Fenerich et al 2020a - Revision history

Rimni at 08:55, 3 September 2020

2020-09-03T08:55:55Z

Rimni at 08:54, 3 September 2020

2020-09-03T08:54:13Z

Rimni: /* Appendix 1 */

2020-09-03T08:52:08Z

‎Appendix 1

Rimni at 08:45, 3 September 2020

2020-09-03T08:45:33Z

Rimni at 08:44, 3 September 2020

2020-09-03T08:44:57Z

Rimni at 08:44, 3 September 2020

2020-09-03T08:44:09Z

Rimni: /* Appendix 1 */

2020-09-03T08:42:57Z

‎Appendix 1

Rimni at 08:40, 3 September 2020

2020-09-03T08:40:38Z

Rimni at 08:39, 3 September 2020

2020-09-03T08:39:30Z

Rimni: /* Appendix 1 */

2020-09-03T08:37:48Z

‎Appendix 1

@@ Line 657: / Line 657: @@
 In addition to the differential in the use of these techniques, in this work, an alternative way from those usually employed to classify clients as “defaulters” or “non-defaulters” was presented. In other words, a “default index” was used instead of only checking for overdue operations. Another differential lay in defining the features considered to compose the model, which are not related to the behavior of clients in previous operations so, in this way, new clients (without a historical data) of the institution can also be classified using the methodology.
-With the results obtained through the proposed methodology, it was possible to answer the questions outlined at the beginning of the article: a) All the three methods (BN; SVM and DT) presented similar performances for the credit scoring prediction close to 95% (specificity) as can be observed in Table 3, in other words, these models can correctly classify about 95% of clients as “defaulters” or “non-defaulters”. This result reinforces the main intention of a financial institution when using these classifier algorithms (ML): identify as assertively as possible clients who are prone to being “defaulters”, it is preferable to use a technique with greater specificity; (b) Besides, we can say that for the problem approached here, the class unbalancing affect the performance of the model, as we can verify comparing the CM and SM models in Table 3, i.e., it is important that the classes are balanced in order to get better performance results; (c) The feature selection stage contributes to the credit scoring model, showing what are the main features or in which features the specialists must give special attention. In the approached problem, the features “risk assessment” and “loan restrictions” are occupying the main positions, as shown in Figure 3. On the other hand, the features “gross annual revenue” and “number of employees” are less important; (d) As for the practical implications of the present study, it becomes fundamental in a critical scenario of instability of Brazilian companies regarding “default”, since in March 2019 the number of debtors grew by 3.3% (''Serviço de Proteção ao Crédito'', SPC, or Credit Protection Services). In general, the market is slow to act against these high levels of defaulting companies [77, 78].
+With the results obtained through the proposed methodology, it was possible to answer the questions outlined at the beginning of the article: a) All the three methods (BN; SVM and DT) presented similar performances for the credit scoring prediction close to 95% (specificity) as can be observed in [[#tab-3|Table 3]], in other words, these models can correctly classify about 95% of clients as “defaulters” or “non-defaulters”. This result reinforces the main intention of a financial institution when using these classifier algorithms (ML): identify as assertively as possible clients who are prone to being “defaulters”, it is preferable to use a technique with greater specificity; (b) Besides, we can say that for the problem approached here, the class unbalancing affect the performance of the model, as we can verify comparing the CM and SM models in [[#tab-3|Table 3]], i.e., it is important that the classes are balanced in order to get better performance results; (c) The feature selection stage contributes to the credit scoring model, showing what are the main features or in which features the specialists must give special attention. In the approached problem, the features “risk assessment” and “loan restrictions” are occupying the main positions, as shown in [[#img-3|Figure 3]] . On the other hand, the features “gross annual revenue” and “number of employees” are less important; (d) As for the practical implications of the present study, it becomes fundamental in a critical scenario of instability of Brazilian companies regarding “default”, since in March 2019 the number of debtors grew by 3.3% (''Serviço de Proteção ao Crédito'', SPC, or Credit Protection Services). In general, the market is slow to act against these high levels of defaulting companies [77,78].
 Suggestions for future studies include conducting tests with other classifier algorithms, as Random Forest, XGBoost, and others, based on ensemble method, for example, always with a view to improving their performance as much as possible. Another suggestion is related to the random undersampling to solve the unbalanced data problem, which sometimes causes bias results; so, in this way, it would be interesting implement it more than 10 times using the average, standard deviation of accuracy and specificity as the results.

@@ Line 665: / Line 665: @@
 ==Acknowledgements==
-This study was financed in part by PUCPR and the Coordination for the Improvement of Higher Education Personnel – Brazil (CAPES), and by the National Council for Scientific and Technological Development – Brazil.
+This study was partially financed by PUCPR and the Coordination for the Improvement of Higher Education Personnel – Brazil (CAPES), and by the National Council for Scientific and Technological Development – Brazil.
 ==References==

@@ Line 836: / Line 836: @@
 {| class="wikitable" style="margin: 1em auto 0.1em auto;border-collapse: collapse;font-size:85%;width:auto;"
 |-style="text-align:center"
-! rowspan="2" | Features (1) !! rowspan='2' | Levels of features (2) !!  colspan='2' | Types of coding
+! rowspan="2" style="text-align:left" | Features (1) !! rowspan='2' style="text-align:left" | Levels of features (2) !!  colspan='2' | Types of coding
 |-
 !   I – Categorical (3) !! II – Binary (4)

