COST ACCOUNTING

COST BEHAVIOUR AND COST

ESTIMATION/PREDICTION

7/5/23 PROF. WILLIAM COFFIE & DR. TEDDY OSSEI KWAKYE 1

Coverage
Classification of cost by behaviour
• Fixed cost, Variable cost, Mixed cost and Stepped cost
Methods of cost estimation
• Historical methods
• High low, Scatter graphs, Account analysis and Regression
analysis methods
• Non-historical methods
• Engineering and conference methods
7/5/23 PROF. WILLIAM COFFIE & DR. TEDDY OSSEI KWAKYE 2
Learning Outcomes
At the end of this lecture students should be able to:
• Understand and appreciate how cost responds to
activity level i.e. the event that leads to cost.
• Segregate total cost into fixed and variable
components.
• Predict cost using linear quantitative techniques.

7/5/23 PROF. WILLIAM COFFIE & DR. TEDDY OSSEI KWAKYE 3

INTRODUCTION

• Cost Accounting is all about providing cost information to managers

for Planning, Control and Decision Making in an organization.
• To effectively do that, (plan, control and take decisions) managers
must know the Cost (types and nature) in their organization.
• They must know their cost structure.
• Cost Structure is the relative proportion of each type of cost (Fixed and
Variable) to the Total Cost present in a firm.

7/5/23 PROF. WILLIAM COFFIE & DR. TEDDY OSSEI KWAKYE 4

INTRODUCTION
• Analysing the cost structure of a firm means understanding the Cost
Behaviour pattern of the various cost.
• Cost Behaviour is the way or how cost responds to changes in
activity level, i.e. the behaviour of cost to changes in activity level.
• The activity level is also referred to as Activity Base or Cost Driver
i.e. the event or activity that results in cost.
• It is expressed in various forms typical ones being capacity, volume,
output, e.g. machine hours, units produced, units sold, labour hours
etc.
7/5/23 PROF. WILLIAM COFFIE & DR. TEDDY OSSEI KWAKYE 5
INTRODUCTION
• Thus by knowing the cost behaviour, the cost structure can be
analyzed properly to forecast the level of cost at any particular level
of activity.
• This is Cost Prediction, which is essential for planning, control and
decision-making.
• The behavior of cost to changes in activity level forms one basis of
cost classification. In this way cost can be classified into:

7/5/23 PROF. WILLIAM COFFIE & DR. TEDDY OSSEI KWAKYE 6

VARIABLE COST
• This changes in Total and in direct proportion to changes in activity
level or cost driver i.e.
• if activity level increases by 10% the total variable cost is expected to
increase by 10% e.g. Direct Material, Direct labour etc.

7/5/23 PROF. WILLIAM COFFIE & DR. TEDDY OSSEI KWAKYE 7

FIXED COST
• This remains constant regardless of changes in activity level or cost
driver e.g. rent, rates, insurance, executive salaries etc.

7/5/23 PROF. WILLIAM COFFIE & DR. TEDDY OSSEI KWAKYE 8

SEMI-VARIABLE OR MIXED COST
• This cost changes in Total but not in direct proportion to changes in
activity level.
• If activity level changes by 10%, Total Cost may change by 18%.
• It is also called mixed cost.
• It is the cost that has both fixed and variable cost component e.g.
electricity, repairs, telephone, maintenance.
• They vary with activity but not in direct proportion.

7/5/23 PROF. WILLIAM COFFIE & DR. TEDDY OSSEI KWAKYE 9

SEMI-VARIABLE OR MIXED COST
• The fixed portion of a mixed cost represents the basic, minimum
charge for just having a service ready and available for use.
• The variable portion represents the charge made for actual
consumption of the service.
• As expected, the variable element varies in proportion to the amount
of the Service that is consumed.

7/5/23 PROF. WILLIAM COFFIE & DR. TEDDY OSSEI KWAKYE 10

ASSUMPTIONS ON COST BEHAV.
• The classification of cost into Variable and Fixed is valid within a
narrow bound of activity known as Relevant Range.
• Relevant Range is the range of activity within which actual operations
are likely to occur without change in cost.
• It is the range of activity level that fixed cost is assumed to remain
constant or the assumptions about cost behavior is valid.
• What it means is that fixed cost can be fixed at a given range, beyond
that it could change e.g. rent and rates could change.

