{"paper_id":"2509.05065","title":"The Subtle Interplay between Square-root Impact, Order Imbalance & Volatility II: An Artificial Market Generator","authors":["Guillaume Maitrier","Grégoire Loeper","Jean-Philippe Bouchaud"],"paper_date":"2025-09-05","paper_type":"empirical","paper_url":"https://arxiv.org/pdf/2509.05065v1","page_url":"https://thequant.space/flowcharts/the-subtle-interplay-between-square-root-impact-order-imbal/","scores":{"empirical_rigor":7.2,"math_complexity":9.5,"quadrant":"Holy Grail"},"proposes":{"status":"available","statement":"The paper proposes a versatile synthetic market simulator that generates realistic order flow and price dynamics using a generalized propagator model of correlated metaorders.","quote":"First, it offers additional evidence for the robustness of our theoretical model by showing that the approximate analytical treatment proposed in [1] correctly describes the key empirical phenomena. Second, we introduce what we believe to be a versatile and realistic simulation tool that captures the intricate interplay between order flow and price formation, at least at the “meso” scale.","summary":"This work extends and complements our previous theoretical paper on the subtle interplay between impact, order flow and volatility. In the present paper, we generate synthetic market data following the specification of that paper and show that the approximations made there are actually justified…","strategy":null},"assumptions":{"status":"available","items":[{"assumption":"Metaorders are initiated as a Poisson process with rate ν per unit time.","type":"model","stated":"explicit","quote":"To succinctly recall the context, we sum-marize the model as follows: • The order flow is composed of a succession of metaorders, initiated with rate ν per unit time. In the rest of the paper, we will typically impose ν = 1.5 · 10−3 and φ = 2 · 10−3.","support":"full"},{"assumption":"The size of each metaorder follows a power‑law distribution Ψq(s) with a q‑dependent tail exponent µq.","type":"model","stated":"explicit","quote":"The size (i.e., the number of child orders per metaorder) is distributed according to a power law, Ψq(s), with a q-dependent tail exponent µq, where q is the size of the child orders, assumed to be constant within each metaorder.","support":"full"},{"assumption":"Child order volumes within a metaorder are log‑normally distributed with parameters (m, σℓ).","type":"model","stated":"explicit","quote":"Such 2 child volumes are distributed according to a lognormal distribution with parameters (m, σℓ).","support":"full"},{"assumption":"Sign autocorrelation of metaorders follows E[εt εt+τ] = Γτ −γ× for large τ.","type":"model","stated":"explicit","quote":"If εt is the sign of the tth metaorder of the day, we assume that for τ ≫1 E[εtεt+τ] = Γτ −γ×.","support":"full"},{"assumption":"A uniform participation rate φ is used across all metaorders.","type":"execution","stated":"explicit","quote":"For simplicity, we assume a uniform participation rate φ across metaorders. By adjusting ν and φ, one can control the average number of concurrently active metaorders.","support":"full"},{"assumption":"Impact of a child order scales as the square‑root of its volume, consistent with the square‑root law.","type":"statistical","stated":"explicit","quote":"This instrument is crafted to incorporate the three main stylized characteristics of the impact of metaorders (see [3, 6] and refs therein): (i) impact grows on average as the square-root of the number of child orders being executed, (ii) average peak impact at the end of the execution solely depends on the square root of the traded volume, and (iii) average i","support":"full"}]},"datasets":{"status":"available","items":[{"name":"Synthetic market data generated by the proposed simulator","provider":null,"kind":"synthetic","access":"public","period":null,"universe":null,"quote":"In the present paper, we generate synthetic market data following the specification of that paper and show that the approximations made there are actually justified, which provides quantitative support our conclusion that price volatility can be fully explained by the superposition of correlated metaorders which all impact prices, on average, as a square-root of executed volume. Whereas our theoretical model is","support":"full"},{"name":"TSE trade‑by‑trade dataset with real metaorder identifiers","provider":"TSE","kind":"other","access":"unknown","period":null,"universe":null,"quote":"A striking initial finding was that randomly shuffling trading IDs and reconstructing synthetic metaorders still preserves the square-root impact