201227580 六、發明說明： 【發明所屬之技術領域】 本發明係有關於一種企業融資與風險評估方法，特別 是有關於一種應用電腦系統進行企業融資與風險評估方 法。 【先前技術】 隨著社會的進步，個人與企業對於資金周轉的需求， 與時遽增。信用融資的出現’說明了人們重視信用，這對 於提高人們對信用的重視，無疑具有劃時代的㈣。對於 -個先進的社會，信用的約束力必須大幅的提高以提升 個人與企業的信用，常言說得好：“人無信不立”。 和社會可能都擔心信用融資的風險， 用何能有效的管理個人與企業的作 用，並使得金_構進行借貸時的風險降低，不僅可= 金融機構的放款利潤，更可以降低個人或企業借 由於新興國家的企#多為家族方式經營 務報表數字常會透Μ同的方法進行㈣，藉 ^財 上資料，因而大幅地降低傳統僅考慮財務 料:報表 之財務危機預警模型的預測效力。特狀，斜=所建揭 上市公！的財務報表’更因為欠缺有效的管理士概或 現實脫節，以致於可信度堪慮。 、常常與 如何此有效地建立融資風險評估的正確性 對財務狀況不是很透明的公司或企業，做出=別是對 嘴的融資決 4 201227580 定，為金融機構所努力的方向。 【發明内容】 鑒於上述之先前技術中所述，由於金融機構在進行融 資時，需能正確地判斷借款人的還款能力，以減少風險， 因此，如何能由借款人的資料判斷其還款能力，並設計合 適的還款方式，將可有有效地降低融資風險。 本發明之目的之一，係提供一種企業融資與風險評估 Φ 方法，使用於一電腦系統，以有效地對一新的客戶進行違 約風險評估’並決定融資的模式與架構。 根據以上所述之目的，本發明係揭露一種企業融資與 風險評估方法，使用於一電腦系統，可有效地對一新的客 戶進行違約風險評估，並決定融資的模式與架構。此企業 融資與風險評估方法包含有下列步驟： 將歷史評分資料輸入一電腦系統之一資料庫； 利用該電腦系統，計算顯著變數； φ 利用該電腦系統，建立判別模式函數； 利用該電腦系統，建立存活時間函數； 利用該電腦系統，建立違約風險係數分配表； 輸入一新客戶的評分表；以及 計算新客戶的違約風險。 其中上述之歷史評分資料包含財務項目與非財務項目 的評分項目。而上述非財務項目的評分項目更包含經營管 理項目、經濟評分項目與規模項目的評分項目。 201227580 此止業融資與風險評估方法更包含區別歷史評分資料 的組別’以分別進行資料統計。上述之歷史評分資料的組 別可以分為服務業、製造業與具有財務簽證之企業等。 卜上述之判別模式函數更包含一違約判別模式函 模式函數。當違約判別模式函數值大於正 电判別模式函數值時，新的客戶被判定會違約。 當新的客户被判定會違約時，本發明之企 方法更利用存活時間函數，以計算一存活時Π 、違約風險係數分配表，再根據客戶融資額 度決疋一适約金額。 、 之企業融資與風險評估方法更可根據存活時間 =約金額調整新的客戶的-融資合約。上述之調整融資 整一還款期限與調整-融資額度，且調整融 貝額度更包含設計一非直線性還款模式。 因此n$本發明之企業融資與風險評估方法 3效違約風險，並藉由曰積月累的數據 ’ 2 6貝料’以更精確地預估客戶違約的風 險此外，由於本發明不僅根據財務項目的評分項目 項目的評分項目，更可以充分反應企業經營狀 L的真實性，以增加系統預測的準確 評估方法預估新客戶有違約風險== 可根據ΐ =的時間與違約風險’進而計算違約金額，並 融資合存科㈣整還款方式，才與客戶簽訂 二==資:r且擴大可融資的範圍，增 的s業&®。此外’根據不同的違約金額與存 201227580 活時間，更可以採用非線性的還款模式，以進一步降低融 資風險。 【實施方式】 本發明係揭露一種企業融資與風險評估方法，可有效 地評估融資的風險、預估違約的時程，並根據預估違約的 時程與金額制訂合適的還款計晝，以有效地降低金融機構 融資的風險。以下將以圖示及詳細說明清楚說明本發明之 精神，如熟悉此技術之人員在瞭解本發明之較佳實施例 後，當可由本發明所教示之技術，加以改變及修飾，其並 不脫離本發明之精神與範圍。 參閱第1圖，其係繪示本發明之企業融資與風險評估 方法之示意圖。本發明之企業融資與風險評估方法，利用 已知的歷史評分資料與實際還款結果等資料，建立判別模 式函數與存活時間函數，並可不斷地根據新的數據更新資 料庫，以建立更符合實際情況的判別模式函數與存活時間 函數。如圖所示，步驟110，首先將歷史評分資料輸入一 電腦系統之中。步驟120，將輸入的歷史評分資料根據融 資用戶的特性區分為不同的組別。其中，組別可以分為製 造業、服務業及具有財務簽證之企業。分別使用不同的表 格進行填寫，以符合不同組別的特性。例如是，具有財務 簽證之企業可填寫一曱表，而未具有財務簽證之服務業與 製造業可填寫一乙表，再將其區分為不同的組別如乙表製 造業與乙表服務業，亦即，同時包含步驟110與步驟120 的功能，其並不脫離本發明之精神與範圍。 201227580 舉例而言，具有財務簽證之企業所使用之曱表可至少 包含下列的評分項目與其分類的代碼： 甲表： 代碼 評分項目 代碼 評分項目 N1 公司歷史 k9 平均銷貨天數 N2 公司内部是否和諧與員工忠誠度 klO 毛利率 N3a 經營者與保證人之背景與經營理念 kll 淨利率 N3b 經營者與保證人資力 kl2 淨值週轉率 N4 財務報表是否詳實可信 kl3 每股淨值成長率 kl 自有資本率 kl4 營業額成長率 k2 負債比率 N6 重大法規及政策對其影響 k3 固定比率 N7 國内外經濟因素演變對其影響 k4 流動比率 N8 產業展望 k5 速動比率 N9a 公司之經營展望（1·生產） k6 債務償債能力-DSR N9b 公司之經營展望（2.銷售） k7 平均淨值週轉率 N9c 公司之經營展望（3.經營團隊） k8 平均收款天數 N10 同業及客戶對其評價 針對未具有財務簽證之服務業則可使用項目較少的乙 表，但由於一般製造業與服務業即便根據相同的評分項 目，但在實際進行評估時，兩者具有截然不同的結果，故 在實際進行分析時，仍採用不同的組別，以提升預估的準 確性。 乙表： 201227580 代碼 評分項目 代碼 評分項目 P1 公司歷史 f6 營業額成長率 P2 公司内部是否和諳與員工忠誠度 f7 營業額/損益平衡點 P3a 經營者與保證人之背景與經營理念 P5 重大法規及政策對其影響 P3b 經營者與保證人資力 P6 國内外經濟因素演變對其影響 fl 債務償債能力-DSR P7 產業展望 f2 金融負債比 P8a 公司之經營展望〇·生產） β 平均收款天數 P8b 公司之經營展望（2.銷售） f4 毛利率 P8c 公司之經營展望（3.經營團隊） f5 淨值報酬率 P9 同業及客戶對其評價 • 其中，評分項目除了包含有財務項目的評分項目，更 包含有非財務項目的評分項目。特別是針對中小型企業， 由於一般財務狀態並不透明，非財務項目的評分就顯的十 分的重要，其中非財務項目至少包含有經營管理項目的評 分、經濟項目的評分與規模項目的評分。而這些評分項目 更可以進一步利用歷史評分資料進行進一步的分類，例如 將公司歷史，依不同的經營年資，可區分為1至3分。例 如，成立1年以内為1分、1-5年為2分’以及5年以上為 φ 3分。又例如，有關於公司内部是否和諧與員工忠誠度的 評分，則可以依員工流動率等客觀因子進行評分。故上述 之非財務項目的評分可根據資料庫中的歷史評分資料訂定 相應的評分標準。 步驟130，當分組完成後，電腦系統將根據每一個債 務人的還款結果，分組別進行統計，以區別各組別的資料 的狀態。 步驟140,電腦系統進一步根據歷史評分資料、組別、 以及各組別的還款結果，利用逐步迴歸分析、迴歸分析、 201227580 逐步邏輯斯迴歸分析或邏輯斯迴歸分析等統計方法，計算 各組別的顯著變數，本發明之顯著變數較佳地包含有財務 評分項目與非財務評分項目。 步驟150，建立判別模式函數，以曱表為例根據被計 算出來的顯著變數，分別建立違約判別模式函數： y = -1.2772*kll + 1.7343*kl2+1.8044*kl0+1.0035*N8 +3.3259*Nl+0.1831*N3b+2.6197*N9b+3.2051*k7+ 0.8927*N4-14.0358 並建立正常判別模式函數： z = 5.2101*kll-2.5698*kl2+0.8724*kl0+2.0120*N8 +2.4201*Nl+0.7673*N3b+1.8157*N9b+4.2700*k7+ 1.3560*N4-20.6564 其中， 變數項目 違約判別模式 正常判別模式 kll 淨利率 -1.2272 5.2101 kl2 淨值週轉率 1.7343 -2.5698 klO 毛利率 1.8044 0.8724 N8 產業展望 1.0035 2.0120 N1 公司歷史 3.3259 2.4201 N3b經營者與保證人實力 0.1831 0.7673 N9b公司之經營展問（2.銷隹、 2.6197 1.8157 k7 ¥均淨值週轉率— 3.2051 4.2700 N4 財務報表是否詳實可作 0.8927 1.3560 Constant -14.0358 -20.6564 其中，當有新的客戶欲申請融資時，可藉由判別模式 函數分別代入所需的顯著變數的評分，取其大值為判定 值’亦即當y>z時此客> 會被判定為可能違約客戶，反之 則會被判定為正常客戶。 