Resolving Latency and Inventory Risk in Market Making with Reinforcement Learning

Bo Li; Chang Yang; Junzhe Jiang; Xiao Huang; Xinrun Wang; Zhiming Li

REVIEW 5 major objections 5 minor 30 references

Resolving Latency and Inventory Risk in Market Making with Reinforcement Learning

T0 review · 5 major / 5 minor · reviewed 2026-08-15 · deepseek-v4-flash

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗ Review JSON

Pith's one-line read The paper proposes RELAVER, an RL-based market-making method that models exchange latency and batch matching, and claims it outperforms five existing strategies on four Chinese index-futures datasets while reducing inventory risk.

desk verdict New simulator and a hybrid RL method that likely helps, but the SOTA claim is unproven without the missing IMM comparison and a careful re-read of the IM numbers. read the letter →

arxiv 2505.12465 v1 pith:BRZUJITD submitted 2025-05-18 cs.LG

Junzhe Jiang , Chang Yang , Xinrun Wang , Zhiming Li , Xiao Huang , Bo Li This is my paper · ORCID

classification cs.LG

keywords marketmakingreinforcementlearningexecutionlatencyinventoryriskbatchmatchingorderholdtimedynamicprogrammingteachertrendprediction

verification ladder T0 review T1 audit T2 compute T3 formal

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The reading

This paper argues that existing reinforcement-learning market-making strategies fail in real markets because they ignore exchange latency and instant order cancellation, which lets orders miss their intended prices and lets inventory build up. To fix this, the paper builds a simulation with random 30-80 ms order delays and 500 ms batch matching, then proposes RELAVER, an RL agent that adds order-hold time to its action space, uses a dynamic-programming teacher to guide exploration, and employs a market-trend predictor to adjust inventory limits. On historical data from four Chinese index-futures contracts, RELAVER reports higher profit and better profit-per-inventory-risk than five baselines. If correct, the work makes latency-aware market making trainable with standard hardware and a viable target for RL.

What carries the argument

The framework has three coordinated components. First, the augmented action space adds per-side order hold times to traditional quote price and volume decisions, letting the agent set how long a quote remains active in a batch-matched, time-priority order book; this is what lets the policy navigate latency instead of fighting it. Second, the DP teacher computes an optimal action-value table from historical data with full future information, and the RL agent is trained with an additional KL-divergence penalty pulling its policy toward that teacher's actions, reducing random exploration. Third, a LightGBM trend predictor classifies the next 30-step price movement into four categories and maps them to dynamic inventory limits (100% or 50% of the cap), triggering market orders when inventory conflicts with the predicted trend. Together, the environment's random latency and batch-matching mechanics are the testbed on which these three components are evaluated.

What would settle it

Replay the same historical order flow through the actual exchange matching engine, or a high-fidelity simulation with measured latency distributions and cancellation rules, and compare RELAVER against the same baselines; if the profit and PnL/MAP advantages shrink to within noise, or if changing the latency range or batch interval materially flips the ranking, the central claim is not robust. A simpler diagnostic: run RELAVER with latency set to zero; the paper's design predicts meaningful degradation from latency-caused cancellations, so an absence of any degradation would undermine the mechanism.

Watch

Extended reading notes

Core claim

The paper's central claim is that explicitly modeling exchange latency (random 30-80 ms) and batch matching (500 ms) in both the environment and the agent's action space, specifically by letting the agent choose how long orders stay alive, allows a PPO-based market maker to simultaneously improve profitability and reduce inventory risk. It further claims that a dynamic-programming Q-table computed from historical data provides a teacher signal that speeds RL training and raises profit, and that a pre-trained trend predictor that shrinks or expands inventory limits with the predicted market direction cuts average position size without sacrificing returns. The evidence is a set of experiments on four Chinese stock-index futures products in which RELAVER achieves the highest episodic PnL on all four and the highest PnL-to-MAP ratio on three, with ablation studies attributing the profit gain mainly to the Q-teacher and the risk reduction mainly to the trend predictor.

Load-bearing premise

The load-bearing premise is that the simulated matching engine, with random latency U(30,80) ms, matching every 500 ms, and scheduled cancellations, behaves like a real index-futures exchange, and that the baselines were given a fair adaptation to this new environment; no real-exchange validation is supplied.

