{"paper_id":"2508.08152","title":"Optimal Fees for Liquidity Provision in Automated Market Makers","authors":["Steven Campbell","Philippe Bergault","Jason Milionis","Marcel Nutz"],"paper_date":"2025-08-11","paper_type":"empirical","paper_url":"https://arxiv.org/pdf/2508.08152v1","page_url":"https://thequant.space/flowcharts/optimal-fees-for-liquidity-provision-in-automated-market-mak/","scores":{"empirical_rigor":8,"math_complexity":7,"quadrant":"Holy Grail"},"proposes":{"status":"available","statement":"The paper proposes a dynamic reduced‑form model of an AMM operating alongside a CEX to study LP profitability and optimal fees.","quote":"We study the key determinants of LP profitability in a dynamic reduced-form model where an AMM operates in parallel with a centralized exchange (CEX), traders route their orders optimally to the venue offering the better price, and arbitrageurs exploit price discrepancies.","summary":"Passive liquidity providers (LPs) in automated market makers (AMMs) face losses due to adverse selection (LVR), which static trading fees often fail to offset in practice. We study the key determinants of LP profitability in a dynamic reduced-form model where an AMM operates in parallel with a…","strategy":"Optimal AMM fee optimization for liquidity providers"},"assumptions":{"status":"available","items":[{"assumption":"LPs are passive and pre‑commit liquidity for the entire trading horizon.","type":"model","stated":"explicit","quote":"Recall that for our model we assume that LPs are passive and precommit liquidity to the pool for the entire trading horizon."},{"assumption":"The AMM fee η1 is publicly known and fixed to the Uniswap v2 fee (30 bps).","type":"data","stated":"explicit","quote":"Specifically, we fix η1 to the Uniswap v2 fee for the ETH/USDC pool, i.e., η1 = 30 bps."},{"assumption":"The trade size ΔB = |ΔS| = ¯Δ is constant over time.","type":"model","stated":"explicit","quote":"For simplicity, we assume that ∆B = |∆S| =: ¯∆, and that this rate is constant over time."},{"assumption":"CEX price S0t follows an Ornstein‑Uhlenbeck process.","type":"model","stated":"explicit","quote":"by modeling S0t as the exponential of an Ornstein-Uhlenbeck (OU) process"},{"assumption":"Simulation uses discrete periods of equal length δt = T/N.","type":"model","stated":"explicit","quote":"Trading evolves over a time interval [0, T] and is simulated as N discrete periods of equal length δt = T/N."},{"assumption":"Fees are disbursed directly to the LP and treated as part of LP PnL.","type":"execution","stated":"explicit","quote":"We also treat fees as though they are disbursed directly to the LP."}]},"datasets":{"status":"available","items":[{"name":"Binance ETH/USDC 1‑minute price data","provider":"Binance","kind":"market-prices","access":"public","period":"January 2025","universe":"ETH/USDC pair","quote":"Binance and Uniswap, we estimate the model’s three core parameters","known_id":"binance"},{"name":"Uniswap v2 ETH/USDC pool data","provider":"Uniswap","kind":"crypto-onchain","access":"public","period":"January 2025","universe":"ETH/USDC pool","quote":"we fix the liquidity used for calibration to be 2.3·10^7 USDC and 6904 ETH, which roughly corresponds to the average values of the pool."},{"name":"Binance price data reference","provider":"Binance","kind":"market-prices","access":"public","period":null,"universe":null,"quote":"Binance price data: https://data.binance.vision/?prefix=data/spot/monthly/klines/ETHUSDC/1m/.","known_id":"binance"}],"known":[{"id":"binance","name":"Binance","provider":"Binance","kind":"market-prices","access":"public","quote":"That said, our implementation can accept an arbitrary time series as input for S0 t , and Section 5 presents an empirical analysis where S0 t follows a historical time series from Binance instead of a simulated price.","mentions":12,"used_in":2}],"sample":{"start":null,"end":null,"market":"ETH/USDC on Binance","asset_class":"crypto","frequency":"intraday"},"data_resolution":"intraday-bars"},"transaction_costs":{"status":"available","modeled":"yes","headline_net_of_costs":true,"detail":"CEX fee calibrated near posted taker rates; AMM fee η1 = 30 bps treated as LP revenue.","quote":"The calibrated CEX fee converges near posted taker rates, and estimated volatility falls within plausible intraday ranges.","source":"front_matter","rule":"R3","candidates":{"front_matter":"yes","evidence_card":"yes","evidence_card_quoted":true},"card_detail":"CEX fee calibrated near posted taker rates; AMM fee η1 = 30 bps treated as LP revenue.","card_quote":"The calibrated CEX fee converges near posted taker rates, and estimated volatility falls within plausible intraday ranges.","meaning":"costs are part of the reported headline results"},"results":{"status":"available","evaluation_type":"trading-pnl","findings":[{"finding":"Under normal market conditions the optimal AMM fee is competitive with CEX trading costs and remains remarkably stable.","kind":"theoretical","quote":"We find that under normal market conditions, the optimal AMM fee is competitive with the trading cost on the CEX and remarkably stable, whereas in periods of very high volatility, a high fee protects passive LPs from severe losses."