201227580 步驟160,接著同時建立存活時間函數，其可以準確 的預估存活月，例如 m = 0.787*Nl+0.935*N3a+0.621*kl+0.759* k8 調整後R2 = 0.931 其中， 代碼 評分項目 N1 公司歷史 N3a 經營者與保證人之背景與經營理念 K1 自有資本率 K8 平均收款天數 當上述之判別模式函數與存活時間函數分別建立完成 後，電腦可利用其他的融資合約的資料進行驗證，以評估 所建立的判別模式函數與存活時間函數是否達到一定的預 估準確率。其中，判別模式函數的歷史樣本的正確率需達 75%以上，而測試樣本正確率需達70%以上。並利用KS 值檢定，且要求KS值至少>30，以增加新客戶預估的準 確率。其中KS值為正常客戶評分的累積機率分配與違約 客戶評分的累積機率分配最大距離，且KS值愈大，越能 區別正常客戶與違約客戶。而存活時間函數則可分別使用 不同統計方法建立存活其模型，選用誤差值較小的函數。 步驟170，建立違約風險係數分配表，分組建立歷史 違約樣本的存活期資料統計分配檢定，找出過往存活期可 能的統計分配，以用來預估為可能發生的違約點的風險係 數。以18個月的統計資料可得，一違約風險係數分配表： 201227580 違約模 蛮斗B； 頁估存活期月I S〇 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 Gamm 嫌 (判別 爲違約 資料） R4 甲表S R7 100% 100% 99% 95% ΐω% 100% 100% 98% 100% 100% 98% 明嘮 87% 77% 66% 54% 43% 34% 25% 19% 14% 10% 7% 5% 3% 2% 2% 94% 87% 78% 67% 57% 46% 37% 29% 22% 16% 12% 9% 6% 4% 3% 85% 76% 65% SS% 4fi% 37% 30% Ύ\% 18% 14% 11% 8% 6% 5% R4 乙表^ R7 100% 100% 99% 96% !00%!00% 99% 97% !00%!00% 98% 95% 91% 83% 73% 63% 53% 43% 35% 27% 21% 16% 12% 9% 7% 5% 3% 92% 84% 75% 64% 54% 44% 35% 28% 21% 16% 12% 9% 6% 5% 3% 89% 82% 73% 64% 55% 47% 39% 32% 26% 21% 17% Π% 10% 8% R4 乙表製R5 造業 R6 R7 100% 100% 99% 95% 100% 100% 99% 97% 100% 100% 100% 98% im%inn%〇Q% 〇7% 89% 81% 72% 62% 53% 44% 36% 29% 23% 18% 14% 11% 8% 6% 5% 92% 85% 77% 67% 58% 48% 40% 32% 26% 20% 16% 12% 9% 7% 5% 94% 87% 79% 69% 59% 49% 40% 32% 25% 19% 15% 11% 8% 6% 4% 93% 88% 81% mn %% 48% 41% m 78% m 19% 15% 1?.% 10¾ R4 乙表服R5 務業R6 R7 100% 100% 96% me 76% 63% 50% 38% 29% 21% 15% 10% 7% 5% 3% 2% 1% 1% 1% Norma 分配< 判別1 正常f 料） 甲表 乙表製 造業 乙表服 mm 100% 100% 100% 1009 100%100%100%1009 100%100%100%100^ 100%100%100%100%99% 98% 96% 92% 85% 75% 64% 50% 37% 25% 15¾ 100%100%100%100%100%100%100%100%1継96% 83% 54% 23% 6% 1% 100% 100% 100% 100% 100% 99% 96% 90% 79% 63% 45% 28% 15% 6% 2%201227580 VI. Description of the invention: [Technical field to which the invention pertains] The present invention relates to a method for enterprise financing and risk assessment, and more particularly to an application computer system for enterprise financing and risk assessment methods. [Prior Art] With the advancement of society, the demand for capital turnover by individuals and businesses has increased. The emergence of credit financing has shown that people attach importance to credit, which is undoubtedly epoch-making for raising people's attention to credit (4). For an advanced society, the binding force of credit must be greatly improved to enhance the credit of individuals and enterprises. It is often said that "people do not believe in standing." And the society may be worried about the risk of credit financing, how to effectively manage the role of individuals and enterprises, and reduce the risk when the loan is borrowed, not only = the lending profits of financial institutions, but also the personal or corporate lending Since the enterprises in emerging countries are mostly in the family business, the number of business statements is often carried out in the same way (4), and the financial information is used to greatly reduce the forecasting effectiveness of the traditional financial crisis warning model. Special shape, oblique = built and unveiled listed public! The financial statements are more credible because of the lack of effective management or reality. Often, and how to effectively establish the correctness of the financing risk assessment, the company or the company that is not very transparent to the financial situation, make a decision on the financing of the financial institution. SUMMARY OF THE INVENTION In view of the above-mentioned prior art, since a financial institution needs to be able to correctly judge the borrower's repayment ability in order to reduce the risk, how can the borrower's information be used to determine the repayment amount. Capabilities