Financial Planning using Excel

Forcasting Planning and Budgeting Techniques

Sue Nugus

AMSTERDAM • BOSTON • HEIDELBERG • LONDON

NEW YORK • OXFORD • PARIS • SAN DIEGO
SAN FRANCISCO • SINGAPORE • SYDNEY • TOKYO

CIMA Publishing is an imprint of Elsevier

CIMA Publishing
An imprint of Elsevier
Linacre House, Jordan Hill, Oxford OX2 8DP
30 Corporate Drive, Burlington, MA 01803

First published 2005

Copyright © 2006, Sue Nugus. All rights reserved

The right of Sue Nugus to be identified as the author of this work has been asserted
in accordance with the Copyright, Designs and Patents Act 1988

No part of this publication may be reproduced in any material form (including

photocopying or storing in any medium by electronic means and whether
or not transiently or incidentally to some other use of this publication) without
the written permission of the copyright holder except in accordance with the provisions
of the Copyright, Designs and Patents Act 1988 or under the terms of a licence issued by the
Copyright Licensing Agency Ltd, 90 Tottenham Court Road, London, England
W1T 4LP. Applications for the copyright holder’s written permission to reproduce any
part of this publication should be addressed to the publisher

Permissions may be sought directly from Elsevier’s Science and Technology Rights
Department in Oxford, UK: phone: (⫹44) (0) 1865 843830; fax: (⫹44) (0) 1865 853333;
e-mail: permissions@[Link]. You may also complete your request on-line via
the Elsevier homepage ([Link] by selecting ‘Customer Support’ and
then ‘Obtaining Permissions’

British Library Cataloguing in Publication Data

A catalogue record for this book is available from the British Library

Library of Congress Cataloguing in Publication Data

A catalogue record for this book is available from the Library of Congress

ISBN-10 0-7506-6355-3
ISBN-13 978-0-7506-6355-7

For information on all CIMA Publishing Publications

visit our website at [Link]

Typeset by Integra Software Services Pvt. Ltd, Pondicherry, India

[Link]
Printed and bound in Great Britain

Working together to grow

libraries in developing countries
[Link] | [Link] | [Link]
Contents

Preface vii

About the Author ix

Part 1 1

1 Forecasting 3
Introduction 5
Approaches to forecasting 5

2 Collecting and Examining the Data 9

Contents
Data collection 11
Using statistical measures 15
Summary 21

iii
3 Smoothing Techniques 23
Introduction 25
Moving averages 26
Weighted moving average 27
Adaptive filtering 31
Exponential smoothing 32
Summary 35

4 Regression Analysis 37
Introduction 39
Calculating the least square line 43
Function vs command 46
The graphic approach 46
Standard error 47
Multiple linear regression 49
Summary 52

5 Time Series Analysis 53

Introduction 55
Multiplicative time series analysis 56
A time series analysis model 57
Summary 62
 6 Expected Values 63
Introduction 65
A model for analysing expected values 65
Summary 70

7 Selecting and Evaluating Forecasting Techniques 71

Introduction 73
Selecting the right technique 73
Accuracy and reliability 75
Charts 75
Statistical methods 75
Subjective methods 77
Summary 78

Part 2
Contents

79

8 Business Planning 81
Introduction 83
iv Approaches to business planning 84
Summary 90

9 Spreadsheet Skills for all Types of Planning 91

Introduction 93
Spreadsheet 1: Getting started 94
Spreadsheet 2: Ownership and version 95
Spreadsheet 3: Formatting 96
Spreadsheet 4: Documentation 99
Spreadsheet 5: Minimising absolute values 100
Spreadsheet 6: Separating growth and cost factors 101
Spreadsheet 7: Optimizing layout 103
Spreadsheet 8: Arithmetic cross-checks 105
Spreadsheet 9: Charts 106
Spreadsheet 10: Mutiple sheets 106
Templates 108
Data input forms 109
Summary 110

10 Developing a Financial Plan 113

Introduction 115
The model 115
Getting started 116
 Developing the Profit and Loss Account 117
Developing the Profit and Loss Appropriation Account 121
Developing the balance sheet 122
Funds flow statement 124
Ratio analysis 125
Cash flow statement 125
Summary 126

11 Business Plans 129

Introduction 131
Capital investment appraisal 131
Developing a capital investment appraisal plan 133
Learning curve costing 139
Break-even analysis 142
Economic order quantities (EOQ) 144

Contents
Sales campaign appraisal 147
Summary 149

12 What-if Analysis 151

v
Introduction 153
Three approaches to what-if analysis 154
Manual what-if analysis on opening assumptions 154
Data tables 156
Backward iteration or goal seeking 161
Summary 165

13 Risk Analysis 167

Introduction 169
Preparing a plan for risk analysis 170
Incorporating the RAND function 172
The results worksheet 173
Frequency distribution 175
Using the risk analysis model 176
Summary 177

Part 3 179

14 Budgeting 181
Introduction 183
Scope of budgeting 183
Benefits of budgeting 184
Different approaches to budgeting 184
 Budget preparation 185
Spreadsheets for budgets 186
Summary 186

15 A Spreadsheet Budgeting System 187

Introduction 189
Preparing the budget template 190
Preparing the actual template 191
Preparing the variance report template 192
Preparing the year-to-date report template 193
Summary 195

16 Consolidating Data 197

Introduction 199
Using the Consolidate command 199
Contents

Summary 202

Index 203

vi
Preface
The objective of this book is to help financial planners improve
their spreadsheet skills by providing a structured approach to
developing spreadsheets for forecasting, financial planning and
budgeting.

The book assumes that the reader is familiar with the basic opera-
tion of Excel and is not intended for beginners.

Readers using a different Windows spreadsheet will find that the

techniques explained in the book are equally relevant, although it is
possible that some command sequences might be slightly different.

The book has been divided into three parts covering the areas of
Forecasting, Planning and Budgeting separately. Although it is rec-

Contents
ommended that readers follow the book from the beginning, the
text is also intended as a reference book that will be a valuable aid
during model development.
vii
The CD-ROM that accompanies the book contains all the examples
described. Instructions for installing and using the CD-ROM are
supplied on the CD itself and it is recommended that readers con-
sult the README file contained on the CD.
This Page Intentionally Left Blank
About the Author
Sue Nugus has been conducting seminars and workshops for account-
ants and other executives for nearly 20 years. She has worked with
the Chartered Institute of Management Accountants and the Institute
of Chartered Accountants in England and Wales, and also with the
equivalent institutes in Ireland and Scotland.

These seminars and workshops have mostly involved helping

accountants and financial managers get the most from their
spreadsheets.

The course on which this book is based runs for Management and

About the Author

Chartered Accountants and other executives at least 12 times a year.

In addition to her teaching she has authored and co-authored some

20 books on a wide range of IT subjects that have been published
by McGraw-Hill, NCC-Blackwell, and Butterworth-Heinemann.

Sue Nugus also offers consultancy services to those who need ix

assistance in developing advanced spreadsheets.

sue@[Link]
This Page Intentionally Left Blank
Part 1
This Page Intentionally Left Blank
1
Forecasting
This Page Intentionally Left Blank
 I never think of the future; it comes soon enough.

– Boyadjian and Warren, RISKS, Reading Corporate Signals, 1987.

Introduction
The objective of a business forecast is to predict or estimate a future
activity level such as demand, sales volume, asset requirements,
inventory turnover, etc. A forecast is dependent on the analysis of
historic and/or current data to produce these estimates. Having

Financial Planning using Excel

accurate forecasts can play an important role in helping an organi-
sation to operate in an efficient and effective manner.

However, before being in a position to create a forecast it is necessary

to look carefully at what has happened in the past. As well as examin-
ing historic data it is also important to be aware of the organisation’s
position in its industry and the industry’s position in the global mar-
ketplace. This is equally true for not-for-profit organisations, which
are likely to be more interested in budgeting costs as opposed to profit. 5

Approaches to forecasting
The process of forecasting can be broadly categorised into two
approaches: objective or quantitative forecasts and subjective or
qualitative forecasts.

Subjective forecasts
Subjective or qualitative forecasts rely to a large extent on an in-
depth knowledge of the activity being forecast by those responsible
for producing the forecast. The forecast might be created by reading
reports and by consulting experts for information and then using
this information in a relatively unspecified or unstructured way to
predict a required activity. A forecasting method discussed in
Chapter 10, called the composite of individual estimations, is based
on essentially subjective information. The main problem with this
approach is that there is no clear methodology which can be
analysed to test how a forecast may be improved in order that past
mistakes are avoided. As a subjective, or qualitative, forecast is very
dependent on the individuals involved, it is prone to problems
 when the key players responsible for the forecasting process
change. This method of forecasting does not usually require much
mathematical input and therefore a spreadsheet will play an
accompanying role as opposed to a central role.

Objective forecasts
An objective or quantitative approach to forecasting requires a
model to be developed which represents the relationships deduced
from the observation of one or more different numeric variables.
Financial Planning using Excel

This is generally achieved by first recording historic data and then

using these historical facts to hypothesise a relationship between
the items to be forecast and the factors believed to be affecting it.
The spreadsheet is clearly an ideal tool for this type of analysis and
thus can play a central role in the production of such forecasts.

Objective forecasting methods are sometimes considered to be more

dependable than subjective methods because they are less affected
6 by what the forecasters would like the result to be. Furthermore,
forecasting models can incorporate means of assessing the accuracy
of the forecast by comparing what actually happened with what
was forecast and adjusting the data to produce more accurate
figures in the future. Most of the forecasting examples in the book
would be described as objective or quantitative forecasts.

Of course, it is important to appreciate that there has to be an ele-

ment of subjectivity in all forecasting techniques. At the end of the
day what the forecasters know about the business will affect the
choice of a particular forecasting technique, and subsequently an in-
depth knowledge of the activity being forecast is likely to affect how
the forecast data is used to predict activity within the organisation.

Time
Whether a forecast is largely subjective or objective, one of the
more common features of a forecast is time, i.e. how far into the
future is a forecast designed to look. In this case there can be short-
term forecasts, medium-term forecasts and long-term forecasts. The
time-span a forecast is considered to fall in will depend on the cir-
cumstances and the type of industry involved. In general business
terms, short-term forecasts would involve periods of up to one year,
 medium-term forecasts would consider periods of between one and
five years and long-term forecasts would be for longer periods.
There are several examples of time-based forecasts in this book,
including the adaptive filtering model and the multiplicative time
series model, discussed in Chapter 10.

Forecast units
Whether forecasts are categorised in terms of time or level of objec-
tivity, the forecast unit is also an important variable. For example, a

Financial Planning using Excel

forecast might seek to estimate the level of sales, either as sales
units or as sales revenue; or a forecast might seek to establish a
level of probability, such as a service level of 99%. It might be
appropriate to forecast activity levels such as the numbers of cus-
tomer service enquiries that are expected between 10 and 11am. In
a not-for-profit situation the forecast might be concerned with the
expenditure on staff over the forthcoming period.

Finally, any forecast must also be seen in terms of whether it is a 7

one-off estimation or a repetitive calculation. One-off forecasts are
normally concerned with large projects and thus may be performed
with the aid of considerable financial resources.

A common requirement of those responsible for the budgeting func-

tion in an organisation is the need to create ongoing forecasts where
there is a need for continuous adjustments to previously forecast fig-
ures. These forecasts need to be developed in such a way that actual
data can be entered into the model in order that a comparison can
be made between the forecast and the actual data. The accuracy of
the forecast can then be assessed and adjustments can be made in
order to attempt to make the next forecast more accurate.

Forecasting and Excel

As mentioned above, the spreadsheet has a valuable role to play
in a range of different forecasting activities, although clearly the
objective or quantitative approach particularly lends itself to the
numeric analysis tools offered by the spreadsheet. Indeed, in
Excel today, as well as the ability to build formulae by referenc-
ing data that has been entered into the spreadsheet cells, there
are a large number of built-in functions that make the task even
 easier and open the door to the ability to perform a wider range
of analyses.

The problem of course is first discovering just what is available and

then learning how to apply the tool to the task in hand. It is hoped
that by working through the exercises in this book the reader will
be better placed to produce useful Excel-based forecasts.
Financial Planning using Excel

8
2
Collecting and Examining the Data
This Page Intentionally Left Blank
 Without systematic measurements, managers have little to guide
their actions other than their own experience and judgment. Of
course, these will always be important; but as businesses becomes
more complex and global in their scope, it becomes increasingly
more difficult to rely on intuition alone.

– J. Singleton, E. McLean and E. Altman, MIS Quarterly, June 1988.

Data collection

Financial Planning using Excel

Before selecting a forecasting technique it is necessary to have an
appropriate amount of data on which to base the forecast. The
quantity and type of data that constitutes ‘appropriate’ will vary
depending on the activity that is to be forecast. At some point, how-
ever, historic data will no longer be relevant and it is important for
those involved with the forecasting to agree on what constitutes
usable data from the outset.

The periodicity of the data is also important. In general terms the

11
input data, i.e. the historic data, should be entered into the spread-
sheet using the same periodicity as the output, i.e. the forecast. The
main reason for this is that if data is entered into the spreadsheet as
monthly figures, for a quarterly forecast, then a degree of calculation is
required at this base level. This can lead to an opportunity for GIGO.

The acronym GIGO is a long-standing computer term which generally

stands for Garbage In Garbage Out – implying that if you are not care-
ful with the information you put into a computer you will only get
rubbish out. In the spreadsheet environment there can be a slightly
different definition, which is, Garbage In Gospel Out – because it is OPPORTUNITY
FOR GIGO
not difficult for the spreadsheet to look right, even though the contents
may actually be rubbish! I will therefore point out opportunities for
GIGO throughout the book with accompanying ways of avoiding them.

In the case of the periodicity of data, if the base data or the end
result has to be divided by four in order to convert it from monthly
to quarterly figures, this requires someone to remember to always
do this and to ensure that all updates and amendments that are
made to a plan have been adjusted. It is this type of activity that
can lead to errors being deeply embedded into spreadsheet systems.

Of course, it is not always possible to have the input and the output
data in the same periodicity and if this is the case, it is important to
 make it clear on the spreadsheet that a change is being made. There
is a further discussion on documenting a spreadsheet in Chapter 9.

Examining the data

Once entered into the spreadsheet, the historic data should be exam-
ined to ascertain the presence of any obvious patterns. For example,
is there evidence of trend, seasonality or business cycle? The quan-
tity of data will affect the types of patterns to be sought. For exam-
ple, in order to establish the presence of seasonality a sufficient
Financial Planning using Excel

number of periods of data must be available, and business cycles

can be considered only by looking at a large number of periods.

First draw a graph

The data shown in the following examples represents historic sales
data from which forecasts are to be produced and can be found on
[Link] the file named [Link] on the CD accompanying the book.
12
The periodicity of the data in each of the examples is monthly, but
the number of periods differs in each example. In order to simplify
the examination of the data, line graphs have been produced. This
is a good example of using simple graphs to look at spreadsheet
data which immediately highlights the presence or absence of
patterns in the data that would otherwise require mathematical
analysis of a set of numbers.

No trend or seasonality
The first data set in Figure 2.1 shows 24 months of historic income
values. By looking at the chart it is clear that there is no strong
trend, no apparent seasonality and the number of periods is too few
to be able to perceive a business cycle. Based on these observations
the next period is as likely to increase, decrease or remain the same.

Some evidence of trend

Figure 2.2 shows another set of 24 months of historic income.
From the chart it can be seen that looking across the 24 periods the
income is increasing, although there are fluctuations in the data.
This would indicate an upward trend. Of course, a trend may not
 Financial Planning using Excel
Figure 2.1 Historic data for 24 monthly periods showing no trend

13

Figure 2.2 Historic data for 24 monthly periods showing some trend

always be favourable and it is important to be able to explain the

reason for any trend; for example growth in the market or the suc-
cess of a marketing plan. On the other hand the data might be rep-
resenting an increase in costs, causing a downward trend.

Seasonality
Looking at the chart in Figure 2.3, which shows three years of
monthly historic income data, in addition to an upward trend,
there is a strong indication of seasonality. There appears to be a
Financial Planning using Excel

Figure 2.3 Historic data for 24 monthly periods showing evidence of seasonality

similar peak in the data between June and September in all three
years, which could indicate a seasonal pattern. To confirm this it is
14
important to refer back to the activity being forecast to ensure that
this is indeed the case.

Business cycle
The last set of data to be examined here is shown in Figure 2.4
and consists of quarterly data for the number of conference

Figure 2.4 Quarterly data for 20 years indicating a business cycle

 participants over the past 20 year period. From the chart it can
be seen that there appears to be a five year cyclical pattern to
the data. As a business cycle implies a cyclical trend pattern
over a longer period of time, the data in this example suggests
five yearly peaks and troughs in the number of people that
attend conferences, and by looking at the overall business situa-
tion at this time this may correspond with periods of growth and
recession.

Financial Planning using Excel

Using statistical measures
Although charts are a useful way of obtaining an overall view of the
movement of data, it is also important to be able to describe or sum-
marise the data using statistical measures.

Descriptive statistics
The descriptive statistics included here are the mean, the mode, the 15

median, the standard deviation, the variance and the range.

Figure 2.5 shows the number of emails that are sent each month [Link]
by day. This is the data from which the descriptive statistics are
measured.

Figure 2.5 Number of emails sent each month by day

 The mean, median and mode are described as measures of
central tendency and offer different ways of presenting a typical or
representative value of a data set. The range, the standard devia-
tion and the variance are measures of dispersion and refer to the
degree to which the observations in a given data set are spread
about the arithmetic mean. The mean, often together with the stan-
dard deviation, are the most frequently used measures of central
tendency. Excel has a series of built-in functions that can be used
to produce descriptive statistics.
Financial Planning using Excel

TECHNIQUE
In the DESCRIP worksheet the area B4:F15 has been named DATA.
TIP! Any rectangular range of cells can be assigned a name in Excel,
which has the benefit of offering a description of a range of cells
and also can make the referencing of the range easier. To name a
range first select the area to be named and then type the chosen
name into the Name box, which is located to the left of the edit line
at the top of the screen.

Figure 2.6a shows the result of the descriptive statistic functions

16
which can be found on Sheet B of the file DESCRIP. In Figure 2.6b
the spreadsheet has been set to display the contents on the cells
in order that the reader can look at the functions and formulae
that have been used. This is achieved by holding down the
TECHNIQUE
CTRL key and then pressing ` key (usually this key also has ¬
TIP!
and symbols on it).

(a) (b)

Figure 2.6 Results of descriptive statistics

Number of observations
The number of observations can be counted through the use of the
⫽COUNT function. This function has a number of variations:

⫽COUNT( . . . ) counts cells containing numbers and numbers

entered into the list of arguments.
⫽COUNTA( . . . ) counts all non-blank cells and numbers
entered into the list of arguments.
⫽COUNTBLANK( . . . ) counts blank cells in the list of arguments.
⫽COUNTIF( . . . )

Financial Planning using Excel

counts cells in the list of arguments that
satisfy a specified criteria.

Mean

The mean or arithmetic mean is defined as follows:

total of a number of sample values

sample mean ⫽
the number of sample values 17

To calculate the sample arithmetic mean of the production weights

the AVERAGE function is used as follows in cell B4.

⫽AVERAGE(DATA)

It is important to note that the AVERAGE function totals the cells OPPORTUNITY
containing values and divides by the number of cells that contain FOR GIGO
values. In certain situations this may not produce the required
results and it might be necessary to ensure that zero has been
entered into blank cells in order that the function sees the cell as
containing a value.

Sample median

The sample median is defined as the middle value when the data
values are ranked in increasing, or decreasing, order of magnitude.
The following formula in cell B5 uses the MEDIAN function to calcu-
late the median value for the production weights:

⫽MEDIAN(DATA)
 Sample mode

The sample mode is defined as the value in an argument which

occurs most frequently. The following is required to calculate the
mode of the production weights.

⫽MODE(DATA)

The mode may not be unique, as there can be multiple values that
return an equal, but most frequently occurring value. In this case the
mode function returns the first value in the argument that occurs
Financial Planning using Excel

most frequently. Furthermore, if every value in the sample data set

is different, there is no mode and the function will return an N/A
result.

Minimum and maximum

It is often useful to know the smallest and the largest value in a

data series and the MINIMUM and MAXIMUM functions have been
18
used in cells B7 and B8 to calculate this as follows:

⫽MIN(DATA)

⫽MAX(DATA)

By including a value within the argument for the MIN and MAX
functions it is possible to ensure that the value in a cell is within
TECHNIQUE
specified boundaries. For example, to return the lowest value in a
TIP!
range, but to ensure that the result was never higher than 500, the
following could be used:

⫽MIN(B4:B16, 500)

In this instance the system will look at the values in the range
B4:B16 and will also look at the value 500 and if 500 is the lowest
value in the range this will be the result.

The range

The range is defined as the difference between the largest and

smallest values in a data series. The following formula can be used
to calculate the range of the number of emails sent by referencing
the already calculated minimum and maximum values:

⫽B8–B7
Sample standard deviation

The sample standard deviation s is obtained by summing the squares

of the differences between each value and the sample mean, dividing
by n ⫺ 1, and then taking the square root. Therefore the algebraic for-
mula for the sample standard deviation is:

√
(兺x)2
兺x2 ⫺
n
s⫽
n⫺1

Financial Planning using Excel

In other words the standard deviation is the square root of the vari-
ance of all individual values from the mean. The more variation in
the data, the higher the standard deviation will be. If there is no
variation at all, the standard deviation will be zero. It can never be
negative.

To calculate the standard deviation for the number of emails sent

the following can be entered into cell B10:
19
⫽STDEV(DATA)

Note that this function assumes a sample population. If the data

represents the entire population then STDEVP() should be used.

Sample variance

The sample variance is the square of the standard deviation. The

formula required to calculate the sample variance of the number of
emails sent in cell B11 is:

⫽VAR(DATA)

In the same way as the standard deviation, the above function

assumes that sample data is being used. For the entire population,
the function VARP() is required.

Data Analysis Command

A quick way of producing a set of descriptive statistics on a range
of data is to use the TOOLS : DATA ANALYSIS : DESCRIPTIVE STATISTICS
command. Figure 2.7 shows the result of this on the number of
emails sent during January.
Financial Planning using Excel

Figure 2.7 Results of descriptive statistics command

20

In addition to the statistics already described this command provides

the standard error, the kurtosis and the skewness.

The standard error of a sample of sample size n is the sample’s stan-

dard deviation divided by √n. It therefore estimates the standard
deviation of the sample mean based on the population mean (Press
et al. 1992, p. 465)1.

The kurtosis can be defined as the degree of peakedness of a

distribution.

