ABSTRACT

Value-at-Risk (VaR) is a very famous and popular model which has been widely used to measure the potential exposure of the value of loss of an underlying asset or an investment portfolio at a certain confidence level and holding period. The main objective of this paper is the implement all of the three approaches applicable to estimate VaR namely non-parametric, parametric, and Monte-Carlo simulation VaR on the synthetic investment portfolio which consists of HKL’s bond and five stocks listing and trading in CSX besides the securities the portfolio also includes the FX and commodity, such as, gold and crude oil. At the position date the initial market value of this portfolio is KHR 591,514,539. With the confidence level of 95% and the holding period of 1 day VaR is KHR 6,198,453, KHR 5,523,467 and KHR 5,354,189 estimated by the non-parametric, parametric and Monte-Carlo simulation respectively. This research also indicates that the non-parametric VaR is very simple to implement; therefore, this approach is highly recommended for the investors who intention is the estimate the risk exposure of the value of the assets or portfolio. On the other, the parametric and Monte[1]Carlo simulation approaches, which is perceivably more difficult than the non-parametric, are highly recommended for the study which intention is to seek high accuracy.

Keywords: VaR, CSX, Monte-Carlo Simulation, investment portfolio.