-------
-------
I. GENERAL ANALYTIC APPROACH
A. Introduction
The financial responsibility regulations under RCRA reoroposed on
May 19, 1980 included a financial test as an option for demonstrating a
firm's capability to meet the costs of its facility closure and post-
closure obligations. The proposed financial test for closure and oost-
closure care included elements designed to verify the current ability of
a firm to pay such costs (net working capital equal to or greater than
twice the estimated closure and/or post-closure costs: a minimum of S10
million in net worth), and a financial ratio designed to indicate the
overall viability of a firm (a total liabilities to net worth ratio of
less than 3 to 1). The financial test provision prompted many public
comments. Several commenters expressed general reservations about the
validity of the financial ratio used as an indicator of viability, or
questioned the net working capital or net worth reauirements, and
suggested that an evaluation of financial test performance should be
carried out to determine the effectiveness of financial tests. The
Agency agreed with this proposal, and conducted a rigorous and compre-
hensive evaluation of over 300 alternative financial tests.
This Appendix describes the methodology employed in performing this
evaluation and summarizes the results of the analysis which provide a
set of "best" candidate tests. The results of special analyses,
performed to determine the comparative effectiveness of one-year versus
three-year eligibility requirements, and the need for specialized,
1-1 o
-------
industry-specific ratio tests, are also summarized in this Appendix.
The "best" candidate tests were further evaluated against a set of
algorithms designed to estimate the total direct public and private
costs that would result from their implementation. The results of the
cost calculations performed, and their implications for the final selec-
tion of a financial test, are discussed in Appendix B of this Background
Document.
B. Possible Elements of Financial Tests
The ideal financial test would pass all firms capable of meeting
their closure and post-closure obligations while failing all firms that
would enter bankruptcy and not be able to meet those obligations.
Unfortunately, no method of financial forecasting is capable of
achieving this ideal. It is therefore useful to examine a ranee of
possible financial tests. Some of these tests have the advantage of
eliminating almost all firms that would be unable .to meet their closure
and post-closure obligations, but at the same time they also eliminate a
large number of firms that could meet those obligations: other tests
have the advantage of passing a much higher percentage of the firms that
could meet those obligations but they eliminate fewer firms that could
not meet their obligations.
A firm may be unable to meet its closure and post-closure
obligations for a variety of reasons. In addition to the ordinary
business misfortunes which may lead a firm into bankruptcy, a firm
owning a hazardous waste treatment, storage or disposal facility (TSDF)
might fail to meet those specific obligations as a result of unique
1-2
-------
events. For example, a firm that might otherwise not have entered
bankruptcy may be forced into bankruptcy if its permit is suddenly
withdrawn and it suddenly has to put up the necessary funds for closure
and post-closure. A firm also might enter bankruptcy unexpectedly as a
result of large cleanup or repair expenditures unique to a hazardous
waste TSDF or of a large liability judgment.
This Appendix analyzes the performance of different possible finan-
cial tests, with respect to the percentage of firms which pass the tests
and could meet their financial obligations, and the percentage of firms
that could not meet their obligations but still pass a given test. The
question of what test is to be preferred must then be determined by the
relative costs associated with the different tests.
Three types of elements which might be used as criteria in^finan-
cial tests are considered in this Appendix:
(1) Financial ratios: These are ratios of the financial variables
found in the income statements or balance sheets of firms.
(2) Multiples: In this text, these are measures of unencumbered
assets and/or liquidity as multiples of the specific financial obliga-
tion considered (closure and post-closure).
(3) Net worth: This is a requirement, independent of the multiple
requirement described above, that a firm must have net worth above a
specified level.
Each of these possible elements of financial tests has a ootential
role in limiting the number of firms that will fail to meet their finan-
cial obligations. By requiring firms to meet specific requirements with
1-3
-------
respect to their financial ratios in order to pass a financial test, it
is possible to limit the percentage of firms which pass the financial
test but which ultimately enter bankruptcy. Multiples serve to ensure
that the financial obligation itself does not? have so great an effect on
the firm that it is a cause for bankruptcy. Without multiples as a part
of the test, a small firm with excellent financial ratios might still be
unable to meet its closure and post-closure obligations because of the
large size of their obligations relative to the size of the firm
itself. A minimum net worth requirement serves two purposes: (1) it
has been found that the larger the firm the less likely it is to fail,
and (2) significant size enables the firm to withstand unusual
contingencies associated with the ownership of hazardous waste TSDFs.
Financial ratios and minimum net worth requirements both can help to
ensure that the failure rate of the firms which will pass the financial
test is lower than the average failure rate for all firms. The multiple
element of the test and the minimum net worth requirement help to ensure
that the failure rate of the firms considered is not higher than that
for firms in all industries, due to the unusual size of these special
financial obligations relative to the size of the firm or due to contin-
gencies associated with the ownership of hazardous waste TSDFs.
C. Overview of the Methodology
This Section, and Figure 1-1, provide an overview of this Appendix
by outlining the major steps of the methodology employed. Detailed
discussions of the topics outlined in this section are then presented in
the following sections of this Appendix. Figure 1-1 illustrates in
1-4
-------
Effect of
Multiples
I
Ul
Share of
All Klrms
Wlilch Fall
(F)
F x
Percentage
of Bankrupt
Sample That
Pass Test
(Up = Mls-
classlfled
bankrupt
firms)
Average
Failure
Rate for
All Firms
I
FxM,,
+ (F x Mp)
(Firms which do go
bankrupt as a
share of all firms
passing test)
x 10,000
Effect of
Net Worth
Requirement
\
Share of
All Firms ..
Percentage
of Non-Bank-
nv . rupt Sample
Not Falling l '"NB That Pass
(1-F) Test
(ANB>
Bankrupt
Firm
Sample
Tests
Ratios
Employed
- in tests -
and Cut-
off Points
Evaluation
of
Individual
Ratios
Against
Samples
Comments
Non-
Bankrupt
Firm
Sample
FIGURE 1-1
APPENDIX A METHODOLOGY
-------
particular the steps which were used to derive the effectiveness measure
(Ep) of any given test, that, is, the most probable estimate of the
number of firms per 10,000 which pass a given financial test but later
fail without providing alternative financial assurance. ,
1. Determination of Baseline Failure Rate
In order to compare the effectiveness of various tests, with
respect to the relative costs associated with these tests, it is neces-
sary to determine the failure rates of the set of firms which may use a
financial test as evidence of financial responsibility. The baseline
failure rate (F) is derived by first determining the average failure
rate for all firms based on the historical data, and then by adjusting
this rate to account for the effect of a net worth requirement on the
average failure rate for firms which may be eligible for a financial
test. In addition to determining the failure rate of firms, a financial
test includes a multiple requirement to provide assurance that at the
time a firm ceases to pass the test it will continue to possess assets
sufficient so that it is able to pay for closure and post-closure
costs. Otherwise, it is possible that the attempt to pay these costs
will be sufficient to force the firm into bankruptcy before these
closure and post-closure costs have been fully paid. In the
nomenclature adopted in this Appendix, the fraction of all firms consid-
ered which will go bankrupt is designated as F; the fraction of all
firms considered which will not go bankrupt is equal to 1-F.
1-6
-------
2. Selection of Financial Ratios
The financial ratios to be evaluated as candidate ratios for the
financial test were selected from three sources: (1) comments received
on the May 19, 1980 regulation; (2) surveys of opinions of bond ratine
services and credit analysts; and (3) existing literature on bankruptcy
forecasting. The list of ratios assembled was extensive and varied
widely in content and estimated effectiveness.
An initial set of financial ratios was chosen for further evalu-
ation because they satisfied three basic conditions: (1) they produced
significant predictive results in the prior literature; (2) they were
frequently identified by bond rating services and credit analysts as key
Darameters: and (3) their values were readily available from corporate
balance sheet data.
3. Sample Selection
In order to determine the effect of financial ratio tests, the
Agency gathered samples of non-bankrupt and bankrupt firms. Non-
bankrupt firms, also referred to as viable firms, are firms that did not
enter bankruptcy within three years of the time period for which data
were gathered. For bankrupt firms, financial data were gathered for
three years prior to the time the firm entered bankruptcy. The sample
of bankrupt firms was gathered from a variety of sources. The sample of
non-bankrupt firms was gathered using the Industry Index of Moody's
Industrial Manual to determine members of industrial categories that
generate, treat, store, or disnose of hazardous waste on-site.
1-7
-------
4. Evaluation of Financial Ratios
These ratios were first tested individually against the primary
bankrupt and non-bankrupt firm samples, using a variety of pass-fail
cutoff points for each ratio derived from the bankruptcy forecasting
literature. A second set of financial ratios was then selected and
tested against a sample of bankrupt and non-bankrupt firms which were
consistently misdassified by the original ratios, to determine if there
were additional financial ratios which would improve the predictive
ability of a test.
5. Design of Financial Tests
The most promising ratios were combined to form 151 alternate
multi-ratio tests. The ratios were combined in three ways: (1) two-
ratio tests (firms must pass both elements of the test to pass); (2)
three-ratio tests (firms must pass all three elements of the test to
pass); and (3) three-ratio contingent tests (firms must pass two of
three conditions to pass). The tests were then retested against the
primary sample to derive for each test the percentage of bankrupt firms
of the sample that pass the test and are therefore misclassified by the
test (Mp) and the percentage of the non-bankrupt firms sampled which
pass the test (Ajjg).
6. Determining the Effectiveness Measure (Ep) for a Given Test
The effectiveness measure for any given test is calculated accord-
ing to a two-step process (see Figure 1-1 above.) First, the fraction
of all firms which will go bankrupt (F) is multiplied by the percentage
of the bankrupt firm sample that is misclassified by a given test (Mp),
1-8
-------
that Is, pass the test but later enter bankruptcy. This provides the
fraction of all firms considered that pass a particular test and later
go bankrupt (F x Mp). Similarly, the share of all firms that will not
fail (1-F) is multiplied by the percentage of the non-bankrupt firm
sample which pass the test (A^). This provides the fraction of firms
that pass a given test and do not go bankrupt (l-F)AIjB.
Finally, the Agency uses F x Mp and (l-F)ANg to calculate what
fraction of all firms that pass a given test are firms which will go
bankrupt (in Figure 1-1 this operation is represented by F x Mp divided
by the total of all firms which pass the test ((1-F)ANB + (F x Mp)).
This fraction is then multiplied by 10,000 to yield the number of firms
per 10,000 that pass a given financial test that will go bankrupt. This
number (Ep) is the measure of effectiveness of a given test.
7. Use of Performance Curves to Select the Best Financial Test
As discussed above, a financial test has two basic performance
measures: (1) the percentage of non-bankrupt firms that can pass the
test (Ajjj): and (2) the number of firms per 10,000 that pass the test
that will later enter bankruptcy without providing alternative financial
assurance (Ep). The Agency developed performance curves consisting of
those tests which pass the maximum percentage of non-bankrupt firms for
any given percentage of firms that pass the test and later enter bank-
ruptcy. The Agency used these curves in the analysis described in
Appendix B to evaluate the relative costs of alternative tests and to
determine a best single financial test.
1-9
-------
8. Comparison to Other Tests
The resulting performance curves for the tests developed In this
study were then compared to the May 19, 1980 proposed financial tests
and the results of other bankruptcy forecasting studies.
9. Industry-Specific Problems
Although a given financial test may serve quite well for most
industries, it is possible that unique characteristics of a specific
industry may cause a high percentage of the viable firms in that
industry to fail a given test. In order to determine the importance of
such problems, the Agency examined the financial characteristics of
firms in various manufacturing industries likely to dispose of hazardous
waste, electric utilities, and hazardous waste management firms.
10. The Role of Bond Ratings in a Financial Test
Section VII of this Appendix discusses the use of bond ratings as
either a substitute for or a supplement to other elements of the finan-
cial tests considered elsewhere in this Appendix.
D. Comparison to Alternative Methodological Approaches
The methodology adopted by the Agency and outlined in Section I.C.
is not the only methodology that might be employed to derive a set of
financial tests. This Section considers three alternative approaches:
(1) the use of functions of financial ratios rather than cutoff points
for financial ratios: (2) the development of industry-specific financial
tests rather than single tests for all industries: and (3) the use of
population sampling instead of the sampling procedure employed in this
analysis.
1-10
-------
1. The Use of Functions of Financial Ratios
The approach used by the Agency was to examine a series of finan-
cial tests consisting of requirements that specific financial ratios be
at certain levels. An alternative approach would be to design financial
tests that consisted of linear or nonlinear functions of a set of finan-
cial ratios (for example, a linear function of financial ratios would be
a test that required that (a x Rj) + (b x R2) + (c x R3) > d, where a,
b, c, and d are constants and Rp R2, and R-j are ratios). Three methods
have been proposed in the literature for developing financial tests that
consist of functions of financial ratios.
One approach is to rely on expert judgment to decide which ratios
should be included and what values should be associated with them (see
Tamari, 1966). This approach has the disadvantage that there is no
statistical method of determining whether the ratings or the financial
ratios chosen have been appropriately determined.
A second approach that has been proposed is to develop a functional
model of the factors that lead to business failure and to use this model
to estimate the probabilities of business failure for any specific
firm. Though theoretically elegant, this approach has only been
attempted once in the literature reviewed for this study, and that
attempt did not include any statistical validation of the model (Wilcox,
1971, as discussed in McWilliams, 1977).
The Agency decided not to develop a functional model for two
reasons. First, the time available in which to carry out the study of a
financial test option was relatively short and might not have allowed
1-11
-------
sufficient research and refinement of the model to produce useful
results. Second, because the technique was untested in this applica-
tion, the Agency was concerned that it might prove unsatisfactory, and
that, if it were not effective, insufficient time would be available to
carry out an alternative analysis.
The third approach, and that most commonly used in the bankruptcy
forecasting literature, is that of multi-discriminant analysis. This is
a statistical search technique for arriving at the best linear or non-
linear function for a set of chosen financial ratios for discriminating
between bankrupt and non-bankrupt firms. This is the most statistically
sophisticated approach to the problem. The use of this approach was
rejected in this study for two reasons: (1) a function of financial
ratios would be more difficult to compute, more difficult to administer,
and more difficult to check quickly than a simple set of ratio require-
ments; and (2) it has been pointed out several times in the literature
on bankruptcy forecasting that multi-discriminant analysis has not
yielded more accurate forecasts of bankruptcy than the simpler approach
adopted here (Deakin, 1972 and McWilliams, 1977). A comparison of the
results of the financial tests developed in this Appendix to financial
tests developed in bankruptcy forecasting studies that employed multi-
discriminant analysis is included in Section V.
2. Use of Industry-Specific Tests
A number of commenters suggested that the Agency adopt specific
tests for different industries. The problem with this approach is the
extreme difficulty of gathering sizable samples of bankrupt firms. An
1-12
-------
extensive search (described in Section III) for firms which had gone
bankrupt and which had publicly available financial statements yielded a
total of only 95 firms, of which many were retail firms and none were
electric utilities. To further subdivide this sample into, for example,
two-digit SIC codes would result in samples of bankrupt firms too small
to be significant. Thus, while industry-specific tests might be theore-
tically desirable, they cannot be developed from the data currently
available. A review of the bankruptcy forecasting literature found only
one study which had developed industry-specific tests (Altman, 1973).
This study was of the extremely bankruptcy-prone railroad industry, and
still had to use bankruptcies over a period of 50 years in order to
develop an adequate sample size.
3. Use of Population Sampling
As outlined in Section I.e., the performance of a financial test
was developed in two steps: first, the development of a baseline
failure rate, and second, the determination of how this baseline failure
rate was reduced by the performance of a specific test. The performance
of each test was measured against the samples of bankrupt and non-
bankrupt firms. An alternative procedure would be to gather relevent
data for a large population of firms, determine which ones later entered
bankruptcy and which ones did not, and then examine each test against
this entire sample for its ability to forecast bankruptcy. The problem
with this approach is that bankruptcy is a relatively rare event. The
average business failure rate in the post-war period has been approxi-
mately 44 per 10,000 firms per year. This implies that in order to
1-13
-------
obtain a sample of bankrupt firms as large as that used in this study,
it would have been necessary to obtain data on a total of 15,000 firms.
Such an effort would have been beyond the scope of this study, and has
not been attempted in any previous study of bankruptcy forecasting. The
structured sample employed in the study should provide, however,
approximately equivalent results.
1-14
-------
II. BASELINE FAILURE RATES AND DETERMINATION OF MULTIPLES
A. Baseline Failure Rates
This Section explains how the Agency determined a baseline failure
rate applicable to large firms from the existing data on failure rates
by size of firm. For the purposes of this study, "large firms" were
considered to be firms with over $10 million in net worth.
Table II-l is a compilation by Dun & Bradstreet of historical
failure rates per 10,000 firms, incorporating all sizes of firms. As
shown in the Table, the arithmetic mean for the period 1950-1978 is 44
business failures per year per 10,000 firms. This is probably too high
a baseline failure rate for the kinds of large firms considered in this
study. Greater assets and net worth reduce the probability of corporate
failure. The failure rates shown in the Table for all firms thus may
overstate the rate of failure for large firms. Altman suggests a reason
for this phenomenon:
[Mjany of these companies are not permitted to fail. Except in
the event of fraud, or where the failing company is simply too
large, we rarely observe in the decades before 1970 firms of
over $25 million in assets actually going bankrupt. In many
cases, the financially troubled firms can expect to be wooed by
a highly liquid or managerially rich firm, usually resulting in
a merger absorption before insolvency in a bankruptcy sense
occurs. (Altman, 1971)
During the period from 1969 to 1974, for example, Douglas Aircraft,
Ling-Temco-Vought, Lockheed Aircraft, and Mohawk Data Sciences came very
close to failing, but failure was prevented by government intervention
or private sector efforts (Altman et al., 1977).
Precise quantitative data on failure rates by size of firm are not
available. Dun & Bradstreet, which maintains the only extensive failure
II-l
-------
TABLE II-l
ANNUAL BUSINESS FAILURE PER 10,000 FIRMS
Years
1970-78
1960-69
1950-59
1940-49
1930-39
1950-1973
1930-1978
Arithmetic
Mean
36
52
42
26
86
44
49
Maximum
Year
44
64
56
63
154
64
154
Number
of Years
Over 60
0
2
0
1
8
2
11
SOURCE: Derivation from Dun & Bradstreet (1979).
II-2
-------
rate data now published, does not regularly provide data on failure
rates by size of assets, sales, or net worth. Although a good deal of
indirect evidence suggests that larger firms have substantially lower
failure rates, a combination or comparison of the different data bases
is difficult. The estimate in this study of the baseline failure rate
for large firms was derived from the following sources:
(1) Dun & Bradstreet publications on the relationship between
credit ratings and failure rates;
(2) Quantitative data from Altman (1971) combined with IRS
statistics on the number of corporations above a given
size; and
(3) A study by Richard C. Edwards on long-term failure rates
for very large firms.
1. Failure Rates by Credit Ratings and Firm Size
Dun & Bradstreet credit ratings are composed of two elements —
size and creditworthiness. Size is defined as the companies' tangible
net worth. Creditworthiness is rated from 1 (high) to 4 (limited).
Table 11-2 provides the key to the ratings classifications.
A study of the relationship between business failures and these Dun
& Bradstreet credit ratings has been published. The results of the
study are presented in Table I1-3. Because the size categories are
approximately equal to the tangible net worth of the firm, two conclu-
sions can be drawn from the results in this Table. First, firms in the
top two categories for financial strength (representing firms of above
$50 million in tangible net worth and firms of between $10 and $50
II-3
-------
TABLE II-2
KEY TO RATINGS
- •
5A
4A
3A
2A
1A
BA
BB
CB
CC
DC
DD
EE
FF
GG
HH
-Estimated Financial
Over
$10,000,000 to
1,000,000 to
750,000 to
500,000 to
300,000 to
200,000 to
125,000 to
75,000 to
50,000 to
35,000 to
20,000 to
10,000 to
5,000 to
Up to
Strength
$50,000,000
50,000,000
10,000,000
1,000,000
750,000
500,000
300,000
200,000
125,000
75,000
50,000
35,000
20,000
10,000
5,000
Composite
High
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
Good
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
Credit
Appraisal
Fair Limited
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
SOURCE: Dun & Bradstreet.
II-4
-------
TABLE II-3
RELATIONSHIP OF D&B CREDIT RATINGS WITH BUSINESS FAILURES
Rating
Category
5A1/
4Ar,
3A-
2A
1A
BA
BB
CB
CC
DC
DD
EE
FF
GG
HH
Percent of
Failures
Rated High
1971
0.00%
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.01
0.01
0.01
0.00
0.05
0.00
1972
0.00%
0.00
0.01
0.00
0.00
0.00
0.00
0.01
0.01
0.00
0.02
0.00
0.02
0.06
0.00
Percent of
Failures
Rated Good
1971
0.00%
0.00
0.00
0.09
0.00
0.05
0.02
0.04
0.05
0.05
0.06
0.08
0.08
0.08
0.07
1972
0.00%
0.00
0.02
0.04
0.00
0.03
0.02
0.05
0.04
0.05
0.06
0.08
0.09
0.08
0.06
Percent of
Failures
Rated Fair
1971
0.00%
0.00
0.33
0.37
0.36
0.30
0.35
0.31
0.39
0.37
0.41
0.46
0.46
0.43
0.21
1972
0.00%
0.00
0.50
0.23
0.25
0.15
0.18
0.28
0.36
0.43
0.52
0.52
0.49
3.63
3.50
Percent of
Failures
Rated Limited
1971
0.00%
0.00
0.00
0.00
4.00
0.31
1.50
1.17
1.65
1.46
2.05
1.89
1.70
1.33
0.50
1972
0.00%
0.00
0.00
0.00
0.00
0.00
0.00
1.02
1.48
2.46
2.04
0.00
1.70
1.40
3.49
SUMMARY
Percent Failures to Listings Rated
1971
1972
High
Good
Fair
Limited
Numeral
Blank
Total Failures
per 10,000 listings
0.004%
0.07
0.40
1.42
0.06
0.75
0.007%
0.06
0.40
1.51
0.07
0.64
42
38
U,
-Category represents firms with over $50 million in tangible
net worth.
2/
-Represents firms with between $10 and $50 million in tangible
net worth.
-Represents firms with between $1 and $10 million in tangible
net worth.
SOURCE: Sanzo (1974).
II-5
-------
million in net worth) have substantially fewer failures for every credit
appraisal rating than do firms in lower categories, including firms with
from $1 to $10 million in net worth. The Table shows that the percent
of failures is .00 percent for the top two credit ratings. However, the
failure rate for .firms in the financial strength category of 3A or
lower, i.e., firms with less than $10 million in tangible net worth,
increases, especially for those firms with a "fair" credit rating. The
"limited" rating is given to firms for which insufficient data are
available to provide a more specific rating. As the Table shows, the
majority of firms with a "limited" rating are smaller firms in the lower
financial strength categories.
One problem in using these data is that when very large firms, such
as W. T. Grant, are believed to be approaching failure, they are not
given a "limited" rating but rather are given a special rating indicated
by a double hyphen. This means that if Dun & Bradstreet thinks that
failure is a possibility for a very large firm, it is not rated at
all. Nevertheless, this tends to be a relatively small category (Backer
and Gosman, 1978).
Table II-3 also shows that Dun & Bradstreet haa greater difficulty
accurately rating small firms than large firms. For every credit
rating, there are significantly more failures in the small-firm cate-
gories than in the respective large-firm categories.
2. Quantitative Data from Altman Analysis and IRS Statistics
In 1970, 12 firms of over $10 million in assets filed for bank-
ruptcy (Altman 1971). At that time, there were 23,300 corporations of
II-6
-------
over $10 million in assets (U.S. Internal Revenue Service, Statistics of
Income, Corporate Income Tax Returns). Thus, the implied bankruptcy
rate in 1970 for firms of over $10 million in assets is approximately 5
per 10,000. Given that the failure rate for all firms is over 44 per
10,000, this implies that the failure rate for firms of over $10 million
in assets is only about 11 percent of that reported for all firms.
3. Long-Term Failure Rates for Very Large Firms; Data from Edwards
Analysis
In a study conducted in 1975, Richard C. Edwards traced the corpo-
rate development of 225 very large firms from the years 1919 to 1969.
Very large firms in the Edwards study were defined as firms with over
$50 million (in" 1967 dollars) in assets in the year 1919. This is
approximately equivalent to $100 million in assets in 1980 dollars. The
purpose of Edwards' study was to determine the extent to which large
size alone might effectively prevent the possibility of business
failure. Table II-4 shows the results of this study. For comparison,
the average failure rate for all firms for this period, as derived from
Dun & Bradstreet data, is 61 firms per 10,000. As is shown in the
Table, the failure rate for large firms as a percentage of the failure
rate for all firms varies from 18 to 32 percent, depending on which
class of firms is examined. This wide variance is largely due to the
inclusion in the sample of 15 urban transit firms, virtually all of
which failed during the 1920's and 1930's.
4. Summary of Failure Rate Data Bases and Conclusions
Table 11-5 summarizes the results discussed above. These results
strongly suggest significantly lower failure rates for larger firms, but
II-7
-------
TABLE II-4
FAILURE RATES OF VERY LARGE FIRMS
Class of Firm
Manufacturing
All Firms Except
Urban Transit
All Large Firms
Number
°f I/
Firms^
110
210
225
Number
°f I/
Failures^
8
11
22
Implied
Annual
Rate Per
10,000
Firms^/
15
11
19.5
Implied
Failure
Rate as a
Percentage
of Dun &
Bradstreet
Failure Rate
For all Firms
(61/1Q,000>3/
24
18
32
— Number of firms and failures are from Edwards (1975).
— Implied failure rate calculated as:
number of failures
1/50
1 -
number of firms
3/
— Derivation from Dun & Bradstreet data.
SOURCE: Derivation from Edwards, 1975.
II-8
-------
TABLE II-5
SUMMARY OF EVIDENCE ON FIRM FAILURE RATES
BY SIZE OF NET WORTH
Size Category
Failure Rate
as a Percentage
of Failure Rate
for All Firms
+$10 Million in Net Worth with a D&B Fair Credit
Rating (1971-1972) '. •
+$25 Million in Net Worth (1970)
+$100 Million in Assets (1919-1969)
11
18-32
-------
a precise estimate is impossible to determine given the quality of the
data bases. For example, the estimate of a 1 percent failure rate for
firms with over $10 million in net worth as a percentage of the failure
rate for all firms based on firms with a Dun & Bradstreet "fair" credit
rating is probably an underestimate. Since Dun & Bradstreet provides no
data on the distribution of firms by size, it is impossible to determine
from Table II-3 if the entries of .00 percent, which Dun and Bradstreet
states may be any percentage less than or equal to .004, represent a
significant number of firms. Further uncertainty arises from the fact
that large firms that are expected to fail are not rated at all.
Although it is impossible to derive a precise failure rate for
large firms from the data available, all of the data suggest that the
failure rate is at least 50 percent lower than the failure rate for all
firms. The Agency concluded that an estimated baseline failure rate of
22 per 10,000 for firms with over $10 million in net worth was reason-
able, assuming that the failure rate for large firms would be at least
50 percent lower than the estimated failure rate of 44 per 10,000 for
all firms. Considering the quality of the data, however, even smaller
failure rates than 22 per 10,000 are possible.
B. Ability to Pay
The fact that a firm has not actually entered bankruptcy does not
ensure that it has adequate funds available to meet financial responsi-
bility obligations. This Section therefore describes how the Agency
determined if adequate funds would be available to those financially
weakest firms which pass a financial test to meet their financial
11-10
-------
responsibility obligations. If a non-bankrupt firm does not have
adequate funds, then such obligations may themselves be a cause of
bankruptcy. This Section also describes how the Agency determined what
additional requirements are necessary to ensure that such obligations
can be met without increasing the failure rate. These determinations
were made primarily through the examination of the characteristics of
several firms that would pass many pos.sible financial tests and enter
bankruptcy within three years of passing the test. Such firms were the
financially weakest of all firms that could pass financial tests.
1. Availability of Funds
To analyze the adequacy of the funds available to financially weak
firms immediately prior to bankruptcy, the Agency examined a sample of
12 bankrupt firms. These firms passed many ratio tests (see Sections
III and IV) three years prior to bankruptcy, and tests which were
stringent enough to eliminate them also eliminated a large percentage of
viable firms. The fact that several of the 12 firms also passed a
number of tests two years prior to bankruptcy indicates that their
financial status deteriorated very rapidly. They therefore represent
the type of firms most likely to fail a financial test only at the point
where they lack sufficient funds to establish an alternate form of
financial assurance. The Agency examined financial data for these firms
to determine whether they would have been able to meet financial respon-
sibility obligations at the time they were eliminated by a financial
test.
11-11
-------
Table II-6 shows, for the 12 firms, the changes in net working
capital, net worth and quick assets occurring over the three years
preceding bankruptcy. As can be seen from this Table, even if these
firms did not fail a financial test until two years prior to bankruptcy,
all of the firms still would have had positive net worth and some quick
assets. In fact, the quick assets of most of these firms increased
between the third and second year prior to bankruptcy. Therefore, if
these firms had failed a financial test two years prior to bankruptcy,
they would have had significant amounts of cash and other liquid assets.
Current assets would have exceeded current liabilities for two-thirds of
the firms, and total assets would have exceeded total liabilities for 11
of the 12 firms.
If the establishment of an alternative means of financial assurance
for these firms were delayed by, for example, legal proceedings, Table
II-6 shows that 66 percent of the firms (8 out of 12) still would have
had positive net working capital one year prior to bankruptcy. All
except one firm would have had positive net worth, and all would have
had significant quick assets. Therefore, with a delay up to one year
prior to bankruptcy, 66 percent of the firms would have maintained
adequate cash and net working capital to pay at least some of their
obligations. Furthermore, negative net working capital alone does not
necessarily imply that a firm cannot pay any of its debts. A conser-
vative conclusion would be that if legal proceedings normally took one
year or less, at least 50 percent of the firms that pass the test up to
two years prior to bankruptcy would still have sufficient funds to
establish an alternative form of financial assurance.
11-12
-------
TABLE I1-6
CHANGES IN KEY FINANCIAL PARAMETERS FOR 12 BANKRUPT FIRMS LIKELY
TO BE MISCLASSIFIED BY A FINANCIAL TEST
Company
Name
Walt ham
Bowmar
Polarad
Beck
Unishops
Uniservices
Remco
Maule Industries
M.H. Fishman
National Video
Dolly Madison
King Resources
Percent of
Decline From
3 Yrs To 2 Yrs
Prior to Bankruptcy
NWK
+ ±'
+
60.9
+
+
+
39.7
-
4.9
NA
+
89.5
NW
+
+
41.9
+
11.5
+
9.0
+
5.7
+
+
+
QA
+
+
+
+
+
+
+
+
13.6
51.6
+
NA
Percent of
Decline From
2 Yrs to 1 Yr
Prior to Bankruptcy
NWK
_ I/
+
+
+
-
66.7
75.6
-
2.6
21.1
83.8
-
NW
92.8
76.4
16.7
+
-
48.5
84.0
11.6
4.6
21.2
5.8
65.7
QA
35.5
12.9
26.1
+
+
27.3
12.3
50.5
+
26.9
+
NA
Percent of
Decline From
3 Yrs to 1 Yr
Prior to Bankruptcy
NWK
—
+
34.8
+
-
40.0
85.3
-
7.4
21.1
75.7
-
NW
90.2
50.7
51.6
+
-
32.0
85.4
-»
10.0
18.9
+
38.6
QA
7.3
+
15.0
+
-
20.0
6.5
38.1
+
64.6
+
NA
I
M
Lo
~ A plus sign indicates that there was an increase in value rather than a decline.
21
— A negative sign indicates that there was .a negative value so calculating the
percentage decline is meaningless.
NA: Not available
-------
To test this hypothesis, the Agency examined the net working
capital and net worth of the entire sample of 66 bankrupt firmsi Table
II-7 shows that over 88 percent of all of these bankrupt firms had
positive net working capital and positive net worth three years prior to
bankruptcy. In fact,' even one year prior to bankruptcy, 89 percent
still had positive net worth and 72 percent still had positive net
working capital.
2. Multiple Requirements
Adequate funds will only be available if the net working capital,
net worth and quick assets of a firm are large enough to cover its
required financial responsibility obligation, without the expenditures
themselves causing bankruptcy. Financial test criteria which require
net working capital and net worth to be equal to some multiple of the
costs of closure and post-closure would help ensure the availability of
adequate funds. Table II-6, which gives the rates of deterioration of
net working capital, net worth and quick assets for the 12 bankrupt
firms discussed above, shows that the quick assets for most of these
firms deteriorated relatively slowly. This suggests that a financial
test requiring minimum levels of quick assets will not necessarily
ensure that adequate funds are available when needed. The percentage
declines in net working capital and net worth are significant, and
suggest that a test should require that net working capital and net
worth be multiples of closure and post-closure costs.
