U.S. patent application number 14/339703 was filed with the patent office on 2015-01-22 for system and method for predicting consumer credit risk using income risk based credit score.
The applicant listed for this patent is SCORELOGIX LLC. Invention is credited to Suresh K ANNAPPINDI.
Application Number | 20150026039 14/339703 |
Document ID | / |
Family ID | 43781378 |
Filed Date | 2015-01-22 |
United States Patent
Application |
20150026039 |
Kind Code |
A1 |
ANNAPPINDI; Suresh K |
January 22, 2015 |
SYSTEM AND METHOD FOR PREDICTING CONSUMER CREDIT RISK USING INCOME
RISK BASED CREDIT SCORE
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.
Inventors: |
ANNAPPINDI; Suresh K; (Bear,
DE) |
|
Applicant: |
Name |
City |
State |
Country |
Type |
SCORELOGIX LLC |
New Castle |
DE |
US |
|
|
Family ID: |
43781378 |
Appl. No.: |
14/339703 |
Filed: |
July 24, 2014 |
Related U.S. Patent Documents
|
|
|
|
|
|
Application
Number |
Filing Date |
Patent Number |
|
|
12582507 |
Oct 20, 2009 |
8799150 |
|
|
14339703 |
|
|
|
|
61247421 |
Sep 30, 2009 |
|
|
|
Current U.S.
Class: |
705/38 |
Current CPC
Class: |
G06Q 40/08 20130101;
G06Q 40/02 20130101; G06Q 40/04 20130101; G06Q 40/025 20130101;
G06Q 40/00 20130101 |
Class at
Publication: |
705/38 |
International
Class: |
G06Q 40/02 20120101
G06Q040/02 |
Claims
1. A computer-implemented method to predict a consumer's credit
risk; wherein said credit risk is the probability of consumer
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 consumer's ability to pay; wherein the said credit
risk also means the consumer's capacity to pay; wherein the said
income risk is predicted using consumer data; wherein the said
income risk is predicted using consumer's unemployment risk;
wherein the said income risk is derived from consumer's
unemployment risk; wherein the said income risk is derived from
consumer's income loss risk; wherein the said income risk is
derived from consumer's income reduction risk; wherein the said
income risk is derived from consumer's 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
consumer's credit risk, payment default risk, payment behavior, and
ability to pay; wherein the said income risk is based on economy's
impact on consumer's income; wherein the said income risk is based
on correlations between consumer's' 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 consumer's response behavior,
purchasing propensity, and ability to pay; wherein the said income
risk is transformed into an income risk based credit score to
predict consumer's 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 consumer's 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'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 an 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 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 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
data, 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,
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 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, 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
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 is 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 or non-equal weights
for each; customizing by the computer the income risk based credit
score, credit bureau scores and consumer risk, 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 business 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 segmentation 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 where the personal
data further comprises 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,
local 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 by 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;
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. (canceled)
Description
RELATED APPLICATION
[0001] This claims the benefit of U.S. Provisional Application No.
61/247,421, filed Sep. 30, 2009, the entire contents of which are
incorporated herein by reference.
BACKGROUND OF THE INVENTION
[0002] 1. Field of the Invention
[0003] The present invention relates generally to the field of
consumer credit scoring and credit risk prediction, and, more
particularly, the present invention relates to the utilization of a
novel income risk based credit scoring system using an individual's
unemployment risk probability and income loss risk, and factoring
the impact of economy on consumers' credit risk, to increase the
accuracy of consumer credit risk forecasts resulting in credit loss
reductions, increase in acquisitions, increase in portfolio credit
quality, and an increase in profitability in the consumer credit
industry.
[0004] 2. Description of the Background
[0005] Individual borrowers pay their loans or loan installments
when they have the ability to pay. The ability to pay largely
depends on a person's disposable income. And if a person's
disposable income disappears due to the loss of his job, or due to
income reduction resulting from a pay cut or a change in job or due
to underemployment, then the person assumes a much higher risk of
defaulting on his loan repayments simply because the person has no
money and therefore has no ability to pay. That is why it is
critical to predict a person's ability to pay based on his future
probability of loss of income or a reduction in income in order to
make a superior prediction of his creditworthiness. Today, the
standard approach to credit scoring is through traditional credit
scores but the problem is that they are increasingly becoming
inaccurate, simply because they don't predict future ability to
pay. They are essentially reactive scores, meaning they change
after borrowers default, and do not factor changes in the economy,
and purely rely on credit histories and consumers' past ability to
pay.
[0006] The problem this invention solves is that traditional credit
bureau scores are not very accurate and have many significant
limitations. Specifically, there are 3 problems with credit bureau
scores. First problem is that credit bureau scores are reactive
scores. The reason credit bureau scores are reactive is because
they change only after the borrower defaults. The second problem is
that credit bureau does not consider borrowers' income risk and
that is why they can never be very accurate in predicting credit
risk. The third problem with credit bureau scores is that they
cannot score about 70 m people. This is because credit bureau
scores can only be generated for people, who have long credit
histories, but some 70 million people do not have credit histories
or have very limited credit histories, and hence credit bureau
model cannot score them. This means most lenders are not able to do
business with these 70 million people.
[0007] To appreciate credit bureau score limitations let's take a
look at credit bureau factors. The five key factors and their
contribution to the overall credit bureau score are: payment
history (35%), amount owed (30%), length of credit history (15%),
types of credit (10%), and new credit (10%). As can be seen, credit
bureau scores are entirely based on past credit behavior and does
not factor future income risk or impact of economy on consumer's
ability to pay. So, essentially credit bureau score is a measure of
past credit risk and would work only for those people whose risk
profile and income risk has not changed or been affected because of
changes in the economy and business conditions.
[0008] A person is only able to repay a loan if the monetary
sources are available which is usually dependent on consumer's
continuance of present income and on consumer's intent to pay;
thus, in effect the consumer's total credit is a function of both
the willingness to pay and the "ability to pay." Since an
individual's ability to pay is directly related to continuance of
income, defining that individual's credit risk using income loss
risk and unemployment probability greatly increases the accuracy
and effectiveness of credit risk prediction. Although, the
consumer's income risk is a critical driver of credit risk it is
not a factor in existing credit bureau scores.
[0009] The ability to pay is a critical factor in predicting credit
risk, because a borrower must have both the willingness and the
ability to repay a loan. If any one factor is missing then lenders
will not get their payment. So the bottom line is that credit risk
equals willingness to pay plus the ability to pay. And while it is
useful to know the past willingness and ability, what really
matters is the future willingness and future ability. And the
future ability to pay depends on the borrower as well as the
economic conditions, just as `accident risk` depends on both the
`driver` and the `driving conditions`. Since ability to pay is such
a critical driver of consumer creditworthiness, considering
consumers' income risk and ability to pay in addition to the credit
histories and payment histories will greatly enhance the predictive
power of credit scoring models.
