Assume
we want to analyze the behavior of total expenses based upon sales volume. The
following 12 months of data is available to us. The following is an Excel file.
A
B
|
Units |
Total |
|
Sold |
Expenses |
1 |
76,506 |
$348,670 |
2 |
75,877 |
$352,090 |
3 |
78,679 |
$363,520 |
4 |
70,470 |
$347,290 |
5 |
67,699 |
$319,030 |
6 |
72,413 |
$331,630 |
7 |
80,629 |
$384,280 |
8 |
78,737 |
$411,730 |
9 |
74,299 |
$365,800 |
10 |
73,814 |
$416,980 |
11 |
67,870 |
$380,740 |
12 |
82,398 |
$424,210 |
Assume
the following results are obtained from a regression analysis:
SUMMARYOUTPUT |
|
Regression
Statistics |
|
Multiple
R |
0.568832 |
|
Coefficients |
t
Stat |
Intercept |
67637.23289 |
0.487542 |
X
Variable 1 |
4.040871218 |
2.187122 |
Given
the above data, the behavior of total expenses can be expressed as follow:
That
is, $67,637 is the amount of monthly expenses regardless of sales volume (also
called fixed expenses). The variable expense per unit is $4.04 per unit sold.
Additional
Analysis
According
to the value of "R Square" (57%) total expenses has a relatively a
moderate relationship with sales units. Note that, the value of R Square is
between 0 and 100%. The closer the
value of R Square to 100%, the higher is the reliability of the
relationship.
According
to the “t Stat.” Values, it is very likely that total expenses will change
$4.04 when sales volume changes (increases or decreases) by one unit. As
a generally accepted practice, if the absolute value of "t Stat" is
greater than 2, then the estimated value is highly reliable. However, the
likelihood of having $67,637 of fixed expenses per month is low as the “t
Stat.” for it is less than 2.