Optimal Trading Speed on Blockchain
Listen
Radio episode about this paper
Transcript
Introduction to the show: ident: Quantitative Finance. The day in quantitative finance: models of markets, risk and prices.
Ruth: We're looking at the paper: "Optimal Trading Speed on Blockchain".
Warren: Here's the gist of the paper. On blockchain-based exchanges, faster blocks fundamentally alter how priority is allocated, shifting competition from a priority auction toward a first come, first served mechanism for arbitrageurs.
Ruth: This change reshapes the competitive landscape: faster blocks weaken priority-fee competition among arbitrageurs by shortening the time between transactions arriving in the same block, moving outcomes closer to first come, first served.
Warren: Consequently, this shift redistributes rents from the sequencer to the arbitrageurs.
Ruth: For liquidity providers (LPs), faster blocks introduce new opportunities and risks.
Title and authors: Ruth: Let's talk about the title and who wrote this thing, "Optimal Trading Speed on Blockchain." It’s pretty straightforward, but it sets up a deep look into how speed matters in digital finance.
Warren: It suggests that optimizing the speed of those blocks isn't just a technical detail; it directly impacts market structure and profitability for everyone involved.
Iris: The authors are Ruizhe Jia and Shihao Yu, and they’ve built this model based on the existing landscape of tokenized stocks trading on public blockchains like Base.
Ruth: They are using real-world data from things like the Base Flashblocks upgrade to test these ideas, showing that faster blocks do indeed weaken priority fee competition among arbitrageurs.
Warren: That means the time between transactions arriving in a single block gets shorter, which naturally favors those who can act first over those who have to wait for a priority slot.
The paper's summary: Ruth: Now, let's look at what the paper actually says about the results. They summarize that faster blocks weaken the priority fee competition among arbitrageurs and shift rents from the block producer to them.
Warren: So, in simple terms, it means faster blocks reduce those competition fees for arbitrageurs because they trade closer together within the same block, making it more like a first come first served situation.
Iris: And this leads to a big change in who benefits—the rents move from the person controlling the sequencing to the arbitrageurs themselves.
Ruth: That's right. It also looks at liquidity providers, or LPs, and shows they have new ways to play when blocks are faster. They can strategically withdraw capital ahead of price trades while still sharing fees from other things.
Warren: They can essentially cream-skim passive providers by timing their withdrawal optimally so it executes before an arbitrage trade happens, protecting their capital and letting them earn more overall.
The paper's improvements: Ruth: The authors suggest several ways to make this model better. They propose modeling the competition as either a "clash" or a "race" to better predict how those rents are actually distributed between the sequencer and the arbitrageurs.
Warren: That would be useful because it lets us predict which type of competition dominates depending on how fast those blocks are running relative to message speed.
Iris: They also refine the active liquidity management strategy for automated market makers by dynamically determining when to pull capital out or put it back in based directly on the block speed.
Ruth: Specifically, they suggest using a threshold, like that B sub two we talked about, which marks the point where the gains from withdrawing capital start to offset the fees they give up <ref:2611.13616#pg1>.
Warren: That sounds like a practical tool for anyone running an automated market maker to decide when to act on their liquidity position.
Conclusion: Ruth: To wrap up, this paper on "Optimal Trading Speed on Blockchain" shows that block speed isn't just a technical parameter; it jointly determines the allocation of rents across arbitrageurs, liquidity providers, and the sequencer.
Warren: The main implication is that faster blocks lead to higher net profit for arbitrageurs and allow active LPs to manage their risk better by cream-skimming toxic flow.
Iris: And they found that while price efficiency improves unambiguously, the effective spread widens by about zero point zero seven basis points on Base relative to Arbitrum when block speeds get higher.
Ruth: So, we see a trade-off: faster blocks help price discovery but can thin out the liquidity pool and widen those effective spreads at certain speeds.
Warren: That B sub d threshold where liquidity peaks is key—it tells us exactly when that tension between price discovery and pool depth starts to bite.
Iris: We also see that in an operating equilibrium with positive passive capital, the mean inverse depth equals one over p naught, which gives us a concrete number for how deep the pools need to be to maintain balance.
Ruth: So, as we look at this paper on "Optimal Trading Speed on Blockchain," it’s not just about speed; it’s about understanding how speed reshapes incentives for everyone in this new financial infrastructure.
