So, what’s your rat­ing?

Mint Asia ST - - News -

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.

Rep­u­ta­tion in­fla­tion

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

Newspapers in English

Newspapers from Malaysia

© PressReader. All rights reserved.