5-1

5-2

Learning Objectives
Cost Estimation LO 5-1 Understand the reasons for estimating fixed and variable costs.
LO 5-2 Estimate costs using engineering estimates.
LO 5-3 Estimate costs using account analysis.
Chapter 5 LO 5-4 Estimate costs using statistical analysis.
LO 5-5 Interpret the results of regression output.
LO 5-6 Identify potential problems with regression data.
LO 5-7 Evaluate the advantages and disadvantages of alternative
PowerPoint Authors:
cost estimation methods.
Susan Coomer Galbreath, Ph.D., CPA
Jon A. Booker, Ph.D., CPA, CIA LO 5-8 (Appendix A)
Cynthia J. Rooney, Ph.D., CPA Use Microsoft Excel to perform a regression analysis.
LO 5-9 (Appendix B)
Understand the mathematical relationship describing
the learning phenomenon.

Copyright © 2017 by The McGraw-Hill Companies, Inc. All rights reserved.

1 2

LO
5-3 5-4
5-1

Why Estimate Costs? Basic Cost Behavior Patterns

Managers make decisions and need to compare LO 5-1 Understand the reasons for estimating fixed and variable costs.
costs and benefits among alternative actions.
Costs

Fixed costs Variable costs

Cost estimates can be an important element Total fixed costs do not Total variable costs
In helping managers make decisions. change proportionately change proportionately
as activity changes. as activity changes.

Per unit fixed costs Per unit variable cost

change inversely as remain constant as
activity changes. activity changes.

3 4

Methods Used to Estimate

LO LO
5-5 5-6
5-1 5-2

Cost Behavior Engineering Estimates

LO 5-2 Estimate costs using engineering estimates.
Charlene, owner of Charlene’s Computer Care
(3C), wants to estimate the cost of a Cost estimates are based on measuring and
new computer repair center. then pricing the work involved in a task.

Identify the activities involved:

– Labor
– Rent
1. Engineering estimates – Insurance
2. Account analysis
3. Statistical methods Estimate the time and cost for each activity.

5 6
 5-2

LO LO
5-7 5-8
5-2 5-3

Engineering Estimates Account Analysis Method

LO 5-3 Estimate costs using account analysis.

Details each step required to perform an operation Review each account comprising the total
Permits comparison of other centers with similar operations cost being analyzed.

Costs for totally new activities can be estimated without Identify each cost as either fixed or variable.
prior activity data.

Fixed Variable

Can be quite expensive to use

Based on optimal conditions

7 8

LO LO
5-9 5-10
5-3 5-3

Account Analysis Method Account Analysis Method

Costs for 360 repair-hours Fixed costs + (Variable cost/unit × No. of units) = Total cost
Variable Fixed
Account Total cost cost
Cost at 360 repair-hours:
Office rent $ 3,375 $1,375 $2,000 $6,216 + ($12.25 × 360) = $10,626
Utilities 310 100 210
Administration 3,386 186 3,200
Supplies 2,276 2,176 100 Cost at 480 repair-hours:
Training 666 316 350 $6,216 + ($12.25 × 480) = $12,096
Other 613 257 356
Total $10,626 $4,410 $6,216
Per repair hour $12.25

9 10

LO LO
5-11 5-12
5-3 5-4

Account Analysis Method Statistical Cost Estimation

LO 5-4 Estimate costs using statistical analysis.

Managers and accountants are familiar with company
operations and the way costs react to changes in
activity levels.
Managers and accountants may be biased.
Analyze costs within a relevant range, which is
the limits within which a cost estimate may be valid.
Relevant range for a projection is usually between
the upper and lower limits (bounds) of past activity
levels for which data is available.

Decisions often have major economic consequences

for managers and accountants.

11 12
 5-3

LO LO
5-4
Overhead Cost Estimation 5-13
5-4
5-14

for 3C Scattergraph

These data will be used to estimate costs using Does it look like a relationship exists
a statistical analysis. between repair-hours and overhead costs?

13 14

LO LO
5-15 5-16
5-4 5-4

Scattergraph Hi-Low Cost Estimation

This is a method to estimate cost based on two cost
observations, the highest and lowest activity levels.

