bayesian models

The purpose of customer relationship management is to develop profitable customer relationships and increase company value. Based on this, this research identifies factors affecting competitive advantage. The current research is applied in terms of purpose and in terms of survey method with model development approach. The time frame of the research is five years (2016-2016). For this purpose, information on the indices of 33 factors affecting competitive advantage in Tejarat Bank were entered into Bayesian averaging models (BMA), (TVP-DMA) and (TVP -DMS) Based on the error rate, the BMA model had the highest accuracy. After estimating the model, 8 main variables were identified. which includes the long-term account balance; the amount of use of mobile bank; the amount of internet bank usage; real customers; legal clients; special or normality of the customer; The type of job and education. In the next step, customers are entered into the LRFM model and categorized based on these variables, finally, the effect of non-fragile variables in the Markov switching model on profitability was analyzed. The results indicated that the majority of the variables have a positive and significant effect on the level of profitability, and by moving from the high prosperity to the deep recession, the impact of the variables has increased. As a result, it can be said that profitability in the state of economic recession has a higher sensitivity to explanatory variables.

This research aims to develop a new approach to modeling systematic and unsystematic risks as well as geopolitical risks, in financial solvency within the insurance industry in Iran. The objective is to improve the accuracy of prediction models used in the industry..

The research follows developmental-practical approach and unilizes a descriptive-survey method. Data from 2011 to 2021, covering an 11-year period, were collected and analyzed. A total of 33 risk indicators affecting the financial solvency of insurance companies were examined using BMA, TVP-DMA, TVP-DMS, and BVAR models.

The BMA model demonstrated the highest accuracy based on error rate. Through the analysis, 11 main variables were identified as significant factors influencing financial solvency including economic growth, inflation uncertainty, exchange rate, sanctions, KOF index, return on working capital, cash adequacy ratio, total debt-to-equity ratio, loss factor, Herfindahl-Hirschman index, and geopolitical risk. The results The results highlight the complex nature of financial solvency prediction in the insurance industry, emphasizing the need for a comprehensive and systematic approach.

This study emphasizes the limitations of relying on a single conceptual model in financial solvency modeling and decision-making. The multiplicity of factors influencing financial solvency requires a systemic perspective in managing insurance companies. Additionally, it is important to consider a wide range of variables rather than relying on a specific model or set of variables to ensure a comprehensive understanding of financial solvency in the industry.

 Credit risk assessment is one of the most challenging issues in the banking sector. The main problem is the correct modeling of this type of risk. Variety of variables affecting credit risk and The lack of a specific model is one of the most important reasons for the failure of traditional models. Based on this, the aim of this research is to model the credit risk of Tejarat Bank in different economic regimes.

 The time period of the research is (2017-2021) and it has been selected from among 105 thousand customers of Tejarat Bank. information on indicators affecting credit risk was entered into BMA, TVP-DMA and TVP-DMS models. After estimating the model, 8 main variables were identified.

 the effect of non-fragile variables on the credit risk in the Markov switching model was analyzed. most of the variables, have a negative and significant effect on the level of credit risk, and by moving from high prosperity to deep recession, the impact of variables on credit risk has increased.

Originality / Value: 

The results showed that the optimal credit risk model is different in each bank. Based on this, an optimized model was designed Only for credit risk in Tejarat Bank based on Bayesian averaging patterns. Credit risk in the state of economic recession has a higher sensitivity to its explanatory variables.Also, based on the results, it was determined that indicators related to fintech have a significant effect on the level of credit risk in Tejarat Bank.