AI Case Study
Baidu does underwriting for consumers with limited credit history using machine learning
Baidu partners Zest Finance to use their machine learning platform to underwrite credit risk for consumers with little credit history. The platform analyses data such as payment and purchase history, customer support data etc. They also analyse variables such as how customers fill out forms, how they navigate websites, whether they are being honest about reporting income etc.
"Unlike traditional underwriting methods, ZAML uses machine learning to analyze tens of thousands of nontraditional and traditional variables to more accurately score borrowers, including thin-file and no-file borrowers. ZAML can analyze vast amounts of data they already have in-house, such as customer support data, payment histories, and purchase transactions. The platform can also add traditional credit information and nontraditional credit variables, such as how a customer fills out a form, how they navigate a lender’s site, and more.
Baidu, the leading Internet search provider in China, partnered with ZestFinance in its effort to turn Baidu's search, location, and payment data into credit scores. They are using this data to assess credit worthiness of Chinese millennials and other thin file customers.
Among the things Zest evaluated was how well a person’s self-reported income matched up against their “modeled income,” what Zest calculates that person actually earned based on other behavior. Just as important as how much discrepancy there is between reported and modeled income is when they report the inflated income (in other words, income that's higher than what the model implies they're actually making) and how much they inflated it".
According to Zest:
* Since 2009, more than 300 million customers scored
* Less than ten seconds to render a credit score
Thin-file or No-file consumers with limited credit history have less access to financial institutions. In China, only 20 percent of the population has any known credit history.
"Since 2009, more than 300 million customers scored and more than three thousand variables considered in a standard ZAML
model. For each borrower, the system assesses 100,000 different data points."