[0010] Consumer credit has traditionally been regarded to have
three components: Collateral, Capacity, and Character (or
Willingness). However, there is no collateral in cases of unsecured
loans such as credit cards, capacity is equated with current income
level, and willingness is judged based on past payment behavior.
While credit bureau scores are based on the idea that a borrower's
past payment behavior is indicative of their future payment
behavior, a person's previous ability to pay is a less reliable
predictor of future ability to pay compared to future continuance
of income. Therefore existing credit scoring models fail to take
into account consumer's true "capacity" to pay or ability to pay
which depends on consumer's future continuance or income risk. But
the present invention addresses this unmet need by providing a
method to determine a consumer's income risk and the dependent
credit risk.
[0011] As of September 2009, the Applicant is the only provider of
income risk based credit score in the industry. No other invention
has been able to so accurately calculate an unemployment
probability and ability to pay and, more importantly, incorporate
income risk into a credit scoring system to offer new, better
credit risk insights resulting in effective and accurate consumer
credit risk predictions.
[0012] One embodiment of the income risk based credit score is the
Job Security Score which is generated by a novel credit scoring
system complaint with the Equal Credit Opportunity Act's (ECOA)
Regulation B. As defined in Regulation B, a "credit scoring system"
is a system that evaluates an applicant's creditworthiness
mechanically, based on key attributes of the applicant and aspects
of the transaction. It determines, alone or in conjunction with an
evaluation of additional information about the applicant, whether
an applicant is deemed creditworthy. 12 C.F.R.
.sctn.202.2(p)(1).
[0013] Also, the Job Security Score qualifies as "an empirically
derived, demonstrably and statistically sound, credit scoring
system" as defined by Reg B. The Regulation B states: [0014] To
qualify as an empirically derived, demonstrably and statistically
sound, credit scoring system, the system must be-- [0015] i. based
on data that are derived from an empirical comparison of sample
groups or the population of creditworthy and noncreditworthy
applicants who applied for credit within a reasonable period of
time; [0016] ii. developed for the purpose of evaluating the
creditworthiness of applicants with respect to the legitimate
business interests of the creditor utilizing the system (including,
but not limited to, minimizing bad debt losses and operating
expenses in accordance with the creditor's business judgment);
[0017] iii. developed and validated using accepted statistical
principles and methodology; and [0018] iv. periodically revalidated
by use of appropriate statistical principles and methodology and
adjusted as necessary to maintain predictive ability. Id. The
regulation goes on to state: [0019] A creditor may use an
empirically derived, demonstrably and statistically sound, credit
scoring system obtained from another person or obtain credit
experience from which to develop such a system. Any such system
must satisfy the criteria set forth in paragraph (p)(1)(i) through
(iv) of this section; if the creditor is unable during the
development process to validate the system based on its own credit
experience in accordance with paragraph (p)(1) of this section, the
system must be validated when sufficient credit experience becomes
available.
[0020] The current system predicts consumer creditworthiness by
predicting an individual's income risk and by empirical comparison
of income risk and credit experiences of a large population of
creditworthy and non-creditworthy applicants or accounts. The key
difference between traditional credit scores and current invention
is that traditional credit scoring systems compare an applicant's
credit profile to credit experiences of others whereas the current
scoring system compares an applicant's income risk profile to
credit experiences of others. Consumers who have more stable income
outlook because they have more job security are likely to be more
creditworthy, which is proven by the fact that unemployed
individuals default on their payment obligations a lot more than
employed individuals. The current invention uses an innovative
approach of using consumers' income risk in predicting their credit
risk and has created a credit scoring system through empirical
comparison and analysis of income loss experiences and credit
default experiences.
[0021] Current bureau scoring models only take into account
previous consumer credit transactions when creating a credit score
and do not attempt to factor a key driver of credit risk which is
lack of sufficient income. Current credit bureau scoring models
predominantly use payment history, amounts owed on account, length
of credit history, new credit inquiries, and types of credit used,
and do not use probability of income continuance. They have not yet
developed a forecasting method capable of generating future income
predictions of consumers, and therefore, have no way to analyze a
consumer's ability to pay. In addition, existing credit scoring
models are unable to score consumers with little-to-no credit
history, leaving a wide gap in its current scoring
capabilities.
[0022] Other companies have attempted to supplement the credit
scoring bureaus, but none have succeeded to the level of the
current invention. This is due to the fact that all are based on
credit data and payment data. None include a forecast of future
income risk or unemployment probability as a factor in consumer
credit risk assessment. Thus, they are restricted in their ability
to make accurate credit risk predictions.
[0023] The current invention is a novel income loss based credit
scoring model that is different from all known credit scoring
models, and was constructed based on the personal data, employment
and unemployment histories, and financial stress experiences of
individuals from a national sample between hundred thousand and one
million people and credit behavior data from actual borrowers
numbering between one million and fifteen million borrowers. It
takes into account the impact of the changing economy on consumers'
income risk and the dependent credit risk by considering: national
and local macroeconomic attributes such as the gross domestic
product, unemployment rates, retail sales, inflation, bankruptcies,
foreclosures, money supply, and energy prices; and attributes that
pertain to a group of individuals, such as type of employer and
occupation; data for individuals, such as income, years at present
job, and years at present residence; and by finding patterns and
mathematical relationships between historical macroeconomic data
and economic conditions and individuals and their historical income
risk, ability to pay, and credit risk. The model uses various
modeling techniques to predict the likelihood of unemployment and
credit risk up to thirty-six months in advance. The income risk
based credit score can be used alone or in conjunction with other
scoring models, e.g. FICO, for functions such as deciding whether
to grant or deny a credit, setting credit limits, or reviewing the
performance of an existing account.
[0024] Traditional credit scores, such as FICO scores, are
generated entirely from the credit bureau's files, but Job Security
Score primarily uses consumer's loan application data to generate
income loss risk and then to make a prediction of consumer's
creditworthiness. Since, the income risk based credit score does
not rely on credit histories it can score everyone including those
consumers who have limited or no established credit histories.
Currently in the U.S. there are 40 to 70 million consumers who do
not have any credit histories or have very little credit histories
which means that traditional credit bureau scores cannot be
meaningfully computed for them. However, the income risk based
credit score and one of its embodiments, the Job Security Score, is
easily able to score all these consumers. This allows lenders to
offer credit to "thin-file" and "no-file" applicants. For consumers
with sufficiently long credit histories and meaningful credit
bureau scores, the income risk based credit score can still be used
in combination with FICO or credit scores to add new risk insights
and to improve the accuracy and effectiveness of consumer payment
default evaluations.
