Constructing Efficient Frontiers in Cryptocurrency Market Using Long-Run GARCH Volatility

Ginting, Josep (Department of Accounting, President University, Jawa Barat, Indonesia, 17550)
Staenly, Staenly (Department of Actuarial Science, President University, Jawa Barat, Indonesia, 17550)

Zero : Jurnal Sains, Matematika, dan Terapan
Vol 9, No 2 (2025): Zero: Jurnal Sains Matematika dan Terapan

Cryptocurrency portfolio optimization faces challenges from persistent volatility. This study aims to construct efficient frontiers using long-run GARCH (1,1) volatility estimates to improve portfolio stability in crypto markets. Daily prices of Bitcoin (BTC), Ethereum (ETH), and Solana (SOL) from January 2023 to December 2024 are analyzed. After confirming stationarity and modeling conditional volatility, GARCH-based risks are applied in portfolio optimization. Results show BTC is the least volatile (25.54%), SOL the most (43.56%), and ETH exhibits strong volatility persistence but a weaker risk-adjusted return. The maximum Sharpe ratio portfolio favors BTC and SOL (Sharpe ratio = 3.692), while the minimum variance portfolio concentrates on BTC. Compared with traditional variance-based methods, the GARCH approach produces more stable and realistic efficient frontiers. These findings suggest that volatility-aware modeling enables investors to design more resilient crypto portfolios under persistent structural risk.