@@ Line 833: / Line 833: @@
 <div class="center" style="font-size: 75%;">'''Table A.''' Feature coding</div>
-<div id='tab-1'></div>
+<div id='tab-A'></div>
 {| class="wikitable" style="margin: 1em auto 0.1em auto;border-collapse: collapse;font-size:85%;width:auto;"
 |-style="text-align:center"

@@ Line 828: / Line 828: @@
 ==Appendix 1==
 <div class="center" style="font-size: 75%;">'''Table A.''' Feature coding</div>
@@ Line 836: / Line 838: @@
 ! rowspan="2" | Features (1) !! rowspan='2' | Levels of features (2) !!  colspan='2' | Types of coding
 |-
-!   I – Categorical (3) !! II – Binary(4)
+!   I – Categorical (3) !! II – Binary (4)
 |-
 |  rowspan='12' style="text-align: left;width:180px;"|(1) Legal Status

@@ Line 838: / Line 838: @@
 !   I – Categorical (3) !! II – Binary(4)
 |-
-|  rowspan='12' style="text-align: left;width:210px;"|(1) Legal Status
+|  rowspan='12' style="text-align: left;width:180px;"|(1) Legal Status
 |  style="text-align: left;"|(1) Limited partnership
-|  style="text-align: left;"|1
+|  style="text-align: center;"|1
 |  style="text-align: center;"|100000000000
 |-

@@ Line 838: / Line 838: @@
 !   I – Categorical (3) !! II – Binary(4)
 |-
-|  rowspan='12' style="text-align: left;width:200px;"|(1) Legal Status
+|  rowspan='12' style="text-align: left;width:210px;"|(1) Legal Status
 |  style="text-align: left;"|(1) Limited partnership
 |  style="text-align: left;"|1
@@ Line 903: / Line 903: @@
 |  style="text-align:center;"|4
 |  style="text-align:center;"|0001
-|-
+|-style="border-top: 2pt solid black;"
-|  rowspan='4' style="text-align: left;border-top: 2pt solid black;"|(3) Number of  Employees
+|  rowspan='4' style="text-align: left;"|(3) Number of  Employees
 |  style="style="text-align: left;"|(17) 0 a 1
 |  style="text-align: center;"|1
@@ Line 920: / Line 920: @@
 |  style="text-align:center;"|4
 |  style="text-align:center;"|0001
-|-
+|-style="border-top: 2pt solid black;"
-|  rowspan='19' style="text-align: left;border-top: 2pt solid black;"|(4) Company Activity
+|  rowspan='19' style="text-align: left;"|(4) Company Activity
 |  style="style="text-align: left;"|(21) Real estate management
 |  style="text-align: center;"|1
@@ Line 997: / Line 997: @@
 |  style="text-align:center;"|19
 |  style="text-align:center;"|0000000000000000001
-|-
+|-style="border-top: 2pt solid black;"
-|  rowspan='4' style="text-align: left;border-top: 2pt solid black;"|(5) Incorporation Data
+|  rowspan='4' style="text-align: left;"|(5) Incorporation Data
 |  style="text-align: left;"|(40) 15/02/1939 a 01/06/1997
 |  style="text-align: center;"|1
@@ Line 1,014: / Line 1,014: @@
 |  style="text-align:center;"|4
 |  style="text-align:center;"|0001
-|-
+|-style="border-top: 2pt solid black;"
-|  rowspan='4' style="text-align: left;border-top: 2pt solid black;"|(6) Time as Client
+|  rowspan='4' style="text-align: left;"|(6) Time as Client
 |  style="text-align: left;"|(44) 23/06/1975 a 29/10/2003
 |  style="text-align: center;"|1
@@ Line 1,031: / Line 1,031: @@
 |  style="text-align:center;"|4
 |  style="text-align:center;"|0001
-|-
+|-style="border-top: 2pt solid black;"
-|  rowspan='7' style="text-align: left;border-top: 2pt solid black;"|(7) Market Sector
+|  rowspan='7' style="text-align: left;"|(7) Market Sector
 |  style="text-align: left;"|(48) Small Company
 |  style="text-align: center;"|1
@@ Line 1,060: / Line 1,060: @@
 |  style="text-align:center;"|7
 |  style="text-align:center;"|0000001
-|-
+|-style="border-top: 2pt solid black;"
-|  rowspan='5' style="text-align: left;border-top: 2pt solid black;"|(8) Risk Assessment
+|  rowspan='5' style="text-align: left;"|(8) Risk Assessment
 |  style="text-align: left;"|(55) A
 |  style="text-align: center;"|1
@@ Line 1,081: / Line 1,081: @@
 |  style="text-align:center;"|5
 |  style="text-align:center;"|00001
-|-
+|-style="border-top: 2pt solid black;"
-|  rowspan='12' style=" text-align: left;border-top: 2pt solid black;"|(9) Borrower’s Credit Limit (hierarchical decision level)
+|  rowspan='12' style=" text-align: left;"|(9) Borrower’s Credit Limit (hierarchical decision level)
 |  style="text-align: left;"|(60) Level 1
 |  style="text-align: center;"|1
@@ Line 1,130: / Line 1,130: @@
 |  style=" text-align: center;"|12
 |  style="text-align:center;"|000000000001
-|-
+|-style="border-top: 2pt solid black;"
-|  rowspan='3' style=" text-align: left;border-top: 2pt solid black;"|(10) Loan Restrictions
+|  rowspan='3' style=" text-align: left;"|(10) Loan Restrictions
 |  style="text-align: left;"|(72) Impeditive restriction
 |  style="text-align: center;"|1