WorksEconomics, Econometrics and Finance
Learning AI Agents, Market Makers, and Stress in Automated Markets
Abstract
Automated markets increasingly route order flow from learning trading agents through automated market makers (AMMs). We model trader archetypes as AI agents defined by their reward functions and analyze the feedback they create on a constant-product AMM. Herd-dominated and arbitrage-dominated markets of equal size have identical return distributions and turnover, so no rule based on volatility, value-at-risk, stress frequency, or volume can tell them apart; yet their stress episodes differ in length by a factor unbounded in the feedback-to-noise ratio . Serial statistics identify the feedback sign from observations, and this rate is tight. When agents trade together, profit-seeking agents ride herds rather than lean against them, leading to all-or-nothing participation. Spread-earning market makers restore stability but make liquidity fragile, with discontinuous collapse and hysteresis; maker–taker pricing, not a uniform fee, is the effective instrument.
Introduction
A growing share of trading is done by automated, increasingly autonomous agents, and a growing share of matching, above all in digital assets, happens against automated market makers: deterministic pricing formulas such as the constant-product rule [1, 3]. Arbitrageurs, retail trend-followers, and front-runners [7] all trade in the same pools, and each leaves a different footprint on the frequency and persistence of stress. As AI agents take on these roles, proportionate regulation requires knowing which objectives stabilize a market and whether a regulator can determine this from data.
Two obstacles stand in the way. Behavior types are not labeled in public data, and the statistics regulators usually monitor (volatility, value-at-risk, the frequency of large losses, turnover) are properties of the one-period return distribution. We show that such statistics are provably blind to the property that matters most, identify what is not blind to it at an optimal rate, and analyze what learning agents and market makers do when they trade together. We take an objective-first view, following the empirical framework of [2]: each agent is defined by the reward it maximizes, so assumptions are explicit while strategy is emergent.
Our results
We model the pool’s log return as , where is the price impact of the period’s net order flow and is exogenous news with scale . Herd-followers trade in the direction of the last move, which we show is optimal under the herd-follower’s reward (Lemma 3.2). In the simplest case , where is net positive feedback (herding) and net negative feedback (arbitrage).
Volatility cannot see feedback (§ 4). Markets with feedback and have the same one-period return law and the same turnover, but their expected stress-episode lengths differ by , which is unbounded in (Theorem 4.2). Therefore, every functional of the return marginal and turnover treats a destabilizing and a stabilizing market identically (Corollary 4.3). For Gaussian news with , episodes last 6.3 periods under herding and 1.2 under arbitrage (Figure 1a).
Figure 1. (a) Theorem 4.2: expected stress-episode length (, Gaussian news) in a herd-dominated and an arbitrage-dominated market. The two have identical volatility and stress frequency at every . (b) Theorem 6.1(v): expected down-run length against rider capital (, , , ). Market makers exit at and return only below ; in the shaded band both outcomes are stable.
Serial statistics see it at an optimal rate. Autocorrelation and the variance ratio have the sign of (Proposition 4.4). A sign-agreement test identifies with error from observations, and no test using the full return path needs fewer than (Theorem 4.5). A herd-follower’s own stress episodes, by contrast, last at most two periods on every path (Proposition 4.6), so strategy-level measurements see herders as brief sources of stress and miss their effect on market persistence.
Agents playing together (§ 5). On a path-independent pool, a trader’s own impact cancels from its round-trip payoff (Lemma 5.1); only trading ahead of predictable herd flow pays. Leaning against the herd is therefore strictly dominated, riding with it is a strategic complement, and under any sign-preserving learning dynamics, including replicator dynamics and softmax policy gradient, participation tips to zero or to full depending on the starting point; full participation strictly lengthens stress, and no uniform fee can induce stabilizing behavior (Theorem 5.2, Corollary 5.3).
