Understanding DAO Governance Token Impact on Crypto Portfolios: A Machine Learning Forecasting and Forecast-Based Optimization Approach.

Indrijo Putra, Aqsha Naufaldy (2025) Understanding DAO Governance Token Impact on Crypto Portfolios: A Machine Learning Forecasting and Forecast-Based Optimization Approach. Other thesis, Institut Teknologi Sepuluh Nopember.

Text 5031211090-Undergraduate_Thesis.pdf - Accepted Version Restricted to Repository staff only Download (2MB) | Request a copy

Abstract

Dalam studi ini, kami menilai nilai inkremental dari integrasi token tata kelola Decentralized Autonomous Organization (DAO) ke dalam portofolio mata uang kripto dengan menggabungkan peramalan Gated Recurrent Unit (GRU) spesifik aset dengan kerangka kerja optimasi rasio Sharpe Monte Carlo–Dirichlet. Kami memanfaatkan data OHLCV per jam dan data indikator teknis dari Januari 2018 hingga Desember 2024 untuk lima koin Layer-1 (BTC, ETH, BNB, SOL, AVAX) dan lima token DAO (LDO, UNI, AAVE, MKR, SNX) untuk membangun indeks DAO turunan PCA dan melatih model GRU untuk prediksi harga satu jam ke depan. Prakiraan ini menginformasikan alokasi portofolio dinamis yang memaksimalkan rasio Sharpe yang diharapkan dan dibandingkan dengan pendekatan mean-variance historis tradisional. Analisis empiris kami menunjukkan bahwa, meskipun strategi yang digerakkan oleh prakiraan menghasilkan rasio Sharpe rata-rata yang lebih tinggi dan alokasi yang lebih gesit, strategi ini juga menghasilkan rasio Sharpe yang lebih besar relatif terhadap tolok ukur historis. Analisis korelasi yang memperhatikan rezim lebih lanjut mengungkapkan bahwa eksposur DAO memberikan properti safe haven yang sesungguhnya selama pasar bearish, memperkuat diversifikasi di bawah korelasi antaraset yang rendah, dan menawarkan keuntungan taktis dalam rezim netral melalui sinyal beta bersyarat. Temuan ini memperluas Teori Portofolio Modern, optimasi yang sensitif terhadap drawdown, dan CAPM bersyarat dengan menanamkan dinamika treasury yang digerakkan oleh tata kelola ke dalam konstruksi portofolio. Bagi para praktisi, kami merekomendasikan aturan pembobotan yang selaras dengan rezim, penyaringan token berbasis tata kelola dan treasury, serta protokol penyeimbangan ulang yang memperhatikan biaya transaksi untuk secara efektif mengintegrasikan token DAO ke dalam sistem alokasi algoritmik.
======================================================================================================================================
In this study, we assess the incremental value of integrating Decentralized Autonomous Organization (DAO) governance tokens into cryptocurrency portfolios by merging asset-specific Gated Recurrent Unit (GRU) forecasting with a Monte Carlo–Dirichlet Sharpe-ratio optimization framework. We leverage hourly OHLCV and technical-indicator data from January 2018 to December 2024 for five Layer-1 coins (BTC, ETH, BNB, SOL, AVAX) and five DAO tokens (LDO, UNI, AAVE, MKR, SNX) to construct a PCA-derived DAO index and train GRU models for one-hour-ahead price predictions. These forecasts inform dynamic portfolio allocations that maximize expected Sharpe ratio and are benchmarked against a traditional historical mean-variance approach. Our empirical analysis demonstrates that, while the forecast-driven strategy yields higher average Sharpe ratios and more agile allocations, it also incurs greater Sharpe ratio relative to the historical benchmark. A regime-aware correlation analysis further reveals that DAO exposure provides genuine safe-haven properties during bear markets, strengthens diversification under low interasset correlation, and offers tactical upside in neutral regimes through conditional beta signals. These findings extend Modern Portfolio Theory, drawdown-sensitive optimization, and conditional CAPM by embedding governance-driven treasury dynamics into portfolio construction. For practitioners, we recommend regime-aligned weighting rules, governance- and treasury-based token screening, and transaction-cost–aware rebalancing protocols to effectively incorporate DAO tokens within algorithmic allocation systems.

Item Type:	Thesis (Other)
Uncontrolled Keywords:	Crypto Portfolio Optimization, DAO Governance Tokens, GRU Forecasting, Sharpe-Ratio Maximization, Monte Carlo Dirichlet Sampling. Optimasi Portofolio Kripto, Token Tata Kelola DAO, Peramalan GRU, Maksimalisasi Rasio Sharpe, Pengambilan Sampel Dirichlet Monte Carlo.
Subjects:	H Social Sciences > HG Finance > HG4529.5 Portfolio management
Divisions:	Faculty of Creative Design and Digital Business (CREABIZ) > Business Management > 61205-(S1) Undergraduate Thesis
Depositing User:	Aqsha Naufaldy Indrijo Putra
Date Deposited:	28 Jul 2025 09:44
Last Modified:	28 Jul 2025 09:44
URI:	http://repository.its.ac.id/id/eprint/122295

Actions (login required)

View Item