INSOLVENCY PREDICTION IN THE PRESENCE OF DATA INCONSISTENCIES

• Author: A. L. Martinez,A. Mendes,R. L. Cardoso,F. R. Ferreira,P. C. Mário
• Date: 01 July 2014
• Published date: 01 July 2014
• DOI: http://doi.org/10.1002/isaf.1352

INSOLVENCY PREDICTION IN THE PRESENCE OF DATA

INCONSISTENCIES

A. MENDES,

a

*R. L. CARDOSO,

b

P. C. MÁRIO,

c,d

A. L. MARTINEZ

e

AND F. R. FERREIRA

f

a

School of Electrical Engineering and Computer Science, The University of Newcastle, Callaghan, NSW, Australia

b

The Brazilian School of Public and Business Administration, Getúlio Vargas Foundation, Rio de Janeiro, RJ, Brazil

c

School of Economics, FederalUniversity of Minas Gerais, Belo Horizonte, MG, Brazil

d

UNA University Center, Belo Horizonte, MG, Brazil

e

FUCAPE Business School, Vitória, ES, Brazil

f

Accounting School, Centro Universitário do Estado do Pará,Belém, PA, Brazil

SUMMARY

In this paper we use data inconsistencies as an indicator of financial distress. Traditional models for insolvency

prediction normally ignore inconsistent data, either by removing or replacing it. Instead of removing that information,

we propose a new variablet oc apturei t; using it togetherwi tht raditional accounting variables (based on financial ratios)

for the purpose of insolvency prediction.

Computational tests use three datasets based on the financial results of 2033 Brazilian Health Maintenance

Organizations over 7 years (2001 to 2007). Sixteen classification methods were used to evaluatewhether or not the

new variable impacted sol vency prediction. Tests show a statistically sig nificantimprovementin classification accuracy

–average results improve 1.3 (p=0.003 ) and 1.8 (p= 0.006) percentage points, for 10-foldand leave-one-out cross-

validations respectively. In addition, the analysis of false positives and false negatives shows that the new variable

reduces the potentially harmful misclassification of false negatives (i.e. financially distressed companies being classified

as financially healthy) and also reduces the estimated overall error rate.

Regarding the extensibility of the results, even though this work uses data from Brazilian companies only, the

calculation of the financial ratios variables, as well as the inconsistencies, could be extended to most companies

worldwide subject to governmental accounting regulations aligned with the International Financial Reporting

Standards. Copyright © 2014 John Wiley & Sons, Ltd.

Keywords: data mining; insolvency prediction; classification; inconsistency

1. INTRODUCTION

This paper addressesan important topic in the analysis of insolvency prediction,namely the inconsistency

of accounting databy companies. Insolvencyand bankruptcy have been studiedin the areas of accounting

and finance for severaldecades. Most of these studies address theseelements under different perspectives,

either by tryingto predict them (Altman, 1968; Ohlson,1980; Newton, 2003; Altman & Hotchkiss,2006),

or by analysing the processes that occur during an insolvency crisis or bankruptcy (Aghion, Hart, &

Moore, 1993; Hart, 2000). As a side note, in the literature, insolvency, failure and bankruptcy usually

appear as synonyms; however, they refer to different moments. Insolvency is linked to a state, failure to

an act, and bankruptcy has a legal meaning, as in a judicial process.

* Correspondence to: A. Mendes, School of Electrical Engineering and Computer Science, The University of Newcastle,

Callaghan, NSW, Australia. E-mail: Alexandre.Mendes@newcastle.edu.au

Copyright © 2014 John Wiley & Sons, Ltd.

INTELLIGENT SYSTEMS IN ACCOUNTING, FINANCE AND MANAGEMENT

Intell. Sys. Acc. Fin. Mgmt. 21, 155–167 (2014)

Published online 18 March 2014 in Wiley Online Library (wileyonlinelibrary.com) DOI: 10.1002/isaf.1352