Measuring Market Risk in EU New Member States (CROSBI ID 532826)
Prilog sa skupa u zborniku | izvorni znanstveni rad | međunarodna recenzija
Podaci o odgovornosti
Žiković Saša
engleski
Measuring Market Risk in EU New Member States
The author in this paper examines different ways of calculating VaR in transitional economies of EU new member states. Majority of the EU new member states are all exposed to very similar processes of strong inflow of foreign direct and portfolio investments, and offer possibilities of huge profits for investors. These countries represent a very interesting opportunity for foreign and domestic banks, investment funds, pension funds, insurance companies and other investors. Banks and investment funds when investing in these financial markets employ the same risk measurement models for measuring market risk and forming of provision as they do in the developed markets. This means that risk managers in banks operating in EU new member states de facto presume similar or even equal characteristics and behaviour in these markets, as they would expect in developed markets. Using the VaR models, that are created and suited for developed and liquid markets, in developing markets raises a serious dilemma: Do the VaR models developed and tested in the developed and liquid financial markets apply to the volatile and shallow financial markets of EU new member states? Do the commonly used VaR models adequately capture market risk of these markets or are they only giving a false sense of security? In this paper the author also develops a new semi parametric VaR model that combines ARMA-GARCH volatility forecasting with bootstrapping, which should be more appropriate for turbulent transitional capital markets. Ten VaR models are tested on ten stock indexes from EU new member states. Performance of analysed VaR models is tested by Kupiec test, Christoffersen unconditional coverage test, Christoffersen independence test and Christoffersen conditional coverage test. To determine the models that are conditionally superior to the other tested models the following statistics are used: Lopez test, Blanco-Ihle test, Root Mean Squared Error (RMSE) and Mean Absolute Percentage Error (MAPE). The obtained results show that VaR models based on ARMA-GARCH volatility forecasts are superior to other tested types of VaR models. The findings show that common VaR models that are widely used in mature markets, such as historical simulation, variance-covariance model and RiskMetrics system are not well suited to transitional capital markets.
Market risk; VaR; GARCH; Bootstrapping; EU New Member States
nije evidentirano
nije evidentirano
nije evidentirano
nije evidentirano
nije evidentirano
nije evidentirano
Podaci o prilogu
1-39-x.
2007.
objavljeno
Podaci o matičnoj publikaciji
13th Dubrovnik Economic Conference
Podaci o skupu
13th Dubrovnik Economic Conference
pozvano predavanje
27.06.2007-30.06.2007
Dubrovnik, Hrvatska