Editorial extensions

If this is right

Market-making strategies that assume zero latency and instant cancellation overestimate their performance and are unsuitable for live markets; the paper's latency-aware batch-matching environment provides a more realistic benchmark.
Including order hold time in the action space allows a market maker to respect time-priority matching and reduce unintended cancellations, a capability existing RL market-making methods lack.
A dynamic-programming Q-teacher computed offline from historical data meaningfully accelerates RL convergence and raises profitability, with about a 60 percent EPnL gain in the IC ablation.
A pre-trained trend predictor reduces inventory risk substantially, with about a 66 percent lower MAP in the IC ablation, improving PnL per unit of inventory.
On IC, IH, IF, and IM index-futures data, RELAVER attains the highest episodic PnL among the compared strategies and the highest PnL/MAP on three of the four products.
The environment introduces random latency, batch matching, and scheduled cancellations, making the simulation substantially closer to real exchange operation than the zero-latency, order-by-order environments used in prior work.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

Because the DP teacher relies on future information, its large boost to backtested profit may shrink in live trading; the trend predictor, which only uses current state, may be the component most responsible for any real-time edge.
The framework targets moderate-frequency market making (one decision per 500 ms), so its benefits probably do not extend to microsecond-latency competing firms; the economic value lies in making market making feasible without top-tier infrastructure.
If the batch-matching model is accurate, the approach should transfer to other venues that use periodic batch auctions, such as certain options or commodity exchanges, with minimal changes; this is testable but not demonstrated in the paper.
The 30-80 ms latency range is a single configuration; heavier-tailed real-world delays would likely require re-tuning the hold-time action range, and the performance margin over baselines may depend on that choice.

Share X Bluesky LinkedIn Reddit HN

Signed reviews

No signed human review yet.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Desk Editor's Note

read the letter

Colleague,

You asked for a quick read on arXiv:2505.12465. The paper builds a new market-making simulator with latency and batch matching and proposes RELAVER, which adds hold-time to the action space, a DP-derived Q-teacher, and a trend predictor. The environment is the most original piece; the hold-time action is a small but sensible extension, and the ablation shows both extra components help on the IC dataset. The DP teacher uses future prices in training, which they disclose, and the out-of-sample evaluation is separate, so that's not circular. The trend predictor is a heuristic, but it's tested.

The problem is the headline claim. The abstract says RELAVER 'significantly improves the performance of state-of-the-art RL-based MM strategies,' but Table 2 does not include IMM (Niu et al. 2023), the most relevant RL market-making baseline from their own lineage. The FAQ says IMM is designed for continuous matching and immediate execution, so 'incompatible' with batch matching and latency. That may be true, but they don't report even an attempted adaptation (for example, giving IMM the same hold-time parameterization). As it stands, the experiments show RELAVER beats the stripped PPO-LSTM (which per FAQ Q9 is just RELAVER without the new components) and a few rule-based baselines. That establishes the components help, but not the SOTA claim.

Also, the 'across multiple metrics' line doesn't hold on IM: there LIIC has better MAP and PnLMAP than RELAVER, even though RELAVER has higher EPnL. The reward coefficients are referenced rather than given, which is a reproducibility gap. The environment itself is unvalidated against real CFFEX execution dynamics, though that is a lesser issue if they open-source the code, which they promise.

Overall, this is a serious attempt at a real problem. The missing IMM comparison is fixable, and the overclaim is easily softened. The paper deserves a serious referee, with the clear request to either adapt IMM or drop the SOTA language, and to report the reward coefficients. I'd bring it to a reading group to argue about the baseline question. I'd cite it if the code and parameters appear.

Recommendation: send to peer review, conditional on those revisions.

Referee Report

5 major / 5 minor

Summary. The paper proposes RELAVER, an RL-based market-making method for environments with exchange latency and batch matching. The three claimed contributions are an augmented state-action space that adds order hold time to price and volume, a DP-computed Q-table that supervises PPO training, and a LightGBM trend predictor that adjusts inventory limits. The method is evaluated on four Chinese index futures datasets (IC, IH, IF, IM) against five baselines (FOIC, LIIC, AS, DRL_os, PPO-LSTM), with ablations on the IC dataset. The paper reports the highest episodic PnL for RELAVER on all four datasets and argues that the Q-teacher and trend predictor jointly improve profitability and inventory risk control. The use of future prices inside the Q-teacher is disclosed, and the out-of-sample test set is separate, so the main comparison is not circular; however, the headline 'state-of-the-art' claim is weakened by the exclusion of the closest SOTA baseline, IMM, and by several internal inconsistencies in the reward definition and reported numbers.