7/5/23 PROF. WILLIAM COFFIE & DR. TEDDY OSSEI KWAKYE 11

POINTS TO NOTE ABOUT COST
BEHAVIOUR
• Again even though the classification is based on changes in total cost
to changes in level of activity, the reverse is the case when it comes to
per unit cost.
• Variable costs vary in total as volume changes, but is constant per unit.
• Fixed costs are fixed in total but vary per unit as volume changes i.e.
as volume increases fixed cost per unit decreases.
• Semi-variable cost changes both in Total and per Unit.

7/5/23 PROF. WILLIAM COFFIE & DR. TEDDY OSSEI KWAKYE 12

POINTS TO NOTE ABOUT COST
BEHAVIOUR
• Variable cost is volume related i.e. it changes with volume.
• Fixed cost is time related i.e. it changes with time.
• There is also an assumption of linearity i.e. if total variable and total
fixed costs are plotted on a graph in relation to activity level, they
appear as unbroken straight line.
• This is in contrast to the Economist where the lines are curvilinear.

7/5/23 PROF. WILLIAM COFFIE & DR. TEDDY OSSEI KWAKYE 13

POINTS TO NOTE ABOUT COST
BEHAVIOUR
• Even though other cost drivers may affect cost, there is always the
assumption of only one cost driver.
• The influences of other possible cost drivers on total costs are held
constant or deemed to be insignificant.

7/5/23 PROF. WILLIAM COFFIE & DR. TEDDY OSSEI KWAKYE 14

FIGURES – VARIABLE COST
(a) (b) (c) (b  a)
• No. Of Bread Produced Total Cost Cost per unit
• (Activity Level)
• 250 7500 30
• 500 15000 30
• 750 22500 30
• 1000 30000 30
• Variable cost increases, as volume of production increases but per unit cost
remained constant.
7/5/23 15
FIGURES – FIXED COST
• No. of Bread Produced Total Cost Cost per unit (Activity
Level)
• 250 1000 4
• 500 1000 2
• 750 1000 1.3
• 1000 1000 1.0
• Fixed cost remains constant in total regardless of changes in activity
level but decreases as volume of production increases i.e.
• Cost per Unit decreases as activity level increases.
7/5/23 16
FIGURES – SEMI-VARIABLE COST
• No. of Bread Produced Total Cost Cost per unit

• (Activity Level)
• 250 8500 34
• 500 16000 32
• 750 23500 31.3
• 1000 30000 30
• Total Cost increases but not in direct proportion to increase in activity
level. Per Unit Cost decreases because of the Fixed Cost element.
7/5/23 17
GRAPHICAL – VARIABLE COST
• Total Variable Cost Per unit variable cost
• (Average variable cost)
• y y

• Cost Cost

•
x x
• Level of Activity Level of Activity
7/5/23 18
GRAPHICAL – FIXED COST
• Total Fixed Cost Per unit Fixed cost
• (Average Fixed Cost)
• y y

• Cost Cost

• x x

• Level of Activity Level of Activity

7/5/23 19
GRAPHICAL – SEMI-VARIABLE COST
• Total Cost
• y
• Variable Cost

•
• Cost Fixed Cost
•
• Level of Activity x
7/5/23 20
STEP COST
• These are cost which remain constant for a range of activity i.e. are
fixed within the relevant range, then when activity increases further,
the cost has to be increased by a significant amount e.g. supervision
cost, rent etc.
• For a range of activities for example one supervisor will be sufficient,
but there comes a point when an additional supervisor has to be
appointed because the scale of activities has increased so cost
increases.

7/5/23 21
GRAPHICAL – STEP COST

Cost

Level of Activity
7/5/23 22
COST ESTIMATION/PREDICTION
• Managers on daily basis plan, control and take decisions on products.
(Production and Procurement) activities, resources consumption and
cost resulting there from.
• To effectively do this, managers must establish the relationship
between activities and cost. Cost do not just happen.
• They are cause by activities.
• Cost estimation is the process of establishing the relationship between
cost and cost drivers. (What causes cost).

7/5/23 23
COST ESTIMATION
• The estimation process answers the following questions:
(i) What causes the cost (cost drivers)
(ii) How do costs relate to the cost drivers (activities that causes or
drives the cost)
(iii) How do cost response to changes in activities?
• The foregoing means to estimate the cost, the manager must know
how the organizational cost behave to changes to activities.
• Cost behavior therefore underline estimation.

