Cost Management and

Analysis

Lecture 2

Cost Behaviour, Cost Drivers

and Cost Estimation

Cost Management and Analysis: 22753 University of Technology, Sydney

 Lecture Outline
• Introduction

• What is a cost?

• Basic cost terminology

• Cost behaviour patterns

o Relevant Range

• Cost estimation

o High-low method

o Regression analysis

• Evaluating and choosing cost drivers

• Lecture demonstration problems: high-low method

2
Cost Management and Analysis: 22753 University of Technology, Sydney
 Introduction
This topic focuses on the analysis of costs and how they behave in relation to
changes in a particular organisational activity (such as production volume).

Knowing how costs vary by identifying the drivers of costs and being able to
distinguish fixed from variable costs are essential to good management
decision making. Many managerial functions in the planning and control areas
require and rely on knowing how costs behave in relation to activities (e.g.
production output/volume). Analysis of cost behaviour patterns allow
managers to prepare standard costs, formulate budgets and to predict and
control costs.

But first, what is a cost?

3
Cost Management and Analysis: 22753 University of Technology, Sydney
What is a cost?

Rent
Stationary Office
Furniture
Machines (Depreciation)
(Depreciation)

Resources
used to Electricity
generate some
Labour
(Time!) beneficial
output
Materials

Spare Parts

Insurance
Water Computers
(Depreciation)

4
University of Technology, Sydney
What is a cost?

If it costs $200 to manufacture one

unit of a product, how much will it
cost to make two units?

5
University of Technology, Sydney
 What is a cost?

Although costs move with output changes, the cost movement tends to be less
sensitive than the activity change, that is, if output doubles, total costs will
increase, but not by 100 per cent.

Production Level Total Costs

1 $200

2 $300

Why in the above example do the costs of producing two units not double the
cost of producing just one unit? (i.e. Why is the cost not $400?)

6
Cost Management and Analysis: 22753 University of Technology, Sydney
Basic Cost Terminology
Cost Pools: Groups of a particular type of cost (e.g. material costs, labour costs,
electricity costs, depreciation costs). Cost pools vary by the type of costs
included and the level of aggregation.

Cost Objects: Factors or items in an organisation that we want to have a

separate measure of their costs (e.g. products, departments, jobs, events).
Costs are assigned from cost pools to these items.

Cost Drivers: Activities or factors that drive (cause) costs to occur. Drivers allow
cost pools to be allocated to cost objects.

Cost Cost Cost

Pool Driver Object(s)

e.g. Electricity Machine Hours

Product
7
Basic cost terminology
Cost behaviour: The relationship between a cost and the level of activity that
causes this cost (i.e. cost driver).

Volume-based driver: A cost driver that assumes that costs are

driven by production or a factor related to production (e.g.
machine hours)

Non-volume-based driver: A cost driver not directly related to

production volume (e.g. number of deliveries)

The basic (linear) cost function can be represented in the following manner:
Y = a + bX
Total Cost = FC + VC x Activity
8
University of Technology, Sydney
 Cost behaviour patterns
Cost Behaviour: The relationship between a cost and the level of activity (i.e.
production volume) or cost driver. There are a number of cost behavior
patterns:

Variable Costs (VC)

Fixed Costs (FC)

o Committed Costs versus Discretionary Costs

Stepped-fixed Costs

Semi-variable (mixed) Costs

o Semi-variable costs can often be broken into variable and fixed
components)

Curvilinear Costs
o We assume in this subject that all costs are linear

9
Cost Management and Analysis: 22753 University of Technology, Sydney
Cost behaviour patterns
Variable costs: Costs that change in total in direct proportion to a change in
activity.

Cost ($)

Variable Cost (VC)

X
Activity (e.g. units produced)
10
University of Technology, Sydney
Cost behaviour patterns
Fixed costs: Costs that remain unchanged in total as the level of activity
varies.

o Discretionary cost: Based on a managerial decision that can be changed

relatively easily.

o Committed cost: Results from the basic organisation structure and difficult to
adjust in the short term.

Y
Cost ($)

Fixed Cost (FC)

X 11
Activity (e.g. units produced) University of Technology, Sydney
Cost behaviour patterns
Step-fixed costs: Costs that remain fixed within a certain range of activity, but
change to a different amount outside that range.

Cost ($)

Step-Fixed
Cost

X
Activity (e.g. units produced)

12
University of Technology, Sydney
Cost behaviour patterns
Semi-variable (mixed) costs: Costs that contain both fixed and variable components.