Overhead Repair-
Month costs hours
High $12,883 568
Low $ 9,054 200
Change $ 3,829 368
We use “eyeball judgment” to determine the
intercept and slope of the line.

15 16

LO LO
5-17 5-18
5-4 5-4

Hi-Low Cost Estimation Hi-Low Cost Estimation

Variable cost per unit (V) =
Variable cost ($12,883 – $9,054) $3,829 $10.40
= = =
(Cost at highest activity level – Cost at lowest activity level) per RH (V) 568 RH – 200 RH 368 RH per RH
(Highest activity level – Lowest activity level)

Fixed cost (F) = Fixed costs (F) = ($12,883 – ($10.40 × 568 RH) = $6,976
Total cost at
– (Variable cost × Highest activity level)
highest activity

Fixed costs (F) = ($9,054 – ($10.40 × 200 RH) = $6,974

Total cost at
– (Variable cost × Lowest activity level)
lowest activity Rounding
difference

17 18
 5-4

LO LO
5-19 5-20
5-4 5-4

Hi-Low Cost Estimation Regression Analysis

How do we estimate manufacturing Regression is a statistical procedure to
overhead with 480 repair-hours? determine the relation between variables.

It helps managers determine how well the

TC = F + VX estimated regression equation describes
the relations between costs and activities.
TC = $6,976 + ($10.40 × 480) = $11,968

19 20

LO LO
5-21 5-22
5-4 5-4

Regression Analysis Regression Analysis

Hi-low method:
Uses two data points

Regression: Y = a + bX
Uses all of the data points
Y = Intercept + (Slope × X)

For 3C:
OH = Fixed costs + (V × Repair-hours)

21 22

LO LO
5-23 5-24
5-5 5-5

Interpreting Regression Interpreting Regression

LO 5-5 Interpret the results of regression output.
The computer output of 3C’s scattergraph
Independent variable: gives the following formula:
– The X term, or predictor
– The activity that predicts (causes) Total overhead = $6,472 + ($12.52 per RH × No. of RH)
the change in costs
Activities:
– Repair-hours Estimate 3C’s overhead with 480 repair hours.
Dependent variable: TC = F + VX
– The Y term
– The dependent variable TC = $6,472 + ($12.52 × 480) = $12,482
– The cost to be estimated

Costs:
– Overhead costs

23 24
 5-5

LO LO
5-25 5-26
5-5 5-5

Interpreting Regression Interpreting Regression

Correlation coefficient (R):
This measures the linear relationship between
variables. The closer R is to 1.0, the closer the
points are to the regression line. The closer R is
to zero, the poorer the fit of the regression line.
T-statistic:
This is the value of the estimated coefficient, b, divided
by its estimated standard error (Seb). Generally, if it is
over 2, then it is considered significant. If significant,
the cost is NOT totally fixed.

From the data used in the 3C regression, the t-statistic is:

Coefficient of determination (R2):
This is the square of the correlation coefficient. t = b ÷ Seb
It is the proportion of the variation in the dependent = 12.5230 ÷ 1.5843
variable (Y) explained by the independent variable(s) (X). = 7.9044

25 26

LO LO
5-27 5-28
5-5 5-5

Interpreting Regression Multiple Regression

Multiple regression:
When more than one predictor (x) is in the model
An 0.91 correlation coefficient means that a linear
relationship does exists between repair hours Is repair-hours the only activity that drives
and overhead costs. overhead costs at 3C?

An 0.828 coefficient of determination means that Predictors:

82.8% of the changes in overhead costs can be X1: Repair-hours
explained by changes in repair-hours. X2: Parts cost

Both have t-statistics that are greater than 2, Equation:

so the cost is not totally fixed. TC = VC(X1) + VC(X2) + FC

27 28

LO LO
5-29 5-30
5-5 5-5

Multiple Regression Output Multiple Regression Output

The adjusted R-squared is the correlation coefficient
squared and adjusted for the number of independent TC = F + V1X1 + V2X2
variables used to make the estimate.

