So, what’s your rating?
term personal loans, has developed its own algorithm to determine the credit-worthiness of its customers, most of whom are millennials with little or no banking history.
“The current banking-based system of giving loans based on credit ratings is a dumb system,” says CASHE founder V. Raman Kumar. “It is entirely dependent on former credit activity. We felt this had to be redefined.”
The company has devised a ‘Social Loan Quotient’ (SLQ), a social behaviour-based alternate credit rating system that Raman says uses 15-200 data points to determine a score on a 1000. The CASHE app uses “rich phone data” to track user behaviour such as phone usage, Linkedin and Facebook profiles, what kind of apps you have, how often you use them, whether you use e-commerce and financial apps etc. SMS gives access to spending history thanks to bank SMSS and OTPS required for online transactions. It tells them whether you have paid EMIS and SIPS, and if you defaulted on your Netflix subscription.
Besides this, they also look at your “cohorts”—for instance, people who work in the same company as you or are from the same educational institutions. Kumar says they are building the SLQ as an industry-wide score that may come to replace or supplement conventional credit scores like Cibil. “Earlier, loans were asset-driven. Banks wanted physical security. But today’s transactions are digital; there’s less focus on owning and more on renting. If banks are not looking at all this, they need to change their behaviour,” says Raman.
When Botsman says reputation is going to become a currency more powerful than our credit scores in the real world, she’s not too far off the mark. But one of the flaws in ratings is systemic. Apart from concerns over privacy and ways in which biases can creep in, there’s the phenomenon of “reputation inflation”.
In a March 2018 paper with the same title, researchers Apostolos Filippas, John J. Horton and Joseph Golden from New York University’s Stern School of Business argued that the effectiveness of post-transaction ratings in the shared economy deteriorates over time.
“The problem is that ratings are prone to inflation, with raters feeling pressure to leave ‘above average’ ratings, which in turn pushes the average higher. This pressure stems from raters’ desire to not harm the rated seller. As the potential to harm is what makes ratings effective, reputation systems, as currently designed, sow the seeds of their own irrelevance,” says the paper.
The researchers note how on Uber and Lyft, it is widely known that anything less than 5 stars is considered “bad” feedback, and this leads to the conclusion that if feedback scores are rising, it could be because of two distinct—but not mutually exclusive—reasons: raters are becoming more satisfied, or raters are lowering their standards. “This second possibility—giving higher scores despite not being more satisfied— can be thought of as a kind of inflation. This ‘ reputation inflation’ potentially makes feedback completely uninformative,” say Filippas, Horton and Golden.
That’s why by making the ratings economy valid criteria for deciding financial and personal worth, we might be putting our trust in a fickle god.
The effectiveness of post-transaction ratings in the shared economy goes down over time