RE-ESTIMATIONS OF THE ZMIJEWSKI AND OHLSON BANKRUPTCY PREDICTION MODELS

doi:10.1016/S0882-6110(03)20004-3

Advances in Accounting

Volume 20, 2003, Pages 77-93

https://doi.org/10.1016/S0882-6110(03)20004-3 Get rights and content

Abstract

Current accounting research uses the Zmijewski (1984) and Ohlson (1980) bankruptcy prediction models as proxies for financial distress/bankruptcy. Such use assumes that the models’ predictive powers transcend to time periods, industries, and financial conditions outside of those used to originally develop the models. The objective of this paper is to address whether the construct validity of the financial distress/bankruptcy proxies (based on the original models) used in those recent studies is possibly open to question. The evidence provided in this study suggests that researchers who use the Zmijewski and Ohlson models using recent data should re-estimate the models’ coefficients to improve the predictive accuracy of the models.

Section snippets

INTRODUCTION AND MOTIVATION

Though the Zmijewski (1984) and Ohlson (1980) models were developed using samples from the 1970s, there is limited evidence addressing the sensitivity of these models to time periods, financial distress situations, and industries outside those of the original samples. Even so, bankruptcy prediction models such as these are still employed in current accounting research to proxy for financial conditions of firms from a variety of industries and time periods (e.g. Altman, 1993; Berger et al., 1996

CONTRIBUTION TO PRIOR RESEARCH

This section summarizes Zmijewski (1984), Ohlson (1980), and other studies that have developed and evaluated bankruptcy prediction models. It explains the contributions of the present study in identifying and resolving the inherent construct validity issues in those earlier studies.

Zmijewski (1984) used financial ratios that measured firm performance, leverage, and liquidity to develop his model. The ratios were not selected on a theoretical basis, but rather on the basis of their performance

RESEARCH DESIGN

This section describes the methodology employed to re-estimate and test the coefficients of the X and Y-score models. In addition, this section describes the selection criteria used to identify the distressed and non-distressed sample companies.

RESULTS AND DISCUSSION

This section reports the findings of the tests used to evaluate the construct validity of the proxies generated by the X and Y-score models. The predictive accuracies of the re-estimated models using the hold-out samples and the stability of their coefficients when re-estimated using the estimation samples are discussed. Evidence related to the models’ sensitivity to non-industrial firms and financial conditions is reported.

SUMMARY

This study evaluated the sensitivity of Zmijewski’s (1984) and Ohlson’s (1980) re-estimated bankruptcy prediction models to samples of distressed and non-distressed companies from time periods, industries, and financial conditions other than those used to develop their models. The findings indicated that the accuracy of the models increased when the coefficients are re-estimated. Zmijewski (1984) and Ohlson (1980) reported 98.2 and 96.4% overall accuracies for their original models using

References (22)

J.S. Grice et al.
Test of the generalizability of Altman’s bankruptcy prediction model
Journal of Business Research
(2001)
E.I. Altman
Financial ratios, discriminant analysis and the prediction of corporate bankruptcy
The Journal of Finance
(1968)
Altman, E. I. (1993). Corporate financial distress and bankruptcy (2nd ed.). New York, NY:...
J. Begley et al.
Bankruptcy classification errors in the 1980s: An empirical analysis of Altman’s and Ohlson’s models
Review of Accounting Studies
(1997)
P.G. Berger et al.
Investor valuation of the abandonment option
Journal of Financial Economics
(1996)
M. Blum
Failing company discriminant analysis
Journal of Accounting Research
(1974)
J.V. Carcello et al.
Temporal changes in bankruptcy-related reporting
Auditing: A Journal of Practice & Theory
(1995)
J.V. Carcello et al.
Audit committee composition and auditor reporting
The Accounting Review
(2000)
K.C.W. Chen et al.
Going concern opinions and the market’s reaction to bankruptcy
The Accounting Review
(1996)
C.W. Chen et al.
Creditors’ decisions to waive violations of accounting-based debt covenants
The Accounting Review
(1993)