Skewness is a measure of the degree of asymmetry of a distribution. If

the left tail (tail at small end of the distribution) is more pronounced
that the right tail (tail at the large end of the distribution), the function
is said to have negative skewness. If the reverse is true, it has positive
skew.2

1 Press, W.H.; Flannery, B.P.; Teukolsky, S.A.; and Vetterling, W.T. Numerical Recipes in
FORTRAN: The Art of Scientific Computing, 2nd ed. Cambridge, England: Cambridge
University Press, 1992.
2 [Link]
Summary
The first step in preparing any forecasting system is to carefully
examine the available data in order to ascertain the presence of a
trend, a seasonal pattern or a business cycle. Simply looking as the
data, or eyeballing it as it is sometimes described, can be good enough
for this purpose. It is then useful to prepare a set of descriptive statis-
tics such as those described in this chapter in order to further under-
stand the situation described by the data. These statistics may be used
as a first step in embarking on an appropriate forecasting technique.

Financial Planning using Excel

21
This Page Intentionally Left Blank
3
Smoothing Techniques
This Page Intentionally Left Blank
 Good judgment is usually the result of experience. And experience
is frequently the result of bad judgment.

– R.E. Neustadt and E.R. May, Thinking in Time, 1986.

Introduction
Smoothing refers to looking at the underlying pattern of a set of data
to establish an estimate of future values. Smoothing can be achieved
through a range of different techniques, including the use of the

Financial Planning using Excel

AVERAGE function and the exponential smoothing formula. To be able
to use any smoothing technique a series of historic data is required.

Estimating a single value for the next period is called univariate

analysis and there are a number of techniques that can be used to
produce the forecast value. These include:

◆ estimation of the value

◆ using the last known value
◆ calculating the average or arithmetic mean of the historic data. 25

Estimation of the value is a subjective approach that depends

entirely on the forecaster knowing the activity being forecast and
being able to judge the outcome for the next period. In some cases
this is referred to as guesstimation.

In a situation where the examination of the historic data has shown

no evidence of a trend then using the last known value can be an
appropriate method of estimation.

Using an average or arithmetic mean produces a value that is typi-

cal or representative of a given set of data. The algebraic formula for
the arithmetic mean is:

Ft ⫽ (1/N) 兺Xt

where Ft ⫽ the forecast, N ⫽ number of observations and Xt ⫽ his-

toric observations. The arithmetic mean or the simple average of a
data set produces a straight line through the data. Although this is
not especially useful as a forecasting tool, having the mean of a
series of historic values is important for comparison purposes.

Figure 3.1 takes the first set of data examined in Chapter 2 and
[Link]
gives examples of an estimation, the last observation and the arith-
metic mean or average.
 The AVERAGE function has been used to calculate the arithmetic mean.

All three of the techniques shown in Figure 3.1 are regularly used
in situations where there is little or no trend to influence the selec-
tion of a different forecasting approach.

The graph in Figure 3.1 shows the arithmetic mean plotted against
the historic data. From this it can be seen that the arithmetic mean
of 42,666 is a measure of central tendency, and there are approxi-
mately as many data points above as there are below the line.
Financial Planning using Excel

26

Figure 3.1 Examples of an estimation, last observation and arithmetic mean

Moving averages
In an environment in which the data does not exhibit any significant
trend and when using an average to calculate a future expected
value, the earlier historic observations may have less relevance than
the more recent observations, especially in cases where there is little
evidence of trend or seasonality. In this case a moving average can be
employed which allows early values to be dropped as later values
are added. The algebraic formula for a moving average is as follows:

Ft ⫽ (1/N)兺X t⫺i

where Ft ⫽ the forecast, N ⫽ number of observations, Xt ⫽ historic

observations and i ⫽ change in Xt variables.
 The greater the value of N, the less the forecast will be affected by
random fluctuations, or the greater the degree of smoothing.
Furthermore, the greater the number of observations the slower
the forecast is to respond to changes in the underlying pattern of
the data. Figure 3.2 illustrates a three month moving average and a
five month moving average.

Financial Planning using Excel

27

Figure 3.2 Three and five month moving averages

The three month moving average shown by the green line in

Figure 3.2 reflects changes in the data and is easily influenced by
irregularities and fluctuations. The five month moving average
shown by the blue line, on the other hand, is smoother and shows
less influence of irregularities and fluctuations.

The formula in cell D6 is ⫽AVERAGE(B4:D4). This formula has then

been extrapolated through to M6.

Weighted moving average

In addition to restricting the number of historic observations that
are incorporated into a moving average, it is sometimes necessary
to place more emphasis on some data points than on others. To this
end a weighted moving average technique can be applied to the
 data. There are a number of different approaches to using weighted
moving averages and the proportional and trend adjusted methods
are discussed here.

Proportional method
With the proportional method each value in the moving average is
multiplied by a specified weight, and the total of the weights usu-
ally equals 1. The algebraic formula for this method is as follows:
Financial Planning using Excel

Ft ⫽ P1X1 ⫹ P2X2 ⫹ ⭈ ⭈ ⭈ ⫹ PnXn

where Ft ⫽ the forecast, Xt ⫽ historic observations and

P1 ⫹ P2 ⫹ ⭈ ⭈ ⭈ ⫹ Pn1 ⫽ 1

Figure 3.3 shows the results of using the proportional method for
calculating a three month weighted moving average which can then
be compared to the previously calculated three month moving aver-
age without weights.
28

Figure 3.3 Three month moving average using the proportional method

The effect of the weights that have been used in the above example
is to place a greater emphasis on the most recent historical observa-
tion. In other words the most recent occurrence is most important
when determining the next occurrence. By looking at the actual
value for the forecast period the weights could be changed in an
attempt to produce a more accurate forecast for the next period.

Figure 3.4 shows the formula required for the weighted moving
average and Figure 3.5 shows the unweighted and weighted moving
averages plotted together on the same graph.
TECHNIQUE
Note that the cell references to the proportional weights are absolute,
TIP!
i.e. $B$9, $C$9 and $D$9. This means that when the formula is
 Figure 3.4 Formulae required for weighted moving average

Financial Planning using Excel

29

Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec

Income 3 month moving avg 3 month weighted moving average

Figure 3.5 Chart showing unweighted and weighted three month moving average

copied the reference to cells B9, C9 and D9 remain fixed, whilst the
other cell references are relative.

Trend adjusted method

If, as a result of examining the data, there is evidence of a trend, then
a trend adjusted method of weighting the average can be applied.
This involves assigning greater weights to more recent observations.
There are a number of approaches to applying trend adjusted weights
and the following is an example:

Ft ⫽ 2Xt⫺1 ⫺ Xt⫺2
 In this example the last observation is doubled before the observa-
tion before that is subtracted. This has the effect of using twice the
increase in the value of the observations from period t ⫺ 2 to t ⫺ 1
in the forecast.

The effect of using this technique on the second set of data from
Chapter 2 which showed some evidence of trend can be seen in
Figure 3.6, and Figure 3.7 shows the results graphically.
Financial Planning using Excel

Figure 3.6 Trend adjusted moving average

30

50000

45000

40000

35000

30000

25000
Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec

Trend adjusted moving average Income - 1 year

Figure 3.7 Trend adjusted moving average chart

The formula required in cell C5, which can be extrapolated for the
remaining periods, is:

⫽(C3*2)⫺B3
Adaptive filtering
Adaptive filtering is a technique used to re-evaluate the individual
weights in a weighted average model to take into account the expe-
rience of actual results. Thus adaptive filtering allows a weighted
average technique to learn from past errors and provides a system-
atic approach to adjusting the weights to the latest information
available. The algebraic formula for adaptive filtering is as follows:

W⬘t ⫽ Wt ⫹ 2K[(Xt ⫺ Ft)/X 2]Xt⫺1

Financial Planning using Excel

where W⬘t ⫽ the updated weight, Wt ⫽ the previous weight, Xt ⫽
the observation at time t, K ⫽ a constant, i.e. the training constant
which may not be greater than 1/n, and X ⫽ the largest of the most
recent n values of Xt. Figure 3.8 is a model for adaptive filtering.

31

Figure 3.8 Model for adaptive filtering

The plan consists of a set of historic quarterly sales figures and

cell B10 is the initial forecast for the first quarter of 1997 which is
calculated by multiplying the historic quarterly data by the opening
proportional weights in row 6. Thus the formula for cell B10 is:

⫽B5*B6⫹C5*C6⫹D5*D6⫹E5*E6

When the actual sales for the first quarter become available the
value is entered into cell B11. The adaptive filtering formula in cell
B15, which is extrapolated into cells C15 through E15, calculates
revised weights which are then referenced in cell C10 when the
 forecast for the second quarter is calculated. The following is the
formula entered into cell B15 and copied across to cell E15.

⫽B6⫹2*0.25*(($B$11⫺$B$10)/$B$5^2)*B5

The forecast formula for the second quarter of 1997 in cell C10 is
therefore:

⫽C5*B15⫹D5*C15⫹E5*D15⫹B11*E15

It is not possible to copy the above formula as the references to the

previous quarters are not in the same row and the revised weights
Financial Planning using Excel

for each quarter are also on different rows.

The adjusted weights will no longer total 1 because if the actual

amount is less than the forecast then the combined weights will
end up being less than 1. The converse is true in that if the actual
amount is more than the forecast, the combined weights will total
more than 1.

In order to test the accuracy of the system, enter an actual value

32 TECHNIQUE
that exactly matches the forecast. In this case the weights should
TIP!
not change, as no adjustment is required.

Exponential smoothing
Exponential smoothing is a weighted moving average technique
which is especially effective when frequent re-forecasting is
required, and when the forecasts must be achieved quickly. It is a
short-term forecasting technique that is frequently used in the
production and inventory environment, where only the next
period’s value is required to be forecast. Because only three num-
bers are required to perform exponential smoothing, this tech-
nique is simple to update. The data required are the historic
observation, the latest data observation and the smoothing coeffi-
cient, or constant.

The smoothing coefficient ␣ is a value between 0 and 1. A small

value of, say, between 0.05 and 0.10 results in a high degree of
smoothing and has the same effect as a large number of observations
in a moving average calculation. A high coefficient value results in
less smoothing and thus a high responsiveness to variations in the
 data. In the extreme, if the coefficient is zero then the next period’s
forecast will be the same as the last period’s forecast and if the coef-
ficient is one, or unity, then the next period’s forecast will be the
same as the current period’s data.

The primary assumption used in the simple form of this smoothing

technique is that the data is stationary, i.e. that there is a clear trend
present. Advanced exponential smoothing techniques are required
if a trend or cycle is present in the data.

The algebraic formula for simple exponential smoothing is:

Financial Planning using Excel

Ft ⫽ Xt ⫹ (1 ⫺ ␣)Ft⫺1

where Ft⫺1 ⫽ the previous forecast, Xt ⫽ the current observation

and ␣ ⫽ the smoothing coefficient. Figure 3.9 shows an example of
exponential smoothing.

33

Figure 3.9 Forecast using the exponential smoothing technique

The smoothing coefficient is first used in the second period of the

forecast and so in Figure 3.9 the formula for cell C6 is:

⫽C4*$B$5⫹(1⫺$B$5)*B6

With a low coefficient value of 0.20 a high degree of smoothing is

expected and this is shown in the graph in Figure 3.10. Figure 3.11
shows the effect of a high smoothing coefficient where 0.80 has
been entered into cell B5.

As exponential smoothing does not require a great deal of historic

data, it is another useful tool for short-term forecasts.

The spreadsheet examples used in this chapter do not require the

use of complicated spreadsheet functions and formulae. Indeed
the AVERAGE function has been the main tool. However, the chapter
 130

125

120

115

110

105
JAN FEB MAR APR MAY JUN JUL AUG SEP OCT NOV DEC
Financial Planning using Excel

Sales volume Forecast

Figure 3.10 Forecast with a low coefficient of 0.20

130

125

34
120

115

110

105
JAN FEB MAR APR MAY JUN JUL AUG SEP OCT NOV DEC

Sales volume Forecast

Figure 3.11 Forecast with a high coefficient of 0.80

has illustrated that having ascertained the algebraic formula for a

particular technique it is not difficult to translate this into a for-
mula that the spreadsheet understands. Once the formulae have
been entered and tested the models can be used again and again
with different data.

As with any forecasting technique it is important to always check

what actually happened with the activity that was being forecast
in order to ascertain how accurate the forecast was, and where
necessary be able to adjust the forecast to better reflect the situa-
tion next time.
Summary
There are many different approaches to averaging and smoothing as
a means of forecasting. Moving averages and weighted moving
averages are useful as a first step, especially when there is no clear
evidence of trend, seasonality or cycles. Adaptive filtering is very
useful in a continuous forecasting situation where the actual data is
available to assist in forecasting the next period.

Financial Planning using Excel

35
This Page Intentionally Left Blank
4
Regression Analysis
This Page Intentionally Left Blank
 “One can’t believe impossible things,” said Alice. “I dare say you
haven’t had much practice,” said the Queen. “When I was your
age, I always did it for half-an-hour a day. Why, sometimes I’ve
believed as many as six impossible things before breakfast.”

– Lewis Carroll, Through the Looking Glass, 1982.

Introduction
Many forecasting methods are based on the assumption that the

Financial Planning using Excel

variable being forecast is related to something else. Thus sales
might vary according to the amount spent on advertising, or the life
of a component might be determined by how long it has been held
in the warehouse. Regression analysis is a widely used tool for
analysing the relationship between such variables and using these
variables for prediction purposes. For example, can variations in
the number of marketing emails sent each month be used to help
predict the income to the conference business?
39
To answer this question the first step is to establish whether there is
indeed a relationship between the number of marketing emails sent
and the income. This is achieved by recording the required information
over a period of time and then plotting the data on a scatter diagram.

Figure 4.1 shows a set of data collected monthly for a period of [Link]

one year.

A B C D
1 CWL Emails sent monthly vs income
2 Month No. emails Income
3 Jan 96 15000
4 Feb 106 15670
5 Mar 198 25300
6 Apr 195 21100
7 May 135 18600
8 Jun 44 12000
9 Jul 120 17230
10 Aug 171 20700
11 Sep 180 20990
12 Oct 50 12400
13 Nov 75 13100
14 Dec 21 10200
15 Total 1391 202290

Figure 4.1 Monthly income and number of emails sent

 Before drawing a scatter diagram from the data it should be sorted
by the number of emails sent. This is achieved by first selecting
the range A2:C14 and then using the Data Sort command to sort on
column B in ascending order (Figure 4.2).

A B C D
1 CWL Emails sent monthly vs income
2 Month No. emails Income
3 Dec 21 10200
4 Jun 44 12000
Financial Planning using Excel

5 Oct 50 12400
6 Nov 75 13100
7 Jan 96 15000
8 Feb 106 15670
9 Jul 120 17230
10 May 135 18600
11 Aug 171 20700
12 Sep 180 20990
13 Apr 195 21100
14 Mar 198 25300
15 Total 1391 202290
40

Figure 4.2 Data sorted by number of emails sent

A scatter diagram is produced by selecting the range B3:C14, clicking

on the Graph icon on the toolbar and selecting XY-scatter. From the
sub-choices of XY-graphs take the option with no lines. Figure 4.3 is
the resulting chart.

30000

25000

20000
Income

15000 Income

10000

5000

0
0 50 100 150 200 250
No. of emails

Figure 4.3 Scatter diagram showing relationship between the number of emails
sent and the monthly income
By looking at the chart in Figure 4.3 it is clear that there is a strong lin-
ear relationship between the two variables. Had this not been the case
the points on the graph would be scattered more widely. Figure 4.4
is scatter diagram showing the number of aircraft sold at different
rainfall levels and it is clear from this graph that there is no relation-
ship between these two variables, as one would of course expect.

Aircraft sales in relation to rainfall

9
8

Financial Planning using Excel

7
6
Aircraft sales

5
4
3
2
1
0 41
0 1 2 3 4 5 6 7 8 9 10
Rainfall
Aircraft sales

Figure 4.4 Scatter diagram showing no discernible relationship between the

variables

Having established a linear relationship between the two variables

it is useful to fit a least square line to the data, and this is referred
to as simple linear regression1. Simple here implies that only one
variable is being used to predict another as opposed to multiple
linear regression which is discussed later in this chapter.

The algebraic formula for fitting a straight line is:

Y ⫽ MX ⫹ c

where Y ⫽ the dependent variable, M ⫽ the slope or gradient of the

line (and is sometimes called the regression coefficient), X ⫽ the

1 The word ‘regression’ originates from the nineteenth century when Galton collected
the heights of fathers and their sons and put forward the idea that since very tall
fathers tended to have slightly shorter sons and very short fathers tended to have
slightly taller sons, there would be what he termed a ‘regression to the mean’
(D. Rees, 1991, Essential Statistics, Chapman and Hall).
 independent variable and c is the constant, or intercept, sometimes
referred to as the Y intercept. The c is the value of Y when X is
zero. It is the base value of Y before any increase or decrease in X is
taken into account. The dependent variable (Y) is the variable to be
predicted which in the above example would be the income, and
the independent variable (X) is the variable used to do the predict-
ing, which in this case refers to the number of emails sent.

In defining the least square line consider the annotated graph in

Figure 4.5. For any given value of X there is a difference between
Financial Planning using Excel

Y1 and the value of Y determined by the line. The difference is

denoted by D, which is referred to as a deviation, residual or error.
The least square line is defined as the line through a data set that
has the property of minimising the sum of any given set of D2s.

16000

15000
42 D1 D2
14000
Deposits

13000 D3

12000

11000

10000
10.00% 10.50% 11.00% 11.50% 12.00% 12.50%
Interest rate
Deposits Forecast

Figure 4.5 Least square line

Given that the least square line approximation has the form
Y ⫽ MX ⫹ c this formula can be broken down as follows:

N兺XY ⫺ (兺X)(兺Y)
m⫽
N兺X2 ⫺ (兺X)2

and

(兺Y)(兺X)2 ⫺ (兺X)(兺XY)
c⫽
N兺X2 ⫺ (兺X)2
 where N ⫽ the number of data points available, X ⫽ the sum of the
X data points, Y ⫽ the sum of the Y data points and XY ⫽ the sum
of the product of each set of data.

Having established the equation representing the least square line,

an estimate of a value for Y corresponding to a value for X may be
obtained. When the least square line is being used in this way it has
a regression line of Y on X, since Y is estimated from a value of X.

Financial Planning using Excel

Calculating the least square line
The least square line can be calculated either through the use of the
FORECAST function or the TOOLS DATA ANALYSIS REGRESSION command.

The function approach

Figure 4.6 shows the results of the Excel FORECAST function. The
following formula was entered into cell D5.
43
⫽FORECAST(A5,$B$5:$B$23,$A$5:$A$23)

Figure 4.6 Calculated straight line using the Excel FORECAST function
 TECHNIQUE Note that in the above formula it is important to reference the
TIP! range for the dependent variable first, followed by the range for
the independent variable. These references are absolute in order
that they remain fixed when the formula is extrapolated for the
remaining cells in the range.

The data in column E can now be plotted as a line onto the scatter
diagram. A simple way to do this in Excel is to select the range,
Copy it to the Clipboard, access and select the chart and then select
Paste. Figure 4.7 shows the forecast line on the scatter diagram.
Financial Planning using Excel

16000

15000

14000
Deposits

13000

12000
44
11000

10000
10.00% 10.50% 11.00% 11.50% 12.00% 12.50%
Interest rate
Deposits Forecast

Figure 4.7 Scatter diagram with regression line

The regression line and regression equation applies only within the
range of the independent or X range of data. The line should not be
extrapolated much below the minimum value of X or above the
maximum value of X. Similarly the regression equation should not
be used for values of X outside the range of data.

Two other Excel functions that are useful in this context are
INTERCEPT and SLOPE which calculate the values for c and M
respectively in the formula Y ⫽ MX ⫹ c. Figure 4.8 shows the
results of these functions. The following formulae were entered
into cells E5 and F5 respectively:

⫽INTERCEPT(B5:B23,A5:A23)
⫽SLOPE(B5:B23,A5:A23)
 Financial Planning using Excel
Figure 4.8 Results of Excel INTERCEPT and SLOPE functions

45

The results of these two functions can be used to calculate the fore-
cast which is useful as an auditing tool to cross-check the result
produced using the FORECAST function. The following formula is
required in cell G5:

⫽$F$5*A5⫹$E$5

The Command Approach

To use the Regression command in Excel it is necessary to have

the Analysis Toolpak2 installed and enabled. Once enabled Data
Analysis becomes the last option on the Tools menu and within
Data Analysis one of the options is Regression. The command
requires references to the X and Y data and a location for the
output. Figure 4.9 shows the results of the command where the
output range was selected as a new worksheet.

2The Analysis Toolpak is a standard part of the Excel package, but if a custom installation
was selected during setup it is possible that this option was not selected. Refer to your
Excel manual for further information on installation.
Financial Planning using Excel

Figure 4.9 Results of Regression command

The regression formula can be created by referencing the intercept

or constant in cell B17 and the X coefficient in cell B18.

46
Function vs command
TECHNIQUE
The previous section has explained how to perform regression
TIP! analysis using spreadsheet functions and spreadsheet commands.
It is generally preferable to apply functions as opposed to com-
mands wherever possible. This is because functions are dynamic
formulae that are recalculated when anything in the spreadsheet
changes. Therefore having set up a model such as the one used in
this chapter, a different set of sample data could be entered into
columns A and B and all the regression formulae and the chart
will automatically reflect the changed data. This is not the case
with a command, because on executing the command only the
OPPORTUNITY resulting values are placed in the spreadsheet without the underly-
FOR GIGO ing formula. Therefore to recalculate after making a change it is
necessary to execute the Regression command again.

The graphic approach

Excel has the ability to plot a trendline onto a chart without having
to calculate the equation beforehand. Figure 4.10 shows the original
scatter diagram on which the data series has been selected and the
command Insert Trendline has been taken.
 Financial Planning using Excel
Figure 4.10 Inserting a trendline onto an Excel chart

16000
15000 y = 130181x – 1250

14000
Deposits

47
13000
12000
11000
10000
10.00% 10.50% 11.00% 11.50% 12.00% 12.50%
Interest rate

Deposits Linear (deposits)

Figure 4.11 Trendline inserted onto an Excel chart

The first option is most suitable for this example, but non-linear
trendlines can also be inserted if appropriate. Figure 4.11 shows the
resulting trendline. Although the individual calculations behind
the line cannot be seen, selecting FORMAT TRENDLINE OPTIONS allows
the formula to be displayed as shown in Figure 4.11.

Standard error
An additional refinement of linear regression is to incorporate the
standard error of the estimate. The standard error has many of the
properties analogous to the standard deviation.
 For example, if lines are drawn on either side of the regression line
of Y on X at respective vertical distances, confidence limits may be
established.