Market makers (§ 6). Spread-earning market makers enter the price process exactly as negative riders, so persistence depends only on the net index of rider capital minus absorbed flow. They supply the negative feedback fee-paying traders will not, but liquidity becomes fragile: when rider capital crosses a threshold market makers exit at once and persistence jumps, and they return only below , so the hysteresis band is exactly as wide as their absorption capacity (Theorem 6.1, Figure 1b). A maker rebate widens the stable region; a taker fee removes riders (Corollary 6.2).
Techniques. Three observations drive the analysis. (i) With symmetric noise, the return sign is a Markov chain whose flips are i.i.d. and independent of the current sign (Lemma 4.1), which decouples the marginal (a function of ) from the serial structure (a function of ). (ii) On a path-independent pool, the own-impact part of a round trip telescopes away in expectation (Lemma 5.1), so payoffs depend on aggregate behavior only through one persistence probability . (iii) is monotone in each population’s net flow, which makes the interaction game one of strategic complements within groups and substitutes across groups [16], so equilibria, stability, and learning basins follow from monotone comparative statics.
Empirical connection. Section 7 relates the results to the companion study [2], which trains these archetypes with deep reinforcement learning on hourly cryptocurrency data. The theory explains its main strategy-level patterns and shows which of them do not survive interaction.
Related work
De Long, Shleifer, Summers, and Waldmann [8] show that rational speculators may amplify positive-feedback traders; Theorem 5.2 derives this for learning agents on an AMM, with equilibrium tipping and learning basins. Agent-based finance shows that heterogeneous traders produce fat tails and crashes [14]. Classical microstructure explains how liquidity providers set prices [13, 9] and absorb temporary imbalances [10]; liquidity can evaporate in spirals [4], as in the 2010 Flash Crash [12]. Our market-maker result gives a closed-form collapse threshold and hysteresis band. Constant-function market makers are analyzed in [3], arbitrage costs against AMMs in [17], and front-running on decentralized exchanges in [7]; our execution model follows frequent batch auctions [5]. Learning algorithms can reach unintended outcomes such as tacit collusion [6]; our learning results use sign-preserving population dynamics [18] and games with strategic complementarities [16]. The variance-ratio test of [15] is a serial statistic of the kind Theorem 4.5 shows a regulator needs.
Organization. Section 2 recalls the archetypes’ rewards and Section 3 gives the AMM feedback model. Section 4 proves the impossibility and detection results, and Sections 5 and 6 analyze agents trading together and market makers. Section 7 connects the results to data, Section 8 discusses implications for regulators, and Appendix A verifies every closed form by simulation.
Trader archetypes as reward functions
We recall the archetypes of [2]. Trading is a discrete exposure-control problem. At each step , the agent observes a feature vector and chooses an action in , which moves its exposure by , , or with . Exposure is flat, half invested, and fully long; there is no short selling. The base reward is a volatility-normalized net return with a stress penalty:
where is the next-period asset return, the proportional transaction cost, trailing volatility, a small constant, and a market crash-regime indicator. Each archetype modifies (1):
Rational arbitrageur. Maximizes risk-adjusted return with additional penalties on realized volatility and turnover. It acts conservatively and reduces exposure as risk rises.
Manipulator. A stylized proxy for front-running and wash trading. It receives a bonus and a short discount horizon, which favors rapid position cycling.
Retail herd-follower. Mimics trend-chasing and fear of missing out. It receives a bonus for aligning exposure with recent momentum,
Sections 3–6 ask what these objectives do to a market when the agents’ trades move prices; the companion study [2] trains them on market data.
A feedback model of an automated market
The AMM. A constant-product pool with reserves quotes price . Depositing units of moves the reserves to with , so
Because the pool is path-independent, a period’s orders aggregate into one net flow , measured in units of the -reserve. We use the first-order impact with ; Remark 4.7 treats the exact impact.
