본문 바로가기
HOME> 논문 > 논문 검색상세

학위논문 상세정보

도산예측모델의 유용성에 관한 실증적 연구 : E.Altman, W.Beaver, J.Horrigan, 한국은행 및 종합금융회사 신용평가 모델적용결과 분석: 한국기업과 일본기업의 표본을 대상으로 원문보기
(An)empirical study on the usefulness of default prediction models

  • 저자

    조정만

  • 학위수여기관

    昌原大學校

  • 학위구분

    국내박사

  • 학과

    경영학과

  • 지도교수

  • 발행년도

    2003

  • 총페이지

    vi, 103장

  • 키워드

    도산 예측[倒産豫測] 도산예측모델 DEFAULT PREDICTION MODEL EALTMAN WBEAVER JHORRIGAN 한국은행 종합금융회사 신용평가 모델적용결과 한국기업 일본기업;

  • 언어

    kor

  • 원문 URL

    http://www.riss.kr/link?id=T9168198&outLink=K  

  • 초록

    The primary objective of this study is to ascertain if three well known financial models selected in the study are effective for predicting corporate failures. The motivation for this study is the lending scandals in the past few years in Korea in which lending institutions have allegedly lent capital to corporations that could have been identified as failing. To time-test the effectiveness of the models and to include a model from each area of research methodology from the literature of accounting and finance, the works of three authors were selected, namely those of Edward I. Altman (1968), William H. Beaver (1966), and James O. Horrigan (1966). If the usefulness of the financial models in predicting the corporate failures is attested, the utility of a certain financial data is also to be positively affirmed albeit in a roundabout way. This study does not support any hypotheses that claim any particular author's work superior to others, nor does it endorse any specific research methodology. Rather this study attempts to prove that the lending institutions could have avoided much bad lending if they had utilized a few classic financial models easily accessible to any students of accounting and finance. Therefore the criteria for selecting the models are not the same as those to prove the effectiveness of particular models. Time-tested well known "textbook" models like Altman's bankruptcy model meet the selecting criteria. The works of Beaver and Horrigan referred to above also meet such criteria for their early contribution to accounting and finance in attesting the utility of financial data in predicting corporate failures and credit ratings, respectively. This study, therefore, simply applies the findings of their original works as they pertain to its objectives. To attest the usefulness of the models and to affirm the utility of the financial data, however, this study employs several research methodologies in analyzing the results of the adopted model's outcome, namely the association analyses: (1) cross-tabulation analysis, (2) Chi-squarer test, (3) Somers's dyx, (4) t-test, and (5) ANOVA. Type I and Type II errors of the model outcomes are discussed as well. As expected the analyses were able to attest the usefulness of the models in a variety of degrees: strong, mixed, and weak. Scores generated by three models above and two credit rating models used by Korean Bank Association and Korean Merchant Bank Association were used for the effectiveness test by a variety of methods, including Somers's dyx. Samples include 49 bankrupt firms and 90 active firms from Korea, 53 bankrupt firms and 100 active firms from Japan. Financial statements used for three financial models cover 3 year periods before their bankruptcy, and financial statements used for two models by Korean Bank Association and Korean Merchant Bank Association cover 2 year periods. The number of bankrupt firms increased two-fold in four years: 4,107 in 1990(5% of the total population) to 11,255 in 1994(11% of the total population), and 13,992 in 1995 (11% of the total population). Could some sort of alternative lending practices have prevented Korea from the economic crisis that required the International Monetary Fund's rescue? There are numerous studies on the development of bankruptcy risk of corporations. "A Classification of Business Failure by Financial Ratios" by Hwang Suk-Ha (1990); "The Corporate Bankruptcy Prediction with Accounting Information and Market Reaction" by Lee Gae-Won (1992); "An Inductive learning-Assisted Neural Network Approach to Bankruptcy Prediction: Comparison with MDA, Inductive Learning, and Neural Network Models" by Lee K. C., Kim M. J., Kim H. (1994); "The Comprehensive Credit Analysis Table of Korean Banks and Financial Distress Prediction" by Huh S. K., Huh T. K. (1997); "ZETA Analysis: A New Model to Identify Bankruptcy Risk of Corporations" by Altman, Haldeman, and Narayanan (1997); "Causes for Corporate Failures and Bankruptcy Prediction Model Analysis: The periods before and after the IMF" by Nam J. H. (1998); "An Application of Survival Analysis for Firm Failure Prediction" by Nam J. W., Lee H. K., Kim D. S. (2001); to name a few. With the exception of the paper by Huh and Huh (1997) all the studies referred to above attempted to develop a bankruptcy model to predict the corporate failures, implying the utility of financial and non-financial data. While their findings contribute to the literature relevant to the prediction of bankruptcy risk of corporations, this study explicitly suggests that lending institutions must utilize a variety of time-tested financial models like the works of Altman and others before lending. While business failures are caused by a number of mishaps, including uncontrollable risk factors at the firm level, as much as firm's ill-preparation for a number of controllable risk types, the findings of this study offer an easy test to predict if a firm may be heading for bankruptcy two to three years ahead of time. Could systematic lending practices of other sorts have prevented Korea from the economic crisis that required the IMF's rescue? To answer this question at a micro-level, the author wishes to inquire about the amounts and forms of actual financing of the firms identified as failing' for the analysis periods. Further, the author wishes to include currently developed bankruptcy models' in a future study if not models with more currently developed methodologies.


 활용도 분석

  • 상세보기

    amChart 영역
  • 원문보기

    amChart 영역