The basic (linear) cost function can be represented in the following manner:

Total Cost (Y) = FC + VC x Activity (X)

Semi-Variable Cost
Cost ($)

Variable Cost
Fixed Cost Component
Component

X
Activity (e.g. units produced)

13
University of Technology, Sydney
Cost behaviour patterns
Curvilinear costs: Costs that exhibit cost behaviour that can be described by a
curved line.
o Convex costs have increasing marginal costs.
o Concave costs have decreasing marginal costs.

Y
Curvilinear Cost
Cost ($)

Concave

Convex

X
Activity (e.g. units produced)
14
University of Technology, Sydney
 Cost behaviour patterns

Relevant Range: Refers to a the range of activity (i.e. the upper and lower limits
of production) for which assumptions about cost behaviour hold.

Assumptions made within one relevant range may not be valid in

another range

o VC per unit may change

o Fixed costs are not constant

Prediction of costs can only be done accurately within a known range

15
Cost Management and Analysis: 22753 University of Technology, Sydney
Cost behaviour patterns

Y
Relevant Range 1 Relevant Range 2
Cost ($)
TC2

 VC per unit
TC1
 FC

X
Activity (e.g. units produced)

16
University of Technology, Sydney
Cost behaviour: Problem 3.32

17
 For each of the following cost items (1 to 11), choose the graph, (a) to (i), that best represents it: Graph
1 Cost of electricity at a university. For low student enrolments, electricity costs increase with enrolment, but at a
decreasing rate. For high student enrolments, electricity costs increase at an increasing rate.

2 Charges for international telephone calls, which are based on the number of minutes per month. The charge is $0.25
per minute, up to 600 minutes. Additional minutes (above 600) are free.

3 Cost of outsourcing pathology testing by a hospital. The hospital pays an independent lab a fee of $10 000 per
month plus $5 for each test performed.
4 Salary costs of the shift supervisors at a courier company. Each shift is eight hours. The depot operates with one, two
or three shift supervisors simultaneously, at various times of the year, depending on the volume of delivery work.

5 Cost of 24-hour security services at a factory, provided by an external security company.

6 Wages of waiters in a restaurant. Waiters are part-time workers who are scheduled to work in three hour blocks.

7 Cost of electricity during peak-demand periods is based on the following schedule:

Up to 10 000 kilowatt-hours (kWh): $0.90 per kWh
Above 10 000 kilowatt-hours: $1.12 per kWh
8 Cost of sheet metal used to manufacture automobiles, priced by weight (per kilogram, or part thereof).
9 Cost of chartering a private aeroplane. The cost is $410 per hour for the first three hours of a flight. Then the cost
drops to $305 per hour.
10 Under a licensing agreement with an Indonesian import/export company, your firm has begun shipping machine tools
to several countries. The terms of the agreement call for an annual licensing fee of $100 000 to be paid to the
Indonesian import company if total exports are under $4 000 000. For sales in excess of $4 000 000, an additional
licensing fee of 7 per cent of sales is due.

11 Tariffs paid by a wine exporter. On one Pacific island, a tariff must be paid by the wine exporter for every case of wine
brought into that country. The tariff schedule is the following
0 to 6000 cases per year: $11 per case
6001 to 12 000 cases per year: $14 per case
Above 12 000 cases per year: $19 per case
18
Cost estimation
Cost estimation is the process of determining the cost behaviour
pattern of a particular cost item.

o Cost prediction is using knowledge of cost behaviour to forecast the

level of cost at a particular level of activity

The most important issue in estimating a cost function is to

determine whether a cause-and-effect relationship exists between
the activity or cost driver (X) and the resulting costs (Y). This may
arise in several ways:

1. Physical relationship with the cost driver (engineered cost)

2. Contractual arrangement

3. Logic and knowledge of operations

19
University of Technology, Sydney
Cost estimation
There are three methods to identify cause-and-effect relationships.

1. Engineering method
Identifies the relationship that should exist between input and outputs.
Relationships are determined through “time and motion” studies which observe
the steps required and time taken to perform particular activities.

2. Managerial judgement

a. Conference method
Estimates cost functions on the basis of opinion about costs and their
drivers gathered from various departments of an organisation – purchasing,
human resources, engineering, manufacturing, etc. This method relies on
the expert knowledge and judgement of managers.

b. Account classification (Account analysis)

The account analysis method estimates cost functions by classifying cost
accounts in the ledger as variable, fixed or mixed with respect to the
identified cost driver. Typically managers use qualitative rather than
quantitative analysis when making these cost classification decisions.
20
Cost estimation
3. Quantitative analysis
These are formal analyses of cost relationships to fit mathematical
equations (functions) to past data. There are six steps involved.

