., and Vanthienen, J. “Benchmarking state-of-the-art classification algorithms for credit scoring,” Journal of the Operational Research Society(54:6), June 2003, pp. 627-635.
[4] Balcaen, S., and Ooghe, H. “35 years of studies on business failure: an
mittee on Banking Supervision “Studies on the Validation of Internal
更 說
點 質性與模型評等效力驗證均優於TCRI；3)
2 類與分成 7 與 10 類的穩定性較
與穩定性均有較的驗證效果；5)違約前二季的樣本資料以及評等級數設定 在 7 級的績效更好，EMCRS 模型可依使用者需求設定評等級數，找出適合的 信用風險評等模型；6)財務變數被選到最多是獲利能力指標(如：資產報酬率(稅 後息前)、資產報酬率(稅後息前折舊前)、淨值報酬率(稅後)、營業毛利率、業 外收支率，其次是每股收益率指標(如：常續性 EPS、每股營業利益、每股稅前 淨利)以及成長能力(如：總資產成長率、總資產報酬成長率)與償債能力(如：
內部保留比率、稅前純益/實收資本)，足見EMCRS 評等模型受整體獲利、每 股收益、成長力以及償債力等變數的影響較大；7)公司治理因素中董監的酬勞 比重與持股以及財測公佈與高階管理者異動均是影響信用風險評等的重要因 素。未來研究建議朝幾個方向進行：配合其他統計變數挑選方式事前檢定後再 加入 EMCRS 或許可能進一步改善評等效能；樣本的數量增加對於模型的建 立，更具代表性；加入其它國際金融評等機構的信用評等模型做比較。
[1] Altman, E. I. “Financial ratios, discriminant analysis and the prediction of corporate bankruptcy,” Journal of Finance(23:4), September 1968, pp.
589-609.
[2] Back, B., Laitinen, T., and Sere, K. “Neural Networks and Genetic Algorithms for Bankruptcy Predictions,” Expert Systems with Applications(11:4), 1996, pp.
407-413.
[3] Baesens, B., Gestel, T. V., Viaene, S., Stepanova, M., Suykens, J
overview of the classic statistical methodologies and their related problems,”
The British Accounting Review(38:1), March 2006, pp. 63-93.
[5] Basel Com
Rating Systems” Basel committee on banking supervision, working paper, No.
14., February 2005.
[6] Deakin, E. B. “A discriminant analysis of predictors of business failure,”
mputer Systems(15:3), April 1999, pp. 433-441.
the US local
aptation in Natural and Artifical Systems,” Society for iley & Sons,
al of Business ction of
“The discovery of experts' decision rules from
recast Financial Distress,” Knowledge Based Systems(19:1),
national Review of Financial
for ordinal response
Failures,” The Journal of
prediction,” Expert Systems With Applications(31:3), Journal of Accounting Research(10:1), Spring 1972, pp. 167-179.
[7] Donate, J. M., Schryver, J. C., Hinkel, C. C., Schmoyer Jr., R. L., Leuze, M. R., Grandy, N. W. “Mining multi-dimensional data for decision support,” Future Generation Co
[8] Elizabeth M., "Handbook of credit scoring: the basics of scorecard development and validation", Lessons Professional Publishing, 2001.
[9] Foreman, R. D. “Logistic analysis of bankruptcy within
telecommunications industry,” Journal of Economics and Business(55:2), March 2003, pp. 135-166.
[10] Holland, J. H. “Ad
Industrial and Applied Mathematics(18:3),1975, pp. 529-530.
[11] Hosmer, D.W., Lemeshow, S. Applied logistic regression, John W New York,1989.
[12] Huyghebaert, N., Gaeremynck Ann, R. F., and Van De Gucht, L. M. “New Firm Survival ： The Effect of Start-up Characteristics, ＂ Journ
Finance and Accounting(27), 2000, pp.627-651.
[13] Keasey, K., and Watson, R.“Non-Financial Symptoms and The Predi
Small Company Failure：A Test of ARGENTIs Hypotheses,＂ Journal of Business Finance and Accounting(14), 1987, pp.335-354.
[14] Kim, M. J., and Han, I.
qualitative bankruptcy data using genetic algorithms,” Expert Systems with Applications(25:4), November, 2003, pp. 637-646.
[15] Ko, P. C., and Lin, P. C. “An Evolution-Based Approach with Modularized Evaluations to Fo
March 2006, pp.84-91.
[16] Laitinen, E. K., and Laitinen, T. “Bankruptcy prediction application of the Taylor’s expansion in logistic regression,” Inter
Analysis(9), 2000, pp.327-349.
[17] Laster, D., "Insurance Company Ratings", Sigma Swiss Re (4), 2003, pp. 23-39.
[18] Limin, F., and Douglas, G. S. “Conditional risk models
data: simultaneous logistic regression analysis and generalized score tests,”
Journal of Statistical Planning and Inference(108), October 2002, pp. 201-217.
[19] Meyer, P. A., Pifer, H. W. “Prediction of Bank Finance(25:4), Septmeber 1970, pp. 853-868.
[20] Min, S. H., Lee, J., and Han, I. “Hybrid genetic algorithms and support vector machines for bankruptcy
October 2006, pp. 652-660.
[21] Ohlson, J. A. “Financial ratios and the probabilistic prediction of bankruptcy,”
Journal of Accounting Research(18:1), Spring 1980, pp.109-131.
search, Vol.
ent International Review(4), 1966, pp. 15-21.
nd Aksoy, H. “Prediction of corporate financial distress in an
H., Tzeng, G. H., Goo, Y. J., and Fang, W. C. “A real-valued genetic [22] Ran, B., Abraham, M., and Douglas, M. K. “Confidence intervals for
controlling the probability of bankruptcy,” Omega(28:5), October 2000, pp.
