Chapter – 12

Forecasting techniques
Forecasts in budgeting
Budgets are based on forecasts. Forecasts might be prepared for:

 the volume of output and sales

 sales revenue (sales volume and sales prices) costs.

The purpose of forecasting in the budgeting process is to establish realistic assumptions

for planning.

A forecast might be based on simple assumptions, such as a prediction of a 5% growth

in sales volume or sales revenue

On the other hand, forecasts might be prepared using a number of forecasting models,
methods or techniques that look to calculate trends and variations over previous years.
The reason for using these models and techniques is that they might provide more
reliable forecasts.
Possible forecasting techniques:
 the high-low method
 linear regression analysis
 time series analysis
 index numbers.

Regression analysis
Regression analysis is concerned with establishing the relationship between a number
of variables. We are only concerned here with linear relationships between 2 variables.

There are a variety of methods available for identifying the relationship:

1. Draw a scatter diagram and plot a line of best fit (see Chapter 3) The data is
plotted on a graph. The y-axis represents the dependent variable, i.e. the
variable that depends on the other. The x-axis shows the independent variable,
i.e. the variable which is not affected by the other variable.
From the scatter diagram, the line of best fit can be estimated. The aim is to use
our judgement to draw a line through the middle of data with the same slope as
the data.
2. The high-low method (see Chapter 4)
3. Least squares regression analysis

110
Least squares regression analysis
Regression analysis finds the line of best fit computationally rather than by estimating
the line on a scatter diagram. It seeks to minimise the distance between each point and
the regression line.
The equation of a straight line is:
y = a + bx
where y = dependent variable
a = intercept (on y-axis)
b = gradient
x = independent variable

𝑛 ∑ 𝑥𝑦 − ∑ 𝑥 ∑ 𝑦
𝑏=
𝑛 ∑ 𝑥 − (∑ 𝑥)

∑ ∑
𝑎= -

n = number of pairs of data

Illustration

111
What does y=300 + 5x explains ?

Mathematical interpretation
If x = 0, then y = 300 and then each time x increases by 1, y increases by 5
Business interpretation
If no money is spent on advertising then sales would still be $300, then for every
additional $1 increase in advertising, sales revenue would increase by $5

Linear regression in budgeting

Linear regression in budgeting

Linear regression analysis can be used to make forecasts or estimates whenever a
linear relationship is assumed between two variables, and historical data is available for
analysis.
The regression equation can be used for predicting values of y from a given x value.
1 If the value of x is within the range of our original data, the prediction is known as
interpolation.
2 If the value of x is outside the range of our original data, the prediction is known as
extrapolation.
In general, interpolation is much safer than extrapolation

Linear regression can also be used:

112
to establish a trend line from a time series. Time series is explained later in this
chapter.
– The independent variable (x) in a time series is time.
– The dependent variable (y) is sales, production volume or cost etc.
as an alternative to using the high-low method in cost behaviour analysis. It should be
more accurate than the high-low method, because it is based on more items of historical
data, not just a ‘high’ and a ‘low’ value.

When a linear relationship is identified and quantified using linear regression

analysis, values for a and b are obtained, and these can be used to make a
forecast for the budget. For example:
a sales budget or forecast can be prepared
costs can be estimated, for a budgeted level of activity.

Q.1)

113
Benefits of simple linear regression
1 Simple and easy to use.
2 Looks at the basic relationship between two sets of data.
3 Can be used to forecast and to produce budgets.
4 Information required to complete the linear regression calculations should be readily
available.
5 Computer spreadsheet programmes often have a function that will calculate the
relationship between two sets of data.
6 Simplifies the budgeting process.
Limitations of simple linear regression
1 Assumes a linear relationship between the variables.
2 Only measures the relationship between two variables. In reality the dependent
variable is affected by many independent variables.
3 Only interpolated forecasts tend to be reliable. The equation should not beused for
extrapolation.
4 Regression assumes that the historical behaviour of the data continues into the
foreseeable future.
5 Interpolated predictions are only reliable if there is a significant correlation between
the data

Correlation
Regression analysis attempts to find the straight line relationship between two variables.
Correlation is concerned with establishing how strong the straight line relationship is.

