Credit Scoring with Social Network Data

摘要: Motivated by the growing practice of using social network data in credit scoring, this study analyzes impact based measures on customer score accuracy and tie formation among customers. We develop a series models to compare scores obtained with without data. also investigate how changes when individuals can strategically construct their networks attain higher scores. find that, if are motivated improve scores, they may form fewer ties focus them more similar partners. The such endogenous consumer is ambiguous. Scores become accurate as result modi fications networks, but improvement come greater fragmentation. threat exclusion endogenously formed provides incentives low type members exert e ffort that improves everyone's creditworthiness. discuss implications for both managers public policy.