[0025] The total yearly consumer credit card losses in the U.S.
amount to over 80 Billion dollars. Thus, there is a great need for
more accuracy in consumer credit risk prediction. One embodiment of
the income risk based credit score, the Job Security Score,
improves risk prediction by up to 30%, where even a 5% reduction in
credit losses will save the credit card industry $4.1 billion
annually (See FIG. 9). The increased ability of lenders,
businesses, and others to forecast the consumer's ability to pay
and credit default risk will enhance profitability by reducing
losses, improving acquisitions and marketing, and by early
identification of high default risk consumers.
SUMMARY OF THE INVENTION
[0026] Every year, millions of consumers face financial hardships
due to income disruption events such as unemployment, income loss,
and income reduction. And a majority of these financially stressed
consumers default on their payment obligations related to credit
card loans, auto loans, mortgage loans, student loans, and other
personal loans; fall behind on various kinds of insurance premium
payments including life insurance, medical insurance, auto
insurance and home insurance; and also are unable to pay their
rental, medical, utilities payments and other purchases, because
the economy and business conditions have impacted their income
continuity or has caused loss of income leaving them with
diminished ability to pay and transforming them into high credit
risk consumers. Therefore income risk, or income disruption risk,
is a very important component and driver of consumer credit
risk.
[0027] While income risk drives consumers' ability to pay, which in
turn affects consumers' creditworthiness, none of the existing
credit scoring models use income risk to predict credit risk, and
hence they are incomplete and inaccurate, and this has been clearly
proved in the current recession where traditional credit scoring
models have failed and credit losses have doubled or tripled over
expected loss rates simply because traditional credit scoring
models and conventional credit bureau scores have not been able to
quantify income risk in a poor and volatile economy, and millions
of consumers with high credit scores have defaulted because they
experienced an income disruption event which adversely impacted
their ability to pay and decreased their creditworthiness.
[0028] Hence, there is a great need for more sophisticated consumer
credit risk assessment model that considers consumers' income risk.
And therefore the present invention of income risk based credit
score is not only a novel method of predicting credit risk arising
out of income risk but it is also solves a major problem faced by
the credit scoring and the lending industry of making better,
complete credit risk predictions and minimizing credit losses.
[0029] Using the income risk based credit score as a primary
decision score or in combination with traditional credit scoring
models is much needed by the credit industry today than ever before
because tens of millions of jobs are lost every year (over 20
million jobs were lost in 2008), primarily due to economic and
business conditions. And, about two-thirds of all unemployed face
some financial difficulties and develop an increased credit default
risk, and a much higher percentage of unemployed actually default
on their debt obligations compared to the employed. In fact, job
loss is the number one reason for credit and mortgage default. With
the use of income risk based credit score, businesses gain the
ability to understand customers' risk of becoming unemployed and
defaulting on payments, allowing them to better target their
products and services to consumers that best fit with their risk
tolerance and strategic goals.
[0030] One preferred embodiment of the current invention is the
ability to generate consumer specific unemployment probabilities
and income risk probabilities by collecting consumers' personal
profile data including employment data, unemployment data,
financial stress history; economic data; and consumers' credit
default data. This preferred embodiment solves the problem of
identifying, segmenting, and targeting consumers with desired level
of income risk in order to create better prospect scores for
increasing response rates and improving marketing efficiencies.
[0031] Another significant limitation that the Applicant's
invention overcomes is that the current scoring models are
reactive, and not truly predictive, since they change or react
after the consumer demonstrates good or bad credit behavior, so in
effect existing models do not predict credit behavior but merely
reflect them and hence are lagging indicators of consumer credit
risk. Essentially, credit bureau scores are retrospective and can
generally have up to 12 month lag time in capturing an increase or
decrease in consumer's default risk simply because they change only
after the consumer defaults on payments or shows some negative
credit behavior, or when the consumer demonstrates good payment
behavior over a long period of time, and it the whole process of
collecting, processing, updating bureau databases, refreshing
scores on a periodic process, and sending them to lenders can be
tedious, error-prone and time consuming; and hence such lagging and
delayed credit risk insights could only be of limited use for
lenders. In effect traditional credit bureau scores tend to make
straight line projections of consumers credit risk, that is, good
consumers will remain good and bad consumers will remain bad in the
near term. However, that is not true for many consumers and good
consumers can go bad quickly and bad consumers can become good
consumers in a very short time, and such consumers will not be
properly identified and scored by traditional scoring models
because of their design limitations.
[0032] Another significant limitation that the Applicant's
invention overcomes is that the current scoring models are unable
to score approximately 40 to 70 million "thin-file" and "no-hit"
credit card portfolios. This inability to score 70 million
consumers due to a lack of sufficient information is a fundamental
flaw of the current credit scoring systems. Hence, a more
comprehensive and accurate credit model that not only considers
past willingness and past ability to pay, but also takes into
account the future ability to pay is greatly needed and provided by
the income risk based credit score. Applicant's invention does not
require consumers' credit histories and can score any individual
thereby providing a 100% scoring coverage.
[0033] The income risk based credit score offers many innovative
and significant improvements over traditional credit scores
because: it does not rely histories which may not be accurate and
current since collection of credit transaction data is a tedious
and error prone process which is evidenced by the fact that a
majority of credit bureau reports have errors and all credit
bureaus have different data on the same individual and that in most
instances they come up with different credit scores for the same
individual; it considers the impact of the economy on consumer's
future income and predicts income risk which is not a factor in
current credit scoring models; it is updated monthly using the
latest economic data since economy impacts consumers' future
income; and it can score every individual irrespective of their
credit histories.
[0034] The income risk based credit scoring model's databases are
updated monthly by using as updated assessment of economic
conditions and how the new conditions are going to impact consumers
future income or income risk, allowing the most current information
to be used by lenders, businesses, and others. Thus, instead of
waiting for negative items to appear on a consumer's credit report,
the income risk based credit score quantifies the source of credit
risk, and that is the interaction between the economy and
consumer's income prospects, enabling lenders to get an accurate
assessment of consumer's potential for defaulting on a payment.
[0035] In another preferred embodiment of the income risk based
credit score, the Job Security Score is usable as a prospect score
which predicts response rates and improves acquisitions by allowing
lenders and businesses to identify better prospects who are more
likely to respond to a marketing offer and become better consumers.
When the Job Security Score is combined with other scores, the
quality of predictions increase and therefore businesses can better
target their products and services to consumers that fit best with
their ability to pay.
[0036] In alternative embodiments, the present invention of income
risk model may also involve the use of input variables such as 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.