Positive and negative correlation

Correlation can be positive or negative.
Positive correlation means that high values of one variable are associated with high
values of the other and that low values of one are associated with low values of the
other.

114
Negative correlation means that low values of one variable are associated with high
values of the other and vice versa.

Degrees of correlation
Two variables might be:
(a) perfectly correlated ( if r = +1 or -1)
(b) partly correlated (if r <+1>-1, ≠ 0)
(c) uncorrelated (if r = 0)

The correlation coefficient

The degree of correlation can be measured by the Pearsonian correlation coefficient, r
(also known as the product moment correlation coefficient).
r must always be between –1 and +1.
If r = +1, there is perfect positive correlation
If r = 0, there is no correlation
If r = –1, there is perfect negative correlation
For other values of r, the meaning is not so clear. It is generally taken that if r > 0.8, then
there is strong positive correlation and if r < – 0.8, there is strong negative correlation.

115
Illustration

Q.2)

Q.3)

116
Q.4)

The coefficient of determination (r^2)

The coefficient of determination, r2 measures the proportion of changes in y that can be
explained by changes in x when a straight line relationship has been established.

Time series analysis

A time series is a series of figures recorded over time, e.g. unemployment over the last
5 years, output over the last 12 months.

Time series analysis is a technique used to:

identify whether there is any underlying historical trend
use this analysis of the historical trend to forecast the trend into the future
identify whether there are any seasonal variations around the trend
apply estimated seasonal variations to a trend line forecast in order to prepare a
forecast season by season.
A time series has 4 components:
Trend
Seasonal variations
Cyclical variations
Residual or random variations.

117
The trend
Most time series follow some sort of long term movement. In time series analysis the
trend is measured by:
1. Inspection.
A graph of the data is produced and the trend line is drawn by eye with the aim of
plotting the line so that it lies in the middle of the data points.

2. Least squares regression analysis.

x represents time (each month would be given a number e.g. January =1, February =2
etc ) and y is the data.

3. Moving averages.
This method attempts to remove seasonal or cyclical variations by a process of
averaging.

Seasonal variations
Once the trend has been found, the seasonal variation can be determined. Seasonal
variations are short-term fluctuations in value due to different circumstances which
occur at different times of the year, on different days of the week, different times of day,
for example traffic is greatest in the morning and evening rush hours.

Seasonal variations are used to forecast future figures by amending the trend.
There are two main models:

The additive model. Here the seasonal variation is expressed as an absolute amount
to be added on to the trend to find the actual result, e.g. ice cream sales in summer are
expected to be $200,000 above the trend.

𝑭𝒐𝒓𝒆𝒄𝒂𝒔𝒕 = 𝑻𝒓𝒆𝒏𝒅 + 𝑺𝒆𝒂𝒔𝒐𝒏𝒂𝒍 𝒗𝒂𝒓𝒊𝒂𝒕𝒊𝒐𝒏

118
The multiplicative model. Here the seasonal variation is expressed as
ratio/proportion/percentage to be multiplied by the trend to arrive at the actual figure,
e.g. ice cream sales are expected to be 50% more than the trend.

𝑭𝒐𝒓𝒆𝒄𝒂𝒔𝒕 = 𝑻𝒓𝒆𝒏𝒅 × 𝑺𝒆𝒂𝒔𝒐𝒏𝒂𝒍 𝒗𝒂𝒓𝒊𝒂𝒕𝒊𝒐𝒏

119
Cyclical variations
Cyclical variations are medium-term to long term influences usually associated with
the economy.

Residual or random variations

Residual or random variations are caused by irregular items, which cannot be predicted,
such as a fire or flood.

Moving averages
Calculating a moving average

A moving average is a series of averages calculated from historical time series data.

