Análise da demanda ao crédito imobiliária com recurso aos modelos de Box–Jenkins. Caso de estudo: Moza Banco da cidade de Maputo (2012 - 2020)

Abstract

The present work to conclude the Degree Course in Statistics, focused on credit and consump- tion, specifically real estate credit, where we described concepts on credit and the respective purpose. However, we use one of the technical concepts of statistics, where we conduct an analysis of the trend and forecast of housing credit in the city of Maputo. The graphical analy- sis of forecast with fluctuations (highs and lows) shows a high demand for the months of June of both years (2012 and 2020) and lows for the months of August and October respectively, con- cluding that the seasonality does not recur in this period. Similarly, as based on the graphical analysis of the series, which was confirmed by the unit root test, it was found that the series in the analysis is not stationary, therefore, the transformation was performed through its box-cox, which helped the removal of this variation. stationary. Once this assumption of stationarity was satisfied and with the help of the Box–Jenkins methodology, among 22 models created by the auto.arima() function, the SARIMA(1,0,2)(0,0,1)[12] model was selected as the 7th most suitable. verified still that, the MAPE of the model was 21.32%. Forecasting the demand for real estate loans from Moza Bank customers in the city of Maputo, for the next thirty-four months. The relevance of this monograph is due to the fact that it will assist bank managers in developing better strategies for credit management and minimizing the risks of investment losses. We can recommend an update to two bank managers on the use of credit management technology systems, as well as reducing credit fees that suffocate young people who intend to acquire a mortgage to start their new Home