Order flow. Time is discrete, , and is the pool’s return. Each period brings exogenous news (for example, moves in an external reference market transmitted to the pool) and endogenous flow from two populations:
mass of herd-followers, each buying one unit if the last return was positive and selling one unit otherwise; and
mass of negative-feedback traders, each trading one unit against the last return.
Writing and , the pool’s return is
Assumption 1. The are i.i.d. and independent of , with for , where has a continuous, strictly increasing distribution function that is symmetric about zero (). The initial sign is uniform on .
Definition 3.1 (Stress). For a threshold , period is stressed if equals one. A stress episode is a maximal run of consecutive stressed periods, and denotes its length. The stress frequency is , the volatility is , and the gross turnover is units per period.
From rewards to feedback. The next lemma shows that the herd-follower’s reward induces the herd rule discussed above.
Lemma 3.2 (The herd reward induces sign-momentum). Let , and suppose the agent is myopic (discount factor 0) with . If
then the unique expected-reward-maximizing action under (1)–(2) is buy when and sell when .
Proof. Write the action as , so , and let . The expected one-step reward is plus terms that do not depend on , where . Relative to holding, buying gains and selling gains . If , the hypothesis gives . Therefore, buying beats holding, and buying also beats selling by . If , then , so selling is the unique maximizer.
With a one-period momentum signal, a population of such agents emits a net flow , the herd term in (3). In § 4, the negative-feedback mass is a modeling device; § 5 asks whether any profit-seeking agent would choose to lean, and § 6 shows who does.
Volatility cannot see feedback
Lemma 4.1 (The sign process). Let . Under Assumption 1, the flips , , are i.i.d. with and independent of . Consequently, is a stationary symmetric Markov chain with uniform marginals.
Proof. . Conditional on , the variable has the law of , by independence and symmetry. So is independent of with . Since and is uniform, each is uniform.
Theorem 4.2 (Volatility equivalence, persistence separation). Under Assumption 1, for every and every :
(i) The law of is , where . It depends on only through . In particular, when has finite variance, and
(ii) Stress-episode lengths are i.i.d. geometric with , so
which is strictly increasing in .
(iii) For ,
and the ratio tends to as .
Proof. (i) By Lemma 4.1, is uniform and independent of , so is an equal mixture of and . Replacing by swaps the components. The variance follows because and are independent with mean , and is the mixture’s distribution function at .
(ii) If , then , so . Hence , whatever the history. Each period of an episode continues it independently with this probability, so is geometric with , which is strictly increasing in because is.
(iii) Apply (ii) at . As , the numerator tends to and the denominator to .
For Gaussian news and , : gives expected episodes of 6.3 periods, against 1.19 for , with identical volatility and stress frequency (Figure 1a).
Corollary 4.3 (Marginal-based rules cannot separate feedback). Let be any functional of the one-period return law and the gross turnover: volatility, value-at-risk, expected shortfall, stress frequency at any threshold, kurtosis, or a volume cap. Then takes the same value on the herd-dominated market and the arbitrage-dominated market , yet their expected stress-episode lengths differ by the factor in Theorem 4.2(iii), which is unbounded in .
Proof. The two markets have and the same gross turnover . Apply Theorem 4.2(i) and (iii).
A regulator must therefore look at serial structure. The simplest serial statistic already works.
Proposition 4.4 (Serial statistics see feedback). If, in addition, has finite variance, then and . Hence the first-order autocorrelation and the two-period variance ratio [15] have the sign of .
Proof. is a symmetric mixture by Theorem 4.2(i), so . Since is independent of with mean , , and because has a continuous law. For a stationary series, .
The next theorem gives the exact cost of reading the sign.
Theorem 4.5 (Tight sample complexity of detecting feedback). Assume Gaussian news () and . Given , the test that outputs “herding” iff identifies with error at most whenever
Conversely, for , every test that observes the full returns has error larger than under or under unless
Proof. Upper bound. By Lemma 4.1, with . Say ; the other case is symmetric. Hoeffding’s inequality [11] gives . For , , and . Hence the error is at most .
