SYSTEM AND METHOD FOR PREDICTING CONSUMER CREDIT RISK USING INCOME RISK BASED CREDIT SCORE

• US 20110078073A1
• Filed: 10/20/2009
• Published: 03/31/2011
• Est. Priority Date: 09/30/2009
• Status: Active Grant

• Current Assignee: Scorelogix LLC
• Original Assignee: Scorelogix LLC
• Inventors: Annappindi, Suresh Kumar
• Granted Patent: US 8,799,150 B2
• Time in Patent Office: Days
• US Class Current: 705/38
• CPC Class Codes: G06Q 40/00 Finance; Insurance; Tax strategies; Processing of corporate or income taxes
• CPC Class Codes: G06Q 40/02 Banking, e.g. interest calculation or account maintenance credit or loans G06Q40/03
• CPC Class Codes: G06Q 40/03 Credit; Loans; Processing thereof
• CPC Class Codes: G06Q 40/04 Trading; Exchange, e.g. stocks, commodities, derivatives or currency exchange
• CPC Class Codes: G06Q 40/08 Insurance

Abstract

Systems and methods are described for scoring consumers'"'"' credit risk by determining consumers'"'"' income risk and future ability to pay. Methods are provided for measuring consumers'"'"' income risk by analyzing consumers'"'"' income loss risk, income reduction risk, probability of continuance of income, and economy'"'"'s impact on consumers'"'"' income. In one embodiment, a method is provided to evaluate an individual'"'"'s creditworthiness using income risk based credit score thereby providing creditors, lenders, marketers, and companies with deeper, new insights into consumer'"'"'s credit risk and repayment potential. By predicting consumers'"'"' income risk and the associated creditworthiness the present invention increases the accuracy and reliability of consumers'"'"' credit risk assessments, results in more predictive and precise consumer credit scoring, and offers a new method of rendering a forward-looking appraisal of an individual'"'"'s ability to repay a debt or the ability to pay for products and services.

Claims

1. A computer-implemented method to predict consumers'"'"' credit risk by determining their income risk;

wherein the said credit risk is the probability of consumers defaulting on their payment obligations including credit card debt, personal loans, automotive loans, student loans, mortgage loans, and other types of consumer loans;

wherein the said credit risk also means the consumers'"'"' ability to pay;

wherein the said credit risk also means the consumers'"'"' capacity to pay;

wherein the said income risk is predicted using consumer data;

wherein the said income risk is predicted using consumers'"'"' unemployment risk;

wherein the said income risk is derived from consumers'"'"' unemployment risk;

wherein the said income risk is derived from consumers'"'"' income loss risk;

wherein the said income risk is derived from consumers'"'"' income reduction risk;

wherein the said income risk is derived from consumers'"'"' probability of continuance of income;

wherein the said income risk is assigned a numerical or qualitative value;

wherein the said income risk is correlated with consumers'"'"' credit risk, payment default risk, payment behavior, and ability to pay;

wherein the said income risk is based on economy'"'"'s impact on consumers'"'"' income;

wherein the said income risk is based on correlations between consumers'"'"' personal data and economic conditions data including unemployment rates, job growth, wages, inflation, trade, GDP, home prices, construction activity, manufacturing activity, retail sales, and others;

wherein the said income risk is transformed into an income risk score;

wherein the income risk score is used to predict consumers'"'"' response behavior, purchasing propensity, and ability to pay;

wherein the said income risk is transformed into an income risk based credit score to predict consumers'"'"' credit risk and payment default risk;

wherein the said income risk based credit score is derived from a risk forecasting computer;

wherein the risk forecasting computer consists of a microprocessor CPU, memory, databases, software programs, analytical and statistical programs, input and output devices, and networking capability; and

wherein the income risk based credit score is an empirically derived, demonstrably and statistically sound credit score predicting consumers'"'"' credit risk that is based on their future income and future ability to pay; and

implementing said data into consumer scoring and scoring systems.