Significance. If the results hold, the paper makes a useful contribution to RL market making: it tackles a real operational constraint (latency and batch matching) that most RL simulators ignore, introduces order-hold-time actions, and provides an ablation study that isolates the effect of each proposed component. The out-of-sample evaluation and the explicit disclosure that the DP teacher uses future information during training are strengths. The paper also ships the environment with an open-source promise, which would benefit the community. The central limitation is that the experiments do not compare against the most relevant SOTA RL market maker, IMM, and the claimed improvement 'across multiple metrics' is contradicted by the IM row of Table 2. The simulation environment is plausible but not externally validated, and several hand-chosen parameters and reward coefficients are unreported or inconsistent. These issues are fixable within the scope of a revision.

major comments (5)

[Section 4.1, Table 2, FAQ Q2 and Q9] The claim that RELAVER significantly improves the performance of state-of-the-art RL-based MM strategies is not established because the closest SOTA baseline in the authors' own lineage, IMM (Niu et al., 2023), is never benchmarked. FAQ Q2 dismisses IMM as incompatible with latency and batch matching, but no adapted version is tested, and FAQ Q9 shows that the PPO-LSTM baseline is exactly RELAVER without Q-teacher and trend predictor. Consequently, Table 2 demonstrates that the two new components help over the stripped version, but it does not demonstrate an improvement over the SOTA named in the abstract. Please provide an adapted IMM baseline, or restrict the claim to the baselines actually compared.
[Table 2 and Section 4.2] The 'across multiple metrics' claim is contradicted by the IM row: LIIC achieves a lower MAP (77.222 vs 111.682) and a higher PnLMAP (0.303 vs 0.109) than RELAVER. In addition, no statistical significance testing is reported, and some gaps are within one standard deviation of the baseline, e.g., IF EPnL: RELAVER 93.282±39.940 vs FOIC 63.952±0.006. Please provide pairwise significance tests or confidence intervals and qualify the multi-metric claim.
[Section 2.3, Appendix D.3, and FAQ Q4] The reward function is defined inconsistently: Section 2.3 states R = PnL − IP + C − ER, while Appendix D.3 states R = PnL + IP + C + ER, and IP is itself written as −η|Q|·I(|Q|>d), making the sign of the inventory penalty ambiguous. Moreover, the coefficients η, β, σ, λ, and the KL coefficient α are never given numerically; FAQ Q4 only says they follow Niu et al. (2023). Since Table 3 attributes a roughly 60% EPnL improvement to the Q-teacher, the exact reward and loss coefficients are needed for reproducibility.
[Section 2.1, Section 4.1, Table 6, Appendix E] The realism of the proposed environment is asserted but not validated. The latency U(30,80) ms, 500ms batch interval, five pending orders per side, inventory limit d=±8 contracts, trend thresholds ±0.01, and 100%/50% exposure limits are hand-chosen heuristics, with no sensitivity analysis or calibration to actual CFFEX auction and latency behavior. The abstract also states 30–100 ms while the body says 30–80 ms. Without external validation, the reported advantages may be artifacts of the simulator, and the claim that prior methods 'cannot be applied in real MM scenarios' is too strong.
[Section 3.2, Algorithm 1] The Q-teacher is a central contribution but is under-specified. The six-dimensional action (δ, φ, W) is compressed into a single index a in Algorithm 1, the '10 distinct states' are not defined, and the recurrence Q*[t,p,a] ← max_{a'} Q*[t+1,a,a'] + a·pb_{t+1} − (p·pb_t + E_t(p−a)) uses an unstated term E_t. Without a precise definition of the states, actions, and transition used by the DP, the teacher cannot be reproduced or checked for optimality. Please add these details or state the simplifications explicitly.

minor comments (5)

[Abstract vs Section 2.1/Table 6] The abstract says the random delay is 30–100 ms, while the body and Table 6 consistently say 30–80 ms; please reconcile these numbers.
[Conclusion and Section 4.1] The conclusion describes the instruments as 'stock index options', but IC, IH, IF, and IM are Chinese stock index futures; please correct the terminology.
[Section 4.2, Table 2] The sentence 'in the IF environment, its MAP is 111.682' is incorrect: Table 2 lists the IF MAP as 125.494, while 111.682 is the IM value. The following comparison with LIIC and DRL_os is also wrong for LIIC.
[Tables 2 and 3] The PnLMAP values in Table 3 are reported without the [10^3] unit used in Table 2; please make the units consistent.
[Figure 2 and typographical errors] The Figure 2 caption has garbled notation ('(s_t^!, s_t^#, s_t^$)' and 'W_t^a,b'), and the text contains typos such as 'Sequctentially' and 'evluate'; please proofread throughout.