The statistics supplied with the output (rounded off) are: TC = $6,416 + ($8.61 × 480) + (77% × $3,500)
– Correlation coefficient (R) = 0.953
– R2 = 0.908 TC = $13,244
– Adjusted R2 = 0.892

29 30
 5-6

LO LO
5-6
Practical Implementation 5-31
5-6
Practical Implementation 5-32

Problems Problems
LO 5-6 Identify potential problems with regression data. The Effect of Nonlinear Relations
$800
Assumed actual cost function
Effect of: $700
– Nonlinear relations $600
– Outliers $500

Cost
– Spurious relations Regression
$400
– Using data that do not fit the assumptions estimate
$300
of regression analysis Relevant
$200 Capacity
range
$100
$0
0 5 10 15 20 25 30 35 40
Volume

31 32

LO LO
5-6
Practical Implementation 5-33
5-6
Practical Implementation 5-34

Problems Problems
The Effect of Outliers on the Computed Regression
Problem:
• Attempting to fit a linear model to nonlinear data
• Likely to occur near full-capacity Computed
regression line

Solution:
• Define a more limited relevant range.
(Example: from 25-75% capacity) “Outlier”
• Design a nonlinear model. True regression line

33 34

LO LO
5-6
Practical Implementation 5-35
5-6
Practical Implementation 5-36

Problems Problems
The Effect of Spurious Relations
Problem:
Outliers move the regression line. Problem:
Using too many variables in the regression
(i.e., using direct labor to explain materials costs).
Solution:
Prepare a scattergraph, analyze the graph, Although the association is very high, actually
and eliminate highly unusual observations both are driven by output.
before running the regression.
Solution:
Carefully analyze each variable and determine
the relationship among all elements before
using in the regression.

35 36
 5-7

LO LO
5-6
Practical Implementation 5-37
5-6
5-38

Problems Learning Phenomenon

The Effect of Using Data That Do Not
Fit the Assumptions of Regression
Learning phenomenon is the systematic relationship
Problem: between the amount of experience in performing
If the assumptions in the regression are not a task and the time required to perform it.
satisfied, then the regression is not reliable.

Solution:
There is no clear solution. Limit time to help
assure costs behavior remains constant,
yet this causes the model to be weaker
due to less data.

37 38

LO LO
5-39 5-40
5-7 5-7

How an Estimation Method is Chosen Data Problems

LO 5-7 Evaluate the advantages and disadvantages
of alternative cost estimation methods.
• Missing data

• Outliers
• Reliance on historical data is relatively inexpensive.
• Computational tools allow for more data to be used • Allocated and discretionary costs
than for non-statistical methods.
• Inflation

• Mismatched time periods

• Reliance on historical data may be the only readily
available, cost-effective basis for estimating costs.
• Analysts must be alert to cost-activity changes.

39 40

LO LO
5-7
Effect of Different Methods 5-41
5-8
Appendix A: Regression 5-42

on Cost Estimates Analysis Using Microsoft Excel

Estimated manufacturing overhead with 480 repair-hours. LO 5-8 (Appendix A) Use Microsoft Excel
to perform a regression analysis.
Total Estimated Estimated
Many software programs exist to aid in performing
estimated fixed variable
Method costs costs costs regression analysis.
Account analysis $12,096 $6,216 $12.25/repair-hour In order to use Microsoft Excel, the Analysis Tool Pak
must be installed.
High-low $11,968 $6,976 $10.40/repair-hour
Data is entered and the user then selects the data and
Simple regression $12,482 $6,472 $12.52/repair-hour type of regression analysis to be generated.

Multiple regression $13,244 $6,416 $8.61/repair-hour The analyst must be well schooled in regression in order
+ 77% of parts cost to determine the meaning of the output.

41 42
 5-8

LO
5-43 5-44
5-9

Appendix B: Learning Curves End of Chapter 5

LO 5-9 (Appendix B) Understand the mathematical relationship
describing the learning phenomenon.
This is the systematic relationship between
the amount of experience in performing a task
and the time required to perform it.
Time to Calculation
Unit produce of time
First unit 100.0 hours (assumed)
Second unit 80.0 hours 80% × 100 hours
Fourth unit 64.0 hours 80% × 80 hours
Eighth unit 51.2 hours 80% × 64 hours

• Impact: It causes the unit price to decrease

as production increases.
• This implies a nonlinear model.

43 44