COST BEHAVIOUR

ANALYSIS OF SEMIVARIABLE COSTS

(OR MIXED COSTS)
 For purposes of various types of cost analyses, semivariable
costs (or mixed costs) must be broken down into their fixed
and variable elements.
 Since semivariable costs contain both fixed and variable
components, the analysis takes the following mathematical
form, which is called a cost-volume formula:
 y =a + bx
 where y = the semivariable cost to be broken up
 x = any given measure of activity, such as production volume,
sales volume, or direct labor hours
 a = the fixed cost component
 b = the variable rate per unit of x
 Several methods are available to separate a semivariable
expense into its variable and fixed components, including
 1. The high--low method
 2. The scattergraph method
 3. The method of least squares (regression analysis)
1. THE HIGH-LOW METHOD
 The high-low method, as the name indicates, uses two
extreme data points to determine the values of a (the
fixed cost portion) and b (the variable rate) in the equation
y = a + bx.
 The extreme data points are the highest representative x-y
pair and the lowest representative x-y pair.
 The activity level x, rather than the mixed cost item y,
governs their selection.
 The high-low method is explained, step by step, as
follows.
 Step 1 Select the highest pair and the lowest pair.
 Step2 Compute the variable rate, b, using the formula
Difference in cost y
Variable rate 
Difference in activity x
 Step 3 Compute the fixed cost portion as
 Fixed cost portion = Total semivariable cost -Variable cost
 Example 1
Flexible Manufacturing Company decides to relate total factory overhead costs to direct
labor hours (DLH) to develop a cost-volume formula in the form of y = a + bx. ? Twelve
monthly observations are collected.
Direct Labor Hours Factory
x Overhead y
Month (000 omitted) (000 omitted)
January 9 hours $ 15
February 19 20
March 11 14
April 14 16
May 23 25
June 12 20
July 12 20
August 22 23
September 7 14
October 13 22
November 15 18
December 17 18
Total hours 174 $225
The high-low points selected from the monthly observations are
X Y
High 23 hours $25 (May pair)
Low 7 14 (September pair)
Difference 16 hours $11
Thus,
Variable rate b = Difference in y/ Difference in x = $11 / 16 hours
= $0.6875 per DLH
The fixed cost portion is computed as
High Low
Factory overhead (y) $25 $14
Variable expense
($0.6875/DLH) (15.8125) (4.8125)
$ 9.1875 $ 9.1875
Therefore, the cost-volume formula for factory overhead is $9.1875 fixed, plus
$0.6875 per DLH. Or,
alternatively,
y’ = $9.1875+ $0.6875*DLH

where y ’ = estimated factory overhead

x = DLH
Note that the reason for using a new symbol y’ (read as y-prime) is that the cost-
volume formula just obtained gives an estimated value of y.
Task 1
The XYZ Tool Manufacturing Co. shows the following factory overhead
costs at various levels
of direct labor hours for the last four months:
Direct Labor Factory
Month Hours x Overhead y
July 2,500 hours $ 7,000
August 1,500 5,000
September 2,000 6,000
October 3,000 8,000
9,000 hours $26,000
Determine the monthly fixed overhead and the variable overhead rate per
direct labor hour (DLH) using)the high-low method.

y’ = $2,000 + $ 2 *DLH
Task 2
 1.The costs of operating the canteen at 'Eat a lot Company' for the past three months are as follows.
 Month Cost Employees
 $
 1 72,500 1,250
 2 75,000 1,300
 3 68,750 1,175
 Variable cost (per employee per month) =______________?
Variable cost per employee = $6,250/125 = $50
 Fixed cost per month =_______________? For 1,175 employees, total cost = $68,750 Total cost = variable cost +
fixed cost $68,750 = (1,175 × $50) + fixed cost
∴Fixed cost = $68,750 – $58,750 = $10,000
 2. Pen Co produced the following units at the following costs during October, November and
December.
 Month Number Total Cost of units $
 October 4,700 252,800
 November 5,500 264,000
 December 9,500 320,000
 The costs could be sub-divided into variable costs of $14 per unit and fixed costs of
$........................................per month. Variable costs = 67,200/4,800 = $14/unit
Fixed costs = Total cost – variable cost
At 9,500 units, fixed cost = $320,000 – (9,500 × $14) = $187,000
2. THE SCATTERGRAPH METHOD
 In the scattergraph method, a semivariable expense is
plotted on the vertical axis (or y axis) and an activity
measure is plotted on the horizontal axis (or x axis). Then,
a regression line is fitted by visual inspection of the
plotted x-y data.
 For purposes of illustration, let us use the data in Example
1, Table 1 . The factory overhead and direct labor hours
are plotted in Fig. 2.
 Since the regression line obtained by visual inspection
strikes the factory overhead axis at the $6 point, that
amount represents the fixed cost component. The variable
cost component is computed as
Factory overhead at 23 hours of direct labor $25
Less: Fixed cost component -6
Variable cost component $19

Therefore, the variable rate per hour is $19/23 hours = $0.8261 per DLH.

In summary, based on the scattergraph method, we obtain

y’ = $6 + $0.8261 * DLH
where y’ = estimated factory overhead
x = DLH
3. THE METHOD OF LEAST SQUARES
(REGRESSION ANALYSIS)
 One popularly used method for estimating the cost-
volume formula is regression analysis.
 Regression analysis is a statistical procedure for
estimating mathematically the average relationship
between the dependent variable y and the independent
variable(s) x.
 Simple regression involves one independent variable, such
as DLH or machine hours alone, whereas multiple
regression involves two or more activity variables.
 (We will assume simple linear regression which means that we
will maintain the y = a + bx relationship.)
 To explain the least-squares method, we define the error as
the difference between the observed value and the
estimated value of some semivariable cost and denote it
by u .
 Symbolically,
 u = y - y’
 where y = observed value of a semivariable expense
 y’ = estimated value based on y’ = a + bx
 The least-squares criterion requires that the line of best fit
be such that the sum of the squares of the errors (or the
vertical distance in below Fig. from the observed data
points to the line) is a minimum, i.e.,
minimum :  u 2   ( y  y`) 2
 The formula for a is a short-cut formula which requires
the computation of b first. This will save a considerable
amount of time. If the data collected are too voluminous
to fit into a calculator, b can be computed using
 To illustrate the computations of b and a, we will once
again refer to the data in Table1. All the sums required are
computed and shown below.