law (SQL): see [6] section 4.2 for details.","support":"full"}],"known":[],"sample":{"start":null,"end":null,"market":null,"asset_class":null,"frequency":null},"data_resolution":"synthetic"},"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":"no","evidence_card_quoted":false},"meaning":"the paper reports no trading or execution result"},"results":{"status":"available","evaluation_type":"none","findings":[{"finding":"Numerical simulations confirm that the analytical approximations of the theoretical model accurately reproduce key empirical phenomena, including the square‑root impact law.","kind":"theoretical","quote":"First, it offers additional evidence for the robustness of our theoretical model by showing that the approximate analytical treatment proposed in [1] correctly describes the key empirical phenomena. We now revisit this observable using simulations based on our model and confirm that the anomalous rescaling ∼T −ω is precisely recovered, validating the theoretical prediction.","support":"full"},{"finding":"The synthetic order flow reproduces the observed correlation structure between generalized order flow and returns.","kind":"statistical","quote":"One of the most striking predictions of our model is the structure of the correlation between generalized order flow and returns, which is observed empirically and reproduced using our synthetic market generator. As predicted by our model, increasing a—which gives greater weight to large-volume orders—reduces the scaling exponent.","support":"full"},{"finding":"Proxy metaorders constructed from simulated data recover the square‑root impact law, matching the theoretical prefactor Ytheory.","kind":"performance","quote":"A striking initial finding was that randomly shuffling trading IDs and reconstructing synthetic metaorders still preserves the square-root impact law (SQL): see [6] section 4.2 for details. For small Q/VD, synthetic metaorders tend to have less concave, but after a crossover value around 10−3, it nicely converges to the expected SQL, which is an input of our simulation.","support":"full"},{"finding":"Scaling exponent of the price diffusion moments decreases with increasing weight a on large‑volume orders, as predicted by the model.","kind":"statistical","quote":"As predicted by our model, increasing a—which gives greater weight to large-volume orders—reduces the scaling exponent. We set m = 6, σℓ= 1 and λσ2 ℓ= 1/8.","support":"full"},{"finding":"The covariance coefficient’s non‑monotonic shape as a function of a is explained by volume fluctuations and the β(q) relation.","kind":"theoretical","quote":"Our theoretical predictions suggest that the non-monotonic shape as a function of a originates from volume fluctuations — particularly in the upward branch, which depends on the parameter λ′ in the relation β(q) = β1 −λ′q (see Eq. (5)).","support":"full"}],"reported_performance":null,"out_of_sample":{"method":"none","headline_out_of_sample":null}},"independent_evidence":{"status":"available","supports":[{"paper":{"paper_id":"2311.18283","title":"The two square root laws of market impact and the role of sophisticated market participants","paper_date":"2023-11-30","authors":["Bruno Durin","Mathieu Rosenbaum","Grégoire Szymanski"],"empirical_rigor":4,"page_url":"https://thequant.space/flowcharts/the-two-square-root-laws-of-market-impact-and-the-role-of-so/"},"question":"What is the scaling law of market impact for large metaorders?","this_paper_finding":"Proxy metaorders constructed from simulated data recover the square‑root impact law, matching the theoretical prefactor Ytheory.","other_paper_finding":"A closed‑form scaling limit of large metaorders is derived, showing a power‑law decay of impact with exponent α.","other_paper_quote":"With these assumptions in force, we can now state our main result. (15) As explained previously, the size of the metaorder scales as TβT and the permanent impact should be linear in the volume.","note":"Both findings report a square‑root impact (α≈0.5) and confirm the theoretical prefactor Ytheory; B validates A with simulated proxy metaorders.","confidence":0.433}],"contradicts":[],"qualifies":[],"compared_with":18,"citations":{"cited_by_in_corpus":2,"by_independent_authors":1,"how_cited":{"used_as_baseline":0,"uses_method_or_data":0,"extends":0,"discusses_findings":0,"mentions":1,"unclassified":1},"independent_papers":[],"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/glatouille/ArtificialMarketSimulator","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":2,"extracted_at":"2026-11-14T19:48:08Z"},"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/"}