2. The method of claim 1, wherein said method for determining income risk based credit score further comprises of the steps:

generating by the computer, an unemployment risk probability for an individual using the individual'"'"'s personal data including age, education, demographic data and employment history, and by using historical and projected unemployment and hiring trends, historical and projected macroeconomic and microeconomic data;

generating by the computer, a probability of income loss for an individual using unemployment risk;

generating by the computer income reduction risk for an individual using unemployment risk;

generating by the computer the probability of continuance of income for an individual using unemployment risk;

correlating by the computer, unemployment risk probabilities and income risk for individuals in a selected geography, or for individuals in a statistically valid sample comprising of hundreds, thousands, or millions of individuals, with their historical payment defaults, credit delinquencies, charge-offs, and bankruptcies;

correlating by the computer, consumers'"'"' unemployment risk probabilities and income risk with consumers'"'"' response rates for marketing offers and their profitability metrics;

correlating by the computer, consumers'"'"' unemployment risk probabilities and income risk with microeconomic and macroeconomic factors,

correlating by the computer, consumers'"'"' unemployment risk probabilities and income risk with consumers'"'"' credit default data, credit histories, and payment histories;

generating by the computer a consumer credit risk model that produces a consumer income risk based credit score by finding mathematical relationships between consumers'"'"' unemployment risk, income risk and payment default data;

generating by the computer a consumer ability to pay prediction model that predicts the likelihood of a consumer being able to buy and repay; and

processing, storing, transmitting, and rendering using the computer and a computer network, a novel consumer unemployment based credit score and a credit scoring system allowing lenders and businesses to score their prospects, applicants, existing accounts and delinquent accounts;

existing portfolios and portfolio segments; and

new portfolios and portfolio segment;

to gain new, predictive insights into consumer credit risk at the individual consumer level and at the portfolio level.

3. The method of claim 1, wherein said income risk based credit score is generated based on data selected from the group consisting of, but not limited to:

individuals'"'"' personal profile data;

individuals'"'"' income risk data;

individuals'"'"' data attributes with risk factors and weighted reason codes;

national, regional, and local employment and unemployment data;

national, regional, and local macroeconomic and microeconomic data;

consumers'"'"' response rates for marketing offers;

consumers'"'"' purchasing behavior and trends; and

consumers'"'"' payment default data, credit default data, delinquencies and bankruptcies data.

4. The method of claim 2, wherein said individual personal data is selected from the group consisting of, but not limited to, education, age, job status, job industry, job type, job tenure, salary, employment and unemployment history, geographical location, income characteristics, and credit characteristics.

5. The method of claim 2, wherein said national employment, unemployment, and economic data is selected from the group consisting of, but not limited to, historical national, regional and local employment and unemployment data, involuntary unemployment data, mass layoffs data, hiring and firing trends, existing and new job postings, unemployed population, underemployed population, discouraged workers population, wage rates, distribution of jobs in industries and occupations, government unemployment insurance claims, government unemployment insurance claim acceptance rates, government unemployment insurance benefit payment rates and amounts, duration of government unemployment insurance claims, federal and state unemployment insurance fund data, and government insurance program policies and guidelines, and non-government data.

6. The method of claim 2, wherein said national, regional, and local macroeconomic and microeconomic data is selected from the group consisting of historical and projected economic indicators including but not limited to:

gross domestic product, sales for retail and food services, durable goods, construction activity, manufacturers'"'"' shipments, inventories, and orders;

manufacturing and trade;

inventories and sales;

monthly wholesale trade, existing home sales, auto sales, new residential construction, new residential sales, construction permits, personal income and outlays, U.S. international trade in goods and services;

U.S. international transactions;

trade deficit, consumer confidence and disposable income, and inflation.

7. The method of claim 1, wherein the step of computing a consumer'"'"'s income risk and income risk based credit score further comprises of the steps of:

segmenting a national workforce population into homogenous risk categories, with each risk category comprising a plurality of homogenous sub risk subcategories;

segmenting dependent, unemployed and non-working individuals into risk categories and sub categories;

assigning a risk factor weight to each of the risk categories and sub risk subcategories;

predicting an unemployment rate for a finite duration of 1 to 10 years, or any other time frame, for each risk category and sub category;

predicting an income loss probability for a finite duration of 1 to 10 years, or any other time frame, for each risk category and sub category;

transforming the said unemployment rate predictions and income loss risk predictions, or any mathematical combinations of these, into a mathematical score on a scale of zero to one thousand or any other similar scale, or into a grade on a scale of A through F or any other similar scale, which may be developed using linear or non-linear mathematical equations;

