Due to technological advancement, Peer-to-Peer (P2P) platforms have allowed significant cost reduction in lending. However, this improved allocation comes at the price of a higher credit risk. In this paper, the authors investigate the effectiveness of credit scoring models employed by P2P platforms with respect to loan default prediction. We claim that, because of differences in risk ownership with respect to traditional lenders, the rating grades obtained from P2P scoring models may not be the best predictors of loan default.