[0037] In short, Applicant provides a comprehensive consumer future
behavior prediction model that employs a number of novel methods to
accurately forecast and implement the ability to pay component in
consumer credit risk scoring models and to increase accuracy of
prospect scoring models by predicting income risk of consumers. The
income risk based credit score alone or in combination with credit
bureau scores, e.g. FICO, and prospects scores will lead to an
enhanced power to discriminate and segment consumers, improving
profitability for businesses. By using the income risk based credit
score, the total dollar losses due to greater identification of
potential charge-off/bankrupt accounts will decrease, the good
accounts volume will increase, and the portfolio performance will
improve. By using the income risk based credit score, decisions on
decreasing the credit line for these accounts can be made before
these accounts become problematic and loss prone. Similarly, using
the combined income risk based credit score and credit bureau
scores, fewer good accounts will be targeted for a reduction in
credit line decreases. Good accounts can also be granted credit
line increases due to the enhanced risk separation and
discrimination. Thus, through the present invention, the Applicant
uniquely addresses the missing component of accurately predicting
consumers' payment default risk, or credit risk, using future
ability to pay using income risk, allowing credit scoring models to
make better, accurate determination of consumers' credit risk.
DESCRIPTION OF THE DRAWINGS
[0038] For the present invention to be clearly understood and
readily practiced, the present invention and its embodiments will
be described with the following figures, wherein like reference
characters designate the same or similar elements, which figures
are incorporated into and constitute part of the specification,
wherein:
[0039] FIG. 1 is a flow chart showing the overall process of the
invention;
[0040] FIG. 2 is a diagram depicting key drivers of consumer credit
risk of which the ability to pay is predicting by the
invention;
[0041] FIG. 3 is a table showing the consumer payment risk
matrix;
[0042] FIG. 4 is a chart demonstrating the two key drivers of
consumers' total credit risk for unsecured loans;
[0043] FIG. 5 depicts exemplary components of the consumer income
risk score generated by the system;
[0044] FIG. 6 is a diagram showing the steps involved in computing
the consumer's income risk based credit score by the invention;
[0045] FIG. 7 is a diagram showing the steps involved in creating a
consumer income risk scoring model by the invention;
[0046] FIG. 8 is a diagram showing the steps involved in creating
an income risk based ability to pay risk model by the
invention;
[0047] FIG. 9 is a diagram showing the steps involved in the
computation of a consumer income risk based credit score;
[0048] FIG. 10 describes the limitations of traditional Credit
Bureau Scores and shows the components of one of its embodiment,
the FICO score;
[0049] FIG. 11 describes the key features and limitations of
alternative credit scores and lists major alternative scores in
existence today;
[0050] FIG. 12 is a table describing existing credit scoring models
and their limitations;
[0051] FIGS. 13 and 13A are charts comparing loss curves for Income
Risk Based Credit Score and a traditional credit score;
[0052] FIG. 14 is a chart showing odds ratios for Income Risk Based
Credit Score;
[0053] FIG. 15 is a table showing superior risk segmentation
capability of Income Risk Based Credit Score over traditional
credit score for payment default risk;
[0054] FIG. 16 is a chart showing Income Risk Based Credit Score
and its ability to predict mortgage insurance claims;
[0055] FIG. 17 is a chart illustrating LEHI's ability to track GDP
because LEHI economic indicator can be used by the invention in
computing Income Risk Based Credit Score;
[0056] FIG. 18 is a chart showing JSI values for sample ZIP codes
which can be used by the invention in computing Income Risk Based
Credit Score;
[0057] FIG. 19 is a chart which shows that credit risk truly
depends on consumers' future ability to pay and their future
willingness to pay;
[0058] FIG. 20 is a chart showing LEHI's ability to predict local
economic health;
[0059] FIG. 21 is a chart showing Income Risk Based Credit Score
(its embodiment as JSS) and its ability to payment default risk
(delinquency rate);
[0060] FIG. 22 is a chart comparing the predictive power of Income
Risk Based Credit Score (its embodiment as JSS) and traditional
credit score (its embodiment as VAN score) for payment default risk
(loss rate or bad rate);
[0061] FIG. 23 is a table comparing the predictive power of Income
Risk Based Credit Score (its embodiment as JSS) and its
effectiveness as a credit score using statistical analysis (KS
stats);
[0062] FIG. 24 is a chart for Income Risk Based Credit Score (its
embodiment as ISS) versus traditional response score (its
embodiment as RESP SCR) for prospect scoring;
[0063] FIG. 25 is a chart comparing Income Risk Based Credit Score
(its embodiment as ISS) and a traditional response score (its
embodiment as RESP SCR) for predicting prospect response rates;
[0064] FIG. 26 is a chart for Income Risk Based Credit Score (its
embodiment as JSS) and traditional credit score (its embodiment as
VAN score) comparing payment default risk (loss rate or bad rate)
for prescreened accounts;
[0065] FIG. 27 is a chart for Income Risk Based Credit Score (its
embodiment as JSS and ISS) versus traditional response score (its
embodiment as RESP and VAN SCR) for prospect scoring;
[0066] FIG. 28 shows two tables comparing K-stats for Income Risk
Based Credit Score (its embodiment as JSS and ISS) and a
traditional response score (its embodiment as RESP and VAN SCR) for
prospect scoring;
[0067] FIG. 29 is a chart showing the relationship between Income
Risk Based Credit Score (its embodiment as JSS_IN) and consumer
delinquencies;
[0068] FIG. 30 is a chart comparing effectiveness of Income Risk
Based Credit Score (its embodiment as ISS) and traditional credit
score (its embodiment as Vantage) in predicting response rates;
[0069] FIG. 31 is a chart comparing effectiveness of Income Risk
Based Credit Score (its embodiment as ISS) and a response score in
predicting response rates;
[0070] FIG. 32 is a chart comparing Income Risk Based Credit Score
(its embodiment as ISS) and traditional credit score (its
embodiment as Vantage) in predicting payment default risk (bad
rates);
[0071] FIG. 33 shows two charts comparing the effectiveness of
Income Risk Based Credit Score (its embodiment as JSS) and
traditional credit score (its embodiment as CBS score) in
predicting payment default risk;
[0072] FIG. 34 shows two charts comparing the effectiveness of
Income Risk Based Credit Score (its embodiment as JSS) and
traditional credit score (its embodiment as CBS score) in
predicting payment default risk;
[0073] FIG. 35 shows tables comparing Income Risk Based Credit
Score (its embodiment as JSS) with traditional credit score (its
embodiment as CBS score) for payment default risk;
[0074] FIG. 36 is a table comparing Income Risk Based Credit Score
(its embodiment as JSS) with traditional credit score (its
embodiment as CBS score) for payment default risk;
[0075] FIG. 37 is a chart comparing Income Risk Based Credit Score
(its embodiment as JSS) versus traditional credit score (its
embodiment as CBS score) for payment default risk;
[0076] FIG. 38 is a chart comparing Income Risk Based Credit Score
(its embodiment as JSS) versus traditional credit score (its
embodiment as CBS score) for payment default risk and ability to
identify high risk accounts (bads);
[0077] FIG. 39 is a chart comparing Income Risk Based Credit Score
(its embodiment as JSS) versus traditional credit score (its
embodiment as CBS score) for cumulative bad rates;
[0078] FIG. 40 shows two charts showing Income Risk Based Credit
Score (its embodiment as JSS) and its ability to predict payment
default risk (delinquencies) in 3 months and 6 months from the time
of booking the accounts;
[0079] FIG. 41 is a diagram showing Income Risk Based Credit Score
(its embodiment as JSS) and its ability to predict payment default
risk (delinquencies) for existing accounts;
[0080] FIG. 42 is a chart showing Income Risk Based Credit Score
(its embodiment as JSS) and traditional credit score (its
embodiment as Riskscore) and their abilities to predict payment
default risk (first payment default or FPD default rates);
[0081] FIG. 43 is a chart showing Income Risk Based Credit Score
(its embodiment as ISS) and its ability to predict customer
conversion rate; and
[0082] FIG. 44 is a diagram of the computer implemented system for
generating and providing an Income Risk Based Credit Score.