Under normal circumstances, a firm with net working capital equal
to twice the estimated closure and post-closure costs would be able to
11-14
-------
TABLE II-7
GENERAL ABILITY TO PAY CHARACTERISTICS
OF FIRMS APPROACHING BANKRUPTCY
Percentage of All Bankrupt
Firms with Positive Net
Working Capital
Percentage of All
Bankrupt Firms With
Positive Net Worth
Years Prior to Bankruptcy
x-3
88
97
x-2
79
97
x-1
72
89
11-15
-------
pay these costs without threatening its continued viability. Inspection
of the financial data of bankrupt firms, however, indicates that net
working capital may often decline precipitously as bankruptcy nears.
Furthermore, firms approaching bankruptcy may frequently experience
negative cash flow, which will place additional burdens on available net
working capital. Nine of the 12 firms sampled, for example, had nega-
tive cash flows in the year prior to bankruptcy. Therefore, a financial
test should require that net working capital be a multiple of closure
and post-closure costs, in order to provide the Agency with a reasonable
probability that a firm will have resources available for the establish-
ment of an alternative form of financial assurance.
The Agency determined the appropriate multiple, by assuming that if
a firm met this multiple requirement three years prior to bankruptcy,
then it would still have sufficient funds for its financial responsi-
bility obligations one year prior to bankruptcy. This requirement can
be derived in the following way:
(1) V3 >_ M x FR
(2) VL 2. FR,
where V-, is the value of net worth or net working capital three years
prior to bankruptcy, V^ is the value of net worth or net working capital
one year prior to bankruptcy, M is the value of the multiple, and FR is
the value of the financial responsibility obligation (closure costs,
post-closure costs, and/or liability requirements). Solving for M
yields:
V.
(3) -:p > M
vl ""
11-16
-------
This multiple has been calculated for each of the 12 firms in the
sample discussed above, with the results shown in Table II-8. The
average value of the multiple for the firms that experienced a decline
in net working capital but still had positive net working capital is
2.8. That is, if net working capital had been equal to 2.8 times the
cost of financial responsibility obligations three years prior to bank-
ruptcy, then adequate funds would still have been available one year
prior to bankruptcy (i.e., after the firm's net working capital had
deteriorated). The average multiple for net worth for firms which had
positive net worth remaining one year prior to bankruptcy was 3.3.
These multiples only ensure that these firms would have had net
working capital and net worth exactly equal to their financial respon-
sibility obligations. As noted above, this provides no margin of safety
after the financial responsibility obligations have been met. If both
of the above multiples are rounded to 3, and then multiplied by 2 to
provide a reasonable margin of safety and to ensure that the firm has
some funds remaining to cover other liabilities which may not be
recorded on the balance sheet, then multiples of 6 result. These
multiples provide a margin of safety to ensure that the failure rates
are not greater than those estimated in the previous Section (22 per
10,000), and to ensure that adequate funds are available even in the
event of lengthy legal proceedings to obtain the funds.
11-17
-------
TABLE I1-8
MULTIPLES OF FINANCIAL PARAMETERS NECESSARY TO INSURE
AGAINST FIRM DETERIORATION
Company Name
Waltham
Bowraar
Polarad
Beck
Unishops
Uniservices
Remco
Maule Industries
M. H. Fishman
National Video
Dolly Madison
King Resources
Average Multiple
No. of firms that
decline in value
% of firms that
decline in value
Net Working Capital
JY
£
1.5
+
-
1.7
6.8
-
1.1
'1.3
4.1
-
2.8
10/12
83%
Net Worth
10.2
2.0
2.1
+
-
1.5
6.3
+
1.1
1.2
+
1.6
3.3
9/12
75%
Quick Assets
1.1
+
1.2
+
+
1.3
1.1
1.6
+
2.8
rf
NA
1.5
6/12
50%
— Negative sign indicates negative value
2/
— A plus sign indicates an increase in value from three
years to one year prior to bankruptcy
NA: Not available
11-18
-------
III. PROCEDURES USED IN SELECTING FIRM SAMPLES AND CANDIDATE
FINANCIAL RATIOS
A. Sample Selection
The Agency assembled an initial sample of 95 firms that filed under
either Chapter 10 or Chapter 11 of the Federal Bankruptcy Act between
1966 and 1979. These firms were identified in two ways. Previous
articles on bankruptcy forecasting by Altman, Haldeman and Narayanan
(1977) and Elam (1975) listed the bankrupt entities investigated in each
study. The Agency included these firms in its sample if financial data
for the three-year period preceding bankruptcy could be located. Addi-
tional bankrupt firms were identified through a two-step process. The
Agency compiled a list of all companies that had ceased to be listed in
Moody's Industrial Manual between 1968 and 1978 (excluding all firms
specifically identified as having undergone name changes, mergers,
etc.). The fate of these firms was then determined by consulting the
company section of Funk and Scott's Annual Periodical Index. Actual
bankruptcy filing dates for firms in this sample were identified either
from Funk and Scott or the Bankruptcy heading of the annual Wall Street
Journal Index.
The Agency then compiled financial data for each sampled firm for
the three years immediately prior to bankruptcy. These data were
collected primarily from Moody's Industrial Manual and supplemented when
necessary by the company Form 10-K reports submitted annually to the
Securities and Exchange Commission. As a result of serious data gaps,
three firms were dropped from the original list of 95. Still later, the
Agency decided to exclude all firms primarily engaged in wholesale,
III-l
-------
retail and/or transportation service activities from the sample, since
these are not likely to be TSDFs, further reducing the final sample size
to 66 firms.
The average net worth of the 66 bankrupt firms sampled (in the
third year prior to bankruptcy) was $14.5 million; only one non-retail
firm had a net worth greater than $100 million. The average asset size
of the bankrupt firm sample was $46.5 million. Thus, as shown in Table
III-l, the Agency obtained a sample with a higher average asset size
than the samples in almost all previous studies (Beaver, 1963; Altaian,
1968; Altman, et al. 1977; Deakin, 1972).
The Agency then assembled a sample of 190 non-bankrupt firms iden-
tified by the industry index of Moody's Industrial Manual as members of
industrial categories that generate .and dispose of large quantities of
hazardous waste on-site (e.g., primary metals, petroleum refining,
chemical and plastics manufacturing, textiles). Data were collected
from Moody's. for these firms for the three-year period 1973-1975. The
recession year of 1975 was deliberately chosen to evaluate the effects
of business cycle fluctuations on financial ratio performance. Twelve
of these firms were later eliminated because they were primarily engaged
in wholesale, retail, or transportation service activities.
Of the 178 firms in the final non-bankrupt sample, slightly more
than half had net worth (in 1973) of less than $100 million. For the
smaller firms, the averages for net worth and asset size were $33.6
million and $69.8 million, respectively; for the larger firms, $519.5
and $985.7 million, respectively.
III-2
-------
TABLE III-l
AVERAGE ASSET SIZE OF BANKRUPT FIRM SAMPLES
(In Millions of Dollars)
Sample
EPA
Altaian,
Deakin,
Altman,
Beaver,
et al. 1977
1972
1968
1966
Number of
Firms
66
53
79
31
79
Average
Asset Size
46.5
I/
96.0-
20.0
6.4
6.3
— Five non-bankrupt firms (Douglas Aircraft, Ling-Tempco-Vought,
Lockheed Aircraft, Memorex Corp., and Mohawk Data Sciences) with
very large assets were included in the bankrupt firm sample because
the firms remained non-bankrupt only due to extraordinary external
support.
III-3
-------
Several bankruptcy forecasting studies have utilized matched bank-
rupt and non-bankrupt firm samples. Each firm in the bankrupt sample
had a counterpart in the non-bankrupt sample that had a relatively equal
asset size and belonged to the same industry category. After careful
deliberation, the Agency decided that use of such a matched sample would
not satisfy the needs of this evaluation. Although the sample of bank-
rupt firms assembled for this study is one of the most comprehensive
ever developed, it contains very few firms with greater than $50 million
in net worth, because very few firms of this size fail. Several of the
most highly publicized business failures of the last decade (W.T. Grant,
Penn Central, United Merchants & Manufacturers) were excluded from the
sample because those firms were primarily involved in activities other
than manufacturing, and this reduced still further the number of large
firms. As a consequence, matched sample techniques would not have
provided information about the effectiveness of the various ratio tests
in accurately classifying the large, viable firms most likely, to make
use of a financial test provision. The Agency did add a representative
sample of large firms to the non-bankrupt group, in order to verify
whether test options were effective over all size ranges.
The Agency also chose not to undertake a one-for-one matching of
sampled firms by industrial category. In order to assemble a sample of
bankrupt firms of sufficient size to provide statistically significant
results, it was necessary to include a substantial number of firms
involved in manufacturing activities not commonly associated with haz-
ardous waste generation (such as bakeries). The Agency determined that
III-4
-------
Che improvements in classification accuracy that would result from
evaluating test alternatives against a set of firms likely to seek to
use a financial test greatly outweighed any potential inaccuracies
engendered by the use of an unmatched sample. To verify this conclu-
sion, the Agency performed a supplementary comparison of industry
financial ratio averages, as described in Section VI of this Appendix.
Many commenters on the May 19, 1980 regulation suggested that
electric utilities and hazardous waste management firms possessed unique
financial characteristics and should be provided with a separate finan-
cial test. To test these premises, special samples of 26 utilities and
two representative large hazardous waste management firms were examined.
These results are described in Section VI.
The fundamental limitation of this analysis with respect to the
samples of firms, particularly the non-bankrupt firm sample, is uncer-
tainty with respect to how well they match the financial characteristics
of firms that own or operate TSDFs. In the absence of information about
the population which the samples are to represent, there is no real way
to check for biases created by the methods used to assemble the samples.
The firms in the non-bankrupt sample were not randomly selected
from firms listed in Moody's. Judgment was applied to ensure that firms
were chosen from the specific industries considered most likely to own
or operate TSDFs and to ensure a greater representation of smaller firms
than a random sample from Moody's would have produced. Because the
sample was not randomly drawn, the possibility exists that the judgment
III-5
-------
employed may have inadvertently resulted in the selection of firms
exhibiting unusually strong or weak financial characteristics. To check
this possibility, the Agency compared the mean values of the non-
bankrupt firm sample to those of other samples used in past studies of
this kind. (This full comparison is given in Section V.D.2.) Where
data on comparable ratios were available, the average values of ratios
for the non-bankrupt sample in this study was either very similar to
that for other studies (within 10 to 20 percent), or fell between the
values given by other studies (average values for cash flow to total
liabilities vary widely among samples in other studies). As a further
.check for possible bias, the distribution of specific ratios for the
non-bankrupt firm sample was compared to the population of all firms in
specific industries reportd in the Expanded Coverage section of the 1980
Moody's Industrial Manual (Volume I) (see Section VI.A of this Appen-
dix) . This comparison showed that most industries had from 5 to 10
percent more firms meeting a given ratio cutoff point than are found in
the non-bankrupt firm sample for 1975. This would indicate that either
due to changes in industry financial performance between 1975 and 1980,
or due to differences between the two samples, the percentage of viable
firms passing a test could be higher than that reported in this study.
B. Holdout Procedures
In an analysis in which numerous combinations of financial ratios
are tested on a randomly drawn set of firms, it is important to avoid
the possibility of search bias. If many combinations of financial data
are tested, the test which is evaluated as the most effective may attain
III-6
-------
this result because it is peculiarly sensitive to the precise charac-
teristics of the sample being analyzed. To ensure that such search
biases did not influence the final-conclusions of this analysis, the
Agency created a "holdout" sample comprised of firms randomly selected
by a computer from the original bankrupt and non-bankrupt samples
(Frank, Massy, and Morrison, 1965). The testing procedure was then
conducted in two phases. All tests were evaluated initially against the
136 non-bankrupt and 42 bankrupt firms which remained in the "primary"
sample. Those tests which provided the best results were then validated
against the 42 non-bankrupt and 24 bankrupt firms in the holdout sample.
C. Selection of Financial Variables to be Tested :
In selecting the financial ratios to be tested in this analysis,
the Agency consulted a broad spectrum of opinion, placing particular
weight on available empirical results. Candidate financial ratios were
drawn from three principal sources: the comments received on the May
19, 1980 proposed regulation, the surveyed opinions of bond rating
service personnel and credit analysts, and the existing literature on
bankruptcy forecasting. The recommendations from each of these sources
are summarized below.
1. Recommendations of Commenters
Several commenters proposed other financial ratios as better indi-
cators of firm viability than the total liabilities/net worth ratio.
Alternative ratios most frequently mentioned involved cash flow, net
income, net worth, total liabilities, debt and equity. Other measures
identified were the quick ratio, estimates of return on investment,
III-7
-------
total fixed assets, retained earnings, and net sales. In addition, a
number of comnenters proposed that qualitative measures of a firm's
financial position, such as corporate bond ratings or an affirmative
auditor's statement in the company's annual Form 10-K report, be used to
supplement or to substitute for a ratio test. None of these commenters
provided quantitative data on the expected classification accuracy of
their preferred alternatives, and only a few suggested what an appro-
priate pass-fail cutoff point would be for the ratios proposed. Conse-
quently, these comments were used to identify candidate ratios, but did
not play a significant role in the critical assessment process.
2. Bond Rating and Credit Analysts
Detailed interviews with senior financial analysts from the bond
rating agencies, leading investment banking firms, and a number of major
urban banks conducted for a study (Backer and Gosman, 1978) for the
National Association of Accountants identified the financial ratios
those analysts considered most important to their credit decisions.
Those ratios are listed in Table III-2. The same authors also conducted
an empirical study of the rate of deterioration in financial variables
for 57 firms whose bond, trade credit, or bank loan ratings had been
downgraded. This analysis indicated that the ratios of cash flow /total
debt, total debt/total net worth, and return on sales showed the most
significant deterioration in the two years prior to downgrading, and
hence had the highest correlation with actual credit actions. The
flow in this Appendix refers to the sum of net income, deprecia-
tion, depletion, and amortization.
III-8
-------
TABLE III-2
FINANCIAL RATIOS CONSIDERED MOST SIGNIFICANT BY
CREDIT ANALYSTS
(based on interview results)
Short-Term Intermediate Term -Long-Term
Bank Loans Trade Bank Loans
Financial Ratios Seasonal Credit Terra Bonds
Accounts Receivable
Turnover X
Cash Flow to Total
Debt* X X
Current Liabilities to
Tangible Net Worth X
Current Ratio X X
Fixed-Charge Coverage X
Inventory to Working
Capital X
Inventory Turnover X
Long-Term Debt to
Working Capital X
Quick Ratio X X
Return on Sales X
Total Debt to Tangible
Net Worth* X X X X
Working Capital to Sales X
Source: Backer & Gosman (1978).
*We have substituted the word "debt" for "liabilities"
because the great majority of interviewees indicated
that they exclude Deferred Taxes and Minority
Interest from liabilities, thereby approximating debt.
III-9
-------
analysis also revealed that many other ratios not identified in the
interview process showed much greater deterioration for the sampled
firms than the ratios listed in Table III-2.
3. Results of Prior Bankruptcy Forecasting Studies
The Agency carefully reviewed the existing literature on bankruptcy
forecasting and identified a large array of financial ratios that have
previously been analyzed for their predictive abilities. These ratios
can generally be organized into four basic categories of financial
information, as described below.
a. Profitability (Income or Cash Flow Based) Ratios
Profitability measures reflect the current response of market
forces to the products manufactured and/or sold by the firm, and may
indicate the ongoing profit-making potential of an operation. Since a
firm's ultimate existence relies on the earning power of its assets,
measures of this type should reflect the future stability of a firm.
The most straightforward measure of profitability is net after-tax
income (NI). Beaver (1968) tested the predictive power of these ratios
of net income to sales, total assets, net worth, and total debt, and
found the net income/total asset (NI/TA) ratio to have the most discri-
minating power. NI/TA has also been employed in a number of other
studies (Deakin 1972, 1977; McWilliams, 1977). Beaver identified cutoff
points for this ratio ranging from 0 to .04, depending on the number of
years prior to bankruptcy the test was applied. McWilliams, in deriving
qualitative descriptors for financial ratio data, assigned a plus to a
firm with NI/TA greater than .04, and a minus to those with NI/TA below
0.
111-10
-------
Altman (1968, 1977) proposed the use of earnings before interest
and taxes CEBIT) as a substitute for net income. Theoretically, EBIT
has the advantage of excluding discontinuities in income trends
resulting from one-time extraordinary charges. However, EBIT fails to
take into account the penalties to profitability suffered by a highly
leveraged firm, as the result of interest payments. More importantly,
Altman has never published results which compare the actual predictive
abilities of EBIT versus NI. Since EBIT values are not always clearly
identified in income account statements (particularly in the case of
smaller firms), EBIT was not included in the list of variables tested.
Measures of cash flow (CF) have been observed in past studies to
provide forecasting ability superior to net income ratios. The two most
popular formulations of cash flow ratios are cash flow/total liabilities
(in which CF = NI + depreciation), and cash flow/total debt (in which CF
= NI + depreciation - dividends, and total debt = long-term debt + the
current portion of long-term debt). Preliminary analysis indicated that
certain key industries (most significantly, electric utilities) per-
formed much worse on the latter of these two ratios. Consequently, the
CF/TL ratio was used in the preliminary evaluation of tests. Beaver
used cutoff points for cash flow/total debt of .03 to .11; Me Williams
considered the cutoff points to be .15. to .35.
b. Liquidity Ratios
Liquidity ratios define the capability of a firm to respond to
short-term or emergency expenditure needs without borrowing. Since the
Agency may call on firms to close their facilities immediately, the
III-ll
-------
availability of liquid assets to cover these costs is particularly
relevant. Four measures of liquid assets are commonly employed: cur-
rent assets (CA); quick assets (QA), defined as current assets minus
inventories; net working capital (NWK), defined as current assets minus
current liabilities (CL); and cash (including marketable securities).
These measures are commonly contrasted to either current liabilities or
total assets (TA).
Altman et al. in 1977 found the current ratio (CA/CL, or CURRAT) to
be the most accurate predictive measure of firm stability, while in 1968
Altman used the ratio NWK/TA as his liquidity measure. Beaver (1968)
obtained equivalent predictive results using both measures, as well as
the quick ratio (QA/CL, or QRAT). Deakin (1972) used all seven vari-
ables (CURRAT, QRAT, Cash/CL, CA/TA, NWK/TA, Cash/TA) in his 14-variable
multi-discriminant equation. In his refined five-variable model (1977),
only CURRAT, CA/TA, and Cash/TA are used.
c. Leverage Ratios
Leverage ratios illustrate the percentage of a firm's total assets
that are tied to short-term or long-term liabilities. A highly
leveraged company may have substantial liquid or long-term assets on
hand, but they may be secured by other creditors if the firm's financial
condition begins to deteriorate. There are several alternate measures
of leverage. Beaver (1968) looked at a number of ratios that contrasted
current and other liabilities to total assets, and found total debt/
total assets to be the most effective. Altman (1968) found that the
ratio of the market value of common and preferred stock to equity
111-12
-------
provided better predictive results than other debt/equity measures,
although the relative performance of the tested ratios was not docu-
mented. Long-term bond analysts cited the total liability (TL)/net
worth (NW) ratio as the most important variable influencing their judg-
ment (Backer and Gosman, 1978). The Agency used a TL/NW ratio as a
measure of a firm's leverage in its preliminary evaluations. The common
application in the business community and ease of calculation of this
ratio make it an appealing candidate for these analytical purposes.
d. Turnover Ratios
Both Beaver (1968) and Altman (1968) tested a number of ratios
relating various asset categories to net sales figures. Such turnover
ratios are commonly used by analysts to measure the sales-generating
capability of assets, or to indicate management's effectiveness in the
competitive marketplace. Although these ratios have traditionally shown
very poor predictive power when used independently, Altman, in his 1968
Z-score model, included a net sales (NS)/TA ratio, and rated it the
second most important contributor.to the overall discriminating power of
his function. However, research by the Agency revealed that such a
ratio severely discriminates against a number of industry categories
(e.g., electric utilities; firms with large extractive interests, such
as mining, gas, oil and timber).
4. Ratios Selected
Seven financial ratios were selected for detailed evaluation as
candidates for inclusion in a financial test. These ratios were chosen
from the many identified because they satisfied three primary
111-13
-------
conditions: (1) they had been quantitatively validated as significant
predictive variables by previous research efforts; (2) they had been
frequently identified by bond rating services and credit analysts as key
parameters; and (3) their values could be readily obtained from Consol-
idated Income Account or Consolidated Balance Sheet data. The seven
ratios selected were:
Cash flow/total liabilities (CF/TL)
Net income/total assets (NI/TA)
Total liabilities/net worth (TL/NW)
Current assets/current liabilities (CURRAT, or current ratio)
Quick assets/current liabilities (QRAT, or quick ratio)
. Net working capital/total assets (NWK/TA)
Net sales/total assets (NS/TA)
D. Summary Financial Data for the Firm Samples
Financial data from the companies included in the bankrupt and non-
bankrupt firm samples are summarized in Tables III-3 through III-6.
Table III-3 summarizes the primary bankrupt sample; Table III-4, the
primary non-bankrupt sample; Table III-5, the holdout bankrupt sample;
and Table III-6, the holdout non-bankrupt sample. The three years of
data collected for each sample are hereafter referred to as years x-1,
x-2, and x-3. For each sample, values for eight financial parameters
are detailed in the tables: net income (NI), cash flow (CF), quick
assets (QA), current assets (CA), total assets (TA), current liabilities
(CL), total liabilities (TL), and net worth (NW). The right-hand
portions of tne Tables for the primary samples (Tables III-3 and III-4)
111-14
-------
indicate each firm's performance in the six major financial ratio cate-
gories used in this analysis: cash flow/total liabilities, total liabi-
lities/net worth, net fixed assets/total assets, net income/total
assets, current assets/current liabilities, and quick assets/current
liabilities. Because the net income/total assets ratio
was not included in any of the tests evaluated against the holdout
samples, this ratio is not included in Tables III-5 and III-6.
111-15
-------
TABLE I11-3
FINANCIAL DATA FOR FIRMS IN PRIMARY BANKRUPT SAMPLE
Name ut
Plru
Alan Wood Steel
Ail led Leisure
American Beef
Packers
Anttiilciiii Bouk
Armac
Industries
Ast rciduta
Harrington
Industries
Hi shop
Industries
buwmar
Computer
Appl Icat Ions
Dlversa
Vear
I978(x-l)
(x-2)
(x-3)
1977(x-l)
(x-2)
(x-3)
1975(x-l)
(x-2)
(x-3)
1973U-1)
(x-2)
(x-3>
1976(x-l)
(x-2)
(x-3)
)970(x-l)
(x-2)
(x-3)
I9>6(x-l)
(x-2)
(x-3)
1970(x-l)
(x-2)
(x-3)
l»75(x-l)
(x-3)
I97()(x-l)
(x-2)
(x-3)
1969(x-l)
(x-2)
(x-3)
Nl
NA
-15.0
-9.4
-3.0
-0.2
-0.3
4.8
3.4
1.1
0
-2.4
0.1
NA
-6.8
2.4
0.2
-6.7
-5.9
NA
0.3
0.1
-1.4
-7.6
0.2
-25.6
2.0
-2.1
0.6
0.7
6.7
-7.3
-7.4
CK
NA
-1.2
-1.4
-2.8
-0.1
-0.2
7.1
5.3
2.3
1.2
-1.2
1.3
NA
NA
3.0
0.7
-6.4
0
NA
NA
NA
-3.0
-7.3
0.4
-24. 1
2.4
-1.4
1.2
1.6
6.9
-6.4
-6.8
QA
NA
18.7
18.4
0.3
1.2
1.3
47.7
49.6
31.3
b.a
11.0
13.6
NA
13.2
10.7
10.7
11.4
9.5
NA
3.6
3.6
4.7
4.5
12.2
20.9
10.7
14.4
14.2
9.9
0.7
17.1
6.0
CA
NA
11.7
34.1
2.2
2.3
3.2
78.2
82.0
40.4
9.'J
12.9
15.5
NA
29.5
30.5
10.6
11.7
16.7
NA
5.9
5.6
8.7
11.1
17.7
41.9
18.2
17.4
17.2
11.3
0.7
22.2
13.7
U.
NA
41.4
29.2
4.4
2.0
1.4
72.2
74.8
35.4
15.5
6.3
6.9
NA
24.0
17.7
5.8
18.0
17.4
NA
5.6
6.9
4.5
10.2
9.1
20.4
10.5
13.3
17.1
7.6
7.6
12.5
27.7
Tl.
NA
H8.0
79.0
4.4
2.0
1.4
97.4
90.3
45.7
12.2
17.5
16.5
NA
27.3
22.3
11.1
23.3
21.7
NA
7.7
9.6
14.0
16.3
15,2
51.5
13.4
33.2
23.0
14.6
39.3
73.1
80.3
NW
NA
32.0
47.0
-1.3
2.1
2.3
22.6
17.7
14.3
6.6
6.6
12.8
NA
5.9
11.7
2.3
-8.7
-1.0
HA
1.2
2.5
2.3
2.2
7.0
3.5
7.1
1.0
11.2
7.2
6.6
3.0
10.4
TA
NA
120.0
126.0
•J.I
4.0
3.7
120.0
108.0
60.0
18.8
24.1
29.3
NA
33.2
34.0
13.4
14.6
20.7
NA
8.9
12.1
16.3
18.5
22.2
55.0
20.5
34.3
34.2
21.8
46.4
76.1
90.7
NFA/TA-'
NA
P
P
r
f
f
?
23.6
P
P
P
P
NA
V
f
f
f
F
NA
F
F
F
F
F
F
'
F
23.6
25.2
1 f
f
T
CF/TIJ?'
NA
F
F
F
F
F
F
F
F
F
F
F
NA
F
13.5
F
V
f
NA
F
F
K
F
F
F
17.9
F
F
11.0
17.1
F
F
•tUmi-1
•NA
F
1.7
P
P
P
F
F
F
l.B
F
1.3
NA
F
1.9
F
F
F
NA
F
F
F
F
F
F
1.9
F
F
F
F
F
F
CURKAT*'
NA
F
F
F
F
P
F
F
F
F
P
P
NA
F
1.7
1.8
F
F
MA
V
F
L.9
F
1.9
P
1.7
F
F
F
F
F
K
QKAT-'
MA
F
F
F
F
F
f
f
F
F
P
P
NA
F
F
P
F
F
NA
F
F
1.0
F
P
1.0
1.0
1 . 1
F
P
F
F
F
NI/TA5'
NA
F
F
F
F
P
4.0
3.1
F
F
F
F
NA
F
P
F
F
F
NA
3.4
F
F
F
F
F
P
F
F
3.2
P
F
F
-'p > 30.0 (passes all tests); F < 20.0 (falls all tests)
- t > 20.0 (passes all tests); F < IO.O (tails all tests)
- P < 1.0 (passes all tests); F > 2.0 (falls all tests)
4/
- P > 2.0 (pusses all tests); F < 1.5 (falls all tests)
- P > 1.2 (passes all tents); F < l.O (falls all tests)
- P > 4.0 (passes all teats); F • 2.0 (falls ull tests)
-------
TABLE III-3
(continued)
M
M
M
I
Niiou: of
Firm
Uully Madison
litigro
Corporation
Fotot:hromG
Inilust ries
General Alloys
Cray
Manufacturing
On.un
Industries
Harvard
Industries
King Resources
Maule
Industries
Michigan
Bakeries
Mllo
Electronics
National Radio
Year
1970(x-l)
(x-2)
(x-3)
1973(x-l)
(x-2)
(x-3)
1970(x-l)
(x-2)
(x-3)
I973(x-l)
-------
TABLE III-3
(continued)
M
M
i
i—•
do
Firm
National Video
Omega Alpha
Patterson
Purcluuent
Pernmnuer
Company
Photon
Pularad
Electruulcs
Heading
Industries
Corpuratlo n
Kemco Industries
Scottux
Sequuy.ih
Induutl liia
Sheffield
Watch
Stellar
Indust rles
Year
1969(x-l)
(x-2)
(x-3)
I971(x-l)
(x-2)
(x-3)
1974(x-l)
(x-2)
(x-3)
I976(x-l)
(x-2)
(x-3>
1972(x-l)
(x-2)
(x-3)
1970(x-l)
(x-2)
(x-3)
1976U-1)
9x-2)
(x-3)
(x-2)
(x-3)
197I(x-l)
(x-2)
(x-3)
1973(x-l)
(x-2)
(x-3)
1971U-I)
(x-3)
1971(x-l)
(x-2)
(x-3)
1975(x-l)
(x-2)
(x-3)
NI
-5.3
2.2
7.3
41.7
0.6
2.0
-2.2
-0.8
-1.3
-15.7
-17.2
-3.0
1.1
-1.0
1.0
0.4
-1.5
0.2
-2.1
-3.3
l.l
-4.8
-0.1
-10.8
-0.7
2.0
-2.4
1.1
0.8
-0.6
-9.5
-10.4
HA
0.7
0.4
-1.7
-4.0
-1.7
CF
-2.1
5.8
9.3
47.4
2.4
6.7
0
0
-0.5
-12.2
-13.6
-0.4
1.8
-0.6
1.2
0.6
-1.4
0.4
-1.2
-2.4
2.0
-4.6
0
-8.8
1-0
3.5
0
0
1.1
0.3
-8.3
-9,2
NA
0.8
0.7
-1.3
-3.5
-1.3
QA
5.7
7.8
16.1
27.9
54.4
39.1
4.4
5.0
6.1
7.8
14.1
18. 5
16.0
14.0
17.4
1.7
2.3
2.0
1.2
4.9
8.8
2.9
2.3
10.0
11.4
10.7
10.1
6.0
3.8
4.7
1.5
1S.O
NA
9.6
S.6
2.4
2.0
3.5
LA
12.7
13.3
23.6
51). 8
95.9
74.5
10.3
11.2
12.1
19.0
36.1
42.5
34.9
31.3
32.4
4.0
5.4
5.1
9.2
11.1
16.6
11.0
4.3
14.0
17.8
13.6
19.3
13.8
7.7
13.3
12.5
29.1
NA
21.6
15.6
3.8
3.9
7.0
CL
5.2
3.8
14.1
40.5
65.5
41.9
5.4
4.6
5.0
13.9
19.1
14.7
8.7
6.2
5.6
1.5
4.5
2.8
H.2
3.6
5.8
8.0
1.7
11.0
13.7
6.8
12.1
7.5
6.7
8.4
12.9
24.3
NA
12.2
9.3
4.5
4.7
6.6
Tl.
8.2
4.5
14.9
145.9
156.5
111.8
11.2
10.2
U.I
67.2
74.6
62.0
23.0
20. 5
21.9
1.6
4.6
2.9 .
16.5
16.9
20.1
a. 7
2.8
16.3
16.9
9.4
22.7
13.4
9.9
29. 0
27.0
39.7
HA
17. 2
10.7
7.9
10.4
10.8
NV
19.3
24.5
23.8
33.1
31.5
63.2
7.3
9.5
10.6
-11.6
4.1
21.3
17.5
15.1
15.7
1.5
1.8
3.1
1.8
5.9
9.2
3.9
1.7
1.1
8.1
8.9
8.0
10.4
2.9
2.8
3.8
15.5
NA
5.3
5.6
0.1
-0.8
2.7
TA
27.5
29.0
38.7
179.0
188.0
175.0
18.5
19.7
21.7
55.6
78.7
83.3
40.5
35.6
37.6
5.1
6.4
6.0
20.3
22.8
29.1
12.6
6.5
17.6
25.0
18.3
30.7
21.8
12.8
31.8
30.8
55.2
NA
22.5
16.3
8.0
9.6
13.5
NFA/TAi'
P
t
r
P
P
P
P
P
P
P
P
P
F
F
F
F
F
F
P
P
P
F
F
F
24.3
22.5
NA
F
F
P
P
28.6
CF/TI^'
F
F
F
F
F
F '
F
F
F
16.7
F
13.8
F
F
F
F
F
F
F
P
F
F
U.I
F
F
F
NA
F
F
F
F
F
Tt/NU?'
P
P
P
F
F
1.8
1.5
1.1
1.0
F
F
F
1.1
1.4
1.4
F
F
P
F
F
F
F
P
F
F
1.1
F
1.3
F
F
F
F
NA
F
1.9
F
F
F
CIJHRAT*-'
V
e
1.7
F
F
1.8
1.9
P
P
F
1.9
P
P
P
P
1.6
F
1.8
F
P
P
F
P
F
e
?
p
1.8
f
1.6
F
R
NA
l.B
1.7
F
F
F
(JRAT-5'
1.1
P
1.1
F
F
F
F
P
P
F
P
P
P
P
P
F
F
F
F
P
P
F
P
F
F
P
f
F
F
F
F
F
NA
F
F
F
F
F
Ml/TA^
F
P
P
P
F
F
F
F
t
f
F
F
1.2
F
2.7
P
F
1.3
F
F
1.8
F
F
F
F
P
F
P
P
F
F
F
NA
1.1
2.5
F
F
F
-------
TABLE III-3
(concluded)
Name of
Firm
Technical
Measurements
Trana- Beacon
Walt ham
Industries .