The algebraic formula for a 95% confidence limit is:

√冤 冥
1 (x0 ⫺ x)2
(a ⫹ bx0) ⫾ tsr ⫹
n (兺x)2
兺x2 ⫺
n
Financial Planning using Excel

Figure 4.12 shows the results of calculating upper and lower confi-
dence limits. Two columns have been inserted into the original
spreadsheet so that the already calculated forecast is in column E
and the lower and upper standard errors will be placed in columns
C and D respectively. The formula required in cell C5 to calculate
the lower confidence limit is as follows.

⫽E5⫺‘LR COMMAND’!$D$18*‘LR COMMAND’!$B$7*SQRT

48
(1⫹1/‘LR COMMAND’!$B$8⫹(A5⫺AVERAGE($A$5:$A$23)^2/
‘LR COMMAND’!$B$8*STDEV($A$5:$A$23)))

Figure 4.12 Results of standard error calculations

 The Regression command output in Figure 4.9 has been placed on a
sheet called LR Command and this is seen in the formula when ref-
erencing the Student t distribution, the standard error and the num-
ber of observations.

To calculate the upper confidence limit in cell D5 only the beginning

of the formula is different to the lower confidence limit. In this case
the formula begins with ⫽E5⫹ . . .

The results of these formulae are shown in Figure 4.12 and

Figure 4.13 shows the lines plotted onto the graph.

Financial Planning using Excel

20000
18000
16000
Deposits

14000
12000
49
10000
8000
6000
10.00% 10.50% 11.00% 11.50% 12.00% 12.50%
Interest rate

Deposits Forecast Lower limit Upper limit

Figure 4.13 Regression line with upper and lower confidence limits

Multiple linear regression

When two or more dependent (X) variables are required for a pre-
diction the analysis is referred to as multiple linear regression.

The simple linear regression equation can be adapted to accommo-

date multiple dependent variables in the following way:

Y ⫽ A0 ⫹ A1X1 ⫹ A2X2 ⫹ ⭈ ⭈ ⭈ ⫹ AnXn

Theoretically there is no limit to the number of independent variables

that can be analysed, but within the spreadsheet the maximum is 75.
This is a far greater number than would ever actually be required in a
business situation.
 Figure 4.14 Multiple regression scenario
Financial Planning using Excel

The following example in Figure 4.14 represents expenditure on

advertising and sales promotion together with sales achieved,
where the advertising and sales promotion are dependent variables
and the sales achieved is the independent variable. The require-
ment is to estimate future sales by entering advertising and promo-
tion expenditure, and applying the multiple regression equation to
50
predict the sales value.

In Excel there are no built-in functions for multiple regression and

therefore the command method is required. Figure 4.15 shows the

Figure 4.15 Results of Regression command for multiple linear regression

 results of the Excel Regression command where the range C3:C12
from Figure 4.14 was specified as the independent variable and
A3:B12 from Figure 4.14 as the dependent variables. The regression
output has been placed on the same sheet commencing in cell A14.
In order to calculate the estimated sales value in cell G5 the data in
cells B30, B31 and B32 shown in Figure 4.15 are required and the
formula for cell G5 is:

⫽(G6*B31)⫹(G7*B32)⫹B30

To use this multiple regression model, data must first be entered

Financial Planning using Excel

into cells G6 and G7. Figure 4.16 shows the estimated sales value if
the advertising expenditure is 250 and the promotion expenditure
is 125.

51

Figure 4.16 Results of multiple regression using the command method

Providing the X and Y data does not change, different values can
be entered into cells G6 and G7 and the estimated sales value
will recalculate correctly. However, because the command
method has been used for the regression, if any of the X or Y
data changes it will be necessary to execute the Regression com-
mand again.

Selecting variables for multiple regression

A problem encountered when applying multiple regression
methods in forecasting is deciding which variables to use. In
the interests of obtaining as accurate a forecast as possible, it is
obvious that all the relevant variables should be included.
 However, it is often not feasible to do this, because, apart from
the complexity of the resulting equation, the following factors
must be considered.

◆ In a forecasting situation the inclusion of variables which do not

significantly contribute to the regression merely inflate the error
of the resulting prediction.
◆ The cost of monitoring many variables may be very high – so
it is advisable to exclude variables that are not contributing
significantly.
Financial Planning using Excel

◆ There is no point including terms that contribute less than

the error variance, unless there is strong prior evidence that a
particular variable should be included.
◆ For successful forecasting, an equation that is stable over a wide
range of conditions is necessary. The smaller the number of vari-
ables in the equation, the more stable and reliable the equation.

52 Summary
Regression analysis is a popular forecasting and estimating tech-
nique. Although many users might find the mathematics involved
quite difficult, the technique itself is relatively easy to use, especially
when a model or template has previously been developed.

However, users who do not understand the underlying mathematics

should obtain some assistance in the interpretation of the results.
5
Time Series Analysis
This Page Intentionally Left Blank
 Technology is the acknowledged master, the engine that pulls all
the rest along and determines where the future shall be, how fast it
shall be attained, what we shall do within it, and who shall pros-
per, who shall languish, who shall fossilize.

– T. Levitt, Marketing in an Electronic Age, 1985.

Introduction

Financial Planning using Excel

A time series is a set of observations recorded over a period of
time. It is a sequence of data that is usually observed at regular
intervals.

Time series analysis predicts the value of an item by studying the

past movements of that item. It is a series that uses time as the
independent or explanatory variable.

Time series data can be analysed for trend, seasonal fluctuations

and residual. Trend analysis involves using a linear regression 55
model with time as the independent variable. The data is also
adjusted for seasonal fluctuations since this can significantly influ-
ence the regression results. The residual is what is left over after
the trend and seasonality has been taken into account, and in this
case consists of a cycle and an error.

Many time series follow a cycle, with either an annual or seasonal

pattern. These cyclical effects are analysed, as are the errors. This is
achieved through decomposition analysis.

Decomposition analysis is the method of reducing a set of time

series data to a trend, a seasonal factor and a residual. This may be
expressed as:

D⫽TSR

where D ⫽ data, T ⫽ trend, S ⫽ seasonality factor and R ⫽ residual

which includes the error.

The decomposition analysis may be either multiplicative or additive.

Additive decomposition takes the following form:

D⫽T⫹S⫹R
 and multiplicative decomposition analysis takes the form:

D⫽T*S*R

The multiplicative model is also referred to as a ration model. It is

the multiplicative model that is used in this chapter.

Multiplicative time series analysis

The decomposition method described here fits a linear regression
Financial Planning using Excel

line to a set of historic quarterly sales data, establishes the trend,

and then uses a centred weighted moving average to isolate the
seasonality. The residual from the difference between the data
trend and the seasonality is then deduced.

The cyclical component of a series may be estimated by comparing

the actual value for each month, with the centred moving average
for the month. Since the centred moving average is based on four
quarters and is equally weighted with data from, before and after
56
the periods in question, the moving average should be free of sea-
sonal effects and should not be affected by trend.

The de-seasonalised series consists of the historic series, but with the
seasonal fluctuations removed. It is a useful aid for judging the overall
behaviour of the historic series because the distraction of the seasonal
fluctuations has been removed. It is also useful for judging whether
one or more values in the historic series deviate appreciably from the
general trend. It is important to detect and, if possible, explain such
anomalies in order to anticipate potential future recurrences.

The trend describes the long-term behaviour of the series after the
removal of the seasonal effects and the irregularities due to short-
term random fluctuations. The trend is shown by means of a
straight line whose slope indicates the long-term increase or
decrease per unit time in the series.

The seasonal component indicates the adjustments that act on the

trend due to seasonality. These are expressed as percentages, and
repeat themselves in each year.

The residuals are also expressed as percentages. They represent

the deviations from the combined series consisting of trend and
seasonal components. Allowances are made for the possibility
that the residuals may exhibit cyclical behaviour.
A time series analysis model
The development of a time series model requires a moderate
amount of statistical knowledge that is beyond the scope of this
book to explain in detail. However, the system shown in this
chapter is a good example of how a complex application can be
developed in such a way that it can be used by less experienced
people to great effect. The template on the CD accompanying this
book which is called TIMESERIES can be used without amendment [Link]

to produce quarterly time series forecasts, and readers wishing to

Financial Planning using Excel

examine the theory behind the formulae can study the Calculation
and Processing worksheet.

The approach to developing the quarterly time series forecasting

system is described in this chapter and the model consists of three
worksheets which have been called Data Input and Output,
Calculation and Processing and Graphics.

Data input and output 57

The Data Input and Output sheet consists of an input form that
captures the historic data. The model has been designed to collect
historic quarterly data for a three-year period. On the same sheet
the final results of the time series analysis are displayed in order
that comparisons can be made with the historic data.

Figure 5.1 shows the completed Data Input and Output sheet
together with some sample data.

Figure 5.1 Input and Output worksheet with data

 Cells B11 through E13 contain references to cells in the Calculation
and Processing sheet as no calculations are performed on the Data
Input and Output sheet. This can be seen in Figure 5.2.
Financial Planning using Excel

Figure 5.2 Cell contents for Data Input and Output worksheet

Calculation and processing

58
The Calculation and Processing sheet contains the formulae for
decomposing the historic data to establish the trend, seasonality
and residual and then to provide the forecast data with standard
deviations. Figures 5.3 and 5.4 show the results section of the
Calculation and Processing sheet, and Figures 5.5 and 5.6 show the

Figure 5.3 Results section of Calculation and Processing worksheet

 Financial Planning using Excel
Figure 5.4 Formulae for results section of Calculation and Processing worksheet

59

Figure 5.5 Work area of Calculation and Processing worksheet

Figure 5.6 Formulae for the work area of the Calculation and Processing worksheet
 work area required to calculate some of the statistics for which
there are no built-in statistical functions.

Rows 6 to 10 of the Calculation and Processing sheet decompose

the historical data to show the trend, the seasonal components and
the residual.

The trend describes the long-term behaviour of the series after

removal of the seasonal effects and the irregularities due to short-
term random fluctuations. It is described by means of a straight
line, the slope of which indicates the long-term rate of increase or
Financial Planning using Excel

decrease per unit time in the series. The trend is computed using
the centred moving average calculated in row 28 which has the
effect of smoothing out seasonal fluctuations and (to some degree)
short-term irregularities.

The seasonal components indicate the adjustments that act on the

trend due to seasonality. These are expressed as percentages and
show, for example, that if the component for season one is 90%
60 then, on average, the value of the series in season one is 90% of the
value indicated by the trend. There are precisely as many different
seasonal components as there are seasons in the data and so in this
example there are four seasons for this quarterly series. The sea-
sonal components repeat each year.

The residuals are also expressed as percentages and represent the

deviations from the combined series consisting of trend and sea-
sonal components. Therefore, for example, if the first historical
value is 75.9 units and the trend and seasonal components associ-
ated with this point in time are 65.0 and 120% respectively, the
corresponding residual is calculated as:

75.9/(65.0*120%) ⫽ 0.973 ⫽ 97.3%

The mean forecast in row 21 can then be calculated by multiply-

ing the trend by the seasonality by the residual (F⫽T*S*R), and
high and low forecast bounds are calculated in rows 19, 20, 23
and 24.

When developing a complex model such as the one described here,

TECHNIQUE a useful form of documentation is to include a brief description of
TIP! each row to the right of the plan. This can be seen in column N of
the TIMESERIES model.
Graphics
Three charts have been created which illustrate the original data
and the seasonally adjusted data, the trend and seasonally adjusted
data and the trendline, and the forecast together with the lines for
plus and minus two and four standard deviations. Figures 5.7, 5.8
and 5.9 are examples of these charts using the sample data from
Figure 5.1.

Financial Planning using Excel

9000
8000
7000
Sales volumes

6000
5000
4000
3000
2000
1000
0 61
Q1 Q2 Q3 Q4 Q5 Q6 Q7 Q8 Q9 Q10 Q11 Q12
Quarters

Historic data Seasonally adjusted data

Figure 5.7 Original data and seasonally adjusted data

10000
9000
8000
7000
Sales volume

6000
5000
4000
3000
2000
1000
0
1 2 3 4 5 6 7 8 9 10 11 12 13
Quarters
Trendline Trend and seasonally adjusted data

Figure 5.8 Trendline and seasonally adjusted data

 16000

14000

12000

Sales volume 10000

8000

6000

4000
Financial Planning using Excel

2000

0
Q13

Q14

Q15

Q16

Q17

Q18

Q19

Q20

Q21

Q22

Q23

Q24
Quarters

Very Low (–4 STD) Low (–2 STD) Mean High (+2 STD) Very High (+4 STD)

62 Figure 5.9 Mean forecast and forecast plus and minus two and four standard deviations

Summary
As already stated the formulae required to develop the multiplicative
time series template are not trivial and in order to understand the
relationships required a sound knowledge of statistics is necessary.

It is beyond the scope of this book to explain in detail all the formulae
used, but rather a general overview of the system has been supplied
which will allow the reader to work with the template supplied on
the accompanying CD.
6
Expected Values
This Page Intentionally Left Blank
 The affair (investment in business) was partly a lottery, though
with the ultimate result largely governed by whether the abilities
and character of the managers were above or below the average.
Some would fail and some would succeed.

– John Maynard Keynes, The General Theory of Employment,

Interest and Money, 1964.

Introduction

Financial Planning using Excel

Quantitative approaches to forecasting as illustrated in the previous
examples depend heavily on data manipulation, whereas qualitative
approaches require the input of human knowledge about the item
to be forecast. One such subjective approach to forecasting is that
of a composite of individual estimates using expected values. The
approach is quite different to the techniques so far described in this
book because it does not depend only on historic data. The expected
value technique requires the opinions of experts as to the likely
occurrence of events in the future. These occurrences are estimated 65

in terms of their magnitude as well as in terms of their probability of

occurrence.

For example, this technique can be used for the evaluation of the
probabilities of future sales where there are a large number of clients
who are expected to re-purchase from time to time. Because the data
will be estimates of expected sales, it is important that the sales
force maintain close and regular contact with clients and prospects
in order that they are in a position to prepare useful estimations.
It is also important that the sales director responsible for the fore-
cast model is familiar with both the clients and his or her sales
team.

A model for analysing expected values

The input required for a model to evaluate the probabilities of
future sales is a list of existing clients’ and prospects’ names,
together with estimates of what they are expected to spend on the [Link]

firms’ products over the forecast period. The probability of the sales
actually occurring, represented as a percentage value, is entered
and the estimated sales value is multiplied by the probability to
 produce the expected value. The expected value for all the clients
and prospects is totalled and this value is used as the forecast of the
total sales revenue for the forecast period.

Figure 6.1 shows the spreadsheet on which the data is entered and
the expected values are calculated. This data can be found in the
file EXPECT on the CD accompanying this book. For the purposes of
this example, only a small sample of data has been used, but in a
real business situation this would be much larger.
Financial Planning using Excel

66

Figure 6.1 Model for analysing expected sales values

This technique relies on the fact that with a large list of clients and
prospects, the probabilities will average themselves out and the end
result will be close to the forecast expected value. The formula for
the expected value is:

Expected value ⫽ Forecast value * Probability

 The model does not contain many formulae. A simple multiplication
in the Expected value column multiplies the forecast sales value
by the probability, and in the totals row the forecast sales and the
expected values have been totalled and the probability has been
averaged using the ⫽average function.

The totals have been entered using the ⫽subtotal function as opposed
the more common ⫽sum function because the ⫽subtotal function
allows cells to be totalled subject to a criteria. The formula in cell E33
is ⫽SUBTOTAL(9,E4:E32).

Financial Planning using Excel

Where 9 means that only cells that are on view are to be summed.
The significance of this will become clear in the next section when
the list is filtered to show only selected clients and the totals will
reflect the totals of the filtered list.

The ⫽subtotal function has options 1–11 which refer to different

functions. For example, ⫽SUBTOTAL(1,D4:D32) will average the cells
that are on view as opposed to summing them. More information
on the use of this function can be found in the Help function 67
within Excel.

Extracting data
Having entered all the data into the spreadsheet it can be used to
select subsets of clients based on specified criteria. For example, it
might be useful to know the clients who are listed with a probability
of more than 0.5.

Excel has a feature called Autofilter that can be used for this purpose.
Autofilter is switched on by placing the cursor on any of the cells
in the list area (A3:E33 in this example) and selecting DATA FILTER
AUTOFILTER. This places a small arrow to the right of each column
heading in the expected value table. By clicking on the arrow in the
Probability column a series of options are displayed which includes
Custom. Figure 6.2 shows the custom dialogue box which allows the
required criteria to be entered – in this case ‘greater than (⬎) 0.5’.

The results of the above query can be seen in Figure 6.3. Notice that
the rows containing probability values that do not satisfy the crite-
ria are hidden and the totals have been recalculated to reflect the
items currently in the list.
Financial Planning using Excel

Figure 6.2 Criteria selection dialogue box

68

Figure 6.3 Results of query on probability greater than 0.5

If a permanent record is required of items satisfying a particular

requirement, a report can be produced using the DATA FILTER ADVANCE
FILTER command.

On a separate worksheet the column headings should be copied

twice – once to be used as a reference for the criteria and once as
the headings of the final report. In the example here, all the head-
ings have been copied across, but it is possible to be selective here
providing the syntax of the headings matches the column headings
on the original list of data. Figure 6.4 shows the prepared work-
sheet and the dialogue box for the DATA FILTER ADVANCED FILTER
command.

The results of the command are shown in Figure 6.5.

Financial Planning using Excel

69

Figure 6.4 Data Filter Advanced Filter command

Figure 6.5 A permanent report produced using Data Filter Advanced Filter
 It should be noted that the results of this command are not dynamic
(rather like the Data Regression command used in an earlier chapter)
OPPORTUNITY
FOR GIGO
and thus if the data in the original list is changed, added to or
deleted from, this report will not automatically update.

Summary
The composite of individual estimates using expected values tech-
nique is particularly useful in situations where there are a large num-
Financial Planning using Excel

ber of clients and prospects, and when there is a sales force who
maintains regular and close contact with the client/prospect base.

In these situations, the large number of clients and prospects

ensure that the expected value averages out over the whole range,
and that the resulting estimation approximately represents the
actual sales.

The main problem with this technique is the subjective values

70 involved in setting the probability levels for each client or
prospect. If great care is not taken over these figures then the
resulting forecast is likely to be of little value.
7
Selecting and Evaluating
Forecasting Techniques
This Page Intentionally Left Blank
 Real problems are hard to spot, especially for managers so involved
in day-to-day operations that they have inadequate perspective to
see the big picture.

– F. Wiersema, Customer Intimacy, 1996.

Introduction
This section of the book will explain in some detail how to
prepare forecasts using a range of different methods, tools and

Financial Planning using Excel

techniques. For any forecast to be useful it has to be accurate
to an acceptable level and the level of accuracy can be affected
considerably by the forecasting approach that is taken. This
chapter provides a table of different forecasting techniques and
indicates when they might be suitable. Some guidelines are given
as to how to measure the level of accuracy after a forecast has
been calculated.

73

Selecting the right technique

Table 7.1 provides a list of the forecasting techniques described in
this book and evaluates them under the following sections:

◆ appropriate use
◆ time horizon
◆ maths sophistication
◆ data required.

Table 7.1 Different forecasting techniques

Technique Use Maths Data

Moving Repeated forecasts A minimal level Historic data

averages without seasonality of competency essential
Adaptive Repeated forecasts Some statistical Historic data
filtering without seasonality knowledge is essential, but
and where the nature required level and
of any trend may detail can
change over time vary

Continued
 Table 7.1 Different forecasting techniques—cont’d

Technique Use Maths Data

Exponential Repeated forecast Some statistical Only recent

smoothing with or without knowledge to set data and
seasonality up a system but current
virtually none to forecast is
operate a well- required once
designed system the smoothing
factor is
established
Financial Planning using Excel

Simple Comparing one Some statistical A sample

linear independent variable knowledge to of relevant
regression with one dependent set up a system observations
variable where but virtually for the
there is a linear none to independent
relationship operate a and
between well-designed dependent
the variables system variables
Multiple When more Some statistical A sample of
linear than one knowledge is relevant
74
regression independent required observations
variables is to be for the
compared with a independent
dependent and
variable dependent
variables
Trend Forecasting over time Some statistical Historic data
analysis in a non-linear way competence is with as much
using simple required detail as
regression possible
Decomposition Identify seasonal Requires statistical Historic data
analysis components knowledge to for the one
as part set up a system variable under
of another but virtually none consideration
forecasting to operate a is required
method well-designed
system
Composite Forecasting using a No statistics Past data
of individual large number of and only basic not required
estimates individual estimates arithmetic
required

The techniques described are not an exhaustive list of forecasting

techniques, but cover those most suited to the spreadsheet environ-
ment, and which have already been explained in this book.
Accuracy and reliability
Every forecast should include an estimate of its accuracy and
an assessment of its reliability. This not only allows for a
reasonable and acceptable margin of error, but also helps to
build confidence in the methods being used, and the results
achieved.

Many forecasts are repetitive, being prepared either weekly or

monthly. This provides an invaluable source of forecast assessment
data, since the forecasts can be measured against the reality, and

Financial Planning using Excel

can then be adjusted with the benefit of hindsight. Explicit analysis
will highlight weak spots in the forecast, and improve the accuracy
of future forecasts.

The three primary approaches to evaluating the accuracy of a forecast

are to use charts, supported by quantitative (statistical) analysis and
subjective methods.

75

Charts
Charts are particularly useful when it is necessary to identify
potential problems as early as possible. Control charts are the most
frequently used graphical method, and these are useful for identi-
fying when a forecast method requires adjustment. Prediction real-
isation diagrams are another graphical method to consider. The
pictorial representation of the data makes it easier to spot changes
in data patterns and can direct the analyst to the data areas that
need to be examined.

Statistical methods
Quantitative or statistical methods are useful for comparing differ-
ent forecasting methods, and in estimating confidence limits. A
confidence limit is a specification that combines a statement about
a point estimate with a measure of the precision of that estimate.
Some useful statistical measures that have been used in the models
in this section include the following.
 Average error
The average error, or the average absolute error or the mean absolute
error, is the simplest of all to compute. The algebraic formula is:

Average error ⫽ (&(F1 ⫺ A1)&⫹&(F2 ⫹ A2)&⫹ ⭈ ⭈ ⭈

⫹&(Ft ⫺ At)&)/T

where Ft ⫽ forecast for period t, At ⫽ actual value for period t, T ⫽

number of forecasts to be used and the modulus ( | ) means that
the absolute value is to be used. In other words if the value is nega-
Financial Planning using Excel

tive, its sign should be changed to positive.

