درخت حوزه‌های تخصصی

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.

Purpose: This study examines the impact of product diversification on asset returns in petrochemical companies using a kernel density approach, addressing contradictory findings in the diversification-performance literature within this capital-intensive sector. Design/methodology/approach: Employing an ex-post facto research design, we analyzed data from petrochemical companies listed on the Tehran Stock Exchange from 2011 to 2021. The study utilized a two-stage analytical approach: first classifying companies into life cycle stages (growth, maturity, decline) using discriminant analysis, then testing hypotheses through multivariate regression and cross-sectional analysis. The kernel density estimation provided a non-parametric assessment of diversification effects. Results: The results demonstrate that product diversification significantly enhances asset returns in petrochemical firms, with the Herfindahl index revealing seven distinct market concentration levels. The positive effect of diversification was particularly pronounced in monopolistic market structures. Empirical evidence confirms that diversification serves as an effective strategy for improving financial performance and mitigating market dependence. Originality/value: This research contributes to the literature by introducing kernel density estimation to diversification analysis, providing a more nuanced understanding of return distributions beyond traditional parametric methods. The study also offers unique insights into the petrochemical sector's dynamics in emerging economies, highlighting how market structure moderates the diversification-performance relationship. Practical implications: The findings suggest that managers should prioritize diversification strategies, especially in monopolistic environments, to enhance asset utilization and financial resilience. Policymakers can use these insights to design industrial policies that encourage strategic diversification in response to Iran's economic conditions.

Climatic conditions are key determinants of agricultural productivity, affecting both production levels and food security. This study aims to decompose total factor productivity growth in Iran’s agricultural sector by explicitly incorporating climatic variables. True Random Parameters Stochastic Production Frontier (TRP-SPF) model and the Maximum Likelihood method were used to estimate parameters in two scenarios, with and without climate variables which are evaluated by  and  models, respectively. The results showed that the variables of cultivated area, tractors, and technology have a positive and significant impact on production in both models, and the highest effect observed for cultivated area. In the  model, the variables of temperature and precipitation also made significant contributions to higher production. The average technical efficiency in the  model was estimated at 56.5%, and in the model with climatic variables, it decreased to 43.3%. These results confirmed the significant impact of climatic variables on production and suggested that agricultural output could be increased by more than 50% with the same resources and technology, minimizing inefficiency factors and the adverse effects of climatic variables. Decomposition of the Climatic Adjusted Total Factor Productivity (CATFP) index into four components, including climatic effects, scale effects, technical efficiency, and technological progress, revealed that climatic effects have the greatest impact, accounting for 26.4%, while the other three components contributed almost equally to improving productivity. Given the significant impact of climatic variables on agricultural productivity, this study recommends strategies adapting better with climatic conditions of different regions.