PointPredictive Launches IncomePASS-A ML Income Validation That Replaces the Work Number
New Solution Enables Lenders to Retire Expensive, Legacy Income Verification Using the Work Number with a Machine-Learning AI-Based Approach That Can Both Clear Good Income Reports and Identify Inflated Incomes
PointPredictive, the company that solves the billion-dollar problems of auto, mortgage, fintech and retail lending risk, announced the general availability of IncomePASS. The solution offers a real-time assessment of an applicant’s stated income to allow highly accurate clearing of validated incomes. This will enable lenders using expensive automated solutions such as The Work Number® or costly paycheck stub manual reviews across their portfolio to drastically reduce their costs while improving the overall accuracy of their income validation efforts to stop defaults.
Results from lenders testing our models in IncomePASS have been extremely impressive. We encourage lenders to do a head-to-head comparison between IncomePASS and The Work Number.
“Traditional legacy tools are expensive and provide incomplete coverage,” says Tim Grace, CEO of PointPredictive. “Employer-based database verification checks can typically verify income on less than 30% of the applicant submissions, leaving 70% of stated incomes to be confirmed by paycheck reviews, many of which are fake paycheck submissions. Replacing these solutions with IncomePASS will enable 100% coverage across a lender’s portfolio and enable significantly less friction between consumers, lenders and dealers in funding loans. We encourage lenders to do a head-to-head comparison between IncomePASS and The Work Number and measure how many inflated incomes are identified by each solution in total and how many accurate incomes are validated in total and at what cost.”
IncomePASS analyzes a borrower’s stated income against millions of reported incomes and salaries from seven diverse sources. Then, using the applicant’s employer, occupation, job title, residence, and estimated years of experience, a sophisticated machine-learning model predicts the borrower’s likely income. When the borrower’s stated income is within 15% of the model’s prediction, the lender is alerted and can allow the application to proceed to underwriting without additional manual income checks using paychecks or bank statements. The solution also provides a likelihood of default prediction and a misrepresentation indicator to allow lenders to assess when additional characteristics of the applications are likely to lead to loss. This solution allows lenders not only to reduce costs of other solutions and reduce errors associated with manually reviewing thousands of paychecks, but also enables lenders to book more quality loans due to the reduced friction in the process.
“Results from lenders testing our models in IncomePASS have been extremely impressive,” says Mike Kennedy, Vice President and head of the Analytic Scientists at PointPredictive. “We have consistently seen in lender tests that 75% of applications can be cleared as valid incomes using our scoring solution with 90 to 97% accuracy rates. The vast sets of data and the millions of income reports in our data sets have enabled our team to significantly change the game in terms of machine learning techniques we have applied and predictive technology that we have created.”
Results of testing at various lenders are showing that, on average, 75% of applications can be cleared as valid income reports at an accuracy rate of greater than 90%. Additionally, IncomePASS is alerting lenders to income inflation and finding up to 40% of inflated incomes in less than 3% of the application population, leaving far fewer applications to review manually while simultaneously reducing the risk of inflated incomes appearing in the population being reviewed, and catching more suspected inflated incomes than existing legacy solutions.
IncomePASS is available to all financial services providers needing to verify stated income on applications. The service is available today for real-time integration into loan origination and underwriting workflows.
For the next 60 days, tests comparing identification of total number of valid incomes and total number of overstated incomes with IncomePASS to the totals identified with The Work Number will be available at no charge from PointPredictive.