COMPARISON OF THE “VALUE AT RISK” WITH EACH UNDER DIFFERENT RISK SCENARIOS

  • นันทรัตน์ ตรีพรชัยศักดิ์
  • นงนภัส แก้วพลอย
  • กนกพรรณ แก้วเนตร
Keywords: VaR

Abstract

This independent study observes the portfolio of financial assets by examining under Markowitz’s model which is an investor would choose to maximize his return for a given level of risk. In this study, the “Value at Risk” will be tested by three different methods, historical simulation, Variance-Covariance Method, and Monte Carlo simulation, at the confident interval of 90%, 95% and 99%. For this case study, the proportion of the investment will break down into four different risky assets which are the SET50 index, the S&P500 index, 10-year government bonds, and gold purity of 96.5%.                   The analyst will be divided into 5 case studies: 70% in SET50 and 10% in each of the other assets, 70% in S&P500 and 10% in each of the other assets, 70% in 10-year government bonds and 10% in each of the other assets, 70% in pure gold of 96.5% and 10% in each of the other assets, and investment under Markowitz’s model. The data in these studies will be daily collected from January 2005 to December 2015 for 2,608 days.              

The results have shown that, by using confident interval at 90%, 95%, 99% in the 3 VaR models with all 5 different investment scenarios, the investment under Markowitz ’s model displays the lowest  VaR when compare to the rest of other investment models. Furthermore, VaR at significance level of 90% under historical simulation has shown better result than any other methods across all case studies. At 95% significance level, all three methods provide similar results in all case studies while at 99% significance level the historical simulation has shown the lower result than other methods. As a result, these case studies do not agree with the assumption that the VaR of all methods yields the same result at every significant level. In addition, the difference in VaR ​​for all case studies is not significantly different to each other. Therefore, the result is inconsistent with the assumption that the low risk portfolios will have a low VaR value.

Published
2017-09-17
Section
Engineering and Technology Articles

Most read articles by the same author(s)

<< < 1 2