臺灣博碩士論文加值系統

Abstract
The recent bankruptcies of some construction companies have underlined the importance of default prediction in this industry. It now seems to be more necessary than ever to develop early warning systems that can help prevent or avert corporate default and that facilitate the selection of firms to collaborate with or invest in.
One of the most important original evaluation tools for the strength of a firm is its financial statements. This research will provide a framework of the default probability valuation relying on the financial ratios by using the Logit model. First of all, according to some paper, a total of 21 ratios will be taken into the calculation of the default probability. These ratios will be gathered into 5 groups according to the characteristics. Next, by using the Logit model for single variable analysis to validate the data following to the leave-one-out method, and then the fitting of the model will be validated by using ROC curve. The results will be the default probability, the area under ROC curve and the optimal cut-off point of each variable correlatively. Relying on the experimental results, the thesis will point out seven highlight factors, which have strong effect and can be used as a stand-alone factor for the prediction of a firm’s default probability. They are ROA (AUC = 0.783), Debt ratio (AUC = 0.6529), Account payable turnover (AUC = 0.6922), Quick ratio (AUC = 0.56), Current ratio AUC = 0.5832), Net working capital to total asset (AUC = 0.6123), Turnover of total assets (AUC = 0.5989). After the impact of the single ratios will be recognized, some of them will be selected for gathering into some combination for multi-variables analysis. Finally, the importance of the market factor as well as the effectiveness of considering the financial factors simultaneously to the default probability prediction will be proven. The criterion points also are calculated for warning those who concern about the default probability of a firm by analyzing its financial statement.

Table of contents
Acknowledgement i
Abstract ii
Table of contents iv
List of figures vii
List of tables ix
Chapter 1: Introduction 1
1.1 Background 1
1.2 Motivations and problem statements 2
1.3 Purposes and contributions 4
1.4 Procedure of the research 5
1.5 Framework of the thesis 7
Chapter 2 : Literature review 9
2.1 Characteristics of construction industry 9
2.1.1 General characteristics 9
2.1.2 Accounting principle 10
2.1.3 External environment 10
2.1.4 The financial characteristics 10
2.2 Researches relative to the default probability prediction 11
2.3 Summary 15
Chapter 3 : Methodology 17
3.1 The Logit model 18
3.1.1 The statement of Logit model 18
3.1.2 The Logistic regression function 19
3.1.3 Application of the Logit regression 20
3.1.4 Maximum likelihood method 20
3.1.5 Log-likelihood function 20
3.2 Cross-validation 21
3.2.1 Over-fitting problems 21
3.2.2 Cross-validation 21
3.3 ROC curve 22
3.3.1 Definition and methodology of ROC curve 22
3.3.2 Using of the ROC curve in the validation of the model 24
3.3.3 Determine the optimal cut-off point 24
3.4 Summary 25
Chapter 4 : Data collection 26
4.1 Data collection 26
4.1.1 Source and authenticity of the input data 26
4.1.2 Data collection principles 26
4.1.3 Summary of the input data 27
4.2 Data arrangement 28
4.2.1 Collection of Financial ratios data 28
4.2.2 Arrangement of selected financial ratios 29
4.3 Financial ratios definition and statistical characteristic 30
4.4 Maturities of the prediction 32
4.5 Restrictions of the input data 33
4.6 Data analysis process and algorithm 34
4.6.1 Tools of data analysis 34
4.6.2 The algorithm chart of the data analysis process 35
4.7 Summary 36
Chapter 5 : Single variable analysis and empirical result 37
5.1 Correlation between the ratios and the default probability 37
5.2 Validation of the correlation between single ratio and its default probability 38
5.3 Application of the ROC curve in the model validation 45
5.3.1 The area under ROC curve (AUC) 45
5.3.2 Selection of the valid single variable and determine optimal cut-off point 57
5.4 Summary 60
Chapter 6 : Multivariate analysis and the empirical result 61
6.1 Multivariate analysis 61
6.1.1 Principles and reasons for gathering the combos 61
6.1.2 The combo selection 63
6.2 Correlation between the default probability and the combo’s value 63
6.3 Application of the ROC in the model validation 66
6.3.1 ROC curve area and accuracy 66
6.3.2 Determination of the optimal cut-off point 69
6.4 Comparison with the relative research 71
6.5 Summary 73
Chapter 7 : Conclusions and Suggestions 74
References 76
Appendix 79
A.1 Sample total of companies 79
A.2 The optimal cut-off line of the valid variables 83
A.3 The optimal cut-off point of the combos 92