7/5/23 24
COST ESTIMATION
• By knowing how organizational cost behaves to its activities,
managers would be able to estimate, predict or forecast what their cost
would be at any activity level.
• At least within the relevant range.
• The classification of cost based on behavior basically put cost into
two; namely variable and fixed.
• Semi variable or mixed cost is a combination of the two.
• For cost estimation in an organization all cost should be segregated
and classified into either variable or fixed.
7/5/23 25
COST ESTIMATION
• This simply involves segregating all Cost into Variable and Fixed
within the relevant range.
• There are two broad approaches or methods used to segregate Total
Cost into Variable Cost & Fixed Cost in order to be able to predict
cost. The methods have variations.
• They are
• Estimation method using Historical (Past) Data
• Estimation method using Non-Historical Data.

7/5/23 26
ESTIMATION METHOD USING
HISTORICAL DATA
• With this method, Cost Estimation is based on Historical data.
• Here the relationship between cost and activity level as observed in the
past is studied and segregated into Variable and Fixed Cost.
• The variation under this method are;
• Accounts Classification Method
• High-Low Method
• Visual Fit Method (Scatter graph)
• Least Square Method
7/5/23 27
ACCOUNT CLASSIFICATION METHOD
• It is also called Account Analysis.
• This involves a careful examination of the organization’s ledger
accounts.
• The cost analysts based on his knowledge and experience of the
organization’s activities and cost, classifies each cost item in the
ledger as Fixed, Variable and Semi-variable e.g.

7/5/23 28
ACCOUNT CLASSIFICATION METHOD
• Direct material as Variable, Depreciation as Fixed and Utilities as
Semi-variable.
• Once cost is classified, the analyst may then examine, job cards,
labour time cards and other source documents to predict cost.
• This method has obvious drawbacks of being time consuming and
lacks objectivity as it leaves room for the analyst’s judgement.

7/5/23 29
HIGH – LOW METHOD
• This requires that the cost involved be observed at both the high and
low levels of Activity within the relevant range.
• The high and low levels are chosen from available data set.
• The difference in Cost observed at the two extremes is divided by the
change in Activity in order to determine the amount of variable cost
involved e.g. Assume that the maintenance costs for Andoh & Co have
been observed as follows within the relevant range of 5000 to 8000
Direct Labour Hours (DLH)

7/5/23 30
HIGH – LOW METHOD
• Month Direct Labour Hours Maintenance Cost Incurred
• (Activity Level)
• January 5500 750
• February 7000 900
• March 5000 700
• April 6500 850
• May 7500 950
• June 8000 1000
• July 6000 800
7/5/23 PROF. WILLIAM COFFIE 31
HIGH – LOW METHOD
• It seems obvious that there is some variable cost elements present as
maintenance increases activity level increases but not in direct
proportion.
• To separate the variable cost element from fixed cost element, relate
the change in Direct Labour Hours between the High and Low points
to the change in Costs of maintenance.

7/5/23 32
HIGH – LOW METHOD
• Direct Labour Hours Maintenance Cost Incurred
• ¢
• High Point Observed 8000 1000
• Low Point Observed 5000 700
• change 3000 300
• Variable Cost = Change in Cost
• Change in Activity

• = ¢ 300 = ¢ 0.10 per Direct Labour Hour

• 3000
7/5/23 33
HIGH – LOW METHOD
• With the Variable cost per activity level determined the Fixed Cost
element can be computed.
• This is done by taking total cost either at the High or the Low Point
and deducting the Variable cost element.
• The formulae is FC = TC – VC per Unit (Activity Level). Using the
figures: -

7/5/23 34
HIGH – LOW METHOD
Fixed Cost Element = Total Cost – Variable Cost
= ¢ 1000 – (¢ 0.10 x 8000) Labour Hours
= ¢ 1000 – 800
= ¢ 200

7/5/23 35
HIGH – LOW METHOD
• Both the fixed and variable cost are now isolated.
• The cost of maintenance within the relevant range analyzed can be
expressed as being ¢200 plus ¢0.10 per direct labour hours.
• This is called Cost Formula. i.e. ¢.10x + 200

7/5/23 36
HIGH – LOW METHOD
• This can be used to predict Total Cost of any Activity Level within the
Relevant Range.
• This method overcomes the lack of objectivity with the Accounts
Classification Methods and it is also simple.
• The drawback is that it utilizes only two points in determining the cost
formulae. The rest of the data is ignored.
• Thus accuracy in the cost analysis is affected.

