Cost behaviour: Analyses and Estimation

 On completion of this study unit you should

be able to:
 define the basic cost behaviour patterns;
 classify costs as either fixed, variable or
mixed;
 distinguish between mixed cost and step-
fixed costs;
 explain the influence of a change in activity
level on the cost behaviour patterns;
 explain why mixed costs must be separated
into their fixed and variable cost elements.
 use the high-low method, scatter-graph
method and the least square regression
method to analyse mixed costs into their
fixed and variable cost components;
 to do cost estimation by using a cost
equation.
 Period cost?

 Product cost?

 Variable cost??

 Fixed cost??
 Costs are classified as fixed of variable
based on cost behavioural paterns
 Cost behaviour has to do with the way in

which costs react to changes in the

business activity
 Cost behaviour is important for accurate

cost predictions
 Variable costs:
costs
◦ Total variable costs change in relation to the
activity level
◦ Variable cost per unit stays constant within the
relevant range
 Step fixed costs:
costs
◦ A range of increases in fixed costs over relative
short ranges of production volumes within the
relevant range
◦ Increases in jumps
1 TRUCK – 2 ton maize
N$ 5 000 – 1 TRUCK's rent
1 to 8 ton maize – relevant range

Rent
2 ton maize – N$5 000
3 – 4 ton maize – N$10 000
5-6 ton maize – N$15 000

It changes in STEPS
 Fixed costs:
costs
◦ Total fixed cost stays constant within the
relevant range irrespective of the change in
activity level
◦ Fixed cost per unit changes if the activity level
changes
 Mixed costs (semi-variable costs):
Costs with a fixed and variable
component, eg electricity, telephone.
 The relevant range
Identifying costs as fixed or variable is
only valid within a specific range, thus the
range is where the total fixed costs and
variable cost per unit do not change.
 Mixed cost consists of a fixed and
variable component
 Can be plotted on a graph
 Can also be expressed as a function
 Graph: Draw fixed cost line first
200
180
160
140
Cost (Rand)

120
100
Fixed cost
80
R75.00
60
40
20
0
0 20 40 60 80
 Graph: Draw variable cost line
200
180
160
140
Cost (Rand)

120
100
80
Variable
60
cost
40 R2.00/minut
20 e
0
0 20 40 60 80
 Graph: Graphs combined
200
180
160
140
Cost (Rand)

120
100
Fixed cost
80
R75.00
60
Variable
40 cost
20 R2.00/minut
0 e
0 20 40 60 80
 Mixed cost consists of a fixed and
variable component
 Graph: 200
180
Total mixed
160
cost
140
Cost (Rand)

120
100
Fixed cost
80
R75.00
60
40 Variable
20 cost
R2.00/minut
0
e
0 20 40 60 80
 Function: y = a + bX
where y = Total mixed cost
a = Total fixed cost
b = Variable cost per cost driver
X = Activity level (Cost driver)
 Thus for our example:
y = R75.00 + (R2/min x X)
 And for 50 minutes
y = R75.00 + (R2/min x 50 min)
y = R175.00
 Four methods
• High-low method
• Scattergraph method
• Least squares regression method
• Multiple regression analysis
 Based on linear relationship between total
mixed cost and the activity level
 Use only two observations – highest and

lowest activity levels

 Method assumes that fixed cost will stay

constant during the period within the

relevant range
 The difference in total mixed cost can thus

only be as a result of a change in variable

cost.
 Month Activity level Total
(machine electricity cost
hours)
January 3 500 R6 625
February 2 800 R6 100
March 3 400 R6 550
April 4 200 R7 150
May 3 700 R6 775
June 3 000 R6 250
July 2 500 R5 875
August 2 200 R6 650
September 2 000 R5 500
October 3 700 R6 775
November 4 500 R7 375
December 4 100 R7 075
39 600 R77 700
 Month Activity level Total
(machine electricity cost
hours)
January 3 500 R6 625
February 2 800 R6 100
March 3 400 R6 550
April 4 200 R7 150
May 3 700 R6 775
June 3 000 R6 250
July 2 500 R5 875
August 2 200 R6 650
September 2 000 R5 500
October 3 700 R6 775
November 4 500 R7 375
December 4 100 R7 075
39 600 R77 700
 Month Activity level Total
(machine electricity cost
hours)
Highest: 4 500 R7 375
November
Lowest: 2 000 R5 500
September
2 500 R1 875
 The difference in total mixed cost (R1 875) is as a
Difference
results of a change in the activity level (2
500mh).
 The variable cost can be calculated as follows:
 Variable cost per ∆ in mix cost
cost driver = ∆ in act level