and designing appropriate repayment methods will effectively reduce financing risks. One of the objects of the present invention is to provide a method for financing and risk assessment of enterprises Φ, which is used in a computer system to effectively conduct a risk assessment of a new customer' and to determine the mode and structure of financing. Based on the above, the present invention discloses a method for enterprise financing and risk assessment, which is used in a computer system to effectively perform a default risk assessment on a new customer and determine the mode and structure of financing. The enterprise financing and risk assessment method includes the following steps: inputting historical score data into a database of one computer system; using the computer system to calculate significant variables; φ using the computer system to establish a discriminant mode function; using the computer system, Establish a survival time function; use the computer system to establish a default risk factor allocation table; enter a new customer's score sheet; and calculate the new customer's default risk. The historical score data mentioned above includes the scoring items of financial items and non-financial items. The scoring items of the above non-financial items include the management project, the economic scoring project and the scoring project of the scale project. 201227580 This method of financing and risk assessment further includes a group that distinguishes historical score data' to conduct data statistics separately. The above historical score data can be divided into service industry, manufacturing industry and enterprises with financial visas. The discriminant mode function described above further includes a default discriminating mode function mode function. When the default discriminant mode function value is greater than the positive discriminant mode function value, the new client is judged to be in default. When a new customer is judged to be in default, the method of the present invention further utilizes the survival time function to calculate a survival time 违, default risk coefficient allocation table, and then determine a suitable amount based on the customer financing amount. The corporate finance and risk assessment method can also adjust the new client's financing contract based on the survival time = approx. The above-mentioned adjustment financing, the entire repayment period and the adjustment-financing amount, and the adjustment of the financing amount include the design of a non-linear repayment mode. Therefore, the invention financing and risk assessment method of the invention is three-way default risk, and by hoarding the accumulated data '26 6 shells' to more accurately estimate the risk of customer default. In addition, the present invention is not only based on finance The scoring project of the project's scoring project can fully reflect the authenticity of the business operation L, to increase the accurate estimation method of the system forecast, and estimate the new customer's default risk == according to the time of ΐ = and the risk of default' The amount of default, and the financing co-existence section (four) the entire repayment method, only signed with the customer two == capital: r and expand the scope of financing, increased s industry & In addition, according to the different default amount and the saving time of 201227580, a non-linear repayment mode can be adopted to further reduce the financing risk. [Embodiment] The present invention discloses a method for enterprise financing and risk assessment, which can effectively evaluate the risk of financing, estimate the time course of default, and formulate an appropriate repayment plan according to the time and amount of the estimated default. Effectively reduce the risk of financing financial