Once computed, the average error for a particular forecasting

method can be compared with the average error for another method.
Alternatively, the average error for different time periods can be
calculated and compared. Although the average error does not have
any intrinsic statistical property that makes it useful for the estima-
tion of confidence limits, it is a useful indication as to the accuracy
76 of a forecast.

Standard error
The most widely used measure of forecasting accuracy is the standard
error and this technique was used in Chapter 4 in conjunction with
the regression analysis.

The standard error is particularly useful because it can be used to

compute confidence limits and, because the terms are squared, it
places more emphasis on large errors. There are however some
pre-requisites for the use of standard errors:

◆ The error terms are normally distributed. A normal distribution is

a model of a continuous variable whose value depends upon the
effect of a number of factors, each factor exerting a small positive
or negative influence.
◆ The average forecasting error is zero, i.e. there is no tendency to
regularly over or under-estimate.
◆ The errors are not serially correlated, i.e. the error terms do not
show a pattern, in terms of direction or size, over time.

If all these conditions are met, then 68% of the outcomes will be
within plus or minus one standard error, 95% will be within plus
 or minus twice the standard error and 99.7% will be within plus or
minus three times the standard error.

The control chart will show whether there is any pattern in the
error terms, and calculation of the error terms will show whether
they meet the necessary conditions as a basis for computation of
the confidence limits.

Thiels U

Financial Planning using Excel

Thiels U compares the accuracy of a forecasting model to a naïve
model that uses the last period’s outcome to predict the present
period’s outcome. This is using what is already known to predict the
future and is thus a measure of relative accuracy. U is defined by:

Standard error of forecasting model

U⫽
Standard error of naïve model

Values of U greater than one indicate that the method of forecasting 77

selected was worse than the naïve model; values less than one indi-
cate that it was better.

U can also be used to compare the accuracy of various forecasting

methods, or even to compare the accuracy of one method over differ-
ent time-spans. This is useful when selecting the optimum time-span
or period for a forecast.

Subjective methods
Subjective forecasting methods are based on common sense. They
involve using a measure of judgement and self-expertise in the
evaluation of a forecast.

When applying subjective methods it is important to avoid spurious

correlations which refers to correlations where the associations are
real enough, but misinterpretation of the results, or unwise jumping
to conclusions, has led to erroneous conclusions being drawn from
the forecast.

The original assumptions under which the predictions were made,

and upon which the model was built, should be kept in mind. If a
 factor is considered to be particularly critical, then different fore-
casts can be prepared using alternative assumptions.

No matter how well the forecast has performed in the past, the
model itself and its underlying assumptions should be periodi-
cally re-examined in order to ensure that the same conditions
prevail.

Summary
Financial Planning using Excel

As success at forecasting is an important contributor to the firm’s

performance, time and effort should be applied to ensure that a
suitable forecasting technique is being used. This will certainly
result in costs of various types being minimised.

All three methods discussed in this chapter are important and can
be applied individually or collectively depending on the precise
situation.
78
Part 2
This Page Intentionally Left Blank
8
Business Planning
This Page Intentionally Left Blank
 Planning, of course is not a separate, recognizable act . . . . Every
managerial act, mental or physical is inexorably intertwined with
planning. It is as much a part of every managerial act as breathing
is to the living human.

– C. George, The History of Management, 1972.

Introduction
Business forecasting involves the prediction of what could happen

Financial Planning using Excel

in the future and business planning involves the calculation of
what is required to fulfil the forecast, i.e. make the future happen as
it is forecast. Thus a business plan uses a forecast as an initial point
of departure and then adds the detail to reflect how the particular
process or project is intended to proceed. Business plans are gener-
ally set in the future and involve making a series of assumptions
about how the variables in the plan will behave in future periods.

Business plans are helpful with the decision-making process of an

83
organisation. There are many types of business plan, which range
from simple verbal or written descriptions of what is to be achieved,
through diagrammatic representations such as Gantt charts to highly
analytical numeric reports.

There are many definitions of planning, but perhaps one worth

quoting is from Ackoff1 who said:

Planning is required when the future state that we desire involves a

set of interdependent decisions; that is, a system of decisions . . .
The principal complexity in planning derives from the interrelated-
ness of the decisions rather than from the decisions themselves . . .

In the context of this book, a business plan is a statement of quan-

tifiable objectives, together with the resources required to achieve
them. The objective may be profit, market share, return on invest-
ment or cash flow, to mention only a few, and the resources could
be manpower, materials, capital, cash, etc.

There are many different ways of classifying business plans,

including short-term plans, long-term plans, profit plans, marketing
plans, production plans, personnel development plans, capital

1 R.L. Ackoff, A Concept of Corporate Planning, Wiley, New York, 1970.

 investment plans, merger and take-over plans, etc. The level of
analysis in the plan reflects the ultimate use for which it is
intended. For high-level strategic decisions, plans generally do not
contain very much detail, while for operational planning a large
amount of detail is required. The contents will reflect the nature
of the project or process being planned, but it is not unusual for
operational plans to involve dozens of line items. These types of
plan often form the basis of a financial budget, which in some cases
is no more than a plan divided into specific responsibility points
and specific time periods.
Financial Planning using Excel

An important aspect of business planning is what-if analysis. This

is a process of changing one or more of the variables or assump-
tions on which a business plan has been based in order to see what
effect such a change in circumstances would have on the results.
This subject is covered in detail in Chapter 12.

84 Approaches to business planning

There are three generically different approaches to developing
business plans, which are described as deterministic2, stochastic3
or probabilistic and optimising.

Deterministic planning
Deterministic planning is the most frequently encountered. It is the
basis on which virtually all corporate planning and budgeting is
performed. A deterministic plan takes single values that are the best
estimate for the variables in the plan and uses this value with a set
of logic rules to produce the plan or budget. The single start value
for a line item in a plan is referred to as a point estimate. This might
be the result of a forecasting model, and although it is recognised by
planners and budgeting officers that accurate point estimates are
almost impossible to obtain, it is usually possible to derive a value
that is accurate enough to be useful.

2 According to the American Heritage Dictionary of the English Language, Third Edition,
deterministic refers to the fact that there is an inevitable consequence of antecedents.
3 According to the American Heritage Dictionary of the English Language, Third

Edition, stochastic refers to involving a random variable or variables and the chance or
probability of their occurrence.
The method of planning is called ‘deterministic’ because once the
point estimates are derived, the outcome of the plan may be
uniquely determined by the logic, i.e. there will be only one
answer. For example, consider the following:

Sales volume 200

Unit price 20
Revenue 4000 (Sales volume * Unit price)

The only way a different result can be obtained in the above example
is by changing one or both of the single point estimates, i.e. the data

Financial Planning using Excel

for the sales volume and the unit price. Of course, one of the great
advantages of the spreadsheet is that making changes to opening
data, growth rates, cost factors, etc. causes all directly and indirectly
affected cells to recalculate and thus different scenarios can be calcu-
lated. However, the basic principle is that the outcome is uniquely
determined by the logic.

Figure 8.1 is a profit and loss account for CWL that has been devel-
oped as a deterministic plan. Figure 8.2 shows the formulae used 85
for the plan.

CWL-P AND [Link]

Figure 8.1 A typical deterministic plan

Financial Planning using Excel

Figure 8.2 Formulae for the deterministic plan

86
The example shown above is a profit and loss account for a profit-
making organisation. However, the deterministic approach to busi-
ness planning is equally applied to not-for-profit situations when
an organisation might be planning, for example, an allocation of
funds. Figure 8.3 is such an example.

[Link]

Figure 8.3 Allocation of funds plan

Stochastic planning
A stochastic or probabilistic plan recognises that because point
estimates are difficult to accurately establish, it is can be preferable
to establish ranges into which it is thought values will fall. For
example, in a stochastic plan, rather than using a point estimate of
£50,000 as an opening sales value, a range consisting of values of
between, say £45,000 and £52,500 could be entered. Similar ranges
could be entered for other line items in the plan.

Probabilities are then associated with the data input ranges and the

Financial Planning using Excel

computer recalculates the plan thousands of times using values within
the specified range to produce a large number of results, or scenarios,
which correspond to different possible outcomes. This method is
referred to as the Monte Carlo method. It is frequently used to evaluate
large capital investment alternatives of complex marketing situations.
Because this type of planning is concerned with attributing the proba-
bility of a particular outcome, it is also referred to as risk analysis.

There are a number of different probability distribution techniques 87

that can be applied to stochastic planning, including uniform,
triangular and beta. The examples shown in this book will assume
a uniform, or normal, distribution. This implies that when the com-
puter generates input data there is an equal chance of any value
within the specified range being selected. The other distributions
mentioned, i.e. triangular and beta distributions do not have this
uniform probability pattern.

Because stochastic planning requires a reiterative process to collect

thousands of alternative results, it requires considerable computing
power that was not used to be available on desk-top computers and
thus was not considered a spreadsheet application. However, with
the speed and capacity of the PC today, stochastic planning has
become a relatively simple application to develop.

The detail of how to set up and work with a stochastic plan for risk
analysis is given in Chapter 13, but the underlying logic required to
apply stochastic analysis to the deterministic plan shown in Figure
8.5 using a uniform or normal distribution is as follows:

No. course attendees Random * (max ⫺ min) ⫹ min

Avg. conference fee Random * (max ⫺ min) ⫹ min
Revenue Attendees * Fee
 The input data for the above model might be:

Max Min

No. course attendees 1200 800

Avg. conference fee 250 175

The use of a random selection from the ranges provides a normal,

uniform or rectangular distribution, but other distributions such as
Financial Planning using Excel

binomial, poisson or triangular could be specified.

Figure 8.4 shows a typical data input form for a stochastic plan and
Figure 8.5 shows some of the formulae required. The results of risk
analysis are usually best viewed graphically and an example is
shown in Figure 8.6.

88

[Link]

Figure 8.4 Input form for risk analysis

Figure 8.5 Formulae required for risk analysis

 Year end profit/loss

300
250
200
Frequency

150
100
50

Financial Planning using Excel

0
50

65

10

45

30

15

00
38

89

07

56
93

48

20

65

10

55
0

–1

33

67
3

–7

–3

10

13

17

20
–1

–1

Profit/loss
Year end profit/loss

Figure 8.6 Graph showing results of risk analysis

89
From the graph in Figure 8.6 it can be seen that, given all the
assumptions in the underlying model, the most likely outcome is a
net profit of about £20,000. However, given the ranges of input data
it is possible for there to be a net loss of £139,350 or there could
possibly be a profit as high as £205,000.

There are a number of factors that need to be taken into considera-

tion when interpreting the results of a stochastic plan and this is
discussed in more detail in Chapter 13.

Optimising models
Optimising models attempt to find a unique course of action that
will produce the best result, given a set of restraints. One of the
most frequently encountered optimising models in business is the
economic order quantity (EOQ) model, whereby the most cost-
effective order quantities are calculated. An optimising model
always attempts to match an objective within a given set of
restraints and the spreadsheet Solver feature makes this type of
analysis possible within the spreadsheet environment.

Backward iteration or goal seeking is another optimising technique

whereby the computer calculates the course of action that should
 be undertaken in order to achieve a particular objective. In this
case, it is usual to change only one variable at a time and con-
straints are not always necessary. In Excel the Goal Seek command
can be applied for this type of analysis. Both Solver and Goal Seek
are discussed in more detail in Chapter 12.

Table 8.1 shows the three types of business plans discussed in this
chapter together with their primary use and development features.
Financial Planning using Excel

Table 8.1 Summary of business planning approaches

Type of plan Primary use Features

Deterministic Income and expenditure, Extrapolation of formulae

budgeting and constraints
Risk or stochastic Complex capital Exploration of
investment appraisal multiple scenarios
Optimising Finding the unique Solving a series of
course of action to equations representing
90 produce the optimal result operational constraints

Summary
Before embarking on the development of any spreadsheet applica-
tion it is important to take time to consider what type of plan is
required and the best approach for the design and structure of the
plan. The following chapters in this book describe the techniques
required for the development of deterministic, stochastic and
optimising models, as well as how to get the most out of the plans
through what-if analysis.
9
Spreadsheet Skills for all
Types of Planning
This Page Intentionally Left Blank
 The most popularly claimed pitfall of planning concerns commit-
ment. The assumption is that with the support and participation of
the top management, all will be well. But the questions must be
asked: well with what and well for whom? For planners? To be
sure. But for the organisation?

– Henry Mintzberg, The Rise and Fall of Strategic Planning, 1994.

Introduction

Financial Planning using Excel

Whether a spreadsheet is being developed as a forecasting plan, a
profit and loss account or a marketing plan it is essential that due
care and attention be given to the design and structure of the plan.
Establishing some rules as to how all the spreadsheets in a depart-
ment or organisation are developed enables different people to look
at different plans and feel familiar with the layout, style, reports,
charts, etc. This is in much the same way as users feel familiar with
software applications that have a similar interface such as those in
93
the Microsoft Office suite of products.

The objectives of good design in spreadsheet terms are exactly the

same as those required for any other software development:

◆ To ensure that the spreadsheet is as error free as possible

◆ To ensure that the spreadsheet can be used without much training
or control
◆ To minimise the work required to enhance or change the spread-
sheet.

If care is taken to ensure sound structure and good design, a

spreadsheet will be straightforward to develop, easy to read, sim-
ple to use, not difficult to change and will produce the required
results.

The plan developed over a number of developmental stages in this

chapter illustrates a variety of aspects of the principles of spread-
sheet design and development. The series begins with a plan that
has had little or no thought put into its design and layout, and as
the chapter proceeds ways of improving and enhancing the plan
are identified and explained. These plans can be found on the
CD accompanying the book under the names STYLE01 through
STYLE10.
 Spreadsheet 1: Getting started
The spreadsheet in Figure 9.1 is a simple profit projection that
may be of use to the author, but is unlikely to be helpful to any-
one else. This is clearly a quick one-off plan which has been pre-
pared with very little care and which may well not even be saved
on the disk.
Financial Planning using Excel

[Link]

Figure 9.1 Simple profit projection

94

Problems with this spreadsheet

The immediately obvious problems with this spreadsheet are that it
has no title, it is not clear what the columns represent, i.e. they are
different periods or perhaps different products, and the author is
unknown.

With regards the data itself, the figures are hard to read as there are
varying numbers of decimal places. Whilst perhaps there has been
a growth in sales and price, the percentage has not been indicated.
The costs line could be misleading as no indication of where the
costs have been derived is supplied.

Positive aspects of this spreadsheet

If the author of the spreadsheet required a quick profit estimation
based on known data and growth rates for sales units, price and
costs then the spreadsheet has supplied that information quickly
and in a more concise form than would have been achievable using
a calculator and recording the results on a paper.
Spreadsheet 2: Ownership and version
In Figure 9.2 the three major shortfalls of the first spreadsheet have
been remedied. The plan has also been given a title and author
details have been included. It is important that every business plan
have a clear owner who is responsible for overseeing the accuracy
and maintenance of the system. A name plus some form of contact
details should always be included.

Financial Planning using Excel

95

[Link]

Figure 9.2 Incorporating some annotation

Problems with this spreadsheet

The construction of the data and results is still unclear and the lack of
formatting makes the figures hard to read. The costs remain grouped
together.

Positive aspects of this spreadsheet

In addition to the owner details having been added to the plan, the
date when the plan was written is a useful feature. The date
becomes particularly important when the question of spreadsheet
versions arise. Note that the date has been entered here as text. If a
date function had been used, it would be continually updated each
time the file is retrieved, whereas here it is the date of the last
update that is required. The ruling lines above and below specific
sections of the spreadsheet are also quite helpful. This can be
 quickly achieved using the automatic formatting features. These are
accessed via the FORMAT AUTOFORMAT command.

Spreadsheet 3: Formatting
In Figure 9.3 the data for the four quarters is totalled and reported
as an annual figure. The values in the plan have also been format-
ted with the majority of figures being formatted to zero decimal
places and the price line to two decimal places.
Financial Planning using Excel

[Link]

96

Figure 9.3 Formatting the plan

One of the automatic formatting options has been selected to shade

and outline the plan.

Problems with this spreadsheet

By looking at the plan in Figure 9.3 it can be seen that the sales and
the costs both increase over time. However it is not clear by how
much because the sales growth factor and the increase in costs have
been incorporated into the formulae as absolute references.

The inclusion of absolute values in formulae is not recommended

and can lead to GIGO. To change the sales growth factor in Figure 9.3
two processes are required. First, cell C7 is accessed and the edit key
OPPORTUNITY
is pressed. The growth factor is changed and enter is pressed. This
FOR GIGO
has changed the formula in this one cell, but only once the formula
 has been extrapolated across into cells D7 and D7 is the amendment
complete. It is not difficult to see that there is room for error here in a
number of different ways.

Positive aspects of this spreadsheet

Having a current date and time indicator displayed on the spread- TECHNIQUE

sheet ensures that a hard copy report will reflect the date, and per- TIP!
haps more importantly the time it was printed. This is achieved
through the NOW() function, which can be formatted with a range of

Financial Planning using Excel

different display options. Because it is likely that a spreadsheet will
be recalculated, even if it is set to manual calculation, before print-
ing, the date and time will always be up to date. It is of course pos-
sible to include the date and time in headers and footers, but dur-
ing the development phase of a system the page layout is of less
relevant than printing the section being worked on and so thought
should be given to the positioning of the NOW() function.

The cells in this plan have now been formatted, which makes the 97
data easier to read. When formatting a spreadsheet it is important to
consider the entire plan and not just the cells that are currently
being worked on. The entire spreadsheet should be formatted to the
degree of accuracy the majority of the plan is to be and those cells
that need to be different, such as percentages, can be reformatted
when necessary. This is quickly achieved by right clicking on the
top left corner of the spreadsheet at the intersection between the col-
umn letters and row numbers and then select format cells. Whatever
formatting is now applied will affect the entire worksheet.

It is important to understand that formatting cells only changes the

display and does not affect the results of calculations that are still
performed to the full degree of accuracy, which is usually 16 signif-
icant decimal places. This is why a cell containing the sum of a
range of cells might display an answer that does not agree with the
result of visually adding the values in the range.

The ROUND function is the only safe way to ensure that the results of a
calculation are actually rounded to a given number of decimal places.
Figure 9.4 shows two tables representing the same extract from a
profit and loss account. In both cases all the cells have been formatted
to zero decimal places, but in Table B the ROUND function has been
incorporated in the formulae for cells F15 through F19.
 Table A – Formatting only
Financial Planning using Excel

Table B – Using ROUND Function

98
Figure 9.4 Difference between rounding and formatting cells

The formula entered into cell F15, which can then be copied for the
other line items is:

⫽ROUND(SUM(B15:E15),0)

The effect of the ROUND function can be seen in cell F20. By visually
adding up the numbers in the range F15 through F19 the result is
78,111 whereas the formatting of these cells without the use of the
ROUND function in Table A returns a value of 78,112 in cell F20.
Having applied the ROUND function to a cell any future reference
made to that cell will use the rounded value.

Excel does offer an alternative to the ROUND function in the

Precision as displayed option within TOOLS OPTIONS CALCULATION.
This command assumes that calculations will be performed to the
level of accuracy currently displayed. The danger of using this
command is that when data is changed or added to the spreadsheet
the command is no longer valid and it is then necessary to repeat
OPPORTUNITY
the command to update the spreadsheet – this is another invitation
FOR GIGO
to GIGO.
Spreadsheet 4: Documentation
Spreadsheet developers are notoriously bad at supplying documen-
tation and other supporting information about the plan. There are a
number of features offered by Excel to assist in the documenting of
plans including the INSERT COMMENT command. Figure 9.5 shows a
comment being entered onto a plan – notice how the user name
of the comment author is included. This is useful when a team of
people are working on a system. The presence of a comment is
indicated by a small red triangle on the cell. To read the comment

Financial Planning using Excel

move the cursor over the cell and it will automatically be dis-
played. A word of caution concerning the use of comment boxes –
they take up a considerable amount of space and if used widely
they can make a noticeable difference to the size of a file. To clear
all the comments use the EDIT CLEAR COMMENTS command.

99

[Link]

Figure 9.5 Inserting comments

The provision of a hard copy report showing the logic used to create
a plan is also helpful as this is the ultimate reference point if a
formula has been overwritten and needs to be reconstructed.

In Excel there is a shortcut key to display the formulae which is

TECHNIQUE
CTRL ⫹ ` (accent grave). Alternatively, this can be achieved through
TIP!
the TOOLS OPTIONS VIEW command and then check the Formulas box.

In addition to providing documentation for a spreadsheet system,

looking at the contents of the cells as opposed to the results can
also be a helpful auditing tool. For example, Figure 9.6 highlights
Financial Planning using Excel

Figure 9.6 Report showing formulae

the fact that there are still values embedded in formulae which is
not good practice and is addressed in the next version of the plan.

A third form of documentation which can be particularly useful for

large systems is the ‘sentence at the end of the row’ technique.
100 Requiring less file space than comment boxes, and always on view
it can be useful to have a brief description of the activity taking
place in each row of a plan.

Spreadsheet 5: Minimising absolute values

One of the reasons that spreadsheets have become such an integral
part of the way we do business is the fact that they facilitate quick,
easy and inexpensive what-if analysis. What-if analysis may be
defined as the process of investigating the effect of changes to
assumptions on the objective function of a business plan.

Performing what-if analysis on the opening sales assumption or the

opening price assumption is quite straightforward, involving plac-
ing the cursor on the figure and entering the new value. On press-
ing Enter the spreadsheet is re-evaluated and all cells which refer to
the changed values, either directly or indirectly, are updated.

The success of performing even the simplest what-if analysis is

dependent on the spreadsheet having been developed with the cor-
rect series of relationships. For example, changing the opening
sales value in Figure 9.7 would automatically cause the other quar-
ter sales values to recalculate, as well as the revenue, costs and
 [Link]

Financial Planning using Excel

Figure 9.7 Absolute values restricting what-if analysis

profit lines, because they relate, through the cell references in the
formulae, either directly or indirectly to the sales value in cell B5.

However, as already mentioned this plan incorporates absolute val-

ues in the formulae for sales and costs growth. Furthermore, the
price is a fixed value and has been entered once into cell B6 and the 101
value has then been copied into the other periods. This presents
problems when what-if analysis is required on any of these factors.

Problems with this spreadsheet

Because no growth in the price is required, the opening value of
12.55 has been copied for the four quarters. Whilst this is fine all
the time a price of 12.55 is required, it presents a problem if the
price needs to be changed. With this spreadsheet it would be neces-
sary to overwrite the price in the first quarter and then copy the
new value for the remaining three quarters. The same applies if the
sales growth or the cost factors need to be changed.