Circularity Check

0 steps flagged · score 0.0 of 10

No circularity found: the DP Q-teacher's lookahead is disclosed training-time supervision, and the out-of-sample evaluation plus transparent ablation keep the central claim non-circular.

full rationale

The paper's derivation chain—realistic latency and batch-matching environment, augmented state-action representation with hold time, DP Q-teacher, pretrained trend predictor, and PPO-LSTM base—does not reduce to its own inputs by construction. The only lookahead element is the Q-teacher: Algorithm 1 constructs Q* using future price data (line 7 uses pb_{t+1}), and Section 3.3 openly states that the Q-teacher 'inherently presupposes access to future information.' That supervision is applied only during training via the KL term in the loss L(theta) = L_RL + alpha KL(pi_theta || pi*); the final evaluation is on a separate out-of-sample period (August-October 2023, Section 4.1), so the reported performance is not the teacher's hindsight value. The trend predictor is a fitted LightGBM model with hand-chosen thresholds and inventory limits; it is an empirical heuristic, not a quantity derived from the target results. FAQ Q2's exclusion of IMM is a benchmark completeness concern rather than circularity, because the paper does not redefine IMM's result as its own. FAQ Q9 explicitly equates RELAVER without Q-teacher and trend predictor to the vanilla PPO-LSTM baseline, making the ablation honest instead of a renamed-known-result. Citations to Niu et al. 2023 for reward coefficients and benchmark conventions are external prior work, not load-bearing self-citations that force the conclusion. Overall, no equation or fitted parameter makes the central claim equivalent to its inputs by definition.

Assumptions & free parameters 8 free parameters · 5 assumptions · 0 invented entities

The central claim rests on an unvalidated simulation environment, an under-specified DP teacher, a trend predictor with no reported accuracy, and a reward function with unspecified coefficients. These are assumptions or fitted heuristics rather than externally grounded facts.

free parameters (8)

latency distribution = U(30,80) ms (abstract says 30-100 ms)
Chosen to model exchange delays; not validated against real latency data.
batch matching interval = 500 ms
Chosen to simulate batch auctions; no sensitivity analysis.
inventory limit d = ±800 units (8 contracts)
Set per exchange constraints; no tuning reported.
trend thresholds = ±0.01 for bull/bear, 0 to ±0.01 for steady
Hand-picked classification cutoffs for the trend predictor; no justification given.
inventory exposure limits = 100% for strong trends, 50% for weak trends
Hand-chosen risk management rules without sensitivity analysis.
reward coefficients (eta, beta, sigma, lambda, alpha) = not reported
The paper cites Niu et al. 2023 for the base reward and introduces a stacking execution risk term, but gives no numeric values for any coefficient.
action discretization = price ±0.06 step 0.02; volume 100-300 step 100; wait time 0.5-2.5 step 0.5
Design choices for the action space; not justified.
maximum pending orders = 5 per side
Heuristic from prior MM practice.

assumptions (5)

domain assumption The simulated market environment with U(30,80) ms latency and 500 ms batch matching faithfully represents real market execution.
No validation against real latency or matching data; the environment is built from scratch by the authors.
ad hoc to paper The DP Q-teacher's backward induction over the training data provides a useful supervision signal for the policy.
The recurrence in Algorithm 1 is under-specified and assumes a simplified 10-state model; its optimality is asserted.
domain assumption LightGBM trained on June-September 2022 can predict 30-step-ahead price direction classes accurately enough to guide inventory control.
No accuracy or calibration results for the trend predictor are reported.
domain assumption The reward function, including the stacking execution risk term, correctly captures the market maker's utility.
Derived from prior work (Niu et al. 2023) plus an ad hoc ER term; the sign of ER is inconsistent between Section 2.3 and Appendix D.3.
domain assumption Batch matching at 500 ms with price-time priority approximates the auction mechanism used by some exchanges.
Cited from general exchange practice, but no specific exchange or empirical latency distribution is cited.