Lower bound. Let be the laws of under . The law of is the same under both (Theorem 4.2(i)). Conditional on the past, : two normals whose means differ by , so each conditional KL divergence equals , and by the chain rule . The Bretagnolle–Huber inequality [19] gives, for any test , . So the larger of the two errors is at least , which exceeds unless .
The sign-agreement test reads only , yet it matches the full-information lower bound up to a constant. This is the theoretical basis for using learned policies as measurement devices: a policy whose action depends on is, in effect, this test.
Proposition 4.6 (A herd-follower’s own stress is short). Consider a herd-follower with one-period momentum, exposure , step , net return , and strategy stress with . On every return path, its stress episodes last at most periods. With , that is two periods.
Proof. Strategy stress at requires , so and , and the agent then sells: . Exposure therefore falls by in each stressed period, and stress is impossible at zero exposure, where .
Proposition 4.6 and Theorem 4.2(ii) together explain a pattern found empirically in [2] (§ 7). At the strategy level, herd-followers show frequent but brief stress. At the market level, more herding lengthens stress episodes, because increases in . Strategy-level measurements capture the former; regulation should care about the latter.
Remark 4.7 (Exact constant-product impact). Let each population trade a net units of the -reserve per period. With the exact impact , a buy shifts the return by and a sell by , where . The sign chain stays Markov, and the proof of Theorem 4.2(ii) goes through state by state. In the herd market the continuation probability of a stress episode is ; in the arbitrage market it is . The ratio in part (iii) becomes , which is larger still. Part (i) and Corollary 4.3 hold up to terms.
Agents playing together
Section 4 fixed the mass of each feedback type. We now let profit-seeking agents choose their behavior in a pool where herd-followers already trade, and ask what learning converges to.
Players. A mass of herd-followers trades every period; by Lemma 3.2, a large enough momentum bonus makes this their behavior whatever it earns. A continuum of nonatomic rational traders with total mass each choose one of three rules: ride (), opening a one-unit position in direction ; lean (), opening a one-unit position in direction ; or stay out (). Positions opened in batch are closed in batch , and each round trip pays a fee . Write for the masses choosing and , and .
Prices. The batch- net flow is the herd flow plus the rational traders’ openings minus their closings, with impact
Orders within a batch execute in uniformly random order against linear impact, so each order’s expected fill is the batch midpoint , in the spirit of batch auctions [5]. A rider earns , and a leaner earns its negative.
The sign process. Call period continuing if and fresh otherwise. The sign persists with probability
because the rational openings and closings add only right after a flip. Whether a period is fresh is therefore a two-state Markov chain, and the stationary probability that the sign persists is
which is strictly increasing in and reduces to , the stay probability of Lemma 4.1, when .
Lemma 5.1 (Own impact cancels). In the stationary regime, the expected per-period payoffs are
They depend on the rational traders’ own flow only through .
Proof. . Let ; by stationarity it is also . Then
and because is independent of . Averaging the two, the terms cancel, and the expected gross payoff from riding is . Leaning is the opposite position. Subtracting the fee gives both formulas.
The terms cancel because a round trip on a path-independent pool cannot profit from its own impact. A rational trader profits only by trading ahead of predictable herd flow.
Theorem 5.2 (Rational agents join the herd; learning tips). Let with , and suppose .
(i) No equilibrium leans. Leaning is strictly dominated, , so in every Nash equilibrium. Without an external price anchor, negative feedback is never a rational response to herd flow.
(ii) Strategic complements. is strictly increasing in . Hence is an equilibrium iff ; is an equilibrium iff ; and at most one interior equilibrium exists, solving , and it is unstable. In particular, for both and are equilibria.
(iii) Learning tips. Consider any sign-preserving learning dynamics, with when and when , at speed bounded away from zero wherever the payoff is. This class includes replicator dynamics and expected softmax policy gradient [18]. Then . If , then ; if , then . With , the basin boundary is .