7/5/23 37
VISUAL FIT OR SCATTER GRAPH
METHOD
• Here the available set of data is plotted on a graph i.e. the Cost at the
various levels of activity.
• The analyst then visually fit a line to these data by laying a ruler on the
plotted points.
• The line is positioned so that a roughly equal number of plotted points
(coordinates) lie above and below the line.
• The graph is known as Scatter Graph and the line fitted to the plotted
points is known as a Regression Line.

7/5/23 38
VISUAL FIT OR SCATTER GRAPH
METHOD
• At the Y axis i.e. vertical axis (the cost) where the regression line
intercepts the axis is the Fixed Cost.
• With the Fixed Cost known the Variable Cost at all levels of activity
can be computed.
• The advantage of this method is that unlike the High and Low method,
where only two data is utilized, all data set available is used.
• Accuracy is therefore enhanced.
• The drawback is the lack of objectivity as two analysts can draw two
Regression Lines using the same data.
7/5/23 39
VISUAL FIT OR SCATTER GRAPH
METHOD
• Cost y
• * * Regression Line
• * *
• * *
• * * Fixed Cost
• x
• Activity Level

7/5/23 40
LEAST – SQUARES METHOD
• This is a more sophisticated approach to the scatter graph idea.
• Rather than fitting a Regression Line through the scatter graph data by
simple visual inspection, the Least – Square method fits the line by
statistical analysis.
• The method is based on computations that find their foundation in the
equation for a straight line.

7/5/23 41
LEAST – SQUARES METHOD
• . A straight line can be expressed in equation form as Y = a + b x
where
• Y - is the dependent variable (total cost)
• a - is the fixed element
• b - is the degree of variability or slope of the line (per
Unit Variable Cost)
• x - is the independent variable (activity level)

7/5/23 42
LEAST – SQUARES METHOD
• From this basic equation and a given set of observations two
simultaneous linear equations can be developed that will fit a
regression line to a linear array of data.
• The equation is
• -- (1)
• eg. -- (2)

7/5/23 43
LEAST – SQUARES METHOD
• Where
• x= Activity Measure, e.g. Hours
• y = Total Mixed Cost Observed
• a = Fixed Cost
• b = Variable Rate
• n = Number of Observations.
• The equation is used to solve a and b for the elements in the Cost
Prediction formulae of y = a + bx.
7/5/23 44
LEAST – SQUARES METHOD
• The equation can be rearranged to solve for a and b as follows:

7/5/23 45
POINTS TO NOTE ABOUT USING
HISTORICAL DATA
• Regardless of the method used, the resulting cost estimates would be
only as good as the data upon which it is based.
• The following affects the data and consequently the accuracy of the
Estimation.
• Missing Data – misplaced source documents or failure to record a
transaction can result in missing data.
• Outliers – These are observations that fall far outside the others. This
could be an error or the data could be correct but due to unusual
circumstances.
7/5/23 46
POINTS TO NOTE ABOUT USING
HISTORICAL DATA
• Inflation – During periods of inflations, historical cost data may not
affect future cost behavior unless data is deflated with Price Index
which itself has its own problems.
• The effect of the Experience curve – Labour and other variable cost
declines by certain percentage as output doubles.

This effect is not given recognition in the Cost Estimation as the

assumption is that Variable Cost increases as activity level increases.

7/5/23 47
POINTS TO NOTE ABOUT USING
HISTORICAL DATA
• The single factor assumption – Other factors may affect the
variability of a total cost other than the cost driver.

For example apart from total mileage travelled that affect total
maintenance cost, age of the vehicles, being maintained the man hours
mechanics require to perform the maintenance number of fleet of
vehicles etc would affect the maintenance cost.

But only one of these factors would be assumed with the rest being
held constant.
7/5/23 48
COST ESTIMATION USING NON
HISTORICAL DATA
• With this, the process of cost estimation is to study the process that
result in cost incurrence.
• There are two variations
i. Engineering Method
ii. Conference Method

7/5/23 49
COST ESTIMATION AND INFLATION
• The behavior of Cost to changes in activity level is referred to as Cost
Behavior.
• Cost Behavior classifies organizational cost into Variable Cost and
Fixed Cost.
• Within the Relevant Range, Fixed Cost remains constant regardless of
changes in activity level (volume) whiles Variable Cost changes in
total in Direct proportion to changes in activity level.