By using moving averages, the variations in a time series can be eliminated leaving a
‘smoothed’ set of figures which is taken as the trend

120
Illustration

121
–

122
Illustration – Regression in time series

Q.6)

123
 Q.7)

Advantages and disadvantages of time series analysis

The advantages of forecasting using time series analysis are that:
forecasts are based on clearly-understood assumptions
trend lines can be reviewed after each successive time period, when the most recent
historical data is added to the analysis; consequently, the reliability of the forecasts can
be assessed
forecasting accuracy can possibly be improved with experience.
The disadvantages of forecasting with time series analysis are that:
there is an assumption that what has happened in the past is a reliable
guide to the future
there is an assumption that a straight-line trend exists
there is an assumption that seasonal variations are constant, either in
actual values using the additive model (such as dollars of sales) or as a
proportion of the trend line value using the multiplicative model.

The product life cycle and forecasting

The product life cycle (seen in Chapter 11) can also be used during the forecasting
procedure

Index numbers

An index number is a technique for comparing, over time, changes in some feature of a
group of items (e.g. price, quantity consumed, etc) by expressing the property each year
as a percentage of some earlier year.

The year that is used as the initial year for comparison is known as the base year. The
base year for an index should be chosen with some care. As far as possible it should be

124
a ‘typical year’ therefore being as free as possible from abnormal occurrences. The
base year should also be fairly recent and revised on a regular basis.

Illustration

125
Types of index numbers
Index numbers are used in a variety of situations and to measure changes in all sorts of
items. As the uses of index numbers are so diverse a number of different types of
indices have been developed.
We shall deal below with the following:
simple indices
chain based indices
multi-item (or weighted) indices.

[Link] index numbers

A simple index is one that measures the changes in either price or quantity
of a single item

There are therefore two types of simple indices:

a price index
a quantity index.

The formulae for calculating simple indices are:

𝑃1
𝑆𝑖𝑚𝑝𝑙𝑒 𝑝𝑟𝑖𝑐𝑒 𝑖𝑛𝑑𝑒𝑥 = ∗ 100
𝑃0
𝑄1
𝑆𝑖𝑚𝑝𝑙𝑒 𝑞𝑢𝑎𝑛𝑡𝑖𝑡𝑦 𝑖𝑛𝑑𝑒𝑥 = ∗ 100
𝑄𝑜

Where:
P0 is the price at time 0
P1 is the price at time 1
Q0 is the quantity at time 0
Q1 is the quantity at time 1

Illustration

126
[Link] base index numbers

A chain base index number expresses each year’s value as a percentage of the value
for the previous year

Illustration

[Link]-item (weighted) index numbers

A weighted index measures the change in overall price or overall quantity of a number
of different items compared to the base year.

For example, an organisation might produce three different products and an index is to
be constructed to measure the selling price changes of all three products. In order to do
this the percentage change in each of the three selling price must first be calculated

127
individually and the results must then be weighted to reflect the relative importance of
each of the three products.
For a price index:
Step 1 Calculate the simple price index for each of the items.
Step 2 These price indices must then be weighted in some suitable manner in order to
produce an overall price index.
Similarly if a quantity index is to be calculated:
Step 1 Calculate the simple quantity index for each of the items.
Step 2 These quantity indices must then be weighted in some suitable manner in order
to produce an overall quantity index.

Illustration

128
 Q.8)

9 Advantages and disadvantages of index numbers

Advantages of index numbers
They aid the management understanding of information presented to them.
Indices present changes in data or information over time in percentage terms.

Index numbers and forecasting

The accuracy of forecasting is affected by the need to adjust historical data and future
forecasts to allow for price or cost inflation.
When historical data is used to calculate a trend line or line of best fit, it should ideally
be adjusted to the same index level for prices or costs. If the actual cost or revenue data
is used, without adjustments for inflation, the resulting line of best fit will include the
inflationary differences.
When a forecast is made from a line of best fit, an adjustment to the forecast should
be made for anticipated inflation in the forecast period.

Illustration

129
Q.9)

130
Q.10)

131