<![CDATA[Cryptocurrencies offer high potential returns but expose investors to extreme price fluctuations, making the measurement of downside risk essential. This study analyzes the investment risk of five major cryptocurrencies (Bitcoin, Ethereum, BNB, Solana, and XRP) and an equally weighted portfolio using the Value at Risk (VaR) method. Daily data from 2 January 2020 to 30 June 2026 are used to estimate one day VaR at the 95% and 99% confidence levels with five approaches: historical simulation, the variance covariance (normal) method, the Cornish Fisher modified VaR, Monte Carlo simulation with Student t innovations, and a GARCH(1,1) model with Student t errors. Expected Shortfall is computed as a complementary measure. The accuracy of each approach is evaluated through out of sample backtesting over 1,642 days using the Kupiec unconditional coverage test and the Christoffersen conditional coverage test. The results show that cryptocurrency returns are strongly non normal, with excess kurtosis for all assets. At the 99% level, the one day VaR ranges from 5.87% to 11.93% under historical simulation, implying a potential daily loss of up to Rp11.928.115 on an investment of Rp100 million in Solana. The normal VaR underestimates extreme losses and is rejected at the 99% level for 3 of six series, whereas the GARCH t model passes all backtests at both confidence levels. The portfolio reduces VaR by about 28% compared with the average individual asset. The findings recommend that crypto investors and platforms use fat tailed, volatility sensitive VaR models rather than the normal approach.]]>
Copyrights © 2026