transforming the said unemployment rate score and income loss risk score into an income risk based credit score by correlating them with individuals'"'"' ability to pay and credit data;

predicting an ability to pay risk for a finite duration of 1 to 10 years, or any other time frame, for each risk category and sub category and converting it into an ability to pay score;

predicting credit default risk for a finite duration of 1 to 10 years, or any other time frame, for each risk category and sub category and converting it into an income risk based credit score; and

providing a quantitative and qualitative explanation and narrative of the contributing risk factors, relative ranking of a said income risk based credit score'"'"'s by comparing it with other scores and score groups including, but not limited to, national and regional risk scores, industry and sub-industry scores, education scores, and scores grouped based on economic, credit and payment behavior, and demographic similarities and other common attributes.

8. The method of claim 7, wherein said unemployment risk categories are selected from the group consisting of education, industry, age, gender, occupation, state, region, income, work experience, training level, work performance, job change frequency, industry change frequency, historical unemployment data, unemployment severity, job necessity, debt-to-income ratio, expenses-to-income ratio, and job confidence.

9. The method of claim 7, wherein said forecasted unemployment rates are generated based on a mechanism selected from the group consisting of national, regional and local unemployment rates, layoff data, job hiring trends, consumer price index, producer price index, interest rates, trade balance, housing starts, industrial production, currency exchange rates, retail sales, personal income and credit, consumer expenditure, industry capacity utilization, government spending, capital spending, consumer confidence and non-government data.

10. The method of claim 1, wherein the utilizing the income risk based credit score includes quantifying by the computer the credit risk of the individual to predict credit default risk, payment and repayment behavior, payment default risk, delinquency risk, charge-off risk, bankruptcy risk, likely spending trends, likelihood of on-time payments, and the effectiveness of products or services to provide a more accurate assessment of an individual consumer'"'"'s credit risk.

11. The method of claim 1, wherein the computation of income risk based credit score further comprises of the steps of:

determining by computer the probability of unemployment risk of an individual consumer or a borrower;

determining by computer the probability of income loss risk of an individual consumer or a borrower;

correlating by computer the income loss risk of an individual consumer or a borrower with credit default data and payment default data;

analyzing by computer the correlation and statistical relationships between actual unemployment data, income loss data, and credit default data for hundreds, thousands or millions of individuals;

correlating and comparing by computer the actual and projected income loss risk with actual and projected credit risk and payment default data for thousands or millions of individuals;

correlating by computer the income loss risk with future ability to pay for an individual consumer or a borrower;

establishing statistical relationships and mathematical equations between unemployment risk, income loss risk, ability to pay, and credit risk using historical credit default data;

establishing statistical relationships and mathematical equations between ability to pay and credit risk through retro tests and back tests by working with relevant banks, financial institutions, credit unions, credit bureaus, data warehousing companies, and other companies which have individual level data;

transforming by computer the income loss risk and ability to pay factors into a probability of payment default for an individual consumer or a borrower;

transforming by computer the probability of payment default into an income risk based credit score consisting of a three-digit or four-digit number, or any mathematical score or grade, or any other similar embodiment that quantifies and reflects different default probabilities into discernible patterns;

developing by computer a statistically valid and empirically sound credit scoring model that predicts an individual'"'"'s credit risk by using the income risk based credit score;

validating the income risk based credit score'"'"'s predictive power and risk separation capability by testing the score using actual consumer data;

assigning the said income risk based credit score appropriate marketing and product names, such as the Job Security Score, Income Risk Score, Income View Score, Income Credit Score, Income Prospect Score, and Income Continuance Score;

correlating by computer the predictive power of the income risk based credit score with performance characteristics of a specific loan portfolio, or with many different loan portfolios;

establishing odds ratios and loss curves for income risk based credit score for different loan portfolio types including marketing, acquisitions, account management and collections,

developing income risk based credit score usage strategies for a multitude of lending and lending related decisions including credit approval decisions, credit line decisions, up sell and cross sell decisions, rewards strategies, account treatment strategies, collection strategies, delinquency management strategies, charge off and loss mitigation strategies, portfolio sale and purchase strategies, portfolio asset valuation strategies, portfolio securitization strategies, and forecasting losses and revenues; and

developing and rendering income risk based credit score for use by marketers, lenders and companies for use as a prospect score, primary credit score, and as a secondary credit score to be used in conjunction with other types of credit scores and alternative credit scores.