institutions. The spirit and scope of the present invention will be apparent from the following description of the preferred embodiments of the invention. The spirit and scope of the present invention. Referring to Figure 1, there is shown a schematic diagram of the method of corporate finance and risk assessment of the present invention. The enterprise financing and risk assessment method of the invention uses the known historical score data and the actual repayment result to establish a discriminant mode function and a survival time function, and can continuously update the database according to the new data to establish a more conformance The discriminant mode function and the survival time function of the actual situation. As shown in the figure, in step 110, the historical score data is first input into a computer system. In step 120, the input historical score data is divided into different groups according to the characteristics of the financing user. Among them, the group can be divided into manufacturing, service industry and enterprises with financial visas. Use separate forms to fill in to match the characteristics of different groups. For example, a company with a financial visa can fill out a form, and a service industry and manufacturing industry that do not have a financial visa can fill out a form and then divide it into different groups, such as the manufacturing industry and the service industry. That is, the functions of step 110 and step 120 are included at the same time, without departing from the spirit and scope of the invention. 201227580 For example, a form used by a company with a financial visa can contain at least the following rating items and their classification codes: A table: Code rating item code rating item N1 Company history k9 Average sales days N2 Whether the company is harmonious internally Employee loyalty klO gross profit margin N3a Operator and guarantor background and business philosophy kll Net interest rate N3b Operator and guarantor capital kl2 Net value turnover rate N4 Financial statements are detailed and credible kl3 Net value per share growth rate kl Self capital ratio kl4 Turnover Growth rate k2 Debt ratio N6 Major regulations and policies affecting it k3 Fixed ratio N7 Domestic and foreign economic factors evolved impact k4 Current ratio N8 Industry outlook k5 Quick ratio N9a Company's business outlook (1·production) k6 Debt solvency -DSR N9b's business prospects (2.Sales) k7 Average net worth turnover rate N9c Company's business outlook (3.Management team) k8 Average collection days N10 peers and customers evaluate their services for non-financial services Use less B-tables, but due to general manufacturing The industry and the service industry, even if they are based on the same rating project, have very different results when actually conducting the assessment. Therefore, different groups are still used in the actual analysis to improve the accuracy of the estimation. Table B: 201227580 Code Scoring Project Code Scoring Project P1 Company History f6 Revenue Growth Rate P2 Internal Responsibility and Employee Loyalty in the Company f7 Turnover / Profit and Loss Balance P3a Background and Business Philosophy of Operators and Guarants P5 Major Regulations and Policies Impact on P3b Operators and guarantor resources P6 Domestic and foreign economic factors evolved impacts on debt Debt solvency - DSR P7 Industry Outlook f2 Financial liabilities ratio P8a company's business outlook 生产 · Production) β Average collection days P8b Company's operations Outlook (2. Sales) f4 Gross profit margin P8c Business outlook of the company (3. Management team) f5 Net rate of return P9 peers and customers evaluate it • Among them, the rating project includes not only financial projects, but also non-finance The rating project for the project. Especially for small and