The relationship between Big Data and Credit Score: the Brazilian case
big data analytics; banks, credit score.
Society is generating and consuming large amounts of data, and companies that use Big Data Analytics must be aware of the challenges of this reality. The search for personal data allows private companies and public entities to use them on an unprecedented scale. In this context, this study proposes an analysis of how Credit Scoring systems are affected by Big Data and its impact on borrowers and financial institutions that operate with credit. The methodology of this study will be quantitative, using logistic regression and the chi-square independence test, and the econometric analysis will be carried out in the R software. As for the examination from the perspective of the financial institution, three models will be used to evaluate: of Credit Scoring; the costs/benefits of using Credit Scoring; and Credit Scoring, Big Data and Loan/Sell Ratio. With regard to borrowers, a model will be used to understand the relationship between debt and indebtedness. To optimize the evaluation of the hypotheses formulated by the models that will be proposed, the principle of parsimony will be applied, in general, the hypothesis should choose the simplest model if it can explain the event. It will be verified which proposed model will have greater statistical significance. Closed questionnaires will be published, using a Likert scale, when possible, in order to assess the respondents' level of indebtedness. The study will look to see whether the use of Big Data significantly affects credit scores.