Measuring the capital at risk in the portfolio investment under the extreme scenario plays a vital role to enable the traders to foreseen the potential maximum capital loss at a particular time frame. The Value-at-Risk (VaR) is extensively adopted by the numerous financial institutions, investors and creditors as a risk assessment method to measure the maximum capital loss to an investment portfolio or risky assets over a period of time under the provided confidence interval. Soon after, it was introduced by J. P. Morgan in 1998 in their RiskMetrics which purposefully aims to publish the volatility and correlation information for stocks listed on the major markets in the world (Kaura, n.d.). Pafka & Kondor (2001) argued that the popular RiskMatrics is the artifice of the choice of risk assessment. Provided that the exceptional performance of volatility estimates is because of the short forecasting horizon and the satisfactory performance in obtaining the VaR is because of the choice of the confidence level.
The VaR method, however, is comprehensively conducted to determine the exposure of the capital to the potential market risks which is the extensively used by the creditors, such as, commercial and investment banks to study about the exposure of their portfolio investment to risk over a particular time to ensure that their capital and cash reserve can cover the value-at-risk without putting the firms at the financial distress. (Kaura, n.d.), Koch (2006), Goorbergh & Vlaar (1999), Shirazi (n.d.), Gregory & Reeves (2008), Hong, Hu, & Liu (2014), Linsmeier & Pearson (2000), Borgdan, Baresa, & Ivanovic (2015), Jorion (2007) and Wong, Cheng, & Wong (2003) have all comprehensively employed the VaR method in studying the exposure of the market risks on the portfolio investment. The Value at Risk can be computed by 3 methods, namely, the historical analysis, the parametric VaR and the Monte Carlo Simulation with each method offers certain pros and cons.
The historical analysis, first and foremost, adopts the historical data from the market ratio or prices to empirically analyse the value at risk. Considered as the easiest method to measure the value at risk, this nonparametric method uses essentially the empirical distribution of portfolio returns, and is not required to fulfil any distributional assumptions (Goorbergh & Vlaar, 1999). The realistic historical information of the past event enables the researchers to accurately predict the possible future event (Kuara, n.d.). The readily available data also adds more simplicity to the method. For example, the historical trading data, such as, securities, is publicly available (Borgdan, Baresa, & Ivanovic, 2015). Only predetermining the time horizon of the data is required, and no mapping is required in comparison with the parametric method. On the contrary, the major drawback of this method is if the composition of the portfolio investment changes over time, collecting large sample size is unmanageable. Therefore, making this method becomes less feasible (Koch, 2006). The historical simulation approach using the historical asset returns data, however, is applicable to dealt with this problem. Yet, intensive computation is required for the large portfolio investment (Kuara, n.d.).
Save your time!
We can take care of your essay
- Proper editing and formatting
- Free revision, title page, and bibliography
- Flexible prices and money-back guarantee
The parametric VaR method, which is also called by other names, including variance-covariance, and linear or delta normal VaR, is another popular method to measure the value-at-risk. According to Lausbch (1999), the parametric method also uses the historical data to measure the potential risk. Unlike the previous method, this method does not require long historical data which allows this method to be quickly and easily calculated. The mean value of the yield rate and the standard deviation of the same data are the two major variables used by the parametric method in the calculation. The primary requirement of the parametric method, however, is the data has to be normal distribution (Value-at-Risk, n.d.). Meaning that the mean value, arithmetic mean, mode and median are the same size and it has a bell shape. Lausbch (1999), on the other hand, stated that the hypothesis of the normal distribution is main disadvantage of the parametric model which makes it less feasible for the nonlinear portfolios and distorted distribution. Jackson, Maude & Perraudin (1997) which VaR was applied on the trading book of an anonymous bank, have concluded that the simulation approach provides more accurate measures of tail probabilities comparing to the parametric VaR. This can happend due to the arise of a serious non-normality of financial return. Lausbch also anticipated that the major limitation of the parametric model is the constancy of the computed standard deviation and correlation coefficients, in which value changes throughout the time. Hence, if the researchers fail to modify the computation due to the extreme values of VaR, it will result in the misinterpretation of the results.
Last but not least, Monte Carlo Simulation is last method for forecasting VaR. The Monte Carlo, basically, is a justify name for the stochastic method for computing VaR. Due to the fact that the method involves the computer simulation of various influences on the observed portfolio of securities (Borgdan, Baresa, & Ivanovic, 2015). Similar to the historical method, this method involves complex computation of the historical data to predict the future risk and potential loss with a statistical confidence interval. The complex computation which involves hundreds or thousands of possible scenarios and generates the feasible solution makes this method to be the most reliable method to compute VaR (Borgdan, Baresa, & Ivanovic, 2015). This method, additionally, can be employed to calculate both the value of stochastic and non-stochastic. Vose (1997) indicated that Monte Carlo is the mathematical risk analysis techniques which describe the impact of risk and uncertainty on the problem. The uncertain parameters in the model are characterised by distribution of probabilities. While that shape and size of these distribution describes range of values that parameters can have with their relative probabilities. Ostojić, Pokorni, Rakonjac, & Brkićm (2012) and Lausbch (1999) agreed that a major advantage for Monte Carlo method would be its effectiveness to accurately calculate the risk value of various financial instruments, yet this method does not necessarily require large historical data. Significantly, the Monte Carlo method support the use of different distribution, including T-distribution, normal and similar. While the major drawbacks for this method are the requirement for complex analysis and really time consuming. Finally, selecting the proper distribution is also vital to quantify the risk of thickened tail distribution.
VaR, in conclusion, is the maximum potential loss to a portfolio investment at a particular period of time. This risk assessment method is very handy for the investors and creditors to estimate the potential loss due to its applicability and simplicity, and the model itself has passed numerous modifications which aim to improve the precision to forecast the value-at-risk. Hendricks (1996) applied the VaR on 1,000 randomly selected foreign exchange portfolios from 1983-94. The study suggested that among the twelve approaches which was applied. None is perceived to have more superiority over the others. The choice on the confidence level, however, appears to have significant influence on the performance of VaR. Borgdan, Baresa, & Ivanovic, (2015), on the other hand, claimed that besides the many advantages that this model contains, the model should be applied with some precautions, for example, the model focus mainly on the portfolio losses but cannot entirely forecast the future losses. Most importantly, the dramatic price fluctuations can probably influence the computed value-at-risk and generate false security, such as, undervalued or overvalued risk. Hence, the VaR model has the best applicability in the stable market conditions. In this paper, the VaR model will be applied to study the value-at-risk of a portfolio investment in The Cambodia Securities Exchange Market (CSX).
Stuck on your essay?
