Apparatus and method for total loss prediction
A Total Loss Predictor Score assesses the risk of Total Loss of new credit applications by leveraging the most powerful insights from business databases. The Total Loss Predictor Score built on segmentation and algorithms that use the commercial information from business databases, powered by analytics, delivers a unique solution for credit risk management to help business creditors with origination decisions.
1. A system for predicting whether extending credit to a first party which lacks a credit history due to no or limited trade data will result in a total loss to a second party that extends credit to the first party, the system comprising a processor and a memory that has instructions that are readable by the processor and cause the processor to obtain a financial risk score for the first party by performing the following steps in order:
collect non-traditional data relating to the first party, wherein said non-traditional data is at least one selected from the group consisting of:
publicly available data,
demographic data,
intelligent engine signals,
deliver sequence files,
velocity of data changes,
velocity of historical inquiries pertaining to the first party, and
address delivery data;
access stored weights to be assigned to each item of the non-traditional data;
apply the assigned weight to each item of non-traditional data to generate weighted non-traditional data;
process the weighted non-traditional data to determine a score for each item of non-traditional data;
add the score for each item of non-traditional data to obtain a score total;
add the score total to a baseline score to obtain a financial risk score; and, after the financial risk score is obtained,
compare the financial risk score to historical financial risk scores to obtain a risk class between 1 and 9 for the first party, wherein 1 represents the lowest risk of never paying and 9 represents the highest risk of never paying, to determine whether the first party has no ability or no intention to pay the second party and whether extending credit to the first party will result in a total loss to the second party, wherein total loss is defined as the first party having at least eighty percent of an outstanding credit balance associated with accounts that have non-decreasing balances for more than one hundred twenty one days past due in the payment history, and wherein said processor that performs the steps is able to predict the total loss to the second party which is otherwise not available because the first party lacks a credit history.
2. The system of claim 1 , further comprising instructions that cause the processor to access a plurality of scorecards after the retrieve step to compute the score total, wherein a scorecard is selected from the plurality of scorecards based on the type of commercial information data available.
3. The system of claim 2 , wherein the scorecard is selected based on whether the first party has or does not have trade data.
4. The system of claim 3 , wherein, if the first party has trade data, the scorecard is selected based on one of a, b and c, wherein:
a. the trade data is detailed trade data and there is no prior severe delinquency;
b. the trade data is thick trade data, and the first party has prior severe delinquency; and
c. the trade data is thin trade data, and the first party has prior severe delinquency.
5. The system of claim 4 , wherein the trade data is thin trade data if there are one or two trades and the trade data is thick trade data if there are three or more trades.
6. A method for predicting whether extending credit to a first party which lacks a credit history will result in a total loss to a second party that extends credit to the first, the method comprising providing a processor and a memory that has instructions that are readable by the processor and cause the processor to obtain a financial risk score for the first party by performing the following steps in order:
collecting non-traditional data relating to the first party, wherein said non-traditional data is at least one selected from the group consisting of:
publicly available data,
demographic data,
intelligent engine signals,
deliver sequence files,
velocity of data changes,
velocity of historical inquiries pertaining to the first party, and
address delivery data;
accessing stored weights to be assigned to each item of the non-traditional data;
applying the assigned weight to each item of non-traditional data to generate weighted non-traditional data;
processing the weighted non-traditional data to determine a score for each item of non-traditional data;
adding the score for each item of non-traditional data to obtain a score total;
adding the score total to a baseline score to obtain a financial risk score; and, after the risk score is obtained,
comparing the financial risk score to historical financial risk scores to obtain a risk class between 1 and 9 for the first party, wherein 1 represents the lowest risk of never paying and 9 represents the highest risk of never paying, to determine whether the first party has no ability or no intention to pay the second party and whether extending credit to the first party will result in a total loss to the second party, wherein total loss is defined as the first party having at least eighty percent of an outstanding credit balance associated with accounts that have non-decreasing balances for more than one hundred twenty one days past due in the payment history, and wherein said processor that performs the steps is able to predict the total loss to the second party which is otherwise not available because the first party lacks a credit history.
7. The method of claim 6 , further comprising instructions that cause the processor to perform the step of accessing a plurality of scorecards after the collecting step to compute the score total, wherein a scorecard is selected from the plurality of scorecards based on the type of commercial information data available.
8. The method of claim 7 , wherein the scorecard is based on whether the first party has or does not have trade data.
9. The method of claim 8 , wherein, if the first party has trade data, the scorecard is selected based on one of a, b and c, wherein:
a. the trade data is detailed trade data and there is no prior severe delinquency;
b. the trade data is thick trade data, and the first party has prior severe delinquency; and
c. the trade data is thin trade data, and the first party has prior severe delinquency.
10. The method of claim 9 , wherein the trade data is thin trade data if there are one or two trades and the trade data is thick trade data if there are three or more trades.
11. A computer readable non-transitory storage medium storing instructions of a computer program which when executed by a computer system results in performance of steps of a method for predicting whether extending credit to a first party which lacks a credit history will result in a total loss to a second party that extends credit to the first party, the instructions causing the computer system to obtain a financial risk score for the first party by performing the following steps in order:
collecting non-traditional data relating to the first party, wherein said non-traditional data is at least one selected from the group consisting of:
publicly available data,
demographic data,
intelligent engine signals,
deliver sequence files,
velocity of data changes,
velocity of historical inquiries pertaining to the first party, and
address delivery data;
accessing stored weights to be assigned to each item of non-traditional data;
applying the assigned weight to each item of non-traditional data to generate weighted non-traditional data;
processing the weighted non-traditional data to determine a score for each item of non-traditional data;
adding the score for each item of non-traditional data to obtain a score total;
adding the score total to a baseline score to obtain a financial risk score; and, after the risk score is obtained
comparing the financial risk score to historical financial risk scores to obtain a risk class between 1 and 9 for the first party, wherein 1 represents the lowest risk of never paying and 9 represents the highest risk of never paying, to determine whether the first party has no ability or no intention to pay the second party and whether extending credit to the first party will result in a total loss to the second party, wherein total loss is defined as the first party having at least eighty percent of an outstanding credit balance associated with accounts that have non-decreasing balances for more than one hundred twenty one days past due in the payment history, and wherein said processor that performs the steps is able to predict the total loss to the second party which is otherwise not available because the first party lacks a credit history.
12. The storage medium of claim 11 , further comprising instructions that cause the processor to perform the step of accessing a plurality of scorecards after the collecting step to compute the score total, wherein a scorecard is selected from the plurality of scorecards based on the type of commercial information data available.
13. The storage medium of claim 12 , wherein the scorecard is based on whether the first party has or does not have trade data.
14. The storage medium of claim 13 , wherein, if the first party has trade data, the scorecard is selected based on one of a, b and c, wherein:
d. the trade data is detailed trade data and there is no prior severe delinquency;
e. the trade data is thick trade data, and the first party has prior severe delinquency; and
f. the trade data is thin trade data, and the first party has prior severe delinquency.
15. The storage medium of claim 14 , wherein the trade data is thin trade data if there are one or two trades and the trade data is thick trade data if there are three or more trades.