Designated Market Makers (DMMs)

Updated 10 February 2026

Designated Market Makers (DMMs) are specialized liquidity providers on electronic exchanges that post continuous two-sided quotes to maintain market stability and narrow spreads.
They are evaluated using agent-based limit order book models and econometric methods like GLM and GAMLSS to measure liquidity replenishment speed and price efficiency.
Incentive schemes such as maker rebates and fee reductions motivate DMMs to meet strict contractual obligations, balancing competition and resilience, especially under market stress.

Designated Market Makers (DMMs) are specialized liquidity providers on contemporary electronic order-driven exchanges. In distinction to anonymous limit order traders, DMMs operate under explicit contractual obligations to continuously supply both buy and sell liquidity within prescribed maximum spreads and minimum sizes. In compensation, exchanges offer DMMs fee reductions, rebates, or analogous incentives. The dual role of DMMs is to ensure the resilience of core market quality metrics—including spread, depth, fulfillment, and price efficiency—and to act as buffers during sudden imbalances between supply and demand (Zhou, 2024, Bellia et al., 2 Feb 2026, Panayi et al., 2015).

1. Institutional Roles, Obligations, and Incentives

DMMs’ primary contractual duties include:

Posting continuous two-sided quotes: DMMs must provide limit orders on both sides of the book for a minimal fraction of trading hours (e.g., active on both bid and ask ≥95% of continuous trading time or ≥90% of the trading day depending on venue) (Bellia et al., 2 Feb 2026, Panayi et al., 2015).
Maximum spread constraints: The quoted spread is bounded (e.g., not exceeding 2.5% for the most liquid stocks on Xetra).
Minimum displayed size: Obliged to quote at or above specified volumes (e.g., at least €5,000 at inside quotes; up to €20,000 for top-tier assets) (Panayi et al., 2015).
Replenishment obligations: While traditional models focus on “time presence,” recent research argues that DMMs should also be required to replenish liquidity with a minimum speed after depletions, best measured by the “Threshold Exceedance Duration” (TED), which quantifies the time required for the order book’s liquidity (e.g., spread) to return below a critical threshold following an adverse shock (Panayi et al., 2015).

Incentive schemes typically comprise:

Maker rebates: For instance, up to 26 bps per traded volume, with empirical studies often citing around 20 bps as an optimal midpoint (Zhou, 2024).
Fee waivers or reductions for liquidity provision: DMMs might be charged lower or zero fees on passively executed trades (e.g., 0.30 bps vs. 0.55 bps for liquidity taking under the SLP program) (Bellia et al., 2 Feb 2026).
Additional privileges: In rare cases, special protections during extreme volatility events may be granted (Panayi et al., 2015).

2. Modeling DMMs: Agent-Based and Econometric Approaches

Simulation-based and empirical econometric approaches provide comprehensive frameworks for analyzing DMM impact and behavior:

Agent-Based Limit Order Book (LOB) Models: DMMs are represented as agents who:
- Post both bid and ask quotes at each decision point, using a risk-adjusted reservation price (Avellaneda–Stoikov) based on mid-price, their own inventory $q_t$ , intraday volatility estimate $\sigma^2$ , and time-horizon $T_h$ :
  
  $r_t^{ask} = S_t + (1-2q_t)\frac{\sigma^2}{2}T_h, \quad r_t^{bid} = S_t - (1-2q_t)\frac{\sigma^2}{2}T_h$
  
  Limit prices $P_t^{ask}$ and $P_t^{bid}$ are then shifted by a tick size/aggresiveness parameter $\delta$ (Zhou, 2024).
- Coexist with several trader types: informed (future price knowledge), uninformed (symmetric), and momentum-based agents.
Liquidity-Supply Metrics: Econometric studies measure net liquidity provision and its time evolution using trading imbalances (TI), together with splits between aggressive and passive imbalance, executing multivariate autoregressive regressions with event windows tied to “drift bursts” (sharp, directional price moves) (Bellia et al., 2 Feb 2026).
Liquidity Replenishment Speed (TED): Regression models, including log-linear OLS, GLM, and GAMLSS with flexible parametric distributions (gamma, Weibull, generalized gamma), model the (log-)duration between liquidity breaches and recovery, conditioned on LOB states and event covariates (Panayi et al., 2015).

3. Impact on Market Liquidity and Quality Metrics

DMMs affect multiple dimensions of market quality:

Bid–ask spreads: The presence and number of DMMs narrow the quoted spread from 0.001712 (one DMM) to 0.001572 (five DMMs) at 20bps incentive, with diminishing returns beyond $N\approx3$ (Zhou, 2024).
Depth at best quotes: Increases with the number of DMMs—rising from 14.61 to 16.57 in simulations—as DMMs compete for order flow.
Fill ratios and execution metrics: Order-fulfillment ratio (OFR) and execution-to-submission ratios improve with moderate increases in provider count and rebate, but elevated rebates above 20bps may lead to wider spreads as DMMs become less competitive, profiting at the expense of takers (Zhou, 2024).
Realized spreads and adverse selection: Realized spreads (RS5, RS10) and adverse selection costs fall with additional DMMs, but improvements plateau for $N > 3$ .
Shock response and market resilience: After a 30% fundamental value shock, price convergence is fastest and least volatile for $N = 3$ , with further increases in DMM count producing reversal or increase in correction times and noise.
Liquidity replenishment speed: Regression-based calibration of TED shows that prior liquidity droughts and initial breach magnitude are strong predictors of replenishment lag; market orders induce faster refill than cancellations (Panayi et al., 2015).

