Professional Documents
Culture Documents
Accounting Research Center, Booth School of Business, University of Chicago
Accounting Research Center, Booth School of Business, University of Chicago
Accounting Research Center, Booth School of Business, University of Chicago
Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at .
http://www.jstor.org/page/info/about/policies/terms.jsp
.
JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of
content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms
of scholarship. For more information about JSTOR, please contact support@jstor.org.
Wiley-Blackwell and Accounting Research Center, Booth School of Business, University of Chicago are
collaborating with JSTOR to digitize, preserve and extend access to Journal of Accounting Research.
http://www.jstor.org
This content downloaded by the authorized user from 192.168.82.207 on Sun, 25 Nov 2012 08:18:15 AM
All use subject to JSTOR Terms and Conditions
Research Reports
AccountingRatios and the Predictionof
Failure: Some BehavioralEvidence
ROBERT LIBBY*
This content downloaded by the authorized user from 192.168.82.207 on Sun, 25 Nov 2012 08:18:15 AM
All use subject to JSTOR Terms and Conditions
PREDICTING FAILURE 151
This content downloaded by the authorized user from 192.168.82.207 on Sun, 25 Nov 2012 08:18:15 AM
All use subject to JSTOR Terms and Conditions
152 JOURNAL OF ACCOUNTING RESEARCH, SPRING, 1975
Method
USEFULNESS
CONSISTENCY
DESIGN
This content downloaded by the authorized user from 192.168.82.207 on Sun, 25 Nov 2012 08:18:15 AM
All use subject to JSTOR Terms and Conditions
PREDICTING FAILURE 153
This content downloaded by the authorized user from 192.168.82.207 on Sun, 25 Nov 2012 08:18:15 AM
All use subject to JSTOR Terms and Conditions
154 ROBERT LIBBY
TABLE 1
SubjectDescriptions
Mean Minimum Maximum
1. Age 39 23 60
2. Years of experience 9 0 32
3. Estimated N of applicationsprocessed * 51 >250
annually
4. Highestgradein school completed 16.2 12 17
5. Averageclienttotal assets (in000's) C-U $127 -
bankers -$25 $700,000
6. Average client total assets (in 000's) $146,198 -
Phila. bankers
* Computationof the mean responseto the questionwas inappropriatesince it
requiredchoice of one of six responsesrangingfrom"0 to 50" to "more than 250"
applicationsper year.
analysisoftherotatedfactormatrix,netincome/total assets,currentassets/
sales, currentassets/currentliabilities,currentassets/totalassets, and
cash/totalassets were chosen respectivelyto representthe fivefinancial
dimensions.'
The predictiveabilityofthe 14-ratioset and ofthereducedset werethen
compared.The original14 ratioswereable to predictcorrectly
54 of60 cases
based upon the derivationsample and 41 of 60 predictionsbased upon
double cross-validation.7
The reducedset predicted51 and 43 correctof 60
based upon the derivationsampleand double cross-validation
respectively.
This slightreductionin predictiveabilitywas deemedsatisfactoryforthis
experiment.
TASK AND INSTRUCTIONS
This content downloaded by the authorized user from 192.168.82.207 on Sun, 25 Nov 2012 08:18:15 AM
All use subject to JSTOR Terms and Conditions
PREDICTING FAILURE 155
Results
USEFULNESS
This content downloaded by the authorized user from 192.168.82.207 on Sun, 25 Nov 2012 08:18:15 AM
All use subject to JSTOR Terms and Conditions
156 ROBERT LIBBY
49-51 * * * * * * * * * 9
46-48 * * * * * * * * 8
43-45 * * * * * * * * * * * * * * 14
40-42 * * * * * * * * 8
37-39 * 1
34-36 * * 2
31-33
28-30
25-27 * 1
Mean 44.4
Std. deviation4.5
Maximum 50.0
Minimum 27.0
FIG. Scores(Number
Achievement
1. Prediction outof60Responses)
Correct
This content downloaded by the authorized user from 192.168.82.207 on Sun, 25 Nov 2012 08:18:15 AM
All use subject to JSTOR Terms and Conditions
PREDICTING FAILURE 157
10/10 * * * * * * * * * * * * * 13
9/10 * * * * * * * * * * * * * * * * 16
8/10 * * * * * * * * * * * * 12
7/10 * 1
6/10 * 1
Mean 8.9
Std. deviation .9
Maximum 10.0
Minimum 6.0
FIG. 2. Test-RetestReliabilityof Failure Predictions
CONSISTENCY
This content downloaded by the authorized user from 192.168.82.207 on Sun, 25 Nov 2012 08:18:15 AM
All use subject to JSTOR Terms and Conditions
158 ROBERT LIBBY
RELATED ISSUES
This content downloaded by the authorized user from 192.168.82.207 on Sun, 25 Nov 2012 08:18:15 AM
All use subject to JSTOR Terms and Conditions
PREDICTING FAILURE 159
DEBRIEFING
Limitations
The major limitationsof the study relate to the generalizability of the
resultsofthe studywithrespectto (1) actors,(2) otherexperimental situa-
tions,(3) the real worldsituationsrelatedto the experimental task and (4)
otherdefinitionsof failure.Participantsin the study were not randomly
selected.The two groupswere specificallychosenbased upon a prioridif-
ferencesin lendingexperienceconstrainedby subjectavailability.Although
valid statisticalinferencecan only be made to otherresponsesmade by
the same subjects,the similarityin resultsbetweenthe two a prioridiffer-
ent subgroupssuggestsa higherdegreeof generalizability.