how reviews work

0 comments

Cite this review

Pith. "Pith review of Resolving Latency and Inventory Risk in Market Making with Reinforcement Learning." pith.science (2026). https://pith.science/paper/BRZUJITD

@misc{pith2026250512465,
  author       = {Pith},
  title        = {Pith review of: Resolving Latency and Inventory Risk in Market Making with Reinforcement Learning},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/BRZUJITD}},
  note         = {Machine review of arXiv:2505.12465}
}

read the original abstract

The latency of the exchanges in Market Making (MM) is inevitable due to hardware limitations, system processing times, delays in receiving data from exchanges, the time required for order transmission to reach the market, etc. Existing reinforcement learning (RL) methods for Market Making (MM) overlook the impact of these latency, which can lead to unintended order cancellations due to price discrepancies between decision and execution times and result in undesired inventory accumulation, exposing MM traders to increased market risk. Therefore, these methods cannot be applied in real MM scenarios. To address these issues, we first build a realistic MM environment with random delays of 30-100 milliseconds for order placement and market information reception, and implement a batch matching mechanism that collects orders within every 500 milliseconds before matching them all at once, simulating the batch auction mechanisms adopted by some exchanges. Then, we propose Relaver, an RL-based method for MM to tackle the latency and inventory risk issues. The three main contributions of Relaver are: i) we introduce an augmented state-action space that incorporates order hold time alongside price and volume, enabling Relaver to optimize execution strategies under latency constraints and time-priority matching mechanisms, ii) we leverage dynamic programming (DP) to guide the exploration of RL training for better policies, iii) we train a market trend predictor, which can guide the agent to intelligently adjust the inventory to reduce the risk. Extensive experiments and ablation studies on four real-world datasets demonstrate that \textsc{Relaver} significantly improves the performance of state-of-the-art RL-based MM strategies across multiple metrics.

Figures

Figures reproduced from arXiv: 2505.12465 by the authors.

**Figure 2.** Illustration of the MM environment and R [PITH_FULL_IMAGE:figures/full_fig_p004_2.png] view at source ↗

**Figure 3.** Illustration of action mapping in the limit order book. Two [PITH_FULL_IMAGE:figures/full_fig_p004_3.png] view at source ↗

Figures from the paper (3 more)

**Figure 4.** Figure 4: Market trends and corresponding inventory limits [PITH_FULL_IMAGE:figures/full_fig_p005_4.png]

**Figure 5.** Figure 5: Up and Down Trend Performance Percentage (MAP) from 243.952 to 81.439, a decrease of approximately 66%. Even when lacking either Q-Teacher or Trend Expert, RELAVER still outperformed the fully ablated version (RELAVERw/o qt, te): the model without Q-Teacher excelled …

**Figure 6.** Figure 6: Testing days of model’s daily adverse selection ratio and [PITH_FULL_IMAGE:figures/full_fig_p007_6.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

30 extracted references · 27 canonical work pages

[1]

Adaptive market making via online learning

[Abernethy and Kale, 2013] Jacob Abernethy and Satyen Kale. Adaptive market making via online learning. Ad- vances in Neural Information Processing Systems , 26,

work page 2013
[7]

Market making under order stacking framework: A deep reinforcement learning ap- proach

[Chung et al., 2022] Guhyuk Chung, Munki Chung, Yong- jae Lee, and Woo Chang Kim. Market making under order stacking framework: A deep reinforcement learning ap- proach. In Proceedings of the Third ACM International Conference on AI in Finance, pages 223–231,

work page 2022
[9]

Citadel’s market maker posts record revenue; hedge fund surges,

[Doherty and Burton, 2023] Katherine Doherty and Kather- ine Burton. Citadel’s market maker posts record revenue; hedge fund surges,

work page 2023
[12]

Op- timal market making in the presence of latency

[Gao and Wang, 2020] Xuefeng Gao and Yunhan Wang. Op- timal market making in the presence of latency. Quantita- tive Finance, 20(9):1495–1512,

work page 2020
[15]

Dealing with the inventory risk: a solution to the market making problem

[Gu´eant et al., 2013] Olivier Gu ´eant, Charles-Albert Lehalle, and Joaquin Fernandez-Tapia. Dealing with the inventory risk: a solution to the market making problem. Mathematics and financial economics, 7:477–507,

work page 2013
[16]