(iv) Joining amplifies stress. At threshold , down-runs have expected length
which is strictly increasing in . The high-participation equilibrium therefore has strictly more persistent stress than herding alone.
Proof. (i) From Lemma 5.1, because and .
(ii) is strictly increasing in because is, and , so is strictly increasing. At , . The three cases are the equilibrium conditions for an increasing payoff against the outside option 0. An interior root is unstable because is negative below it and positive above it.
(iii) Since everywhere, decreases to 0. Write . In the first case , so rises while falls; then rises, and since is increasing it stays positive, so . In the second case , so falls; then for all , stays negative, and .
(iv) At , a down-run starts with a flip, so its first continuation is a fresh period, which persists with probability ; every later continuation is a continuing period, which persists with probability . Hence with , and is strictly increasing in .
Corollary 5.3 (What a uniform fee can and cannot do). A uniform round-trip fee removes every equilibrium with rational participation. No uniform fee induces leaning: by part (i), for every . Negative feedback against herds must come from outside this game, for example from arbitrage to an external reference price or from agents who are paid to take the other side.
Proof. For , for every , so in equilibrium. The second claim restates part (i).
Theorem 5.2 connects the archetypes. The return-maximizing agent, placed in a pool with herds, learns to trade ahead of herd flow, which is the behavior the manipulator archetype was designed to proxy. How a profit-seeking agent behaves depends on who else is in the pool, so the archetypes cannot be studied in isolation. With , , , and , the basin boundary is ; softmax policy-gradient learners starting from 15% riders converge to 0.9%, and from 40% to 98%, while down-runs lengthen from 2.62 periods (herds alone) to 3.59 (Appendix A).
Adding profit-maximizing market makers
Corollary 5.3 states that negative feedback must come from agents who are paid to take the other side. We now add them: competitive market makers (MMs) who supply liquidity next to the pool and earn a spread for absorbing flow, as wholesalers internalize retail orders. An MM differs from a leaning trader in one respect only: it is paid to trade rather than paying to trade.
Market makers. A competitive mass of MMs absorbs a share of herd flow in batch , at the batch midpoint, and earns a net spread per unit absorbed. MMs carry the inventory for one period and unwind it in batch . Riders are as in Section 5, with mass (leaning remains dominated for them) and fee .
Prices. The herd’s net impact in batch is , plus the MMs’ unwinding . Adding the riders’ round trips,
So MMs enter exactly as negative riders, and everything depends on the net ignition kick : rider capital minus the herd flow absorbed by MMs. The fresh-period stay probability becomes , the continuing-period probability is unchanged, and .
Theorem 6.1 (Market makers). In the stationary regime:
(i) Payoffs. Riders earn per unit; MMs earn per unit absorbed. Own impact cancels for both groups.
(ii) Persistence depends only on . At , , strictly increasing in . MMs shorten stress, and when stress is shorter than with herds alone. MMs dampen the start of a trend but not an established one, since is unaffected.
(iii) Complements within groups, substitutes across groups. is increasing in and decreasing in ; is increasing in and decreasing in .
(iv) Corner equilibria. is an equilibrium iff ; iff ; iff ; and iff . An equilibrium with interior requires , and one with interior requires ; both are unstable within their own group under sign-preserving dynamics.
*(v) Liquidity collapse and hysteresis. Let and , so riders are always active, and define and by
assuming both exist. Since is strictly increasing, : the hysteresis band is exactly as wide as the MMs’ absorption capacity. For the full-liquidity equilibrium exists and is stable. At it disappears, MMs exit, and jumps upward by
For both and are stable, so once liquidity is lost it returns only when rider capital falls below .
Proof. (i) Riders follow the proof of Lemma 5.1 with replaced by ; the terms cancel. An MM holds from midpoint to midpoint , and the same computation gives gross payoff per unit, to which the spread is added.
