Exposing Stealthy Wash Trading on Automated Market Maker Exchanges
Why this work is in the frame
A frame that forgets how it found something cannot be audited. These are the routes that admitted this work.
Bibliographic record
Abstract
Decentralized Finance (DeFi), a pivotal component of the emerging Web3 landscape, is gaining popularity but remains vulnerable to market manipulations, such as wash trading. Wash trading is an illegal practice, where traders buy and sell assets to themselves within cryptocurrency exchanges to artificially inflate trading volumes and distort market perceptions. However, current research primarily focuses on traditional exchanges based on the Order-book mechanism (similar to stock markets), while ignoring the Automated Market Maker (AMM) exchanges, which dominate over 75% of the market and represent a significant innovation within the DeFi. This study utilizes entity recognition technology to detect wash trading on AMM exchanges within Ethereum-like systems, based on the understanding that colluding addresses (perceived as the same entity) must use ETH for transaction fees and exhibit direct or indirect ETH transfer links. We identify wash trading when addresses with transfer connections almost simultaneously buy and sell assets while their total asset holdings remain nearly constant. This comprehensive blockchain network analysis, compared to focusing solely on transactions within exchanges, unveils covert wash trading activities. Our detection method achieves a 95.9% recall and a 96.7% true negative rate in identifying pools affected by wash trading, demonstrating its superiority over existing methods. Furthermore, we apply our method to 98,945 pools from Uniswap V2 & V3 (the most popular AMM exchanges on Ethereum) and identify 1,070,626 abnormal transactions, totaling $27.51 billion in trading volume. Analysis of these transactions uncovers insights into wash traders’ behaviors, including the utilization of multiple addresses and the dual roles of certain addresses as wash traders and liquidity providers. These insights are crucial for developing more effective strategies to combat fraudulent activities in the DeFi ecosystem and enhance financial scrutiny.
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
Full frame distilled prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. Learned from the 10,348 direct Codex labels and 10,348 direct Gemma labels. Candidate is the union of thresholded teacher heads; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels or direct frontier model labels.
Codex and Gemma teacher scores by category
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.000 | 0.000 |
| Meta-epidemiology (narrow) | 0.000 | 0.000 |
| Meta-epidemiology (broad) | 0.000 | 0.000 |
| Bibliometrics | 0.001 | 0.002 |
| Science and technology studies | 0.000 | 0.000 |
| Scholarly communication | 0.000 | 0.000 |
| Open science | 0.002 | 0.000 |
| Research integrity | 0.001 | 0.001 |
| Insufficient payload (model declined to judge) | 0.000 | 0.000 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from it