OPTIMAL ALLOCATION OF BANK RESOURCES AND RISK REDUCTION THROUGH PORTFOLIO DECENTRALIZATION

Arezoo Mohammadi; Mehrzad Minnoei; Zadollah Fathi; Mohamamd Ali Keramati; Hossein Baktiari

doi:10.52950/ES.2022.11.2.007

Authors

Arezoo Mohammadi IAU, Iran Author
Mehrzad Minnoei Department of Industrial Management, Central Tehran Branch, Islamic Azad University, Iran Author
Zadollah Fathi Department of Industrial Management, Central Tehran Branch, Islamic Azad University, Iran Author
Mohamamd Ali Keramati Department of Industrial Management, Central Tehran Branch, Islamic Azad University, Iran Author
Hossein Baktiari Department of Industrial Management, Faculty of Islamic Studies and Management, Imam Sadegh University, Tehran, Iran Author

DOI:

https://doi.org/10.52950/ES.2022.11.2.007

Keywords:

Risky and non-risky assets, New portfolio, Bank deposits, Risk , PSO, PSA

Abstract

The main concern of all economic companies is the allocation of resources across different economic sectors with the aim of maximizing profit and minimizing risk. Decentralization is one of the important factors that reduce investment risk. Investors plan their investments by carefully collecting sufficient information on the economic situation and analyzing the conditions of various industries. As economic enterprises, banks seek short- and long-term investments in types of loans such as bailment of capital, civil participation, reward, etc., which guarantee the return of their capital. In this paper, considering the condition of a bank as an economic enterprise, a model is presented which not only increases profit but also reduces risk. Two objective functions have been defined: the first objective is to minimize risk and the second objective function is to maximize the bank’s profit, which is achieved using robust programming and the Malvi Sim model. In this paper, we investigate the risky and non-risky portfolio and the optimal portfolio of bank assets based on a scenario-based solution of the model using PSO and the Genetic Optimization Algorithm. At all confidence levels, the optimal values of risk based on the estimation of the SPP-CVAR method obtained by the Particle Swarm Algorithm (PSA) are lower than those obtained by the Genetic Algorithm (GA), which indicates better performance of PSA compared to GA. Also, the optimum wealth obtained from the PSA solution is higher at all confidence levels than the corresponding value of the Genetic Algorithm (GA), which further confirms the superior performance of PSO compared to GA. The values of the first objective function obtained from the PSO algorithm for all confidence levels are lower than those of the genetic algorithm. The optimum wealth obtained from PSA is higher than that obtained from the genetic algorithm. At the 0.9 level, the value of the LR Kupiec statistic for the SPP-CVAR method was lower than the chi-square critical value, which indicates that the model is acceptable.

Data:
Received: 23 Aug 2022
Revised: 8 Oct 2022
Accepted: 10 Nov 2022
Published: 24 Nov 2022