One Rising Ship Sinks Other Ships: Cross-Chain Negative Spillovers in Crypto Markets

Battle Count: 68

Highly relevant for quantitative trading in crypto markets. The findings directly inform: (1) cross-chain portfolio diversification strategies exploiting negative correlations, (2) attention-based trading signals using extreme returns and chain activity metrics, (3) risk management for multi-chain portfolios, (4) understanding capital flow dynamics between competing blockchain ecosystems, and (5) timing strategies around token launches and protocol upgrades. The half-day frequency and factor model framework are directly implementable for systematic trading strategies. However, the paper is primarily academic and does not provide backtested trading strategies or execution details.

Moderate to high complexity. Requires: (1) On-chain data extraction from multiple blockchains via APIs and full nodes, (2) Market-cap weighted portfolio construction with multiple variants (All, CEX, non-CEX, Local), (3) ARIMA model fitting for extracting unexpected components, (4) GJR-GARCH estimation for volatility modeling, (5) Linear and non-linear factor model estimation with interaction terms, (6) Extreme return dummy construction, (7) Global market variable alignment to half-day UTC frequency. The econometric framework is standard but the data pipeline for comprehensive on-chain portfolio construction is non-trivial.

The paper uses publicly available data sources (Coingecko API, CoinDesk API, stakingrewards.com, marketwatch.com, euribor-rates.eu, investing.com, QuickNode full nodes, AAVE Protocol API). However, the exact portfolio construction code, ARIMA parameter selection process, and full data pipeline are not explicitly provided as a repository. The methodology is well-documented with equations, but reproducing the exact on-chain data extraction and portfolio weighting would require significant effort. No GitHub repository is mentioned.