{"paper_id":"2309.00632","title":"Improving Capital Efficiency and Impermanent Loss: Multi-Token Proactive Market Maker","authors":["Wayne Chen","Songwei Chen","Preston Rozwood"],"paper_date":"2023-08-17","paper_type":"empirical","paper_url":"https://arxiv.org/pdf/2309.00632v2","page_url":"https://thequant.space/flowcharts/improving-capital-efficiency-and-impermanent-loss-multi-tok/","scores":{"empirical_rigor":3,"math_complexity":9,"quadrant":"Lab Rats"},"proposes":{"status":"available","statement":"The paper proposes the Multi-Token Proactive Market Maker (MPMM), a multi‑token generalization of DODO Exchange's proactive market maker.","quote":"We explore this paradigm primarily through the construction of a generalization of PMM, the multi-token token proactive market maker (MPMM).","summary":"Current approaches to the cryptocurrency automated market makers result in poor impermanent loss and capital efficiency. We analyze the mechanics underlying DODO Exchange’s proactive market maker (PMM) to probe for solutions to these issues, leading to our key insight of multi-token trading pools.…","strategy":null},"assumptions":{"status":"available","items":[{"assumption":"The ratio PQ/PB remains fixed, so token balances naturally return to equilibrium.","type":"market","stated":"explicit","quote":"In summary, assuming the ratio PQ/PB remains fixed, token balances will naturally return to equilibrium.","support":"full"},{"assumption":"Oracles providing PB and PQ must quickly adapt to price fluctuations.","type":"model","stated":"explicit","quote":"A caveat is that since the PMM pricing curve uses PB and PQ provided from oracles, the oracles must quickly adapt to price fluctuations.","support":"full"},{"assumption":"Initial equilibrium balances B0 and Q0 are set equal to LPs' deposited token amounts.","type":"model","stated":"explicit","quote":"DODO sets their PMM B0 and Q0 equal to the token amounts deposited by LPs [14] to ensure the pool tracks a holding strategy, thus reducing impermanent loss.","support":"full"},{"assumption":"Liquidity providers can supply liquidity using a single token type.","type":"execution","stated":"explicit","quote":"Additionally, LPs can supply liquidity using single token types, reducing barriers to participation.","support":"full"},{"assumption":"The parameter k interpolates between constant‑sum and constant‑product pricing curves.","type":"model","stated":"explicit","quote":"We analyze PMM’s use of a parameter k to “interpolate” an exchange rate curve between the naive constant sum (CSMM) and constant product (CPMM) market makers.","support":"full"},{"assumption":"Historical hourly price data can be linearly interpolated between daily observations to generate hourly prices.","type":"data","stated":"explicit","quote":"3.1 Real World Price Simulations We use hourly historical price data from a Kaggle dataset for 14 token types [9] spanning from April 8th, 2021 to September 10th, 20231 and daily per-token transaction volumes from CoinGecko to generate swapping traffic.","support":"full"}]},"datasets":{"status":"available","items":[{"name":"CoinGecko per‑token transaction volumes","provider":"CoinGecko","kind":"alternative","access":"public","period":null,"universe":null,"quote":"3.1 Real World Price Simulations We use hourly historical price data from a Kaggle dataset for 14 token types [9] spanning from April 8th, 2021 to September 10th, 20231 and daily per-token transaction volumes from CoinGecko to generate swapping traffic.","support":"full","known_id":"coingecko"}],"known":[{"id":"coingecko","name":"CoinGecko","provider":"CoinGecko","kind":"market-prices","access":"public","quote":"3.1 Real World Price Simulations We use hourly historical price data from a Kaggle dataset for 14 token types [9] spanning from April 8th, 2021 to September 10th, 20231 and daily per-token transaction volumes from CoinGecko to generate swapping traffic.","mentions":1,"used_in":1},{"id":"kaggle","name":"Kaggle","provider":"Kaggle","kind":"other","access":"public","quote":"3.1 Real World Price Simulations We use hourly historical price data from a Kaggle dataset for 14 token types [9] spanning from April 8th, 2021 to September 10th, 20231 and daily per-token transaction volumes from CoinGecko to generate swapping traffic.","mentions":1,"used_in":1}],"sample":{"start":null,"end":null,"market":null,"asset_class":null,"frequency":null},"data_resolution":"intraday-bars"},"transaction_costs":{"status":"none_found","modeled":"unclear","headline_net_of_costs":null,"detail":null,"quote":null,"source":"none","rule":"R0b","candidates":{"front_matter":"unclear","evidence_card":"no","evidence_card_quoted":false},"meaning":"no cost treatment was found in the passages the extractor read; that is not evidence that costs are not modeled"},"results":{"status":"available","evaluation_type":"none","findings":[{"finding":"Simulations show that MPMM has lower impermanent loss and higher capital efficiency than comparable market makers across a variety of market scenarios.","kind":"performance","quote":"We show via simulations that MPMM has better impermanent loss and capital efficiency than comparable market makers under a variety of market scenarios.","support":"full"},{"finding":"Across a range of k values, MPMM consistently achieves a lower median impermanent loss that is less volatile compared to PMM.","kind":"performance","quote":"Anther noteworthy comparison is that MCPMM performs worse than CPMM while MPMM outperforms PMM across a range of k values (MPMM consistently has a lower median impermanent loss that is less volatile compared to PMM).","support":"full"},{"finding":"In the real‑world price simulation, CPMM, CMPMM, CSMM, and MCSMM suffer significantly more impermanent loss than PMM and MPMM.","kind":"performance","quote":"Overall, we see that CPMM, CMPMM, CSMM, and MCSMM all suffer significantly more than PMM and MPMM.","support":"full"},{"finding":"During a bull‑market simulation, PMM’s impermanent loss improves markedly, while MPMM performs worse than PMM under those conditions.","kind":"performance","quote":"Now, MCPMM and MPMM perform worse than CPMM and PMM, respectively. PMM’s impermanent loss significantly improves (e.g. −3.946e−6 vs −1.328e−5 when k = 0.05 and −3.792e−6 vs −1.450e−5 when k = 0.75 for the “bull market” vs the “real world price” simulation).","support":"full"},{"finding":"A spike in capital‑efficiency inefficiency is observed around the 100,000th swap in the simulations.","kind":"performance","quote":"Finally, when plotting the capital efficiency in 7, we also observe a spike in inefficiency surrounding the 100,000th swap.","support":"full"}],"reported_performance":null,"out_of_sample":{"method":"unclear","headline_out_of_sample":null}},"independent_evidence":{"status":"none_found","supports":[],"contradicts":[],"qualifies":[],"compared_with":16,"citations":{"cited_by_in_corpus":0,"by_independent_authors":0,"how_cited":{"used_as_baseline":0,"uses_method_or_data":0,"extends":0,"discusses_findings":0,"mentions":0,"unclassified":0},"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":"2026-11-14T20:01:40Z"},"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/"}