Raden Aswin Rahadi
Sekolah Bisnis dan Manajemen, Institut Teknologi Bandung, Bandung, Indonesia

Published : 1 Documents Claim Missing Document
Claim Missing Document
Check
Articles

Found 1 Documents
Search

Optimal Cross-Asset-Class Portfolio Construction Using the Markowitz Mean-Variance Framework: Evidence from a U.S. Dollar-Based Foreign Investor Perspective Rafli Ahmad Fauzan; Subiakto Soekarno; Raden Aswin Rahadi
Journal Development Manecos Vol. 4 No. 2 (2026): Journal Development Manecos
Publisher : Scieclouds Publishing

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.71435/742327

Abstract

Foreign investors price Indonesian assets in U.S. dollars (USD), not rupiah, and that distinction carries weight. Between 2021 and 2024 the rupiah lost roughly 19% against the dollar while foreign capital left the equity market in waves, eroding dollar returns even when local prices held up. A portfolio built only from the Jakarta Composite Index (JCI) and the S&P 500 spreads risk across two markets yet guards against neither force. The question this paper pursues is whether gold and Bitcoin can close that gap. A four-asset portfolio of Bitcoin, gold, JCI, and the S&P 500 is assembled with all returns converted to USD, then optimised on monthly data from January 2021 to December 2024 through the Markowitz mean-variance model in Excel Solver. The Sharpe ratio serves as the objective, bound by three limits: weights sum to one, no short selling, and a 5% floor per asset. For the risk-free rate, the U.S. Treasury Bond and Indonesian Government Bond yields are blended by the optimal allocation. The resulting weights are then carried into an out-of-sample 2025 test across three management strategies. The optimal mix of 52% gold, 28% S&P 500, 15% Bitcoin, and 5% JCI reaches a Sharpe ratio of 0.812. The equity-only portfolio manages only 0.227, the S&P 500 alone 0.615, and the JCI, once measured in dollars, sinks to −0.360. Quarterly rebalancing earns the most in 2025, at 37.85%. The contribution is a dollar-based allocation template that moves past the rupiah-denominated evidence dominating earlier Indonesian studies.