The standardized discriminant function coefficients and the discriminant function
are as shown in Exhibit 10.21. The classification accuracy of the model on the de-
velopment sample was 97% overall. With the same level of accuracy for Type I and
II. Tests on data 2 years prior yielded a Type I accuracy of 91% and Type II accuracy
of 93%. No hold-out test results were reported.
10.24 SUMMARY. We have attempted to review and compare a relatively large
number of empirical failure classification models from over twenty countries. Much
of the material is derived from little-known sources and as such we hope that the
study will stimulate a greater transnational discussion. Indeed, as financial institu-
tions and government agencies in countries such as Canada, the United States, Brazil,
France, and England wrestle with the specter of large firm failures in the future, the
knowledge that prior work has been done with respect to early warning models may
help obviate the consequences or reduce the number of these failures.
We expect the quality and reliability of models constructed in many of the afore-
mentioned countries to improve (1) as the quality of information on companies is ex-
panded and refined, (2) as the number of business failures increase, thereby provid-
ing more data points for empirical analysis, and (3) as researchers and practitioners
become more aware of the problems and potential of such models. Where sufficient
data do not exist for specific sector models, for instance, manufacturing, retailing,
and service firms, the application of industry relative measures, for example, Altman
and Izan (1983), can perhaps provide a satisfactory framework for meaningful analy-
sis. Of course, this requires that government or private agencies build reliable indus-
try databases for comparison purposes.
SOURCES AND SUGGESTED REFERENCES
Abrahams, A., and R. A. I. van Frederikslust. “Discriminant Analysis and the Prediction of
Corporate Failure.”European Finance Association 1975 Proceedings. R. Brealey and G.
Rankine (eds.). Amsterdam: North Holland, 1976.
X 6 : Quick assets>current debtX 5 : Quick assets>inventoryX 4 : Total debt>total assetsSOURCES AND SUGGESTED REFERENCES 10 • 47Coefficient
(absolute value The Relative
of the difference Importance of the
Ratio Coefficient of the means) Ratio (%)
X 1 18.11 5.4029 52.04
X 2 1.64 1.0365 9.98
X 3 –1.21 0.1078 1.04
X 4 1.21 0.1806 1.74
X 5 –0.96 0.3890 3.75
X 6 5.85 3.2663 31.45
Exhibit 10.21. Discriminant Function Coefficients.