medium-sized enterprises, because the general financial status is not transparent, the scores of non-financial items are obviously important. The non-financial items include at least the scores of management projects, the scores of economic projects and the scale of scale projects. These scoring items can be further further classified by historical scoring data. For example, the company history can be divided into 1 to 3 points according to different business years. For example, it is 1 point within 1 year, 2 points for 1-5 years, and φ 3 points for 5 years or more. For example, if there is a score on whether the company is harmonious or employee loyalty, it can be scored according to objective factors such as employee turnover rate. Therefore, the scores of the above non-financial items can be determined according to the historical score data in the database. In step 130, after the grouping is completed, the computer system will perform statistics according to the repayment result of each debtor to distinguish the status of each group of data. In step 140, the computer system further calculates the respective groups by using statistical methods such as stepwise regression analysis, regression analysis, 201227580 stepwise logistic regression analysis or logistic regression analysis according to historical score data, groups, and repayment results of each group. Significant variables, the significant variables of the present invention preferably include financial rating items and non-financial rating items. Step 150: Establish a discriminant mode function, and use the table as an example to establish a default discriminant mode function according to the calculated significant variables: y = -1.2772*kll + 1.7343*kl2+1.8044*kl0+1.0035*N8 +3.3259*Nl +0.1831*N3b+2.6197*N9b+3.2051*k7+ 0.8927*N4-14.0358 and establish the normal discriminant mode function: z = 5.2101*kll-2.5698*kl2+0.8724*kl0+2.0120*N8 +2.4201*Nl+0.7673*N3b+ 1.8157*N9b+4.2700*k7+ 1.3560*N4-20.6564 Among them, variable item default judgment mode normal discrimination mode kll net interest rate -1.2272 5.2101 kl2 net value turnover rate 1.7343 -2.5698 klO gross profit rate 1.8044 0.8724 N8 industry outlook 1.0035 2.0120 N1 company history 3.3259 2.4201 N3b operator and guarantor strength 0.1831 0.7673 N9b company's business exhibition question (2. sales, 2.6197 1.8157 k7 ¥ net worth turnover rate - 3.2051 4.2700 N4 financial statement is available for 0.8927 1.3560 Constant -14.0358 -20.6564 which, when there is new When customers want to apply for financing, they can substitute the scores of the required significant variables by the discriminant mode function, and take the large value as the judgment value 'that is, when y>z > will be judged as a possible default customer, and vice versa will be judged as a normal customer. 201227580 Step 160, then establish a survival time function, which can accurately estimate the survival month, such as m = 0.787*Nl+0.935*N3a+ 0.621*kl+0.759* k8 After adjustment R2 = 0.931 Among them, code scoring project N1 company history N3a operator and guarantor background and business philosophy K1 own capital ratio K8 average collection days when the above discriminant mode function and survival time function After the establishment is completed, the computer can use other financing contract data to verify whether the established discriminant mode function and the survival time function reach a certain estimation accuracy. Among them, the correct rate of the historical sample of the discriminant mode function needs to be determined. More than 75%, and the test sample correct rate needs to reach more than 70%. And use KS value verification, and require KS value at least > 30, in order to increase the accuracy of new customer estimates. The KS value is the accumulation of normal customer ratings. The probability