Spreadsheet 6: Separating growth and cost factors

To prevent the problems in Spreadsheet 5 from arising, a different

approach to the development of the plan needs to be taken.

In the first instance, all growth and cost factors should be repre-
sented in a separate area of the spreadsheet – even on a different
sheet altogether in the case of a large system with a lot of input.
 The factors can then be referenced from within the plan as and
when they are required. Figure 9.8 shows the adapted layout for
this plan after extracting the sales growth and costs factors.
Financial Planning using Excel

[Link]

102
Figure 9.8 Using cell references for non-changing values

Having the growth and cost factors in separate cells means that the
formulae need to be changed to pick up this information. Figure 9.9
shows the amended formulae for this plan.

Figure 9.9 Amended formulae to take account of extracted growth and cost
factors

Note that the references to cells D15, D16 and D17 are fixed refer-
ences. This is achieved by placing the $ symbol before the column
 letter and row number, i.e. $D$15, and means that when the for- TECHNIQUE
mula is copied the reference to cell D15 remains fixed. A shortcut TIP!
key to add the $ symbols to a cell reference is F4.

In this plan an option in the growth factors has been included for
the price, despite the fact that in this plan the price does not
change. It is important to always think ahead when developing
any plan and although the price does not currently change, it
might be necessary to include a percentage increase in the future.
Having the facility for change built-in to the plan could save

Financial Planning using Excel

time later – and for the time being the growth factor is simply set
to zero.

Removing the growth and cost factors from the main body of a busi-
ness plan is the first step in developing a data input form which
will ultimately separate all the input data from the actual logic of
the spreadsheet. This separation of the data allows the logic cells to
be protected from accidental damage. This is discussed further in
the ‘Template’ section of this chapter.
103

Spreadsheet 7: Optimizing layout

The amount of detail supplied in the plan so far is clearly insuffi-
cient for any real decision-making process.

More detail of the firm’s cost structure would constitute an obvious

improvement and Figure 9.10 shows how this might be incorpo-
rated. This sample plan is obviously small, and even the expanded
plan fits onto a screen, but wherever possible do try and divide
plans into sections that fit comfortably in the screen area, and indi-
cate if there is more information to follow by writing in a prompt
such as press page down for cost analysis.

Some people might find the spreadsheet in Figure 9.10 disjointed

with revenue in one section and the costs in another. Although
good spreadsheet design is essential, personal preference will
always have a role to play in the finished result, and of course the
way in which a system is used and for what purpose will play an
important role in deciding which variables are grouped together.
For example, it is not difficult to restructure this plan as shown in
Figure 9.11.
 [Link]
Financial Planning using Excel

104

Figure 9.10 Expanded cost structure

[Link]

Figure 9.11 Alternative approach for expanding costs

Spreadsheet 8: Arithmetic cross-checks
As already mentioned spreadsheet users are not inherently good at
auditing plans as thoroughly as perhaps they should, and therefore an
important aspect of spreadsheet design is to build into the system
checks on the arithmetical accuracy that will raise the alarm if things
begin to go wrong. This might include validating input data through
the use of an IF function, or performing a cross-check on a calculation.

When creating calculation checks the first step is to select a number

of key items from the model, whose result can be calculated using a

Financial Planning using Excel

different arithmetic reference. For example, in Figure 9.12 below
the year end gross profit has been calculated by referencing the
individual total values in column F and then by totalling the values
in the gross profit row. An IF function is then applied to compare
the two results and if they are not the same the word ‘error’ is
displayed in cell D8.

105

[Link]

Figure 9.12 Cross-check control box

The formulae required in cells D6 and D7, which calculate the year
end gross profit from the plan illustrated in Figure 9.11 are:

For the vertical total

⫽‘P&L ACCOUNT’!F10⫺‘P&L ACCOUNT’!F17⫺‘P&L ACCOUNT’!F19

⫽F10⫺F17⫺F19

and for the horizontal total

⫽SUM(‘P&L ACCOUNT’!B23:E24)

The formula in cell D8 is an IF function that compares the two cells

as follows:

⫽IF(D6⬍⬎D7,“ERROR”,“OK”)
 A macro could also be created that alerts the user should the arith-
metic not balance, probably by a beep and going to a suitable message
screen.

Spreadsheet 9: Charts
It is useful to support the information supplied in business plans
with charts. In the profit and loss account used in this chapter, vari-
ous charts might be useful, for example, to show the relative impact
Financial Planning using Excel

of price and sales volume figures. Although charts can be placed on

the same worksheet as the plan, it is usually preferable to keep
graphs on separate chart sheets. The exception might be if it is
appropriate to view changes on a chart at the same time data in the
plan is changed, or if a spreadsheet is to be copied into a manage-
ment report being created in Word. Figure 9.13 is an example of the
type of chart that might be produced from the plan used in this
[Link] chapter. This chart has been received using the data in Figure 9.11.
106

35000
30000
25000
20000
Value

Revenue
15000
Total costs
10000 Gross profit/loss
5000
0
1 2 3 4
Quarters

Figure 9.13 Three-dimensional graph

Spreadsheet 10: Mutiple sheets

The plan used in this chapter has been a simple quarterly plan, but
in many cases business plans will be larger and more complex.
Figure 9.14 is an extract from a five year quarterly plan. Although it
is not obvious by looking at Figure 9.14, each year this report has
been formatted with a different colour font. This is a useful tech-
nique when working with large models because it enables the user
to quickly know which part of the plan is being viewed or worked
on, without having to scroll around the spreadsheet to see the titles.
This colour coding can then be carried over to summary reports,
and other reports pertaining to the different parts of the plan.

From a design point of view it is preferable to place different

reports associated with a plan on separate worksheets. The report
in Figure 9.15, which has been placed on a separate sheet called
Summary is created by referencing the cells from the yearly totals

Financial Planning using Excel

in the main plan.
TECHNIQUE
To assign a name to a worksheet simply double click on the sheet
TIP!
reference at the bottom of the screen and type in the required name.

107

[Link]

Figure 9.14 Five year extended plan

Figure 9.15 Summary report

 Templates
A business plan that requires time and effort to design and imple-
ment is likely to be in regular use for some time. In addition, the
data in the plan will almost certainly change, either as situations
within the business change, or on a periodic basis. In such circum-
stances it is advisable to convert the developed plan into a tem-
plate, into which different data can be entered whenever necessary.

[Link] A template is a plan that contains the logic required, i.e. the
formulae, but from which the data has been removed. When new
Financial Planning using Excel

data is entered the formulae will be calculated. Figure 9.16 shows

the simplest approach to creating a template. Taking the one-year
quarterly plan used in this chapter the input data and growth fac-
tors have been removed and these cells have been highlighted by
shading the cells. The file can be found on the CD accompanying
the book under the name TEMPLAT1.

When the input cells are set to zero all other cells that are directly
108 or indirectly related to those cells should also display zero. The
only exception to this is if there are division formulae in which
case a division by zero error will be displayed. The act of removing

Figure 9.16 Plan converted to a template

 the data is in itself a useful auditing tool, because if values are
found in any cells this indicates that there is an error in the way
that the plan was developed which can be rectified.

When the template is complete the spreadsheet should be pro-

tected, specifying only the cells into which data can be entered as
unprotected cells. This is a two-step process. First, the cells into
which data can be entered are unprotected using the FORMAT
CELLS PROTECTION command and removing the tick on the Locked
box. The second step is to then enable protection by selecting

Financial Planning using Excel

TOOLS PROTECTION PROTECT SHEET.

It is also important to save the file now as a Template file as opposed

to a Worksheet file. This is achieved by selecting FILE SAVE AS
Template (*.XLT) in the File Type box. The location of the template
file defaults to the directory where other Microsoft Office template
files are located. To use the template, File New is selected which
accesses the Template directory and when a file is selected a copy of
it is opened, leaving the original template unchanged on the disk.
109

Data input forms

A further enhancement that makes working with templates easier
to control is to remove all the data from the main plan and place it
on one or more data input forms which will normally be located
on separate worksheets. Figure 9.17 is a data input form for the

[Link]

Figure 9.17 Data input form

Financial Planning using Excel

Figure 9.18 Amended formulae to reference data input form

quarterly plan, and Figure 9.18 shows the amended formulae in

the plan which picks up the data from the input form.

There are many benefits to be derived from using data input forms
including the fact that the data can be checked more easily.
110
Sometimes it might be possible to design an input form that is com-
patible with a forecasting or accounting system so that the data can
be electronically picked up from the other system without having
to type it in again. Even if this is not possible, the order of items in
the data input form does not have to be the same as the order in
which they are referenced in the logic, which means that the data
input form can be created to be as compatible with the source of the
input data as possible. Furthermore, the worksheet containing the
logic for the plan can be protected, and if necessary made read-
only, in order to maintain the integrity of the system.

It is not a trivial task to change existing systems to be templates

with data input forms, and it will also take a little longer to develop
a new system in this way, as opposed to incorporating the data with
the logic. However, the ease of data input and ongoing maintenance
should make the additional effort worth-while.

Summary
This chapter has considered some of the principal design elements
that should be considered when embarking on the creation of any
business model or plan, be it a financial statement, a budgetary
control system, a marketing model or a forecasting plan. A small
plan has been used for demonstration purposes, and many of the
techniques illustrated become essential when working with larger
plans. Taking time to consider the layout and design of a system
before embarking on its development has been proven by many
users to pay considerable dividends in the long term. In addition it
is worth talking with colleagues who might find a plan useful
before starting development to see whether some additional lines
should be incorporated, as it is always more difficult to add to a
spreadsheet later.

Financial Planning using Excel

111
This Page Intentionally Left Blank
10
Developing a Financial Plan
This Page Intentionally Left Blank
 Planning unifies diverse activities, providing a ‘road map’ for the
undertaking of complete tasks that must be well co-ordinated,
accomplished over extended time frames and inclusive of many
people.

– R.K. Wysocki and J. Young, Information Systems – Management

Principles in Action, 1990.

Introduction

Financial Planning using Excel

Financial planning is a popular spreadsheet application, which
often involves acquiring a sales forecast, perhaps using one of the
models or methods described in Part 1 of this book, and then speci-
fying all the resources necessary to ensure that the forecast targets
are reached. Financial planning can be performed for an entire
organisation, or for a department or division. When separate divi-
sional profit plans are developed they may subsequently be incor-
porated into a corporate plan.
115

The model
The financial plan described in this chapter is a one year quarterly
plan for a manufacturing company. It is a deterministic plan and
assumes point estimates for the opening input values, as well as the
growth and cost factors. In this example all growth rates and cost
factors remain the same for the duration of the plan. The completed
model can be found on the CD accompanying this book under the
name FINPLAN.

The plan consists of six separate worksheets. The first is a Data

Input sheet. For this example all the input data has been grouped
onto one sheet, but in some situations, if there is a large amount of
input, it might be preferable to break the input down into multiple
sheets. The second worksheet is a Profit and Loss Account
that shows how the firms’ income will be made and what costs will
be incurred in generating this income. The Profit and Loss
Appropriation Account, which provides an estimate of tax payable,
the dividends payable and the funds to be transferred to reserve
accounts for future growth is also on this worksheet.

The third worksheet is a Balance Sheet, followed by Funds Flow

Statement, some Ratio Analysis and finally a Cash Flow Statement.
 Getting started
The first step when creating a multi-sheet plan such as this is to col-
lect as much information as possible in terms of the line items, vari-
ables, growth and cost factors, etc., and then to name worksheets and

[Link]
Financial Planning using Excel

116

Figure 10.1 Data Input worksheet (Input)

 enter titles, column headings and line item descriptions. This forms
the basic structure of the plan and makes cross-sheet referencing pos-
sible when formulae are developed. Figures 10.1 through 10.6 show
the six worksheets that make up the full plan with sample data.

Financial Planning using Excel

117

Figure 10.2 Profit and Loss Account worksheet (P&L)

The worksheets have been named as follows: Input, P&L, Balsheet,

Funds Flow, Ratios and Cash Flow. These names will be referenced
by formulae when cross-sheet references are required.

Developing the Profit and Loss Account

The data input sheet in Figure 10.1 shows all the assumptions that
have been made for the Profit and Loss Account. For example, the
sales volume begins at 6000 and increases by 3% per period, the
unit price begins at 10 and increases at 2%, the cost of raw materials
is £4 per unit, etc.

Having prepared the structure for this part of the plan the logic can
now be entered. The opening sales volume value is required in cell
Financial Planning using Excel

118

Figure 10.3 Balance Sheet worksheet (Balsheet)

C4 of the P&L sheet and therefore a reference to the Input sheet, cell
B5 is entered as follows:

⫽INPUT!$B$5

In the second and subsequent periods the sales volume is to grow

by the value entered in cell C5 of the Input sheet. The following
formula is therefore required in cell D4 of the P&L sheet:

⫽C4*(1⫹INPUT!$C$5)

This formula can then be extrapolated for the remaining two quarters.
Note that the use of the absolute reference ensures that the reference
to cell C5 in the Input sheet remains fixed.
 Financial Planning using Excel
119

Figure 10.4 Funds Flow worksheet (Funds Flow)

The total sales volume in cell G4 of the P&L sheet is calculated with
the SUM function as follows:

⫽SUM(C4:F4)
Financial Planning using Excel

120 Figure 10.5 Ratio Analysis worksheet (Ratios)

Figure 10.6 Cash Flow Statement worksheet (Cash Flow)

The remaining lines in the Profit and Loss Account follow the same
principles and Figure 10.7 shows the formulae for this part of the
financial plan.
 Financial Planning using Excel
Figure 10.7 Formulae for Profit and Loss Account

Developing the Profit and Loss Appropriation 121

Account
The assumptions for the profit and loss appropriation account are
again entered on the Input sheet and they involve the dividends and
reserves. As these values can change for each period, the lines on the
Input sheet have allowed for four different values to be entered. In
this example, dividends of 2100 have been entered in the 2nd and 4th
quarters and reserves of 3500 have been entered in the 4th quarter.

The lines in the Appropriation Account on the P&L sheet need to

refer to the appropriate cells on the Input sheet. Figure 10.8 shows
the formulae required for this part of the financial plan.

Figure 10.8 Formulae for P&L Appropriation Account

 Developing the balance sheet
The balance sheet has been developed following the North–South
convention, which means that the assets and liabilities are placed in
what amounts to a vertical column. This is the more usual convention
today with the more traditional, or old-fashioned East–West conven-
tion, where the assets and liabilities are placed side by side, being
much less frequently used.
Financial Planning using Excel

Assets
The opening balances and the assumptions with regard to debtors
and creditors are again found on the Input sheet. The assumptions
made for the debtors and creditors in the balance sheet are that they
are paid in full in the following period. The only cash payments
made are the labour and salary costs.

The fixed assets and the current assets are taken from the Input sheet
122
and do not change over the four quarters of the Balance Sheet. The
Depreciation in the first quarter is calculated by multiplying the open-
ing balance by the Depreciation value for the first quarter in the P&L
sheet. This formula can then be copied for the remaining periods.

The debtors are calculated by taking the turnover for the current
period.

The logic for the cash calculation in the first quarter is as follows:

(Opening cash balance ⫺ Labour ⫺ Salaries) ⫹ Previous periods’

sales)

Therefore, the formula required in cell C16 of the Balsheet sheet is:

⫽(B16⫺‘P&L’!C9⫺‘P&L’!C19)⫹INPUT!B57

Unfortunately it is not possible to simply extrapolate this formula

for the remaining quarters because in the second quarter it is neces-
sary to incorporate the turnover for the first quarter less the trade
creditors. Therefore the formula required in cell D16 is:

⫽(C16⫺‘P&L’!D9⫺‘P&L’!D19)⫹‘P&L’!C6⫺C20

The formula in cell D16 can then be copied to the fourth quarter in
cell F16.
Capital and liabilities
The current liabilities consist of the trade creditors, tax and divi-
dends. The opening balance for the trade creditors is taken from the
Input sheet, but in the first quarter the creditors are calculated by
taking the total direct costs less the labour (which is paid in cash)
plus the total other costs less the salaries and depreciation.
Therefore the formula required in cell C 20, which can be copied for
the remaining periods is:

⫽‘P&L’!C11⫺‘P&L’!C9⫹‘P&L’!C20⫺‘P&L’!C19⫺‘P&L’!C16

Financial Planning using Excel

The opening tax and the dividends are both taken from the Input
sheet and in the first quarter the opening values are added to the
provision for tax made and the dividends respectively in the profit
and loss account. Therefore the formulae required in cells C20 and
C21, which can be copied for the remaining periods are:

⫽B21⫹‘P&L’!C24
123
and

⫽B22⫹‘P&L’!C28

The working capital is the total current assets less the total current
liabilities and therefore the formula required in cell B25, which can
be copied for the remaining periods is:

⫽B17⫺B23

The long-term liabilities consist of debenture, loan and mortgage

payments, which are all taken from the Input sheet and remain the
same for the year.

The total net assets are the total assets less the total long-term lia-
bilities and therefore the formula required in cell B34, which can be
copied for the remaining periods is:

⫽B26⫺B32

Capital consists of share capital, retained profit and reserves. Share

capital is taken from the Input sheet and carried across for the four
quarters. The opening retained profit is also taken from the Input
sheet, but in the first quarter the retained profit from the first quar-
ter of the profit and loss account is included. Therefore the
 formula required in cell C39, which can be copied for the remain-
ing periods is:

⫽B39⫹‘P&L’!C31

The same procedure is required for the reserves by picking up the

opening reserves from the balance sheet section of the Input sheet
and then adding in the reserves from the profit and loss account in
the subsequent quarters.

The total capital and liabilities can then be calculated by adding

Financial Planning using Excel

together the total net assets and the owners’ equity. Therefore the
formula required in cell B43, which can be copied for the remaining
periods is:

⫽B41⫹B32

Funds flow statement

124 The funds flow statement is a tool for understanding how funds are
being used in the organisation. It indicates into which assets money
is being invested or disinvested as well as showing how these funds
have been financed or which sources of finance are being repaid.

The statement of source and application of funds is calculated by

subtracting two consecutive balance sheets from each other. This
will then indicate which assets are increasing or decreasing, and
which sources of funds are being used or being paid back.

The layout of a funds flow statement differs from country to country

and company to company, and the following shows the principles
of how to set up a statement, but readers will need to format the
final report to suit individual requirements.

The basic statement can be quickly created on the spreadsheet by

copying the headings and titles from the balance sheet (Balsheet)
onto the Funds Flow sheet. The first column of the source and appli-
cation of funds statement will represent the second quarter, being the
first quarter values from the balance sheet less the opening balance,
and thus the titles for the opening balance and the first quarter can
be deleted. The following formula can then be entered into cell D5:

⫽BALSHEET!D5⫺BALSHEET!C5

This formula can be copied for all the remaining cells in the statement.
Ratio analysis
There are many different ratios that can be applied to a financial
statement and Figure 10.5 provides a selection that has been calcu-
lated on the Ratios sheet. Figure 10.9 shows the formulae for the first
quarter ratios which have been copied for the remaining periods.

Financial Planning using Excel

125

Figure 10.9 Formulae for ratio analysis

Cash flow statement

Traditional financial statements produced by an organisation are
essentially historic in nature. They show what assets have been
acquired and how they have been funded, as well as showing what
income has been made and how it has been dispersed. Although all
this information is considerable value to management, it does suffer
from this strong historical perspective. To counter this, organisa-
tions developed techniques of forecasting, planning and budgeting
and one of the most important tools available to management is the
cash flow plan.

The cash flow plan shows what money the organisation expects
to receive and from what sources as well as how it intends to
spend it.
 The cash balance row of the cash flow plan is the same as the cash
row of the balance sheet and therefore the formula in cell C3 that is
copied for the remaining periods is:

⫽BALSHEET!B16

The cash-in section includes sales, investments and cash receipts.

In this example there are no investments to be considered and no
payments are received in cash, but the sales are the turnover from
the previous period. Therefore the opening sales figure is a refer-
ence to the opening sales in the Input sheet:
Financial Planning using Excel

⫽INPUT!B57

In the following period the sales from the profit and loss account in
cell C6 can be referenced and this can then be copied for the
remaining periods.

The cash-out section of the cash flow plan includes the cash
expenses, the creditors and repayments. The cash expenses include
126 labour and salaries which means the following formula is required
in cell C11 and can be copied for the remaining periods.

⫽‘P&L’!C9⫹‘P&L’!C19

The opening creditors in cell C12 can be picked up from Cell B20 in
the balance sheet and the reference can be copied for the remaining
periods.

There are no tax or dividend payments, although accommodation for

these has been included in the cash flow plan. The cash balance can
finally be calculated by taking the opening cash balance, adding the
total cash in and subtracting the total cash out. Therefore the formula
required in cell C15, which can be copied for the remaining periods is:

⫽C3⫹SUM(C6:C8)⫺SUM(C11:C14)

Summary
Having spent a considerable amount of time and effort developing
an integrated financial statement such as the one described here, it
is likely that it will be used periodically with different data.
Because the input has been held separately on the Input sheet it is
not difficult to convert this plan into a template. This involves
removing all the data in the Input sheet, or entering zeros into the
cells, which will cause the formulae in the other worksheets to
return zero results (with the exception of the Ratios sheet which
will return division by zero errors). The file can then be saved as an
.XLT template file. The section on templates in Chapter 9 gives
more details on how to do this.

Financial Planning using Excel

127
This Page Intentionally Left Blank
11
Business Plans
This Page Intentionally Left Blank
 A plan . . . is ‘tangible evidence of the thinking of management.’ It
results from planning.

– J.O. McKinsey quoted by G.A. Steiner in Top Management

Planning, Macmillan, 1969.

Introduction
This chapter consists of a selection of different business plans that
illustrate a range of spreadsheet techniques which readers may find

Financial Planning using Excel

helpful when developing their own systems. The business applica-
tions that have been selected for discussion in this chapter represent
only a small number of the possible business spreadsheet applica-
tions and the areas chosen are:

◆ capital investment appraisal (CIA)

◆ break-even point analysis
◆ learning curve costing
◆ economic order quantity 131
◆ sales campaign analysis.

For each of these examples the spreadsheet development approach

and the key formulae are explained in detail in this chapter.
However, it is strongly recommended that readers study these plans
in conjunction with the accompanying CD.

Capital investment appraisal

The CIA model described in this chapter uses both discounted and
non-discounted techniques.

Discounted techniques are based on the time value of money. By

this it is meant that cash received today is more valuable than
cash received at future dates. The rationale of this assertion is
that cash can be invested as soon as it is received and thus imme-
diately begin earning a return. Therefore, the longer it takes to
receive a sum of money, the less value that sum represents in
today’s terms.