555-565.
[23] Shin, K. S., and Lee, Y. J. “A genetic algorithm application in bankruptcy prediction modeling,” Expert Systems with Applications(23:3), October 2002, pp. 321-328.
[24] Shin, K. S., and Han, I. “A case-based approach using inductive indexing for corporate bond rating,” Decision Support Systems(32:1), November 2001, pp.
41-52.
[25] Sjur, W. and Nico, V. D. W. “Default probabilities in a corporate bank portfolio:
a logistic model approach,” European Journal of Operational Re 135(2), December 2001, pp. 338-349.
[26] Tamari, M. “Financial ratios as a means of forecasting bankruptcy,”
Managem
[27] Tseng, F. M., and Lin, L. “A quadratic interval logit model for forecasting bankruptcy” Omega(33:1), February 2005, pp. 85-91.
[28] Ugurlu, M., a
emerging market: the case of Turkey,” Cross Cultural Management: An International Journal(13:4), 2006, pp. 277-295.
[29] Wu, C.
algorithm to optimize the parameters of support vector machine for predicting bankruptcy,” Expert Syst ems With Applications(32:2), February 2007, pp.
397-408.
附錄一
【表8】變數表
自變數 變數名稱 自變數 變數名稱
財務變數
x1 資產報酬率(稅前息前折舊前)% x31 稅後淨利成長率%
x2 資產報酬率(稅後息前) % x32 經常淨利成長率─稅後%
x3 資產報酬率(稅後息前折舊前) % x33 常續性利益成長率─稅後%
x4 淨值報酬率(稅後) % x34 總資產成長率%
x5 淨值報酬率－常續利益% x35 淨值成長率%
x6 營業毛利率(常續性利益) % x36 折舊性固定資產成長率%
x7 已實現銷貨毛利率% 總資產報酬成長率%
x8 營業利益率% 內部保留比率%
率% x39 流動比率%
%
x11 業外收支率%每股營業利益 1 利息支出率%
元） 本%
元） 收資本%
） 款/淨值
次）
)
元） 次)
元） (次)
) 毛利成長率%
） 基本資料變
x37 x38 x9 稅前淨利
x10 稅後淨利率 x40 速動比率%
x4
x12 常續性利益率(稅後)% x42 總負債/總淨值%
x13 員工人數(人) x43 負債比率%
x14 營業費用率% x44 淨值/總資產%
x15 現金流量比率% x45 長期資金適合率(A)%
x16 有息負債利率% x46 借款依存度%
x17 稅率(A)% x47 或有負債/淨值%
x18 每股淨值(B)（元） x48 利息保障倍數 x19 每股淨值(A)（ x49 營業利益/實收資 x20 每股淨值(C)（ x50 稅前純益/實 x21 常續性EPS（元 x51 存貨及應收帳 x22 每股現金流量（元） x52 總資產週轉率（
x23 每股營業額（元） x53 應收帳款週轉率(次 x24 每股營業利益（ x54 存貨週轉率(
x25 每股稅前淨利（ x55 固定資產週轉率
x26 營收成長率% x56 淨值週轉率(次) x27 營業毛利成長率% x57 淨營業週期(天
x28 已實現銷貨 x58 淨收款週轉率%
x29 營業利益成長率% x59 每人營收（千元
x30 稅前淨利成長率% x60 每人營業利益（千元）
數 x62 公司年齡
會計師變數 x63 是否更換會計師事務所
總體經濟變數 x64 台灣地區失業率
x65 基準放款利率
x66 本國一般銀行對民營事業放款餘額 x67 臺灣加權股價指數月均值
x68 出超
公司治理變數
x69 調整後負債權益比 x89 盈餘分配%
x70 董監持股% x90 股份盈餘偏離倍數
x71 大股東持股% 席次盈餘偏離倍數 92 次股份偏離倍數
x73 家族總持股% 超額持股%
94 係人銷貨%
95 係人財產交易損益%
96 測次數 97 報重編次數 佔董監事席次比% 98 事長異動次數
x79 外部人士佔董監事席次比 總經理異動次數
次比% 次數
理
淨利% %
酬勞百萬
%
資 x91
x72 經理人持股% x 席
x93
x74 外部持股% x 關
x75 股權集中度 x 關
x76 法人持股% x 財
x77 自然人持股% x 財
x78 經理人 x 董
% x99
x80 獨立董監佔董監事席 x100 財務主管異動
x81 董事長兼總經 x101 發言人異動次數
x82 監察人內部化 x102 內部稽核異動次數
x83 董監質押% x103 員工流動%
x84 董監酬勞佔稅前 x104 轉投資佔淨值
x85 平均每位董監 x105 轉投資佔資產%
x86 董監席次控制 x106 背書保證%
x87 控制持股% x107 員工平均年
x88 直接持股% x108 是否為集團公司
研究成果
本計劃是透過信用風險評估機制的建置，建立一客觀的評估系統，有效 管控企業信用風險。本計劃在研究過程中提出一個多重組合羅吉斯迴歸 模型模型結合遺傳演算法，根據不同企業進行風險值預測並評等，以便
相關授信、貸款額度設定。後續研究 SOA)，建
立一更具彈性與成本效率的互通平台，其中 能有效整
合各功能模組與公司內部間資訊；透過整合、分享的方式，使得公司內 部的文件資料達到即時標準化，提升公司整體控管與溝通績效。本計劃
的 投稿到資管學報，目 入
自評:
可以運用服務導向架構(
Web Service 技術
結案報告已經 前已進 修改階段。