References
1. Ohlson (1980). Financial ratios and the probabilistic prediction of Bankruptcy. Journal of Accounting reseach , Vol.18, 109-131.
2. Abidali, Adnan Fadhil and Harris, Frank. (1995). A Methodology for predicting company failure in the Construction industry. Construction management and economics , Vol.13, 189-196.
3. Akobeng (2006). Understanding diagnostic test 3: Receiver operating characteristic curves. Journal compilation 2007 foundation Acta paediatrica , 644-647.
4. Altman and Edward (1968). financial ratios, Discriminant analysis and the prediction of corporate bankruptcy. Journal of finance , Vol.23, No.4, pp.589-609.
5. Biology at the university of alaska Fairbanks. (2009). Retrieved from http://mercury.bio.uaf.edu/: http://mercury.bio.uaf.edu/courses/wlf625/readings/MLEstimation.PDF
6. Brockman, Paul and Turtle (2003). A Barrier option framework for corporate security valuation. Journal of financial economics , 67, 511-529.
7. Chung (2008). Insolvency prediction model using multivariate discriminant analysis and artificial neural network for the finance industry. International Journal of business and management , 19-29.
8. Donald S.Barie and Boyd C.Paulson. (1992). Professional Construction management including C.M. Design-Contruct and General Contracting (Third edition ed.). McGray-Hill international editions.
9. Douglas W.Dwyer and Yi-Yun Wang. (2005). Moody''s KMV Riskcalc V3.1 France.
10. Edum-Fotwe, Francis, Price, Andrew and Thorpe, Antony. (1996). A Review of Financial ratio tools for predicting Contractor Insolvency. Construction management and Economics , Vol.14, 189-198.
11. Ekanayake. (2008, 11). Contractor insolvency- be prepared. Retrieved 4 2010, from FSI: http://www.fsilaw.com/sitecore/content/Global/content/Mads%20insolvency.aspx
12. Enrique F. Schisterman, Neil J.Perkins, Aiyi Liu and Howard Bondell. (2005). Optimal cut-off point and its Corresponding Youden index to discriminate individuals Using pooled blood samples. Epidemiology , 16, 73-81.
13. Evan carmichael. (2010). Retrieved from http://www.evancarmichael.com/: http://www.evancarmichael.com/Management/3327/UNDERSTANDING-FINANCIAL-AND-OPERATING-RATIOS.html
14. Beaver (1966). Financial ratios as predictors of failure. Journal of Accounting research , Vol.4, 71-111.
15. Halpin, Daniel (1985). Financial and cost concepts for construction management. John wiley & Sons Inc.
16. Hayden (1999). Estimation of a rating model for coporate exposues. financial report.
17. Huang (2009). Prediction of contractor default probability using structural models of credit risk: an empirical investigation. Construction Management and Economics , 581-596.
18. Invest sign. (2010). Retrieved 4 2010, from http://www.investsign-home.com: http://www.investsign-home.com/Publications_Financial_ratios.htm.
19. James M.W.Wong and S.Thomas NG (2010). Company failure in the construction industry: A critical review and a future reseach agenda. Sydney-Australia.
20. Jian-Hua LUO and Han-yun LEI. (2008). Empirical study of corporation credit default probability based on Logit model. China.
21. John Concato, Alvan R.Feinstein and Theodore R.Holford (1993). The risk of determining risk with multivariable models. Annal of internal medicine , 118, 201-210.
22. Kangari (1988). Business failure in construction industry. Journal of construction Engineering and management , 172-190.
23. Kangari Farid and Elgharib (1992). Financial performance analysis for construction industry. Journal of Construction Engineering and Management ASCE , 349-361.
24. lord (2010). Investopia. Retrieved from http://www.investopedia.com/: http://www.investopedia.com/university/ratios/.
25. M.Stein (2002). Benchmarking Default prediction models: Pitfalls and Remedies in Model validation. Technical report #020305.
26. MARTIN (1977). Early warning of bank failure: A logit regression approach . Journal of banking and finance 1 , 249-276.
27. Payam Refaeilzadeh, Lei Tang and Huan liu. (2008). Cross - Validation. Arizona State University.
28. Rodriguez (2007). http://data.princeton.edu/wws509/notes/c3.pdf - Logit models for binary Data. Retrieved from http://data.princeton.edu .
29. Ros, Westerfield and Jordan. Essentials of corporate finance (6 ed.).
30. The National academies press Retrieved from http://www.nap.edu: http://www.nap.edu/openbook.php?record_id=12717&page=1.
31. Thomajean Johnsen and Ronald W.Melicher. (1994). Predicting corporate Bankruptcy and financial Distress: Information Value added by Multinominal Logit Models. Journal of Economics and Business , 269-286.
32. Wikipedia. (2002). Retrieved 2010, from http://en.wikipedia.org: http://en.wikipedia.org/wiki/Overfitting.