12. The method of claim 1, wherein the said income risk based credit score can be modified or customized by:

combining by the computer, the income risk based credit score with credit bureau scores and consumer risk scores using equal on non-equal weights for each;

customizing by the computer, the income risk based credit score, credit bureau scores and consumer risk scores, and a selected portfolio'"'"'s credit performance data; and

producing at the computer, a more comprehensive consumer credit risk score a and credit scoring system for a specific pool of accounts, portfolio, or lender, allowing businesses to obtain a more predictive and accurate assessment of their consumers'"'"' credit risk.

13. The method of claim 1, further comprising of the step of updating the income risk based credit score frequently and periodically such as monthly, quarterly or yearly, and any other suitable frequency, to capture the latest and best possible measure of economy'"'"'s impact on consumers'"'"' income risk and credit risk.

14. The method of claim 11, further comprising the step of combining the income risk based credit score with any of the existing credit bureau scores or risk scores in order to increase approvals by redistributing the population through segmention and differentiation based on advanced credit risk prediction capabilities.

15. The method of claim 1, wherein in one embodiment of income risk based credit score, the Job Security Score, predicts a payment default risk for an individual for a finite future, such as a period of up to thirty six months or any other finite period from the time of scoring, by using a set of input variables selected from the group consisting of but not limited to age, personal income, total debt, debt ratio (debt/available debt), number of times delinquent in last two years, savings account information (if one exist), residency (city, state, and zip code), years at current residence, own/rent status, total yearly income, highest level of education, education discipline/concentration, year attained, educational institution, years of full time work experience, current employer, length of time with present employer, self-employment (if any), part-time/full-time status, work city, state and zip code, job occupation area, employer'"'"'s industry (name, SIC code), and total employees at place of work.

16. The method of claim 1, wherein in one embodiment of income risk based credit score, the Job Security Score, becomes an FCRA compliant credit score predicting consumer credit risk for an individual for a finite future using only FCRA compliant input variables.

17. The method of claim 1, wherein in one embodiment of income risk based credit score, the Income Stability Score, predicts ability to pay and buy for an individual for use as a prescreening score and as a prospect score for identifying prospects and for predicting response rates for marketing purposes.

18. The method of claim 1, wherein the income risk based credit score is generated using a system consisting of a computer-readable medium having computer-executable instructions for performing a method comprising of:

creating databases to store national, regional and local employment and unemployment data, economic data, and consumers'"'"' personal data;

adding and refreshing new data into said databases;

updating said databases with derivative data and predicted data using mathematical processes periodically; and

forecasting unemployment risk and income risk for homogeneous risk groups in order to measure and predict an individual'"'"'s income risk based credit score and Job Security Score.

19. The method of claim 18, wherein the process of computing, generating and rendering income risk based credit scores further comprises of the steps including establishing a computer-based method and system based on scoring and processing elements selected from the group consisting of mathematical equations, statistical relationships, algorithms, computer software, computing systems, mathematical models, advanced programs, electronic databases, analytical tools, statistical software, computer networks, data transfer protocols, internet and intranet, web based user interface, VPN, EDI, security protocols, dynamic handshake methods, batch scoring, real time scoring, partner network, partner'"'"'s computers and servers, and ERM systems and processes.

20. A computer-implemented method to compute a short term and a long term employment based creditworthiness score and index utilizing an individual consumer'"'"'s personal unemployment probability, income risk, and probability of continuance of income on a mechanism selected from the group consisting of unemployment risk scores, projected unemployment rates for short term and long term, current income, expected income growth for short term and long term, expected duration of employment for short term and long term, current and expected education level, expected job changes, current and future cost of living projections, job change history, and income history.

21. A computer-implemented method of utilizing an individual consumer'"'"'s personal unemployment probability and income risk to generate a credit score, said method comprising:

predicting payment defaults, delinquencies and charge off for credit card loans;

predicting mortgage payment defaults for mortgage loans;

predicting payment defaults for auto loans;

predicting payment defaults for student loans;

predicting payment defaults for payday loans;

predicting payment defaults secured and unsecured personal loans; and

predicting premium payment defaults for insurance policies including vehicle insurance, home insurance, life insurance and medical insurance.