(ii) This is the proof of Theorem 5.2(iv) with .
(iii) is strictly increasing in .
(iv) Each condition says that every active group earns at least 0 and every inactive group at most 0. At an interior point, a group’s payoff is increasing in its own share, so a small deviation grows.
(v) For , , and by assumption it exceeds , so both groups earn strictly positive payoffs at the corner, which is therefore stable. For , for every , since is increasing in ; so MMs exit, and jumps from to , and is the difference in (ii). For , , so is a strict, stable equilibrium. On both corners are strict.
Corollary 6.2 (Maker–taker pricing). Corollary 5.3 said no uniform fee induces leaning. With market makers, a direction-blind maker rebate that raises does. Raising raises and , which widens the range of rider capital over which full liquidity is an equilibrium and can be recovered after a collapse. A taker fee removes riders, which sets and minimizes .
Proof. is strictly increasing, so and solve at points that increase with . The taker-fee claim follows from (iv) and (ii).
What changes with market makers. Stabilizing behavior now arises in equilibrium: leaning is still dominated for traders who pay to trade, but profitable for MMs whenever , so what separates a stabilizer from a non-participant is who pays the spread. Persistence is set by one index, , so each unit of absorption offsets one unit of rider capital, and MMs alone ( in the example of Figure 1b) cut expected down-runs from 2.62 to 2.00 periods. The market becomes fragile: tipping now happens in liquidity, with a discontinuous jump in stress persistence and hysteresis, as in liquidity spirals [4] and the Flash Crash [12]. And the effective instrument is a maker–taker schedule, which is behavior-specific, rather than a uniform fee.
Connection to empirical evidence
The companion study [2] trains the three archetypes of § 2 with DQN and REINFORCE on hourly bars for AAVE, ETC, and SUI from March 2024 to February 2026, as price-takers in separate environments, and attributes the learned policies with SHAP. Each policy relies on the signals its objective implies: momentum for the herd-follower, drawdown and cross-asset returns for the rational agent, and short-term volatility and position for the manipulator. Three of its findings bear on the theory.
Herders: brief own stress, longer market stress. The study finds that herd-follower strategies have the most frequent but shortest stress episodes. This is Proposition 4.6: with exposure steps of , a herd-follower’s own episodes last at most two periods on every path. Theorem 4.2(ii) shows that the same agents lengthen market stress, an effect that strategy-level metrics cannot detect.
Rational agents: stabilizing alone, not together. The study’s rational agents, trading alone, show the most stable risk-adjusted returns, and its manipulators the rarest but most persistent stress. Theorem 5.2 shows that a return-maximizing agent sharing a pool with herds never leans against them and, above a computable participation threshold, learns to ride them, which is the front-running behavior the manipulator archetype proxies, and which lengthens stress. Reading rational agents as stabilizers is therefore valid only where something else supplies negative feedback, such as market makers (Theorem 6.1) or arbitrage to an external price.
What market-level data show. In the study’s data, the two-hour variance ratio is 1.010–1.017, statistically indistinguishable from one. Therefore, by Proposition 4.4, the full sample shows no net feedback at the hourly horizon. Stress nonetheless clusters: runs of hours with returns below −2% average 1.09–1.13 hours, against 1.02–1.04 under independence, and absolute returns have autocorrelation 0.22–0.27. This persistence comes from volatility clustering, which Assumption 1 excludes. Empirical tests should therefore apply the serial statistics of § 4 conditionally, relating variation in sign persistence to archetype intensity after controlling for volatility rather than to full-sample levels. The theory yields three such predictions: herd intensity raises market stress persistence without raising stress frequency at fixed gross flow (Theorem 4.2); the sign of tracks the net index (Proposition 4.4, Theorem 6.1); and market-maker participation falls discontinuously as trend-chasing capital rises and recovers only after a larger decline (Theorem 6.1(v)).