distribution and the cumulative probability of default customer ratings are assigned the maximum distance, and the larger the KS value, the more distinguishing between normal customers and defaulters. Household. The survival time function can be used to establish a surviving model using different statistical methods, and a function with a small error value is selected. In step 170, a default risk coefficient allocation table is established, and the survival data statistics allocation test of the historical default sample is grouped to find out the possible statistical distribution of the past survival period, and is used to estimate the risk factor of the possible default point. Available in 18 months of statistics, a default risk factor allocation table: 201227580 Default model B; Page estimated survival month IS〇1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 Gamm Suspect (identified as default information) R4 A table S R7 100% 100% 99% 95% ΐω% 100% 100% 98% 100% 100% 98% Alum 87% 77% 66% 54% 43% 34% 25% 19% 14% 10% 7% 5% 3% 2% 2% 94% 87% 78% 67% 57% 46% 37% 29% 22% 16% 12% 9% 6% 4% 3% 85% 76% 65% SS% 4fi% 37% 30% Ύ\% 18% 14% 11% 8% 6% 5% R4 Table B^ R7 100% 100% 99% 96% !00%!00% 99% 97% !00 %!00% 98% 95% 91% 83% 73% 63% 53% 43% 35% 27% 21% 16% 12% 9% 7% 5% 3% 92% 84% 75% 64% 54% 44% 35% 28% 21% 16% 12% 9% 6% 5% 3% 89% 82% 73% 64% 55% 47% 39% 32% 26% 21% 17% Π% 10% 8% R4 R5 Manufacturing R6 R7 100% 100% 99% 95% 100% 100% 99% 97% 100% 100% 100% 98% im%inn%〇Q% 〇7% 89% 81% 72% 62% 53% 44 % 36% 29% 23% 18% 14% 11% 8% 6% 5% 92% 85% 77% 67% 58% 48% 40% 32% 26% 20% 16% 12% 9% 7% 5% 94 % 87% 79% 69% 59% 49% 40% 32% 25% 19% 15% 11% 8% 6% 4% 93% 88% 81% mn %% 48% 41% m 78% m 19% 15% 1?. % 103⁄4 R4 B Service R5 Business R6 R7 100% 100% 96% me 76% 63% 50% 38% 29% 21% 15% 10% 7% 5% 3% 2% 1% 1% 1% Norma Distribution < Judgment 1 normal f material) A table and a table of the manufacturing industry B table service mm 100% 100% 100% 1009 100%100%100%1009 100%100%100%100^ 100%100%100%100%99% 98% 96% 92% 85% 75% 64% 50% 37% 25% 153⁄4 100%100%100%100%100%100%100%100%1継96% 83% 54% 23% 6% 1% 100 % 100% 100% 100% 100% 99% 96% 90% 79% 63% 45% 28% 15% 6% 2%
當判別模式函數、存活時間函數以及違約風險係數分 配表分別建立完成後，則可輸入新客戶的評分表，並計算 新客戶的違約風險，步驟180與步驟190。 接著參閱第2圖，其係揭露本發明之企業融資與風險 籲 評估方法新客戶申請融資之流程示意圖。步驟210，一新 客戶欲進行融資的申請。步驟220，填寫客戶資料。步驟 230,根據客戶填寫的客戶資料，計算評分表，並將這些評 分資料傳送至具有第1圖所揭露之風險評估模組的電腦之 中，以計算新客戶的違約風險，步驟240。 以曱表舉例而言，參閱步驟150的說明，違約判別模 式函數： y = -1.2772*kll + 1.7343*kl2+1.8044*kl0+1.0035*N8 12 201227580 +3.3259*Nl+0.1831*N3b+2.6197*N9b+3.2051*k7+ 0.8927*N4-14.0358 正常判別模式函數： z = 5.2101*kll-2.5698*kl2+0.8724*kl0+2.0120*N8 +2.4201 *Nl+〇.7673*N3b+l.8157*N9b+4.2700*k7+ 1.3560*N4-20.6564 其中，kll=2、kl2=2、kl0=2、N8=2、N卜2、N3b=2、 N9b=2、k7=2。 φ 因此，y=12,9472 z= 11.6512 y>z所以被判定可能違約之客戶。 步驟250，判斷案件是否違約。若z>y則判斷客戶並 不會違約，若y>z則可判定客戶可能違約。 當客戶並不會違約時，進入步驟260，與客戶簽訂融 資合約。當客戶可能會違約時，進入步驟270進一步計算 此融資之存活時間與違約風險係數。 接著利用存活時間函數計算存活時間 m = 0.787*Nl+0.935*N3a+0.621*kl+0.759* k8 其中 ’ Nl=2、N3a=2、kl=2、k8=2。 因此，m=6.204 所以，預估此新客戶可能之違約時間為第6.2個月。 然後’查詢步驟170中的違約風險係數分配表，以曱 13 201227580 表R5第7個月違約比率為54%。此時電腦系統可進一步 計算此時客戶預估已還款的金額，與此時本金的餘額。其 中，表中R6與R7則分別表示不同評等所使用之違約風險 係數，其可以依客戶的信用評等進行選擇。一般而言，新 的客戶較佳地採用R5作為查詢的標準，當客戶曾經出現過 延遲還款等信用風險時，可調整評等，以更正確地估計此 客戶的違約風險係數。因此，將此違約時點之本金餘額* 違約比率為54%即可算出，預估的違約金額，步驟280。 例如，當客戶融資18萬分12個月償還，亦即每個月償還 • 本金1.5萬，當第7個月違約時，已償還6個月的本金， 約9萬元，尚餘本金9萬未清償。而根據歷史資料的機率 統計，此種客戶在第7個月違約的機率甚高，約有違約54 %比率會違約，故預估的違約金額為9萬乘54%約等於 4.86萬，亦即以歷史資料的統計而言，若金融機構融資給 此客戶約可能會發生4.86萬的本金虧損的機率。 此時，步驟290，可根據違約可能發生的月份，調整 還款的模式，例如是將本金償還由每月平均償還，修改成 • 初期數個月須償還較高的本金，例如前1/4的時間需償還 1/2的本金，或者是縮短融資期限，將12個月的融資縮短 為6個月的融資。亦或者是，同時縮短融資的金額與融資 的期限。還可以是非直線性的還款方式，以使客戶可在較 短的期限下還清融資的金額。如此，可有效地降低融資違 約的風險，提高金融機構融資的意願。步驟300，金融機 構可根據調整後的還款方式與客戶簽訂融資合約，並依照 修正後之合約内容要求客戶調整還款方式，以縮短融資期 14 201227580 限。 然若，其可能發生的違約金額實在過高，則金融機構 亦可以據此直接拒絕核貸，步驟320。相反地，步驟330， 若經計算其可能發生的違約金額十分的低，此時，金融機 構可考慮部份已回收的本金與利息後，此違約金額並不會 造成金融機構的損失，亦可以直接核貸，以增加金融機構 的獲利機會與能力。 