Prediction-oriented PLS path modeling in microfinance research

By using a prediction-oriented PLS path model, this study predicts the financial performance (FP) of Microfinance Institutions (MFIs) by means of their social impact (SI) and portfolio quality (PQ). The empirical study uses a dataset, obtained from the Mix Market database, for the year 2012 with 563 MFIs worldwide. Firstly, the findings show that the higher the SI, the higher the FP in the microfinance sector. Secondly, the results suggest that the PQ increases substantially the ability of the model to predict the FP of MFIs, providing a full mediation in the relationship between SI and FP. Thirdly, in consequence, both SI and PQ are powerful and accurate predictors of the FP of MFIs since the average R2 is 58%, with a standard deviation of 0.18 (by using a 10-fold cross-validation procedure) and an SRMR of 0.05. The main contribution of the current study is to show that MFIs can continue serving the poorest people (achieving their SI) while obtaining high financial results via increasing the quality of their portfolio, since in the microfinance sector great poverty implies lower default risks.