Implications for regulation
Corollary 4.3 shows that limits on volatility, value-at-risk, loss frequency, or turnover cannot distinguish stabilizing from destabilizing populations of agents. Theorem 4.5 shows that the distinction is cheap to read from serial statistics at a rate no method can beat. Corollaries 5.3 and 6.2 show that a uniform fee can deter riders but never induce stabilizing trades, while a maker rebate does and widens the range of trend-chasing capital over which liquidity survives and can be restored. Instruments should act on feedback rather than on one-period statistics.
Conclusion
Defining trading agents by their objectives traces a path from rewards to order flow to market stress. On that path, the one-period statistics regulators watch are provably blind, while serial statistics separate stabilizing from destabilizing feedback at an optimal rate. When agents trade together, profit-seeking agents amplify herds, and market makers that are paid to absorb flow restore stability at the cost of fragile, hysteretic liquidity. Instruments that act on feedback, such as maker–taker pricing, are the effective response.
Numerical verification
Every closed form in § 4–§ 6 was checked by Monte Carlo simulation with Gaussian news and . Table 1 reports simulated values with theoretical values in parentheses.
| Result and setting | Simulated (theory) |
| Thm. 4.2, , , periods | 1.6398 (1.6400); 0.2282 (0.2283); 1.7247 (1.7263); 0.7883 (0.7881) |
| Thm. 4.2, , | 1.6398 (1.6400); 0.2287 (0.2283); 1.0372 (1.0373); 0.2117 (0.2119) |
| Prop. 4.6, , periods | longest own-stress episode 2 (); mean 1.08 |
| Lemma 5.1, , , periods | rider gross 0.1847 (0.1854); 0.6181 (0.6179); 2.6164 (2.6172) |
| 0.2102 (0.2108); 0.7026 (0.7027); 3.3608 (3.3638) | |
| 0.2155 (0.2164); 0.7215 (0.7214); 3.5860 (3.5891) | |
| 0.1173 (0.1163); 0.3874 (0.3877); 1.6295 (1.6333) | |
| Thm. 5.2(iii), , , softmax policy gradient | from shares : ; from : ; basin boundary |
| Thm. 6.1(i)–(ii), , | rider gross 0.1699 (0.1701); MM gross (); 2.308 (2.309) |
| , | 0.1988 (0.1995); (); 2.996 (2.984) |
| , | 0.2168 (0.2157); (); 3.571 (3.558) |
| Thm. 6.1(v), , , | , ; jumps ; learning dynamics lose MMs at on the way up and recover them at on the way down |
Table 1. Simulated versus theoretical values. Payoffs are gross of fees, per unit per period.
Simulation design. § 4 simulates (3) directly. § 5–§ 6 simulate the batch model with impact , fill each order at the batch midpoint, and compute payoffs on round trips from batch to . For Theorem 5.2(iii), each of 400 iterations estimates the three payoffs from a fresh simulation of periods at the current shares and applies the expected softmax policy-gradient update with step size 4. For Theorem 6.1(v), rider and MM shares follow logistic sign-preserving dynamics on the closed-form payoffs, run to convergence at each while is swept up from 0.05 to 0.8 and back. Learning slows near each threshold, so the simulated collapse and recovery points lie outside the band rather than on its edges, as stability predicts. The code is provided as supplementary material.
References
- [1]Hayden Adams, Noah Zinsmeister, and Dan Robinson. Uniswap v2 core. Whitepaper, 2020.
