Stalled Cross-chain Transactions: Scale, Impacts, and Causes

Abstract

Cross-chain bridges are critical for asset mobility across fragmented blockchain ecosystems and the backbone of multi-chain DeFi. However, cross-chain transaction asynchrony introduces an underexplored risk: transaction stalling. Unlike active threats (e.g., security breaches), stalled transactions trap user funds, undermine liquidity, and distort transparency. Despite these consequences, they remain largely overlooked in existing research. This paper presents the first systematic study on cross-chain transaction stalling, addressing three key questions: how widespread stalling is, how it impacts cross-chain ecosystems, and why it occurs. We build a dataset covering nine mainstream cross-chain bridges through transaction matching with unique identifiers, cross-verification via bridge explorers and on-chain logs, and a lifecycle-based tracing framework to identify root causes. Our findings show notable asymmetries: OP Stack-based bridges have stalling ratios up to 41.33% for L2-to-L1 withdrawals (due to user-dependent workflows), while L1-to-L2 flows show near-zero stalling. Millions of dollars in assets are stranded across protocols, eroding liquidity and distorting proof-of-reserve (PoR) mechanisms, because stranded assets are not excluded from reserve calculations, which inflates reported ratios. For example, Base Bridge's top 50 affected tokens see 0.2%–50% PoR distortion. Root causes include inadequate parameter validation, insufficient relayer incentives, and incomplete user actions. This work identifies cross-chain transaction stalling as a systemic risk, providing insights for improving cross-chain reliability.