Aliasghar Tehranipour; ebrahim abbasi; Barat Karimi; Firooz Yazaloo
Abstract
The purpose of the present research is to design a credit portfolio optimization model in the banking industry using a meta-heuristic algorithm. Risk is one of the basic concepts in financial markets, which has a certain complexity. Due to the lack of an accurate picture of risk realization, financial ...
Read More
The purpose of the present research is to design a credit portfolio optimization model in the banking industry using a meta-heuristic algorithm. Risk is one of the basic concepts in financial markets, which has a certain complexity. Due to the lack of an accurate picture of risk realization, financial markets need risk control and management approaches. The current study is descriptive in terms of data collection and developmental-applicative in terms of purpose. The statistical population of this research includes all the facility files over the last 10 years, as well as the financial statements of the branches of one of the commercial banks in Iran, which were selected through the census method. The risk criteria used in the models are the values at risk of the fuzzy average. Research models were implemented using Pareto Strength Evolutionary Algorithms (SPEA-II), Non-Globular Ranking Based Genetics (NSGA-II) and Multi-Objective Particle Swarm Optimization (MOPSO). The software used in the research is MATLAB. The results show that the NSGA-II algorithm has a better performance compared to the other two algorithms in terms of the quality metric, the diversity metric, and the spacing metric, both in small and large sizes. Also, the SPEA-II algorithm performs better than the other two algorithms in terms of the Mean Ideal Distance in both small and large scales, and the MOPSO algorithm in terms of time.