DESCRIPTION OF THE PREFERRED EMBODIMENTS
[0083] It is to be understood that the figures and descriptions of
the present invention have been simplified to illustrate elements
that are relevant for a clear understanding of the invention, while
eliminating, for purposes of clarity, other elements that may be
well known. Those of ordinary skill in the art will recognize that
other elements are desirable and/or required in order to implement
the present invention. However, because such elements are well
known in the art, and because they do not facilitate a better
understanding of the present invention, a discussion of such
elements is not provided herein. The detailed description will be
provided herein below with reference to the attached drawings.
[0084] Generally speaking, the present invention provides systems
and methods for a novel income-risk based credit scoring system to
predict consumers' credit risk by using their income risk, which is
key driver of credit risk.
[0085] Consumers are able to pay their debt obligations when they
have money (FIG. 3), and consumers typically rely on a steady
source of income to be able to manage their financial obligations.
When their source of income disappears, reduces, or is adversely
impacted, which is usually because of job loss or change, then the
consumer's ability to pay is diminished and their probability of
payment defaults increases, which results in the fact that they
become less creditworthy and present higher credit risk to their
lenders.
[0086] FIG. 3 is a table showing the consumer payment risk matrix
which shows that consumers make a payment if they have both the
willingness and ability (301) and do not make a payment if they
don't have the willingness or the ability (302 and 304) or have
neither the willingness nor the ability (303).
[0087] Until recently, consumers' income risk and their ability to
pay risk were not defined or included in credit scoring models
(FIG. 11 and FIG. 12), making it impossible to predict a consumer's
true total credit risk. FIG. 11 describes the key features and
limitations of alternative credit scores (1101) and lists major
alternative scores in existence today (1102). FIG. 12 is a table
describing existing credit scoring models and their limitations
which consist of credit bureau scores (1201), lenders' internal and
custom credit scores (1202), and alternative credit scores
(1203);
[0088] The present invention is the first to develop a computer
implemented system for quantifying (FIG. 44) consumers' credit risk
due to income loss by providing a credit scoring system using
unemployment risk probability, and using income risk (FIG. 7),
called the income risk based credit score (FIG. 1 and FIG. 9), of
which the Job Security Score is a preferred embodiment. The Job
Security Score predicts the probability of an individual defaulting
on credit obligations by factoring in the probability of income
loss.
[0089] FIG. 44 is a diagram of the computer implemented system for
generating and providing an Income Risk Based Credit Score and its
embodiments and it comprises of a database bank (4401) consisting
of a consumer database, economic database and a modeling database;
a data processing unit (4402) consisting of a modeling server, a
communication server, and a database server, a memory storage unit
(4403); a consumer score generating computer (4404) which produces
the Income Risk Based Credit Score and its embodiments; an
administrator workstation computer (4405) which manages the access,
read, and write privileges to various user groups in the system
connected through an internal LAN (4408); an internet and VPN
connection (4406); access to client computer (4409) for exchange of
consumer data and scores; and access to external databases (4407)
for exchange of consumer data and scores.
[0090] FIG. 1 is a flow chart showing the overall process of the
invention. Consumer income risk (101) is the probability that a
consumer will have the necessary income to pay their debts, i.e.
consumer's ability to pay risk (102). Factored into the risk
calculations are the consumer's payment default data (103), which
can be comprised of at least the following: credit card default
data (106), mortgage default data (107), auto loan default data
(108), and other debt default data (109). Block 104 depicts the
step whereby the invention analyzes and correlates consumers'
payment default data with their income risk using statistical
models predicting consumer credit risk (105). The invention
generates a consumer income risk based credit score (110).
[0091] FIG. 7 is a flow chart depicting the steps to be carried out
in the consumer data analysis process in order to generate the
novel consumer income risk scoring model (707). The consumer data
(701) is used to place the consumer in a risk group, wherein
consumers with similar attributes are placed in risk groups (703),
and is analyzed with historical unemployment data (704), historical
income disruption data (705), and historical economic data and many
economic indicators (702). The invention then analyzes, correlates,
and establishes historical mathematical relationships (706) between
consumer's unemployment and income disruption data and economic
data resulting in the consumer income risk scoring model (707).
[0092] FIG. 9 is a diagram showing the steps involved in the
computation of a consumer income risk based credit score (909)
which comprises of developing a consumer income risk scoring model
(901) to generate an income risk score (902), developing a consumer
ability to pay scoring model (903) to generate an ability to pay
score (904), using consumers' historical payment default data for a
large number of actual loans (905), establishing mathematical
relationships between income risk, ability to pay, and payment
default data (906), predicting payment default risk for consumer
risk groups, categories and subcategories (907), and developing a
consumer income risk based credit scoring model (908).
[0093] Two fundamental concepts of consumer credit risk are the
"willingness" to make repayment on a loan and the "ability" to
repay (FIG. 2 and FIG. 4). While willingness can be judged by past
credit behavior, the ability to pay is dependent upon external
factors beyond the consumers' control (e.g. loss of income, medical
problems, divorce, unemployment, etcetera) of which the probability
of continuance of future income, or income risk, due to
unemployment is the one of the biggest factors.