Discounting techniques for CIA reduce amounts paid and received

in the future to the equivalent amount paid and received today.
 This is achieved through the application of the following formula
that is multiplied by the sum of money concerned:

1/((1⫹I)n)

where I ⫽ discount rate and n ⫽ number of periods in the future.

There are several Discounted Cash Flow (DCF) measures available in

the spreadsheet, with the Net Present Value (NPV), the Profitability
Index (PI) and the Internal Rate of Return (IRR) being some of the
more frequently used.
Financial Planning using Excel

The CIA plan shown here calculates the NPV and the PI for both a
fixed discount rate (FDR) and for a variable or inflation adjusted
discount rate (VDR). Non-discounted cash flow measures of pay-
back and rate of return are also produced.

The results of the FDR and VDR calculations are presented in such
a way that the investment measures dependent on a discount rate
can be compared and the difference between a fixed and a variable
132 discount rate on the NPV and PI can be evaluated. The model can
be found on the CD accompanying this book under the name CIA.

Net present value

The purpose of the NPV is to calculate the balance between the
trade-off in investment outlays and future benefits in terms of time-
adjusted present monetary values. Thus the formula for NPV is:

NPV ⫽ present value of cash flow in ⫺ present value of investment

NPV is a straightforward way of establishing whether, during the

economic life of a project, a minimum earnings standard can be
obtained.

The NPV may be defined as the difference between the sum of the
values of the cash-in flows, discounted at an appropriate cost of
capital, and the present value of the original investment. Provided
the NPV is greater than or equal to zero, the investment will earn
the firm’s required rate of return. The size of the NPV may be con-
sidered as either a measure of the surplus which the investment
makes over its required return, or as the margin of error which may
be made in the estimate of the investment amount before the invest-
ment will be rejected.
 The interpretation of the NPV should be based on the following rules:

◆ If NPV ⱖ 0 then invest

◆ If NPV ⬍ 0 then do not invest.

Profitability index
The PI is defined as the sum of the present values of the cash-in
flows divided by the present value of the investment. This produces
a rate of return expressed as the number of discounted pounds and

Financial Planning using Excel

pence, or dollars and cents, or any other currency that the invest-
ment will earn for every unit of currency originally invested. The
formula for the PI is:

兺 Present value of benefits

PI ⫽
Present value of investment

Internal rate of return

133
The IRR is the rate of interest which will cause the NPV to be zero.
It is the internally generated return that the investment will earn
throughout its life. It is also sometimes referred to as the yield of
the investment. The formula for the IRR is:

IRR ⫽ i, such that NPV ⫽ 0

The IRR is the most complex of the three discounted cash-flow sta-
tistical measures described here and it needs to be used with care.
The IRR may produce incorrect results if the investment shows neg-
ative cash flows after the first year of the project.

Developing a capital investment appraisal plan

Figure 11.1 shows the completed CIA plan with some sample data. [Link]

The investment reports

The investment reports consist of the simple payback and dis-
counted payback in years and months, the rate of return, the NPV,
the PI and the IRR at a fixed discount rate, as well as the payback,
NPV and the PI at a variable discount rate.
Financial Planning using Excel

Figure 11.1 Capital investment appraisal plan

Some of these reports can be calculated using built-in spreadsheet

functions, but others, such as the payback and the variable discount
rate reports, require some additional calculations that have been
134
grouped together in a separate area of the spreadsheet. Figure 11.2
shows the work area for the payback calculations.

Figure 11.2 Payback work areas

Simple payback
The simple payback refers to the amount of time it takes for the
original investment to be paid back in nominal terms.
Rows 26 through 29 is a lookup table from which the data for the sim-
ple payback is derived. The reason for repeating the year numbers
will become clear shortly, but it is to do with how the HLOOKUP func-
tion operates.

The formulae in the range B31 through D34 could be combined into
a single, nested formula. However, it is preferable to avoid long
complex formulae and to break an operation down into a number of
smaller modules. This makes auditing and amending the formula
much easier.

Financial Planning using Excel

The first step in calculating the payback is to ascertain the year in
which the payback occurs in cell B31. The HLOOKUP function is
used. This function requires three pieces of information.

1. What is to be looked up. In this case it is the amount invested in

cell C3.
2. The location of the lookup table, which is B28 through F29. The
system will then look for the value in C3 across the first row of
the specified table range. If an exact match cannot be found it 135
will pick up the next lowest value. For this function to work
properly in this context, it is necessary for the values in the first
row of the table to be sorted in ascending order.
3. The number of rows to count down for the result, where the first
row of the table is 1.

The following is the formula required for cell B31.

⫽HLOOKUP(C3,B28:F29,2)

Given the data in the plan this formula will lookup 350,000 in the
first row of the table range B28 through F29 and, not finding an
exact match, will find 275,000 as the next lowest, coming down 1
row in the table will return 3 as the result. In other words, payback
is received in year 3.

In order to ascertain which month in year 3 payback occurs, it is

necessary to know the funds received by the beginning of year 3.
The following formula is required in cell D31:

⫽HLOOKUP(B31,B26:F28,3)

This formula looks up the value 3 in the table range B26 through
F28 and then returns the value 2 rows down, which is 275,000.
 The amount of money outstanding is calculated in cell D32 by sub-
tracting the value in cell D31 from the investment in cell C3 which
in this example is 75,000.

To be able to calculate what proportion of a year this represents, the

amount of cash due in year 4 must be calculated as follows:

⫽HLOOKUP(B31⫹1,B26:F29,2)

This produces a result of 180,000. The number of months can then

be calculated by dividing 75,000 by 180,000 and multiplying by 12
Financial Planning using Excel

in cell D34 as follows:

⫽D32/D33*12

Cell B14 in the main investment report area is simply a reference to

cell B32 in the work area and cell D14 is a reference to cell D34 in
the work area.

The calculation of the payback is a good example of when a sub-

stantial amount of work is required to calculate something that can
136
be visually calculated quite easily just by looking at the data on the
screen. However, to ensure that an up-to-date payback is always
reported the logic must be built-in to the plan.

Discounted payback
The simple payback as described above is not regarded as
being an adequate measure of investment performance because
it does not take into account the time value of money. The
discounted payback, which is calculated after the future cash
flows have been discounted, is regarded as a much stronger indi-
cation of the realistic period required to recover the investment.
This calculation requires each cash flow to be individually
discounted and the new discounted values are then used in the
same way as was shown in the simple payback. The formula
required for the discounted cash flow in row 40 of Figure 11.2 is
as follows:

⫽B39/(1⫹$B$37)^B38

As with the simple payback cells B15 and D15 are references to
cells B45 and D48 in the work area.
Rate of return
The rate of return is calculated by taking the average of the cash-in
flows and dividing by the original investment, expressed as a per-
centage. Thus the following formula is required in cell B16:

⫽AVERAGE(D4:D8)/C3

NPV at a fixed discount rate

The NPV function is used to calculate the net present value. This

Financial Planning using Excel

function requires reference to the discount rate and the cash-in
flows. The following formula is required in cell B17:

⫽NPV(B9,D4:D8)–C3

Note that it is necessary to subtract the original investment outside

of the NPV function.

Profitability index (PI) 137

The PI also uses the NPV function, but the result is divided by the
original investment as can be seen in the formula in cell B18:

⫽NPV(B9,D4:D8)/C3

Internal rate of return

The IRR function is used to calculate the IRR in cell B19. This func-
tion requires reference to cash-in flows, but the first cell in the range
must be the original investment expressed as a negative value.
Furthermore, after specifying the data range the function requires a
‘guess’. Thus the formula for cell B19 is:

⫽IRR(F3:F8,0.3)

Note that the original investment and cash-in flows were copied into
column F, with the investment represented as a negative value and it
is this range that has been used for the IRR calculation.

A ‘guess’ is entered for the IRR because the calculation of the IRR is
a reiterative process and the system requires a ‘seed’ from which to
base the calculation. If no guess is entered the system will assume
 0.1 or 10% and this is usually sufficient for a result to be calculated
within 0.00001%. If IRR cannot find a result that works after 20
tries, the #NUM! error value is returned.

If IRR gives the #NUM! error value, or if the result is not close to
what you expected, try again with a different value for guess.

NPV and PI at variable discount rate

A separate work area is required to be able to calculate the NPV
Financial Planning using Excel

using a discount rate that varies from year to year. Figure 11.3
shows the work area required.

138

Figure 11.3 Variable discount rate work area

The NPV function assumes a constant rate of interest over the dura-
tion of the investment and therefore to accommodate an interest or
discount rate that can change each year, the PV function needs to be
used. In this context, the PV function is used one year at a time and
the discounted value is picked up year by year by a new PV func-
tion which adjusts it appropriately. In the example, the cash-in
flows have to be discounted over the five years.

In cells B51 to F51 the nominal cash-flow amounts are discounted

for one year using the following formula:

⫽PV(B$11 ⫹ $B$9,1,⫺B50)

Note that the PV function assumes that the cash-flow value is

negative (i.e. a credit) and in the above formula this has been
reversed by the negative reference to B50. The reference to B11
has the row fixed with the $ which means that the formula can be
copied holding the reference to row 11 absolute but changing the
column reference. Figure 11.4 shows the formulae in the first two
columns.
 Financial Planning using Excel
Figure 11.4 Formulae for calculating NPV at a variable discount rate

In rows 52 and 55 each future cash flow from year 2 to 5 is again

discounted using a unique interest rate for each year until the cash
flow in year 5 has been discounted five times and the cash flow in
year has been discounted four times, etc. The individual cash
streams are then summed and in the usual way the investment is
subtracted from this amount to produce the net present value.
139
There are clearly other financial reports that could be calculated
from this capital investment plan, but those discussed here are
intended to give the reader some exposure to some of the more fre-
quently encountered issues.

Learning curve costing

The learning curve costing model is a deterministic plan that demon-
strates how unit cost of production will vary as a result of improve-
ments in the production process. The underlying assumption of this
model is that the efficiency of labour and material utilisation will
improve as the organisation learns from its experience in the manu-
facture of the product. Thus, unit costs will decrease as improve-
ments in labour efficiency and material usage are effected. The
model is on the CD accompanying this book under the name LEARN.

Developing the plan

The plan commences with a data input form in which all the key
variables are specified. The first year production figure is required
that can be held constant over time in order to see the effect of the
 improvements in material and labour. Alternatively, the produc-
tion figure can be increased in order to demonstrate the total effect
of an increase in the scale of production, as well as improvements
in efficiency.

The input data required for the plan are the first year’s production,
material costs per unit, labour cost per unit, fixed costs, annual
growth production, learning curve effect for materials and labour
and the price. The horizon for this plan has been set to 10 years.

The learning curve effect for materials is the percentage improve-

Financial Planning using Excel

ment in the use of materials that can be achieved per annum

over the life of the project. This figure should also include
improvement that the firm may achieve through a better buying
policy, either due to eventual economies of scale or from finding
less expensive sources of the materials. The learning curve
effect for labour includes increases in efficiency due to better
skills, a higher degree of automation, as well as improved internal
procedures. Figure 11.5 shows the plan with some sample data.
140

[Link]

Figure 11.5 Learning curve costing model

The sample data used in Figure 11.5 has no annual growth rate in
production. The effect of this is to highlight the impact on the
profit of the learning curve effect alone. If a growth rate in produc-
tion is specified then the overall profit improvement would be even
more dramatic. Therefore a formula is entered into cell C14 and
copied for the remaining years to accommodate a possible growth
in production which is:

⫽C13*(1⫹$B$7)

The formula for the variable costs in column E begins in year 1 by

adding the material costs and labour costs and multiplying by the
production. Thus the formula in cell E13 is:

⫽($B$4⫹$B$5)*C13

Financial Planning using Excel

In year 2 the learning curve effect for materials and labour is taken into
consideration when calculating the variable costs, and therefore the
formula in cell E14, which can be copied for the remaining years is:

⫽($B$4*(1⫺$B$8)^B13⫹$B$5*(1⫺$B$9)^B13)*C14

The unit cost is calculated by adding the fixed and variable costs,
and dividing by the production. Thus the formula in cell F13,
which can be copied for the remaining years is:
141
⫽(D13⫹E13)/C13

The revenue is the price multiplied by the production and therefore

the formula in cell F13, which can be copied for the remaining years is:

⫽C13*$B$10

The profit is the revenue less the production multiplied by the unit
cost and therefore the formula in cell G13, which can be copied for
the remaining years is:

⫽G13⫺(F13*C13)

Finally, the profit improvement percentage is calculated in column H.

Clearly, no improvement can be produced in year 1 and so a zero is
entered into cell H13. In year 2 the improvement percentage is calcu-
lated by subtracting the profit in the previous year from the profit in
the current year and dividing by the total costs. Therefore the formula
required in cell H14, which can be copied for the remaining years is:

⫽(H14⫺H13)/(D14⫹E14)

The cumulative effect of the learning curve can be seen not only by
the percentage profit improvement, but also by the reduction in
unit costs. The chart in Figure 11.6 illustrates this well.
 100.00

90.00

80.00

Unit cost
70.00

60.00

50.00
Financial Planning using Excel

40.00
1 2 3 4 5 6 7 8 9 10
Years

Figure 11.6 Learning curve costing chart

Break-even analysis
142
The break-even analysis model is a deterministic plan that calcu-
lates the volume at which the total costs are equal to the total
revenue. The model is on the CD accompanying this book under
the name BREAKEVEN. This level of volume is defined as the break-
even point. The break-even point is derived by calculating the
contribution per unit sold, which in turn is defined as the unit
selling price less the unit variable cost. The unit contribution is
then divided into the fixed costs and the result is the number of
units that must be sold for the contribution to absorb the total
fixed costs.

The break-even point is not a stationary concept. The volume

required in order to pay the total cost continually changes over
time due to changes in various costs and prices.

The plan shown here is designed to demonstrate the effect of infla-

tion on the break-even point, which is achieved by providing a
growth factor for the fixed costs, the variable costs and the price.
When these figures have been entered the break-even point will
automatically be extrapolated for four years.

The completed model is shown in Figure 11.7.

 [Link]

Financial Planning using Excel

Figure 11.7 Break-even point analysis

The formulae required to calculate the break-even point for

the plan in Figure 11.7 are not complex and are shown in
Figure 11.8.
143

Figure 11.8 Formulae for break-even analysis

The break-even analysis shown here assumes a single-product situ-

ation and frequently this is not the case. Where multiple products
are involved the fixed costs or overheads must first be apportioned
and then a break-even point calculated for each product or product
category. The result will be a break-even point statement for the
firm as a whole, which will include a series of volumes, one for
each product.
 Economic order quantities (EOQ)
There are a range of mathematical techniques available to assist with
most aspects of production management, and particularly with the
management and control of stock or inventory levels.

The EOQ model calculates the most efficient lot size in which inven-
tory should be purchased. It is a production management technique
that is used in many firms. The model is on the CD accompanying
the book under the name EOQ.
Financial Planning using Excel

This is an optimising model which is based on the assumption that

the larger the order placed the lower the costs of ordering will be.
At the same time, the larger the order placed the higher the carrying
costs of stock will be. The EOQ model calculates an optimal level
between these two conflicting cost curves.

The input required for the EOQ model are the annual usage in units
(AUU), the unit price (UP), the variable cost per order (VCO) and the
144
holding costs as a percentage (HC%). The formula for determining
the EOQ is:

EOQ ⫽ √(2*AUU*VCO/UP*HC %)

In addition to indicating the lot size, this model is also used to

evaluate a supplier’s special offer to see whether or not it is worth-
while to purchase in volume in order to obtain a discount.

Developing the EOQ plan

The plan considers three scenarios. The first is the EOQ with no
supplier discount policy, the second calculates the EOQ with a sup-
[Link] plier discount policy and the third brings in some additional
assumptions to calculate the EOQ with lead times and safety stock.
Figure 11.9 shows the completed plan.

EOQ with no supplier policy

The required input for calculating the EOQ with no supplier
policy is the annual usage in units, the unit price, the holding
 Financial Planning using Excel
Figure 11.9 Economic order quantity

cost as a percentage and the variable cost per order. Using

the equation for EOQ described above the formula required in
cell B8 is: 145

⫽ROUND(SQRT((2*B3*B5)/(B4*B6)),0)

The ROUND function has been incorporated in the above

formula to ensure that the nearest whole number is returned as
the EOQ.

EOQ with a supplier policy

One additional item of input is required for the calculation of the
EOQ with a supplier policy and that is the minimum order require-
ment. For this example the unit price has been reduced from 60 to
47.50, but the supplier demands a minimum order of 1250 units.
The formula required in cell C8 is therefore:

⫽ROUND(MAX(SQRT((2*C3*C5)/(C4*C6)),C7),0)

The MAX function in the above formula compares the result

of the SQRT part of the calculation with the minimum order
and returns the larger number. Therefore if the annual usage is
very large it will be more economical to purchase more than
1250 units.
 Cost of inventory
It is important to know the cost of inventory both with and with-
out the supplier discount policy in order that a comparison can be
made to ascertain whether it is worthwhile accepting the supplier
discount. Figure 11.10 shows the formulae for this part of the
plan.
Financial Planning using Excel

146

Figure 11.10 Formulae for calculating the cost of inventory

As can be seen from Figure 11.10, although the cost of inventory is

higher for the supplier policy scenario, the overall costs are still
lower due to the lower selling price.

Lead times and safety stock

As it is likely that there will be a lead time between placing and
receiving an order, this should be taken into consideration and a
safety stock level be set which ensures that an order will be placed
at the right time. The implications of this are calculated in the
third part of this plan. Three further assumptions are made; the
lead time in weeks, the possible delivery delay in weeks and
the possible percentage unforeseen demand. The calculations here
have been done for no supplier policy, but could equally well
include the supplier policy. Figure 11.11 shows the formulae
required for this part of the plan.
 Financial Planning using Excel
Figure 11.11 Formulae for calculating lead times and safety stock

Sales campaign appraisal

147
The sales campaign appraisal model is a deterministic plan that
evaluates the cost of conducting a promotional sales campaign over
a number of weeks. The plan reports on the likely number of
prospects that will attend sales presentations and how many will
be converted to sales. The model is on the CD accompanying the
book under the name CAMPAIGN.

The sales campaign plan has been designed for 10 weeks of analysis
after the initial promotion and requires the following input:

◆ initial promotion cost

◆ duration of primary promotion in weeks
◆ cost of a sales presentation
◆ revenue per unit sold
◆ cost per unit
◆ conversion rate to sale at presentation as a percentage
◆ estimate of the percentage of enquiries converted to sales pres-
entation
◆ estimate of the number of enquiries per week after promotion
◆ estimate of the number of prospects attending sales presentation
who will eventually purchase.

Figure 11.12 shows the completed model with sample data.

 [Link]
Financial Planning using Excel

Figure 11.12 Sales campaign appraisal model

Developing the sales campaign plan

The first step in the plan is to enter all the headings so that there is a
148 structure for the model. Some sample data should be entered in order
that the logic for the formulae can be checked as it is being developed.

The number of enquiries received each week after the initial promo-
tion is an estimate. In this example it is assumed that the effects of the
promotion increase over a five-week period and then begin to drop off
until there are only a few enquiries at the end of the ten weeks.

The estimated numbers of sales presentations are assumed to be

linked with the numbers of enquiries. In this example it is esti-
mated to be between 30 and 40% of the previous week’s enquiries.
The values in the model are derived by a random number generator
used in conjunction with a one-period lag on the number of
enquiries. The formula entered into cell D13, which is copied for
the remaining weeks is:

⫽INT(RAND()*0.1)⫹0.3)*B12

The RAND function in the above formula generates a random num-

ber between 0 and 1. However, by manipulating the result as shown
in the above formula the result will always be between 30 and 40%
of the previous week’s number of enquiries.

The problem with the RAND function is that it is re-evaluated every

time the spreadsheet is recalculated and therefore different values
will be returned. Therefore having produced a series of estimated
 values these cells should be converted to values only. This is
achieved by copying the range with the Edit Copy command and
then selecting Paste Special Values. This removes the underlying
formula leaving only the results in the cells.

The units sold are calculated by multiplying the number of sales pre-
sentations by the conversion to sale percentage. The following formula
is required in cell F13, which can be copied for the remaining weeks:

⫽INT(D13*$E$7)

Financial Planning using Excel

The INT function has been incorporated into the formula to ensure
that the result is rounded down to a whole number.

The cost per presentation is calculated by multiplying the cost per

sales presentation by the number of presentations, adding that to
the total cost of promotion and dividing by the cumulative sales
presentation. The formula in cell H13, which is copied for the
remaining weeks is:

⫽($E$4⫹(D13*$E$6))/E13 149

The marketing cost per sale is the promotion cost divided by the
cumulative sales, which means the following formula is required in
cell I13:

⫽$E$4/G13

The total costs are calculated as the promotion costs plus the cost
of the sales presentations that week plus the production costs for
units sold. The formula required in cell J13 is therefore:

⫽$E$4⫹($E$6*D13)⫹(G13*$E$9)

The total revenue is the number of units sold multiplied by the rev-
enue per unit and the profit on the campaign is therefore calculated
as the total costs less the total revenue.

Summary
In this chapter a number of different business plans have been devel-
oped in order to demonstrate some of the applications for which a
spreadsheet can be used. It is hoped that readers will find some of the
models directly applicable but some of the techniques and functions
shown can also be applied in a number of complimentary areas.
This Page Intentionally Left Blank
12
What-if Analysis
This Page Intentionally Left Blank
 Good judgment is usually the result of experience. And experi-
ence is frequently the result of bad judgment.

– R.E. Neustadt and E.R. May, Thinking in Time, 1986.

Introduction
What-if analysis may be defined as the technique of asking specific
questions about the result of a change or of a series of changes to
assumptions in a model or a business plan.

Financial Planning using Excel

What-if analysis has been performed manually for decades before
the arrival of computers. In the spreadsheet environment, however,
it is a direct product of the fact that once a model or plan has been
entered into the computer, it may be recalculated again and again.
It allows the user to change assumptions concerning input data or
input relationships, and to recalculate a model to see the impact of
these changes on critical output values.

Typical what-if questions might be to ask what effect will a 2% 153

increase in direct labour costs have on profit and return on invest-

ment? What effect will a further 30 day delay in receiving cash from
the debtors have on the overdraft and/or return on investment?

What-if questions may be considered one at a time, or several at a

time. If it is necessary to investigate the effect of two simultaneous
changes in the assumptions, then it is usually advantageous to also
consider these changes in isolation, i.e. one at a time, so that their
individual effects, as well as their joint effects, will be known.
Irrespective of whether single or multiple changes to input are
made, several factors or objectives will usually be monitored.

In addition to what-if analysis, there are two other related concepts

which should also be considered. These are goal seeking or backward
iteration and sensitivity analysis.

