DeXposure-FM: A Time-series, Graph Foundation Model for Credit Exposures and Stability on Decentralized Financial Networks
Researchers introduce DeXposure-FM, the first foundation model for time-series and graph analysis in decentralized finance (DeFi). Using a dataset of over 43 million points from 4,300 protocols across 602 blockchains, the model predicts credit exposures and systemic risk. DeXposure-FM’s novel encoder and multi-task learning enable early warnings and stress tests. While it relies on Total Value Locked (TVL) and excludes off-chain data, its open-source release marks a major advance in DeFi risk monitoring.
What it examines
This paper introduces DeXposure-FM, a time-series graph foundation model designed to measure and forecast credit exposures in decentralized finance (DeFi) networks. Using large-scale on-chain data, it aims to improve risk monitoring, stress testing, and understanding of systemic risk in the rapidly evolving DeFi ecosystem.
What it concludes
DeXposure-FM outperforms existing models in predicting DeFi credit exposures and systemic risk. Its tools support macroprudential monitoring, early warning signals, and stress testing. Applications include financial stability oversight and policy analysis. Limitations include data coverage and temporal resolution; future work will expand data sources and model capabilities.
Evidence objects
DeXposure-FM debuts as the first time-series, graph foundation model for DeFi, analyzing over 43 million data points from 4,300+ protocols across 602 blockchains to forecast credit exposures and systemic risk.
key_findings bullet 1 · key_findings · validation V0
The models novel graph-tabular encoder and multi-task learning outperform state-of-the-art methods, enabling early warnings, systemic importance rankings, and stress-test scenarios for network-wide contagion in volatile DeFi ecosystems.
key_findings bullet 2 · key_findings · validation V0
Despite its groundbreaking scope, DeXposure-FM relies on Total Value Locked (TVL) as a risk proxy and excludes off-chain or centralized exchange data, but its open-source release promises ongoing updates and broad impact.
key_findings bullet 3 · key_findings · validation V0
DeXposure-FM pioneers a time-series, graph foundation model for forecasting inter-protocol credit exposures in DeFi, leveraging a vast, unique dataset ($43.7$ million entries, $4,300+$ protocols, $602$ blockchains). Its advanced graph-tabular encoders, open-source release, and superior performance establish a novel, impactful benchmark for DeFi risk analytics and macroprudential tools.
key_findings bullet 4 · key_findings · validation V0
Raw abstract and provenance
Abstract: Credit exposure in Decentralized Finance (DeFi) is often implicit and token-mediated, creating a dense web of inter-protocol dependencies. Thus, a shock to one token may result in significant and uncontrolled contagion effects. As the DeFi ecosystem becomes increasingly linked with traditional financial infrastructure through instruments, such as stablecoins, the risk posed by this dynamic demands… ▽ More Credit exposure in Decentralized Finance (DeFi) is often implicit and token-mediated, creating a dense web of inter-protocol dependencies. Thus, a shock to one token may result in significant and uncontrolled contagion effects. As the DeFi ecosystem becomes increasingly linked with traditional financial infrastructure through instruments, such as stablecoins, the risk posed by this dynamic demands more powerful quantification tools. We introduce DeXposure-FM, the first time-series, graph foundation model for measuring and forecasting inter-protocol credit exposure on DeFi networks, to the best of our knowledge. Employing a graph-tabular encoder, with pre-trained weight initialization, and multiple task-specific heads, DeXposure-FM is trained on the DeXposure dataset that has 43.7 million data entries, across 4,300+ protocols on 602 blockchains, covering 24,300+ unique tokens. The training is operationalized for credit-exposure forecasting, predicting the joint dynamics of (1) protocol-level flows, and (2) the topology and weights of credit-exposure links. The DeXposure-FM is empirically validated on two machine learning benchmarks; it consistently outperforms the state-of-the-art approaches, including a graph foundation model and temporal graph neural networks. DeXposure-FM further produces financial economics tools that support macroprudential monitoring and scenario-based DeFi stress testing, by enabling protocol-level systemic-importance scores, sector-level spillover and concentration measures via a forecast-then-measure pipeline. Empirical verification fully supports our financial economics tools. The model and code have been publicly available. Model: https://huggingface.co/EVIEHub/DeXposure-FM. Code: https://github.com/EVIEHub/DeXposure-FM. △ Less
Source row: 599 · abstract type: unknown