[0094] FIG. 2 is a diagram displays the three main categories of
the credit risk. They are collateral (201), willingness to repay
(202) and ability to repay (203), only two of which (collateral and
willingness) have been previously incorporated in prior art. As
credit cards are not secure loans, the willingness to pay alone
will not tell the lender all the pertinent information about a
consumer. A reliable indicator of future economic ability is
necessary in order to determine the true credit risk of the
consumer. Through the invention, the prediction of consumers'
ability to pay makes lenders' decisions more accurate.
[0095] FIG. 4 is a chart demonstrating the two key drivers of
consumers' total credit risk (403) for unsecured loans comprising
of willingness (401) and ability (402).
[0096] Since credit risk is truly a measure of payment default
probability in future, future willingness and future ability to pay
are the real drivers of consumer credit risk than past willingness
and past ability (FIG. 19). Thus, there is an unmet need in the
marketplace to incorporate the ability to pay risk in credit
scoring models in order to decrease credit lending risks.
Recognizing this unmet need, the present invention provides a
highly predictive model that predicts the ability to pay component.
This results in a new consumer risk score, the income risk based
credit score, of which the Job Security Score is a preferred
embodiment, which predicts consumers' payment default risk using
consumers' income risk as a factor.
[0097] FIG. 19 is a chart which shows that total credit risk (1903)
truly depends on consumers' future ability to pay (1902) and their
future willingness to pay (1901) rather than on their past
willingness and past ability, and this is a critical distinction
that needs to be understood because credit risk by definition is an
assessment of future payment default probability so what really
matters is the future willingness and ability of the consumer to
make payments.
[0098] By monitoring economic conditions, and establishing
relationships between economic activity, consumers' income,
consumers' financial behavior, consumers' ability to pay,
consumers' ability and willingness to buy, and consumers' wellbeing
(FIG. 8 and FIG. 9), the present invention has been able to
quantify consumers' credit risk that is dependent on their income
risk, and this approach is a novel one which credit bureaus have
yet to conceive.
[0099] FIG. 8 shows an exemplary capacity to generate an ability to
pay risk model. Consumer profile data (801), consumer unemployment
histories (802), and income disruption histories (803) were the
first elements used in the process to predict consumers' income
disruption risk. Then, consumers financial stress data is
correlated with income disruption risk (804) and mathematical
relationships are established between economic conditions and
consumer financial stress (805). Said combination creates the
statistical model predicting consumers' financial stress for
present and forecasted economic conditions (806). Said historical
and forecasted economic data (807) integrated with consumer profile
data results in the income risk based ability to pay risk model
utilized by the invention. Said model can then be used by lenders
in making consumer credit based decisions.
[0100] Jobs and unemployment incidents are affected by the economy.
During weak cycles, demand is down, whereas during strong cycles,
demand is up. From an individual perspective, the job loss risk is
a function of the supply and demand in the labor market, which is
driven by economic conditions (FIG. 17, FIG. 18, and FIG. 20).
[0101] FIG. 17 is a chart showing LEHI's (Local Economic Health
Indicator) ability to track GDP as LEHI, which can be optionally
used in computing Income Risk Based Credit Score. Such current
economic information being updated monthly by the invention
produces the most current statistics on which lenders can base
their decisions. The lower the LEHI of an area, the higher the
payment default risk is for that area.
[0102] FIG. 18 is a chart showing how JSI (Job Security index)
varies by ZIP codes. This is valuable information which can be used
in computing Income Risk Based Credit Score.
[0103] FIG. 20 is a chart showing LEHI's ability to predict local
economic health. This chart demonstrates that national economic
health and regional economic health, such as for Stockton,
California's MSA can be quite different and economic variations
such as these can be incorporated into the present invention for
predicting consumer income risk.
[0104] Credit bureau models do not adapt or reflect the changing
economy. A delay of around 12 months or more is common in the
credit bureau models and the changes are only considered as a
derivative effect resulting from consumers' altered credit
behavior. Yet, with a continually growing amount of consumer data
being correlated into the system, Applicant predictions become more
accurate, giving the lender greater ability to analyze consumer
risk (FIG. 5 and FIG. 6). Thus, models that incorporate the Job
Security Score will adapt quickly to reflect the latest economic
conditions and forecasts, providing a more accurate and "true"
credit risk prediction.
[0105] FIG. 5 depicts exemplary elements of the consumer income
risk score (or income disruption risk) (501) analyzed by the
invention. The consumer income risk score is the probability of the
consumer's future employment (and therefore income) calculated
twelve (12) months in advance and reduced into a score. Said score
is a measure of the likelihood the consumer will be able to repay a
debt and comes in the form of a number between -1000 to +1000, or
any other numerical or non-numerical range or scale. Said score is
independently assessed to each individual consumer and it varies
based on consumer's income risk and credit risk. Such an accurate
indicator of ability to pay for each consumer is essential is a
core aspect of the current invention. The following elements are
incorporated into the income risk score: The consumer's
unemployment risk (502), consumer's income loss risk (503),
consumer's income reduction risk (504), and consumer's income
variability and volatility risk (505). The consumer's probability
of a becoming unemployed and experiencing a reduction in their
income is then weighted and incorporated into the final consumer
income risk score.
[0106] FIG. 6 is a flow chart showing the advantages of the current
invention. The invention takes a selected consumer (601) and
collects consumer profile data (602) (further described in FIG. 8)
on said customer. This information is then used by the method to
compute the consumer's income disruption risk (603) using outputs
from method described in FIG. 5. Using (601), (602), and (603), the
invention computes the consumer's income risk score (604). Said
credit score can be used for the following purposes: To predict
consumer's payment default risk and credit risk for loans (605)
through the consumer's income risk based credit score (606); and to
estimate consumer's response rate for credit card offers and other
marketing offers (607) through the consumer's income risk based
prospect score (608). The current invention allows lenders the much
needed ability to pay insights for their consumers and potential
consumers. No existing credit scoring model incorporates a
prediction of consumer's income risk in predicting payments
defaults and credit risk.
[0107] Income loss due to unemployment incidents are related to
several factors that can be categorized into three groups:
macroeconomics (e.g. GDP, money supply, M1, M2, energy prices,
etc.), macro-demographics (i.e. factors that pertain to a
population of individuals such as occupation industry, occupation
type, zip code, etc.) and micro-demographics (e.g. income, years at
residence, highest level of education, etc.).