E.B. Deakin

A discriminant analysis of predictors of business failure

Journal of Accounting Research

(1972)

Cited by (31)

Bankruptcy prediction models' generalizability: Evidence from emerging market economies
2018, Advances in Accounting
Citation Excerpt :
For this study, financial distress is defined as two consecutive years of negative income by existing firms (see also Altman, 1968; DeAngelo & DeAngelo, 1990; Hill, Perry, & Andes, 1996; Li & Sun, 2008; Oz & Yelkenci, 2015). By adopting this measure of financial distress, which represents a uniform outcome variable, this study seeks to validate the classification results more effectively (Oz & Yelkenci, 2017), whereas previous generalizability studies have not included such a measure (Grice & Dugan, 2001, 2003; Grice & Ingram, 2001; Lawrence, Pongstat, & Lawrence, 2015). With a constant firm sample over time, this research also attempts to overcome the bankruptcy proportion problem, as noted by Zmijewski (1984), which arises from the mismatch between the number of bankruptcies in the estimation sample versus the actual (smaller) number of firms that experience bankruptcy in reality.
The effects of financial distress: Evidence from US GDP growth
2018, Economic Modelling
Citation Excerpt :
Conventional predictors include accounting-based ratios computed from a firm's accounting statements. Grice Jr and Dugan (2003) re-examine the performance of original accounting-based models when applied to periods other than the period studied by pioneer authors and show that these models are sensitive to time. Predictors using stock indices are also common (examples include Agarwal and Taffler (2008) Hillegeist et al. (2004), and these market-based models rely on the Black and Myron (1973) and Merton (1974) option pricing models.
This study provides a dynamic characterization of the link between financial distress risk and the real economy. Using a large dataset of firm-level observations, new ex-ante measures of financial distress are developed at the sector level and used to examine growth trends in the US economy. More specifically, we develop a comprehensive set-up for predicting ex-ante financial distress risk, then examine the effects of ex-ante financial distress risk on GDP growth. Our results show that over the period of 1970–2012, ex-ante financial distress risk contracts GDP growth by up to 9 per cent. The results also reveal greater contractions in exports and investment. The results remain unchanged when internal and external instruments are used to address endogeneity issues.
Ensemble boosted trees with synthetic features generation in application to bankruptcy prediction
2016, Expert Systems with Applications
Citation Excerpt :
Following this line of thinking, Altman (1968) proposed to use multidimensional analysis to predict corporate bankruptcy that was further developed by others (Altman & Loris, 1976; Blum, 1974; Deakin, 1972; Edmister, 1972; Ketz, 1978; Koh & Killough, 1990; Laitinen, 1991; Libby, 1975; Meyer & Pifer, 1970; Pettway & Sinkey, 1980; Rujoub, Cook, & Hay, 1995; Sinkey, 1975; Wilcox, 1973). In parallel, a great interest was paid to the generalized linear models that can be used in both decision making and providing certainty of the prediction (Aziz, Emanuel, & Lawson, 1988; Grice & Dugan, 2003; Hopwood, McKeown, & Mutchler, 1994; Koh, 1991; Li & Miu, 2010; Ohlson, 1980; Platt & Platt, 1990; Platt, Platt, & Pedersen, 1994; Zavgren, 1983; Zmijewski, 1984). Additionally, the generalized linear models are of special interest because estimated weights of the linear combination of economic indicators in the model can be further used to determine importance of the economic indicators.
Bankruptcy prediction has been a subject of interests for almost a century and it still ranks high among hottest topics in economics. The aim of predicting financial distress is to develop a predictive model that combines various econometric measures and allows to foresee a financial condition of a firm. In this domain various methods were proposed that were based on statistical hypothesis testing, statistical modeling (e.g., generalized linear models), and recently artificial intelligence (e.g., neural networks, Support Vector Machines, decision tress). In this paper, we propose a novel approach for bankruptcy prediction that utilizes Extreme Gradient Boosting for learning an ensemble of decision trees. Additionally, in order to reflect higher-order statistics in data and impose a prior knowledge about data representation, we introduce a new concept that we refer as to synthetic features. A synthetic feature is a combination of the econometric measures using arithmetic operations (addition, subtraction, multiplication, division). Each synthetic feature can be seen as a single regression model that is developed in an evolutionary manner. We evaluate our solution using the collected data about Polish companies in five tasks corresponding to the bankruptcy prediction in the 1st, 2nd, 3rd, 4th, and 5th year. We compare our approach with the reference methods.
