Beyond HODL: Can You Trade with On-Chain Data
summary
Claim: on-chain data produces tradable signals. A tree-based model trained only on dozens of on-chain metrics runs a Bitcoin long-only strategy in out-of-sample and live paper trading, improves risk-adjusted returns and achieves a 50% drawdown reduction versus buy-and-hold. The most important features are profitability and flow metrics (MVRV, entities in profit, short-term realized P/L), and a two-metric “Goldilocks” rule reconstructed from the model alone outperforms Bitcoin on return and volatility.
AI-written summary · about
chapters (8)
- 0:00On-chain data trading intro
- 1:14Fundamental analysis with on-chain data
- 2:18Market on-chain data categories
- 4:02Proof of on-chain trading value
- 5:51Machine learning model results
- 7:18Actionable on-chain metrics
- 9:02Two-metric trading strategy
- 10:29Conclusion and research
related talks
28mBeyond the Balance Sheet: Transactional Data as a Window to Company Health
01mOn-Chain Data Reliability — Which Data Can You Trust?
27mStop Writing the Analysis Script: Let the Data Decide What to Look At
32mThe Trust Problem in Onchain Insights: What Data Can We Believe?
16mData Science in Asset Management
15mWhat's Happening On-Chain? Current Trends in On-Chain Data
29mDrilling for Data: Powering DeFi Applications
29mUnleashing the Power of Data-Driven Crypto Investing