[0108] The present invention provides a unique risk scoring model
of unemployment incidents using vast amounts of economic data and
actual consumer data (FIG. 5 and FIG. 6). With a proprietary data
collection originating 6 years ago, the present invention is able
to collect data on consumers' historical unemployment incidents,
prevailing economic conditions and historical trends, consumers'
post unemployment financial stress situation, unemployment severity
and payment defaults data. Using sophisticated data-mining
techniques and statistical algorithms, the predicative analytical
model finds patterns and relationships between economic indicators
and an individual's profile to predict the person's likelihood of
income loss due to unemployment within the next 12 months (FIG. 8
and FIG. 9). The Job Security Score has been developed using
thousands of actual individual profiles, hundreds of macroeconomic
variables covering decades of local, regional, and nation economic
trends, and the credit behavior of millions of actual borrowers
(FIG. 7). Analysis proved that the unemployment risk scores used by
the present invention are over 85% accurate in predicting
unemployment risks twelve months in advance and are better
predictors of consumers' payment default risk.
[0109] Previously, the consumer's ability to pay risk was not
defined or included in credit scoring models and that they had many
limitations (FIG. 10) making it impossible for them to predict
consumers' income risk and therefore they were unable to predict
consumers' true credit risk. Inclusion of income risk by lenders
and credit card issuers in their credit matrix will significantly
improve the accuracy and effectiveness of their credit risk
prediction capabilities. FIG. 10 describes the limitations of
traditional Credit Bureau Scores (1001) and shows the breakdown of
one of its embodiment, the FICO score (1002).
[0110] In one preferred embodiment, the invention provides an
application of statistical algorithms to find patterns and
relationships between economic indicators to predict the likelihood
of future events with high levels of accuracy. The invention
quantifies the inherent income risk in the form of an income risk
score. This information can then be used by the income risk based
credit score, of which the Job Security Score is a preferred
embodiment, to predict consumer behavior, delinquency, charge-off
risk, spending trends, likelihood of on-time payments,
effectiveness of products or services, and virtually any other
factor that can be statistically analyzed and related to consumer's
income in making superior assessment of credit risk over
traditional credit scores (FIGS. 13, 13A, 15, 16, 21, 22, 23, 26,
27, 29, 30, 31, 32, 33, 34, 35, 36, 37, 38, 39, 40, 41, and
42).
[0111] FIG. 14 is a chart shows odds ratios for payment default
risk for the Income Risk Based Credit Score (its embodiment as
JSS). As shown, the JSS is able to predict and rank order payment
default risk very well.
[0112] FIG. 15 is a table which shows how the Income Risk Based
Credit Score (its embodiment as JSS) is able to identify more good
accounts without increasing a lender's loss rate. The JSS allows
the lender to decrease their existing credit score cutoff (its
embodiment as Custom Score) and yet not increase their loss risk
because of JSS's ability to segment good accounts (in blue) and bad
accounts (in red) that was not possible before.
[0113] FIG. 16 is a chart showing Income Risk Based Credit Score
(its embodiment as JSS) and its ability to predict mortgage
insurance claims (Payment defaults). In using the consumer score in
connection with mortgages, lenders can acquire a more informed
rational for a mortgage decision. The higher the consumer score,
the more likely they are to repay their debts.
[0114] FIG. 21 is a chart showing Income Risk Based Credit Score
(its embodiment as JSS) and its ability to predict payment default
risk (delinquency rate). As can be seen, the JSS is able to predict
and rank order delinquencies very well.
[0115] FIG. 22 is a chart showing loss curves for Income Risk Based
Credit Score (its embodiment as JSS), a credit bureau score (its
embodiment as VAN score), and for a combined JSS+VAN score. As can
be seen, the JSS is able to increase good accounts by 15% and
decrease bad accounts by 11% clearly demonstrating that JSS offers
new credit risk insights.
[0116] FIG. 23 is a table showing KS-stats (a higher KS indicates
better predictive power) for Income Risk Based Credit Score (its
embodiment as JSS), a credit bureau score (its embodiment as VAN
score), and for a combined JSS+VAN score. As can be seen, the JSS
is able to increase KS-stats by 50% over existing VAN score.
[0117] FIG. 26 is a chart showing loss curves for Income Risk Based
Credit Score (its embodiment as JSS), and a credit score (its
embodiment as VAN score) and for a combined JSS+VAN score. As can
be seen, the JSS is able to increase good accounts by 25% and
decrease loss rates by 5% clearly demonstrating that JSS offers new
credit risk insights and has superior risk separation
capabilities.
[0118] FIG. 27 is a chart showing loss curves for Income Risk Based
Credit Score (its embodiment as JSS and ISS), a response score (its
embodiment as RESP score), a credit score (its embodiment as VAN
score), and for a combined ISS+RESP score As can be seen, the ISS
can be used to approve good accounts that were declined by the use
of existing credit scores.
[0119] FIGS. 29 to 42 show various comparisons between Income Risk
Based Credit Score (its embodiment as JSS and ISS), a response
score (its embodiment as RESP score), a credit score (its
embodiment as FICO and VAN score) and demonstrate how the Income
Risk Based Credit Score is able to add new consumer insights in
credit scoring and prospect scoring.
[0120] The present invention predicts income-loss risk by measuring
how economy is changing and how it impacts consumers' income and
income risk to forecasts future paying capacity--or paying
ability--of consumers. This offering of new and in-depth
comprehension into future consumer credit behavior has not yet been
captured by any credit bureau scores or other credit scores. In
capturing economic impact on an individual consumer's ability to
pay, the present invention utilizes elements not previously
available in the field. Further, the Job Security Score can be used
as the primary and sole credit score to predict credit risk or it
can be used in conjunction with current scoring methods (FIGS. 13
and 13A). By creating a way to predict a separate aspect of
consumer risk originating from their income risk, the present
invention greatly enhances the accuracy of consumer credit risk
assessment. FIGS. 13 and 13A are charts comparing Income Risk Based
Credit Score (its embodiment as Job Security Score or JSS) against
a traditional credit score (its embodiment as FICO) for payment
default risk and they show that JSS improves risk prediction. The
Job Security Score (or JSS) is an embodiment of the invention and
measures the ability to pay, and credit risk, of an individual.
This chart shows loss curves for JSS and FICO and it demonstrates
the superior ability of the JSS over FICO in predicting and
segmenting good and bad accounts.
[0121] The income risk based credit score, of which the Job
Security Score is a preferred embodiment, is also a prospect score
and capable of scoring any consumer in the marketplace. All
existing scoring models and methods are reliant on credit data
and/or payment data, but none include income risk except present
invention. In implementing a scoring model that is accurate in its
credit risk assessments, lenders can improve risk assessment by an
average of 11%. Testing shows an average of 9% lift for thick file
credit card portfolios and 30% lift for thin-file and no-hit credit
card portfolios. Considering the approximately 70 million consumers
who are not scored have a much high delinquency rate than the
average consumer, the income risk based credit score, of which the
Job Security Score is a preferred embodiment, will greatly benefit
lenders.