因此，藉由本發明之企業融資與風險評估方法，不僅 可有效地評估新客戶的違約風險，並藉由新的數據資料， ® 更新歷史評分資料，以更精確地預估客戶違約的風險。此 外，由於本發明不僅根據財務項目的評分項目，更包含非 財務項目的評分項目，以充分反應企業經營狀態的真實 性，以增加系統預測的準確度。值得注意的是，當本發明 之企業融資與風險評估方法預估新客戶有違約風險時，更 可以計算新客戶存活的時間與違約風險，進而計算違約金 額，並可根據違約金額與存活時間調整還款方式，再與客 戶簽訂融資合約，有效降低融資風險。而還款方式，更可 • 以採用非線性的方式進行，以進一步降低融資風險。 如熟悉此技術之人員所瞭解的，以上所述僅為本發明 之較佳實施例而已，並非用以限定本發明之申請專利範 圍。凡其它未脫離本發明所揭示之精神下所完成之等效改 變或修飾，均應包含在下述之申請專利範圍内。 【圖式簡單說明】 為讓本發明之上述和其他目的、特徵、優點與實施例 15 201227580 能更明顯易懂，所附圖式之說明如下： 第1圖係為本發明之企業融資與風險評估方法之建構 流程是示意圖；以及 第2圖係為本發明之企業融資與風險評估方法新客戶 申請融資之流程示意圖。 【主要元件符號說明】 110〜330 :步驟After the discriminant mode function, the survival time function, and the default risk coefficient allocation table are respectively established, the new customer's score table can be input, and the new customer's default risk is calculated, step 180 and step 190. Referring to Figure 2, it is a schematic diagram of the process of applying for financing by a new customer for the corporate finance and risk appeal evaluation method of the present invention. Step 210, a new customer wants to apply for financing. In step 220, fill in the customer information. Step 230: Calculate the score table according to the customer data filled in by the customer, and transmit the score data to the computer having the risk assessment module disclosed in FIG. 1 to calculate the default risk of the new customer, step 240. For example, referring to the description of step 150, the default discriminant mode function: y = -1.2772*kll + 1.7343*kl2+1.8044*kl0+1.0035*N8 12 201227580 +3.3259*Nl+0.1831*N3b+2.6197*N9b +3.2051*k7+ 0.8927*N4-14.0358 Normal discriminant mode function: z = 5.2101*kll-2.5698*kl2+0.8724*kl0+2.0120*N8 +2.4201 *Nl+〇.7673*N3b+l.8157*N9b+4.2700*k7+ 1.3560*N4-20.6564 where kll=2, kl2=2, kl0=2, N8=2, Nb2, N3b=2, N9b=2, k7=2. φ Therefore, y=12,9472 z= 11.6512 y>z is therefore a customer who is judged to be in default. In step 250, it is determined whether the case is breach of contract. If z>y, it is judged that the customer will not default, and if y>z, it can be determined that the customer may default. When the customer does not default, proceed to step 260 to enter into a financing contract with the customer. When the customer may default, proceed to step 270 to further calculate the survival time and default risk factor for the financing. The survival time is then calculated using the survival time function m = 0.787 * Nl + 0.935 * N3a + 0.621 * kl + 0.759 * k8 where ' Nl = 2, N3a = 2, kl = 2, k8 = 2. Therefore, m=6.204 Therefore, it is estimated that the default time for this new customer is 6.2 months. Then, the default risk factor allocation table in step 170 is queried to 54 13 201227580 Table R5 The 7th month default ratio is 54%. At this point, the computer system can further calculate the amount of the customer's estimated repayment at this time, and the balance of the principal at this time. Among them, R6 and R7 in the table respectively indicate the default risk coefficient used by different ratings, which can be selected according to the customer's credit rating. In general, new customers prefer R5 as the standard for enquiries. When a customer has experienced a credit risk such as delayed repayment, the rating can be adjusted to more accurately estimate the default risk factor of the customer. Therefore, the default balance of the principal balance* at the time of default can be calculated as 54%, and the estimated default amount is step 280. For example, when the customer finances 180,000 points and 12 months to repay, that is, repays the principal amount of 15,000 per month, when the contract expires in the 7th month, the principal of 6 months has been repaid, about 90,000 yuan, and the remaining principal. 