{"paper_id":"2406.16573","title":"An Improved Algorithm to Identify More Arbitrage Opportunities on Decentralized Exchanges","authors":["Yu Zhang","Tao Yan","Jianhong Lin","Benjamin Kraner","Claudio Tessone"],"paper_date":"2024-06-24","paper_type":"methodological","paper_url":"https://arxiv.org/pdf/2406.16573v1","page_url":"https://thequant.space/flowcharts/an-improved-algorithm-to-identify-more-arbitrage-opportuniti/","scores":{"empirical_rigor":6,"math_complexity":7.5,"quadrant":"Holy Grail"},"proposes":{"status":"available","statement":"The paper proposes a new arbitrage detection method that combines a line‑graph transformation with a modified Moore‑Bellman‑Ford algorithm (MMBF) to find more arbitrage loops and non‑loop opportunities on DEXs.","quote":"In this paper, we develop a new method to solve these problems by combining the line graph and a modified Moore‑Bellman‑Ford algorithm (MMBF). This method can help to find more arbitrage loops by detecting at least one arbitrage loop starting from any specified tokens in the DEXs and can detect the non‑loop arbitrage p","summary":"In decentralized exchanges (DEXs), the arbitrage paths exist abundantly in the form of both arbitrage loops (e.g. the arbitrage path starts from token A and back to token A again in the end, A, B,…, A) and non-loops (e.g. the arbitrage path starts from token A and stops at a different token N, A…","strategy":"modified Moore-Bellman-Ford algorithm (MMBF) on line graph"},"assumptions":{"status":"available","items":[{"assumption":"Gas fees are ignored in the profit calculations.","type":"execution","stated":"explicit","quote":"Another limitation of this paper is that the gas fee is not considered. The gas fee is a significant factor that deserves to be considered in on‑chain transactions."},{"assumption":"The token graph is constructed from liquidity pool reserves and token addresses as provided by The Graph.","type":"data","stated":"explicit","quote":"which provides detailed information for each token of the liquidity pools, such as tokens’ addresses, names, symbols, reserves, and decimals."},{"assumption":"Only the largest four hundred liquidity pools (about 90% of TVL) are used, representing roughly one hundred tokens.","type":"data","stated":"explicit","quote":"largest four hundred liquidity pools in TVL take about 90% of the total TVL and correspond to about one hundred different tokens. This is why we chose one hundred tokens and about four hundred liquidity pools in our research."},{"assumption":"Arbitrage profit is computed on daily snapshot data, assuming static reserves within each snapshot.","type":"model","stated":"implied","quote":"To identify potential arbitrage opportunities with time, we retrieve historical daily snapshots from The Graph, which encompass the daily trading volume in USD and reserves of each token."},{"assumption":"The bisection method can efficiently find the optimal input amount for maximal arbitrage profit.","type":"model","stated":"explicit","quote":"it is very easy and efficient to use the bisection method to calculate the optimal input of the starting token to get maximal arbitrage profit."},{"assumption":"Negative cycles may exist in the token graph, but the MBF combined algorithm cannot handle non‑loop paths when they do.","type":"model","stated":"explicit","quote":"The MBF combined algorithm is not suitable directly on token graphs for looking for non‑loop paths between a pair of nodes if negative loops exist in the network."}]},"datasets":{"status":"available","items":[{"name":"Uniswap V2 token graph","provider":"The Graph","kind":"crypto-onchain","access":"public","period":"historical daily snapshots (date range n","universe":"≈100 tokens and ≈400 liquidity pools","quote":"we chose one hundred tokens and about four hundred liquidity pools in our research.","known_id":"the-graph"}],"known":[{"id":"the-graph","name":"The Graph (subgraphs)","provider":"The Graph","kind":"crypto-onchain","access":"public","quote":"Their informa-tion on Uniswap V2 is obtained from the The Graph 4, which provides detailed information for each token of the liquidity pools, such as tokens’ addresses, names, symbols, reserves, and decimals.","mentions":3,"used_in":2}],"sample":{"start":null,"end":null,"market":"Uniswap V2","asset_class":"crypto","frequency":null},"data_resolution":"unclear"},"transaction_costs":{"status":"available","modeled":"no","headline_net_of_costs":null,"detail":null,"quote":"Another limitation of this paper is that the gas fee is not considered. The gas fee is a significant factor that deserves to be considered in on‑chain