Wustgate
California
Year
1968(x-l)
(x-2)
(x-3)
1970U-1)
(x-2)
(x-3)
1971(x-l)
(x-2)
(x-3)
1972U-1)
(x-2)
(x-3)
Nl
1.9
-2.1
-1.8
NA
0.2
0.3
-13.9
0.3
1.6
-2.6
2.8
3.6
CF
2. B
-i.a
-1.5
MA
0.3
0.1
0
1.6
2.4
2.7
7.2
6.8
QA
6.7
2.6
4.9
NA
1.5
1.3
8.9
13.8 .
9.6
22.8
46.3
19.6
CA
10.0
5.0
9.1
NA
1.6
1.4
27.0
34.7
18.6
41.3
63.8
32.1
CL
2.7
4.5
5.6
NA
1.5
1.2
30.8
21.9
8.1
34.0
76.9
32.2
TL
3.1
6.0
7.5
NA
2.6
3.7
43.9
47.9
21.1
143.6
182.0
116.7
NU
16.5
0.5
5.5
NA
0.7
-0.8
1.1
15.3
11.2
30.4
33.0
34.3
TA
19.6
6.5
13.0
NA
3.3
2.9
45.0
63.2
32.3
174.0
215.0
151.0
NFA/TAr-
P
F
f
HA
25.2
P
26.8
28.1
29.2
P
P
P
CF/TL-7
P
F
F
NA
F
F
F
F
11.4
F
F
F
TL/NW^
P
F
1.3
NA
F
F
F
F
1.9
F
F
F
4/
CUKKAT-
P
F
1.6
HA
F
K
F-
1.6
P
F
K
F
QKA^'
P
F
F
NA
1.0
1.1
F
F
1.2
F
F
F
Nl/TA^'
P
F
F
NA
P
P
F
F
P
F
F
2.4
-------
TABLE III-A
FINANCIAL DATA FOR FJRMS IN THE PRIMARY NON BANKRUPT SAMPLE
I
N>
o
-'for all firms, x-l In 1975, x-2 Is 1974. and x-3 Is 1973
-'p • 30.0 (passes all tests); F • 20.0 (falls all tests)
-'P ' 2U.O (passes all I oil:;); F < 10.0 (falls all tests)
- P < 1.0 (p:isse« all testa); f > 2.0 (falls nil test::)
- P > 2.0 (passes all teals); F ' 1.5 (falls all li-.sls)
-'e '' 1.2 (pistes all lusts); F • 1.0 (falls all tests!
-'p •' 4.0 (P..SSUU all tvato); F -' 2.0 ((alia all tests)
N.UQ.- l.f
Firm
Alc-.ii
Al lugliuny 1. ml tun
All mail
A 1 lima x
Ani.'tx
Ainu r on
AiuuLi:k
-
Ampeo-PUiiiburgli
/Vastcd
Ankuu
Artncn SlUiil
ASAKCO
Y..ri'
(x-l)
(x-2)
(x-3)
(x-l)
(x-2)
(x-3)
(x-l)
(x-2)
(x-3)
(x-l)
(x-2)
(x-3)
(x-l)
(x-2)
(x-3)
(x-l)
(x-2)
(x-3
(x-l)
(x-2)
(x-3)
(x-l)
(x-2)
(x-3)
(x-l)
(x-2)
(x-3)
(x-l)
(x-2)
(x-3)
(x-l)
(x-2)
(x-3)
(x-l)
(x-2)
(x-1)
HI
65
173
104
3D
45
31
1
.7
3
21
36
16
134
148
105
8
7
6
13
11
9
6
9
5
35
17
15
.7
.6
1
117
204
107
26
126
113
CK
235
338
263
50
64
49
2
2
3
23
36
16
193
194
142
14
12
11
16
15
13
H
10
6
45
26
24
1
1
2
209
290
190
62
161
140
'I*
400
474
453
119
153
135
8
13
9
117
99
81
451
261
417
41
48
35
43
32
10
14
16
13
98
71
63
4
4
4
438
601
438
195
182
171
CA
1(121
1040
905
325
379
320
16
26
21
217
179
139
677
404
535
70
87
60
85
78
70
3 8
43
35
157
129
117
7
7
7
1001
1124
851
442
408
364
r.i.
37/
472
308
141
184
146
3
12
8
54
1,1)
51
333
207
209
29
39
25
35
36
29
12
17
13
75
61
50
2
3
3
415
554
172
174
234
196
Tl.
1845
1651
1412
318
352
317
4
16
12
It) I
118
71
1116
826
812
67
82
56
54
57
48
15
22
18
79
65
61
2
3
3
1276
1268
1114
641
466
375
MM
1575
1547
1409
320
304
270
17
15
15
279
259
201
1364
942
840
68
62
58
67
59
53
50
45
38
179
153
144
10
9
9
1330
1274
1145
86)
863
774
TA
3420
3198
2821
63tt
656
588
21
31
27
462
377
272
2480
1768
1652
135
144
114
121
117
102
65
67
56
258
218
205
12
12
12
2606
2542
2259
1502
1329
1149
NFA/TA-'
P
P
V
V
r
V
22.3
F
20.9
f
F
f
P
P
P
P
P
P
29.6
P
29.2
25.8
20.7
21.4
P
P
P
P
P
P
P
P
P
P
P
P
CF/TI.-'
12. H
P
18.6
15. B
18.2
15.5
P
10.3
P
12.5
P
P
17.3
P
17.5
P
14.9
18.6
1'
P
P
P
P
P
Y
Y
Y
P
P
P
16.. 3
P
17.1
F
P
P
•ri./Nu^'
1.17
1.06
1.00
P
1.16
1.17
P
1.01
P
P
P
P
P
P
P
P
1.32
P
P
P
P
P
P
t
. t
t
P '
P
P
P
P
P
P
P
P
P
CUKKA r -
t
p
P
f
P
r
p
p
p
p
p
p
p
1.95
p
p
p
p
p
p
p
p
• p
p
1.54
p
p
p
p
p
p
p
p
p
1.74
1.86
ns&t-'
1.06
1.00
H
F
f
F
P
1.05
I. 10
P
P
P
P
P
P
P
P
P
P
f
1.03
P
1.04
1.02
P
1.16
P
V
P
P
1.06
1.08
1.18
1.12
K
F
NI/TA-'
F
P
3.7
P
P
P
P
2.3
P
P
P
P
P
P
P
P
P
P
P
P
P
P
P
P
P
P
P
P
P
P
P
P
P
F
P
P
-------
TABLE I11-4
(continued)
I
ro
N;itnu of
Firm
ASPRO
Alhlone
A.I ft 1 c
Bernz-0-Matlc
Bethlehem Steal
Bliss & Laughlln
Brooks & Perkins
Buntly Corporal ion
Carpenter
Ceco
Cerro
i:FU
Cleveland-Cliffs
Yoari'
(x-l)
(x-2)
(x-3)
(x-l)
(x-2)
(x-3)
(x-l)
(x-2)
(x-3)
(x-l)
(x-2)
(x-3)
(x-l)
(x-2)
(x-3)
(x-l)
(x-2)
(x-3)
(x-l)
(x-2)
(x-3)
(x-l)
(x-2)
(x-3)
(x-l)
(x-2)
(x-3)
(x-l)
(x-2)
(x-3)
(x-I)
(x-2)
(x-3)
(x-l)
(x-2)
(x-3)
(x-l)
(x-2)
(x-3)
NI
2
2
2
12
13
7
3
4
2
-2
.8
2
21,2
342
207
10
9
8
1
.5
.5
4
.3
8
22
17
14
9
11
10
-2
16
15
38
35
14
31
26
21
r.r
4
3
3
15
16
10
5
6
5
-1
1
2
476
553
403
14
14
12
1
.8
.a
8
4
II
23
24
20
14
17
15
11
24
22
64
60
33
34
26
21
QA
13
11
10
43
50
32
10
14
11
4
7
6
768
1172
934
37
24
29
5
4
3
19
19
21
47
49
40
6(1
50
46
208
247
185
85
79
62
58
72
70
CA
23
19
18
85
96
77
25
27
24
12
14
14
1388
1682
1376
64
56
56
11
8
7
29
29
33
113
117
100
98
100
92
311
318
253
127
115
92
75
81
78
Cl.
12
9
a
32
45
30
8
9
a
4
5
6
719
1032
713
14
19
16
4
4
1
10
11
U
34
40
28
26
31
31
147
154
80
71
46
53
)2
30
18
Tl.
23
21
18
109
122
112
22
21
23
IU
11
9
1180
2023
1677
52
50
56
7
5
3
21
23
25
55
65
55
58
63
65
380
396
298
127
107
117
68
56
33
NU
15
13
12
50
48
40
27
25
22
8
9
9
2612
2490
2242
68
62
57
6
5
6
50
49
52
139
124
112
92
87
80
341
369
362
Ib2
181
168
216
203
185
TA
38
34
30
159
170
152
49
46
45
18
20
17
4592
4513
3919
120
112
113
13
10
8
71
72
78
194
189
167
150
150
145
/21
765
660
309
288
285
284
259
218
?/
NFA/TA-'
t
t
t
27.5
24.1
25.7
P
P
P
P
23.5
F
P
P
P
21.9
24.1
23.6
F
F
20.2
P
P
P
P
P
P
P
30.0
P
NA
F
F
P
P
P
F
F
F
CF/T1.2'
16.3
13.7
15.6
13.7
13.3
?
P
P
P
t
11.7
P
P
P
P
P
P
P
18.6
17.4
P
P
16.2
P
P
P
P
P
P
P
F
11.1
K
P
V
P
P
P
P
TI./NW-'
1.53
1.55
1.48
F
F
F
P
P
1.02
1.30
1.19
1.01
P
P
P
P
P
P
1.17
P
P
P
P
P
P
P
P
V
P
P
P
1.07
P
P
P
P
P
P
P
ClIHKAT-
1.95
P
P
P
P
P
P
P
"'
P
P
p
1.93
1.63
1.93
P
P
P
P
1.83
P
P
P
P
P
P
P
p
P
i>
P
I'
P
1.78
P
1.72
P
P
P
•IKA-i-fc'
1.06
P
P
P
1.11
1.07
P
P
P
1.05
P
1.11
1.07
1.14
P
P
P
P
1.15
F
P
r
p
p
p
p
p
p
p
p
p
p
p
1.20
p
1.15
P
P
P
NI/'IA?'
P
P
P
P
P
P
P
P
P
F
3.9
P
P
P
P
P
P
P
P
P
P
P
F
P
P
P
P
P
P
P
P
2.1
2.3
P
P
P
P
P
P
-------
TABLE III.-4
(continued)
I
Ni
Hi) nit* of
Kinu
Coiiunerc J n 1 Mt-ta 1 s
Conayne
Copper Riinge
Co|>|i*-rwuld
1
t
Crown Industries
•
1 Cyclops
j
Driver Harris
.
.
: Earth Resources
I
tlASLO
Ea^turn
tl.'IKA
Florlilu Steel
Cultural Steel
rui.pl'
(x-l)
(x-2)
(»-3)
(x-l)
(x-2)
(x-3)
(x-l)
(x-2)
(x-3)
(x-l)
(x-2)
-------
TABLE ITI-4
(continued)
i
NJ
u>
Naint: of
Firm
lUiimu
liar sco
llofitiaim
Indust r les
Houvur Ba 1 1
& Bt:arlnu
IMCO
Inland Steel
Insp i rat Ion
Consul Idated
Interl jke
Knistr Aluminum
& Chemical
Kaiser Industries
K.ity Industries
Kiiweckl-Burylco
vcar!'
(x-l)
(x-2)
(x-3)
(x-1)
(x-2)
(x-3)
(x-1)
(x-2)
(x-3)
(x-1)
(x-2)
(x 3)
(x-1)
(x02)
(x-3)
(x-l)
(x-2)
(x-3)
(x-1)
(x-2)
(x-3)
(x-l)
(x-2)
(x-3)
(x-l)
(x-2)
(x-3)
(x-1)
(x-2)
(x-3)
(x-2)
(x-3)
(x-1)
(x-2)
(x-3)
(x-1)
(x-2)
(x-3)
NI
47
21
23
36
25
22
.7
3
1
12
11
14
166
57
26
83
148
83
-4
10
15
34
39
17
95
104
45
79
31
45
67
53
10
10
12
-1
7
5
CF
57
30
33
59
48
42
2
3
2
21
17
21
199
79
41
164
222
154
7
20
21
54
55
32
149
157
95
213
91
97
111
96
15
15
17
3
10
a
QA
79
75
60
135
98
91
7
8
8
60
39
33
267
267
183
309
323
250
28
18
36
98
124
87
354
401
312
212
219
219
143
144
35
35
38
22
21
21
CA
100
97
81
224
194
162
18
16
16
87
72
58
469
346
260
627
664
515
60
39
47
216
250
164
825
799
618
305
334
313
244
224
75
75
70
74
72
61
Cl.
49
49
31
59
90
64
12
9
11
37
31
27
284
150
128
243
302
243
13
17
15
107
155
66
446
452
394
189
202
191
134
135
34
34
40
18
14
15
Tl.
92
114
86
152
137
107
16
14
16
69
70
59
630
435
305
896
819
710
54
65
68
216
254
148
1304
855
754
657
692
688
394
422
98
98
97
64
57
40
NW
293
283
274
206
182
164
14
13
11
79
73
62
458
309
250
9/1
939
849
115
91
88
264
242
214
798
1202
1060
632
577
550
395
336
68
68
60
53
55
50
TA
385
397
360
358
319
271
30
27
27
149 1
143
122 |
1088 1
744
555 1
1867
1758
1559
169
156
156
480
496
362
2102
2057
1814
1289
1269
1238
789
758
167
167
156
116
112
90
IIFA/TA-
F
F
F
P
P
P
P
P
P
P
P
f
P
P
'
P
P
P
P
P
P
P
P
P
P
P
P
P
P
P
P
P
20.9
20.9
20.8
P
29.0
26.6
CF/TI.~
P
P
P
P
P
P
12.5
P
11.9
P
P
P
P
18.2
13.4
18.3
P
P
11.3
P
P
P
P
P
11.4
18.4
12.7
P
13.1
14.1
P
P
15.4
15.4
17.8
f
17.3
19.3
TL/NU^'
P
1.16
1.04
1.43
P
P
P
1.38
1.41
1.22
P
P
P
P
P
P
P
1.05
P
1.63
P
P
1.04
1.20
1.25
P
1.26
1.44
1.44
1.64
1.21
1.64
1.03
CURRAT-
P
1.96
P
P
P
P
1.5B
1.89
I. 56
P
P
P
1.65
P
P
P
P
P
P
P
P
P
1.61
r
.85
.77
.57
.61
.65
.64
.83
.66
P
P
1.74
P
P
P
QKA'I-
,.
P
P
P
1.09
P
F
F
F
P
P
P
F
P
P
P
1.07
1.03
P
1.07
P
F
F
P
F
F
F
1.12
1.08
1.15
1.07
1.07
1.02
1.02
F
P
P
P
NI/TA-7
P
P
P
P
P
P
2.3
P
P
P
P
P
P
P
P
P
P '
P
F
P
P
P
P
P
P
P
2.5
P
2.4
3.6
K
F
P
P
-------
TABLE II1-4
(continued)
M
Is)
N^mu ut
Firm
Kumiccot t:
Keybtont?
Consul Idatttil
Kerr-HcCue
Ki-wanee
Kluibi-Tly-Cliirk
Kin-Ark
Koppurs
Kraft co
l.tjliigh Valley
Intlusl rlus
Lilly & Co.
I.TV Corporation
l.ubrlzol
likens SLbul
v I/
Year-
(x-l)
(x-2)
(x-3)
(x-1)
(x-2)
(x-1)
(x-1)
(x-2)
(x-3
(x-1)
(x-2)
(x-3)
(x-1)
(x-2)
(x-3)
(«-l)
(x-2)
(x-3)
(x-1)
(x-2)
(x-3)
(x-1)
(x-2)
(x-3)
(x-1)
(x-2)
(x-3)
(x-1)
(x-2)
(xU3)
(x-1)
(x-2)
(x-3)
(x-l)
(x-2)
(x-1)
(x-1)
(x-2)
(x-3)
N[
49
211
1S9
11
10
5
131
116
63
28
29
17
103
95
82
.9
1
1
60
48
30
140
95
103
2
2
3
101
179
156
13
112
50
47
51
17
11
11
7
CF
71
302
240
20
15
12
190
187
120
43
42
29
150
143
124
2
2
2
91
Bl
SB
192
144
151
4
4
4
212
207
182
70
197
130
59
61
46
12
HA
HA
QA
171
317
331
41
39
17
338
298
200
101
66
49
237
249
228
4
4
3
214
193
1S1
318
337
295
8
a
a
611
553
450
367
645
498
116
112
101
31
39
31
CA
446
606
550
135
124
102
567
496
333
1H1
116
86
493
486
402
7
7
4
369
339
255
1024
1110
854
32
32
32
947
850
667
848
1097
871
193
191
138
60
73
58
1:1.
IB3
287
241
41
55
36
285
294
128
63
42
30
219
199
169
4
6
7
150
144
107
466
582
411
IB
IB
19
',16
390
268
585
647
515
64
80
49
27
16
25
Tl.
812
767
670
128
122
108
579
510
108
174
74
60
575
562
1.58
15
17
16
311
336
246
6 SB
860
583
53
53
55
476
417
290
1602
1686
1552
74
89
55
54
66
55
NM
1411
1442
1107
112
103
95
809
655
559
183
170
147
730
668
610
12
11
1O
368
312
274
1009
850
BOB
6
6
4
958
849
730
361
344
233
231
203
170
109
102
94
TA
2223
2209
1295
240
22S
202
I38B
1164
867
357
244
207
1305
1229
1068
27
27
25
680
648
520
1667
1710
1391
59
59
59
1434
1265
1020
1963
2031
1785
3IIS
292
22S
161
168
149
NFA/TA-
P
P
P
P
P
P
P
P
t
t
P
P
P
P
P
P
P
P
P
P
P
P
P
P
23.7
23.8
23.1
P
29.3
P
P
P
P
P
P
P
P
P
P
CF/Tl.-7
F
P
P
15.9
12.2
10. B
P
P
P
P
P
P
P
P
P
14.6
13.9
14.2
P
P
P
P
16.7
P
F
f
f
t
r
r
r
11.7
t
t
r
r
p
15.0
15.0
L 1
TI./HW-
P
P
P
1.14
1. IB
1.14
P
P
P
P
P
P
P
P
P
1.25
1.54
1.60
P
1.08
P
P
1.01
P
F
F
F
P
P
P
F
F
F
P
P
P
P
P
P
UWKAT-
P
P
P
P
P
P
1.99
1.69
P
P
P
P
P
P
P
1.61
F
F
P
P
P
P
1.94
P
l.Bl
1.81
1.74
P
P
P
F
1.69
1.69
P
P
P
P
P
P
QUA-!*'
F
1. 10
t
1.01
F
1.01
1.19
1.01
P
V
f
I1
I.OB
P
!•
1.08
F
F
P
P
P
F
F
F
F
F
F
P
P
P
F
f
F
P
P
P
P
1.09
P
HI/TA-'
F
P
P
P
P
2.4
P
P
P
P
P
P
P
P
P
3.3
P
P
P
P
P
P
>•
P
3.9
3.9
P
P
P
P
F
P
2.8
P
P
P
P
P
P
-------
TABLE I.I 1-4
(continued)
i
NJ
in
Name of
Kiroi
MarAildrtrws
Forbus
M.mgood
Corporut Ion
Marathon
Manufacturing
Martin
Marietta
Mclntouh
Corporal ion
Mcl.onll!
Me He 11
Corporation
Mead
Merck
Mlirhluan Seamless
Tub.:
MiUlaiid-Koss
Miles
Laboratories
JM
Ycari'
(x-l>
(x-2)
(x-3)
(x-l)
(x-2)
(x-3)
(x-1)
(x-2)
(x-3)
(x-1)
(x-2)
(x-3)
(x-1)
(x-2)
(x-3)
(x-1)
(x-2)
(x-3)
(x-1)
(x-2)
(x-3)
(x-1)
(x-2)
(x-3)
(x-1)
(x-2)
(x-3)
(x-1)
(x-2)
(x-3)
(x-1)
(x-2)
(x-1)
(x-1)
(x-2)
(x-3)
(x-1)
(x-2)
(x-1)
NI
3
2
6
.2
.1
.2
14
7
-19
55
81
57
1
4
4
6
22
16
4
5
6
51
82
50
229
211
183
10
7
4
21
21
12
15
16
17
262
102
296
CF
6
5
a
.5
F
F
22
13
-12
115
139
109
3
5
5
23
41
15
6
7
8
94
120
82
281
257
223
11
P
P
11
12
21
19
26
26
417
411
405
Q.A
7
8
10
4
6
5
42
42
52
207
215
225
10
10
10
55
75
68
11
37
30
245
219
190
447
312
135
14
15
1!)
140
•J9
76
ai
60
74
795
675
741
CA
28
26
27
1
9
7
126
lit)
106
393
192
371
15
18
15
136
110
118
95
92
77
171
181
136
857
721
580
12
29
26
209
186
141
144
149
124
1589
1577
1110
Cl.
2
10
12
4
5
3
64
68
66
161
169
145
5
9
5
79
65
55
16
15
27
192
211
152
155
161
238
15
10
9
104
89
57
77
105
80
620
921
565
Tl.
25
21
27
a
9
6
112
160
117
510
518
569
7
11
9
154
144
138
61
55
40
560
552
486
624
420
279
21
27
19
169
157
110
200
198
171
1194
1146
768
NU
11
29
28
4
4
5
81
68
6)
609
584
501
21
22
20
175
177
162
7)
69
66
512
509
459
950
821
710
42
14
29
182
167
152
146
.117
128
1821
1695
1511
TA
56
52
54
12
13
11
213
228
198
1139
1121
1072
10
36
28
129
321
299
131
124
106
1092
1061
945
1574
1241
989
65
60
48
151
124
282
146
315
299
3017
2841
2281
NFA/TA-'
V
t
t
f
V
V
r
p
p
p
p
p
p
p
p
p
p
p
25.1
23.2
23.9
P
P
P
V
V
P
P
P
!•
26.1
27.6
P
P
P
P
P
P
P
CK/T1.-'
P
P
P
F
F
F
16.4
F
F
P
P
19.2
P
P
P
14.8
P
P
10.5
13.0
P
16.8
P
16.8
P
P
P
P
P
P
18.5
P
15. B
13.1
13.2
15.3
P
P
P
TlW
P
P
t
1.86
F
1.11
1.63
F
F
P
P
1.13
P
P
P
P
P
P
P
P
P
1.05
1.09
1.06
P
P
P
P
P
P
P
P
P
1.37
1.45
1.34
P
P
P
C11RKAT-
P
P
P
P
1.84
P
1.97
P
1.61
P
P
P
P
P
P
1.70
P
P
P
P
P
1.91
1.79
P
P
1.99
P
P
P
P
P
P
P
1.88
F
1.55
P
1.71
P
QKAT-'
F
F
F
1.20
P
P
F
F
F
P
P
P
P
1.18
P
F
1.15
P
F
1.04
I. 11
P
1.01
P
p
F
p
F
P
P
P
1.11
P
I.Ob
F
K
P
F
P
NI/TA-"
P
P
P
F
F
F
P
3.1
F
P
P
P
P
P
P
P
P
P
2.7
1.7
P
P
P
P
P
P
P
P
P
P
P
P
P
p
P
P
p
P
P
-------
TABLE III-4
(continued)
Nuoiu of
firm
N t r i-o
Aluminum
Mobl 1 Oil
Mulyci>ri>
Morton Norwich
Nalco Chemical
Mat lonal
Clieoiuuarch
H.i I ional
Ultulllcrs
Nul ional
Industries
N.I 1 1 mi j)
SurvJce
National
Sturch & Chum.
National Steel
Nc-w £11^1 anij
tine 1 u.i r
N4:wim>nt Mining
Y.arl'
(x-l)
(x-2)
(x-3)
(x-l)
(x-2)
(x-3>
(x-l)
(x-2)
(x-3)
(x-l)
(x-2)
(x-3)
(x-l)
(x-2)
(x-3)
(x-l)
(x-2)
(x-3)
U-l)
(x-2)
(x-3)
(x-l)
(x-2)
(x-3)
(x-l)
(x-2)
(x-3)
(x-l)
(x-2)
(x-3)
(x-l)
(x-2)
(x-3)
(x-l)
(x-2)
(x-3)
(x-l)
(x-2)
(x-3)
NI
3
3
4
BIO
1047
849
8
14
8
18
25
24
32
27
25
17
14
10
62
90
72
15
17
14
18
23
23
19
18
17
58
176
98
3
3
2
53
114
103
CF
5
5
6
1522
1560
1282
16
22
14
35
41
39
44
36
32
19
14
10
89
117
101
22
HA
NA
29
33
32
25
25
23
17k
286
207
4
4
3
85
140
123
QA
26
17
16
3954
3861
2986
10
12
14
107
105
96
77
58
71
64
50
42
328
315
34)
93
100
66
87
90
87
58
42
16
318
578
42J
8
II
9
129
123
111)
CA
51
49
43
6156
5827
3939
30
32
30
201
191
164
118
100
100
86
67
52
629
641
629
IB2
197
154
139
161
138
100
89
70
763
857
681
11
13
11
298
266
186
Cl.
10
9
18
5234
5206
3375
a
15
13
59
89
61
47
41
30
34
21
15
163
181
174
110
143
101
45
74
55
32
33
28
503
581
431
3
3
3
126
124
72
TL
28
27
22
8209
7638
4976
42
50
42
222
208
186
58
49
35
43
29
20
429
454
504
189
219
182
74
104
85
84
69
47
1201
1096
958
4
4
3
481
440
379
MU
43
42
40
6841
6436
5715
92
77
75
238
231
217
158
138
122
74
59
48
595
565
505
92
86
81
155
148
140
130
116
102
1209
1192
1067
19
16
13
648
637
564
TA
71
69
62
15050
14074
10690
134
127
117
460
439
403
216
187
157
117
88
68
1024
1019
1009
281
305
261
229
252
225
214
185
149
2410
2289
2024
23
20
16
1129
1077
944
NKA/TA?'
27.8
29.0
P
r
P
f
f
f
r
f
?
r
p
V
27.4
20.0
p
If
p
V
f
f
f
21.2
p
r
p
p
p
p
p
p
p
p
p
28.8
p
p
p
CF/TL-'
19.2
19. 1
P
18.5
P
P
P
P
P
15.7
19.7
P
P
P
P
P
P
P
P
P
P
11.7
10.0
10.0
P
P
P
P
P
P
14.2
P
P
P
P
P
17.7
P
P
TL/mi*'
P
P
P
1.20
1.19
P
P
P
P
P
P
P
P
P
P
P
P
P
P
P
P
F
F
F
P
P
P
P
P
P
P
P
P
P
CIIRHAT-'
P
P
P
F
F
F
P
P
P
P
P
P
P
P
P
P
P
P
P
]>
r
1.65
F
1.53
P
P
P
P
P
P
1.52
F
1.58
P
P
P
P
P
P
O.KAT-'
P
P
F
F
F
F
1.14
F
1.12
P
1.17
P
l>
P
P
P
p
P
P
P
P
F
F
F
P
P
P
P
P
P
F
F
F
P
P
P
1.02
F
P
NI/TA-'
P
P
P
P
P
P
P
P
P
3.8
P
P
P
P
P
P
P
P
P
P
P
P
P
P
P
P
P
P
P
P
2.4
P
P
P
P
P
P
P
P
-------
TABLE III-4
(continued)
Name of
f 1 rut
Nl. Iiulu-it r Jus
Northwest
Industries
NW Steel 6 Wirfc
Hurt on Co.
Nucor
Corporation
Oaklte
Products
Occidental
l*t-troluiuu
Ulln
Cttrporut iu n
P6F Industries
Park Chiiiolcul
Park
Elect rochciofcal
Peerless Tube
PenilUii 1 1
Yeari/
(x-l)
(x-2)
(x-J)
(x-l)
(x-2)
(x-1)
(x-l)
(x-2)
(x-3)
(x-l)
(x-2)
(x-3)
(x-l)
(x-2)
(x-3)
(x-l)
(x-2)
(x-3)
(x-l)
(x-2)
(x-3)
(x-l)
(x-2)
(x-3)
(x-l)
(x-2)
(x-3)
(x-l)
(x-2)
(x-3)
(x-l)
(x-2)
(x-3)
(x-i)
(x-2)
(x-3)
(x-l)
-------
TABLE II1-4
(conLinued)
I
oo
Name uf
Film
PeilllZOil
-
Phelpa Doil^u
Pgh-Dub Mollies
Sletl
I'll CalMH|;li
Forging*
Plant
Industries
Portec
j
Kedwan
Industtl us
Republic Sltul
Kevere ^oupfcr
& Brass
Reynolds MetalB
Kohl in Industries
Kusco Indutit r i*-'s
Season-All
Induut r Ics
Yea pi'
(x-l)
(x-2)
(x-3)
(x-l)
(x-2)
(x-3)
(x-l)
(x-2)
(x-3)
(x-l)
(x-2)
(x-3)
(x-l)
(x-2)
(x-3)
(x-l)
(x-2)
(x-3)
(x-l)
(x-2)
(x-3)
(x-l)
(x-2)
(x-3)
(x-l)
(x-2)
(x-3)
(x-l)
(x-2)
(x-3)
(x-l)
(x-2)
(x-3)
(x-l)
(x-2)
(x-3)
(x-l)
(x-2)
(x-3)
Nl
107
128
84
71
122
109
4
2
3
9
3
2
2
1
1
6
5
2
-23
-24
-15
72
171
87
-31
17
3
60
111
45
2
5
-3
.4
.2
.9
.8
.8
.6
CF
215
218
155
106
158
145
7
4
5
12
6
5
5
t
3
7
6
4
-22
-22
-13
157
252
166
-15
34
19
133
188
120
5
7
-2
1
1
2
1
1
1
QA
297
218
245
145
121
229
62
58
49
25
24
16
11
10
8
18
17
8
22
22
31
282
502
352
64
61
84
294
399
221
21
20
21
9
10
10
4
3
3
CA
445
338
321
336
324
362
73
73
57
52
52
37
20
IB
16
41
44
27
33
33
45
614
759
613
lay
194
173
845
902
717
31
32
37
21
22
19
8
7
6
Cl.
247
146
228
189
231
145
48
51
42
26
31
18
19
16
11
11
18
10
20
20
39
269
409
333
88
83
84
292
340
250
13
14
18
9
9
9
4
3
3
Tl.
1453
1281
1135
759
601
454
56
59
43
A2
42
30
33
26
20
22
28
19
89
89
110
792
810
755
321
329
333
1173
1525
1409
36
3D
49
22
24
22
6
6
5
MM
573
515
628
893
892
815
39
36
35
42
36.
32
16
16
16
35
29
24
11
11
35
1279
1232
1107
137
170
153
831
799
709
18
17
13
7
7
7
6
5
5
TA
2026
1798
1762
1652
1493
1269
96
95
78
84
78
62
48
44
36
56
57
13
100
101
146
20/1
2042
1862
458
499
486
2204
2324
2118
54
55
61
3O
32
29
12
12
10
NFA/TA^'
P
P
P
P
P
P
20.6
F
2O. 6
F
F
F
P
P
P
25.3
21.1
27.9
f
f
f
V
P
P
P
P
P
P
P
P
28.0
28.3
26.0
24.3
24.1
26.2
P
P
P
CF/TI.-'
14.8
17.0
13.7
13.9
P
P
11.7
F
10.9
P
13.8
18.2
LS.I.
12.4
13.5
P
P
18.9
r
f
F
19.8
P
P
F
10.4
F
F
12.3
F
12.8
17.2
F
F
F
F
20.0
16.7
19.6
TI./NW*'
F
F
1.81
P
P
P
1.42
1.64
1.23
1.00
1.18
P
F
1.81
1.28
P
P
P
F
F
F
P
P
P
F
1.94
F
1.65
1.91
1.99
1.97
F
F
F
F
F
1.05
1.21
1.11
OJKKAT"'
1.80
P
F
1.78
F
P
1.51
F
F
P
1.68
P
f
r
t
r
f
e
I.b8
1.65
F
P
1.86
1.84
P
P
P
P
P
P
P
P
1.99
P
P
P
P
P
1.84
(jHA-ri7
1.20
P
1.07
F
F
P
P
1.14
1.17
F
r
F
F
F
F
P
F
F
1.13
1. 10
F
1.05
P
1 . 06
K
V
1.01
1.01
1.17
P
P
P
1.14
i.t>:;
1.07
1.05
i. ii
F
F
NI/TA--'
2.7
P
P
P
P
P
P
2.0
3.5
P
3.5
3.9
f
3.2
3.1
P
t
r
K
K
F
3.5
P
P
F
3.4
F
P
P
2.1
P
P
F
F
F
3.1
P
P
P
-------
TABLE III-4
(continued)
Name of
Pirn
Sharon Si i*e 1
Signal Companies
1
PsiBnoile
' Corporation
i
; st. joe
Minerals
j Standard
1 Alliance
'.
