臺灣博碩士論文加值系統

近年財務危機對公司之影響間接震盪了全球經濟，並逐漸受到重視；若組織之債務無法履行亦或履行困難，即會被認定為財務危機。本研究以群啟發演算法(SIP)探究營建產業中發生財務危機公司之特徵，並計算其權重，用以分析之各項財務比率提供足夠之財務資訊，能供投資者與分析者評斷該企業運行之狀態。本研究提供管理者一個警惕，部分未被重視之特徵仍具有其重要性；透過上述特徵，即能使用SIP作為預測財務危機的新分析工具。本研究同時併用PCA對特徵之權重進行調整，確保其適用於營建相關公司，最終定義出可能造成財務危機之主要變數。本研究所取得之資料包含發生財務危機與未發生財務危機之建設公司資料，並著重發生危機公司之分析，其中包含了55家公司1998至2008有效財務報表共1615筆。利用25項財務比例已PCA單獨進行演算，差異百分比為73.4%。而在分析前以SIP進行分群，再使用PCA分類，即得3群之差異百分比，分別為91.4%、88.5%、93.5%，平均為90%，高於原先單獨分析之73.4%。

Financial crisis has raised concerns for years and its effect on companies influence economies globally. Financial distress of an organization is defined as a condition where obligations are not met, or are met with difficulty. This study explores and weights features for Financially Distressed Companies in the Construction Industry using Swarm Inspired Projection (SIP) algorithm. The financial ratios involved provide useful quantitative financial information to both investors and analysts so that they can evaluate the operation of a firm and analyze its position within a sector over time. This research brings awareness to managers as to which features they have to focus on at the same time not neglecting other features. All the ratios involved, each play a crucial role. It employs the SIP algorithm as a new analysis tool for forecasting financial distress. In this paper, the SIP algorithm is combined with the Principal Component Analysis (PCA) to determine the weights of the features and to adjust these weights to suit the profitability of these construction companies. The analyses identifies the most likely variables responsible for financial distress in these companies. It makes use of 25 different ratios; profit margin, return on assets, after tax rate of return, operating profit to after-tax rate of return, operating profit to paid-in capital ratio, earning per share, operating margin, revenue growth rate, growth rate of total assets, equity ratio, receivables turnover, total assets turnover etc., covering a maximum of fifty five construction companies in Taiwan.

The data available in this study covers both the ‘Failed’ and ‘non-failed’ construction companies but the focus is on failed construction companies. The combination of the two techniques used in this research not only identifies the parameters or features responsible for financial distress but also enhances the variance percentage of the results obtained. The variance percentage in this case measures the percentage variability of the ratios in the selected components with the rest of the other components. The study made use of 1615 effective financial reports from the 55 construction companies over the last decade (1998-2008) for the analysis in this paper. The data used is derived from the Taiwan Economic Journal (TEJ) which provides accurate and reliable data on companies throughout Asia. Based on the 25 ratios used, the PCA, without incorporating the SIP algorithm, initially achieved a variance percentage of 74.3%. Incorporating the SIP algorithm model in the analysis first to cluster, then the PCA to classify the data, raised variance percentages in the three clusters 1, 2, and 3 respectively, enhancing performance to 91.4%, 88.5%, and 89.3%. This gives us an average of 90%.
This method, compared to other methods most commonly used in financial analysis provides better reliability in the identification of the principal features in bankruptcy analysis. Corporate financial distress is a major concern to business sectors worldwide, therefore using both clustering and statistical techniques in unison is a better basis in mitigating bankruptcy to both practitioners and researchers.

Abstract II
1. Introduction XII
2. Literature review 5
3. Methodology 9
3.1A. Financial Ratios used in the Research 11
3.1B. Data Collection (Basic Analysis) 13
3.2 Clustering 19
3.3 SIP 19
A. DSOM Algorithm; 20
B. Estimation of the Cluster Number. 25
3.4 Feature Weighting (PCA) Data Reduction 36
4. Adjustment (Partial Adjustment Model) 57
4.1: Weights and their explanations 65
5. Conclusion 67
5.1. Contributions to Research 68
5.2. Suggestions 68
References 70
Appendix 75