},{"finding":"Lowering the AMM fee below the effective CEX level improves LP profitability on real price data.","kind":"performance","quote":"Importantly, when testing LP performance on real price data, we find that lowering the AMM fee below the effective CEX level improves LP profitability—a conclus…"},{"finding":"A threshold‑type dynamic fee schedule (low fee in normal times, high fee in extreme volatility) is robust and improves LP outcomes.","kind":"theoretical","quote":"These findings suggest that a threshold-type dynamic fee schedule is both robust enough to market conditions and improves LP outcomes."},{"finding":"Calibration reproduces the empirical log‑price‑ratio distribution using only three parameters (CEX fee, volatility, fundamental demand).","kind":"pricing-calibration","quote":"With only these inputs, the model reproduces a price ratio distribution that closely aligns with empirical observations."},{"finding":"Increasing the AMM fee widens the profit region, reduces turnover, and contracts the price‑ratio distribution around its center.","kind":"statistical","quote":"As the AMM fee increases, the widening of the profit region and the decline in the AMM’s competitiveness restrict participation to trades that remain viable despite the higher cost."}],"reported_performance":{"strategy":"Optimal AMM fee optimization for liquidity providers","sharpe":null,"sharpe_benchmark":null,"annual_return_pct":null,"total_return_pct":null,"max_drawdown_pct":null},"out_of_sample":{"method":"unclear","headline_out_of_sample":null}},"independent_evidence":{"status":"available","supports":[],"contradicts":[{"paper":{"paper_id":"2302.05219","title":"Decentralized Exchanges: The Profitability Frontier of Constant Product Market Makers","paper_date":"2023-02-10","authors":["Tobias Bitterli","Fabian Schär"],"empirical_rigor":7.5,"page_url":"https://thequant.space/flowcharts/decentralized-exchanges-the-profitability-frontier-of-constant-product-market/"},"question":"How does changing the AMM transaction fee affect the size of the profitable allocation region?","this_paper_finding":"Increasing the AMM fee widens the profit region, reduces turnover, and contracts the price‑ratio distribution around its center.","other_paper_finding":"Higher relative transaction fees shift the profitability frontier inward, narrowing the set of profitable allocations.","other_paper_quote":"The shift leads to a narrower section of the k‑indifference curve between P1 and P1.","note":"Finding A reports higher fees shrink the profitable set, while Finding B reports higher fees expand it, so they contradict each other.","confidence":0.914}],"qualifies":[],"compared_with":47,"citations":{"cited_by_in_corpus":6,"by_independent_authors":4,"how_cited":{"used_as_baseline":0,"uses_method_or_data":0,"extends":0,"discusses_findings":2,"mentions":2,"unclassified":2},"independent_papers":[{"paper":{"paper_id":"2603.09669","title":"Competition between DEXs through Dynamic Fees","paper_date":"2026-03-10","authors":["Leonardo Baggiani","Martin Herdegen","Leandro Sanchez-Betancourt"],"empirical_rigor":3.5,"page_url":"https://thequant.space/flowcharts/competition-between-dexs-through-dynamic-fees/"},"how":"discusses_findings","quote":"... characterizes the optimal fee as a function of market conditions such as volatility and trading volume, suggesting a threshold-type dynamic fee schedule that is stable in normal conditions and changes with volatility."},{"paper":{"paper_id":"2403.18177","title":"Growth rate of liquidity provider's wealth in G3Ms","paper_date":"2024-03-27","authors":["Cheuk Yin Lee","Shen-Ning Tung","Tai-Ho Wang"],"empirical_rigor":2,"page_url":"https://thequant.space/flowcharts/growth-rate-of-liquidity-providers-wealth-in-g3ms/"},"how":"discusses_findings","quote":"... finding that the optimal fee is stable under normal market conditions but should be raised during periods of high volatility --- a finding that complements our characterization of the fee tier that maximiz"}],"scope":"Citations between papers in this corpus only; citations from outside it are not counted."}},"implementation":{"status":"available","public":true,"code_url":"https://github.com/JasonSome/cpmm-trading","note":null},"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":"2026-11-14T17:30:21Z"},"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/"}