Sid. Pressed
Steel
Sunbeam
• Corporation
j
1 Sunshine
1 Mining
SynaJ toy
Tf-iineUiice
Forging
Texasfculf
U.S. Reduction
Company
Utah
Intvrnut iona I
»«,*'
(x-l)
(x-2)
(x-l)
(x-l)
(x-2)
(x-3)
(x-l)
(x-2)
(x-3)
(x-l)
(x-2)
(x-3)
(x-l)
(x-2)
(x-3)
(x-l)
(x-2)
(x-3)
(x-l)
(x-2)
(x-3)
(x-l)
(x-2)
(x-3)
(x-l)
(x-2)
(x-3)
(x-l)
(x-2)
(x-3)
(x-l)
(x-2)
(x-3)
(x-l)
(x-2)
(x-3)
(x-l)
(x-2)
(x-3)
NI
IS
47
12
1
176
59
17
24
22
82
89
38
2
3
2
li
23
31
28
4
5
3
3
2
.9
1
4
1
101
147
74
8
a
i
12
97
55
CF
25
58
22
78
201
80
32
39
35
113
113
56
2
3
2
8
16
12
38
45
41
5
6
4
3
2
2
2
.6
2
129
188
109
10
10
3
IbO
135
91
QA
89
140
62
485
464
254
70
69
75
126
146
82
10
13
12
26
37
27
199
188
152
26
21
24
9
8
6
7
9
4
118
229
103
14
22
8
118
80
60
CA
146
1U9
104
1149
1038
684
157
165
127
200
190
112
20
25
22
80
98
72
466
398
322
36
36
32
20
20
16
17
17
7
291
327
182
28
34
18
193
132
110
Cl.
50
86
49
443
403
273
39
54
39
107
119
65
10
12
11
27
43
34
213
198
126
9
11
9
6
8
5
a
13
6
105
142
68
16
23
12
176
118
84
Tl.
96
137
89
1058
747
672
99
115
100
195
197
151
19
22
21
56
71
54
340
272
207
31
34
30
12
15
13
29
24
13
527
417
336
19
29
19
505
461
428
NU
182
171
127
BOB
786
686
174
164
146
361
299
229
14
13
11
75
75
66
267
256
237
27
25
23
17
15
14
13
11
5
628
560
440
25
17
9
541
448
380
. 1
278
308
216 1
1866
1533
1358
273
279
247
556
496
380
33
35
32
131
146
120
607
528
444
58
59
53
30
30
26
42
35
18
1155
977
776
44
46
28
1046
909
828
NFA/TA-'
" P "
P
P
21.5
F
F
P
27.8
27.2
P
P
P
22.7
t
r
P
P
P
F
F
F
21.8
20.8
20.4
22.3
22.7
26.5
P
P
P
P
P
P
P
2.52
P
P
P
P
CF/TI.-'
P
P
P
F
P
11.9
P
P
P
P
P
P
r
11.2
11.3
13.8
P
P
11.2
16.4
19.7
16.8
17.9
14.1
P
15.4
13.6
F
P
15.3
P
P
P
P
P
13.5
P
P
P
ul
•n./nv-
f
p
p
1.31
p
p
p
p
P
p
p
p
1.38
1.76
1.99
P
P
P
1.27
1.06
P
1.15
1.33
1.28
P
P
P
F
F
F
P
P
P
P
1.69
F
P
1.03
1.13
CUHRAT--
P
P
P
P
P
P
P
t
P
.87
.60
.72
.99
.99
P
p
l>
p
P
P
P
P
P
P
P
P
P
F
F
P
P
P
1.78
F
F
F
F
F
QKAT-7
P
P
P
1.0'J
1.15
F
P
P
P
I. IB
P
P
F
1.01
I. 10
F
F
F
F
F
P
P
P
P
P
1.04
P
F
F
F
1.12
P
P
F
F
F
F
F
F
Nl/TA-'
P
P
P
2.2
P
P
P
P
P
P
P
P
P
P
P
•>
I
3.8
P
P
P
P
P
P
P
l>
3.1
P
P
P
P
P
P
P
P
P
P
P
-------
TABLE III-A
(concluded)
U)
o
Nairn: u f
Firm
Vu 1 Icy I n Jus t r luy
Vim Porn
Company
Vulcan
MatL*rtals
WuHhingCou
Steel
Uluniling-
Pittsburgh
Youngs town
Steul Door
»eari'
(x-l)
(x-2)
(x-3)
(x-l)
(x-2)
(x-3)
(x-l)
(x-2)
(x-3)
(x-l)
(»-2)
(x-l)
(x-l)
(x-2)
(x-3)
(x-l)
(x-2)
(x-3)
HI
IS
25
5
3
4
4
28
30
23
3
6
4
.6
73
7
1
3
2
CF
16
26
6
6
7
6
51
60
42
3
6
4
31
102
37
2
4
1
-------
TABLE II1-5
FINANCIAL DATA FOR FIRMS IN THE HOLDOUT BANKRUF1' SAMPLIi
I
U)
Name of
firm
Ai-.ine-llainlltoii HI'B.
Aurodex
AIM Companies
Capeliart Corporation
Cuimnmltire Corporation
Commonweal til Oil
Refining
Dynamics Corporation
tcologlcal Science
lilcor Chemical
Klcctrospace
Kill. Inc.
C-arcia Corporation
Year
197/U-U
1976(x-2)
1975(x-3)
1975(x-l)
1974(x-2)
1973- 3)
1969U-1)
1968(x-2)
1967(x-3)
I970(x-l)
1969(x-2)
1968(x-3)
1972(x-l)
1971(x-2)
1970(x-3)
I977U-1)
1976(x-2)
1975(x-3)
1977(x-l)
!977(x-2)
197S(x-3)
NI
NA
-0.9
0.)
NA
.0(16
0.3
-2.9
1.3
-1.1
-4.0
3.2
1.8
1.1
0.8
-3.0
-18.0
-32.1
-24.2
-6.3
-9.9
2.3
2.8
2.5
0
-39.5
-8.8
-4.7
1.8
2.0
0.9
1.6
0.8
-8.7
-10.0
-4.5
0.3
CF
MA
-0.4
O.S
NA
0.6
1.0
NA
2.1
-0.6
-2.7
3.9
2.3
1.9
1.6
-2.2
-6.1
-19.8
-13.3
-4.3
-7.4
4.3
3.6
3.0
0.3
-38.6
-7.4
NA
4.0
2.5
1.2
3.6
2.9
-7.8
-7.6
-2.4
2.2
l)A
NA
5.1
4.0
NA
2.5
5.2
4.1
4.6
4.5
14.5
15.2
8.6
17.0
13.6
9.3
129.0
110.0
151.0
43.3
51.1
51.6
18.4
13.5
3.1
2.9
6.7
9.7
NA
11. 1
8.1
23.7
21.8
21.2
28.5
33.0
40.4
CA
HA
11.2
10.4
NA
8.8
11.2
13.4
a. 5
8.5
41.0
39.0
14.1
26.6
20.4
17.3
232.6
248.2
252.6
63.7
80.9
82.0
26.8
18.1
3.5
2.9
8.2
11.5
33.9
30.1
27.1
36.3
31.7
33.8
61.9
61.0
75.7
a.
NA
9.0
7.1
NA
6.5
7.8
9.6
4.0
4.5
43.1
32.9
7.9
21.4
14.2
12.0
401.9
221.2
208.5
39.8
51.7
40.9
16.9
7.9
1.2
11.4
11.6
12.4
19.4
18.3
15.8
14.0
12.7
15.9
55.7
46.5
60.0
TL
NA
15.3
13.0
NA
13.4
15.8
19.6
10.6
14.2
33.9
39.3
10.1
37.1
29.9
27.0
409.1
420.9
394.2
29.9
26.7
63.7
66.1
37.6
20.6
43.5
53.9
62.7
30.5
29.7
27.0
44.2
40.3
44.4
68.7
59.8
68.6
NW
NA
4.3
5.2
NA
0.9
0.8
6.9
7.6
6.4
7.1
11. 1
7.9
9.8
8.6
9.4
127.5
145.2
179.3
50.0
73.9
37.1
22.7
16.2
7.3
-32.0
6.5
3.9
13.8
11.9
10.1
10.8
10.4
9.8
12.1
22.1
26.7
TA
NA
19.6
18.2
NA
14.3
16.6
26.5
18.2
20.6
41.0
50.4
18.0
46.9
38.5
36.4
536.6
566.1
573.5
79.9
100.6
100.8
88.8
53.8
27.9
11.5
60.4
66.6
44.3
41.6
37.1
55.0
50.7
54.2
80.8
81.9
95.3
NFA/TAi'
NA
f ,
P
NA
F
22.9
21.7
25.6
24.8
f
f
r
f
r
r
f
t
t
F
F
r
V
f
V
e
r
t
r
f
r
r
t
20.7
r
f
r
CF/TI.*'
NA
F
F
NA
f
F
NA
19.8
F
f
F
f
F
F
F
F
F
F
f
F
F
F
F
F
F
F
NA
13.1
F
r
r
F
F
F
F
F
TI./NW -'
NA
F
F
NA
F
V
F
1.39
F
F
F
1.28
F
F
F
F
V
F
H
P
1.72
F
F
F
F
F
F
F
F
F
F
F
f
F
F
F
OIKIIAT-'
NA
F
F
NA
F
F
F
H
1.89
F
F
1.78
F
F
F
F
F
F
1.60
1.56
P
1.59
t
P
F
F
F
1.75
1.64
1.72
t
P
P
F
F
F
HUM-
MA
F
F
NA
F
F
F
1. 15
1.00
F
F
1.09
F
F
F
F
F
F
1.09
F
1.26
1.09
P
P
F
F
F .
F
F
F
P
P
P
F
F
F
•
• P "' 30.11 (passes all Itstu); F < 2U.O (falla al
- P > 20.0 (p.iiisuu all testa); F < 10.0 (falls al
— P < 1.0 (passes all tests); F > 2.0 (falls all
all lesttl)
all tests)
tests)
-v > 2.0 (passes all tents); F < 1.5 ((alls all tests)
- P > 1.2 (passes all testu); F < 1.0 (falls all tests)
-------
TABLE II1-5
(concluded)
M
ii
I
to
Niimf. of
Klrui
CKT Corporation
l.ulsure Croup, Inc.
M^istur Hr.in
Puller liiHlriiuiuiics
K. II. iu
Sltkln
TMA Co.
iranso«rain
Ul Icox-Clbbs
Huberts Company
Stulbi-r
Plcduitjnt Industrie
Year
1978(x-l)
1977(x-2)
1976(x-3)
I970(x-l)
1969(x-2)
1968(x-3)
19/OU-l)
1969(x-2)
1968(x-3)
1974(x-l)
1973(x-2)
1972(x-3)
1968(x-l)
1967(x-2)
1966U-3)
1977(x-l)
1976(x-2)
1975(x-3)
I970(x-l)
1969U-2)
I968(x*3)
1970
-------
I
LO
U>
TABLE II1-6
FINANCIAL DATA FOR FIRMS IN THE HOLDOUT NON-BANKRUPT SAMPLE
Naiuu of
Kliui
Attains Ml 1 1 is Corp.
Air Products & Chemicals
Akzoiia
Aiutitafie
American Pt*trofina
Alias Corp.
Avoiidale Mills
Hall Corporation
Bu 14 1 ng-llciul nway
boftj Warner
bucliLcr Corporation
Year"
x-l
x-2
x-J
x-l
x-2
x-3
x-l
x-2
x-3
x-l
x-2
x-3
x-l
x-2
x-3
x-l
x-2
x-3
x-l
x-2
x-3
x-l
x-2
x-3
x-l
x-2
x-3
x-l
x-2
x-3
x-l
x-2
x-3
HI
1.7
-5.9
1.3
54.2
39.7
21.1
7.9
33.4
35.7
8.1
13.5
9.8
40.2
86.7
37.0
3.0
1.6
-2.4
4.0
6.3
8.1
14.1
9.5
7.2
3.2
2.8
2.5
44.5
50.8
71.3
0.7
1.1
0.9
CF
3.8
-3.6
3.5
107. 6
85.5
61.6
41.3
63.4
63.6
15.5
20.3
16.7
71.3
IJ7.2
56.2
4.8.
3.1
-0.8
11.8
14.2
16.0
23.3
18.2
NA
5.1
4.7
4.2
87.3
93.9
111. 8
1.5
1.9
1.7
()A
11). 0
7.1
11.2
171.0
168.0
111.0
124.0
99.0
118.0
44.0
40.7
44.3
177.0
170.0
118. 0
12.9
12.7
9.5
40.0
49.8
38.4
511.6
40.6
35.7
20.2
15.1
18.0
253.0
245.0
271.0
3.5
4.5
4.5
CA
20. 3
19.8
24. 8
248.0
223.0
144.0
265.0
258.0
228.0
101.2
93.5
88.1
282.0
280.0
172.0
27.8
26.5
20.8
66.8
74.2
65.7
129.6
105.1
75.1
45.9
42.9
41.0
590.0
688. 0
642.0
11.8
12.7
9.6
Cl.
8.8
7.8
8.3
193.0
162.0
106.0
94.0
80.0
87.0
49.3
35.4
33.2
1S5.0
148.0
100.0
21.2
17.6
11.8
32.9
41.6
33.8
71.5
54.3
33.4
14.0
13.4
17.2
244.0
264.0
291.0
4.0
5.1
5.4
TL
16.6
24.0
26.5
475.0
411.0
309 . 0
341.0
301.0
264.0
92.6
79.1
75.2
238.0
230.0
175.0
25.6
22.1
16.0
35.4
43.8
36.0
122.3
89.4
64.8
29.0
29.4
29.7
504.0
615.0
530.0
10.5
12.3
10.4
NW
19.0
17.0
23.0
3U1.0
248.0
211.0
315.0
323.0
304 . 0
90.0
86.8
84. B
364.0
345.0
276.0
27.2
24.3
23.1
74.5
74.3
70.3
102.8
91.7
84.8
33.1
30.8
29.4
b89.0
670.0
642.0
5.3
4.6
3.5
TA
35.4
41.1
49.8
776.0
659.0
520.0
656.0
624.0
51.8.0
182.6
165.9
160.0
602.0
575.0
451.0
52.8
46.4
39.1
109.9
118.1
106.3
225.1
181.1
149.6
62.1
60.2
59.1
1193.0
1285.0
1172.0
15.8
16.9
13.9
2 /
NFA/TA-
P
f
r
f
r
r
t
p
p
p
p
p
p
p
p
p
p
p
p
p
t
p
p
p
20.7
22.5
25.3
P
P
P
23.0
22.0
26.0
CF/TL-'
P
F
11.5
P
P
19.9
12.1
P
P
16.7
P
P
P
P
P
18.8
14.0
F
P
32.4
P
19.0
20.4
11.1
17.6
16.0
P
17.3
15.3
P
14.3
15.4
16. 3
TI./NW4-'
I'
1.41
1.15
l.SB
1 .66
l.4(>
1 . 0«
I'
1'
i .03
P
V
1'
P
f
e
p
p
p
H
r
1.19
p
p
p
V
1.01
p
p
p
i.'ja
F
F
CUliRAT-'
P
P
P
F
F
F
P
P
P
P
|>
P
1.82
1.89
1.72
F
1.51
1.76
1'
1.79
1.94
1.81
1.94
P
P
|»
P
P
P
P
P
P
1.78
IJRA'I*'
P
F
I1
F
1.04
1.05
P
P
P
F
1.15
P
1.14
1.15
1.18
F
F
F
P
1.20
1.14
F
F
1.07
P
1.13
1.05
1.04
F
F
F
F
F
-For all firms, x-l Is 1975.
-V > 30.0 (passes all testa)
- P > 20.0 (passes all tests)
- P < 1.0 (passes all tests);
-- P > 2.0 (passes all tests);
- I' > 1.2 (passes all tests);
x-2 Is 1974, anil x-3 Is 19/3.
F < 20.0 (falls all lur.ls)
F < 10.0 (falls all tests)
F > 2.0 (falls all tests)
F < 1.5 (falls all tests)
F < 1.0 (falls all lestii)
-------
TABLE IIT.-6
(continued)
V
CO
Name ul
t'iria
Hurl lug tun
Carter Wai lace
Chemut run
Clark Oil
CulL Industries
Coiicltuuico
Cone Hills Corp.
Crompcon Co.
Crown Central
Daliliitruni Corporation
Damon Corporal Ion
Daniel Industries
Dan River Inc.
ra.rl'
X-1
X-2
x-3
x-1
K-2
x-1
x-1
x-2
x-3
x-1
x-2
x-3
x-1
x-2
x-1
x-1
x-2
x-1
x-1
x-2
x-3
x-l
x-2
x-1
x-1 .
x-2
x-3
x-1
x-2
x-3
x-1
x-2
x-3
x-1
x-2
x-3
x-1
x-2
x-3
NI
39.8
9'J.5
82.4
8.4
3.1
10.4
10.4
22.3
10.4
5.2
-7.1
30.5
52.1
77.3
26.7
0.3
1.8
1.8
24.2
15.3
9.7
-2.6
U.I
5.8
5.5
10.2
8.4
0.9
1.8
-4.1
1.8
8.1
9.7
4.4
3.0
1.3
-2.9
7.0
10.4
CF
132.9
191.7
173.1
11.1
5.6
12.5
46.9
38.1
26.1
19.1
6.1
43.2
79.8
103.8
52.2
1. 3
2.7
2.6
37.8
27.3
21.6
1.5
2.5
7.9
13.0
16.6
14.4
1.5
2.2
-3.5
4.5
10.7
11.9
5.2
3.6
1.9
11. U
19.7
22.1
I(A
485.0
496.0
459.0
52.7
71.5
68.7
99.0
92.1
68.5
60.0
76.7
84.4
205. (I
227.0
205.0
11.1
11 .1
10.5
73.5
78.7
81.9
29.1
22.6
26.4
5O.6
50.0
'ia.3
4.5
3.5
S.4
52.3
59.0
55.5
13.0
9.6
7.8
89.4
79.2
88.3
CA
853.0
932.0
aao.o
83.0
94.0
99.0
190.2
173.5
146. 1
120.7
141.1
141.4
497.0
497.0
422.0
19.1
26.0
21.1
162.2
163.1
162.1
41.7
14.9
31.8
82.7
76.9
61.1
9.5
6.0
9.1
73.1
81.0
72.1
32.6
23.9
16.7
188.7
198.7
210.4
Cl.
237. 0
319.0
293.0
20.5
25.3
24.7
54.0
60.1
55.4
95.3
123.6
101.5
187.0
172.0
119.0
6.8
11.4
7.2
51.5
65.5
67.9
24.2
11.8
12.5
79.0
69.3
47.2
6.5
2.8
5.5
20.5
25.3
24.7
16.6
11.4
8.5
57.2
76.9
73.1
Tl.
671.0
757.0
756.0
30.9
19.9
22.5
161.9
158.6
141.4
208.3
202.9
253.5
491. 0
438.0
196.0
12.1
19.5
15.0
96.0
95.5
102.8
51.9
42.1
21.4
1 34 . 6
112.2
91.2
14.5
7.6
12.5
44.0
48.9
49.4
22.5
15.6
9.6
161.6
166.9
165.6
NH
H'i'j.O
887.0
825.0
124.3
119.0
118. /
209. 4
183.6
167.4
100.5
98.8
109.5
175.0
140.0
275.0
22.5
22.5
21.2
177.5
156.4
145.6
18.9
41.5
42.7
71.7
66.2
57.0
6.9
6.0
4.2
62.5
61. 8
52.7
21.9
16.9
14.3
144.0
148. 0
144.0
TA
1566.0
1644.0
1581.0
155.2
118.9
141.2
371.3
342.2
308.8
10H.8
301.7
263.0
866.0
778.0
671.0
34.8
42.0
36.2
273.5
241.9
248.4
90.8
81.6
64.1
206.3
178.4
148.2
21.4
11.6
16.7
106.5
110.7
102.1
44.4
12.5
21.9
308.0
315.0
310.0
NFA/TA2-
F
F
P
F
20.9
F
f
r
t
t
t
P
F
P
P
F
27.1
29.0
F
F
P
P
P
F
F
P
P
P
P
F
25.1
22.4
24.4
25.5
26.2
29.3
P
P
P
CF/TI.3'
19.8
25.1
22.9
t
t
e
r
p
18.4
F
F
17.0
16.3
P
13.2
F
13.8
17.3
P
P
21.0
F
F
P
F
14.8
15.8
10.3
F
F
10.2
F
P
P
V
19.8
6.7
11.8
13.3
TI./HW-'
F
f
V
r
f
p
f
p
p
F
F
F
1.31
1.29
1.44
P
F
P
P
P
F
1.33
1 .01
P
1 .SB
1.69
1.60
K
1.27
F
F
P
P
1.03
P
F
1.13
1.13
1.15
CUKUAT-'
P
P
P
F
F
P
F
P
P
F
F
F
P
V
l>
P
P
P
P
P
P
1.81
F
P
F
F
F
F
P
1.65
F
P
t
1.96
P
1.96
F
F
r
b/
O.KAT-
P
F
P
P
F
P
P
P
F
F
F
F
1.10
P
F
F
F
P
P
P
P
F
F
P
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
-------
TABLE II1-6
(continued)
I
(Jl
Num.: "f
Firm
l)e Sol,,, Inc.
IHvursu
KjBlu-Pithcr
Klcoi Clii.-mlcal (Jorp.
Kiuury liuliistriuB
KLhyl Curp.
Fab Ind.
Fairwuul Chemical
Fcrro
Flltrul
Fuule Mineral Co.
Y,ari'
x-l
x-2
x-l
x-l
x-2
x-3
x-l
x-2
x-3
x-l
x-2
x-3
x-l
x-2
x-3
x-l
x-2
x-l
x-l
x-2
x-3
x-l
x-2
x-3
x-l
x-2
x-3
x-l
x-2
x-3
x-l
x-2
x-3
x-2
x-3
Nl
2.B
0.6
8.1
4.0
3.6
2.5
IB. 7
18. 0
14.8
5.2
2.5
0.4
10.7
10.2
7.0
61.0
74.3
52.9
2.1
0.9
1.9
0.1
1.1
0.2
15.2
19.9
15.9
5.5
4.4
3.1
9.8
11.3
2.8
-4.8
-1.0
CF
9.2
6.5
11.9
6.4
5.5
4.1
27.8
27.2
23.4
6.1
3.1
0.8
15.2
14.7
11.1
103.1
111. 1
B5.3
3.3
2.0
3.0
O.5
1.3
0.5
21.8
25.9
21.1
7.6
5.6
3.6
14.8
15.6
6.9
-4.6
-0.5
O.A
32.1
32.0
31.3
23.4
18.7
16.0
74.9
60.0
57.1
15.9
9.3
12.0
35.8
30.5
26.7
246.0
179.0
274.0
15.2
9.9
10.5
1.4
2.6
1.3
84.7
77.3
73.3
17.6
13.6
22.0
18.7
26.7
24.0
4.1
5.1
CA
75.4
87.5
97.0
15.6
32.2
24.6
IJI.l
129.0
111.2
16.7
10.0
12.1
61. 8
52. 1
45.7
371.0
128.0
365.0
24.7
20.5
19.7
2.9
4.2
2.4
I39.B
147.1
121.6
2b.O
20.0
25.2
49. B
45.2
45.0
5.4
6.4
Cl.
51.7
62.2
62.4
21.5
21.9
14.5
41.5
46.7
38.4
9.2
7.0
10.4
23.6
21.2
13.7
105.0
128.0
95.0
24.7
6.2
6.5
0.5
1.8
0.7
54.7
69.9
54.1
9.9
6.2
5.0
15. B
21.3
19.1
2.5
2.3
Tl.
64.9
75.7
74.8
33.3
31.2
23.0
81.8
BB.5
81.0
14.7 ••
14.2
21.7
61.4
49.1
40.8
440.0
383.0
348.0
10.2
11.7
11.9
0.7
2.1
l.l
88. 0
101.4
86.1
28.6
24.4
5.0
52.4
26.7
29.5
2.6
2.5
MU
70. 1
69.2
72.0
22.6
19.5
16.7
136.9
124.0
111.1
12.2
7.0
4.6
81. 7
74.3
67.4
436.0
400.0
348.0
14.6
17.8
17.2
4.0
3.8
2.8
126.3
115.1
99.4
31. b
27.9
26.1
74.4
67.2
58.7
4.4
9.1
TA
135.0
144.9
146.8
55.9
50.7
19.7
218.7
212.5
192.1
26.9
21.2
26.1
143.1
123.4
108.2
876.0
783.0
691.0
19.8
29.5
29.1
4.7
5.9
1.9
212.1
216.5
185.5
60.4
52.1
11.1
126.8
93.9
88.2
7.0
11.6
NFA/TA-'
P
t
V
24.5
21.5
22.4
P
P
P
t'
21.3
22.8
P
f
P
P
P
P
P
P
P
27.6
20.3
P
29.9
27.6
28.8
P
P
f
P
P
P
F
F
CF/T.,5-'
14.2
F
18.6
19.2
17.6
17.8
P
P
P
P
P
F
P
t
e
p
p
p
12.4
17.1
P
P
P
P
P
P
t
P
p
p
p
p
P
F
F
TI./MU--
.O'l
.04
.47
.6
.18
1'
p
P
1.21
F
F
V
p
p
1.01
P
1.00
p
p
P
P
p
P
P
P
P
V
P
P
P
P
P
P
P
UIKKAT-
.55
.06
.7
P
P
P
1.82
F
F
P
P
P
P
P
"'
P
P
P
P
P
P
P
P
p
P
P
P
P
P
''
F
P
WAT*7
F
F
F
1.09
F
1.1
P
P
P
P
F
F
P
P
P
P
P
P
P
P
P
P
P
l>
1>
1.11
P
P
P
P
1.18
P
P
P
P
-------
TABLE II1-6
(concluded)
M
M
i
LO
Nairn; uf
firm
(In If Resources 6
Chemical
tluiuly & Herman
Hunt Qiumical Corp.
Hy drome La la, Inc .
.lioi Walter
Youngs tuwu Sliect &
Tube
Yeari'
x-l
x-2
x-3
x-l
x-2
x-3
x-1
x-2
x-3
x-l
x-2
x-3
x-l
x-2
x-3
x-l
x-2
x-3
NI
28.7
36.2
7.4
12.7
12.2
5.4
3.6
6.1
6.L
3.4
3.7
1.0
69.3
62.9
53.6
40.9
96.4
43.3
CF
40.7
43.7
12.7
15.7
14.8
7.8
NA
NA
NA
4.8
5.2
4.5
98.4
88.6
76.2
105.2
156.8
97.9
QA
81). 6
79.0
36.9
56.6
54.5
50.1
19. 0
18. 0
15.2
14.4
12.8
15.3
618. 1)
618.0
528.0
208.0
319.0
251.0
CA
135.9
111.0
61.4
111.5
127.0
100.7
28.9
26.9
24.4
50.2
52.2
45.8
813.0
824.0
693.0
471.0
520.0
199.0
CL
63.1
46.6
24.3
59.6
82.0
72.0
6.5
6.9
5.6
14.3
18.7
13.8
519.0
627.0
498.0
250.0
304.0
198.0
TL
158.9
142.3
102.6
90.1
111.9
92.4
6.5
6.9
5.6
28.1
34.1
30.0
875. 0
862.0
731.0
546.0
560.0
561.0
IIW
118.4
76.1
41.7
56.4
45.4
35.1
38.7
36.7
32.1
33.2
29.7
26.5
432.0
398.0
351.0
713.0
692.0
664.0
TA
277.3
218.4
144.3
146.5
157.2
127.5
45.2
43.6
37.7
61.3
63.8
56.5
1309.0
1260.0
1082.0
1259.0
1252.0
1125.0
NFA/TA-
P
t
r
21.6
F
F
P
e
e
F
F
F
p
p
p
p
p
p
CF/TL-'
P
P
12.4
17.4
13.2
f
P
P
P
17.1
15.2
15.0
11.2
10.3
10.4
19.3
P
17.4
TL/MU-'
1.34
1.87
If
1.60
F
F
P
F
P
f
1.15
1.13
F
F
F
P
P
P
CIJKIIAT-
V
V
f
1.87
1.55
F
P
P
P
P
P
P
l.bO
F
F
l.BH
1.71
P
6/
(
-------
IV. EVALUATION CRITERIA AND DEVELOPMENT OF RATIO TESTS
All test alternatives were initially evaluated against both the
primary non-bankrupt and bankrupt samples. In this evaluation the
Agency assumed that a firm would have to meet the financial criteria
required by a test for its most recent fiscal year in order to pass the
test (i.e.,.a one-year eligibility requirement). The seven financial
ratios selected for detailed analysis were first individually tested;
for each ratio examined several different pass-fail cutoff points were
investigated. The most promising individual ratio tests were then
combined into a series of 120 two-ratio tests (a firm must pass both
ratio elements to pass), three-ratio tests (a firm must pass all ratio
elements to pass), and contingent three-ratio tests (a firm must pass
two of three ratio elements to pass).
To further validate whether these tests represent the most effec-
tive indicators of future firm viability, the Agency performed a
supplementary analysis. Other financial variables identified in the
preliminary literature review but not included in the set of 120 tests
were evaluated against a subset of 32 non-bankrupt and 12 bankrupt firms
from the primary sample that had proven particularly difficult to
classify. The ratio that performed best in this supplementary analysis,
net fixed assets/total assets, was then combined with the other candi-
date financial ratios, and 31 additional three- and four-ratio tests
were evaluated.
All ratio tests were examined with two variants: a one-year eligi-
bility requirement and a three-year eligibility requirement. The former
IV-1
-------
requires that a firm meet the requirements of a financial test based on
its most recent annual report in order to pass the test. The latter
requires that a firm meet the requirements of a financial test based on
its three most recent annual reports in order to pass the test.
A. Evaluation Criteria
Five evaluation criteria have been developed to assist the Agency
in comparing the test results. These measures are defined and described
below.
(1) ANB represents the percentage of sampled non-bankrupt firms
that pass a given test. For tests with a one-year eligibility require-
ment, A^JTJ is measured as the percentage of firms in the non-bankrupt
firm sample which passed the test in the year 1975. For tests with a
three-year eligibility requirement, Ajjg is measured as the percentage of
firms in the non-bankrupt firm sample which passed the test for all of
the years 1973-1975.
Because the non-bankrupt firm sample created for this analysis is
composed of companies from industry categories which are most likely to
have on-site treatment, storage and/or disposal facilities requiring
financial assurance, and contains a representative sample of firms from
all size classes, the Agency believes that the ANB results can be used
as a reasonable surrogate measure of the overall percentage of viable
firms that would be able to pass a financial test. Therefore, tests
which have a. very low ANB will generate large private sector expendi-
tures, because many viable firms will be required to establish alter-
native forms of financial assurance. Alternatively, tests with a high
Ajjg will result in low private costs of complying with RCRA.
IV-2
-------
(2) M represents the percentage of sampled bankrupt firms which
would fail a given financial test with insufficient time remaining prior
to bankruptcy to ensure that alternative financial mechanisms are avail-
able for facility closure, post-closure, and liability requirements.
Because firms in financial distress often suffer a rapid deterio-
ration in their liquid assets in the two to three years prior to bank-
ruptcy, and because of possible delays in enforcement and litigation,
the Agency has concluded that a one-year lead time would not be suffi-
cient to guarantee that sufficient funds would be available to cover the
costs of closure, post-closure, and liability requirements. Three
different assumptions as to what constituted "sufficient lead time" were
analyzed by the Agency:
Best case - If a test fails a firm at least two years prior to
bankruptcy, there will still be sufficient time to ensure the
funding of alternative mechanisms; for this case the value of
M is designated as Mg.
Worst case - If a test fails a firm at least three years prior
to bankruptcy, there will still be sufficient time to ensure
the funding of alternative mechanisms; for this case the value
of M is designated as MW.