Goal seeking is a technique whereby a model calculates the value of

an input variable that is required in order to achieve a stated output
objective. For example, using goal seeking the system could calcu-
late the level of sales required for a return on investment of 25%.
Thus the goal-seeking procedure requires an input variable, which
is usually considered the models’ output, and the result is the value
for a variable that is normally input to the model. Excel has a built-
in command for this, which will be discussed later in this chapter.
 Sensitivity analysis is a technique that ascertains the relative impor-
tance of specified input variables in a business scenario or plan. This
is achieved by calculating the result of a number of relatively small
changes in the specified input factors on the objective function, or
output of the business model. These input and output changes are
then compared to ascertain which variables have the greatest impact.
The user of sensitivity analysis is concerned to establish whether a
5% increase in raw material costs, or a 2% increase in labour costs
will have the worse impact on the profit of the business.
Financial Planning using Excel

Sensitivity analysis and what-if analysis are quite different tech-

niques, although the terms are often used interchangeably.

Three approaches to what-if analysis

With spreadsheet models there are several possible levels of what-if
and associated analysis available, which can be described in terms
of the following three categories:
154
1. manual what-if analysis on opening assumptions
2. data tables
3. backward iteration or goal seeking.

Each of these will be considered separately in this chapter.

Manual what-if analysis on opening assumptions

Manual what-if analysis on opening values is the simplest case and
is performed by placing the cursor on the appropriate input cell
and entering a new assumption. Because spreadsheets are devel-
oped using cell references, all cells which refer directly or indi-
rectly to the cell being changed will be recalculated based on the
revised input. Providing recalculation is set to automatic, the
spreadsheet is immediately recalculated and the effect of the new
assumption can be seen. This approach to the what-if technique is
especially suitable when first-period data assumptions such as
price, opening sales volumes, or growth factors and cost factors are
to be changed.

Figure 12.1 shows the 12 month business plan that will be used
throughout this chapter and Figure 12.2 shows the effect that
reducing the opening sales volume has on the net profit. This plan
can be found on the CD accompanying the book under the name
BUSPLAN.

[Link]

Financial Planning using Excel

155

Figure 12.1 Business plan before performing what-if analysis

Figure 12.2 Business plan after reducing opening volume

 Providing any growth or cost factors have been separated from the
model, any input assumption can be changed using this tech-
nique. Furthermore if a data input form has been created as was
shown in Chapter 9, it is easier to see the input data that can be
changed.

Of course in some situations it may be necessary to make multi-

ple changes to the data to derive a particular result. In this case it
can be difficult to know which change is having the greatest
impact on the result. In addition care needs to be taken not to
Financial Planning using Excel

accidentally overwrite the file if it is still required in the pre-

what-if version.

Occasionally it might be necessary to see the effect of changing the

actual logic of a plan. Extreme care must be taken in this case
because changing the logic of plans that have been tried and tested
can have knock-on effects that might not be obvious to the individ-
ual making the changes.

156

Data tables
Excel provides a powerful what-if facility referred to as data
tables. Data tables allow a range of what-if questions to be calcu-
lated at one time, by setting up a table in which a range of possi-
ble input values are specified together with a reference to
required output. For example, a data table can be established to
report the effect of changing the sales volume growth rate on the
gross profit, net profit, return on investment and cash in bank for
all integer growth rates of sales volume between 0.5 and 10%. A
key advantage to managing what-if questions with data tables is
that the table sits alongside the original plan, which is not altered
in any way.

There are two main types of data table, which are referred to
as the one-way table and the two-way table. A one-way table
allows a single input factor to be analysed against as many output
variables as the user requires. The input factor is a cell in the
plan that is to change and the output variables are those cells
on which you want to see the effect of the change. For example,
in the business plan used as the example in this chapter, the
input factor might be the opening volume and the output variables
 might be the gross profit and net profit figures for July and
December.

In the case of a two-way table, two input factors and a single output
variable can be specified. In the business plan for example, the
input factors might be the opening sales volume and the opening
price and the output variable might be the year-end net profit.

Creating a one-way data table

Financial Planning using Excel

In the case of a one-way data table, a range of input values must
first be entered into a suitable area of the spreadsheet, and refer-
ences to the formulae on which the analysis is to be performed
must be made.

Figure 12.3 shows the outline of a table to analyse varying volume

rates against the gross profit and net profit before tax for December
in the business plan.

N.B. Remember to re-load the original [Link] before com- 157

mencing this exercise.

[Link]

Figure 12.3 Outline of a one-way data table

The values in the range A43 through A53 can be entered either by typ-
ing in the values, using a Fill technique or they maybe the result of a
formula. In this example the values are in ascending order at regular
intervals, but this is not necessary for the command to work. Cells B42
and C42 contain references to cells M14 and M26 in the business plan,
which are the gross profit and net profit figures for December.
 The next step is to highlight the table range, which in this case is
A42 through C53. The DATA TABLE command is then selected which
will bring up the dialogue box shown in Figure 12.4.
Financial Planning using Excel

Figure 12.4 Excel data table dialogue box

158
This example is a one-way table and the input data is referenced
in a column. Therefore it is necessary to make a reference to the
cell in the main model into which the input values are to be
entered. In this case, a reference to the opening volume in cell B5
is required.

On clicking OK the table is calculated and will be displayed as

shown in Figure 12.5.

Figure 12.5 Results of a one-way data table

 The results produced by the table is the equivalent of having
entered 8000 into cell B5 and then recording the gross profit and
net profit figures for December, then 8250 is entered into cell B5
and the gross profit and net profit figures for December is again
recorded. This reiterative process would continue until all the
required values for sales volume had been entered and the profit
results recorded. The DATA TABLE command performs this reitera-
tion extremely quickly.

As has been mentioned previously, it is important to always have a

Financial Planning using Excel

way of cross-checking the results of any analysis in order to validate
the figures. The table in Figure 12.5 can be checked by looking at
row 41 which gives the results when the opening volume is 9750 –
as it is in the main plan – and these values should correspond to the
results in cells B33 and C33. It can be seen from Figure 12.5 that the
results do correspond.

Creating a two-way data table

159
A two-way data table requires input data for two variables. For
example, to analyse a combined variation of opening sales
volume figures and unit prices on the net profit before tax
in December in the business plan, the table specifications in
Figure 12.6 are required. In this case only one ouput variable is
allowed – which is a reference to cell M26 (December net Profit)
in cell B60.

Figure 12.6 Input data for a two-way data table

 The table area is selected in the same way as for the one-way table,
but when the dialogue box is accessed it is necessary this time to
specify the column input cell as the sales volume in cell B5, and the
row input cell as the unit price in cell B6. The results of this table
are shown in Figure 12.7.
Financial Planning using Excel

160
Figure 12.7 Completed two-way data table

The results of this table are interesting, as it is clear that if the

unit price is reduced to £45, a minimum of 9500 units must be
produced to be in profit by the end of the year. But, if the price
can be held at 50, the opening sales volume can be as little at 8000
and still leave a net profit at the end of the year.

In validating the accuracy of this table by taking a sales volume

figure of 9750 and a price of 50, which is apparently the data in the
main model, the resulting net profit in the table should be the same
as the net profit in the model (which is shown in cell A55 of the
table). Looking at the table, this is not the case as the net profit is
showing 56,120 as opposed to the 50,701 that would be expected.
Closer examination of the main plan will show that the unit price is
actually 49.5, but the row has been formatted to a whole number
and it therefore displaying 50. If the opening price in the model is
changed to 50, the net profit for December in the model will be
56,120 as it is in the table.

Unusually the Data Table command places a formula using the

TABLE function into each of the cells in the table range. Therefore if
 any changes are made to data in the main plan, or to the input or
output ranges, the table will automatically be recalculated to reflect
the changes. This also means that there can be multiple data tables
on a spreadsheet and all can be kept up to date simultaneously. In
fact sometimes when working with large tables, or many tables, it is
not always desirable to have the tables recalculate every time an
adjustment is made to the main plan and therefore by selecting
TOOLS OPTIONS CALCULATIONS and then selecting Automatic except
Tables means that it will be necessary to press F9 to recalculate the
data tables.

Financial Planning using Excel

One possible disadvantage of using data tables in Excel is the fact
that they need to be located on the same worksheet as the input data,
which effectively means that it is not possible to have a separate
worksheet on which all the data tables for an application are placed.

Backward iteration or goal seeking

161
Having developed a business plan for the purposes of profit plan-
ning, the results calculated are not always entirely satisfactory to
the decision makers. A typical example might be when the profit
plan shows a projected 10% return on investment (ROI), but the
management require a 15% ROI. In these circumstances what-if
analysis can be performed on the basic assumptions in the model,
in order to determine what volume of sales, for example, would be
required to achieve the desired ROI for the business.

This procedure can, however, be a lengthy one if a large number

of changes to the assumptions have to be made before the model
produces the required results. An alternative method of finding
appropriate input values for a required result or outcome are to use
the Goalseek and Solver tools provided by Excel.

Using the goal seeking feature

In the original business plan the net profit in December is £50,701.
The Goal Seek command can be used, for example, to consider
what opening sales volume would be required in order for the
December net profit to be £60,000 (taking all the other assumptions
made in the plan into consideration).
 Goal Seek is accessed by selecting TOOLS GOAL SEEK and the dia-
logue box shown in Figure 12.8 is displayed and has to be filled out
with the appropriate information.
Financial Planning using Excel

Figure 12.8 Excel Goal Seek dialogue box

On clicking OK the spreadsheet is recalculated and the opening

162
volume required in order to attain a year-end net profit of £60,000
is displayed in cell B5. In this case the answer is 10,120.

Goalseek is a quick to apply and useful tool, but is limited in func-

tionality because it can only be applied to a single variable and the
target value must be set as an actual value – no formulae reference
or conditions are possible.

Use Solver for optimising

If a more complex analysis is required, the SOLVER command may
be useful. With Solver a number of different cells in different parts
of the spreadsheet can be changed and constraints can be specified
to ensure certain parameters are met, such as units produced in a
production model cannot exceed a given number or advertising
expenditure cannot be a negative value. The output cell can be a
specified value or it can be the maximum possible solution or the
minimum possible solution, so that, for example, the maximum
profit for varying units of production can be found or the minimum
profit within the same constraints could be returned.

Solver requires the model to be set up with the required data and
constraints before the analysis can be performed. Figure 12.9
shows a salesman’s productivity model that will be used in this
example. This file can be found on the CD accompanying this book
under the name SALESMAN. The aim is to find the maximum profit
that can be made by the sales team. Note that the constraints
in Figure 12.9 are at this stage only text entries for information
purposes.

[Link]

Financial Planning using Excel

163

Figure 12.9 Salesman’s productivity model before optimising

Each sales person has a minimum quota that he or she is

required to attain and the firm is dependent on a total sales fig-
ure of 150 units. Each sales person is responsible for a different
product and the profit per sale, which is different for each prod-
uct, has been entered into the model in column C. The adjustable
cells are B5 through B8 which represent the number of sales
per person.

Some values must be entered into these cells before using Solver,
but they will be replaced during the analysis.

Having set up the spreadsheet, TOOLS SOLVER is selected and the

Solver Parameters dialogue box is displayed as shown in Figure 12.10.

The target cell is the total sales in cell D9 and the intention is to
maximise this value by changing cells B5 through B8. These
changes are subject to the constraints that have been specified on
the spreadsheet in Figure 12.11.
Financial Planning using Excel

Figure 12.10 Excel Solver dialogue box

164

Figure 12.11 Spreadsheet after using Solver

Figure 12.11 shows the spreadsheet after Solve is selected from the
dialogue box.

It is sometimes possible to specify a problem that has no solution,

which means that there is no set of adjustable cell values that will
satisfy all of the constraints made. For example, if an additional
constraint of total profit to be greater than 10,000 is added to the
above example, Solver attempts to find a solution and then reports
that no feasible answer could be found.
 The example used here to illustrate Solver is a simple one and
serves merely as an introduction to a feature that can perform very
sophisticated analysis. Possible applications include solving simul-
taneous linear and non-linear equations, optimising investment
yield, production level planning, staff scheduling models, etc.

Summary
The ability to perform what-if analysis on a business plan provides

Financial Planning using Excel

the necessary flexibility that enables financial managers and account-
ants to become power users of Excel. The speed of recalculation and
the ease of change make this perhaps the single most important
reason for the success of spreadsheet technology. It is important to
remember, however, that the success of any kind of what-if or sensi-
tivity analysis is entirely dependent on the correct development of
the basic plan and the ability to interpret what the results might
mean.
165
The goal seeking and solver features allow for powerful optimising
techniques to be applied to a range of business plans.
This Page Intentionally Left Blank
13
Risk Analysis
This Page Intentionally Left Blank
 There is an intrinsic impermanence in industry and indeed the man-
agement task is to recreate the company in a new form every year.

– Sir John Harvey-Jones, Making it Happen – Reflections on

Leadership, 1988.

Introduction
Most business plans are deterministic, which means that they rely on
the use of single point estimates for input data and assumptions.

Financial Planning using Excel

Under conditions of uncertainty, which is the most common environ-
ment in which business plans are developed, it can be difficult to pro-
duce accurate estimates using a single point approach and in such
situations it would be preferable to specify input data as ranges.

For example, to say that the sales volume for the next period will be
between 8500 and 12,500 will offer a greater probability of being
right than a single point estimate of, say, 10,000. Similarly, to specify
the average sales price as being between 45 and 52 will often have a 169
greater chance of producing useful results than having to depend on
a single projection of 50.

The structure of a deterministic plan, by its very nature, cannot cope

with input data specified as ranges. However, it is possible to develop
a model which enhances a deterministic, single-point estimate plan to
allow data in the form of ranges to be incorporated, and which in
effect converts the plan from a deterministic to a probabilistic, stochas-
tic or risk analysis model. Risk analysis is also sometimes called prob-
abilistic modelling, stochastic modelling or Monte Carlo modelling.

The principle of this type of modelling is to produce a probability dis-

tribution of the required result. This is achieved by randomly select-
ing values between the specified ranges and collecting the result after
each calculation. This recalculation of the plan is repeated many
times (hundreds or thousands) and a frequency distribution is then
calculated on the output. The results of probabilistic plans are usually
best accessed in the form of a chart and by examining the shape of the
curve and the extent of the spread of the output.

For example, to see the effect of a range of input data for the
investment amount in the CIA model on the NPV at a fixed interest
rate, it would be necessary to recalculate the model using different
 investment amounts and to collect the NPV result for each calcula-
tion. After a considerable number of recalculations, preferably
thousands, a frequency distribution of the results is created and a
graph is drawn. This graph will, in general, be a bell-shaped curve
and the precise shape of the curve will reflect the degree of risk
that is present in the investment based on the input data ranges.

Preparing a plan for risk analysis

Financial Planning using Excel

It is usual to begin risk analysis from a fully tested deterministic plan

or model. This is because it is difficult to develop a plan and be able
to test the results of the underlying logic when ranges of data are used.

However, by carefully designing the original plan it is not difficult

to adapt it for risk analysis and the example in this chapter uses the
CIA plan developed in Chapter 11.

The completed risk analysis model will use four separate work-
170 sheets. The first sheet, called CIA Model, is the original CIA plan.
The amended data from the original plan is on a sheet called Risk
Model, the input form worksheet is called Input, the risk analysis
results worksheet is called Results and the chart is on a worksheet
called Risk Chart. The file can be found on the CD accompanying
this book under the name RISK.

Changing the original plan

Two changes are required to the basic plan. These are for the input
form to accommodate ranges of data and for the logic of the plan to
incorporate a random number generator using the range data.

The selection of the data for insertion in the plan is based on either a
[Link] probability distribution or, in the case of the risk analysis described
in this book, a random number generator. The specification of proba-
bility distributions for risk analysis is beyond the scope of this
book, and thus it is assumed that the data will be specified as simple
maximum and minimum values and that all possible obtainable
results are of equal probability. This is referred to in statistical jargon
as rectangular distributions.

Figure 13.1 shows the data input form for risk analysis on the CIA
model. For this example the variable data is the investment amount,
 Financial Planning using Excel
Figure 13.1 Input form for risk analysis

the cash-in flows, the fixed and the inflation adjusted discount rates.
When entering data for the risk analysis it may not be appropriate to
171
use all these variables and in this case the same value can be
entered for the minimum and the maximum value. In this example,
an option for performing risk analysis on three variables has been
given. These are NPV at a fixed interest rate, IRR and NPV at a vari-
able discount rate.

Cells and ranges in the output selection part of the input form need to
be named for future reference in the Results sheet. Therefore using the
Insert Name Define command names representing the three variables
that the risk analysis can be performed on are assigned as follows:

E17 NPVF
E18 IRR
E19 NPVV

References to the formulae for these variables in the main model on

the Risk Model sheet should be entered into cells F17, F18 and F19
as follows:

F17 ⫽‘RISK MODEL’!C19

F18 ⫽‘RISK MODEL’!C21
F19 ⫽‘RISK MODEL’!C24

At this point, cells F17, F18 and F19 will be displaying the current
values for the net present value, the IRR and the net present value
 at a variable discount rate. As it is not necessary for the user to see
the value in these cells, selecting them and changing the font
colour to white can hide them.

The range D17 through F19 is named ‘looktab’ and will be refer-
enced as a lookup table in the Results sheet.

Incorporating the RAND function

In order to select values at random from within the specified ranges
Financial Planning using Excel

the referencing cells in the main model use the RAND function in
the following way:
TECHNIQUE INT(RAND * (MAXIMUM ⫺ MINIMUM) ⫹ MINIMUM)
TIP!
The RAND function generates a value between 0 and 1, never actu-
ally being 0 or 1. Looking at Figure 13.1 the range of values for the
investment has been specified as between 350,000 and 400,000 and
so if RAND returned a value of 0.56125, the above formula would
172
calculate as:

INTEGER OF (0.56125 * (400,000 ⫺ 350,000) ⫹ 350,000)

which will give an answer of 378,062. In fact the result will always
be a value between the specified ranges.

Figure 13.2 shows some of the formulae in the main model incorpo-
rating the random number generator.

Figure 13.2 Incorporating a random number generator

 The system generates a different random number every time the
spreadsheet is recalculated and thus by pressing F9 a different set
of values will be returned in the main model and in turn the invest-
ment reports will be recalculated using different data each time.

The results worksheet

The results of recalculating the model are collected on the results
worksheet together with some summary statistics and a frequency

Financial Planning using Excel

distribution table. Figure 13.3 shows a section of the results spread-
sheet after performing the analysis on the IRR.

173

Figure 13.3 Results worksheet

In order that the risk analysis will be performed on the variable

marked with an X on the Input worksheet, formulae are required in
cells B3 and B4. A VLOOKUP function is used in both cases. Cell B3 is a
reference to the variable label and the following formula is required:

⫽IF(ISERR(VLOOKUP(“X”,LOOKTAB,2)),“MAKE A SELECTION”,
VLOOKUP(“X”,LOOKTAB,2))

The VLOOKUP in this formula looks for X in the first column of the
table range looktab and returns the contents of the cell one column
to the right. If something other than X is entered, the VLOOKUP
function will return an error. For this reason the formula begins
with ISERR which means that if the result of the VLOOKUP is an
error, the text “Make a selection” will be returned, otherwise the
result of the VLOOKUP function will be returned, which in the case
of Figure 13.3 is IRR.
 A similar formula is required in cell B4 to pick up the cell reference
to the variable data which in the case of Figure 13.3 is 0.18:

⫽VLOOKUP(“X”,LOOKTAB,3)

The labels in cells E3 and H4 are references to cell B3.

Collecting the results

Financial Planning using Excel

The results of the risk analysis will be collected using a data

table. The more reiterations of the model the better the results,
and so for this example the plan has been set up to perform 2000
recalculations, thus collecting 2000 different results. Numbers 1
through 2000 are entered into cells A5 through A2004. To calcu-
late the table the range A4 through B2004 is highlighted and the
Table command selected by Data Table. This is a one-way table
requiring column input and the reference can be to any blank
174 cell, such as A3. When OK is clicked the table is calculated by
placing the number 1 in cell A3, recalculating the model and
placing the resulting IRR in cell B5. The number 2 is then entered
in cell A3, the model is recalculated, which due to the RAND func-
tion causes all input and output to change and a different IRR is
placed in cell B6. The computer continues this process 2000
times at which point there are 2000 IRR results in the range B5
through B2004.

Note that it is important at this point to set recalculation to

Automatic except Tables from the Tools Calculation box otherwise
the table will be recalculated every time something is entered into
the spreadsheet.

Summary statistics
Some useful statistics about the results have been entered into the
range D5 through E10. To make the referencing of the data easier
the range B5 through B2004 is named ‘output’. A range is named by
selecting the required range and then clicking on the name
box located to the left of the edit line at the top of the screen and
 typing in the name. The following formulae have been entered into
column E:

E5 ⫽AVERAGE(OUTPUT)
E6 ⫽STDEV(OUTPUT)
E7 ⫽E9–E8
E8 ⫽MIN(OUTPUT)
E9 ⫽MAX(OUTPUT)
E10 ⫽COUNT(OUTPUT)

Financial Planning using Excel

The results of the risk analysis are most clearly viewed on a chart,
but 2000 data points are too many to plot onto a graph. Therefore a
frequency distribution of the data is created and a graph is drawn
using this data.

For this example, 11 data points have been chosen, beginning with
the minimum value returned as a result and then 10 further points
at equal intervals finishing with the maximum value returned as
a result. This is achieved with the following formulae in cells G5
175
and G6:

G5 E8
G6 G5⫹($E$7*0.1)

The formula in cell G6 is then copied to cell G15.

Frequency distribution
The FREQUENCY function is used to return the number of times the
results fall between the specified ranges. The form of the FREQUENCY
command in cell H5 is:

⫽{FREQUENCY(B5:B2004,G5:G15)}

The {} brackets mean this is an array function and as such has to

be entered in a special way. First, using the mouse select the
range in which the formulae are required – H5 through H15. Type
in the formula without the {} brackets. Then enter the formula by
holding the CTRL key and pressing SHIFT ENTER. The full range
will be calculated in one operation.
 Displaying the results on a chart
The results can be displayed graphically by selecting the range G4
through H15 and using the Chart button to automatically create a
chart such as the one shown in Figure 13.4.

350

300

250
Financial Planning using Excel

Frequency

200

150

100

50

0
0.16 0.17 0.18 0.19 0.20 0.20 0.21 0.22 0.23 0.24 0.24
176
Distribution

Figure 13.4 Graphical results of risk analysis on IRR

This graph illustrates quite a low level of risk because the most
likely outcome is a return of 21.20% with a standard deviation of
1.6%. Furthermore, even if all the most unfavourable estimates
occur, i.e. maximum investment costs, lowest cash-in flows and
highest cost of capital, this investment will still be expected to pro-
duce an IRR of 16.4%. On the other hand, if the investment is kept
low and the highest cash-in flows are achieved with a low cost of
capital, this investment could produce a return of 24.3%.