 Variable cost per R1 875

machine hour = 2 500 mh

= R0.75/machine hour
 Now that the variable cost per machine hour
is known, the total fixed cost can be
calculated by means of the cost function:
 Y = a + bX
 Substitute any of the two points in the
function:
R7 375 = a + (R0.75/mh x 4 500mh)
a = R7 375 – R3 375
= R4 000
 Cost function: Y = R4 000 + R0.75X(mh)
 A range of observations are plotted on a
graph, and then a regression line is drawn.
 The regression line represents the average

of all the observations, thus also the

average total cost of the mixed cost
 The Y-intercept represents the fixed cost

element
 The slope of the line represents the

variable cost
 8000
7000
Total cost (Rand)

6000
5000
4000
3000
2000
1000
0
0 1000 2000 3000 4000 5000
Activity level (Machine hours)
 8000
7000
Total cost (Rand)

6000
5000
4000
3000 R4 000
2000 Y-intercept
represents
1000
the fixed cost
0
0 1000 2000 3000 4000 5000
Activity level (Machine hours)
 The regression line intercepts the y-axis at
R4 000. Thus the fixed cost is R 4 000.
 The slope (b) can now be calculated by

substitution in the function – NB The point

chosen for substitution should be as close
as possible to the regression line:
 8000
7000

Total cost (Rand)

6000
5000 4 500; R7 375
4000
3000
2000
1000
0
0 1000 2000 3000 4000 5000
Activity level (Machine hours)
 Formula: Y = a + bX
R7 375 = R4 000 + b(4 500)
4 500b = R3 375
b = 0.75
 Thus the variable cost is R0.75/mh
 Thus the cost function is as follows:

Y = R4 000 + R0.75/mh
 Statistical method that takes all points into
account.
 Easy with calculator or computer (Excel)
 Formulas:
A. ΣXY = aΣX + bΣX²
B. ΣY = na + bΣX
Formulas:
A. b = n(ΣXY) – (ΣX)(ΣY)
n(ΣX²) – (ΣX)²

B. a = (ΣY) – b (ΣX)
n
 X = Activity level
 Y = total mixed cost for the activity
level
 A = total fixed cost
 B = variable cost per cost driver
 N = the number of observations
 Σ = sigma – means the sum of
 Month Activity level Total XY X²
(machine electricity
hours) (X) cost (Y)
January 3 500 R6 625 23 187 12 250
500 000
February 2 800 R6 100 17 080 7 840 000
000
March 3 400 R6 550 22 270 11 560
000 000
April 4 200 R7 150 30 030 17 640
000 000
May 3 700 R6 775 25 067 13 690
500 000
June 3 000 R6 250 18 750 9 000 000
000
July 2 500 R5 875 14 687 6 250 000
500
August 2 200 R5 650 12 430 4 840 000
 b = n(ΣXY) – (ΣX)(ΣY)
n(ΣX²) – (ΣX)²

= (12 x 261 765 000) – (39 600 x 77 700)

(12 x 137 820 000) – (39 600)²

= 3 141 180 000 – 3 076 920 000

1 653 840 000 – 1 568 160 000

= 64 260 000
85 680 000

= 0.75
 a = (ΣY) – b (ΣX)
n

= 77 700 – (0.75 x 39 600)

12

= 48 000
12

= 4 000

 Thus the cost function is: Y = R4 000 +

R0.75/mh
Evaluation of regression line
Economic plausibility Is there a
relationship between cost and driver
Goodness of fit Does line fit data,
vertical distance from line, R₂ give indication
Significance of independent variable
steep slope show strong relationship
 High-low method
• Only focused on two activity levels and thus not
representative of the range
• Only highest and lowest points are used – can
be unusually high or low
 Scattergraph method
• More representative of range
• Drawn by hand – can thus be inaccurate; also
difficult to determine the precise Y-intercept
 Least square regression method
• All the data are taken into account
• Data statistically calculated
 Use the cost function to predict costs at a
certain activity level
 Example:
• Cost function – at activity of 5 000 mh
• Y = a + bX
Y = R4 000 + R0.75X(activity)
Y = R4 000 + R0.75 (5 000mh)
Y = R7 750
 Thus at an activity level of 5 000mh the
estimated electricity cost would be R7 750
 Managerial activities are dependant on accurate
cost estimation – it is necessary to understand
cost behaviour
 To understand cost behaviour costs should be
classified into fixed and variable cost
 A relevant range exists within which behavioural
paterns of costs are valid
 Mixed costs consists of a variable and a fixed
component and should be analysed by one of the
following techniques:
◦ High Low method
◦ Scattergraph method
◦ Least square regression method
 Cost equation can be set