Most probable case - All firms that are first eliminated by a
test three years prior to bankruptcy will provide alternative
financial assurance. One-half of the firms that are first
eliminated two years prior to bankruptcy will provide alterna-
tive financial assurance. For this case the value of M is
designated as Mp.
c' ' IV-3
-------
The implications of all three of these assumptions on the inter-
pretation of possible classification results are presented in Table IV-
1. Although the posibility of injunctive relief under Section 7003 of
RCRA might render all three of these assumptions conservative estimates,
the Agency has decided to use the "most probable case" definition of
lead time requirements. (See, Hazardous Waste Section, Land and Natural
Resources Division, United States Department of Justice, Annual Report
(October 1980.)
(3) C represents the percentage of sampled bankrupt firms that
fail a financial test with sufficient time remaining prior to bankruptcy
to ensure that alternative financial mechanisms are available for
facility closure, post-closure, and liability requirements. C is thus
measured as:
C = 100 - Mp
(4) D reoresents the difference between the percentage of non-
bankrupt firms passing a test and the percentage of bankrupt firms
passing the test, and is called the "discriminating power" of a test. D
is calculated according to the formula:
D = ANB ~ %
The higher the D score, the better the test discriminates between bank-
rupt and non-bankrupt firms. A test receiving a D score of 100 would
perfectly discriminate between bankrupt and non-bankrupt firms, passing
all non-bankrupt firms and failing all bankrupt firms. A D score of
zero would indicate that the same percentage of bankrupt firms pass a
test as non-bankrupt firms, suggesting that the test does not discrim-
inate between bankrupt and non-bankruot firms. A negative D score
indicates that more bankrupt firms pass a test than non-bankrupt firms.
IV-4
-------
TABLE IV-1
ESTIMATED PROBABILITY THAT A FIRM WILL NOT SET UP ALTERNATIVE FINANCIAL MECHANISMS
FOR VARIOUS PATTERNS OF PASSING AND FAILING FINANCIAL TESTS
Test Result—
1 Year
Prior to
Bankruptcy
P
P
P
P
F
F
F
F
2 Years
Prior to
Bankruptcy
P
P
F
F
P
P
F
F
3 Years
Prior to
Bankruptcy
P
F
P
F
P
F
P
F
Worst Case
Probability
for
One-Year
Eligibility
Requirement
1.0
1.0
1.0
1-0
1.0
1.0
1.0
0
Probability
for
Three-Year
Eligibility
Requirement
1.0
0
1.0
0
1.0
0
1.0
0
Most Probable Case
Probability
for
One-Year
Eligibility
Requirement
1.0
1.0
1.0
1.0
1.0
1.0
.5
0
Probability
for
Three-Year
Eligibility
Requirement
1.0
0
.5
0
1.0
0
.5
0
Best Case
Probability
for
One-Year
Eligibility
Requirement
1.0
1.0
1.0
1.0
1.0
1.0
0
0
Probability
for
Three-Year
Eligibility
Requirement
1.0
0
0
0
1.0
0
0
0
—'P = Passes Test; F = Fails Test.
-------
(5) E represents the number of firms per 10,000 which pass a given
financial test and will enter bankruptcy without providing alternative
financial assurances. E, like M, can be calculated for three different
assumptions as to what constitutes sufficient lead time.
Best case - If a test fails a firm at least two years prior to
bankruptcy, there will still be sufficient time to ensure the
funding of alternative mechanisms; for this case the value of
E is designated as Eg.
Worst case - If a test fails a firm at least three years prior
to bankruptcy, there will still be sufficient time to ensure
the funding of alternative mechanisms; for this case the value
of E is designated as Ew.
Most probable case - All firms that are first eliminated by a
test three years prior to bankruptcy will provide alternative
financial assurance. One-half of the firms that are first
eliminated two years prior to bankruptcy will provide alter-
native financial assurance. For this case, the value of E is
designated as Ep.
Ep is used for most analytic purposes. A detailed formula for Ep, as
noted in Section I, is:
Ep . F(MP)
(U-F) x Aj^) -I- (F x Mp)
However, because of the very small values of F, a less complicated
formula approximates this completely correct measure with an error of
less than 1 percent for all tests examined. Ep is therefore calculated
as:
IV-6
-------
ANB
B. Development of Ratio Tests
1. Evaluation of Single Ratio Tests
Each ratio listed in Section III.C.4 was tested against the primary
bankrupt and non-bankrupt samples. Several pass-fail cutoff points were
analyzed for each ratio since the midpoint values used in the historical
literature often vary widely between studies. These single-ratio tests
are summarized in Table IV-2.
A.
In choosing ratios and cutoff points to employ in further tests,
the Agency did not use the same sample as it used in the rest of this
study. As noted in Section III, a larger primary sample which contained
retail firms was employed in the earlier phases of this study. The
evaluations of individual ratios and cutoff points shown in Table IV-2
were conducted using this larger sample and are reported in terms of
that larger sample in this Section. This sample consisted of 60 bank-
rupt firms and 147 non bankrupt firms. M and C were evaluated using a
three-year eligibility assumption, whereas Ajjg was evaluated using a
one-year eligibility assumption. However, results for the tests
presented in Sections V and VIII are derived from the sample with retail
firms removed.
As Table IV-2 illustrates, the cash flow/total liability ratio is
the most significantly predictive single ratio, attaining the two
highest D scores, the three highest C scores, and the four highest E
ratings (i.e., the lowest number of firms per 10,000 that will fail
without providing alternative financial assurance). The margins of
i
IV-7
-------
TABLE IV- 2
PERFORMANCE OF SINGLE -RATIO
Test M 21
Description ; "P
CF/TL > .05
CF/TL > .1
CF/TL > .15
CF/TL > .2
CF/TL > .3
TL/NW < 3.0
TL/NW < 2.0
TL/NW < 1.5
TL/NW < 1.2
TL/NW 1.0
NI/TA > 0
NI/TA > .02
NI/TA > .04
NWK/TA > 0
NWK/TA > .2
NWK/TA > .25.
CURRAT > 1.2
CURRAT > 1.5
CURRAT > 2.0
QRAT > 1.0
QRAT > 1.2
NS/TA > 1.0
2 /
^
t
30.4
12.5
5.4 '
5.4
5.4
61.7
35.0
26.7
16.7
11.7
50.0
21.7
11.6
83.3
46.7
30.0
66.7
36.7
18.3
18.7
13.6
71.7
NS/TA > 1.5 1 41.7
58.9
32.1
17.9
8.9
3.6
81.7
61.7
41.7
31.7
25.0
68.3
43.3
31.6
90.0
60.0
50.0
76.7
55.0
31.6
33.9
25.5
80.0
48.4
3/
^B
88.5
83.8
69.9
52.0
24.5
96.0
90.7
83.4
67.5
60.9
92.1
83.4
72.8
100.0
74.8
63.6
98.7
97.3
74.9
76.2
53.1
77.4
36.9
21
C
55.3
77.7*
88.3*
92.8*
95.5*
28.3
51.7
65.8
75.8*
81.6*
40.9
67.5
78.4*
13.3
46.6
60.0
28.3
54.1
75.0*
73.7*
80.4*
24.1
54.9
4/
D
29.6
51.7*
52.0*
43.1*
20.9
14.3
29.0
41.7*
35.8
35.9
23.8
40.1*
41.2
10.0
14.8
13.6
22.0
42.3*
43.3*
42.3*
27.6
[-2.6]
4/
P
11.1
5.9*
3.7*
3.0*
4.0*
16.4
11.7
9.0*
7.9*
6.6*
14.1
8.6*
6.5*
19.1
15.7
13.8
16.0
10.4*
7.3*
7.6*
8.1*
21.6
[-11.5] 26.9
^
values expressed in % terms except Ep which represents the number
of firms (per 10,000 passing the test) that will fail without providing
alternate financial assurance. All results are based on the original
samples that included retail firms.
2/
—Based on three-year eligibility requirement.
-Based on one-year eligibility requirement.
-^Based on three-year eligibility requirement for M and one-year eligi-
bility requirement for A^.
*Exceeds minimum performance cutoff point.
IV-8
-------
difference between the scores of the best cash flow ratios and those of
other ratios are also substantial. Thirteen of the ratios tested
reduced the normal failure rate for large (greater than $10 M in net
worth) firms by more than 50 percent (from 22 per 10,000 to less than 11
per 10,000). Ten ratios eliminated more than 70 percent of bankrupt
firms in sufficient time to establish alternative financial guarantee
mechanisms: nine had a discriminatory power (D) of greater than 40
percent. If these three conditions are viewed to be the minimum
acceptable values of test effectiveness, TL/NW, NI/TA, CURRAT, and ORAT
all have cutoff scores that satisfy these criteria; only the net working
capital to total assets and net sales to total assets ratios failed to
provide significant results.
Those parametric values which satisfied one or more of the above
minimum criteria (13 in all) were used to develop multi-ratio tests; the
others were dropped from further consideration. There were two excep-
tions to this general classification rule: (1) the cash flow/total
liability greater than .3 ratio was dropped, because the E ratings
demonstrated that this test was already dominated by the .2 and .15 cash
flow cutoff points, and its extremely high rate of non-bankrupt firm
rejections made it a poor candidate for use in multi-ratio tests and (2)
the total liabilities/net worth cutoff point of 2.0 was added to the
list of ratios for detailed consideration, despite the fact that it
failed to meet the minimum reauirements, because preliminary investiga-
tions revealed that this ratio was one of the few that could be used
effectively to classify electric utilities (see Section IV. B).
IV-9
-------
2. Evaluation of Multi-Ratio Tests
The 13 financial ratios selected from the initial evaluation
process were linked in various combinations and then retested against
the primary bankrupt and non-bankrupt samples. The ratios were combined
in three ways: two-ratio tests (firm must pass both elements of the
test to pass the test), three-ratio tests (firms must pass all elements
to pass the test), and three-ratio contingent tests (firms must pass 2
of 3 elements to pass the test). Since CF/TL and NI/TA are alternative
methods of measuring a firm's rate of return on its assets, these ratios
were not included within the same test; the same procedure was followed
with CURRAT and QRAT, the two liquidity measures being evaluated.
In all, 120 alternative tests were investigated in this phase.
Section VIII presents the results of all tests both with one-year and
three year liability requirements in Tables VIII-1 to VIII-4 and VIII-5
to VIII-8, respectively.
As illustrated in these Tables, two-ratio tests are more accurate
than single-ratio tests in the C and E ratings. This pattern continues,
at a somewhat lower rate of increase, for the more stringent three-ratio
tests (passage of all elements required to pass). The three-ratio (pass
2 out of 3) tests evaluated barely out-performed the single-ratio cash
flow tests on the C, 0, and E measures. Although test formulations of
the three-ratio type greatly increase the eligibility of firms in the
non-bankrupt sample, they also increase misclassifications of bankrupt
firms by a similar (and sometimes greater) amount.
IV-10
-------
The major problem with the multi-ratio tests evaluated in this
phase is that the discriminating power (D score) of the best tests does
aot increase greatly. The greater levels of bankrupt firm identifi-
cation and overall test effectiveness indicated in these results
apparently are attained mainly by excluding a larger number of viable
firms. Consequently, the Agency performed supplementary tests to estab-
lish whether there were other financial variables that could be added to
these tests to improve their discriminating power, while retaining high
levels of bankrupt firm detection.
3. Test of Supplementary Financial Ratios
Based on the results of the initial round of testing, a set of 32
non-bankrupt and 12 bankrupt firms was identified that were consistently
misclassified by the tests evaluated in the first phase. The results
achieved in testing this set of firms against a second set of financial
ratios are summarized below.
The following ratios used in this auxiliary analysis were selected
from the financial ratios tested in prior bankruptcy forecasting
studies:
Retained earnings/total assets (RE/TA)
Earnings before interest and taxes/total assets (EBIT/TA)
Cash flow/net sales (CF/NS)
Balance sheet value of preferred and common stock/net worth (PC/NW)
Balance sheet value of preferred and common stock/current and long-
term debt (PC/CL -1- LTL)
Net fixed assets/total assets (NFA/TA)
IV-11
-------
Cash/total assets (Cash/TA)
Cash/current liabilities (Cash/CL)
Net sales/total assets (NS/TA)
Net working capital/total assets (NWK/TA)
Values of each of the ratios were computed for the firms described
above. The data were then analyzed to determine the values for each
ratio that would provide the most accurate overall classification of the
tested firms. The percentage of bankrupt and non-bankrupt firms that
passed the resulting test were then compared to derive an estimate of
the discriminating power (D1) of each ratio. For this analysis, the
value of D' , defined as the difference between A^g and Mp, was used.
(This is similar to D, but substitutes Mp for My). These results are
presented in Table IV-3.
In many cases, the ratio tested had higher values of Mp than of
Ajjg, resuitng in negative values for D' . A single ratio test with a
negative discriminating power would not be likely to add to the overall
effectiveness of financial tests combining several ratios. Three ratios
— RE/TA, NS/TA, and Cash/CL — had some positive impact on firm cate-
gorization, with D' values of 15 to 24 percent. These results, however,
were far exceeded by the classification results achieved by using the
NFA/TA ratio. Three different cutoff points for NFA/TA were tested,
with resulting D' ratings of 30 to 44 percent. More significantly, the
ratio of NFA/TA greater than .3 correctly classified 9 of the 12 bank-
rupt firms as non-viable three years prior to bankruptcy.
IV-12
-------
TABLE IV-3
TESTS OF SUPPLEMENTARY FINANCIAL RATIOS-1
II
Test
Variables
RE/TA >
RE/TA >
EBIT/TA >
NS/TA > 1
NS/TA > 1
CF/NS >
CF/NS >
PC/NW >
PC/TL >
NFA/TA >
NFA/TA >
NFA/TA >
Cash/TA >
Cash/CL >
Cash/CL >
NWK/TA >
.2
.25
.12
.25
.4
.04
.05
.1
.1
.15
.25 -
.3
.04
.15
.2
.25
»P
41.7
25.0
41.7
41.7
41.7
91.7
83.3
66.7
50.0
50.0
33.3
25.0
50.0
50.0
33.3
75.0
v •
62.5
37.5
43.8
65.6
50.0
62.5
43.8
40.6
40.6
93.5
61.3
54.8
65.6
46.9
34.4
56.3
D'^/
20.8
12.5
2.1
23.9
8.3
-29.2
-39.5
-26.1
-9.4
43.5
28.0
29.8
15.6
-3.1
1.1
-18.7
I/
Tests run against a portion of the EPA primary bankrupt and non-
bankrupt samples comprising 32 non-bankrupt and 12 bankrupt firms
that were frequently misclassified by the initial set of candidate
financial tests.
I/
D1 = Difference between the percentage of viable firms passing a
test and the percentage of non-viable firms passing the same test
in year x-3.
IV-13
-------
The NFA/TA ratio represents the portion of a firm's assets tied to
long-term tangible property, excluding the portion of these fixed assets
that has already been depreciated. It represents the reserve of assets
that a company can call on in a time of financial difficulties, either
as a reliable source of earnings or a. potential source (through sale) of
needed capital. This ratio has not been extensively evaluated in
previous bankruptcy forecasting studies; however, several studies have
•>.
examined the ratio that represents to some extent its inverse — current
assets/total assets (CA/TA), the fraction of total assets in the form of
cash, inventories, and short term receivables. Both Edmister (1972) and
Deakin (1972) have found CA/TA to be a significant indicator of finan-
cial stability; and both studies have indicated that a high CA/TA is
negatively correlated with continued firm solvency. A NFA/TA test also
can be readily passed by almost all electric utilities, a category of
hazardous waste disposers that encounters great problems in passing
candidate tests using other more common financial ratios (see Section
VI). For these reasons, the Agency decided to evaluate a number of
tests incorporating NFA/TA as a variable against the entire primary
sample. The results of these tests with one-year and three-year eligi-
bility requirements are presented in Section VIII, Tables VIII-4 and
VIII-8, respectively. As will be discussed in the next Section, most of
the tests incorporating NFA/TA (Tests 121-151) dominate tests which do
not incorporate this ratio.
IV-14
-------
V. TEST PERFORMANCE
A. Construction of the Performance Curves
The performance of a test is measured by two values: Ep, the number
of firms per 10,000 which pass the test that later enter bankruptcy
without providing alternative financial assurance; and ANB, the percent-
age of non-bankrupt firms passing the test. Both of these values are
used to determine the relative costs of alternative tests. In order to
determine the best tests, the Agency ascertained, for particular values
of Ep, the test which allows the highest value of ANB. A test which,
for any given value of Ep has the highest value of Ajjg, is termed a
dominant test.
Using the results for all tests-of a given type (see Section VIII
for complete test results), a graph was constructed for each of the
eligibility requirements, depicting on one axis the number of bankrupt
firms per 10,000 which pass the test, (E?) and on the other axis the
percentage of non-bankrupt firms which pass the test (ANB).
Figure V-l illustrates the method used to determine if a point C is
on the performance curve. If a test is to be on the performance curve,
there must be no other tests which dominate that test.
V-l
-------
ANB
BETTER
DIFFERENT
DIFFERENT
WORSE
FIGURE V-l
TECHNIQUE FOR DETERMINING IF A POINT IS ON THE PERFORMANCE CURVE
For example, a particular test which passes 6 per 10,000 of bank-
rupt firms and passes 70 percent of non-bankrupt firms would have the
coordinates of (6, 70%) in Figure V-l. To compare the performance of
other tests to this test, as illustrated in the Figure, any tests repre-
sented by points in the northwest quadrant, using point C as the point
of origin, dominate point C because both the value of Ep is lower (i.e.,
fewer bankrupt firms pass the test) and the value of AJJB is higher
i
(i.e., a higher percentage of non-bankrupt firms pass the test.) A test
with the coordinates (4, 80%) would fall in the northwest quadrant, and
would indicate that 4 of 10,000 bankrupt firms and 80 percent of non-
bankrupt firms pass the test. This makes it superior to the test repre-
sented by point C. Thus, if there were any tests with coordinates in
the northwest quadrant, the test represented by point C would not be on
V-2
-------
the performance curve. Similarly, any point in the southeast quadrant
is absolutely inferior to point C.
Points in the other two quadrants cannot be classified as domi-
nating point C, but can only be classified as different. For example, a
point with coordinates (8, 80%) means that 80 percent of non-bankrupt
firms pass the test, which is absolutely superior to point C; however, 8
out of 10,000 bankrupt firms also pass, which is inferior to point C.
There is thus no way to classify any points in the northeast or south-
west quadrants as dominant to point C. Points in those quadrants may or
may not be on the performance curve; they would have to be evaluated in
the same manner as point C.
Once the performance curve is constructed, it is sometimes useful
to compare specific tests to the performance curve as a whole. If a
test is plotted above the performance curve, then it dominates at least
one test on the performance curve. If it is below the performance
curve, then it is dominated by at least one test on the performance
curve.
B. Performance Curves for the Primary Sample
Figure V-2 shows the performance curve for the set of 151 ratio
tests with one-year eligibility requirements. Table V-l lists the
components of the tests and the values of Ep and ANB for each test.
Seventeen tests lie on the performance curve, thus reducing the set of
tests that need to be further considered from 151 to 17. The perfor-
mance curve includes tests ranging from a test with an Ep of 0 and an
of 49 percent to one with an Ep of 10.1 and an ANB of 96 percent.
V-3
-------
<5
100-
95-
NB
50-
45-
40-
35-
139
FIGURE V-2
PERFORMANCIi CURVE FOR TESTS WITH
ONE-YEAR ELIGIBILITY REQUIREMENTS
143
T22
Ho .4
2.'o 2'.5 3.'o
5.'o 5.S
65
7.s
9 .*5 lo'.O lo'.5
-------
TABLE V-l
DOMINANT TESTS
(One-Year Eligibility Requirement)
Test
Number
139
122
141
143
138
151
134
113
150
Test
Variables
NFA/TA > . 3
CF/TL > .1
CURRAT >2.0
NFA/TA > .3
CF/TL > .1
TL/NW < 1.2
NFA/TA > .2
CF/TL > . 1
CURRAT > 1.5
NFA/TA > .2
TL/NW < 1.2
QRAT >1.0
NFA/TA > .2
CF/TL > .1
CURRAT >1.5
NFA/TA > . 3
CF/TL > .1
TL/NW < 1.2
CURRAT >2.0
NFA/TA > .3
CF/TL > .1
TL/NW < 1.0
CF/TL > .1
TL/NW < 1.0
QRAT > 1.0
NFA/TA > .3
CF/TL > .1
TL/NW < 1.5
CURRAT >2.0
Number of
Variables
Required
To Pass
All
All
All
All
All
3 of 4
2 of 3
2 of 3
3 of 4
EP
0
1.4
1.9
2.6
2.8
3.7
4.6
5.2
5.5
^B
(%)
49
53
55
58
76
77
79
80
81
1
1
V-5
-------
TABLE V-l
(concluded)
Test
Number
135
149
136
120
137
98
146
100
Test
Variables
NFA/TA > .3
CF/TL > .1
TL/NW < 1.2
NFA/TA > .3
CF/TL > .1
TL/NW <2.0
CURRAT >2.0
NFA/TA > .3
CF/TL > . 1
TL/NW < 1.5
CF/TL > . 1
TL/NW < 2.0
QRAT > 1.0
NFA/TA > .3
CF/TL > . 1
TL/NW < 2.0
CF/TL > . 1
TL/NW < 2.0
CURRAT > 2.0
NFA/TA > .2
CF/TL > . 1
CURBAT > 2.0
CF/TL > . 1
TL/NW < 2.0
CURRAT > 1.5
Number of
Variables
Required
To Pass
2 of 3
3 of 4
2 of 3
2 of 3
2 of 3
2 of 3
2 of 3
2 of 3
EP
5.6
6.0
7.9
9.1
9-1
9.3
9.9
10.1
\B
(Z)
83
87
89
92
92
93
95
96
V-6
-------
The tests on the performance curve had certain ratios in common.
All tests on the performance curve included the cash flow to total
liabilities ratio. Thirteen of the 17 tests on the performance curve
included the net fixed assets to total assets ratio. Thirteen out of .17
tests included a total liabilities to net worth ratio.
Figure V-3 shows the performance curve (indicated by the heavy
line) for the set of 151 ratio tests with a three-year eligibility
reauirement. The lighter line shows, for purposes of comparison, the
performance curve for tests with a one-year eligibility requirement.
Descriptions of these tests and their values for Ep and ANB are given in
Table V-2. Nine tests with a three-year eligibility requirement lie on
.this performance curve. The tests range from one with an Ep of 0 and an
ANB °^ -^ percent to one with an Ep.of 7.3 and an Ajjg of 86 percent. All
of these tests include the cash flow to total liabilities ratio and the
net fixed assets to total assets ratio.
Comparing these two curves shows that the general effect of a
three-year eligibility requirement on any given test is to lower the
values of both Ep and ANB. The overall effect is relatively slight,
with the following two exceptions: (1) Test 139 markedly improves with
a one-year eligibility requirement: (2) only tests with a one-year
eligibility requirement can achieve values of Ajjg greater than 84
percent.
The choice between a one-year and three-year eligibility require-
ment cannot be based solely on performance, however, since the Agency
must also consider the administrative burdens associated with different
V-7
-------
FIGURE V-3
<
CO
100-
95-
90-
85-
80-
75-
NB
COMPARISON OF PERFORMANCE CURVE FOR TESTS WITH
THREE-YEAR ELIGIBILITY REQUIREMENTS WITH PERFORMANCE
CURVE FOR TESTS WITH ONE-YEAR ELIGIBILITY REQUIREMENTS
Key: Dark Line - Three-Year Tests
Light Line - One-Year Tests
1.0 l!5 2.0 2.5 3:0 3.5 AiO 4.5 5.0 5.5 6iO 6.5 7.0 7.5 8.0 8.5 9.0 9.5 10
-------
TABLE V-2
DOMINANT TESTS
(Three-Year Eligibility Requirement)
Test
Number
139
141
125
138
127
134
135
136
137
Test
Variables
NFA/TA > .3
CF/TL > .1
CURRAT > 2.0
NFA/TA > .2
CF/TL > .1
CURRAT > 2.0
NFA/TA > .25
CF/TL > .1
TL/NW < 1.5
NFA/TA > .2
CF/TL - > .1
CURRAT > 1.5
NFA/TA > .2
. CF/TL > . 1
TL/NW < 2.0
NFA/TA > .3
CF/TL > . 1
TL/NW < 1.0
NFA/TA > .3
CF/TL > . 1
TL/NW < 1.2
NFA/TA > .3
CF/TL > .1
TL/NW < 1.5
NFA/TA > .3
CF/TL > .1
TL/NW < 2.0
Number of
Variables
Required
To Pass
All
All
All
All
All
2 of 3
2 of 3
2 of 3
2 of 3
EP
0
1.2
1.4
1.6
1.9
2.9
4.1
5.4
7.3
**B
(%)
32
44
56
64
67
72
77
82
86
V-9
-------
tests. In order to assess the performance of one-year and three-year
tests together, a merged performance curve, consisting of all the
dominant tests without regard to eligibility requirements, was con-
structed. Such a curve is the envelope curve of the two performance
curves shown in Figure V-3. Figure V-4 presents this envelope curve and
Table V-3 presents a full description of the tests on this curve. This
merged performance curve contains 16 tests, of which four have a three-
year eligibility requirement, and 12 have a one-year eligibility
requirement.
G. Performance with Respect to the Holdout Sample
As noted in Section III.B. the Agency used a holdout sample in this
study to help ensure the statistical validity of the results. Figure
V-5 compares the performance on the primary sample and on the holdout
sample of those one-year eligibility tests that appeared on the primary
sample performance curve; Figure V-6 provides this same comparison for
three-year eligibility tests. table V-4 compares the results of all
tests with a one-year eligibility requirement and Table V-5 of all tests
with a three-year eligibility requirement. Tables V-6 and V-7 provide
complete holdout sample results for a larger set of tests.
For one-year eligibility tests, 11 of the 17 tests yield results
for the holdout sample which lie above the performance curve for the
primary sample. The remaining six tests are only slightly below the
primary sample performance curve. This result strongly suggests that
search bias has not led to tests which are incapable of validly discrim-
inating between bankrupt and non-bankrupt firms when applied to samples
V-10
-------
f
100-
95-
90-
85-
80-
75-
V
39 (1)
45-
40-
35-
FIGURE V-4
PERFORMANCE CURVE OF 1JEST FINANCIAL TESTS WITH EITHER
ONE-YEAR OR THREE-YEAR ELIGIBILITY REQUIREMENTS
9U(1)
(1)
120(1)
0
2.*0 2*.5 3.'o 3?
5.'o 5.'s 6.'o 6.*5 ?!o 7.*5 S.'o 8.!5 9.'o 9 .*5 lo'.O lo'.5
-------
TABLE V-3
DOMINANT FINANCIAL TESTS WITH EITHER ONE-YEAR
OR THREE-YEAR ELIGIBILITY REQUIREMENTS
Test
Number
139
125
138
127
138
151
134
136
135
Number
Of Years
Required
To 3e
Eligible
1
3
3
3
1
1
1
3
1
Test
Variables
NFA/TA > . 3
CF/TL > .1
CURRAT > 2.0
NFA/TA > . 25
CF/TL > . 1
TL/NW < 1.5
NFA/TA > . 2
CF/TL > . 1
CURRAT > 1.5
NFA/TA > . 2
CF/TL > . 1
TL/NW < 2.0
NFA/TA > .2
CF/TL > . 1
CURRAT > 1.5
NFA/TA > . 3
CF/TL > . 1
TL/NW < 1.2
CURRAT > 2.0
NFA/TA > . 3
CF/TL > . 1
TL/NW < 1.0
NFA/TA > . 3
CF/TL > . 1
TL/NW < 1.5
NFA/TA > . 3
CF/TL > . 1
TL/NW < 1.2
Number of
Variables
Required
To Pass
All
All
All
All
All
3 of 4
2 of 3
2 of 3
2 of 3
SP
0
1.4
1.6
1.9
2.8
3.7
4.6
5.4
5.6
ANB
49
56
64
67
76
77
79
82
83
7-12
-------
TABLE V-3
(concluded)
Test
Number
149
136
120
137
98
146
100
Number
Of Years
Required
To Be
Eligible
1
1
1
1
1
1
1
Test
Variables
NFA/TA > .3
CF/TL > . 1
TL/NW < 2.0
CURRAT > 2.0
NFA/TA > . 3
CF/TL > . 1
TL/NW < 1.5
CF/TL > . 1
TL/NW < 2.0
QRAT > 1.0
NFA/TA > . 3
CF/TL > . 1
TL/NW < 2.0
CF/TL > . 1
TL/NW < 2.0
CURRAT > 2.0
NFA/TA > . 2
CF/TL > . 1
CURRAT > 2.0
CF/TL > .1
TL/NW < 2.0
CURRAT > 1.5
Number of
Variables
Required
To Pass
3 of 4
2 of 3
2 of 3
2 of 3
2 of 3
2 of 3
2 of 3
EP
6.0
7.9
9.1
9.1
9.3
9.9
10.1
^B
87
89
92
92
93
95
96
V-13
-------
FIGURE V-5
f
100-
95-
NBH
45-
PERFORMANCE CURVE V. HOLDOUT SAMPLE PERFORMANCE
FOR TESTS WITH ONE-YEAR ELIGIBILITY REQUIREMENTS
• 137
• 136
120
143
146)
137)
(120)
(98)
(100)
• 122
40-
35-
139
H0 .5
2.'0 2*5 3.*0
5.'o 5.*5
6.*5
7.*5
9.'o 9 .'s lo'.O lo'.5
-------
100-
95
FIGURE V-6
PERFORMANCE CURVE V. HOLDOUT SAMPLE PERFORMANCE FOR
TESTS WITH THREE-YEAR ELIGIBILITY REQUIREMENTS
<
90
85-
80-
75-
« 136
(137)
134*127
• 138
.5 1.0 15 20 2.5 3.0 3.5 410 4.5 5.'0 5.5 6.0 615 710 7/5 8JO 8.5 9.0 915 10.0 10.5
-------
TABLE V-4
f
COMPARISON OF PRIMARY AND HOLDOUT SAMPLE PERFORMANCE FOR TESTS ON THE ONE-YEAR PERFORMANCE CURVE
(Subscript 11 Indicates Holdout Sample Result, No Subscript Indicates Priniary Sample Result)
Test
Number
98
100
113
120
122
134
135
136
137
US
139
Test
Variables
CF/TL > . 1
T1./NW < 2.0
CURRAT > 2.0
CF/TL > . 1
TL/NW < 2.0
CURRAT > 1.5
CF/TL > . 1
TL/NW < 1.0
QRAT > 1.0
CF/TL > . 1
TI./NU < 2.0
QKAT > 1.0
NFA/TA > . 3
CF/TL > . 1
TL/NW < 1.2
NFA/TA > . 3
CF/TL > . 1
TL/NW < 1.0
NFA/TA > . 3
CF/TL > . 1
TL/NW < 1.2
NFA/TA > . 3
CF/TL > . 1'
TL/NW < 1.5
NFA/TA > . 3
CF/TL > . 1
TL/NW < 2.0
NFA/TA > . 2
CK/TL > . 1
CURRAT > 1.5
NFA/TA > .3
CF/TL > . 1
CURRAT > 2.0
Number of
Variables
Required
To Pass
2 of 3
2 of 3
2 of 3
2 of 3
All
2 of 3
2 of 3
2 of 3
2 of 3
All
All
•Vii
20.8
31.3
14.6
18.8
0
4.2
4.2
12.5
16.7
10.4
0
"P
39.2
44.0
19.0
38.0
3.5
16.6
21.4
32. 1
38.0
9.5
0
Net
Change*
(V'W
+ 18.4
+ 12.7'
+4.4
+ 19.2
+3.5
+12.4
+17.2
+19.6
+11.3
-0.9
0
*NBI1
84
91
74
91
44
81
86
91
98
70
37
^B
93
96
80
92
53
79
83
89
92
76
49
Net
Change*
(ANBirANB>
-9
-5
-6
_,
-9
+2
+3
+2
+6
-6
-12
EP»
5.4
7.6
4.3
4.5
0
1. )
1. 1
3.0
3.7
:
3.3
,
i 0
El-
9.3
10.1
5.2
9. 1
1.4
4.6
5.6
7.9
9.1
2.8
U
Net
Change*
+3.9
+2.5
+0.9
+4.6
+1.4
+3.5
+4.5
+4.9
+5.4
-0.5
0
-------
TABLE V-4
(concluded)
I
M
-~J
Test
Number
141
143
146
149
150
151
Test
Variables
NFA/TA > .2
CF/TL > . 1
CURRAT > 2.0
NFA/TA > . 2
TL/NW < 1.2
QRAT > 1.0
NFA/TA > .2
CF/TL > . 1
CURRAT > 2.0
NFA/TA > . 3
CF/TL > . 1
TL/NW < 2.0
CURRAT > 2.0
NKA/TA > . 3
CF/TL > . 1
TL/NW < 1.5
CURHAT > 2.0
NFA/TA > . 3
CF/TL > . 1
TL/NU < 1.2
CURRAT > 2.0
Number of
Variables
Required
To Pass
All
All
2 of 3
3 of 4
3 of 4
3 of 4
MPII
4.2
4.2
31.3
6.3
6.3
U
»P
4.7
7.1
42.8
23.8
20.2
13.0
Net
Change*
(V^,,)
+0.5
+2.9
+ 11.5
+ 17.5
+ 13.9
+ 13.0
*NBII
51
47
93
86
74
72
ANB
55
58
95
87
81
77
Net
Change*
(ANBlfANB)
-4
-11
-2
-1
-7
-5
EPII
1.8
2.0
7.4
1.6
1.8
0
EP
1.9
2.6
9.9
6.0
5.5
3.7
Net
Change*
+0.1
+0.6
+2.5
+4.4
+3.7
+3.7
Mn all cases, a positive value in the Nee Change column represents an improvement in test performance in the holdout over the primary sample.