transactions.","source":"evidence_card","rule":"R2","candidates":{"front_matter":"unclear","evidence_card":"no","evidence_card_quoted":true},"card_detail":null,"card_quote":"Another limitation of this paper is that the gas fee is not considered. The gas fee is a significant factor that deserves to be considered in on‑chain transactions.","meaning":"a trading or execution result is reported without costs"},"results":{"status":"available","evaluation_type":"trading-pnl","findings":[{"finding":"The proposed MMBF method discovers many more arbitrage loops and non‑loops than the MBF combined algorithm.","kind":"performance","quote":"The number of arbitrage loops and non‑loops by our method is much larger than that by the MBF combined algorithm, which denotes that our method can find more arbitrage opportunities than the MBF combined algorithm."},{"finding":"The largest arbitrage profit identified by the new method reaches about one million dollars, versus about one hundred thousand dollars for the MBF combined algorithm.","kind":"performance","quote":"The largest profit by our method can be a million dollars, compared to about one hundred thousand by the MBF combined algorithm."},{"finding":"23,868 arbitrage paths with profit > $1,000 are found by the new method, while only 19 such paths are found by the MBF combined algorithm.","kind":"statistical","quote":"23,868 arbitrage paths with more than a thousand dollars are found by our method, the number is only 19 by the MBF combined algorithm."},{"finding":"Arbitrage path lengths found by the new method are mostly between 7 and 11 hops, with a near‑symmetric distribution.","kind":"statistical","quote":"The distribution of arbitrage paths by our method is nearly symmetric, and we find that the length of most arbitrage paths ranges from 7 to 11."},{"finding":"Running time of the algorithm on the selected 100 tokens and 400 pools is about 8–10 seconds, indicating practical feasibility.","kind":"performance","quote":"the running time is about 8-10 seconds, which means that it can be used in practical arbitrage detection."}],"reported_performance":{"strategy":"modified Moore-Bellman-Ford algorithm (MMBF) on line graph","sharpe":null,"sharpe_benchmark":null,"annual_return_pct":null,"total_return_pct":null,"max_drawdown_pct":null},"out_of_sample":{"method":"none","headline_out_of_sample":false}},"independent_evidence":{"status":"none_found","supports":[],"contradicts":[],"qualifies":[],"compared_with":16,"citations":{"cited_by_in_corpus":1,"by_independent_authors":0,"how_cited":{"used_as_baseline":0,"uses_method_or_data":0,"extends":0,"discusses_findings":0,"mentions":0,"unclassified":1},"independent_papers":[],"scope":"Citations between papers in this corpus only; citations from outside it are not counted."}},"implementation":{"status":"available","public":false,"code_url":null,"note":"No repository presented by the authors as their own was found in the paper. Absence here is not proof that no code exists."},"replication":{"status":"none_found","card":null,"note":"No replication of this paper has been run by thequant.space yet."},"extraction":{"text_source":"fulltext","model":"openai/gpt-oss-120b","version":1,"extracted_at":"2026-11-14T18:07:18Z"},"provenance":{"claims":"Everything here is what the authors state in the paper. Nothing was re-run or reproduced by the platform.","quotes":"Assumptions, datasets, cost detail and findings were extracted by a language model. Every quote is verbatim paper text. Where an item has `support`, \"full\" means every figure in it appears in the quoted sentences and \"partial\" means a figure is stated elsewhere in the paper; items with a figure found nowhere in the paper were discarded.","datasets_known":"datasets.known lists well-known vendors and public datasets by a stable id, found by a dictionary over the full text where the sentence says the data is used; it is independent of the model and covers only names in the dictionary.","flags":"transaction_costs.modeled, results.evaluation_type and results.out_of_sample.method are platform classifications (structured-decision models); see https://thequant.space/score-guide/#penalty-flags","comparisons":"supports / contradicts / qualifies are a language model's reading of two papers' quoted findings, restricted to closely related papers that share no author. They are pointers to read, not verified replications, and coverage is partial.","implementation":"code_url is a repository the paper itself presents as its own (rule-scored, ambiguous cases judged by a model), checked for liveness when it was recorded."},"docs":"https://thequant.space/agents/"}