- [2]Irene Aldridge, Lourenço Belleau-Bowhnik, James Chan, Josh Chang, Kelly Deng, Pranav Dhingra, Xinda Ding, Haodong Fang, Satya Hanumanchi, Carter He, Shuang He, Yadong Hou, Vallab Karanam, Rithvik Kasarla, Tishya Khanna, Jihyun Kim, Sarah Konety, Abhinav Kotta, Anderson Lo, Jake Luther, Omar Mian, Ishan Patwardhan, Matteo Perona, Chenyue Qi, Shashank Ramachandran, Shubham Rateria, Jake Swain, David Thai, Michael Tversky, Susan Wu, Dean Xu, Helenna Yin, and Jiachun Zheng. Agentic AI in automated markets. arXiv:2610.22522, 2026.arxiv.org/abs/2610.22522
- [3]Guillermo Angeris and Tarun Chitra. Improved price oracles: Constant function market makers. In Proceedings of the 2nd ACM Conference on Advances in Financial Technologies (AFT ’20), pages 80–91. ACM, 2020.DOI
- [4]Markus K. Brunnermeier and Lasse Heje Pedersen. Market liquidity and funding liquidity. Review of Financial Studies, 22(6):2201–2238, 2009.
- [5]Eric Budish, Peter Cramton, and John Shim. The high-frequency trading arms race: Frequent batch auctions as a market design response. Quarterly Journal of Economics, 130(4):1547–1621, 2015.
- [6]Emilio Calvano, Giacomo Calzolari, Vincenzo Denicolò, and Sergio Pastorello. Artificial intelligence, algorithmic pricing, and collusion. American Economic Review, 110(10):3267–3297, 2020.DOI
- [7]Philip Daian, Steven Goldfeder, Tyler Kell, Yunqi Li, Xueyuan Zhao, Iddo Bentov, Lorenz Breidenbach, and Ari Juels. Flash boys 2.0: Frontrunning in decentralized exchanges, miner extractable value, and consensus instability. In Proceedings of the 2020 IEEE Symposium on Security and Privacy, pages 910–927, 2020.DOI
- [8]J. Bradford De Long, Andrei Shleifer, Lawrence H. Summers, and Robert J. Waldmann. Positive feedback investment strategies and destabilizing rational speculation. Journal of Finance, 45(2):379–395, 1990.
- [9]Lawrence R. Glosten and Paul R. Milgrom. Bid, ask and transaction prices in a specialist market with heterogeneously informed traders. Journal of Financial Economics, 14(1):71–100, 1985.DOI
- [10]Sanford J. Grossman and Merton H. Miller. Liquidity and market structure. Journal of Finance, 43(3):617–633, 1988.
- [11]Wassily Hoeffding. Probability inequalities for sums of bounded random variables. Journal of the American Statistical Association, 58(301):13–30, 1963.
- [12]Andrei Kirilenko, Albert S. Kyle, Mehrdad Samadi, and Tugkan Tuzun. The flash crash: High-frequency trading in an electronic market. Journal of Finance, 72(3):967–998, 2017.
- [13]Albert S. Kyle. Continuous auctions and insider trading. Econometrica, 53(6):1315–1335, 1985.DOI
- [14]Blake LeBaron. Agent-based computational finance. In Leigh Tesfatsion and Kenneth L. Judd, editors, Handbook of Computational Economics, volume 2, pages 1187–1233. Elsevier, 2006.
- [15]Andrew W. Lo and A. Craig MacKinlay. Stock market prices do not follow random walks: Evidence from a simple specification test. Review of Financial Studies, 1(1):41–66, 1988.
- [16]Paul Milgrom and John Roberts. Rationalizability, learning, and equilibrium in games with strategic complementarities. Econometrica, 58(6):1255–1277, 1990.DOI
- [17]Jason Milionis, Ciamac C. Moallemi, Tim Roughgarden, and Anthony Lee Zhang. Automated market making and loss-versus-rebalancing. arXiv:2208.06046, 2022.arxiv.org/abs/2208.06046
- [18]William H. Sandholm. Population Games and Evolutionary Dynamics. MIT Press, Cambridge, MA, 2010.
- [19]Alexandre B. Tsybakov. Introduction to Nonparametric Estimation. Springer Series in Statistics. Springer, New York, 2009.