Optimal high-frequency trading with limit and mar- ket orders

[Guilbaud and Pham, 2013] Fabien Guilbaud and Huyen Pham. Optimal high-frequency trading with limit and mar- ket orders. Quantitative Finance, 13(1):79–94,

work page 2013
[17]

Optimization of inventory holding cost due to price, weight, and volume of items

[Gurtu, 2021] Amulya Gurtu. Optimization of inventory holding cost due to price, weight, and volume of items. Journal of Risk and Financial Management , 14(2):65,

work page 2021
[18]

Trading and exchanges: Market microstructure for practitioners

[Harris, 2002] Larry Harris. Trading and exchanges: Market microstructure for practitioners. Oxford university press,

work page 2002

Show all 30 references

[22]

Deep reinforcement learning for market making

[Kumar, 2020] Pankaj Kumar. Deep reinforcement learning for market making. In Proceedings of the 19th Interna- tional Conference on Autonomous Agents and MultiAgent Systems, pages 1892–1894,

2020
[23]

High-frequency market-making with inventory con- straints and directional bets

[Labadie and Fodra, 2013] Mauricio Labadie and Pietro Fo- dra. High-frequency market-making with inventory con- straints and directional bets. Quantitative Finance,

2013
[24]

An end-to-end optimal trade execution framework based on proximal policy optimization

[Lin and Beling, 2021] Siyu Lin and Peter A Beling. An end-to-end optimal trade execution framework based on proximal policy optimization. In Proceedings of the Twenty-Ninth International Conference on International Joint Conferences on Artificial Intelligence , pages 4548– 4554,

2021
[25]

Market microstruc- ture: A survey

[Madhavan, 2000] Ananth Madhavan. Market microstruc- ture: A survey. Journal of Financial Markets , 3(3):205– 258,

2000
[27]

Price dynamics in limit order markets

[Parlour, 1998] Christine A Parlour. Price dynamics in limit order markets. The Review of Financial Studies , 11(4):789–816,

1998
[29]

Per- formance of deep reinforcement learning for high fre- quency market making on actual tick data

[Xu et al., 2022] Ziyi Xu, Xue Cheng, and Yangbo He. Per- formance of deep reinforcement learning for high fre- quency market making on actual tick data. In Proceed- ings of the 21st International Conference on Autonomous Agents and Multiagent Systems, pages 1765–1767,

2022
[30]

ask” or “bid

A Notations Notation Description t Time Step t sl t Latency between order placement and exe- cution te Time of order execution ot Submitted order at timet pt Submitted order price at timet qt Submitted order volume at timet τ “ask” or “bid” side of order OBt Limit Order Book a...

2023
[1981]

Long short-term memory

[Hochreiter and Schmidhuber, 1997] Sepp Hochreiter and J¨urgen Schmidhuber. Long short-term memory. Neural Computation, 9(8):1735–1780,

1997
[1997]

Market making with scaled beta policies

[Jerome et al., 2022] Joseph Jerome, Gregory Palmer, and Rahul Savani. Market making with scaled beta policies. In Proceedings of the Third ACM International Conference on AI in Finance, pages 214–222,

2022
[1998]

Deep reinforcement learning in cryptocurrency market making

[Sadighian, 2019] Jonathan Sadighian. Deep reinforcement learning in cryptocurrency market making. arXiv preprint arXiv:1911.08647,

2019 arXiv
[2000]

IMM: An imi- tative reinforcement learning approach with predictive rep- resentation learning for automatic market making

[Niu et al., 2023] Hui Niu, Siyuan Li, Jiahao Zheng, Zhouchi Lin, Jian Li, Jian Guo, and Bo An. IMM: An imi- tative reinforcement learning approach with predictive rep- resentation learning for automatic market making. arXiv preprint arXiv:2308.08918,

2023 arXiv
[2002]

Optimal dealer pricing under transactions and return uncertainty

[Ho and Stoll, 1981] Thomas Ho and Hans R Stoll. Optimal dealer pricing under transactions and return uncertainty. Journal of Financial economics, 9(1):47–73,

1981
[2005]

Reinforcement learning for market mak- ing in a multi-agent dealer market

[Ganesh et al., 2019] Sumitra Ganesh, Nelson Vadori, Mengda Xu, Hua Zheng, Prashant Reddy, and Manuela Veloso. Reinforcement learning for market mak- ing in a multi-agent dealer market. arXiv preprint arXiv:1911.05892,

2019 arXiv
[2008]