90,000 outstanding. According to the probability of historical data, the probability of such a customer defaulting in the 7th month is very high. About 54% of the default will default, so the estimated default amount is 90,000 by 54%, which is equal to 48,600. In terms of historical data, if a financial institution finances this customer, it may have a probability of a loss of 48,600 yuan. At this time, in step 290, the repayment mode can be adjusted according to the month in which the default may occur, for example, the principal repayment is repaid by the average monthly repayment, and the initial principal is required to repay the higher principal amount, for example, the first one. /4 time to repay the 1/2 principal, or shorten the financing period, shorten the 12-month financing to 6 months of financing. Or, at the same time, shorten the amount of financing and the time limit for financing. It can also be a non-linear repayment method so that the customer can pay off the amount of financing in a shorter period of time. In this way, the risk of financing defaults can be effectively reduced and the willingness of financial institutions to raise funds can be improved. In step 300, the financial institution may sign a financing contract with the customer according to the adjusted repayment method, and request the customer to adjust the repayment method according to the revised contract content, so as to shorten the financing period 14 201227580. However, if the amount of default that may occur is too high, the financial institution may directly reject the nuclear loan accordingly, step 320. Conversely, in step 330, if the amount of default that may occur is calculated to be very low, at this time, after the financial institution can consider part of the recovered principal and interest, the amount of the breach will not cause loss to the financial institution. You can directly lend to increase the profit opportunities and capabilities of financial institutions. Therefore, the corporate finance and risk assessment method of the present invention not only can effectively assess the default risk of new customers, but also update the historical score data with new data, to more accurately estimate the risk of customer default. In addition, the present invention not only according to the scoring items of the financial items, but also the scoring items of the non-financial items, in order to fully reflect the authenticity of the business operation state, so as to increase the accuracy of the system prediction. It is worth noting that when the enterprise financing and risk assessment method of the present invention estimates that the new customer has a default risk, it can calculate the time of the new customer to survive and the risk of default, and then calculate the default amount, and can adjust according to the default amount and the survival time. Repayment method, and then sign financing contracts with customers to effectively reduce financing risks. And the repayment method can be made in a non-linear way to further reduce the financing risk. The above description is only for the preferred embodiment of the present invention, and is not intended to limit the scope of the invention. All other equivalent changes or modifications made without departing from the spirit of the invention are intended to be included within the scope of the appended claims. BRIEF DESCRIPTION OF THE DRAWINGS In order to make the above and other objects, features and advantages of the present invention and the embodiment 15 201227580 more apparent, the description of the drawings is as follows: FIG. 1 is a corporate financing and risk of the present invention. The construction process of the evaluation method is a schematic diagram; and the second figure is a schematic diagram of the process of applying for financing for new customers of the enterprise financing and risk assessment method of the present invention. [Main component symbol description] 110~330: Step
1616