Step 1: Choose the dependent variable (the cost variable you want to
estimate)

o Example: Electricity cost in the manufacturing plant

Step 2: Identify the independent variable(s) (the activity or activities

that cause the cost)
o Example: Machine hours

21
University of Technology, Sydney
Cost estimation
Step 3: Collect data on the dependent variable and the cost driver(s)

o Example: Monthly electricity cost for the last 12 months

Month Electricity Machine Hours

Jan 61,020 25,504
Feb 76,917 30,907
Mar 75,313 30,309
Apr 54,200 20,000
May 68,067 28,325
Jun 63,269 25,322
Jul 59,918 24,518
Aug 69,890 29,707
Sep 72,000 45,000
Oct 75,802 32,159
Nov 63,019 25,862
Dec 74,293 31,356

22
University of Technology, Sydney
Cost estimation
Step 4: Plot the data – use a scatterplot to visually observe the
relationship between the cost and its’ driver(s)

Step 5: Estimate the cost function using High-Low or Regression

analyses
23
University of Technology, Sydney
Cost estimation: High-low method
The high-low method is a simple method for separating semi-variable costs into
their fixed and variable components.

VC calculation: Two periods of data (high and low) are chosen at based on
the levels of activity – both levels should be within the same relevant range.

Total Cost High – Total Cost Low

VC per unit of activity
Activity High – Activity Low

FC calculation: Substitute the VC with either the high or low values into the
total cost formula.

TC = FC + ( VC x Activity )
24
University of Technology, Sydney
Cost estimation: High-low method
Select the highest and lowest values of the cost driver.

Machine Hours Electricity

Difference in
costs are due to
Maximum Value 45,000 $72,000
Minimum Value 20,000 $54,200 variable costs
within the
Difference 25,000 $17,800
relevant range
TC

$72,00 High-Low method

0 calculates the
slope/gradient of
$54,200 the line
Cost at Cost at
Total 20,000 hrs 45,000 hrs
Costs FC

Hours.
Fixed costs are the same at
25
both levels of activity!
University of Technology, Sydney
Cost estimation: High-low method
The dependent variable (Y) is electricity. This is predicted or explained by the
independent variable (X) of machine hours.

Use the high-low method to construct a cost function for electricity:

26
University of Technology, Sydney
Cost estimation: High-low method

There are two simplifying assumptions in the High-Low method:

o Changes in the total costs of a cost object are explained by variations of a

single cost driver.

o Cost behaviour is adequately approximated by a linear cost function of

the cost driver within the relevant range.

More advanced statistical methods can overcome these limitations.

o They are still limited to a single relevant range

27
University of Technology, Sydney
Cost estimation: Regression analysis
Regression analysis is a statistical method that measures the average amount
of change in the dependent variable that is associated with a unit change in
one or more independent variables. It has two advantages over high-low:

o All available data points are used (high-low uses just two data points). The
difference can be seen in the scatterplot below.

o Multiple cost drivers can be used to estimate costs

Regression
estimated line

Outlier

High-low
estimated line

28
University of Technology, Sydney
 Cost estimation: Regression analysis
Regression analysis for the previous example would give the following output:

SUMMARY OUTPUT

Regression Statistics
Multiple R 0.698
R Square 0.487
Adjusted R Square 0.436
Fixed Cost
Standard Error 5565.3
Observations 12
Variable Cost
ANOVA
df SS MS F Significance F
Regression 1 294540335.4 294540335.4 9.5096 0.0116
Residual 10 309729181.4 30972918.14
Total 11 604269516.8

Coefficients Standard Error t Stat P-value Lower 95% Upper 95% Lower 95.0% Upper 95.0%
Intercept 43293.4148 8110.6279 5.3379 0.0003 25221.8 61365.0 25221.8 61365.0
Machine Hours 0.8430 0.2734 3.0838 0.0116 0.2339 1.4521 0.2339 1.4521

Cost Function (Regression):

Cost Function (High-Low):

29
University of Technology, Sydney
 Cost estimation techniques
Regression analysis can also include multiple cost drivers:
SUMMARY OUTPUT

Regression Statistics
Multiple R 0.853
R Square 0.728
Adjusted R Square 0.668
Standard Error 4272.2
Observations 12