Bankruptcy prediction using terminal failure processes
2015, European Journal of Operational Research
Traditional bankruptcy prediction models, designed using classification or regression techniques, achieve short-term performances (1 year) that are fairly good, but that often worsen when the prediction horizon exceeds 1 year. We show how to improve the performance of such models beyond 1 year using models that take into account the evolution of firm’s financial health over a short period of time. For this purpose, we design models that fit the underlying failure process of different groups of firms. Our results demonstrate that such models lead to better prediction accuracy at a 3-year horizon than that achieved with common models.
Predicting restaurant financial distress using decision tree and AdaBoosted decision tree models
2014, Economic Modelling
The restaurant industry has been facing tough challenges because of the recent economic turmoil. Although different industries face different levels of competition and therefore the likelihood of financial distress can differ for firms in different industries, scant attention has been paid to predicting restaurant financial distress. The primary objective of this paper is to examine the key financial distress factors for publicly traded U.S. restaurants for the period from 1988 to 2010 using decision trees (DT) and AdaBoosted decision trees. The AdaBoosted DT model for the entire dataset revealed that financially distressed restaurants relied more heavily on debt; and showed lower rates of increase of assets, lower net profit margins, and lower current ratios than non-distressed restaurants. A larger proportion of debt in the capital structure ruined restaurants' financial structure and the inability to pay their drastically increased debt exposed restaurants to financial distress. Additionally, a lack of capital efficiency increased the possibility of financial distress. We recommend the use of the AdaBoosted DT model as an early warning system for restaurant distress prediction because the AdaBoosted DT model demonstrated the best prediction performance with the smallest error in overall and type I error rates. The results of two subset models for full-service and limited-service restaurants indicated that the segments had slightly different financial risk factors.
Forecasting financial failure using a Kohonen map: A comparative study to improve model stability over time
2012, European Journal of Operational Research
This study attempts to show how a Kohonen map can be used to improve the temporal stability of the accuracy of a financial failure model. Most models lose a significant part of their ability to generalize when data used for estimation and prediction purposes are collected over different time periods. As their lifespan is fairly short, it becomes a real problem if a model is still in use when re-estimation appears to be necessary. To overcome this drawback, we introduce a new way of using a Kohonen map as a prediction model. The results of our experiments show that the generalization error achieved with a map remains more stable over time than that achieved with conventional methods used to design failure models (discriminant analysis, logistic regression, Cox’s method, and neural networks). They also show that type-I error, the economically costliest error, is the greatest beneficiary of this gain in stability.

View all citing articles on Scopus

View full text

RE-ESTIMATIONS OF THE ZMIJEWSKI AND OHLSON BANKRUPTCY PREDICTION MODELS

Abstract

Section snippets

INTRODUCTION AND MOTIVATION

CONTRIBUTION TO PRIOR RESEARCH

RESEARCH DESIGN

RESULTS AND DISCUSSION

SUMMARY

Journal of Business Research

Financial ratios, discriminant analysis and the prediction of corporate bankruptcy

The Journal of Finance

Bankruptcy classification errors in the 1980s: An empirical analysis of Altman’s and Ohlson’s models

Review of Accounting Studies

Investor valuation of the abandonment option

Journal of Financial Economics

Failing company discriminant analysis

Journal of Accounting Research

Temporal changes in bankruptcy-related reporting

Auditing: A Journal of Practice & Theory

Audit committee composition and auditor reporting

The Accounting Review

Going concern opinions and the market’s reaction to bankruptcy

The Accounting Review

Creditors’ decisions to waive violations of accounting-based debt covenants

The Accounting Review

A discriminant analysis of predictors of business failure

Journal of Accounting Research