N.B. Due to the use of the RAND function the results of this exercise
will never be exactly the same as those displayed here.

Using the risk analysis model

Having set the risk analysis model up as shown in this chapter it can
be used with different ranges of input data, in this example to analyse
varying investment amounts with different cash flow and discount
rate scenarios. By putting an X in the appropriate box in the input
 sheet the analysis can be performed on the NPV at a fixed discount
rate, the IRR or the NPV at a variable discount rate. If it is only appro-
priate to specify ranges for some of the input variables the same value
can be entered for the minimum and maximum in the input sheet.

To recalculate the model after new input data has been entered or a
different output variable has been selected, F9 is pressed which
recalculates the data table and the remaining formulae cells are also
re-evaluated.

Financial Planning using Excel

Summary
The risk of any investment is the potential for input and/or output
variables to fluctuate from their original estimates. Risk analysis
accommodates this uncertainty by allowing ranges, as opposed to
single point estimates, to be used. It is generally easier to confi-
dently state that an investment will be between 300,000 and
400,000 than to say it will be 380,000.
177
The methodology applied to this example can be used with most
well designed deterministic plans.
This Page Intentionally Left Blank
Part 3
This Page Intentionally Left Blank
14
Budgeting
This Page Intentionally Left Blank
 You can fool all the people all the time if the advertising is right
and the budget is big enough.

– Joseph E. Levine, quoted in Halliwell’s Filmgoer’s Companion,

1984.

Introduction
A budget is a detailed estimate of future transactions. It can be

Financial Planning using Excel

expressed in terms of physical quantities, money or both. The
essence of a budget is that it is a target set for management to keep
within, achieve or surpass. Thus, a budget is always associated
with a specific departmental responsibility point or centre within
the organisation. This might be a division that has a sales budget,
a factory with a capital budget or an individual with an expense
budget.

183

Scope of budgeting
Budgetary control is not limited to commercial and industrial firms
attempting to produce a profit. The procedures involved are equally
applicable to not-for-profit organisations such as government depart-
ments, universities and charities.

All aspects of the business or organisation can be budgeted.

There might be income and expenditure budgets, cash budgets,
capital budgets, research and development budgets to mention
only a few examples. Budgets can be classified as master budgets,
departmental budgets or functional budgets. Whatever level or
degree of detail, a budget is useless if it does not focus on a point
of responsibility.

Budgeting is a management function that incorporates:

◆ setting objectives
◆ establishing detailed financial estimates
◆ delegating specific responsibility
◆ monitoring performance
◆ reacting to expectations.
 Benefits of budgeting
The benefits of budgetary control can usefully be classified in the
following way:

◆ A budget forces management to express in figures its future

intentions.
◆ It provides a yardstick by which individuals or groups can be
measured and rewarded.
◆ It allows some responsibility and authority to be decentralised
without loss of information required by management for control
Financial Planning using Excel

purposes.
◆ Budgeting provides a mechanism to control in detail the revenue,
costs, cash and capital expenditure of the firm.
◆ It facilitates an atmosphere of cost consciousness.
◆ It helps ensure that ROI is optimised.

184 Different approaches to budgeting

There are various approaches to the preparation of a budgeting sys-
tem, but two popular methods are the traditional approach and the
zero-based budgeting approach.

With the traditional approach some organisations re-compile last

year’s or the previous period’s figures, adjusted for expected growth
or for inflation. This approach to formulating a budget relies on the
notion that all the business variables will remain more or less the
same in relation to one another from one period to the next. For
this type of budget, after preparing an initial set of figures, a mini-
mal amount of maintenance is required.

On the other hand, some organisations adopt the view that the most
important aspect of budgetary control lies in the fact that the cre-
ation of budgets should impose on the firm a strict regime of think-
ing through what the organisation is doing and where it is going.
This approach has been popularised under the title of zero-based
budgeting (ZBB) and assumes that it is necessary to start the budg-
eting process from scratch each time. The main advantage of this
approach is that is ensures a rethink of the basic business assump-
tions on which the organisation relies. With a ZBB system, each
functional section in the business will plan in detail all revenue,
 expenditure and capital items and these items will be ranked in
order of importance to the firm.

An outcome from a ZBB system that is not the case with the tradi-
tional approach is that management are forced to prove the need for
each item of expenditure in the budget. It is not good enough to say
that the promotion account had a particular amount in it last year
and that this figure should be increased by 10% for inflation. Clearly,
it is more expensive to operate a ZBB system than a traditional
system, and therefore a cost-benefit study is appropriate before

Financial Planning using Excel

embarking on a ZBB.

The term ‘ZBB’ is generally associated with the budgeting of indirect

costs. It is therefore used extensively in central service departments,
marketing and distribution, and research and development. ZBB
is not considered to be directly relevant to the manufacturing
process, as detailed expenditure in this area is usually automatically
accounted for on a variable or direct basis.

185

Budget preparation
The procedure and operation of any budgetary control system is
clearly specific to each individual organisation and is a function of
management style and corporate culture. However, there are general
guidelines that are useful to bear in mind:

◆ Establishing the objectives – This usually involves a lengthy

process of analysing the organisations’ strengths and weak-
nesses and matching these to the opportunities in the environ-
ment in which the organisation functions. This is sometimes
referred to as strategic planning. This is a senior management
activity which is an ongoing process coming to a focal point
at budget time. A top-down approach is normally dominant
here.
◆ Forecasting the key business variables – Before any figures can be
derived an activity forecast must be established. In many organi-
sations this means that a sales forecast must be produced which
can be a lengthy and difficult process. It is important to ensure
that a broad spectrum of people are involved and committed
to the sales forecast. Therefore both top-down and bottom-up
approaches are appropriate here.
 ◆ Physical estimates are calculated – The number of people, the
scale of equipment and the volumes of raw materials required
must be established. This is usually a bottom-up procedure.
◆ Detailed costings of each responsibility centre are made – This
requires the involvement of a wide variety of staff and can lead
to a considerable amount of negotiation. Thus, both top-down
and bottom-up approaches are required.

These activities are best conducted in a cyclical fashion with feed-

back being sent up and down the organisation at various times,
Financial Planning using Excel

representing how different groups feel about the suggestions being

made. Thus, the amount of time required to produce a budget can
be substantial.

Spreadsheets for budgets

A spreadsheet can be used to support both the traditional approach
to budgeting as well as the zero-based approach.
186
With the traditional approach, having initially designed and devel-
oped a plan with care, continuous use of the system is largely a
case of changing the input data. Template techniques shown in
Chapter 9 are invaluable in this situation.

Templates can also form the basis of zero-based budgets as, even
when starting from scratch, an outline of the requirements for the
budget is known and a template can form the basis of the new budget.

Summary
Budgeting is an essential part of modern management and is regu-
larly performed in most organisations. It is as much a management
philosophy and technique as an approach to financial accounting.

The spreadsheet is a particularly powerful tool in the development

of budgets. It is also useful for the production of reports such as
budget–actual–variance, year-to-date totals and consolidated results.
15
A Spreadsheet Budgeting System
This Page Intentionally Left Blank
 It is a mistake to look too far ahead. Only one link in the chain of
destiny can be handled at a time.

– Anonymous

Introduction
The budgetary control system described in this chapter illustrates
how files can be linked in order to provide a flexible reporting
facility. Figure 15.1 shows the modules of the system. The system is

Financial Planning using Excel

a quarterly plan for a single department or division and provides for
the collection of quarterly budget figures and actual values. Options
for producing variance reports and year-to-date reports are provided.

Budget Template Actuals Template

189
Variance Year-to-date
Report Report

Figure 15.1 Proposed budgetary control system

Rather than develop a single, large file with the different modules on
separate worksheets, separate files for each module will be created
for this example. There are both advantages and disadvantages to
linking data across files. One of the main advantages is that different
parts of a system can be worked on simultaneously by different peo-
ple. Probably the most significant disadvantage is that when there are
many files to link the formulae can become very long and complex.

When embarking on a spreadsheet system that links cells across

files it is advisable to keep the files in the same directory. The rea-
son for this is that Excel keeps a track of the directory path when
the links are established and problems can arise if files are moved,
directory names or filenames are changed.

For the purposes of explaining how to develop a linked spreadsheet

system for budgetary control, a small summary profit and loss
account will be used as the basis for the budget. Obviously in a real
situation the files will be larger, but the methodology is the same.
 The system is supplied on the CD accompanying this book in a sep-
arate directory called BUDGET. The system then comprises three
files called BUDGET, ACTUALS and VARIANCE.

Preparing the budget template

The summary profit and loss account is a deterministic model.
Figure 15.2 shows the results of the plan after some data has been
entered and Figure 15.3 is the data input form. This model has been
Financial Planning using Excel

designed using the template methodology described in Chapter 9

and has two sheets. The first is called Budget and contains the logic
for the plan and the second is called Input and consists of the data
input form. The file is saved with the name BUDGET. Although any
name can be used it is important to carefully select the required

190

[Link]

Figure 15.2 Budget plan with data

Figure 15.3 Data input form for budget plan

 name, as the other files will reference it in the system. Once the
budgetary control system is complete the four files can be saved as
template files with an .XLT extension in order that they can be used
repeatedly for different sets of data.

Once completed this file is saved with the name BUDGET.

Preparing the actual template

To minimise the work required for the actual file, open BUDGET,

Financial Planning using Excel

copy columns A and B to a new file, remove any existing data and
shade the cells into which input is required. Figure 15.4 shows the
first sheet of the actual file. As this is a quarterly plan, it has to be
possible to enter four separate sets of actual data. Therefore the
information in Figure 15.4 is copied into three further sheets and
the sheets are named QTR1, QTR2, QTR3 and QTR4 respectively.

191

[Link]

Figure 15.4 Template for Actual data with sample data for first quarter
 To clarify which sheet is which, each sheet is also named as can be
seen from Figure 15.4.

Once completed the actuals template is saved in the same directory

as BUDGET with the name ACTUAL.

Preparing the variance report template

The variance report requires the user to select a quarter for which a
report is required. The system will then select the appropriate
Financial Planning using Excel

range from the budget file and the actual file.

Figure 15.5 shows a variance report for the first quarter using the
data for the first quarter’s budget shown in Figure 15.1 and the
actual data shown in Figure 15.4.

192

[Link]

Figure 15.5 Example variance report

The user enters the required quarter number in cell D1 and the
following nested IF function is required in cell B4 to pick up the
appropriate data.

In Excel:

⫽IF($D$1⫽1,[[Link]]BUDGETS!B4,IF($D$1⫽2,[[Link]]
BUDGETS!C4,IF($D$1⫽3,[[Link]]BUDGETS!D4,IF($D$1⫽4,
[[Link]]BUDGETS!E4,“!!!”))))
 This formula can be copied to the remaining cells in the budget col-
umn of the variance report.

If a number other than 1, 2, 3 or 4 is entered, the above formula will

return !!! in the cell. In addition the following IF function has been
entered into cell E1:

⫽IF(OR(D1⬍1,D1⬎4),“ENTER A QUARTER NUMBER BETWEEN 1 AND 4”,“”)

This will prompt the user if an invalid value is entered.

A formula similar to that used for the budgets is required for the actual

Financial Planning using Excel

column of the variance report, but with a reference to the appropriate
sheet of the ACTUAL file. The formula in cell C4 is therefore:

⫽IF($D$1⫽1,[[Link]]QTR1!B4,IF($D$1⫽2,[[Link]]QTR2!B4,
IF($D$1⫽3,[[Link]]QTR3!B4,IF($D$1⫽4,[[Link]]QTR4!
B4,“!!!”))))

This formula can be copied for the remaining cells in column C. To

complete the variance report the actual data is subtracted from the
193
budget data. Therefore the formula for cell D4 is

⫽B4⫺C4

This formula can be copied into the remaining cells in column D.

Preparing the year-to-date report template

The variance report is designed to calculate the variance for any given
quarter. However it is also useful to have a year-to-date report that
shows the cumulative variance for a specified number of quarters.

Figure 15.6 shows a year-to-date variance report for two quarters.

This report works by adding together the budget and actual data for
the number of quarters that are specified in cell E1. As the actual
data is only added to the file when it becomes available, cell H1
reports the first quarter for which there is no data. Therefore with
the above example it is only possible to prepare a year-to-date
report for the first or second quarters as there is not yet any data
available for the third quarter.

The following formula is required in cell H1 that will report the

first quarter in which there is no actual data. If there is data in all
 [Link]
Financial Planning using Excel

Figure 15.6 Year-to-date variance report for two quarters

four quarters, the cell will remain blank. This is achieved through
the use of the “” at the end of the formula:

⫽IF([[Link]]QTR1!$B$4⫽0,1,IF([[Link]]QTR2!$B$4⫽0,2,
IF([[Link]]QTR3!$B$4⫽0,3,IF([[Link]]
194 QTR4!$B$4⫽0,4,“”))))

The following formula is required in cell B4 to produce the year-to-

date budget amount based on the number of quarters entered into
cell E1.

⫽IF($E$1⫽1,[[Link]]BUDGETS!B4,IF($E$1⫽2,SUM
([[Link]]BUDGETS!B4:C4),IF($E$1⫽3,SUM([[Link]]
BUDGETS!B4:D4),IF($E$1⫽4,SUM([[Link]]
BUDGETS!B4:E4),“!!!”))))

If a number other than 1, 2, 3 or 4 is entered into cell E1, the above

formula will return !!!, alerting the user to the fact that a valid quar-
ter number has not been entered. This formula can be copied to the
remaining cells in the budget column, with the exception of the
unit price. It is not appropriate to accumulate a unit price and
therefore the average unit price is calculated in cell B5 by dividing
the revenue by the sales.

A formula similar to that used to calculate the year-to-date budgets

is required for the actuals as shown below:

⫽IF($E$1⬎⫽$H$1,“NO DATA”,IF($E$1⫽1,[[Link]]QTR1!B4,
IF($E$1⫽2,SUM([[Link]]QTR1:QTR2!B4:B4),IF($E$1⫽3,
SUM([[Link]]QTR1:QTR4!B4:B4),IF($E$1⫽4,
SUM([[Link]]QTR1:QTR4!B4:B4),“!!!”)))))
 The first part of the above formula compares the number of quarters
entered into cell E1 with cell H1 which is reporting the actuals
quarter for which there is no data. If E1 is greater than or equal to
H1, the message ‘No data’ is returned, otherwise the appropriate
number of periods are accumulated.

As with the budget column, the formula can be copied to the

remaining cells, but again the unit price is changed to be the aver-
age unit price by dividing the revenue by the sales.

The year-to-date variance is calculated in the same way as in the

Financial Planning using Excel

variance report by subtracting the year-to-date actuals from the
year-to-date budgets. An enhancement is made to this formula to
check the contents of actuals in order that the variance will be left
blank if the ‘No data’ message is returned for the actual. Therefore
the following formula is entered in cell D4 and copied for the
remaining cells in the variance column.

⫽IF(C4⫽“NO DATA”,“”,B4–C4)
195

Summary
The four files that have been developed in this chapter describe a
methodology for producing a flexible budgetary control system.
The flexibility is primarily due to the use of separate files for each
module of the system which means that different people can work
on different parts of the system at the same time and users can
choose to produce reports using selected data.

Before using the system it is important to ensure that all the files
are in the same directory and that the links are correctly referencing
the files. It might be necessary to select Edit Links and amend the
path if the system is installed on another user’s computer.
This Page Intentionally Left Blank
16
Consolidating Data
This Page Intentionally Left Blank
 Man is still the most extraordinary computer of all.

– John F. Kennedy in a speech on 21 May 1963.

Introduction
The budgetary control system described in Chapter 15 is for a single
division or department. In many organisations it is necessary to
consolidate, summate or aggregate the data from separate divisions
to produce a divisional or corporate report.

Financial Planning using Excel

The consolidation options available in the spreadsheet allow for
selected ranges of data from different files to be merged into a
single file with options for summing, subtracting, averaging, etc.

To illustrate the consolidation features the variance report from the

budgetary control system will be used.

199
Using the Consolidate command
Excel provides a powerful set of consolidation tools through the Data
Consolidate command. However, before using the command a set of
files to be consolidated need to be created. To do this open the Variance
file and select a valid quarter number in order to produce a variance
report. To simplify the consolidation procedure the range to be
included in the consolidation is named by selecting the range A3:D14
followed by INSERT NAME DEFINE and call the range DATA_AREA.

To indicate that this is a variance report for a division enter

Division A into cell A2 and then save the file as DIVA. To quickly
produce another two files for this consolidation exercise change
the division reference in cell B2 to Division B and save the file as
DIVB and then change B2 once more to Division C and save the file
as DIVC.

The next step is to create a new file that will form the consolidated
report. Type an appropriate title for the report into cell A1 and then
place the cursor on the cell that will be the top left cell of the con-
solidated range, which for this example will be B3.

Select DATA CONSOLIDATE, which produces the dialogue box shown

in Figure 16.1. The default function for the Consolidate command
Financial Planning using Excel

Figure 16.1 Data consolidate dialogue box

200
is SUM, i.e. the corresponding cells from the selected ranges will be
summed. However, by clicking on the arrow to the right of the func-
tion box the other alternatives can be seen.

The reference box refers to the files or ranges that are to be consoli-
dated. To complete this click on Browse and select the file DIVA.
After the exclamation mark (!) type the range name Data_area and
then change the A in DivA to a question mark (?). This is a wildcard
that will replace any single character for the question mark. Click
Add to put this reference in the list of references to be consoli-
dated. The effect of this reference is that when the OK button is
clicked the system will open each file beginning with DIV and take
the range in those files called Data_area and sum them into the cur-
rent worksheet. The range Data_area does not have to be in the
same position in each file, but it should be the same size in each
file in order that the correct cells are added together.

It is not always possible to name files and ranges with similar

names and if this is the case each file and range to be consoli-
dated must be individually selected and the Add box checked
until the full list is displayed in the dialogue box. To illustrate
this Figure 16.2 shows a completed dialogue box referencing
the files separately with range references instead of the common
range name.
 Financial Planning using Excel
Figure 16.2 Consolidate dialogue box with separate references

Before clicking OK, check the three boxes at the bottom of the dia-
logue box to indicate that the top row and left column of the ranges 201
to be consolidated are labels and that links are required to the
source files. Figure 16.3 shows the results of the consolidation.

Figure 16.3 Results of Data Consolidate command

Because the Create links to source data box was checked in the dia-
logue box, the bar to the left of the report is automatically pro-
duced. By clicking on the number 2 the report is expanded as
shown in Figure 16.4. This shows the data from all the ranges
included in the Consolidate command.
Financial Planning using Excel

202

Figure 16.4 Expanded consolidated report

Looking at the data in Figure 16.3 the references to the three divi-
sional files show the full path to those files and the totals contain a
SUM function.

It is useful to be able to audit the source data of a consolidated report,

but if a large number of files are being consolidated it may represent
too much data. By not selecting the Create links to source data box,
only the results will be produced in the consolidated report. This will
result is a file much smaller in size, but it will not leave an audit trail
back to the individual files.

Summary
The DATA CONSOLIDATE command provide a means of adding data
together from different files without the need for linking files. This
is especially useful when a large number of files need to be
accessed as the file linking procedures require long formulae and
take time to recalculate.
Index
This Page Intentionally Left Blank
Adaptive filtering, 31–2 number of observations, 17
range, 18
Break-even analysis, 142–3 standard deviation, 19
Budgetary control system: variance, 19
preparing actual template, 191–2
preparing budget template, 190–1 Economic order quantities (EOQ):
preparing variance report template, cost of inventory, 146
192–3 developing EOQ plan, 144
preparing year-to-date report lead times and safety stock, 146–7
template, 193–5 with no supplier policy, 144–5
Budgeting: with supplier policy, 145
benefits, 184 Excel, graphic approach with, 46
budget preparation, 185–6 Expected values, model for analysing,
different approaches, 184–5 65–70
scope, 183 Exponential smoothing, 32–4
spreadsheets for budgets, 186
Business planning, 84 Financial plan:
deterministic planning, 84–6 cash flow statement, 125–6

Index
optimising models, 89–90 developing balance sheet:
stochastic planning, 87–9 assets, 122
capital and liabilities, 123–4
205
Capital investment appraisals (CIA), developing profit and loss account,
131–2 117–21
internal rate of return, 133, 137 developing profit and loss
net present value, 132–3 appropriation account, 121
profitability index, 133, 137 funds flow statement, 124
see also CIA plan, developing model used, 115
CIA plan, developing: ratio analysis, 125
discounted payback, 136 Forecasting:
internal rate of return, 133, 137 and Excel, 7–8
investment reports, 133 forecast units, 7
NPV and PI at variable discount objective forecasts, 6
rate, 138 subjective forecasts, 5–6
NPV at a fixed discount rate, 137 time, 6–7
profitability index (PI), 133, 137 Forecasting techniques:
rate of return, 137 accuracy and reliability, 75
simple payback, 134–6 charts, 75
Consolidate command, using, selecting the right technique, 73–4
199–202 statistical models:
average error, 76
Data analysis command, 19–20 standard error, 76–7
Data collection, 11–15 Thiels U, 77
Descriptive statistics, 15–19 subjective methods, 77–8
mean, 17
median, 17 Goal seeking:
minimum and maximum, 18 using goal seeking feature, 161–2
mode, 18 using Solver for optimising, 162–5
 Learning curve costing model, 139–42 summary statistics, 174–5
Least square line: using risk analysis model, 176–7
command function:
command approach, 45–6 Sales campaign appraisal, 147–9
function approach, 43–5 developing sales campaign plan,
function vs command, 46 148–9
Scatter diagram, 40–4
Moving averages, 26 Spreadsheets, 93–111
Multiple linear regression, 49–51 Standard error, 47–9
selecting variables for multiple
regression, 51–2 Templates, 108–9
with Excel, 50–1 data input forms, 109–10
Multiplicative time series analysis, 56 Time series analysis model:
calculation and processing, 58–60
Planning, 83 data input and output, 57–8
see also Business planning graphics, 61–2

Risk analysis: Weighted moving average, 27–30

Index

changing the original plan, 170–2 proportional method, 28–9

collecting the results, 174 trend adjusted method, 29–30
displaying the results on a What-if analysis, 153–6
206
chart, 176 data tables, 156–7
frequency distribution, 175 one-way data table, 157–9
incorporating the RAND function, two-way data table, 159–61
172–3 manual what-if analysis on opening
results worksheet, 173–4 assumptions, 154–6