ESTIMASI RISIKO PASAR PADA DATA RETURN KURS HARIAN DENGAN VALUE AT RISK MENGGUNAKAN MODEL–MODEL VOLATILITAS GARCH

Abstract

Volatility is a critical tool in analyzing financial data. GARCH models are common tools widely use to forecast financial data volatility. This research discusses the measurement of market risk using the Value at Risk volatility model GARCH, IGARCH, and EGARCH for the USD major currency pairs which consist of seven currency pairs with the observation period July 2016 – September 2022. The results of the analysis show that the calculation of returns does not comply with the normal distribution, so the estimation of losses using the normal distribution VaR can be biased, instead this research used student’s t distribution error. The test results show that the IGARCH volatility model at a confidence level of 99% and 95% proves to be valid after the Kupiec test and Conditional Coverage Test are carried out on all tested currencies. Meanwhile, risk measurement using the EGARCH model is invalid on the EUR/USD, USD/CHF and AUD/USD exchange rates. Besides that, risk measurement using the SGARCH model is invalid on the EUR/USD and AUD/USD exchange rates.