Iranian Journal of Finance

Abstract In today's complex economic landscape, accurately predicting events such as customer loan defaults presents a significant challenge for financial institutions. Traditional methods have shown limitations in accuracy, prompting the adoption of data-driven machine learning techniques for enhanced predictive capabilities. This study investigates the efficacy of novel machine-learning algorithms compared with linear models for predicting loan defaults at a major commercial bank. Data from over six thousand customer loan files spanning 2019 to 2022 were collected, cleaned, and clustered based on key loan indicators. The accuracy of predicting loan defaults was first evaluated using popular machine learning classification models, including LightGBM, XGBoost, Multilayer Perceptron, and Logistic Regression, and XGBoost performed best. After that, prediction accuracy was evaluated using various time-series machine learning algorithms, with a particular focus on a combined Gradient Boosting and Long Short-Term Memory (LSTM) approach. Results indicate that the combined algorithm outperforms traditional linear models, showing a substantial 40% improvement over the ARIMA algorithm in predicting loan default behavior. This study underscores the potential of advanced machine learning techniques to enhance predictive accuracy in the banking sector, offering valuable insights for risk assessment and financial decision-making.

Abstract Market participants use different tools basically technical or fundamental analysis to have a higher return in constructing a well-maintained portfolio. Examining the efficiency of technical strategies in creating a portfolio is the main objective of this study. Technical analysis is based on using historical trading data to launch selling and buying rules that maximize return and still control risks of loss. We use the adjusted trading data of 50 active stocks in the Tehran Stock Exchange as our sample which includes daily trading data from 2008 to 2019. We construct two types of portfolio; strategy-based portfolio versus random one. Then we calculate abnormal returns of each type of portfolio, applying the Monte-Carlo technique. Using Independent-Samples T-Test to compare means of the abnormal returns, our findings show that there is a significant positive abnormal return for both strategies applied in constructing a portfolio (0.057 and 0.062 mean difference for the first and second strategy, respectively), confirming the higher efficiency of applying technical strategies in portfolio management. Therefore, it is suggested to have and apply a strategy or combination of strategies for trading as an active participant, instead of constructing, rebalancing and maintaining one’s portfolio only by chance, since there will be undesirable results in the long-run.