TABLE OF FIGURES
Figure 1 RESEARCH PROCEDURE 10
Figure 2 descriptive STATISTICS 13
Figure 3 PROFITABILITY RATIOS 15
Figure 4 LEVERAGE RATIOS 16
Figure 5 POSITION OF THE DOVES BEFORE THE CRUMBS WERE INTRODUCED. 22
Figure 6 THE MOVEMENT OF 25 DOVES ON A TWO DIMENSIONAL ARTIFICIAL GROUND AFTER THE CRUMBS WERE INTRODUCED. 27
Figure 7 CLUSTER PIE CHART 27
Figure 8 MEANS AND STANDARD DEVIATIONS OF STRONG RATIOS 29
Figure 9 MEANS AND STANDARD DEVIATIONS OF WEAK RATIOS 31
Figure 10 PCA PROCEDURE 37
Figure 11 95% CONFIDENCE INTERVAL WITH ERROR BARS. 38
Figure 12 SCREE PLOT CLUSTER 1, 2 AND 3 RESPECTIVELY 42

List of tables
Table 1 Financial Ratios 12
Table 2 Descriptive statistics 17
Table 3 Means, standard deviations and weights of strong ratios 28
Table 4 Means, standard deviations and weights of weak ratios 30
Table 5 The initial means, standard deviations and weights of cluster 2. 33
Table 6 the initial means and standard deviations of cluster 1. 34
Table 7 the initial means and standard deviations of cluster 3. 35
Table 8 Total Variance Explained table for cluster 1. 39
Table 9 Total Variance Explained table for cluster 2. 40
Table 10 Total Variance table for cluster 3. 41
Table 11 KMO and Bartlett's Test 42
Table 12 Monte Carlo PCA for Parallel Analysis 44
Table 13 Communalities 46
Table 14 Extraction Method: Principal Component Analysis. Pattern/structure for coefficients cluster 1. 48
Table 15 Extraction Method: Principal Component Analysis. Pattern/structure for coefficients cluster 2. 49
Table 16 Extraction Method: Principal Component Analysis. Pattern/structure for coefficients cluster 3 50
Table 17 Extraction Method: Principal Component Analysis. Rotation Method: Varimax with Kaiser Normalization. Cluster 1 51
Table 18 Extraction Method: Principal Component Analysis. Rotation Method: Varimax with Kaiser Normalization. Cluster 2 53
Table 19 Extraction Method: Principal Component Analysis. Rotation Method: Varimax with Kaiser Normalization. Cluster 3 55
Table 20 The weights and the weighted means and standard deviations 59
Table 21 Adjusted weights; means and standard deviations 62
Table 22 Adjusted weights, means and standard deviations 64
Table 23 The adjusted weights, means and standard deviations 65

Appendix a
Appendix a 1 TABLE 1.A. CORRELATION MATRIX FOR CLUSTER 1 75
Appendix a 2 TABLE 2.A. CORRELATION MATRIX FOR CLUSTER 2 79
Appendix a 3 TABLE 3.A. CORRELATION MATRIX FOR CLUSTER 3 83

APPENDIX B
APPENDIX B 1 TABLE 1.B INITIAL COMPONENT SOLUTION FOR CLUSTER 1 88
APPENDIX B 2 TABLE 2.B INITIAL COMPONENT SOLUTION FOR CLUSTER 2 89
APPENDIX B 3 TABLE 3.B INITIAL COMPONENT SOLUTION FOR CLUSTER 3 90

APPENDIX C
APPENDIX C 1TABLE 1.C CASE PROCESSING SUMMARY 91
APPENDIX C 2 TABLE 2.C RELIABILITY STATISTICS 91
APPENDIX C 3 TABLE 3.C ITEM STATISTICS 91
APPENDIX C 4 TABLE 4.C ITEM-TOTAL STATISTICS 92
APPENDIX C 5 Table 5.C Scale Statistics 92


Equations
………………………………... (2) 23
Step 6. …….…… (3) 23
……………………………………………………. (4) 23
…………………… (5) 24
………………….… (6) 24
…………………………. (7) 24
…………………………………………………………………… (8) 24
……………………………………………………… (9) 24
y_t=1/n ∑_(i=1)^n▒(y_it ) ………………………….…………………………………………….… (10) 57
y_it-y_(it-1)=α_(i+) β_i (y_t^*-y_(it-1))+Є_it with Є_it≈N(0,σ_(Є_it)^2 )……………………...……. (11) 57
(y ̅_t-y ̅_(t-1) ) =α ̅+β ̅(y_t^*-y ̅_(t-1) )+ε ̅_t ε ̅_t≈(0,(σ_t^2)/N)………………...…………………. (12) 58
y ̅_t=1/N ∑_(i=1)^N▒y_it ……………………………………………………………….…………... (13) 58
y ̅_(t-1)= 1/N ∑_(i=1)^N▒y_(it-1) ……………………………………………….……………………... (14) 58
ε ̅_t=1/N ∑_(i=1)^N▒ε_it …………………………………………………….……………………….(15) 58
β ̅=1/N ∑_(i=1)^N▒β_i ………………………………………………………………..............……. (16) 58