Metric	DMM Increases (to $N=3$ )	Effect of Excess Incentives
Quoted spread	Narrows	Slight re-widening ( $r>20$ bps)
Depth at best quote	Increases	Stable/no consistent effect
Realized spread	Decreases	Slight widening if too many DMMs
Price efficiency	Improves, then plateaus	May reverse at high $N$
TED (replenishment)	Shortens, if obligations	Not imposed in most regimes

4. DMM Behavior in Stress and Extreme Price Movements

Empirical evidence delineates DMM performance under idiosyncratic and systemic market stress:

Single-stock sell-offs (“unsystematic events”): DMMs, especially high-frequency (IB-HFT MMs), supply liquidity as mandated, even at personal trading losses—early-drop $\beta=0.290^{***}$ —providing a stabilizing force (Bellia et al., 2 Feb 2026).
Market-wide shocks (“systematic events”): DMMs often reverse course late in the drop, withdrawing or turning net sellers—late-drop $\beta=-0.373^{**}$ —with non-HFT traders ultimately stepping in to absorb inventory. Passive DMM trading becomes negative, confirming a withdrawal from their liquidity-supplying role (Bellia et al., 2 Feb 2026).
Implications: The “Immediacy-Provision Hypothesis” holds only in isolated asset sell-offs; in systematic downturns, DMMs tend to “lean with the wind,” potentially exacerbating price overshoots. Current DMM contracts (e.g., under SLP) with monthly-averaged obligations lack the granularity to penalize intraday withdrawal in stress, calling for more robust intra-day resilience metrics and obligations.

5. Design and Calibration of DMM Obligations

Traditional DMM contract structures, based on time-in-market and minimum sizes, are increasingly viewed as insufficient:

Uniform “presence” quotas are poorly aligned with time-varying demand: DMMs can avoid quoting during periods of severe illiquidity (e.g., around open or news), yet still meet 90%+ time-based requirements (Panayi et al., 2015).
Incorporation of liquidity-speed metrics: The TED metric is advocated for formal inclusion in DMM obligations; statistical modeling links TED to LOB state variables, enabling exchanges to compute target (mean/quantile) replenishment deadlines for each time-of-day bucket and asset (Panayi et al., 2015):

$\widehat{E}[\tau|x] = \exp(\overline{x}^T\beta)$
Calibration procedure: Historical L2 data, threshold selection ( $c$ ), event detection, and regression fitting (ln-GLM or GAMLSS) yield operational benchmarks. For example, DMMs might be required to replenish to spread ≤ $c$ within 300 ms on average and within 400 ms in 90% of cases during the 10:00–10:15 window.

6. Policy and Practical Implications

Optimal DMM regime design is characterized by the following principles:

Optimal competition: An interior optimum exists around $N=3$ –5 DMMs—neither monopoly nor excessive competition maximizes liquidity, resiliency, and price efficiency (Zhou, 2024).
Rebate calibration: Excessive maker rebates incentivize less competitive quoting, so an intermediate rate ( $\sim$ 20 bps) is empirically optimal.
Resilience-focused obligations: Inclusion of speed-of-replenishment (TED) and granular intra-day obligations is necessary to align DMM quoting with actual liquidity risk, especially at stress junctures (Panayi et al., 2015).
Regulatory adaptation: Introducing minimum “liquidity-at-risk” quotes during stress, refining presence metrics to intra-day windows, and triggering circuit-breakers on directional drift bursts—not merely volatility—are proposed remedies for improving market stability and DMM reliability (Bellia et al., 2 Feb 2026).

A plausible implication is the need for ongoing recalibration of requirements and stress-testing via simulation, as market regimes and algorithmic trading behaviors evolve (Zhou, 2024, Panayi et al., 2015).

7. Controversies and Open Challenges

Persistent controversies in DMM regime efficacy include:

Moral hazard of time-averaged obligations: DMMs can fulfill “presence” yet withdraw when most needed, a fact evidenced in large-scale systematic sell-offs.
Tradeoff between competition and market signal quality: Too many DMMs inject noise, degrade price efficiency, and reduce overall profitability, with a convex market quality response to DMM count (Zhou, 2024).
Insufficient penalties for withdrawal in stress: Current schemes do not penalize brief, strategic lapses in quoting, motivating the call for event-driven and speed-based obligations (Bellia et al., 2 Feb 2026, Panayi et al., 2015).
Complexity of calibration: While flexible models (GAMLSS) offer better fit for TED, empirical gains over simpler log-linear models may be minimal for operational purposes (Panayi et al., 2015).

Ongoing research focuses on designing obligation schedules, penalties, and incentive algorithms that adapt to real-time LOB and volatility state, with agent-based and econometric stress-tests serving as methodological benchmarks (Zhou, 2024, Bellia et al., 2 Feb 2026, Panayi et al., 2015).