-------
TABLE V-5
COMPARISON OF PRIMARY AND HOLDOUT SAMPLE PERFORMANCE FOR TESTS
ON THE THREE-YEAR ELIGIBILITY PERFORMANCE CURVE
Test
Number
125
127
134
135
136
137
138
139
141
Test
Variables
NFA/TA > . 25
CF/TL > . 1
TL/NW < 1.5
NFA/TA > . 2
CF/TL > . 1
TL/NW < 2.0
NFA/TA > . 3
CF/TL > . 1
TI./NW < 1.0
NFA/TA > . 3
CF/TL > . 1
TL/NW < 1.2
NFA/TA > . 3
CF/TL > . 1
TL/NW < 1.5
NFA/TA > . 3
CF/TL > . 1
TL/NW < 2.0
•
NFA/TA > .2
CF/TL > . 1
CIJKRAT > 1.5
NFA/TA > . 3
CF/TL > . 1
CURHAT > 2.0
NFA/TA > . 2
CF/TL > . 1
CURKAT > 2.0
Number of
Variables
Requl red
To Pass
All
All
2 of 3
2 of 3
2 of 3
2 of 3
All
All
All
M.'l.
2.1
2.1
0
4.2
0.3
12.5
2.1
0
0
MP
3.5
5.9
9.5
14.3
20.2
28.5
4.7
0
2.3
Nee
Change*
+ 1.4
+3.8
+9.5
+10.1
+ 11.9
+ 16.0
+2.6
0
+2.3
ANBH
51
63
63
70
81
82
60
33
44
*NB
56
67
72
77
82
86
64
32
44
Net
Change*
(ANBH~*NB)
-5
-4
-9
-7
-1
-4
-4
+ 1
0
EHI1
0.9
0.7
0
1.3
2.3
3.4
0.8
0
0
EP
1.4
1.9
2.9
4. 1
5.4
7.3
1.6
0
1.2
Net
Change*
(VW
+0.5
+ 1.2
+2.9
+2.8
+3.1
+3.9
+0.8
0
+1.2
M
00
*lu all cases, a positive value in Che Net Change column represents an improvement in test performance in the holdout over the primary sample.
-------
TABLE V-6
PERFORMANCE OF BEST TESTS AGAINST HOLDOUT SAMPLE
(One-Year Eligibility Requirement)
Test
Number
11
12
26
67
68
98
100
110
113
120
Test
Variables
CF/TL > .1
TL/NW < 1.5
CF/TL > .1
TL/NW < 2.0
CF/TL > .1
CURRAT > 1 . 5
CF/TL > .1
TL/NW < 1.5
CURRAT > 1.5
CF/TL > .1
TL/NW < 2.0
CURRAT > 1.5
CF/TL > .1
TL/NW < 2.0
CURRAT > 2.0
CF/TL > .1
TL/NW < 2.0
CURRAT> 1.5
CF/TL > .1
TL/NW < 2.0
QRAT > 1.2
CF/TL > .1
TL/NW< 1.0
QRAT > 1.0
CF/TL > . 1
TL/NW < 2.0
QRAT > 1.0
Number of
Variables
Required
To Pass
All
All
All
All
All
2 of 3
2 of 3
2 of 3
2 of 3
2 of 3
E3
2.4
2.2
3.5
2.5
2.4
4.4
6.0
2.0
3.7
3.0
"P.
3.6
3.3
5.2
3.8
3.6
5.4
7.6
4.0
4.3
4.5
.=»
4.8
4.4
7.0
5.1
4.8
6.5
9.1
6.0
5.0
6.0
*NB
77
84
79
72
77
84
91
91
74
91
V-19
-------
TABLE V-6
(continued)
Test
Number
122
.
125
127
132
134
135
136
137
138
Test
Variables
NFA/TA > .3
CF/TL > . 1
TL/NW < 1.2
NFA/TA > .25
CF/TL > . 1
TL/NW < 1.5
NFA/TA > . 2
CF/TL > .1
TL/NW < 2.0
NFA/TA > .3
CF/TL > .15
TL/NW < 1.2
NFA/TA > . 3
CF/TL > .1
TL/NW < 1.0
NFA/TA > .3
CF/TL > .1
TL/NW < 1.2
NFA/TA > . 3
CF/TL > . 1
TL/NW < 1.5
NFA/TA > .3
CF/TL > .1
TL/NW < 2.0
NFA/TA > . 2
CF/TL > .1
CURRAT> 1.5
Number of
Variables
Required
To Pass
All
All
All
2 of 3
2 of 3
2 of 3
2 of 3
2 of 3
All
EB
0
1.6
2.5
1.3
1.1
1.1
2.0
2.8
2.6
i
i
EP
0
2.4
3.1
1.3
1.1
1.1
3.0
3.7
3.3
*V
0
3.1
3-7
1.3
1.1
1.1
4.0
4.7
3.9
^B
44
58
74
72
81
86
•
i
1
91
i
i
i
98
70
V-20
-------
TABLE V-6
(concluded)
Test
Number
139
141
143
146
149
150
151
Test
Variables
NFA/TA > .3
CF/TL > . 1
CURRAT > 2.0
NFA/TA > .2
CF/TL > . 1
CURRAT > 2 . 0
NFA/TA > .2
TL/NW < 1.2
QRAT > 1.0
NFA/TA > .2
CF/TL > .1
CURRAT > 1.5
NFA/TA > .3
CF/TL > .1
TL/NW < 2.0
CURRAT > 2.0
NFA/TA > . 3
CF/TL > .1
TL/NW < 1.5
CURRAT > 2.0
NFA/TA > . 3
CF/TL > . 1
TL/NW < 1.2
CURRAT > 2.0
Number of
Variables
Required
To Pass
All
All
All
2 of 3
3 of 4
3 of 4
3 of 4
EB
0
1.8
2.0
4.9
1.1
1.2
0
EP
0
1.8
2.0
7.4
1.6
1.8
0
^
0
1.8
2.0
10.8
2.1
2.5
0
^B
37
51
47
93
86
74
72
V-21
-------
TABLE V-7
PERFORMANCE OF BEST TESTS AGAINST HOLDOUT SAMPLE
(Three-Year Eligibility Requirement)
Test
Number
i
11
12
26
67
68
98
100
110
113
120
122
Test
Variables
CF/TL > .1
TL/NW <1.5
CF/TL > .1
TL/NW <2.0
CF/TL > .1
CURRAT >1.5
CF/TL > .1
TL/NW <1.5
CURRAT > 1 . 5
CF/TL > .1
TL/NW <2.0
CURRAT >1.5
CF/TL > .1
TL/NW <2.0
CURRAT >1.5
CF/TL > .1
TL/NW <2.0
CURRAT >1.5
CF/TL > .1
TL/NW <2.0
QRAT >1.2
CF/TL > .1
TL/NW <1.0
QRAT >1.0
CF/TL > .1
TL/NW <2.0
QRAT >1.0
NFA/TA > .3
CF/TL > .1
TL/NW <1.2
Number of
Variables
Required
To Pass
All
All
All
All
All
2 of 3
2 of 3
2 of 3
2 of 3
2 of 3
.All
EB
1.5
1.4
0
0
0
3.6
4.5
1.2
1.7
1.2
0
EP
2.9
2.7
2.1
1.5
1.5
4.8
6.2
3.7
2.5
4.3
0
*W
4.4
4.1
4.2
3.0
3.0
5.9
7.9
6.2
3.3
6.2
0
^B
63
67
65
60
60
77
81
74
56
74
40
V-22
-------
TABLE V-7 (Continued)
Test
Number
125
127
132
134
135
136
137
138
139
Test
Variables
NFA/TA > .25
CF/TL >' .1
TL/NW <1.5
NFA/TA > .2
CF/TL > . 1
TL/NW <2.0
NFA/TA > .3 '.
CF/TL > . 15
TL/NW <1.2
NFA/TA > . 3
CF/TL > .1
TL/NW <1.0
NFA/TA > .3
CF/TL > .1
TL/NW <1.2
NFA/TA > .3
CF/TL > . 1
TL/NW <1.5
NFA/TA > .3
CF/TL > .1
TL/NW <2.0
NFA/TA > .2
CF/TL > .1
CURRAT >1.5
NFA/TA > .3
CF/TL > . 1
CURRAT >2.0
Number of
Variables
Required
To Pass
All
All
•
2 of 3
2 of 3
2 of 3
2 of 3
2 of 3
All
All
EB
0
0
0
0
1.3
1.1
2.2
0
0
EP
90
.73
0
0
1.3
2.3
3.4
.77
0
A
1.8
1.5
0
0
1.3
3.4
4.5
1.5
0
**
51
63
58
63
70
81
82
60
33
V-23
-------
TABLE V-7 (Concluded)
Test
Number
141
143
146
149
150
151
Test
Variables
NFA/TA > .2
CF/TL > .1
CURRAT >2.0
NFA/TA > .2
TL/NW <1.2
QRAT >1 . 0
NFA/TA > .2
CF/TL > . 1
CURRAT >1.5
NFA/TA > .3
CF/TL > . 1
TL/NW <2.0
CURRAT >2.0
NFA/TA > .3
CF/TL > .1
TL/NW <1.5
CURRAT >2 . 0
NFA/TA > .3
CF/TL > . 1
TL/NW <1.2
CURRAT <2 . 0
Number of
Variables
Required
To Pass
All
All
2 of 3
3 of 4
3 of 4
3 of 4
EB
0
2.3
4.5
0
0
0
EP
0
2.3
7.4
.66
.71
0
*W
0
2.3
10.2
1.3
1.4
0
^B
44
40
81
70
65
63
V-24
-------
other than the sample used in the test selection. The results for the
three-year eligibility tests further confirm the validity of the tests
selected. All nine three-year eligibility tests for the holdout sample
lie above ,the primary sample performance curve. For the tests on the
merged performance curve, only three of the 16 tests on the performance
curve lie below the merged primary sample performance curve.
Although the holdout sample results attest to the validity of the
selection procedure for the tests on the primary sample performance
curve, they are less encouraging with respect to the accuracy of the
estimates of Ep and A^g. When one-year eligibility tests on the
primary sample performance curve are tested against the holdout sample,
ANB falls by an average of four percentage points while Ep falls by an
average of 2.6. Similar results are obtained for tests with a three-
year eligibility requirement. While small changes in the values of Ajjg
are not surprising given the sample size, the changes in Ep are unex-
pectedly high, given that the primary and holdout sample were randomly
assigned from the same original sample of firms. Apparently, the
resulting holdout sample contained firms with much weaker financial
performance two to three years prior to bankruptcy than the primary
sample.
If the results of the primary sample are used, Test 100 (one-year)
has the highest value of Ajjg of any of the ratio tests examined. If the
results of the holdout sample are used, Test 137 (one-year) is superior
to Test 100 (one-year), with both a higher Ajjg and a lower Ep.
V-25
-------
As a further check on the relative performance of these tests, Ep
and A"B were calculated using a weighted average of the performance for
the primary and holdout samples. These results are shown in Table V-
8. Using these weighted average results, all three tests would lie on a
performance curve, with none absolutely dominating the others.
D. Comparison to the Results of Other Tests
1. Other tests examined
This Section reports the performance of the other types of tests
which were examined in this study in addition to the ratio tests.
One of the other tests evaluated was the financial test proposed by
the Agency on May 19, 1980. That test consisted of the following
elements:
Total liabilities to net worth < 3
Net working capital > 2 X closure plus post-closure costs
Net worth > $10 million
This test had a one-year eligibility requirement. The effect of the net
worth requirement and the total liabilities to net worth ratio in elim-
inating bankrupt firms while admitting viable firms was evaluated using
the same methodology used for deriving the performance of the ratio
tests. The effect of the multiple requirement on the performance of the
test is more difficult to evaluate. A requirement that net working
capital be a multiple of total financial responsibility obligations
means that net working capital must be positive. As a result, the May
19 test was evaluated as a test which required a ratio of total liabil-
ities to net worth of less than three and a current ratio of greater
V-26
-------
TABLE V-8
PERFORMANCE OF SELECTED LESS STRINGENT TESTS
USING WEIGHTED AVERAGES OF HOLDOUT AND PRIMARY
SAMPLE TO DETERMINE Ep AND
Test
Number
137
146
100
Number
of Years
Required
To Be
Eligible
1
1
1
*
Test
Variables
NFA/TA > . 3
CF/TL > . 1
TL/NW < 2.0
NFA/TA > . 2
CF/TL > .1 .
CURRAT > 2.0
CF/TL > . 1
TL/NW < 2.0
CURRAT > 1.5
Number of
Variables
Required
To Pass
2 of 3
2 of 3
2 of 3 :
Tf
EP
7.1
9.0
9.2
NB
93
94
95
V-27
-------
chan one (i.e., the ratio of current assets to current liabilities must
be greater than one for there to be positive net working capital).
The Agency also evaluated tests which were composed only of
multiple elements. In this report, the term "Ability to Pay Test"
refers to those tests that required that net worth and net working
capital each exceed six times the total financial responsibility obli-
gation to be covered through a financial test. The performance of those
tests was evaluated in the same manner as the performance of the May 19,
1980 tests. That is, the net worth multiple was ignored as implicitly
accounted for by the $10 million in net worth requirement, and the
current ratio was required to be greater than one. The Agency evaluated
the Ability to Pay test using both one-year and three-year eligibility
requirements.
The performances of the May 19, 1980 and Ability to Pay tests are
shown in Table V-9. Neither the proposed May 19, 1980 test nor the two
Ability to Pay tests are dominated by other tests, and therefore all are
added to the performance curve. These tests pass almost all viable
firms, but also have very high values of Ep»
The presence in a test of multiple elements of the type included in
the Ability to Pay Tests does not influence the Ep or A^ results. All
ratio tests of any importance fail firms that would have failed a test
including a requirement that the current ratio be greater than 1. As a
result, the addition of such a requirement to a ratio test does not
alter either the A*,,, or E- of the test.
V-28
-------
TABLE V-9
PERFORMANCE OF SELECTED TESTS
WITH RESPECT TO THE PRIMARY SAMPLES
Test Description
May 19, 1980 Test
Ability to Pay Test
Ability to Pay Test
Parameters
Tested
TL/NW < 1
OJRRAT > 1
CURRAT > 1
CURRAT > 1
Eligibility
Requirement
One-Year
One-Year
Three-Year
^
97
100
99
EP
15.1
20.2
18.4
V-29
-------
2. Comparison to Results Obtained by Other Investigators
As noted in Section II, a number of studies have been published on
methods for forecasting bankruptcies. The most commonly used method in
these studies has been multi-discriminant analysis. The disadvantage to
this approach is that tests based upon multi-discriminant analysis would
require much more extensive reporting forms and would be more difficult
to check. Nevertheless, the Agency concluded that if multi-discriminant
analysis would lead to results clearly superior to those of the tests
adopted in this study, it should consider adopting one of the tests from
the published literature rather than one of the tests it had developed.
The Agency therefore compared its results to the results reported in the
literature.
Four of the studies which utilized multi-discriminant analysis were
further examined: Altman (1968), Altman et al. (1977), Deakin (1972)
and Deakin (1976). All of these studies examined the same basic problem
as that examined here: forecasting bankruptcies for relatively large
firms engaged in manufacturing and, in some cases, retail trade.
The first step in the comparison was to examine the results of
these studies as reported, and to use the baseline failure rate employed
in this study to derive a measure of E? comparable to that used in
reporting the results of this study. Multi-discriminant analysis is
particularly prone to search bias, with the result that performance with
respect to the holdout sample is almost always weaker than that with the
primary sample. Whenever possible, the performance reported for these
studies is that obtained against a holdout sample. (This was not pos-
sible for Altman et al. (1977) because no holdout sample was employed.)
V-30
-------
The results for these studies are shown in Table V-10. The ANB and
Ep for each study are plotted on a graph showing the performance curve
and holdout results for one-year eligibility tests in Figure V-7. As
can be seen from the Figure, the reported result for Deakin (1976)tis
dominated by both the primary sample performance curve and by many of
the results from the holdout sample. The Deakin (1972) result is,
however, extremely strong and would be a dominant test even on a perfor-
mance curve consisting only of results from the holdout sample.
The problem with these comparisons is that each study reported used
a different sample of bankrupt and non-bankrupt firms, with the result
that comparisons of performance may simply reflect different sample
characteristics rather than the actual .relative strength of the tests.
Table V-ll compares the means of the samples used in various studies
with the mean of the samples for the Agency primary and holdout samples.
As shown in the Table, the means of the key financial variables for the
Agency bankrupt firm sample are higher than those of firms used in other
studies, which indicates that the firms used by the Agency had greater
average financial strength. This difference could be expected to result
in higher Ep scores for the other tests if the Agency bankrupt samples
had been used in evaluating those other tests. The non-bankrupt sample
of firms shows more mixed results. In some cases the means of the EPA
samples show greater financial strength; in other cases, other studies
show greater financial strength. Thus no strong conclusion is possible
about how A™ for other studies would change if applied to the Agency
samples.
V-31
-------
TABLE V-10
ESTIMATED PERFORMANCE OF TESTS DEVELOPED
BY PRIOR BANKRUPTCY FORECASTING STUDIES
Study Name
Altaian (1968)-
Altman (1977)-
Deakin (1972)-
Deakin (1976)-
Type of Test
5-Variable MDA
7-Variable MDA
14-Variable MDA
5- Variable MDA
MB
17.0
15.1
8.0
NX
U.
w •
29.4
25.5
18.0
17.0
-SB
87.1
89.1
94.0
75.0
EP
5.8
5.0
3.0
5.0
— Performance based upon results reported in Altman (1977). No Holdout
Sample was used in Altman (1968), but this test was evaluated against
the new sample developed in Altman (1977). Note that Altman (1977)
Bankrupt Sample contains 5 firms which did not fail.
2/
— Primary Sample Performance from Altman (1977). No Holdout Sample
was used.
— Holdout Sample results are reported in Deakin (1972).
— M,, based on results reported for M for all firms passing the test
and entering bankruptcy from 1-3 years following passing the test.
V-32
-------
FIGURE V-7
LO
CO
100-
95-
90-
85-
80-
75-
• 151
UUII
70-
65-
60-
55-
50
45-
122
40-
35-
• 139
RESULTS OF OTHER STUDIES PLOTTED WITH EPA ONE-YEAR
PERFORMANCE CURVE AND HOLDOUT SAMPLE RESULTS
• 137
\» 1 36
D'72 •
• 120/ .
• A'77 /
A'68/ ,
146)
•135 • 149
• 134
(138)
• 150
(100)
Key: D172 - Deakin (1972)
D'76 - Deakin (1976)
A168 - Altman (1968)
A177 - Altman (1977)
143
30-,
.4
2.'0 2'.5 3.'o
5.0
5.S 6.'0 6*5
9*0 9 .'s lo'.O
-------
TABLE V-ll
COMPARISON OF SAMPLE MEANS FOR EPA AND OTHER
BANKRUPTCY FORECASTING STUDIES
A. Bankrupt Firms Samples;
f
Financial
CF/TL
TI./NW
CUKRAT
QRAT
NWK/TA
Due a Year
x-1
EPA
Primary
-.046
7.78
1.51
.77
.13
EPA
Holdout
-. 117
3.22
1.35
.67
.09
Deakin
(1972)
-.088
-51.00
.70
.35
-.21
Altman
(1968)
NA
26.78
1.33
NA
-.06
x-2
EPA
Primary
.017
3.67
1.58
.87
.18
EPA
Holdout
-.007
3.67
1.60
.77.
.22
Do akin
(1972)
-.052
10.66
.89
.48
-.07
A 1 tman
(1968)
NA
3.35
1.31
NA
.02
x-3
EPA
Primary
.049
1.84
1.97
1.07
.25
EPA
Holdout
.008
3.99
1.72
.90
.23
Deakin
(1972)
-.001
8.52
1.44
.53
. 18
Al ttoan
(1968)
NA
1.58
1.62
NA
.18
B. Non-Bankrupt Finn Samples:
Financial
CK/TL
TL/HU
ClIKKAT
QRAT
NWK/TA
x-l
EPA
Primary
.261
1.08
2.75
1.38
.30
EPA
Holdout
.239
.997
2.61
1.40
.32
Dc-akin
(1972)
.132
1.15
2.33
1.13
.33
Al troun
(1968)
NA
NA
NA
NA
.41
Al tman
(1977)
.314
NA
2.60
NA
NA
Data Year
man
»77)
314
JA
60
W
JA
x-2
EPA
Primary
.278
1.18
2.29
1.22
.29
EPA
Holdout
.172
1.07
2.26
1.26
.30
Deakin
(1972)
.150
.89
2.37
1.18
.33
EPA
Utilities
.114
1.41
1.26
.91
.02
x-3
EPA
Primary
.271
1.20
2.42
1.37
.28
EPA
Holdout
.232
1. 17
2.50
1.45
.32
Dunk in
(1972)
.115
.91
2.51
1.24
.35
EPA
Utilities
. 113
1.41
1. 11
.74
.01
SOURCE: Hoody's Industrial Manual and company annual Form 10-K reports for EPA samples; original articles cited for other samples.
-------
In order to provide better methods of comparing tests, Deakin
(1972) tests were evaluated using the Agency holdout sample using Ew
rather than Ep. The Agency used E^ because it required gathering only
one year rather than two years of additional data. The result of this
evaluation is shown in Figure V-8 and the results of the Deakin test are
summarized in tabular form in Table V-12. When both the Agency's and
Deakin1s tests are compared against the same sample of firms, the per-
formance is comparable. Seven of the Agency's tests fall below the
performance curve for Deakin1s test and seven are above it. Two of the
one-year eligibility tests are not comparable. (If the line connecting
the last two points representing Deakin's tests were extended, both of
these noncomparable .tests would be above this hypothetical performance
curve.)
V-35
-------
FIGURE V-8
COMPARISON OF EPA AND DEAKIN (1972) TESTS USING EPA HOLDOUT SAMPLE
f
to
NBH
122
40-
35-
139
137
• 100
Key: D1 - Deakin (1972)
146*
3t"b .5 llfl
2*.0
3.0
5.o
EW
7.*5 8.'o 8J5 9.'o
Ifl'.O
-------
TABLE V-12
PERFORMANCE OF DEAKIN 14-VARIABLE TEST
VERSUS EPA HOLDOUT SAMPLES*
Test Cutoff
Score
20.0
25.0
27.0
30.0
EPH
(x-3)
0
9.4
12.9
31.2
^SlBH
49
74
81
38
D
49
66
68
55
%
0
2.4
3.5
8.3
*G and Ep scores cannot be computed because only one year of
data, x-3, was analyzed against the Deakin test.
V-37
-------
VI. THE NEED FOR ALTERNATIVE TESTS FOR SPECIFIC INDUSTRIES
A number of commentsrs argued chat no single financial test could
be selected as an appropriate indicator of firm viability, because
financial ratios varied greatly among industries. Many of these same
commenters proposed that separate financial tests should be formulated
for each industry category. The Agency was generally opposed to such a
concept, because such a structure would greatly increase the administra-
tive complexity of the financial responsibility regulations. Initially,
it would be a difficult task to determine the number and type of cate-
gories that would be appropriate for this purpose. For many industries,
data on bankrupt firms would be so limited that the statistical validity
of a test derived from such data would be very questionable. Further-
more, even if a framework and set of tests could be established, many
large firms with diversified facilities would not fit easily into any
classification scheme. For example, if a parent firm primarily involved
in chemicals manufacturing attempted to use the financial test provision
to guarantee the waste disposal operations of its subsidiary, a pulp
mill, a decision rule would have to be developed on whether the firm
should be classified under the chemicals or paper category. The Agency
believes that the addition of such complexities could only be justified
if there is compelling evidence that the financial test alternatives
evaluated in this analysis all discriminated against specific
industries.
This Section examines the need for industry specific tests for
three categories of industry: manufacturing firms by two digit SIC
classification; utilities; and hazardous waste management firms.
VI-1
-------
A. Manufacturing Industries
To test the hypothesis that specific manufacturing industries may
be unable to pass a. general financial test, the Agency collected finan-
cial data for all the firms listed in the Expanded Coverage section of
the 1980 Moody's Industrial Manual (Volume 1). Using the Dun and
Bradstreet Million Dollar Directory, the Agency identified the 2-digit
SIC classifications corresponding to each firm's major manufacturing
activities. Table VI-1 shows the number of firms examined by 2-digit
SIC classification activities. The Agency then tested this sample
against the four financial ratios which consistently provided the best
classification performance in the initial evaluation of test alterna-
tives: cash flow/total liabilities, total liabilities/net worth, net
fixed assets/total assets, and current assets/current liabilities.
The results of this evaluation, as shown in Tables VI-2 through VI-
5, do not in any way support the argument for industry-specific tests.
Although there are some isolated differences between industries in cases
where stringent ratio cutoff points are assumed, each of the four ratios
has at least one commonly used cutoff value (NFA/TA > .2, CF/TL > .1,
TL/NW < 2.0, CURRAT > 1.5) which at least 75 percent of the firms in
each 2-digit SIC can pass. The discrimination among industry groups
resulting from the use of a single test therefore appears slight and
well within the boundaries of difference illustrated by other financial
indicators (e.g., bond ratings). Consequently, there is no justifi-
cation for requiring industry specific tests for manufacturing
industries.
VI-2
-------
TABLE VI-1
NUMBER OF FIRMS BY SIC CLASSIFICATION
USED IN TABLES VI-2 THROUGH VI-6
SIC
20
22
24
26
28
29
30
32
33
34
35
36
37
38
39
Description of Industry
Food and kindred products
Textile mill products
Lumber and wood products, except furniture
Paper and allied products
Chemicals and allied products
Petroleum refining and related industries
Rubber and miscellaneous plastics products
Stone, clay, glass, and concrete products
Primary metal industries
Fabricated metal products, except machinery
and transportation products
Machinery, except electrical
Electrical and electronic machinery, equipment
and supplies
Transportation equipment
Measuring, analyzing, and controlling instru-
ments; photographic, medical and optical
goods; watches and clocks
Miscellaneous manufacturing industries
Number
of Firms
33
8
15
17
51
13
29
19
26
44
65
38
33
20
14
VI-3
-------
TABLE VI-2
PERCENTAGES OF FIRMS IN MANUFACTURING INDUSTRIES PASSING OR FAILING
GIVEN CASH FLOW/TOTAL LIABILITY (CF/TL) REQUIREMENTS
(Based on Sample Described in Text)
SIC
20
22
24
26
28
29
30
32
33
34
.35
36
37
38
39
Description of Industry
Food and kindred products
Textile mill products
Lumber and wood products,
except furniture
Paper and allied products
Chemicals and allied products
Petroleum refining and
related industries
Rubber and miscellaneous
plastics products
Stone, clay, glass, and
concrete products
Primary metal industries
Fabricated metal products,
except machinery and
transportation products
Machinery, except electrical
Electrical and electronic
machinery, equipment, and
supplies
Transportation equipment
Measuring, analyzing, and
controlling instruments;
photographic, medical and
optical goods; watches and
clocks
Miscellaneous manufacturing
industries
CF/TL
< .10
9.1
18.7
13.3
-
9.8
7.7
3.4
10.5
3.8
6.8
7.7
10.5
9.1
-
7.1
>.1Q
90.9
81.3
86.7
100.0
90.2
92.3
96.6
89.5
96.2
. 93.2
92.3
89.5
90.9
100.0
92.9
>.15
72.7
62.5
73.3
88.2
72.5
76.9
75.9
84.2
61.5
68.2
72.3
81.6
66.7
75.0
64.3
>.20
48.5
50.0
60.0
64.7
47.1
53.8
55.2
68.4
42.3
40.9
50.8
57.9
24.2
55.0
57.1
VI-4
-------
TABLE VI-3
PERCENTAGE OF FIRMS IN MANUFACTURING INDUSTRIES PASSING OR FAILING
GIVEN NET FIXED ASSETS/TOTAL ASSETS (NFA/TA) REQUIREMENTS
(Based on Sample Described in Text)
SIC
20
22
24
26
28
29
30
32
33
34
35
36
37
38
39
Description of Industry
Food and kindred products
Textile mill products
Lumber and wood products,
except furniture
Paper and allied products
Chemicals and allied products
Petroleum refining and
related industries
Rubber and miscellaneous
plastics products
Stone, clay, glass, and
concrete products
Primary metal industries
Fabricated metal products,
except machinery and
transportation products
Machinery, except electrical
Electrical and electronic
machinery , equipment , and
supplies
Transportation equipment
Measuring, analyzing, and
controlling instruments;
photographic, medical and
optical goods; watches and
clocks
Miscellaneous manufacturing
industries
NFA/TA
< .20
9.1
-
13.3
-
7.8
-
10.3
10.5
3.8
13.6
18.5
21.1
12.1
20.0
28.6
>.20
90.9
100.0
86.7
100.0
92.2
100.0
89.7
89.5
96.2
86.4
81.5
78.9
87.9
80.0
71.4
>.25
75.8
93.8
73.3
100.0
80.4
92.3
89.7
78.9
84.6
72.7
70.8
71.1
75 .'8
65.0
50.0
>.'30
66.7
87.5
66.7
94.1
72.5
92.3
86.2
78.9
76.9
63.6
50.8
47.4
51.5
55.0
42.9
VI-5
-------
TABLE VI-4
PERCENTAGE OF FIRMS IN MANUFACTURING INDUSTRIES PASSING OR FAILING
GIVEN TOTAL LIABILITIES/NET WORTH (TL/NW) REQUIREMENTS
(Based on Sample Described in Text)
SIC
20
22
24
26
28
29
30
32
33
34
35
36
37
38
39
Description of Industry
Food and kindred products
Textile mill products
Lumber and wood products,
except furniture
Paper and allied products
Chemicals and allied products
Petroleum refining and
related industries
Rubber and miscellaneous
plastics products
Stone, clay, glass, and
concrete products
Primary metal industries
Fabricated metal products,
except machinery and
transportation products
Machinery, except electrical
Electrical and electronic
machinery, equipment, and
supplies
Transportation equipment
Measuring, analyzing, and
controlling instruments;
photographic, medical and
optical goods; watches and
clocks
Miscellaneous manufacturing
industries
TL/NW
>2.0
9.1
-
20.0
5.9
3.9
7.7
3.4
10.5
3.8
6.8
6.2
5.3
6.1
-
14.3
< 2.0
90.9
100.0
80.0
94.1
96.1
92.3
96.6
89.5
96.2
93.2
93.8
94.7
93.9
100.0
85.7
< 1.5
78.8
93.8
73.3
88.2
80.4
61.5
86.2
84.2
73.1
75.0
83.1
81.6
69.7
85.0
78.6
•<1.2
69.7
75.0
73.3
76.5
47.1
46.2
65.5
63.2
57.7
52.3
58.5
63.2
39.4
80.0
64.3
< 1.0
42.4
56.3
46.7
52.9
31.4
15.4
48.3
57.9
38.5
36.4
46.2
44.7
24.2
40.0
57.1
VI-6
-------
TABLE VI-5
PERCENTAGE' OF FIRMS IN MANUFACTURING INDUSTRIES PASSING
OR FAILING GIVEN CURRENT RATIO (CURRAT) REQUIREMENTS
(Based on Sample Described in Text)
SIC
20
22
24
26
28
29
30
32
33
34
35
36
37
38
39
Description of Industry
Food and kindred products
Textile mill products
Lumber and wood products, except
furniture
Paper and allied products
Chemicals and allied products
Petroleum refining and related
industries
Rubber and miscellaneous plastics
products
Stone, clay, glass, and concrete
products
Primary metal industries
Fabricated metal products, except
machinery and transportation
products
Machinery, except electrical
Electrical and electronic
machinery , equipment , and
supplies
Transportation equipment
Measuring, analyzing, and control-
ling instruments; photographic,
medical and optical goods;
watches and clocks
Miscellaneous manufacturing
industries
CURRAT
< 1.5
6.1
6.2
6.7
5.9
9.8
23.1
3.4 .