ANOVA
df SS MS F Significance F
Regression 2 440007799.2 220003899.6 12.0541 0.0028
Residual 9 164261717.6 18251301.95
Total 11 604269516.8

Coefficients Standard Error t Stat P-value Lower 95% Upper 95% Lower 95.0% Upper 95.0%
Intercept 27576.2781 8352.071 3.302 0.009 8682.580 46469.976 8682.580 46469.976
Machine Hours 0.9880 0.216 4.573 0.001 0.499 1.477 0.499 1.477
Batches 3.2431 1.149 2.823 0.020 0.644 5.842 0.644 5.842

Cost Function (Multiple regression):

30
University of Technology, Sydney
Cost estimation
Step 6: Evaluate the estimated cost function

The four most important criteria in evaluating the cost function are:

1. Economic plausibility – Does the cost function make sense?

2. Goodness of fit – How well does the line fit the data points?

3. Slope of line – A relatively flat line indicates a weak or non-existent

relationship between the activity and the cost. A steep line would indicate a
stronger relationship.

4. Cost v benefit – A subjective estimate might be good enough!

Incorrectly estimating the cost function has repercussions for cost

management and cost control. Managers will have a reduced ability to make
accurate plans and budgets and have less understanding of organisational
processes. 31
University of Technology, Sydney
Lecture demonstration problem 1
Consider the internal auditing section of Macquarie Bank. This section has a small core
staff which is supplemented with accounting staff from other departments when the
workload increases beyond a certain level. Using the time sheets and cost records from
the auditing department, determine the cost behaviour pattern of audit staff costs using
high-low and regression analysis. What is the expected labour cost if the audit section
works 720 hours?
Month Hours Total Working Out:
Charged Labour
Cost
Jan 760 $16,000
Feb 600 $13,500
Mar 500 $13,500
Apr 680 $15,000
May 740 $16,400
Jun 780 $17,200
Jul 760 $16,800
Aug 500 $12,500
Sep 600 $15,000
Oct 680 $14,000
Nov 700 $13,400
32
Dec 800 $17,000
University of Technology, Sydney
 Lecture demonstration problem 1
Using regression analysis:
SUMMARY OUTPUT

Regression Statistics
Multiple R 0.846
R Square 0.716
Adjusted R Square 0.687
Standard Error 913.370
Observations 12

ANOVA
df SS MS F Significance F
Regression 1 21000052.7 21000052.7 25.173 0.001
Residual 10 8342447.3 834244.7
Total 11 29342500

Coefficients Std Error t Stat P-value Lower 95% Upper 95% Lower 95.0% Upper 95.0%
Intercept 6039.2405 1810.287 3.336 0.008 2005.670 10072.811 2005.670 10072.811
X Variable 1 13.3122 2.653 5.017 0.001 7.400 19.224 7.400 19.224

Cost Estimation Function: Total Cost = Fixed Cost + Unit variable Cost * Hours
Total cost for 720 Hours =

33
University of Technology, Sydney
Lecture demonstration problem 2
You are the management accountant at Lion Nathan for the product line “Tooheys
Extra Dry”. You think that there is an economies of scale advantage to production for
certain variable resources, but are unsure what the actual cost effect is. You think that
the relevant range to get this cost advantage starts at around 1500 units per day. Use
the following data to calculate the reduction in variable costs.

Daily Actual Production

Production Cost

1 600 $6,000
2 750 $6,300
3 800 $6,650
4 1000 $7,200
5 1300 $7,750
6 1400 $8,000
7 1750 $9,000
8 1800 $9,200
9 1950 $9,500
10 2000 $9,600
34
University of Technology, Sydney
Lecture demonstration problem 2
Y Change in slope is due to
economies of scale
Cost
TC

VC

FC

X
1500 Production
Working Out:

35
University of Technology, Sydney
Tutorial 2 (Week 3): Cost Behaviour, Cost Drivers and
Cost Estimation
Next week’s tutorial will take place in computer labs.

Prior to the tutorial, you need to:

1) Complete Tutorial 2 preparation tasks (in the Subject Outline)

2) Attempt Tutorial 2 homework :

• Cost Behaviour exercises
• Donut Emporium Case Study (AP2.1) exercise

The following resources are available on UTS Online (Subject

Documents/Tutorial Materials and Solutions/Computer Lab Materials)
• CMA Tutorial 2 Excel Guide
• CMA Tutorial 2 Excel Workbook

You need to submit your group assignment registration form to your lecturer
before the start of this tutorial.

36
University of Technology, Sydney