ANALYSIS OF THE MOST COMMON CALCULATION METHODS “VALUE AT RISK, VAR”

Abstract

The Regulation of the National Bank of Ukraine on Risk Management in Ukrainian Banks stipulates that this Regulation obliges banks to use the VaR methodology to assess market risk. The article shows that VaR is the absolute maximum amount of losses of the market investment portfolio due to fluctuations in prices for financial instruments during a certain fixed period of time (VaR horizon) in normal market conditions at a given level of confidence level (confidence level). A study of the Value at Risk methodology as the absolute maximum size of market investment portfolio losses due to price fluctuations in financial instruments during a certain fixed period of time (VaR horizon) in normal market conditions for a given level of confidence level (confidence level). It was found that there are three main methods of VaR: parametric VaR, or delta normal VaR, historical method, or historical simulation, Monte Carlo, or simulation and other methods and models. The parametric VaR method, or the delta-normal VaR method, is based on the assumption of a normal distribution of a random variable. This means that the data are distributed according to the normal distribution law, ie you can calculate the mean and standard deviation. It is proved that these two indicators are the basis of VaR by the parametric method. It is shown that nonparametric VaR methods do not require the hypothesis of normal distribution. Therefore, the form of distribution is determined by empirical data, and percentiles are stored as empirical percentiles of the historical distribution of profitability. The article considers the nonparametric VaR method in comparison with the parametric one on a practical example. It is proved that in the domestic scientific literature there is little research on the practical application of the VaR method in finance and in particular in banking. Therefore, the practical aspects of applying different VaR models to the NBU exchange rate data are demonstrated and conclusions are drawn. It is shown that the practical application of methods is a modern tool for assessing market risks, but with the expansion of the database, the parameters of the models should be refined in combination with the economic method. This requires monitoring and back-testing.