Reg­u­la­tion alone will not work

To stop bank­ing fraud, re­di­rect at­ten­tion to the prob­lem of in­cen­tive and prom­ise of tech­nol­ogy

Hindustan Times (Gurugram) - - Comment - NIRANJAN RAJADHYAKSHA niranjan.r@livemint.com This is the sec­ond in a se­ries of ar­ti­cles on the re­cent con­tro­ver­sies re­lated to PSU banks in In­dia

There is a fun­da­men­tal prob­lem at the heart of In­dian bank­ing — of per­verse in­cen­tives. It de­serves more at­ten­tion when the Ni­rav Modi case has hit the head­lines, amid a pro­found cred­i­bil­ity cri­sis af­ter years of poor lend­ing prac­tices.

The in­cen­tive prob­lem was high­lighted by econ­o­mists Ab­hi­jit Ban­er­jee, Shawn Cole and Es­ther Du­flo in a re­search pa­per, De­fault and Pun­ish­ment: In­cen­tives and Lend­ing Be­hav­iour in In­dian Banks. The three econ­o­mists pointed out that banks need to pro­vide in­cen­tives to their em­ploy­ees to in­crease lend­ing while avoid­ing de­fault, even while these same bank em­ploy­ees earn a frac­tion of the money they man­age. The quest to lend more is of­ten at odds with the need to pro­tect as­set qual­ity, and a lot de­pends on how front-line bankers are re­warded or pun­ished for their ef­forts.

A lot has been writ­ten since the Pun­jab Na­tional Bank case hit the head­lines — on fail­ure of in­ter­nal con­trols, cor­po­rate gov­er­nance and reg­u­la­tions. These are def­i­nitely not triv­ial is­sues. How­ever, an equally im­por­tant ques­tion is how to pre­vent rogue em­ploy­ees buried deep in the branch net­works from col­lud­ing with fraud­sters.

In­dian bankers have a very poorly de­signed in­cen­tive sys­tem. Pay grades are hostage to the iron laws of se­nior­ity. It is al­most im­pos­si­ble to sack an em­ployee. The threat of trans­fer can be used to pun­ish — but it is of­ten used against hon­est em­ploy­ees who do not play the game. All these fac­tors should be seen in the wider con­text of a pub­lic sec­tor cul­ture where per­for­mance of se­nior bankers used to be judged by quan­tity rather than qual­ity, or how much is lent out to meet the po­lit­i­cal pref­er­ences of the gov­ern­ment of the day, rather than the re­turn on cap­i­tal. The on­go­ing bad loan mess is at least partly the re­sult of a credit bub­ble as well as po­lit­i­cally-directed lend­ing to in­flu­en­tial in­fra­struc­ture com­pa­nies.

Would a bet­ter scheme of in­cen­tives change the way bankers give loans? Cole later teamed up with Martin Kanz and Le­ora Klap­per to ex­am­ine how dif­fer­ent in­cen­tive schemes can al­ter banker be­hav­iour. They re­cruited 209 loan of­fi­cers from var­i­ous In­dian banks for an ex­per­i­ment. These bankers were once again shown a ran­dom sam­ple of loan ap­pli­ca­tions from en­trepreneurs, but were asked to treat them as new ap­pli­ca­tions. Since these were pa­pers of loans al­ready given, the re­searchers knew even be­fore the ex­per­i­ment be­gan which ap­pli­ca­tions had been ac­cepted as well as which loans even­tu­ally went bad.

There were three in­cen­tive struc­tures in play. First, an orig­i­na­tion bonus in which of­fi­cers were given a com­mis­sion for every loan is­sued. Sec­ond, a low-pow­ered in­cen­tive in which of­fi­cers re­ceived small re­wards for loans that per­form well and small penal­ties for loans that per­form badly. Third, a high-pow­ered in­cen­tive in which of­fi­cers re­ceived big re­wards for is­su­ing loans that per­form well while they had to face big penal­ties for loans that de­fault.

The re­sults were in­ter­est­ing. The bankers gave out more loans when the stakes were lower but were more care­ful about the qual­ity of lend­ing when the stakes were higher. The in­cen­tive prob­lem was also high­lighted in a lot of re­search fol­low­ing the North At­lantic fi­nan­cial cri­sis of 2008. Bankers were paid fat an­nual bonuses based on how their bets worked in the short term but their losses over the long term were un­der­writ­ten by tax pay­ers.

The in­cen­tive prob­lem should be seen in tan­dem with the op­por­tu­ni­ties pro­vided by tech­nol­ogy. Global ex­pe­ri­ence shows that a lot of bank­ing fraud emerges from the grey ar­eas where data­bases over­lap. It is thus no sur­prise that the re­cent fraud at the Brady House branch of Pun­jab Na­tional Bank in Mum­bai was pos­si­ble be­cause the data on let­ters of un­der­tak­ing were not in­te­grated with the core bank­ing sys­tem, de­spite reg­u­la­tory di­rec­tives to do so. The mis­use of the SWIFT mes­sag­ing sys­tem is also at one level a fail­ure to in­sti­tute con­trols on the use of tech­nol­ogy.

Bank­ing to­day is about the move­ment of bi­nary dig­its, and bank fraud is more than a forged cheque these days. Banks should be able to use ad­vanced data an­a­lyt­ics to iden­tify breaches of in­ter­nal con­trols on the one hand and out­lier lend­ing or guar­an­tee ac­tiv­ity on the other. Jayant Verma of the In­dian In­sti­tute of Man­age­ment, Ahmed­abad, has even made the rad­i­cal sug­ges­tion that banks should use Blockchain tech­nol­ogy to track in­ter­nal ac­tiv­ity — though he adds that the prin­ci­ple of bank­ing se­crecy would have to be re-ex­am­ined be­fore this can be done.

All the talk right now is about reg­u­la­tion. At least some of it needs to be redi­rected to the prob­lem of in­cen­tive and the prom­ise of tech­nol­ogy.

PTI

The closed of­fice of the Pun­jab Na­tional Bank’s Brady House branch, Mum­bai

Newspapers in English

Newspapers from India

© PressReader. All rights reserved.