The impact of diversification on risk reduction: using a mix of Merton model and random matrix approach to take into account non-stationary

Subject Areas : Operations Management

Zahra Eskandari ¹ , Mirfeiz Fallah Shams ² , Gholamreza Zomorodian ³

1 - Department of Management, Central Tehran Branch, Islamic Azad University, Tehran, Iran
2 - Department of Management, Central Tehran Branch, Islamic Azad University, Tehran, Iran
3 - Department of Management, Central Tehran Branch, Islamic Azad University, Tehran, Iran

Received: 2022-11-25 Accepted : 2023-01-24 Published : 2022-09-01

Keywords: Value at Risk, Non-stationary, Loss Distribution, Random Matrix,

Abstract :

Portfolio diversification has been discussed in many researches and has been proven in different portfolios, including stock portfolios, currency portfolios, etc. In this paper, we are going to investigate the impact of diversification on a credit portfolio risk related to companies listed on the Tehran Stock Exchange. To calculate the risk of the mentioned companies, we use structural Models of credit risk. In fact, the most important factor for assessing the risks of financial markets is the estimation of loss distribution. On the other hand, the estimation of loss distribution is highly dependent on the characteristics of the distribution parameters. One of the characteristics that can affect the loss distribution is non-stationary time series of asset returns. In this research, the data of the adjusted prices of the companies in the Tehran Stock Exchange for the period 2011 to 2019 is used. The loss distribution of credit portfolio is obtained through Merton's model with regard to non-stationary time series of asset returns and the changes of the asset returns’ covariance matrix during the period of 2011-2019. The risk measure used in this paper is the value at risk. According to the results of the model, at lower confidence levels such as 99% and 99.5%, there is not enough evidence for the impact of diversification in reducing risk, but at the confidence level of 99.9% and for the type error of 5%, it can be said that diversification has a significant effect on reducing risk.