The specificinstructions,data presentations,and dependentmeasures
used in the experimental situationmay affectthe resultsofthe study.Such
seeminglyminorchangesas usinga six-pointscale in place of a two-point
scale or seven in place of fiveratiosmay have significanteffects.This type
of externalvaliditycan onlybe determinedthroughadditionalexperimen-
tation.
In the actual lendingdecision,loan officershave an abundanceof multi-
period quantitativeand qualitative informationavailable to them. The
presentexperiment attemptedto testtheusefulnessofonlya smallsegment
This content downloaded by the authorized user from 192.168.82.207 on Sun, 25 Nov 2012 08:18:15 AM
All use subject to JSTOR Terms and Conditions
160 ROBERT LIBBY
Summaryand Conclusions
The major thesisof this studywas that the usefulnessof accountingin-
formationis a functionof the predictiveabilityof the informationand the
abilityof usersto interpretthe data. The predictionachievementcriterion
was developed based upon this thesis.The experimentconductedin this
study directlymeasuredthe predictionachievementof a selected set of
accountingratios and loan officers in the predictionof failure.Also, the
test-retestand interraterreliabilityof the loan officers'predictionswas
measured.
A small empiricallyderived set of accountingratios allowed bankers
rangingwidelyin backgroundto make highlyaccurateand reliablepredic-
tions of business failure.The resultssuggestthat futureresearchwhich
wouldtestthegeneralityoftheresultsoverdecisionmakers,environmental
events,and information sets may be profitable.The presentsituationwas
chosenbecause it was simple,well defined,and the subject of previousre-
search.Testingthe same hypothesesin othersituationswill be morecom-
plex.
REFERENCES
ABDEL-KHALIK, A. R. "The EffectofLinearAggregation ofAccountingData on the
Quality of Decisions.." Unpublisheddoctoral dissertation,Universityof Illinois,
1972.
."The EffectofAggregating AccountingReportson theQualityof the Lending
Decision: An EmpiricalInvestigation,"EmpiricalResearchin Accounting:Selected
Studies,1973.Supplementto JournalofAccounting Research11.
ALTMAN, E. I. "The Predictionof CorporateBankruptcy:A Discriminant Analysis."
Unpublisheddoctoraldissertation,UCLA, 1968a.
. "Financial Ratios, DiscriminantAnalysis and the Predictionof Corporate
Bankruptcy,"JournalofFinance,23 (September1968b):589-609.
BEAVER, W.H. "Financial Ratios as Predictorsof Failure," Empirical Researchin
Accounting: SelectedStudies,1966.Supplementto Journalof AccountingResearch,
4: 71-111.
. "AlternativeAccountingMeasuresas Predictorsof Failure," The Accounting
Review,43 (January1968): 113-22.
12 See footnote4.
This content downloaded by the authorized user from 192.168.82.207 on Sun, 25 Nov 2012 08:18:15 AM
All use subject to JSTOR Terms and Conditions
PREDICTING FAILURE 161
This content downloaded by the authorized user from 192.168.82.207 on Sun, 25 Nov 2012 08:18:15 AM
All use subject to JSTOR Terms and Conditions