Abstract
This paper introduces a novel large-scale dataset derived from blockchain data to measure and characterize the complexity of USD stablecoin activity. Focusing on USDC, USDT, and PYUSD, we develop a framework that links stablecoin transfers to a broader transactional context. The analysis provides new evidence on how stablecoins are used across decentralized finance protocols. The results shed light on composability and atomicity, showing that stablecoin transactions frequently embed multistep financial interactions rather than simple peer-to-peer payments. These findings contribute to the growing policy and academic debate by offering a more nuanced understanding of stablecoin usage and by highlighting the functional integration of stablecoins into broader digital financial market infrastructures.