ZestCash, an online loan provider, claims that it will be able to lend 25 per cent more money by using a big data-based credit assessment method.
The California-based company, founded by former Google CIO Douglas Merrill, will use data from other creditors such as mobile phone companies and payday lenders to create a more accurate system for determining the outcome of loan applications. By analysing multiple sources of data ZestCash will be able to decide how much to lend an applicant and at what rate.
The lender will be using Hollerith, a new set of underwriting models, to offer credit to 25 per cent more customers and increase repayments from borrowers by 20 per cent.
In the past loan providers would have looked at around 15 points of data to make their decision. Now, using just one underwriting model, the ZestCash decision-making infrastructure can run dozens of models to look at up to 1,000 variables and deliver results within seconds.
Shawn Budde, ZestCash’s COO, said: “We've seen incremental improvements in how underwriting is done, but really the fundamental underpinnings of how credit models work are roughly 40 years old.
“All that math, which is really cool, hasn't been applied to the credit space by anybody else up to this point.”
In the 1st Big Data Insight Group Industry Trends Report, David Chan of City University explained how creating new analytic models using big data would help make far more accurate risk assessment models in deciding the chances of a loan applicant defaulting on payments. The report also features a look at how the financial services sector, specifically British Pearl, has plans to use big data in a very similar way to ZestCash.