Financial institutions typically rely on traditional credit scores and bank statements to drive their risk models. These are inadequate when it comes to assessing a borrower’s creditworthiness, or the likelihood of timely loan repayment. Inaccurate assessment of risk has led to a rising number of loan defaults and fraud. Lenders need to find new ways to measure default risk.
The more the financial landscape changes, the more essential accurate risk modeling becomes. Financial institutions are looking to newer, updated, and more relevant data points for data augmentation to better assess their business risk. Even the most sophisticated machine learning models are insufficient without the most relevant risk data signals, pulled from a wide variety of data sources.
Some examples of alternative data for lending are:
By incorporating alternative data into risk models, lenders can retrain existing risk models with new datasets and reduce their default rates. Improved risk modeling can also lead to better operational efficiency; assessing credit risk more accurately means less personnel and budget allocated to recoup bad loans. One of our customers was able to reduce their default rate from 30% to 21%. More importantly, they have been able to vet better potential borrowers and extend safer loans that are more likely to be repaid. The result is a revenue boost of nearly 4% in the six months following the adoption of their new risk model.