7/5/23 50
COST ESTIMATION AND INFLATION
• The forgoing means the Total cost relating to one activity level
(volume) within a relevant range is influenced by the Total Variable
Cost associated with the activity level i.e. to say difference between
Total cost of two activity levels is accounted for by differences in
Variable Cost at least within a relevant range as illustrated.

7/5/23 51
COST ESTIMATION AND INFLATION
• Total Cost of Higher Output = Fixed Cost + Variable Cost of Output

less
• Total Cost of Lower Output = Fixed Cost + Variable Cost of Output

• Difference in Total Cost = Variable Cost of Difference in Output

7/5/23 52
COST ESTIMATION AND INFLATION
• However, in practice because of Inflation (persistence increases in
general price level) Total Cost may increase not because of volume
increase but for inflation.
• Businesses after establishing the relationship between Activity Level
and Cost do estimate cost.
• In order to estimate (determine) correctly and reliably the effect of the
inflation present in the past data should be removed.
• There are two approaches to do that.

7/5/23 53
DEFLATING PAST COST TO A BASE YEAR
• This technique is used when the High-Low Method is used to
estimate cost.
• With the high-low method only two of the past data is utilized.
• The idea behind this approach is that to be able to compare the two
extremes correctly, the related cost should be at the same inflation rate.
• This is done by deflating the past cost to Base Year Index of 100.

7/5/23 54
DEFLATING PAST COST TO A BASE YEAR
• Steps to the approach are as follows:
• (i) Deflate the Total Cost of the two extreme activity level or output.
The formulae is
• Total Cost x 100 Price
Index
• NB: Price index is the index of the Total Cost of the extremes under
consideration.
• (ii) Use High-Low Method to determine the Fixed Cost and Variable
Cost per Unit, using the deflated values of the Total Cost as in (i)
7/5/23 55
DEFLATING PAST COST TO A BASE YEAR
• (iii) Estimate (determine) the Total Cost for the future period as usual
• (iv) Inflate the Estimated Total Cost using the formulae: Total Cost x
Price Index

100
• NB: Price Index is index for the future
• With Deflating process, Price Index is used.
• Therefore where Percentage Inflation is given, it has to be converted
into Index. (See the conversion steps below)
7/5/23 56
DEFLATING PAST COST TO A BASE YEAR
• e.g. You are presented with the following data
• Year Output Total Cost Price Index
• 2000 32,500 72500 100
• 2001 40,000 89500 112
• 2002 45,000 104550 123
• 2003 30,000 100800 144
• 2004 37,500 124000 160
• What Cost should be expected in 2005 if Output is expected to be 42500
and price index is 180
7/5/23 57
DEFLATING PAST COST TO A BASE YEAR
Solution
Output Total Cost
(a) Highest 45000 104550
Lowest 30,000 100800
(b) Deflate Total Cost to Base Year
(i) 104500 x 100 = 85000
123
(ii) 100800 x 100 = 70,000 144
7/5/23 58
DEFLATING PAST COST TO A BASE YEAR
• This means the Total Cost for the two extremes would have been those
values in the year 2000 and makes Total Cost of 2002 and 2003
comparable.
(c) Determining Variable Cost per Unit
Activity Level Total Cost
High 45000 85000
Low 30,000 70,000
15000 15000

7/5/23 59
DEFLATING PAST COST TO A BASE YEAR
Variable Cost per Unit = _CL = 15000
AL = 15000

AL = ¢1
Substituting:
Total Cost of 45000 activity level 85000
Variable Cost (45000 x 1) 45000
Fixed Cost 40,000
7/5/23 60
DEFLATING PAST COST TO A BASE YEAR
(d) Cost in 2005 for 42500 Units:
Variable Cost (42500 x1) 42,500
Fixed Cost 40,000
Total Cost 82,500
(e) Inflate 2005 Total Cost @ 100 Index to 2005 Total Cost Price Index
of 180:
82500 x 180 100
= ¢148500
7/5/23 61
DEFLATING PAST COST TO A BASE YEAR
• If Past Cost is not Adjusted for inflation the Estimated Cost would
have been ¢103925 meaning an under estimation of ¢44575 (42.8%).
• This approach is very simple or less cumbersome but does not
recognize the cumulative effect of inflation or price level changes.

7/5/23 62