{"paper_id":"2306.09421","title":"FLAIR: A Metric for Liquidity Provider Competitiveness in Automated Market Makers","authors":["Jason Milionis","Xin Wan","Austin Adams"],"paper_date":"2023-06-15","paper_type":"empirical","paper_url":"https://arxiv.org/pdf/2306.09421v1","page_url":"https://thequant.space/flowcharts/flair-a-metric-for-liquidity-provider-competitiveness-in-au/","scores":{"empirical_rigor":7,"math_complexity":6,"quadrant":"Holy Grail"},"proposes":{"status":"available","statement":"The paper introduces a novel metric called FLAIR (fee liquidity‑adjusted instantaneous returns) to measure liquidity‑provider competitiveness in AMMs, supplementing existing flow‑toxicity metrics.","quote":"This work introduces a novel metric for LP competitiveness, called FLAIR (short for fee liquidity-adjusted instantaneous returns), that aims to supplement LVR in assessments of LP performance to capture the dynamic behavior of LPs in a pool.","summary":"This paper aims to enhance the understanding of liquidity provider (LP) returns in automated market makers (AMMs). LPs face market risk as well as adverse selection due to risky asset holdings in the pool that they provide liquidity to and the informational asymmetry between informed traders…","strategy":null},"assumptions":{"status":"available","items":[{"assumption":"The AMM pool consists of two assets, a risky asset A and a numéraire asset B, and all values are denominated in B.","type":"model","stated":"explicit","quote":"2 Metric Theory 2.1 Preliminaries Assume a two-asset AMM pool, where, without loss of generality, the first asset 𝐴is a risky asset, and the second asset 𝐵is a num´eraire asset (the unit of account), in the units of which the various portfolio values, fees, profit and loss will be denominated.","support":"full"},{"assumption":"An external market price trajectory {p_t} is given and ex‑post measures are evaluated against this realized path.","type":"market","stated":"explicit","quote":"Whenever our measures are ex-post, this means that there exists a specific realized price trajectory {𝑝𝑡}0≤𝑡≤𝑇, and that we are measuring the past performance against that specific sample path.","support":"full"},{"assumption":"The AMM is a constant‑function market maker (CFMM) defined by a bonding curve f(x,y)=L, with trades constrained to keep f constant.","type":"model","stated":"explicit","quote":"An AMM in the form of a constant function market maker (CFMM) is represented by a bonding curve 𝑓such that the only allowable trades are those that maintain a constant level set of the curve, i.e., 𝑓(𝑥, 𝑦) = 𝐿.","support":"full"},{"assumption":"Technical smoothness assumptions on the bonding curve allow definition of an implied pool price as the negative ratio of partial derivatives.","type":"model","stated":"explicit","quote":"Through this representation, at any state of the pool (𝑥, 𝑦), under suitable 6 technical assumptions, we can define a marginal price, the so-called implied pool (marginal/spot) price that is given by the negative of the ratio of the partial derivatives of 𝑓evaluated at the point (𝑥, 𝑦).","support":"full"},{"assumption":"The theory does not require arbitrageurs to rebalance the pool to the external price; mispricing between implied and external prices is allowed.","type":"model","stated":"explicit","quote":"2For developing the theory of our competitiveness metric, there is no need to assume that arbitrageurs rebalance the pool until the external market price, i.e., as if they would pay no fees for their trades. Our theory works in the completely general case where there is mispricing between the implied pool price ˜𝑝𝑡and the external market price 𝑝𝑡.","support":"full"},{"assumption":"Liquidity distributions are functions of price (piecewise‑constant) rather than single point values.","type":"model","stated":"explicit","quote":"In particular, we say that there exists a piecewise-constant liquidity distribution 𝐿(𝑝) that has the property that 𝐿(𝑝) = 𝐿𝑘for all 𝑝∈(𝑝𝑘, 𝑝𝑘+1) for any 𝑘∈ℐsuch that whenever the implied (marginal/spot) pool price 𝑝∈(𝑝𝑘, 𝑝𝑘+1), the valid reserves (𝑥, 𝑦) contained in the pool satisfy the relationships 𝑥= 𝑥⋆(𝑝, 𝐿(𝑝)) and 𝑦= 𝑦⋆(𝑝, 𝐿(𝑝)) for all 𝑝∈(𝑝𝑘, 𝑝𝑘+1) for any 𝑘∈ℐ.","support":"full"}]},"datasets":{"status":"none_found","items":[],"known":[],"sample":{"start":null,"end":null,"market":null,"asset_class":null,"frequency":null},"data_resolution":"unclear"},"transaction_costs":{"status":"not_applicable","modeled":"n/a","headline_net_of_costs":null,"detail":null,"quote":null,"source":"front_matter","rule":"R1","candidates":{"front_matter":"n/a","evidence_card":"unclear","evidence_card_quoted":false},"meaning":"the paper reports no trading or execution result"},"results":{"status":"available","evaluation_type":"none","findings":[{"finding":"FLAIR captures fee return on capital relative to the pool and aligns with intuitive notions of LP competitiveness, as shown in quadrant visualisations and example cases.","kind":"theoretical","quote":"To illustrate how both flow toxicity, accounting for the sophistication of the counterparty of LPs, as well as LP competitiveness, accounting for the sophistication of the competition among LPs, affect individual LP returns, we propose a quadrant interpretation where all of these characteristics may be readily visualized. FLAIR has particular merit in empirical analyses, and is able to better inform practical","support":"full"},{"finding":"In the fully competitive example with two LPs, the aggregate FLAIR equals 2γ(T−t0), demonstrating that perfectly competitive LPs earn fees without excess capital.","kind":"theoretical","quote":"As a direct consequence, the aggregate FLAIR for the pool would be CMagg(𝑡0, 𝑇) = ∫︁𝑇 𝑡0 fee𝑡 2fee𝑡 𝛾 𝑑𝑡= ∫︁𝑇 𝑡0 2𝛾𝑑𝑡= 2𝛾(𝑇−𝑡0) .","support":"full"},{"finding":"The metric can be applied ex‑post to existing LP positions and for back‑testing prospective strategies, normalising fee income by instantaneous capital deployed.","kind":"theoretical","quote":"2.2 FLAIR for Existing LP Positions As mentioned in the introduction, the ex-post competitiveness metric for existing LP positions should assess the backwards-looking profitability of the position, measured by the fee earned in comparison to the rest of the pool, normalized by the instantaneous amount of capital deployed on the AMM, as this position varies its liquidity distribution over time. In other words, the","support":"full"}],"reported_performance":null,"out_of_sample":{"method":"none","headline_out_of_sample":null}},"independent_evidence":{"status":"none_found","supports":[],"contradicts":[],"qualifies":[],"compared_with":10,"citations":{"cited_by_in_corpus":6,"by_independent_authors":3,"how_cited":{"used_as_baseline":0,"uses_method_or_data":0,"extends":0,"discusses_findings":0,"mentions":2,"unclassified":4},"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":2,"extracted_at":"2026-11-14T20:02:29Z"},"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/"}