Market making via reinforcement learning

[Beysolow II and Beysolow II, 2019] Taweh Beysolow II and Taweh Beysolow II. Market making via reinforcement learning. Applied Reinforcement Learning with Python: With OpenAI Gym, Tensorflow, and Keras , pages 77–94,

2019
[2011]

[Chen, 2019] W. Chen. Inventory risk management of HF market making. Technical report, Huatai Futures,

2019
[2013]

High-frequency trading in a limit order book

[Avellaneda and Stoikov, 2008] Marco Avellaneda and Sasha Stoikov. High-frequency trading in a limit order book. Quantitative Finance, 8(3):217–224,

2008
[2015]

Market making and mean reversion

[Chakraborty and Kearns, 2011] Tanmoy Chakraborty and Michael Kearns. Market making and mean reversion. In Proceedings of the 12th ACM conference on Electronic commerce, pages 307–314,

2011
[2019]

Risk metrics and fine tuning of high-frequency trading strategies

[Cartea and Jaimungal, 2015] ´Alvaro Cartea and Sebastian Jaimungal. Risk metrics and fine tuning of high-frequency trading strategies. Mathematical Finance, 25(3):576–611,

2015
[2020]

Market making with signals through deep reinforcement learning

[Gaˇsperov and Kostanjˇcar, 2021] Bruno Ga ˇsperov and Zvonko Kostanjˇcar. Market making with signals through deep reinforcement learning. IEEE access , 9:61611– 61622,

2021
[2021]

Deep reinforcement learning for market making in corporate bonds: beating the curse of dimensionality

[Gu´eant and Manziuk, 2019] Olivier Gu ´eant and Iuliia Manziuk. Deep reinforcement learning for market making in corporate bonds: beating the curse of dimensionality. Applied Mathematical Finance, 26(5):387–452,

2019
[2022]

Observations on market liquidity enhancement,

[Council, 2022] HK Financial Services Development Coun- cil. Observations on market liquidity enhancement,

2022
[2023]

Limit order book as a market for liquidity

[Foucault et al., 2005] Thierry Foucault, Ohad Kadan, and Eugene Kandel. Limit order book as a market for liquidity. The review of financial studies, 18(4):1171–1217,

2005

Pith tools

Reviewed August 15, 2026 · model on record in the stance chip above.

[1] [1]

Adaptive market making via online learning

[Abernethy and Kale, 2013] Jacob Abernethy and Satyen Kale. Adaptive market making via online learning. Ad- vances in Neural Information Processing Systems , 26,

work page 2013

[2] [7]

Market making under order stacking framework: A deep reinforcement learning ap- proach

[Chung et al., 2022] Guhyuk Chung, Munki Chung, Yong- jae Lee, and Woo Chang Kim. Market making under order stacking framework: A deep reinforcement learning ap- proach. In Proceedings of the Third ACM International Conference on AI in Finance, pages 223–231,

work page 2022

[3] [9]

Citadel’s market maker posts record revenue; hedge fund surges,

[Doherty and Burton, 2023] Katherine Doherty and Kather- ine Burton. Citadel’s market maker posts record revenue; hedge fund surges,

work page 2023

[4] [12]

Op- timal market making in the presence of latency

[Gao and Wang, 2020] Xuefeng Gao and Yunhan Wang. Op- timal market making in the presence of latency. Quantita- tive Finance, 20(9):1495–1512,

work page 2020

[5] [15]

Dealing with the inventory risk: a solution to the market making problem

[Gu´eant et al., 2013] Olivier Gu ´eant, Charles-Albert Lehalle, and Joaquin Fernandez-Tapia. Dealing with the inventory risk: a solution to the market making problem. Mathematics and financial economics, 7:477–507,

work page 2013

[6] [16]

Optimal high-frequency trading with limit and mar- ket orders

[Guilbaud and Pham, 2013] Fabien Guilbaud and Huyen Pham. Optimal high-frequency trading with limit and mar- ket orders. Quantitative Finance, 13(1):79–94,

work page 2013

[7] [17]

Optimization of inventory holding cost due to price, weight, and volume of items

[Gurtu, 2021] Amulya Gurtu. Optimization of inventory holding cost due to price, weight, and volume of items. Journal of Risk and Financial Management , 14(2):65,

work page 2021

[8] [18]

Trading and exchanges: Market microstructure for practitioners

[Harris, 2002] Larry Harris. Trading and exchanges: Market microstructure for practitioners. Oxford university press,

work page 2002