5.3
3.8
6.8
9.2
13.2
15.2
>1.5
93.9
93.8
93.3
94.1
90.2
76.9
96.6
94.7
96.2
93.2
90.8
86.8
84.8
100.0
100.0
>2.0
48.5
68.8
53.3
70.6
56.9
15.4
69.0
68.4
53.8
68.2
58.5
50.0
54.5
70.0
64.3
VI-7
-------
B. Electric Utilities
In their comments on the proposed financial test of May 19, 1980,
firms in the electric utility industry asserted that utilities could not
be analyzed using "normal" financial logic, and should be judged under
alternative criteria. Utilities objected to the use of net working
capital as a required component of a financial test, since many firms in
each category operate regularly from a negative net working capital
position. Both the data submitted by the commenters and an independent
check by the Agency verified that utilities do operate with negative net
working capital. As a result, any test which requires that net working
capital be some multiple of financial responsibility obligations will
fail most utilities.
The Agency also examined the question of the extent to which
specific ratio requirements might consistently fail viable utilities.
To examine this issue, a sample of 26 electric utilities were drawn from
the 1978 and 1979 Moody's Public Utilities Manual (see Table VI-6).
Table VI-7 illustrates the performance of this sample against the key
single-ratio tests identified in Section IV. As can be seen, a number
of these ratios are totally unsuited for analyzing utility viability.
The following financial ratios and pass-fail cutoff points were iden-
tified as potentially acceptable tests:
CF/TS > .1
TL/NW < 2.0, < 1.5
NI/TA > .02, > .04
NFA/TA > .2, > .25, > .3
VI-8
-------
TABLE VT-6
FINANCIAL DATA FOR THE ELECTRIC UTILITY SAMPLE
Maim: of
Firm
Hlai:k III UK
I'nwtT .mil l.lp.lit
lloHton Kill Him
Carol Ilia 1'ower
anil Unlit
iA'iitral Hudson
P
P
F
K
F
F
F
F
• I' III.II (p.-iiisns .-ill LCKts); F '- 2O.O (l.illa nil teals)
21V • 2(1.0 (II.-IK.SCH all ti:HlH); F • 10.0 (falls all tuKtu)
I1 1.0 (II.-IHSUK all lasts); F > 2.0 (falls all tests)
- P 2.0 (|>a:;s.'H all testa); F
V,.
1.2 (passes all tests); F
1.5 (falls all tests)
I.0 (lalIs al1 tests)
-------
TABUS VI-6 (Continued)
Nami- ul
!•' 1 riu
I'lorlila I'uwi.-r ami l.l|;lil:
1 :<•!!,• r a 1 I'ulil lc
Ul 1 1 lly Corp.
01,1,. lull son
IV.uusy 1 van la I'owor
ami I.I t;lit
IVnplirs lila KliKtrlc
C.mi|Mny
I'lfdmoiiL Nat u nil lias
I'url laml CentT.il
lil.tclrl.-
I'oLtimai: Flecirli:
l'uw<:| til.
PiiMIc Sitrvlci; Ci>. uf
Cu l<> ratio
Public Sm'vlcti I'o. ol
1 mil ana
Pill, II.- S.-rvli:,.. <:,.. of
Hi-w llam|i.slilri:
Y.:ar
1977(x-l)
l976(x-2)
1975(x-3)
1977(x-l)
1976(x-2)
1975(x-3)
1977(x-l)
1976(x-2)
!975(x-3)
1978(x-l)
1977(x-2)
1976(x-3)
1978(x-l)
1977(x-2)
1976(x-3)
1978(x-l)
1977(x-2)
1976(x-3)
1978
-------
TAIil.li VI-6
N.-IIIH! of
Hriu
San l)frt;o (>:IH
anj K|i;t:lrlr
Saviiniuili Klurirlt:
Power i.'u.
Ye.ir
I.978U-1)
1977(x-2)
1976(x-3)
1978(x-l)
1977(x-2)
1976(x-3)
Nl
66.8
60.2
50.5
11.4
8.7
6.4
C.F
1.02 . 4
92.8
80.5
17.4
14.3
11.2
QA
NA
NA -
NA
NA
NA
NA
CA
199.0
187.0
146.0
15.8
11.1
10.8
Cl.
163.0
213.0
181.0
16.7
11.0
8.7
1
86(
83
701
16'
13'
12
Tl.
860.0
831.0
706.0
169.0
134.0
127.0
NW
694.0
582.0
481.0
79.0
73.0
63.0
TA
1554.0
1413.0
1187.0
248.0'
207.0
190.0
NFA/TAi/
P
P
P
P
P •
P
CK/T,,^
11.9
11.2
11.4
10.3
10.7
f
TI./NW-'
1.24
1.43
1.47
K
1.84
K
CUKKAT-
f
K
K
F
K
K
(JRAT-
NA
K
K
K
NA
NA
<5
M
-------
TABLE VI-7
APPLICABILITY OF FINANCIAL TESTS TO UTILITIES
, (26 Firm Samples)
1
1
Single
CF/TL
CF/TL
CF/TL
TL/NW
TL/NW
TL/NW
TL/NW
NI/TA
NI/TA
NWK/TA
CURRAT
CURRAT
QRAT
QRAT
NFA/TA
NFA/TA
Ratio Tests
> .1
> .15
> .2
< 2.0
< 1.5
< 1.2
< 1.0
> .02
> .25
> .25
> 1.5
> 2.0
> 1.0
> 1.2
> .25
> .3
w
(One- Year
Eligibility
Requirement)
77
4
0
81
46
12
0
100
54 '
0
12
4
8
4
100
100
^JBU
(Three-Year
Eligibility
Requirement)
54
4
0
77
38
4
0
100
46
0
8
4
4
4
100
100
I/
A™.. = Percent of viable utilities passing a given test.
VI-12
-------
Given this reduced list of acceptable parameters, only 28 of the
151 multi-ratio tests were classified as applicable to utilities. These
tests were then evaluated against the utility sample, with the results
summarized in Table VI-8 and VI-9. For many of the higher Ep score
values, the tests that can be effectively applied to utilities are also
those whi.ch are most effective for other industries. If, however, a
greater than 90 percent reduction in the overall rate of large firm
bankruptcy is sought, the tests which permit an acceptable percentage of
utilities to pass also reject a higher percentage of other viable firms.
This disparity is illustrated in Figure VI-1, which contrasts one-year
tests on the performance curve with utility applicable tests, using only
tests applicable to utilities derived in Section V.
C. Hazardous Waste Management Firms
Because the majority of firms currently in the hazardous waste
management business are either privately held or single-site entities
with assets less than $1 million (Foster D. Snell, 1976), it was not
possible to assemble a sample of hazardous waste management firms which
could be used to evaluate test applicability.
Table VT-10 shows average ratios for hazardous waste management
firms from 1971-1975. These ratios suggest that the average hazardous
waste management firm would have little difficulty with possible net
fixed assets to total assets requirements. However the average firm in
this industry would have had trouble in many of these years with pos-
sible current ratio and total liability to net worth requirements. The
average profits to total debt ratio suggests an uncertain picture with
VI-13
-------
TABLE VI-8
OVERALL EFFECTIVENESS OF UTILITY-APPLICABLE
FINANCIAL TESTS
(One—Year Eligibility Requirement)
Test
Number
11
12
15
16
19
20
97
98
99
100
109
110
Test
Description
Pass All:
CF/TL > .1
TL/NW < 1.5
CF/TL > . 1
TL/NW < 2.0
NI/TA > .04
TL/NW < 1.5
NI/TA > .04
TL/NW < 2.0
NI/TA > .02
TL/NW < 1.5
NI/TA > .02
TL/NW < 2.0
2 out of 3:
CF/TL > .1
TL/NW < 1.5
CURRAT > 2.0
CF/TL > .1
TL/NW < 2.0
CURRAT > 2.0
CF/TL > .1
TL/NW < 1.5
CURRAT > 1.5
CF/TL > .1
TL/NW < 2.0
CURRAT > 1.5
CF/TL > .1
TL/NW < 1.5
QRAT > 1.2
CF/TL > .1
TL/NW < 2.0
QRAT > 1.2
EP
6.1
6.8
6.3
7.3
6.5
7.4
9.7
9.3
10.1
10.1
8.3
8.4
*NB
77
81
66
72
77
82
86
93
91
96
82
87
\BU
42
69
31
46
46
81
42
69
42
69
42
69
71-14
-------
TABLE VI-8 (Continued)
Test
Number
119
120
123
124
125
126
127
130
133
134
135
Test
Description
CF/TL > .1
TL/NW < 1.5
QRAT > 1.0
CF/TL > . 1
TL/NW < .2.0
QRAT > 1.0
Pass All:
NFA/TA > . 3
CF/TL > . 1
TL/NW < 1.5
NFA/TA > . 3
CF/TL > .1
TL/NW < 2.0
NFA/TA > . 25
CF/TL > .1
TL/NW < 1.5
NFA/TA > .25
CF/TL > .1
TL/NW < 2.0
NFA/TA > . 2
CF/TL > .1
TL/NW < 2.0
2 out of 3:
NFA/TA > . 3
CF/TL > .2
TL/NW < 1.5
HFA/TA > . 3
CF/TL > .15
TL/NW < 1.5
NEA/TA > . 3
CF/TL > . 1
TL/NW < 1.0
NFA/TA > . 3
CF/TL > . 1
TL/NW < 1.2
EP
9.5
9.1
3.3
3.1
2.9
3.6
3.8
7.1
6.8
4.6
5.6
SB
88
92
55
59
62
66
75
73
80
79
83
SBU
42
69
42
69
42
69
69
46
46
77
77
"1-15
-------
TABLE VI-S (Concluded)
Test
Number
136
137
146
149
150
Test
Description
NFA/TA > .3
CF/TL > .1
TL/NW < 1.5
NFA/TA > . 3
CF/TL > .1
TL/NW < 2.0
NFA/TA > . 2
CF/TL > .1
CURRAT > 2.0
3 out of 4:
NFA/TA > . 3
CF/TL > .1
CURRAT > 2.0
TL/NW < 1.5
NFA/TA > . 3
CF/TL > .1
CURRAT > 2.0
TL/NW < 2.0
EP
7.9
9.1
9.9
6.0
5.5
\B
89
92
95
87
81
NBU
81
88
77
69
42
VI-16
-------
TABLE VI-9
OVERALL EFFECTIVENESS OF UTILITY-APPLICABLE
FINANCIAL TESTS
(Three-Year Eligibility Requirement)
Test
Number
11
12
15
16
19
20
97
98
99
100
109
110
Test
Description
Pass All:
CF/TL > .1
TL/NW < 1.5
CF/TL > .1
TL/NW < 2.0
NI/TA > .04 .
TL/NW < 1.5
NI/TA > .04
TL/NW < 2.0
NI/TA > .02
TL/NW < 1.5
NI/TA > .02
TL/NW < 2.0
2 out of 3:
CF/TL > .1
TL/NW < 1.5
CURRAT > 2.0
CF/TL > .1
TL/NW < 2.0
CURRAT > 2.0
CF/TL > .1
TL/NW < 1.5
CURRAT > 1.5
CF/TL > .1
TL/NW < 2.0
CURRAT > 1.5
CF/TL > .1
TL/NW < 1.5
QRAT > 1.2
CF/TL > .1
TL/NW < 2.0
QRAT > 1.2
%
3.7
5.2
4.7
6.1
4.9
6.6
7.7
8.6
8.2
9.7
6.5
7.4
^B
70
76
56
60
69
75
75
82
80
84
72
78
\BU
35
58
<20
23
38
77
35
58
35
58
35
58
VI-17
-------
TABLE VI-9 (Continued)
1
Test
Number
119
120
123
124
125
126
127
130
133
134
135
Test
Description
CF/TL > .1
TL/NW < 1.5
QRAT > 1.0
CF/TL > .1
TL/NW < 2.0
QRAT > 1.0
Pass All:
NFA/TA > . 3
CF/TL > .1
TL/NW < 1.5
NFA/TA > . 3
CF/TL > .1
TL/NW < 2.0
NFA/TA > .25
CF/TL > .1
TL/NW < 1.5
NFA/TA > .25
CF/TL > .1
TL/NW < 2.0
NFA/TA > . 2
CF/TL > .1
TL/NW < 2.0
2 out of 3:
NFA/TA > .3
CF/TL > .2
TL/NW < 1.5
NFA/TA > .3
CF/TL > .15
TL/NW < 1.5
NEA/TA > . 3
CF/TL > . 1
TL/NW < 1.0
NFA/TA > . 3
CF/TL > . 1
TL/NW < 1.2
EP
8.0
8.2
1.6
1.5
1.4
2.2
1.9
4.9
4.7
2.9
4.1
V
78
83
49
53
56
59
67
64
72
72
77
\BU
35
58
35
58
35
58
58
38
42
62
62
VI-18
-------
TABLE VI-9 (Concluded)
Test
Number
136
137
146
149
150
Test
Description
NFA/TA > .3
CF/TL > .1
TL/NW < 1.5
NFA/TA > .3
CF/TL > .1
TL/NW < 2.0
NFA/TA > . 2
CF/TL > .1
CURRAT > 2.0
3 out of 4:
NFA/TA > . 3
CF/TL > .1
CURRAT > 2.0
TL/NW < 1.5
NFA/TA > .3 .
CF/TL > .1
CURRAT > 2.0
TL/NW < 2.0
EP
5.4
7.3
9.5
5.0
4.1
\B
82
86
83
74
71
\BU
65
81
62
58
35
VI-19
-------
FIGURE VI-1
PERFORMANCE CURVE FOR TESTS WITH ONE-YEAR ELIGIBILITY REQUIREMENTS
WITH AND WITHOUT UTILITY APPLICABLE TESTS
NB
Key: Dark Line - Utility Applicable
Tests
Light Line - One-Year Tests
40-
35-
O A
0
.4
!5 2.'0 2*.5 3.'o
!o 4.*5 5.'o S.'s 6.'o 6,'s ?!o 7.'s s!o 8.*5 9.'o 9 .'s lo'.O id. 5
c.p
-------
TABLE VI-10
AVERAGE RATIOS OF HAZARDOUS WASTE MANAGEMENT FIRMS, 1971-1975
1971 1972 1973 1974 1975
Current Ratio
a. Firms whose main objective is hazardous 1.58 1.77 1.42 1.33 1.28
waste management
b. Corporations having divisions or sites 1.28 1.61 1.81 2.04 1.83
engaged in hazardous waste management
Total Debt/Equity
a. Firms whose main objective is hazardous 1.20 1.19 1.18 1.60 1.64
waste management
b. Corporations having divisions or sites 1.37 1.66 2.43 1.96 NA
M engaged in hazardous waste management
*" Profits/Total Debt
a. Firms whose main objective is hazardous NA .096 .102 .134 .182
waste management
b. Corporations having divisions or sites .127 .098 .056 .076 .075
engaged in hazardous waste management
Net Fixed Assets/Total Assets
a. Firms whose main objective is hazardous .65 .64 .62 .555 .60
waste management
b. Corporations having divisions or sites .62 .57 .63 .60 .62
engaged in hazardous waste management
SOURCE: Foster D. Snell, 1976.
-------
respect to the ability of the average firm in the industry to meet the
possible cash flow/total liability requirements. This ability would
depend in at least some years on total depreciation, amortization and
depletion allowance.
Thus, firms in the hazardous waste management industry could be
expected to have difficulty meeting many possible financial tests even
without a $10 million in net worth requirement. However, this reflects
the financial weakness of this industry during this period rather than
some unique financial feature of the industry. The study from which the
ratios in Table VI-10 are drawn states:
Since 1973, long term debt has remained stable for the
industry as financial institutions have been reluctant to
invest capital in the industry because: hazardous waste
management firms in general did not have strong financial
positions, and the growth of the industry was deemed risky in
future periods. Short debt financing has increased from $3.2
million in 1971 to $7.2 million in 1975. This high cost debt
financing additionally points out the current problems of the
industry in obtaining lower cost long-term capital. (Foster
D. Snell, 1976).
The share of hazardous waste facilities owned by hazardous waste
management firms with over $10 million in net worth and independently
audited is small. Only 26 facilities are owned by the four largest
hazardous waste management firms which are independently audited. The
next eight to nine largest firms are all privately held and are unlikely
to be independently audited. The majority of off-site facilities are
owned by relatively small firms with less than $10 million in net worth
(Booz, Allen & Hamilton and Putnam, Hayes & Bartlett, 1980). The 26
facilities which are owned by large, independently audited firms thus
represent only about 1 percent of the 2,500 facilities owned by firms
eligible for a financial test.
71-22
-------
A careful examination of the balance sheets of two of the larger
firms in the industry which own 12 of the 26 facilities, Browning Ferris
Industries (BFI) and SCA Services, Inc. showed their performance with
respect to the relevent ratios as summarized in Table VI-11. BFI would
pass 13 of the 16 financial tests on the performance curve for tests
with one-year or three-year elibibility requirements. SCA would pass 12
of these 16 tests.
VI-23
-------
TABLE VI-11
RATIOS FOR BROWNING FERRIS INDUSTRIES (BFI)
AND SCA SERVICES, INC.
Firm Name
BFI
SCA
Year of Data
(x-1) 1978
(x-2) 1977
(x-3) 1976
(x-1) 1978
(x-2) 1977
CF/TL
17.5
15.5
15.7
15.4
14.9
TL/NW
1.13
1.10
1.12
1.52
1.55
NFA/TA
^
P
P
P
P
CURRAT
1.60
F*'
1.64
1.65
F
QRAT
P
P
P
P
P
NI/TA
P
P
P
4.0
3.0
— P = Passes all tests using this variable.
2/
— F = Fails all tests using this variable.
VI-24
-------
VII. BOND RATINGS
Several commenters on the May 19, 1980 reproposed financial respon-
sibility regulations suggested that corporate bond ratings should be
used as an alternative or substitute for the proposed financial test.
The Agency therefore analyzed the possibility of using corporate bond
ratings as an assurance of financial responsibility.
Bonds are a form of long-term debt. They may be unsecured, or
secured by collateral such as equipment, marketable securities, or fixed
assets. Certain bonds include specific promises concerning the use of
special sinking funds. Repayment of other categories of bonds may be
subordinated to repayment of other debt instruments.
Bond ratings provide an appraisal of the ability of firms to repay
these long-term debts. They are used extensively by the financial
community for the purpose of providing a long-term credit risk evalu-
ation of the relative likelihood of timely payment of interest and
repayment of principal of specific bond instruments. Many banks,
including large ones, rely almost exclusively on rating service evalu-
ations of bonds and do not have internal groups which independently
evaluate the riskiness of their bond portfolios (Kaplan and Urwitz,
1979). Because bond ratings measure the ability of firms to meet a
long-term debt, they are better adapted to an analysis of long-term
financial responsibility than are ratings of short-term credit and trade
credit.
The Agency did not attempt to evaluate the performance of ratings
of specific types of bonds. Rather, it considered the ratings
VII-1
-------
performance of all categories of long-term debt that were (1) identified
as bonds by either of the two major bond ratings services and (2) rated
by either of them.
The two major ratings services are Moody's Investors Services, Inc.
(Moody's) and Standard and Poor's Corporation (Standard and Poor's).
Each has long experience in rating corporate bonds. Moody's and
Standard and Poor's rate almost all corporate bond issues of $ 5 million
and over in principal amount with or without request from the issuer.
The bonds of certain issuers (including newly created companies in most
cases) are not rated, and ratings of other issuances are carried out
only upon request and are not published except upon permission of the
issuer.
Both major ratings services provide nine basic ratings of corporate
bonds. Figure VII-1 provides a listing of the ratings given by Moody's
and the interpretation of each rating as suggested by Moody's. Figure
VII-2 provides the ratings and their interpretations used by Standard
and Poor's. The ratings classifications are highly qualitative, and
neither rating service suggests any specific quantitative equivalent for
the ratings assigned.
The top four ratings of both services are assigned to bonds consid-
ered to be of investment grade, while the remaining ratings are assigned
to bonds that are considered to be speculative. Ratings of each bond
issuance are reviewed periodically and may be upgraded or downgraded.
Table VII-1 provides a distribution of corporate bond ratings issued by
VII-2
-------
Aaa
Bonds which are rated Aaa are judged to be of the hest
quality. They carry the smallest degree of investment risk
and are generally referred to as "gilt edge." Interest payments
are protected by a large or by an exceptionally stable margin
and principal is secure. While the various protective elements
are likely to change, such changes as can be visualized are
must unlikely to impair the fundamentally strong position
of such issues.
Aa
Bonds which are rated Aa are Judged to be of high quality
by all standards. Together with the Aaa group they comprise
what are generally known as high grade bonds. They are
rated lower than the best bonds because margins of protec-
tion may not be as large as in Aaa securities or fluctuation
or protective elements may be of greater amplitude or there
may be other elements present which make the long term
risks appear somewhat larger than in Aaa securities.
any great length of time. Such bonds lack outstanding invest-
ment characteristics and in fact have speculative characteristics
as well.
Ba
Bonds which are rated Ba are judged to have speculative
elements; their future cannot be considered us well assured.
Often the protection of interest and principal payments may
be very moderate and thereby not well safeguarded during both
good and bad times over the future. Uncertainty of position
characterizes bonds in this class.
B
Bonds which are rated B generally lack characteristics of the
desirable investment. Assurance of interest and principal payments
or of maintenance of other terms of the contract over any long
period of time may be small.
Bonds which are rated A possess many favorable invest-
ment attributes and are to be considered as upper medium
grade obligations. Factors giving security to principal and
interest are considered adequate but elements may be present
which suggest a susceptibility to impairment sometime in the
future.
Baa
Bonds which are rated Baa are considered as medium
grade obligations, i.e., they are neither highly protected nor
poorly secured. Interest payments and principal security ap-
pear adequate for the present but certain protective elements
may be lacking or may be characteristically unreliable over
Caa
Bonds which are rated Caa are of pour standing. Such issues
may be in default or there may be present elements of danger with
respect to principal or interest.
Ca
Bonds which are rated Ca represent obligations which are
speculative in a high degree. Such Issues arc often in default
or have other marked shortcomings.
C
Bonds which are rated C are the lowest rated class of bonds and
issues so rated can be regarded as having extremely poor prospects of
ever attaining any real investment standing.
FIGURE VII-1
KEY TO MOODY'S BOND RATINGS
-------
H
M
A Standard & Poor's corporate or municipal bond rating is a current assessment of
the creditworlhiness of an obligor with respect to a specific ilcbi obligation. I'll is
assessment may lake into consideration obligors sncli as guarantors, insurers, or
lessees.
The bund rating is not a recommendation to purchase, sell, or hold a security
inasmuch as it does not comment as to market price or suitability for a particular
investor.
The ratings are based on current information furnished by the issuer or obtained
by Standard £ I'oor's from other sources we consider reliable. We do not perform an
audit in connection with any rating and may, on occasion, rely on unaudited financial
information. The ratings may be changed, suspended, or withdrawn as a result of
changes in, or unavailability of, such information, or for other reasons.
The ratings arc based, in varying degrees, on the following considerations:
I. Ukelihood of default—capacity and willingness of the obligor as to
the timely payment of interest and repayment of principal in accor-
dance with the terms of the obligation
II. Nature of and provisions of the obligation.
HI. Protection afforded by, and relative position of, the obligation in the
event of bankruptcy, reorganization, or other arrangement under
the laws of bankruptcy and other laws a flea ing creditor's rights
Bonds rated AAA have the highest rating assigned by Standard & I'oor's to
AAA a debt obligation. Capacity to pay interest and repay principal is extremely
strong.
Bonds rated AA have a very strong capacity to pay interest and repay prin-
cipal and differ from the highest-rated issues only in a small degree.
Bonds rated A have a strong capacity to pay interest and repay principal,
although they are somewhat more susceptible to the adverse effects of
changes in circumstances and economic conditions than bonds in higher-
rated categories.
Bonds rated BBB are regarded as having an adequate capacity to pay interest
and repay principal. Although they normally exhibit adequate protection
parameters, adverse economic conditions or changing circumstances are
more likely to lead to a weakened capacity to pay interest and repay principal
for bonds in this category than for bonds in higher-rated categories.
AA
BB
B
CCC
CC
C
D
NR
BBB
Bonds rated UB, B. CCC. and CC arc regarded, on balance, as predomj.
nanlly s|>eculalive with respect to capacity to pay interest and repay principal
in accordance with the terms of the obligation. UU indicates (lie lowest degree
of S|x.-cnlation and CC the highest degree of s|x:culaiion. While such bond*
will likely have some quality and protective characteristics, these are out-
weighed by large uncertainties or major risk exposures to adverse conditions.
The rating C is reserved for income lionds on which no interest is being paid.
Bonds rated O are in default, and payment of interest and/or repayment of
principal is in arrears.
Plus (+) or Minus (-): The ratings from AA to UB may lie modified by the
addition of a plus or minus sign to show relative standing within the major
rating categories.
Provisional Ratings: The letter "p" indicates that the rating is provisional.
A provisional rating assumes the successful completion oi the project
financed by the bonds being rated and indicates that payment of debt service
requirements is largely or entirely dependent upon the successful and lime!)
completion of the project. This rating, however, while addressing credit qual-
ity subsequent to completion of the project, makes no comment on the like-
lihood of, or the risk of default UJKIII failure of, such completion. The inves-
tor should exercise his or her own judgment with respect to such likelihood
and risk.
Indicates that no rating has been requested, thai there is insufficient infor-
mation on which to base a rating, or that Standard & Poor's does not rate a
particular tyjx: of obligation as a matter of policy.
Debt obligations of issuers outside the United Stales and its territories are
rated on the same basis as domestic corporate and municipal issues. The
ratings measure the credilworlhiness of the obligor, but do not take into
account currency exchange and other unccriaiiuies.
Bond Investment Quality Standards: Under present commercial bank reg-
ulations issued by the Comptroller of the Currency, lx>nds rated in the top
four categories (AAA. AA. A, BBB, commonly known as "investment grade"
ratings) are generally regarded as eligible for bank investment. In addition,
the legal investment laws of various stales impose certain ratings or other
standards for obligations eligible for investment by savings banks, trust com-
panies, insurance companies, and fiduciaries generally.
FIGURE VII-2
STANDARD & POOR'S CORPORATE RATING DEFINITIONS
-------
TABLE VII-1
PERCENTAGE DISTRIBUTION OF PUBLICLY OFFERED
STRAIGHT CORPORATE BONDS
Amount
in billions
Aaa
A
Baa
Not
& Aa
& below
rated
1966
7.8
48%
19
27
6
1967
14.8
42%
19
30
9
1968
10.9
48%
17
23
12
1969
9.5
53%
27
20
0
1970
23.1
50%
38
11
1
1971
21.3
53%
32
11
1
1972
12.3
60%
28
8
4
1973
12.3
61%
29
5
5
1974
24.8
64%
29
6
1
1975-
31.1
57%
36
7
3
I/
— 11 months.
SOURCE: Backer and Gosman (1978).
VII-5
-------
Moody1 s for recent years. Table VII-2 shows the number of outstanding
bond issues rated by Moody's at a point in time in July 1981. According
to this Table, 86 percent of the outstanding bond issues rated by
Moody's as of that time would be acceptable for satisfying the bond
rating portion of the financial test.
Both ratings services insist that a bond rating is based on the
specific bond instrument in question, and does not imply a rating for
other instruments of the firm or a rating of the firm as a whole. This
is due, in part, to the fact that different bonds are secured by dif-
ferent types and amounts of collateral, and by different types of other
so-called "structural credit supports" such as third party guarantees,
segregated cash flow, or insurance. However, the rating agencies do
carry out an analysis of the situation of both the industry and of the
company within the industry and a financial analysis of the company, and
these analyses also affect the rating assigned to the specific bond
instrument.
According to Standard and Poor's Ratings Guide, ratings are based,
to varying degrees, on the following considerations:
I. Likelihood of default—capacity and willingness of the
obligor as to the timely payment of interest and
repayment of principal in accordance with the terms of
the obligation
II. Nature of and provisions of the obligation
III. Protection afforded by, and relative position of, the
obligation in the event of bankruptcy, reorganization,
or other arrangement under the laws of bankruptcy and
other laws affecting creditor's rights
(Standard and Poor's Ratings Guide, 1979)
VII-6
-------
TABLE VII-2
OUTSTANDING BOND ISSUES*
JULY 1981
Moody 's Ratings
Investment Grade
Aaa
Aa
A
Baa
Total Investment Grade
Below Investment Grade
Ba
B
Caa
Ca
C
Total Below Investment Grade
Number
of Issues
750
1140
2029
1012
4931
339
320
95
38
10
802
Percentage
of Issues
13
20
35
18
86
6
6
2
1
0
15
*Includes: Industrials, Utilities, Transportation, Banks and Finance
Companies.
Note: Totals may exceed 1013% due. to rounding error.
SOURCE: Ron Lewis of Moody's Corporate Bond Department.
VII-7
-------
The wide scope of bond rating agency analysis can be better
appreciated from the description given by Brenton W. Harris, a former
President of Standard and Poor's, when listing the factors used as a
basis for assigning bond ratings:
1. Issuing documents: ... In determining a rating, the
indenture is far less important than the company's
earning power, financial resources, and property protec-
tion. This is not to say, however, that the indenture
does not have a great bearing on a bond rating...
2. Earnings: the past record and foreseeable potential are,
in most cases, the single most important factor in credit
rating. High levels of earnings frequently preclude
liquidity problems because access to short-term cash
needs can be readily accommodated. Remembering that bond
ratings turn on timely repayment of principal and
interest, strong cash flows generated by high and con-
tinued earnings, combined with an adequate depreciation,
depletion, and amortization policy where applicable,
contribute a healthy plus factor to the determination of
a bond rating...
3. Asset protection: Asset protection generally is more
important as a long-term consideration than as one
influencing immediate liquidity. The analysis here is
primarily statistical and, hence, highly objective. Of
primary interest are the ratio of its working capital to
its debt; the ratio of its debt to its equity; and the
ratio of its total net tangible assets to its debt. The
relative importance of these major ratios depends on the
type of industry the company is in...
4. Management: Evaluating management is one of the most
difficullt chores a rating agency faces. But management
is one of the most important facets of a successful
operation. It is my opinion that the single most impor-
tant reason for the failure of the Penn Central was the
inability of its management to deal with its problems...
5. Financial resources are, of course, the largest single
area in which liquidity has a direct impact on long-term
debt rating. In looking at the financial resources of a.
company, we are concerned not only with a company's cash
position but also with its ability to obtain cash. This
area of financial resources, which constitutes one of the
five fundamental areas of investigation to determine
VII-8
-------
long-term debt rating, is the test of liquidity ...
Briefly stated, the financial resources we consider are
those alternative sources of borrowing that a firm may
use to raise cash for either long- or short-term debt
repayment. They include the amount of cash reserves on
hand including salable receivables; the short-term
borrowing potential, particularly bank lines; the ability
of. the company to tap the long-term debt market, partic-
ularly at the time of its choosing; the ease with which
the company could sell stock; and the potential sale of
assets—obviously the weakest alternative.
(Quoted in Backer and Gosman, 1978)
The Agency is satisfied that the actual establishment of a rating
is the result of the informed judgment of trained analysts based on a
detailed study of all of the above factors, and does not involve the
mechanical application of formulae. Bond ratings are therefore a dif-
ferent type of analysis of financial liability than a financial test.
Spokesmen for the ratings services emphasize the point that bond ratings
are based on expert judgment (Backer and Gosman, 1978). Although some
investigators have questioned this assertion (Ross 1976, Sherwood 1976),
arguing that bond ratings services rely on explicit cutoff points of
particular financial ratios, a thorough recent study of bond and other
financial ratings procedures prepared for the National Association of
Accountants reported, with respect to the importance of financial ratios
in credit rating evaluations, that:
All interviewees stated the belief that financial ratios were
of moderate to strong importance in their credit rating evalu-
ations. However, almost all those interviewed cautioned that
such financial measures were not the only thing of importance;
additional considerations frequently mentioned included
quality of management, future for product, and general
economic conditions. (Backer and Gosman, 1978).
VII-9
-------
The same study conducted a statistical analysis to determine which
ratios seemed to be most important by comparing a group of bonds that
had not been downgraded and group of bonds that had been downgraded.
The most statistically significant ratio was found to be cash flow to
long-term debt. However, the statistical results also indicated that a
number of other ratios could also be of some importance: return on
sales, return on total assets, return on tangible net worth, long-term
debt to capitalization, net tangible assets to long-term debt, long-term
debt to property and equipment, and cash flow to total debt. This
statistical examination of downgraded bonds compared to bonds that were
not downgraded would probably have revealed a specific formula had one
existed. No such specific formula was found.
The Agency, in order to assess the actual performance of bond
ratings in predicting default, first conducted a literature search for
previous studies of performance. A variety of sources suggest that bond
ratings can be used as a reasonably good indicator of the quality of
corporate debt. (See Lev, 1974; Fraine and Mills, 1961; and Fiedler,
1971.) The only sources for detailed statistical data supporting this
conclusion, however, are two studies by W.B. Hickman, Corporate Bond
Quality and Investor Experience (1958), and Statistical Measures of
Corporate Bond Financing Since 1900 (1960).