[0122] In one preferred embodiment, the income risk based credit
scoring model generates a Job Security Index (JSI) that measures
the job conditions for a MSA/Zip location. By correlating with
consumer spending, credit charge-offs, and delinquencies, the JSI
is a useful prospective score when evaluating how consumer behavior
is likely to change with fluxes in the economy. Because the Job
Security Score is updated on a monthly basis, the JSI is able to
reflect current conditions (FIG. 18). Additionally, the JSI can be
applied to any consumer lists in order to identify better
prospects.
[0123] In another preferred embodiment, the present invention will
improve acquisitions because of better-informed targeting and
segmentation of prospects which the invention provides (FIG. 24,
FIG. 25, FIG. 28, and FIG. 43).
[0124] FIG. 24 is a chart for Income Risk Based Credit Score (its
embodiment as Income Stability Score, or ISS in short) versus
traditional response score (its embodiment as RESP SCR) for
prospect scoring.
[0125] FIG. 25 is a chart showing response rate curves for Income
Risk Based Credit Score (its embodiment as Income Stability Score,
or ISS) and a response score (its embodiment as RESP score). As can
be seen, the ISS is able to increase response rate by 4%.
[0126] FIG. 28 is a table showing KS-stats for Income Risk Based
Credit Score (its embodiment as JSS and ISS), a response score (its
embodiment as RESP score), a credit score (its embodiment as VAN
score), and for a combination of above scores. As can be seen, the
ISS is able to increase KS-stats by 49% and 22% over existing RESP
scores.
[0127] FIG. 43 is chart shows the income risk based credit score
(its embodiment as ISS) and its ability to predict customer
conversion rate. The ISS can be used for alternative purposes, such
as identifying customer conversation rate as displayed.
[0128] When lenders use income risk based credit score, pricing,
loan amounts, and product offering decisions will be more effective
and profitable due to new credit risk insights not available
through conventional credit bureau scoring. Further, delinquencies
and charge-offs will decrease due to early identification of high
risk accounts. Businesses can now market to more promising
prospects and approve more applicants based on the "true" overall
credit risk, rather than basing their decisions on an outdated view
of credit risk offered by credit bureau scores alone. Prospect
identification processes will be more effective as most identified
prospects will match the ideal customer profile requirements. The
offer response rate will improve as well because of the greater
precision in targeting. The approval decision process is enhanced
due to better separation of applicant's credit risk. By gaining
insights into ability to pay risk using Job Security Score,
businesses can now get a complete picture of consumer's credit risk
thereby improving their portfolio's size and quality while giving
them a significant competitive advantage.
[0129] It is noted that there are an infinite number of ways to
create homogenous classes of people with similar risks for the
millions of people nationwide. Because there never has been a
personal unemployment risk score and a consumer income risk score,
there is no actuarial data available by any established risk
classes related to unemployment rates. Therefore, the present
invention also presents a method and model to segment the labor
force into homogenous unemployment risk classes and establishes
empirical relationships between historical unemployment rates and
risk classes and income risk.
[0130] Another aspect of the present invention provides account
management strategies using the Job Security Score. By monitoring
the Job Security Score of accounts on a constant basis (scores are
updated every month) lenders can identify a high risk account
before the account actually becomes a high risk/bad account--unlike
credit bureau scores which deliver the news after the account has
negative items on file. Thus, the new predictive power enables
lenders the time and insights to take strategic initiatives to
manage and mitigate the risk before it is too late.
[0131] In an additional embodiment, the present invention provides
for better-informed credit line decisions. Both credit line
decrease decisions and balance build/balance transfer offer
decisions are most effective and profitable when they are based on
the latest, most complete, and most accurate assessment of
consumer's risk. Since Job Securities Scores change every month,
even for consumers whose profiles remain unchanged, they reflect
the latest economic conditions that may impact jobs and income
prospects. Lenders will always have the latest and most accurate
risk assessment possible, allowing them to deploy more profitable
credit line increase and decrease strategies.
[0132] When the Job Security Score is combined with credit bureau
scores, or other internal risk scores, it redistributes the
population in such a way that lenders can lower their cut-off
credit bureau scores allowing for more approvals, without lowering
the risk threshold (FIGS. 13, 13A, 14, and 15). In fact, at the
same time, the risk profile of the portfolio screened and managed
using a combined score (e.g. Job Security Score+FICO) decreases
because risk assessment is more accurate, leading to an increase in
portfolio quality and a decrease in losses (FIGS. 13, 13A, 14, 15,
16, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32, 33, 34, 35, 36,
37, 38, 39, 40, 41, 42, and 43).
[0133] The present invention also preferably provides improved
marketing capabilities for businesses. With the advanced prospect
scoring, marketing strategies can be more focused and target the
ideal populations. In addition, the present invention enhances any
prescreening of an individual and can be used to strengthen
predictions (FIGS. 24, 25, 28 and 43). This allows for the early
identification of high risk individuals, narrowing the delinquency
probabilities and making marketing more sophisticated.
[0134] The present invention also preferably provides lenders with
the following benefits: a more accurate picture of consumer's
credit risk, in both good and bad economic times; a proactive and
leading indicator of credit risk (unlike credit bureau scores,
which are reactive and lagging); and improved segmenting and
differentiation due to better credit risk prediction capabilities
(FIGS. 13, 13A, 14, 15, 16, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30,
31, 32, 33, 34, 35, 36, 37, 38, 39, 40, 41, 42, and 43).
[0135] As described above, the present invention provides methods
for implementing an unique indicator of consumer credit risk
stemming from the ability to pay risk associated with each and
every individual applying for credit. Its applications are in all
those areas that involve credit risk assessment or predicting
credit dependent behavior.
[0136] It is also to be understood that this invention is not
limited to using the data, records, data elements, variables and
field structures described herein, and other data elements, data,
and physical structures will be equivalent for the purposes of this
invention. The invention has been described with reference to a
preferred embodiment, along with several possible variations;
however, it will be appreciated that a person of ordinary skill in
the art can effect further variations and modifications without
departing from the spirit and the scope of the invention.
[0137] Nothing in the above description is meant to limit the
present invention to any specific materials, geometry, or
orientation of elements. Many part/orientation substitutions are
contemplated within the scope of the present invention and will be
apparent to those skilled in the art. The embodiments described
herein were presented by way of example only and should not be used
to limit the scope of the invention.
[0138] Although the invention has been described in terms of
particular embodiments in an application, one of ordinary skill in
the art, in light of the teachings herein, can generate additional
embodiments and modifications without departing from the spirit of,
or exceeding the scope of, the claimed invention. Accordingly, it
is understood that the drawings and the descriptions herein are
proffered only to facilitate comprehension of the invention and
should not be construed to limit the scope thereof.
* * * * *