Behavioral Credit Scoring Helps Banks Lend in Uncertain Times

Singapore-based fintech CredoLab makes behavioral credit scoring software. It recently raised $7 million in a Series A funding round for its alternative credit scoring technology. By using machine learning software, CredoLab can predict whether a loan applicant will pay off a loan. It does this by checking the applicant’s smartphone data. A loan applicant with no credit history may use their phone like a borrower who has paid back loans. And once the bank sees this behavior data, it can loan the money to the applicant even if a credit score isn’t available.

How Credit Scores Work

Credit reporting agencies record your payment history and use it to predict whether you will pay off a loan. The credit reporting agency provides this information to the bank in the form of a credit score. If you make your payments on time, your score will rise. If you miss a few payments or make them late, your score will drop. And if you declare bankruptcy you’ll receive a very low credit score.

The bank uses your credit score to decide whether you qualify for a loan. It also uses your credit score to determine the proper interest rate for the loan. And it can also use your credit score to determine the amount of money that you can borrow. Banks consider credit scores when they make several types of loans. They check your score when you take out a mortgage, open up a credit card account, or borrow money to buy a car.

Weaknesses of Credit Scores

A credit reporting agency can only calculate your credit score if you’ve taken out a loan before and have a history of making payments on that loan. So if you’re still in school you may not have a credit score yet. And people without bank accounts who borrow money through informal channels may not have a loan repayment history that’s been reported to the credit bureaus.

Additionally, banks often establish other criteria for loans. For example, they’re more likely to offer you a loan if you earn a steady wage from a job. If you’re a freelancer, contractor, affiliate marketer, or another type of entrepreneur, you may not receive a consistent stream of income each month. In some cases, even if you earn more money from this work than you’d earn as an employee, the bank will still calculate that it’s too risky to offer you a loan.

Behavioral Credit Scoring Provides a Solution

With behavioral credit scoring, the bank has another way to predict if you’ll pay back the loan. If you apply for a loan from a bank that uses CredoLab software, the fintech collects data about how you use the phone. CredoLab CEO Peter Barcak explained how the behavioral credit scoring app works at ATM Marketplace. If you give the app the permissions it wants, the fintech can see which entertainment and communications apps are installed on the phone.

It can see if you have lots of work-related software or play a lot of games. It can see if you started talking to lots of new people recently by counting the number of contacts you have. You don’t have to give it access to everything on your phone, but you have more chances to raise your score if you give the app more information. The CredoLab app only collects metadata, it doesn’t read your text messages or emails.

The app collects millions of data points. That makes it harder to raise your score by doing any single activity. If the app only considered a few factors, like the email app installed on the device or the battery level of the phone, it would be easier to see which behaviors led to a high score. But with many factors making small contributions to the score, it’s harder to manipulate your score. It might be easier to improve your traditional credit score than your behavioral credit score because of the ways the scores are calculated.

Behavioral Credit Scores and the Pandemic

The CredoLab also brought up another topic in his article. Government agencies have declared payment holidays to give borrowers more time to pay back their loans. During a payment holiday, borrowers don’t have to make certain loan payments on time. And credit reporting agencies won’t use these payments in their credit score calculations. So a lender won’t know if a loan applicant isn’t making payments to another bank. And the machine learning models that banks use to calculate default risk may not work correctly during this period. Because of the pandemic, the economy is very different from the economy in 2019.

But the CredoLab behavioral scoring software still works the same way as it did before. It can still predict whether you’ll repay a loan based on smartphone data. If you’re worried about this feature being used against you, remember that this fintech app is opt-in. A behavioral credit score may convince a bank manager to make a loan right now, even though the economic situation is very uncertain.


A behavioral credit scoring app already provides several benefits in the modern economy. Lots of people are freelancers, gig workers, and other types of short-term workers. They might not qualify for a traditional loan already. But during the pandemic, fintech apps like the CredoLab app have become even more useful. With traditional credit scoring information unavailable or less relevant to the current situation, this might be one of the few ways to convince a bank to make a loan during uncertain times.

Leave a Reply

Your email address will not be published. Required fields are marked *