The current credit scoring system is broken, and for the first time in decades the topic is getting the attention it deserves. Just last year American Banker covered
The system works in favor of those who are well off, hold a traditional full-time job and are from the U.S. Individuals who don't fit this demographic are punished with high-interest rates and unfavorable terms.
I have experienced problems plaguing our system firsthand. As an immigrant and entrepreneur, I haven't had a traditional job since I was 19. I have struggled with securing a loan, being approved for an apartment and getting favorable deal terms. I'm from Armenia originally but have lived in Switzerland, the U.K. and the U.S. I have started companies and raised millions of dollars in funding, created jobs and value in the U.S. economy, yet can still be rejected on a mortgage application or a car loan.
At the root of the problem is credit algorithms, a fact that has yet to be widely acknowledged or discussed. Current algorithms have not been trained on samples of people like me, or individuals with similar patterns and financial behavior that don't fit the norm.
This makes me, and others like me, a riskier prospect because the outdated tech simply doesn't understand scenarios, context and modern-day behaviors. Credit models are hard to change — it's costly and incredibly slow. If a company aims to update a scoring system, it will take months and sometimes years to get real data on its performance.
There are many companies aimed at solving this problem and all efforts are well intentioned. Nova Credit is translating international credit scores and helping bridge the credit gap. Argyle, Pinwheel and Atomic are expanding the creditworthiness of different types of workers by allowing lenders to access real-time payroll data. Credit builders like Petal and Tomo Credit are helping borrowers earn trust within the traditional credit systems. Challenger lenders and business models are emerging in the form of buy now/pay later startups.
The issue is that no single financial institution, bank or technology company can cause real change on its own. Every organization has a subset of information they're working with, leaving the picture faulty and incomplete. Contextual data around borrower financial behavior is spread across multiple institutions, and every one of them has its own method of collecting, processing and utilizing it.
Algorithms will always need to be trained on historical data, and so collaboration and coordination of multiple institutions, including banks, fintech companies and other vendors is the crucial first step that must take place. The future requires partnership between incumbents and startups to change our system, with support and encouragement from regulators and lawmakers. Now is the time for the industry as a whole to join together and collaborate on building a fairer financial system.