Although Hickman1s work is somewhat dated, covering the period
1900-1943, it does examine a period of economic volatility which
provides a good test of the results of ratings services. The ratings
used are those assigned at the time of issue. Thus, if the bond was
VII-10
-------
downgraded prior to default but after issue, this would not be reflected
in these ratings. Given the volatility of the economy for the period in
question, and the fact that changes in ratings are not accounted for use
of these data results in a strict judgment of bond ratings. In addi-
tion, statistical analysis of the performance of bond ratings is
difficult for recent periods because so few firms with rated bonds have
gone bankrupt that creation of a useful sample is almost impossible.
The Agency noted, for example, that none of the firms in its sample of
bankrupt firms between 1966 and 1979 had an investment grade rated bond
issuance.
In a study incorporating Hickman's findings, T.B. Atkinson compared
prewar and postwar (through 1965) corporate bond quality. He concluded
that:
U.S. corporate bonds defaulting in the postwar period (from
1945 through 1965) averaged less than 0.1 percent of the
volume of outstanding, or about one-half billion dollars in
all. This compares with 1.7 percent of the outstanding bonds
which defaulted during 1900-43. The postwar defaults were
concentrated in bonds of the railroad industry, many of which
had been outstanding before 1920 or had been refundings of the
original bonds issued prior to that year. (Atkinson, 1967)
It should be noted that these percentages of outstanding bonds which
defaulted include bonds of investment and speculative quality.
In order to analyze the effectiveness of agency ratings in
predicting defaults, the Agency applied a general procedure similar to
that used to calculate the effectiveness (E) of various financial tests.
Hickman reported in his study the percentage of the total value of
outstanding bonds of a particular rating defaulted upon by firms with
VII-11
-------
these bonds and the percentage of the total value not defaulted.
Although there is no direct evidence correlation between the default
rates of bonds and firms, it was assumed that as a rough approximation
the accuracy of predicting defaults on bonds is equivalent to the
accuracy in identifying defaulting firms. Therefore, the percentages
reported by Hickraan were used in columns 1 and 2 of Table VII-3 as the
equivalents of the Mp and A.™ measures obtained when a particular
financial test is applied to the bankrupt and to the non-bankrupt firm
samples developed by the Agency. Using two alternative estimates of the
baseline failure rate, effectiveness measures (E) were calculated for
each grade of bond ratings in the top four categories. The results of
this analysis are given in Table VII-3. The first assumption is the
baseline of 22 per 10,000 used earlier in this report. The second
reflects the possibility that the baseline failure rate for the kinds of
firms which would have rated bonds might be as low as 11 per 10,000.
Public bond ratings are generally given to bond issuances from
substantial firms. A significant number of large public bond issues are
not rated at all. This category of unrated public debt issues has a
much higher failure rate than the failure rate of those which are
rated. The unrated issues are excluded from the entire rating system,
and from the percentages shown in the Table.
As can be noted from the Table, the results are quite different
depending upon whether one considers the ratings for public utilities,
industrials, and transportation companies combined, or only the ratings
for industrials (and public utilities, though they are not shown). One
important reason why bond ratings, taken as a whole, did not perform
VII-12
-------
TABLE VI1-3
: OF AGLNCY UAT1NGS (1900-1943)
-------
better during this period is that transportation company bonds, speci-
fically those of railroads, were overrated and suffered very high
failure rates, even for firms in the highest rating categories.
The column of Table VII-3 labeled "Percentage of Non-Defaulting
Bonds With Rating" shows the proportion of non-defaulting bonds by
dollar value included in each set of rating categories. The investment
grade ratings I-IV include the vast majority of all rated bonds. They
include 94 percent of all rated corporate bonds and 93 percent of all
rated industrial bonds. Thus, about 6 percent of corporate bonds and 7
percent of industrial bonds are eliminated by insisting on a rating of
investment grade or better. This means that over 94 percent of the
dollar value of all corporate bonds rated were rated as investment grade
or above.
The top two classes of investment grade ratings are much more
restrictive. For all bonds, only 46 percent had one of the top two
ratings, and only 19 percent had the top rating. The Agency therefore
concluded that the top two bond ratings are roughly equivalent to the
effectiveness of a relatively stringent financial test, assuming that
the baseline failure rate is approximately 11 per 10,000 for the firms
likely to use the test rather than 22 per 10,000. A rating of IV or
above, in contrast, is about equivalent to the effectiveness of a less
stringent test.
The Agency also performed a special study to determine the effec-
tiveness of bond ratings as a means of financial responsibility for
utilities in particular. A sample of 113 utilities was drawn from
VII-14
-------
Moody's Public Utilities Manual. All were large (greater than $10
million in net worth), and were either electric or gas utilities. The
mix of bond ratings for these firms was as follows:
TABLE VI1-4
UTILITY BOND RATINGS
RATING
Aaa
Aa
A
Baa
Unrated
NUMBER OF
UTILITIES
4
22
50
25
12
113
PERCENTAGE
OF TOTAL
3.5
19.4
44.2
22.1
10.7
100.0
PERCENTAG
RATED UTIL
3.9
21.7
49.6
24.8
—
100.
As can be seen, requiring that firms have Aaa or Aa ratings to pass
the test would allow only 23-26 percent of utilities to pass this test.
If utilities with A ratings are included, 67-75 percent of utilities
would be able to pass, and a requirement that a firm have an investment
grade bond rating would allow about 89 percent or more to pass the
test. However, because the sample obtained did not contain any util-
ities with a rating lower than investment grade, the Agency could not
determine the precise effect of requiring an investment grade rating.
VII-15
-------
VIII. RESULTS FOR ALL RATIO TESTS
This Section presents the Tables showing the performance of all of
the ratio tests with respect to the evaluation criteria described in
Section IV.A for the primary sample. Tables VIII-1 through VIII-4 give
results for tests with a one-year eligibility requirement and Tables
VIII-5 through VIII-8 give results for tests with a three-year
eligibility requirement.
VIII-1
-------
TABLE VIII-1
PERFORMANCE OF TWO-RATIO TESTS
(One-Year Eligibility Requirement)
(Tests 1-56)
Test
Number
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
Test
Variables
CF/TL > .2
TL/NW < 1.0
CF/TL > .2
TL/NW < 1.2
CF/TL > .2
TL/NW < 1.5
CF/TL > .2
TL/NW < 2.0
CF/TL > .15
TL/NW < 1.0
•CF/TL > .15
TL/NW < 1.2
CF/TL > .15
TL/NW < 1.5
CF/TL > .15
TL/NW < 2.0
CF/TL > .1
TL/NW < 1.0
CF/TL > .1
TL/NW < 1.2
CF/TL > .1
TL/NW < 1.5
CF/TL > .1
TL/NW < 2.0
NI/TA > .04
TL/NW < 1.0
NI/TA > .04
TL/NW < 1.2
NI/TA > .04
TL/NW < 1.5
*B
4.7
4.7
9.5
9.5
4.7
4.7
9.5
7.1
4.7
4.7
11.9
11.9
4.7
4.7
11.9
KW
11.9
14.2
16.6
16.6
14.2
16.6
19.0
21.4
16.6
21.4
30.9
38.0
14.2
19.0
26.1
^B
38
41
41
41
52
59
62
63
58
71
77
81
54
64
66
C
91.7*
90.5*
87.0
87.0
90.5*
89.3
85.8
85.7
89.3
87.0
78.6
75.0
90.5*
88.1
81.0
D
26.1
26.8
24.4
24.4
37.8
42.4
43.0
41.6.
41.4
49.6*
46.1
43.0
39.8
45.0
39.9
EP
4.8*
5.1*
7.0.
7.0
4.0*
4.0*
5.1*
5.0*
4.1*
4.1*
6.1
6.8
3.9*
4.1*
6.3
VIII-2
-------
TABLE VIII-1 (Continued)
Test
Number
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
Test
Variables
NI/TA > .04
TL/NW < 2.0
NI/TA > .02
TL/NW < 1.0
NI/TA > .02
TL/NW < 1.2
NI/TA > .02
TL/NW < 1.5
NI/TA > .02
TL/NW < 2.0
CF/TL > .2
CURRAT > 2.0
CF/TL > .15
CURRAT > 2.0
CF/TL > .1
CURRAT > 2.0
CF/TL > .2
CURRAT > 1.5
CF/TL > .15
CURRAT > 1.5
CF/TL > . 1
CURRAT > 1.5
CF/TL > .2
QRAT > 1.2
CF/TL > .15
QRAT > 1.2
CF/TL > .1
QRAT > 1.2
CF/TL > .2
QRAT > 1.0
CF/TL > .15
QRAT > 1.0
*B
11.9
4.7
4.7
11.9
11.9
4.7
4.7
4.7
4.7
7.1
11.9
4.7
4.7
4.7
4.7
4.7
*v
35.7
16.6
21.4
33.3
42.8
9.5
14.2
19.0
9.5
19.0
33.3
9.5
11.9
16.6
9.5
14.2
^B
72
58
71
77
82
35
54
63
41
62
81
28
42
50
44
55
C
76.2
89.3
87.0
77.4
72.6
92.9*
90.5*
88.1
92.9*
87.0
77.4
92.9*
91.7*
89.4
92.9*
90.5*
D
36.3
41.4
49.6*
43.7
39.2
25.5
39.8
44.0
31.5
43.0
47.7*
18.5
30.1
33.4
34.5
40.8
*P
7.3
4.1*
4.1*
6.5
7.4
4.5*
3.9*
4.2*
3.8*
4.6*
6.1
5.6*
4.3*
4.7*
3.6*
3.8*
VIII-3
-------
TABLE VIII-1 (Continued)
Test
Number
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
Test
Variables
CF/TL > .1
QRAT > 1.0
TL/NW < 1.0
CURRAT > 2.0
TL/NW < 1.2
CURRAT > 2.0
TL/NW < 1.5
CURRAT > 2.0
TL/NW < 2.0
CURRAT > 2.0
TL/NW < 1.0
CURRAT > 1.5
TL/NW < 1.2
CURRAT > 1.5
TL/NW < 1.5
CURRAT > 1.5
TL/NW < 2.0
CURRAT > 1.5
TL/NW < 1.0
QRAT > 1.2
TL/NW <.1.2
QRAI > 1.2
TL/NW < 1.5
QRAT > 1.2
TL/NW < 2.0
QRAT > 1.2
TL/NW < 1.0
QRAT > 1.0
TL/NW < 1.2
QRAT > 1.0
TL/NW < 1.5
QRAT > 1.0
«,
7.1
7.1
11.9
16.6
16.6
4.7
9.5
19.0
23.8
7.1
7.1
9.5
9.5
7.1
9.5
14.2
M..
26.1
16.6
23.8
30.9
35.7
21.4
28.5
45.2
57.1
11.9
19.0
23.8
26.1
11.9
19.0
26.1
V
66
54
65
69
72
62
76
84 .
91
42
49
51
54
52
63
68
C
83.4
88.1
82.2
76.2
73.9
87.0
81.0
67.9
59.6
90.5*
86.9
83.4
82.2
90.5*
85.8
79.8
D
39.9
37.4
41.2
38.1
36.3
40.6
47.5*
38.8
33.9
30.1
30.0
27.2
27.9
40.1
44.0
41.9
EP
5.6*
4.8*
6.0
7.6
8.0
4.6*
5. "5*
8.4
9.8
5.0*
5.9
7.2
7.3
4.0*
5.0*
6.5
VIII-4
-------
TABLE VIII-1 (Concluded)
Test
Number
48
49
50
51
52
53
54
55
56
Test
Variables
TL/NW < 2.0
QRAT > 1.0
NI/TA > .04
CURRAT > 2.0
NI/TA > .04
CURRAT > 1.5
NI/TA > .02
CURRAT > 2.0
NI/TA > .02
CURRAT > 1.5
NI/TA > .04
QRAT > 1.2
NI/TA > .02
QRAT > 1.2
NI/TA > .04
QRAT > 1.0
NI/TA > .02
QRAT > 1.0
KB
14.2
4.7
9.5
4.7
9.5
4.7
7.1
4.7
7.1
*W
33.3
19.0
30.9
26.1
45.2
14.2
23.8
19.0
30.9
^B
74
57
73
64
83
46
49
62
67
C
76.2
88.1
79.8
84.6
72.7
90.5*
84.6
88.1
81.0
D
40.7
38.0
42.1
.37.9
37.8
31.8
25.2
43.0
36.1
=P
7.1
4.6*
6.1
5.3*
7.2
4.5*
6.9
4.2*
6.2
*Exceeds performance of best single ratio tests.
VIII-5
-------
TABLE VIII-2
PERFORMANCE OF THREE-RATIO TESTS: PASS ALL TO
(One-Year Eligibility Requirement)
(Tests 57-88)
PASS
Test
Number
57
58
59
60
61
62
63
64
65
66
67
Test
Variables
CF/TL > .2
TL/NW < 1.0
CURRAT > 2.0
CF/.TL > .15
TL/NW < 1.0
CURRAT > 2.0
CF/TL > .1
TL/NW < 1.0
CURRAT > 2.0
CF/TL > .2
TL/NW < 1.2
CURRAT > 2.0
CF/TL > .2
TL/NW < 1.5
CURRAT > 2.0
CF/TL > .15
TL/NW < 1.2
CURRAT > 2.0
CF/TL > .15
TL/NW <.1.5
CURRAT > 2.0
CF/TL > .1
TL/NW < 1.2
CURRAT > 2.0 i
CF/TL > .1
TL/NW < 1.5
CURRAT > 2.0
CF/TL > .1
TL/NW < 2.0
CURRAT > 2.0
CF/TL > .1
TL/NW < 1.5
CURRAT > 1.5
*B
4.7
4.7
4.7
4.7
4.7
4.7
4.7
4.7
4.7
4.7
7.1
"w
7.1
9.5
9.5
9.5
9.5
11.9
14.2
11.9
14.2
19.0
23.8
^B
33
46
50
34
34
51
52
59
62
62
76
C
94.1*
92.9*
92.9*
92.9*
92.9*
91.7*
90.5*
91.7*
90.5*
88.1
84.6
D
25.9
36.5
40.5
24.5
24.5
39.1
37.8
47.1*
47.8*
43.0
52.2
S
3.9*
3.4*
3.1*
4.6*
4.6*
3.6*
4.0*
3.1*
3.4*
4.2*
4.5*
VIII-6
-------
TABLE VIII-2 (Continued)
Test
Number
68
69
70
71
72
73
74
75
76 .
77
78
79
Test
Variables
CF/TL > .1
TL/NW < 2.0
CUREAT > 1.5
CF/TL > .2
TL/NW < 1.0
QRAT > 1.2
CF/TL > .15
TL/NW < 1.0
QRAT > 1.2
CF/TL > .1
TL/NW < 1.0
QRAT > 1.2
CF/TL > .2
TL/NW < 1.2
QRAT > 1.2
CF/TL > .2
TL/NW < 1.5
QRAT > 1.2
CF/TL > .15
TL/NW < 1.2
QRAT > 1.2
CF/TL > .15
TL/NW < 1.5
QRAT > 1.2
CF/TL > . 1
TL/NW < 1.2
QRAT > 1.2
CF/TL > .1
TL/NW < 1.5
QRAT > 1.2
CF/TL > .1
TL/NW < 2.0
QRAT > 1.2
CF/TL > .2
TL/NW < 1.0
QRAT > 1.0
"B
7.1
4.7
4.7
4.7
4.7
4.7
4-7
4.7
4.7
4.7
. 4.7
4.7
Mw
33.3
7.1
9.5
9.5
9.5
9.5
11.9
11.9
14.2
14.2
16.6
7.1
^B
80
27
36
38
28
28
41
41
45
46
47
34
C
79.8
94.1*
92.9*
92.9*
92.9*
92.9*
91.7*
91.7*
90.5*
90.5*
89.3
94.1*
D
46.7*
19.9
26.5
28.5
18.5
18.5
29.1
29.1
30.8
31.8
30.4
26.9
*P
5.6*
4.8*
4.3*
4.1*
5.6*
5.6*
4.5*
4.5*
4.6*
4.5*
5.0*
3.8*
VIII-7
-------
TABLE VIII-2 (Concluded)
Test
Number
80
81
82
83
84
85
86
87
88
Test
Variables
CF/TL > .15
TL/NW < 1.0
QRAT > 1.0
CF/TL > .1
TL/NW < 1.0
QRAT > 1.0
CF/TL > .2
TL/NW < 1.2
QRAT > 1.0
CF/TL > .2
TL/NW < 1.5
QRAT > 1.0
CF/TL > .15
TL/NW < 1.2
QRAT > 1.0 |
CF/TL > .15
TL/NW < 1.5
QRAT > 1.0
CF/TL > .1
TL/NW < 1.2
QRAT > 1.0
CF/TL > .1
TL/NW < 1.5
QRAT > 1.0
CF/TL > .1
TL/NW < 2.0
QRAT > 1.0
"B
4.7
4.7
4.7
4.7
4.7
4.7
4.7
4.7
4.7
KW
9.5
9.5
9.5
9.5
11.9
11.9
14.2
16.6
23.8
*m
46
49
36
36
52
54
59
62
&4
C
92.9*
92.9*
92.9*
92.9*
91.7*
91.7*
90.5*
89.3
85.8
D
36.5
' 39.5
26.5
26.5
40.1
42.1
44.8
45.4
40.2
EP
3.4*
3.2*
4.3*
4.3*
3.5*
3.4*
5.3*
3.8*
4.9*
*Exceeds performance of best single ratio tests.
VIII-8
-------
TABLE VIII-3
PERFORMANCE OF THREE-RATIO TESTS: PASS 2 OF 3 TO PASS
(One-Year Eligibility Requirement)
(Tests 89-120)
Test
Number
89
90
91
92
93
94
95
96
97
98
.99
Test
Variables
CF/TL > .2
TL/NW < 1.0
CURBAT > 2.0
CF/TL > .15
TL/NW < 1.0
CURRAT > 2.0
CF/TL > .1
TL/NW < 1.0
CURRAT > 2.0
CF/TL > .2
TL/NW < 1.2
CURRAT > 2.0
CF/TL > .2
TL/NW < 1.5
CURRAT > 2.0
CF/TL > .15
TL/NW < 1.2
CURRAT > 2.0
CF/TL > .15
TL/NW < 1.5
CURRAT > 2.0
CF/TL > .1
TL/NW < 1.2
CURRAT > 2.0
CF/TL > .1
TL/NW < 1.5
CURRAT > 2.0
CF/TL > .1
TL/NW < 2.0
CURRAT > 2.0
CF/TL > .1
TL/NW < 1.5
CURRAT > 1.5
«B
7.1
7.1
7.1
11.9
23.8
11.9
11.9
23.8
26.1
26.1
23.8
*V
21.4
26.2
33.3
26.2
35.7
30.9
35.7
38.0
50.0
52.3
59.5
\B
62
70
76
72
77
74
78
80
86
93
91
C
85.8
83.3
79.8
81.0
70.3
78.6
76.2
69.1
62.0
60.8
58.4
D
40.6
43.8
42.7
45.8
41.3
43.1
42.3
42.0
36.0
40.7
31.5
Sp
5.0*
5.2*
5..8*
5.8*
8.5
6.4
6.7
8.5
9.7
9.3
10.1
VIII-9
-------
TABLE VIII-3 (Continued)
Test
Number
100
101
102
103
104
105
106
107
108
109
110
111
Test
. Variables
CF/TL > .1
TL/NW < 2.0
CURRAT > 1.5
CF/TL > .2
TL/NW < 1.0
QRAT > 1.2
CF/TL > .15
TL/NW < 1.0
QRAT > 1.2
CF/TL > .1
TL/NW < 1.0
QRAT > 1.2
CF/TL > .2
TL/NW < 1.2
QRAT > 1.2
CF/TL > . 2
TL/NW < 1.5
QRAT > 1.2
CF/TL > .15
TL/NW < 1.2
QRAT > 1.2
CF/TL > .15
TL/NW < 1.5
QRAT > 1.2
CF/TL > . 1
TL/NW < 1.2
QRAT > 1.2
CF/TL > .1
TL/NW < 1.5
QRAT > 1.2
CF/TL > .1
TL/NW < 2.0
QRAT > 1.2
CF/TL > .2
TL/NW < 1.0
QRAT > 1.0
" :3
28.5
7.1
7.1
7.1
7.1
14.2
7.1
14.2
7.1
19.0
19.0
7.1
\
59.5
16.6
19.0
26 ..1
21.4
28.5
23.8
30.9
28.5
42.8
47.6
16.6
ANB
96
54
59
62
64
70
69
72
77
82
87
59
C
56.0
88.1
' 87.0
83.4
85.8
78.6
84.6
77.4
82.2
69.1
66.7
88.1
D
36.5
37.4
40.0
35.9
42.6
41.5
45.2
41.1
48.5*
39.2
39.4
42.4
**
10.1
4.8*
4.8*
5.9
4.9*
6.7
4.9*
6.9
5.1*
8.3
-
8.4
4.4
i
VIII-10
-------
TABLE VIII-3 (Concluded)
Test
Number
112
113
114
115
116
117
118
119
120
Test
Variables
CF/TL > .15
TL/NW < 1.0
QRAT > 1.0
CF/TL > .1
TL/NW < 1.0
QRAT > 1.0
CF/TL > .2
TL/NW < 1.2
QRAT > 1.0
CF/TL > .2
TL/NW < 1.5
QRAT > 1.0
CF/TL > .15
TL/NW < 1.2
QRAT > 1.0
CF/TL > .15
TL/NW < 1.5
QRAT > 1.0
CF/TL > .1
TL/NW < 1.2
QRAT > 1.0
CF/TL > .1
TL/NW < 1.5
QRAT > 1.0
CF/TL > .1
TL/NW < 2.0
QRAT > 1.0
KB
7.1
7.1
9.5
19.0
9.5
19.0
11.9
26.1
26.1
\
21.4
30.9
21.4
30.9
26.1
35.7
38.0
50.0
50.0
^B
67
80
68
74
75
79
83
88
92
1
C
85.8
81.0
84.6
75.0
82.2
72.7
75.0
62.0
62.0
D
45.6
49.1
46.6*
43.1
48.9*
43.3
45.0
38.0
42.0
S
4.7*
5.2*
4.9*
7.4
5.2*
7.6
6.6
9.5
9.1
*Exceeds performance of best single ratio tests.
VIII-11
-------
TEST VIII-4
PERFORMANCE OF MULTI-RATIO FINANCIAL TESTS
INCLUDING NET FIXED ASSETS/TOTAL ASSETS
(One-Year Eligibility Requirement)
(Tests 121-151)
Test
Number
121
122
123
124
125
126
127
128
129
130
Test
Variables-
Pass All:
NFA/TA • > -3
CF/TL > .1
TL/NW < 1.0
NFA/TA > .3
CF/TL > .1
TL/NW < 1.2
NFA/TA > . 3
CF/TL > .1
TL/NW < 1.5
NFA/TA > . 3
CF/TL > .1
TL/NW < 2.0
NFA/TA > . 25
CF/TL > .1
TL/NW < 1.5
NFA/TA > . 25
CF/TL > .1
TL/NW < 2.0
NFA/TA "> .2
CF/TL > .1
TL/NW < 2.0
Pass 2 of 3:
NFA/TA . > .3
CF/TL > .2
TL/NW < 1.0
NFA/TA > .3
CF/TL > .2
TL/NW < 1.2
^ O
TL/NW < 1.5
^
2.3
2.3
4.7
4.7
4.7
7.1
9.5
7.1
9.5
19.0
^
4.7
4.7
11.9
11.9
11.9
14.2
16.6
11.9
16.6
28.6
^B
43
53
55
59
62
66
75
62
70
73
C
96.5*
96.5*
91.7*
91.7*
91.7*
89.3
87.0
90.5*
87.0
76.2
D
38.3
48.3*
43.1
47.1*
50.1*
51.8*
58.4*
50.1*
53.4*
44.4
i*
1.7*
1.4*
3.3*
3.1*
2.9*
3.6*
3.8*
3.3*
4.0*
7.1
VIII-12
-------
TABLE VIII-4 (Continued)
Test
Number
131
132
133
134
135
136
137
138
139
140
141
Test
Variables
NFA/TA . > .3
CF/TL > .15
TL/NW < 1.0
NFA/TA > . 3
CF/TL > .15
TL/NW < 1.2
NFA/TA > • 3
CF/TL > .15
TL/NW < 1.5
NFA/TA > • 3
CF/TL > .1
TL/NW < 1.0
NFA/TA > *3
CF/TL > .1
TL/NW < 1.2
NFA/TA > -3
CF/TL > .1
TL/NW < 1.5
NFA/TA > -3
CF/TL > .1
TL/NW < 2.0
Pass All:
NFA/TA > -2
CF/TL > .1
CURRAT > 1.5
NFA/TA > • 3
CF/TL > .1
CURRAT > 2.0
NFA/TA > • 3
CF/TL > .2
CURRAT > 1.5
NFA/TA > • 2
CF/TL > .1
CURRAT > 2.0
^B
7.1
9.5
19.0
9.5
11.9
19.0
21.4
4.7
0
2.3
2.3
\
14.2
19.0
30.9
23.8
30.9
45.2
54.7
14.2
0
2.3
7.1
^B
71
76
80
79
83
89
92
76
49
31
55
C
89.3
85.8
75.0
83.4
78.6
67.9
62.0
90.5*
100.0*
97.7*
95.3*
D
56.8*
57.0*
49.1*
55.2*
52.1*
43.8
37.3
61.8*
49.0*
'28.7
47.9*
n
EP
3.3*
4.1*
6.8
4.6* .
5.6*
7.9
9.1
2.8*'
0*
1.6*
1.9*
VIII-13
-------
TABLE VIII-4 (Concluded)
Test
Number
142
143
144
145
146
147
148
149
150
151
Test
Variables.
NFA/TA > -2
CF/TL > .2
CURRAT > 2.0
NFA/TA > -2
TL/NW < 1.2
QRAT > 1.0
NFA/TA > • 3
TL/NW < 1.0
CURRAT > 1.5
NFA/TA > • 3
TL/NW < 1.5
CURRAT > 2.0
Pass 2 of 3:
NFA/TA > -2
CF/TL > .1
CURRAT > 1.5
NFA/TA > • 3
CF/TL > .2
CURRAT > 2.0
NFA/TA > • 3
CF/TL > .2
CURRAT > 1.5
Pass 3 of 4:
NFA/TA • > .3
CF/TL > .1
TL/NW < 2,0
CURRAT > 2.0
NFA/TA > • 3
CF/TL > .1
TL/NW < 1.5
CURRAT > 2.0
NFA/TA > -3
CF/TL > .1
TL/NW < 1.2
CURRAT > 2.0
*B
2.3
4.7
2.3
4.7
21.4
16.6
21.4
14.2
-
11.9
7.1
*W
4.7
9.5
4.7
7.1
64.2
23.8
30.9
33.3
28.5
19.0
**B
. 32
58
49
49
95
70
80
87
81
77
C
96.5*
92.9*
96.5*
94.1*
57.2
79.8
73.9
76.2
79.8
87.0
D
27.3
48.5*
44.3
41.9
30.8
46.2
49.1*
53.7*
52.5*
58.0*
EP
2.4*
2.6*
1.5*
2.6*
9.9
6.3
7.1
6.0
5.5*
3.7*
*Exceeds performance of best single ratio tests.
VIII-
-------
TABLE VIII-5
PERFORMANCE OF TWO-RATIO TESTS
(Three-Year Eligibility Requirement)
(Tests 1-56)
Test
Number
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
Test
Variables
CF/TL > .2
TL/NW < 1.0
CF/TL > .2
TL/NW < 1.2
CF/TL > .2
TL/NW < 1.5
CF/TL > .2
TL/NW < 2.0
CF/TL > .15
TL/NW < 1.0
CF/TL > .15
TL/NW < 1.2
CF/TL > .15
TL/NW < 1.5
CF/TL > .15
TL/NW < 2.0
CF/TL > .1
TL/NW < 1.0
CF/TL > .1
TL/NW < 1.2
CF/TL > .1
TL/NW < 1.5 .
CF/TL > .1
TL/NW < 2.0
NI/TA > .04
TL/NW < 1.0
NI/TA > .04
TL/NW < 1.2
NI/TA > .04
TL/NW < 1.5
KB
2.3
2.3
4.7
4.7
2.3
2.3
4,7
4.7
2.3
2.3
2.3
4.7
2.3
2.3
4.7
Mw'
9.5
11.9
11.9
11.9
11.9
14.2
14.2
19.0
14.2
19.0
21.4
30.9
11.9
16.6
19.0
SB
31
34
35
36
43
50
52
54
49
62
70
76
41
51
56
C
94.1
92.9
91.7
91.7
92.9
91.7
90.5
88.1
91.7
89.4
88.1
82.2
92.9
90.5
88.1
D
40.5
22.1
23.1
24.1
31.1
35.8
37.8
35.0
34.8
43.0
48.6
45.1
29.1
34.4
37.0
EP
4.2
4.6
5.2
5.1
3.6
3.7
4.0
4.8
3.7
3.8
3.7
.5.2
3.8
4.1
4.7
VIII-15
-------
TABLE VIII-5 (Continued)
Test
Number
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
Test
Variables
NI/TA > .04
TL/NW < 2.0
NI/TA > .02
TL/NW < 1.0
NI/TA > .02
TL/NW < 1.2
NI/TA > .02
TL/NW < 1.5
NI/TA > .02
TL/NW < 2.0
CF/TL > .2
CURRAT > 2.0
CF/TL > .15
CURRAT > 2.0
CF/TL > .1
CURRAT > 2.0
CF/TL > .2
CURRAT > 1.5
CF/TL > .15
CURRAT > 1.5
CF/TL > .1
CURRAT > 1.5
CF/TL > .2
QRAT > 1.2
CF/TL > .15
QRAT > 1.2
CF/TL > .1
QRAT > 1.2
CF/TL > .2
QRAT > 1.0
CF/TL > .15
QRAT > 1.0
XB
4.7
2.3
2.3
4.7
7.1
0
0
0
2.3
2.3
4.7
0
0
0
2.3
2.3
KW
28.5
' 14.2
19.0
26.1
38.0
4.7
9.5
14.2
7.1
14.2
26.1
4.7
7.1
11.9
7.1
11.9
^B
60
49
60
69
75
25
37
49
34
46
64
20
27
30
32
43
C
83.4
91.7
89.4
84.6
77.4
97.7*
95.3
92.9
95.3
91.7
84.6
97.7*
96.5*
94.1
95.3
92.9
D
31.5
34.8
41.0
42.9
37.0
20.3 .
27.5
34.8
26.9
31.8
37.9
10.5
15.1
18.1
24.9
31.1
EP
6.1
3.7
3.9
4.9
6.6
2.0*
2.8*
3.2*
3.0*
4.0
5.3
2.5*
2.9*
4.3
3.2*
3.6
VIII-16
-------
TABLE VIII-5 (Continued)
Test
Number
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
Test
Variables
CF/TL > .1
QRAT > 1.0
TL/NW < 1.0
CURRAT > 2.0
TL/NW < 1.2
CURRAT > 2.0
TL/NW < 1.5
CURRAT > 2.0
TL/NW < 2.0
CURRAT > 2.0
TL/NW < 1.0
CURRAT > 1.5
TL/NW < 1.2
CURRAT > 1.5
TL/NW < 1.5
CURRAT > 1.5
TL/NW < 2.0
CURRAT > 1.5
TL/NW < 1.0
QRAT > 1.2
TL/NW <-1.2
QRAT > 1.2
TL/NW < 1.5
QRAT > 1.2
TL/NW < 2.0
QRAT > 1.2
TL/NW < 1.0
QRAT > 1.0
TL/NW < 1.2
QRAT > 1.0
TL/NW < 1.5
QRAT > 1.0
KB
4.7
2.3
7.1
9.5
11.9
2.3
7.1
9.5
16.6
2.3
2.3
4.7
4.7
4.7
7.1
11.9
*w*
23.8
11.9
19.0
23.8
30.9
19.0
26.1
35.7
50.0
7.1
14.2
19.0
21.4
9.5
16.6
23.8
ANB
52
42
50
55
56
53
66
76
80
26
30
33
33
46
50
5A
C
85.8
92.9
87.0
83.4
78.6
89.4
83.4
77.4
66.7
95.3
91.7
88.1
86.9
92.9
88.1
82.2
D
28.2
30.1
31.0
31.2
25.1
34.0
39.9
40.3
30.0
18.9
15.8
14.0
11.6
36.5
33.4
30.2
EP
6.1
3.7
5.8
6.7
8.4
4.4
5.5
6.6
9.2
4.0
6.1
7.9
8.7
3.4
5.2
7.3
VIII-17
------- |