References

1. Mokhatab Rafiei, F., S.M. Manzari, and S. Bostanian, Financial health prediction models using artificial neural networks, genetic algorithm and multivariate discriminant analysis: Iranian evidence. Expert Systems with Applications, 2011. 38(8): p. 10210-10217.
2. Chaudhuri, A. and K. De, Fuzzy Support Vector Machine for bankruptcy prediction. Applied Soft Computing, 2011. 11(2): p. 2472-2486.
3. West, D., S. Dellana, and J. Qian, Neural network ensemble strategies for financial decision applications. Computers &; Operations Research, 2005. 32(10): p. 2543-2559.
4. Chen, M.-Y., Predicting corporate financial distress based on integration of decision tree classification and logistic regression. Expert Systems with Applications, 2011. 38(9): p. 11261-11272.
5. Lin, F., D. Liang, and E. Chen, Financial ratio selection for business crisis prediction. Expert Systems with Applications, 2011. 38(12): p. 15094-15102.
6. Wu, D., L. Liang, and Z. Yang, Analyzing the financial distress of Chinese public companies using probabilistic neural networks and multivariate discriminate analysis. Socio-Economic Planning Sciences, 2008. 42(3): p. 206-220.
7. Martikainen, T., K. Puhalainen, and P. Yli-Olli, On the industry effects on the classification patterns of financial ratios. Scandinavian Journal of Management, 1994. 10(1): p. 59-68.
8. Achim, M.V., C. Mare, and S.N. Borlea, A Statistical Model of Financial Risk Bankruptcy Applied for Romanian Manufacturing Industry. Procedia Economics and Finance, 2012. 3: p. 132-137.
9. Mironiuc, M. and I.-B. Robu, Empirical Study on the Analysis of the Influence of the Audit Fees and Non Audit Fees Ratio to the Fraud Risk. Procedia - Social and Behavioral Sciences, 2012. 62: p. 179-183.
10. Horrigan, J.O., Some Empirical Bases of Financial Ratio Analysis. The Accounting Review, 1965. 40(3): p. 558-568.
11. Beaver, W.H., Financial Ratios As Predictors of Failure. Journal of Accounting Research, 1966. 4: p. 71-111.
12. Tsai, C.-F., Combining cluster analysis with classifier ensembles to predict financial distress. Information Fusion, 2014. 16: p. 46-58.
13. Chen, M.-Y., A hybrid ANFIS model for business failure prediction utilizing particle swarm optimization and subtractive clustering. Information Sciences, 2013. 220: p. 180-195.
14. Chen, N., et al., Clustering and visualization of bankruptcy trajectory using self-organizing map. Expert Systems with Applications, 2013. 40(1): p. 385-393.
15. De Andrés, J., et al., Bankruptcy forecasting: A hybrid approach using Fuzzy c-means clustering and Multivariate Adaptive Regression Splines (MARS). Expert Systems with Applications, 2011. 38(3): p. 1866-1875.
16. Enke, D., M. Grauer, and N. Mehdiyev, Stock Market Prediction with Multiple Regression, Fuzzy Type-2 Clustering and Neural Networks. Procedia Computer Science, 2011. 6: p. 201-206.
17. Huang, S.-C., Integrating spectral clustering with wavelet based kernel partial least square regressions for financial modeling and forecasting. Applied Mathematics and Computation, 2011. 217(15): p. 6755-6764.
18. Karimi, S. and B. Hemmateenejad, Identification of discriminatory variables in proteomics data analysis by clustering of variables. Analytica Chimica Acta, 2013. 767: p. 35-43.
19. Lai, R.K., et al., Evolving and clustering fuzzy decision tree for financial time series data forecasting. Expert Systems with Applications, 2009. 36(2, Part 2): p. 3761-3773.
20. Sun, J. and H. Li, Data mining method for listed companies’ financial distress prediction. Knowledge-Based Systems, 2008. 21(1): p. 1-5.
21. Chen, J.-H., A hybrid knowledge-sharing model for corporate foreign investment in China’s construction market. Expert Systems with Applications, 2012. 39(9): p. 7585-7590.
22. dos Santos, D.S. and A.L.C. Bazzan, Distributed clustering for group formation and task allocation in multiagent systems: A swarm intelligence approach. Applied Soft Computing, 2012. 12(8): p. 2123-2131.
23. Gajawada, S. and D. Toshniwal, Projected Clustering Using Particle Swarm Optimization. Procedia Technology, 2012. 4: p. 360-364.
