seasonal variation are
i) Climate and weather condition. ii)
Customs and traditional habits.
Measurement of seasonal variation:
The following are some of the methods more popularly used for measuring the seasonal
variations.
1. Method of simple averages.
2. Ratio to trend method.
3. Ratio to moving average method.
4. Link relative method
Method of simple averages
The steps for calculations:
i) Arrange the data season wise
ii) Compute the average for each season.
iii) Calculate the grand average, which is the average of seasonal averages.
iv) Obtain the seasonal indices by expressing each season as percentage of Grand average
The total of these indices would be 100n where n‘ is the number of seasons in the year.
 Quantitative Techniques in Management
Example :
Find the seasonal variations by simple average method for the data given below.

Solution:

Grand average = (47.6 63.6 58.8 54)/4

=224/4 = 56
Seasonal Index for I quarter = (First quarterly Average/ Grand Average) 100
= (47.6/56) 100 = 85
Seasonal Index for II quarter = (Second quarterly Average/Grand Average) 100
= (63.6/56) 100 = 113.6
Seasonal Index for III quarter = (Third quarterly Average/Grand Average) 100
= (58.8/56) 100 = 105
Seasonal Index for IV quarter = (Fourth quarterly Average/Grand Average) 100
= (54/56) 100 = 96.4

Cyclical variations:
The term cycle refers to the recurrent variations in time series that extend over longer period of
time, usually two or more years. Most of the time series relating to economic and business show
some kind of cyclic variation. A business cycle consists of the recurrence of the up and down
movement of business activity. It is a four-phase cycle namely.
1. Prosperity
2. Decline
3. Depression
4. Recovery
Each phase changes gradually into the following phase. The following diagram illustrates a
business cycle.
 Quantitative Techniques in Management

The study of cyclical variation is extremely useful in framing suitable policies for stabilizing the
level of business activities. Businessmen can take timely steps in maintaining business during
booms and depression.

Irregular variation:
Irregular variations are also called erratic. These variations are not regular and which do not
repeat in a definite pattern. These variations are caused by war, earthquakes, strikes flood,
revolution etc. This variation is short-term one, but it affects all the components of series. There
are no statistical techniques for measuring or isolating erratic fluctuation. Therefore the residual
that remains after eliminating systematic components is taken as representing irregular
variations.

3.5 Forecasting
Introduction:
A very important use of time series data is towards forecasting the likely value of variable in
future. In most cases it is the projection of trend fitted into the values regarding a variable over a
sufficiently long period by any of the methods discussed latter. Adjustments for seasonal and
cyclical character introduce further improvement in the forecasts based on the simple projection
of the trend. The importance of forecasting in business and economic fields lies on account of its
role in planning and evaluation. If suitably interpreted, after consideration of other forces, say
political, social governmental policies etc., this statistical technique can be of immense help in
decision making.
The success of any business depends on its future estimates. On the basis of these estimates a
business man plans his production stocks, selling market, arrangement of additional funds etc.
Forecasting is different from predictions and projections. Regression analysis, time series
analysis, Index numbers are some of the techniques through which the predictions and
projections are made. Whereas forecasting is a method of foretelling the course of business
activity based on the analysis of past and present data mixed with the consideration of ensuring
economic policies and circumstances. In particular forecasting means fore-warning. Forecasts
based on statistical analysis are much reliable than a guess work.

Methods of Business forecasting:

There are three methods of forecasting
1. Naive method
2. Barometric methods
 Quantitative Techniques in Management
3. Analytical Methods

1. Naive method: It contains only the economic rhythm theory.

2. Barometric methods: It covers

i) Specific historical analogy
ii) Lead- Lag relationship
iii) Diffusion method
iv) Action –reaction theory

3. Analytical Methods: It contains

i) The factor listing method
ii) Cross-cut analysis theory
iii) Exponential smoothing
iv) Econometric methods

The economic rhythm theory:

In this method the manufactures analysis the time-series data of his own firm and forecasts on
the basis of projections so obtained. This method is applicable only for the individual firm for
Which the data are analyzed, the forecasts under this method are not very reliable as no
subjective matters are being considered.

Diffusion method of Business forecasting

The diffusion index method is based on the principle that different factors, affecting business, do
not attain their peaks and troughs simultaneously. There is always time-log between them. This
method has the convenience that one has not to identify which series has a lead and which has a
lag. The diffusion index depicts the movement of broad group of series as a whole without
bothering about the individual series. The diffusion index shows the percentage of a given set of
series as expanding in a time period. It should be carefully noted that the peaks and troughs of
diffusion index are not the peaks troughs of the business cycles. All series do not expand or
contract concurrently. Hence if more than 50% are expanding at a given time, it is taken that the
business is in the process of booming and vice - versa.
The graphic method is usually employed to work out the diffusion index. The diffusion index can
be constructed for a group of business variables like prices, investments, profits etc.

Cross cut analysis theory of Business forecasting:

In this method a thorough analysis of all the factors under present situations has to be done and
an estimate of the composite effect of all the factors is being made. This method takes into
account the views of managerial staff, economists, consumers etc. prior to the forecasting. The
forecast about the future state of the business is made on the basis of overall assessment of the
effect of all the factors.
 Quantitative Techniques in Management
Chapter-3
Forecasting Techniques
End Chapter quizzes
Answer the following questions-

1. is a relative measure of association between two or more variables

(a). coefficient of correlation
(b). coefficient of regression
(c). both
((d). none of these

2. Correlation coefficient lies between

(a) -1 to +1
(b) 0 and +1
(c) -1 and 0
(d) None of these

3. R is independent of
(a). choice of origin and not of choice of scale
(b). choice of scale and not of choice of origin
(c). both choice of origin and choice of scale
(d). none of these

4. Correlation between temperature and sale of garments

(a). Positive
(b). 0
(c). negative
(d). none of these

5. Covariance can vary from

(a). -1 to+1
(b) –infinity to + infinity
(c). -1 to 0
(d). 0 to +1

6. Karl Pearson‘s coefficient is defined from

(a). Ungrouped data
(b). Grouped data
(c). Both
(d). None

7. If the value of r2 for a particular situation is 0.49. What is the coefficient of correlation
(a). 0.49
(b). 0.7
(c). 0.07
(d). Cannot be determined
 Quantitative Techniques in Management

8. If r=0.6, then the coefficient for non determination is-

(a).0.4
(b).-0.6
(c).0.36
(d).0.64

9. Coefficient of determination is-

(a). r3
(b). r2
(c). 1-r2
(d) 1+r2

10. When r=0 then covariance of (x,y) is equal to

(a). +1
(b). -1
(c). 0
(d). none of these