Panic to Profits: Time Series Evidence Between GPRD and DeFi Token Prices

Abstract

This paper investigates the dynamic relationship between geopolitical uncertainty and decentralized finance (DeFi) token prices using a nonlinear, time-series-based framework. Leveraging the GPRD index as a proxy for global risk sentiment, the study examines seven prominent DeFi tokens representing diverse functional roles within the ecosystem. Through a layered empirical strategy - including Transfer Entropy, Mutual Information, Kernel-based Granger Causality, and Structural Time Series. Modeling - the analysis identifies both predictive and structural dependencies between GPRD and token valuations. The results reveal that tokens associated with financial-layer functions such as lending, collateralization, and liquidity rebalancing (e.g., Maker, Aave, BAL) exhibit stronger and more persistent exposure to geopolitical shocks than exchange-layer tokens like Uniswap or PancakeSwap. Kernel Granger causality confirms significant nonlinear predictive power of GPRD across all tokens, while structural decomposition shows that GPRD systematically depresses the long-term trend component of financial DeFi tokens. These findings indicate that global uncertainty operates not only through shortterm volatility, but also as a sustained driver of DeFi asset repricing. By combining information-theoretic and structural techniques, the study provides a comprehensive empirical lens through which to evaluate systemic risk transmission into DeFi markets. The results underscore the heterogeneous macro-financial sensitivity of decentralized protocols and suggest the need for differentiated risk assessment frameworks in crypto-asset research and governance.