24. Hsieh, T.-J., H.-F. Hsiao, and W.-C. Yeh, Mining financial distress trend data using penalty guided support vector machines based on hybrid of particle swarm optimization and artificial bee colony algorithm. Neurocomputing, 2012. 82: p. 196-206.
25. Huang, C.-L., et al., Hybridization strategies for continuous ant colony optimization and particle swarm optimization applied to data clustering. Applied Soft Computing, 2013. 13(9): p. 3864-3872.
26. Shen, W., et al., Forecasting stock indices using radial basis function neural networks optimized by artificial fish swarm algorithm. Knowledge-Based Systems, 2011. 24(3): p. 378-385.
27. Tsai, C.-Y. and I.W. Kao, Particle swarm optimization with selective particle regeneration for data clustering. Expert Systems with Applications, 2011. 38(6): p. 6565-6576.
28. Su, M.-C., S.-Y. Su, and Y.-X. Zhao, A swarm-inspired projection algorithm. Pattern Recognition, 2009. 42(11): p. 2764-2786.
29. Wang, Y.-J. and H.-S. Lee, A clustering method to identify representative financial ratios. Information Sciences, 2008. 178(4): p. 1087-1097.
30. Aielli, G.P. and M. Caporin, Variance clustering improved dynamic conditional correlation MGARCH estimators. Computational Statistics &; Data Analysis, ().
31. Aviad, B. and G. Roy, Classification by clustering decision tree-like classifier based on adjusted clusters. Expert Systems with Applications, 2011. 38(7): p. 8220-8228.
32. Pattarin, F., S. Paterlini, and T. Minerva, Clustering financial time series: an application to mutual funds style analysis. Computational Statistics &; Data Analysis, 2004. 47(2): p. 353-372.
33. Wang, Y.-J., A clustering system for data sequence partitioning. Expert Systems with Applications, 2011. 38(1): p. 659-666.
34. Chen, X., et al., A feature group weighting method for subspace clustering of high-dimensional data. Pattern Recognition, 2012. 45(1): p. 434-446.
35. Chen, W.-S. and Y.-K. Du, Using neural networks and data mining techniques for the financial distress prediction model. Expert Systems with Applications, 2009. 36(2, Part 2): p. 4075-4086.
36. Li, H. and J. Sun, Empirical research of hybridizing principal component analysis with multivariate discriminant analysis and logistic regression for business failure prediction. Expert Systems with Applications, 2011. 38(5): p. 6244-6253.
37. Li, H. and J. Sun, Principal component case-based reasoning ensemble for business failure prediction. Information &; Management, 2011. 48(6): p. 220-227.
38. When is a correlation matrix appropriate for factor analysis? Some decision rules. 1974, American Psychological Association: US. p. 358-361.
39. Cerny, B.A. and H.F. Kaiser, A Study Of A Measure Of Sampling Adequacy For Factor-Analytic Correlation Matrices. Multivariate Behavioral Research, 1977. 12(1): p. 43-47.
40. Kaiser, H., A second generation little jiffy. Psychometrika, 1970. 35(4): p. 401-415.
41. Tobias, S. and J.E. Carlson, BRIEF REPORT: BARTLETT'S TEST OF SPHERICITY AND CHANCE FINDINGS IN FACTOR ANALYSIS. Multivariate Behavioral Research, 1969. 4(3): p. 375-377.
42. Cattell, R.B., The Scree Test For The Number Of Factors. Multivariate Behavioral Research, 1966. 1(2): p. 245-276.
43. Thurstone, L.L., Thurstone, L. L. Multiple‐factor analysis. Chicago: University of Chicago Press, 1947, pp. 535. $7.50. Journal of Clinical Psychology, 1947. 4(2): p. 224-224.
44. Lev, B., Industry Averages as Targets for Financial Ratios. Journal of Accounting Research, 1969. 7(2): p. 290-299.
45. Gallizo, J.L., P. Gargallo, and M. Salvador, Multivariate Partial Adjustment of Financial Ratios: A Bayesian Hierarchical Approach. Journal of Applied Econometrics, 2008. 23(1): p. 43-64.
46. Wu, C. and S.-J. Ho, Financial Ratio Adjustment: Industry-Wide Effects or Strategic Management. Review of Quantitative Finance and Accounting, 1997. 9(1): p. 71-88.
47. Davis, H.Z. and Y.C. Peles, Measuring Equilibrating Forces of Financial Ratios. The Accounting Review, 1993. 68(4): p. 725-747.
48. Peles, Y.C. and M.I. Schneller, The Duration of the Adjustment Process of Financial Ratios. The Review of Economics and Statistics, 1989. 71(3): p. 527-532.