Ruizhe Jia, Shihao Yu
Stanford University · Singapore Management University
q-fin.TR, econ.GN, q-fin.EC
Submitted: 2026-11-10
Updated: 2026-11-12
License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
Importance score: 91/100
The gist: On blockchain-based exchanges, faster blocks fundamentally alter how priority is allocated, shifting competition from a priority auction toward a first come, first served mechanism for arbitrageurs.
Key concepts
- Arbitrage Competition Shift
- Faster blocks change how arbitrageurs compete for profits. Instead of a priority auction, the competition becomes more like first-come, first-served because transactions arrive faster within the same block. This reduces the advantage of paying higher fees to get priority.
- Active Liquidity Provider Strategy
- Active liquidity providers can use faster blocks to time their capital moves perfectly. They can withdraw funds before arbitrage trades occur, protecting capital from loss, and share fees from non-arbitrage trades. This strategy is optimized by the block speed.
- Equilibrium Block Speed (Bd)
- The model finds a unique optimal block speed, Bd, where pool depth reaches its maximum. If the network goes faster than Bd, liquidity depth falls and spreads widen even if external prices are tracked closely. This indicates a tension between price discovery and pool quality.
- Cream-Skimming
- Active LPs can benefit by avoiding 'toxic flow'—bad trades or flows that might hurt their capital. Faster blocks increase the chance that an LP can withdraw before these trades execute, allowing them to keep a higher share of profitable, non-toxic liquidity.
Terminology
Summary
On blockchain-based exchanges, faster blocks fundamentally alter how priority is allocated, shifting competition from a priority auction toward a first come, first served mechanism for arbitrageurs. p. 4 This change reshapes the competitive landscape: faster blocks weaken priority-fee competition among arbitrageurs by shortening the time between transactions arriving in the same block, moving outcomes closer to first come, first served. p. 4 Consequently, this shift redistributes rents from the sequencer to the arbitrageurs. p. 1
For liquidity providers (LPs), faster blocks introduce new opportunities and risks. p. 3 An active liquidity provider can strategically withdraw capital ahead of stale-price arbitrage trades while simultaneously sharing fees from non-arbitrage trades, effectively cream-skimming passive providers. p. 4 This strategy is supported by the fact that faster blocks allow the active LP to time liquidity provision optimally—her withdrawal is more likely to execute before an arbitrage trade, protecting her capital from loss, and her redeposit can occur sooner once prices align. p. 5
The paper models how block speed dictates rent allocation and liquidity management. p. 9 It establishes that equilibrium liquidity peaks at a unique interior block speed (B d), denoted as B d < B 2. p. 1 Beyond this threshold, faster blocks cause pool depth to fall, effective spreads to widen, and volume to decline, even while pool prices track external market prices more closely. p. 1 This phenomenon is characterized by a tension: faster blocks improve price discovery but can lead to a thinner pool. p. 8
The analysis also details the optimal trading strategies for an active LP based on her fee-to-loss ratio (b f) relative to various cutoffs (theta 0, theta EL, theta LN), determining whether she should always deposit, withdraw early, withdraw late, or permanently exit. Furthermore, the model derives that the common global maximum of pool depth occurs at a specific block rate B d, and the mean inverse depth equals 1/p 0 in an operating equilibrium with positive passive capital. p. 8
Empirical evidence from upgrades like Base Flashblocks confirms these predictions: faster blocks reduce gas-fee competition, increase arbitrageurs' net profit, allow active LPs to cream-skim by avoiding toxic flow while earning more, and ultimately lead to a thinner pool structure. The paper contributes a tractable equilibrium model where block speed jointly determines the allocation of rents across arbitrageurs, liquidity providers, and the sequencer. p. 9
Green — the paper says it, and the page is printed. Red — no passage carries it, so treat that sentence as unverified. The same check runs on your own text: check a summary against its source, $5.
Improvements for AI systems
-
Better prediction of arbitrage rent distribution based on block speed by modeling competition as either a
clash
or arace.
The improved system can predict which type of competition dominates based on current block speed relative to message speed, thereby accurately estimating the sequencer's share versus the arbitrageurs' combined share. -
Enhanced active liquidity management strategy for automated market makers by dynamically determining when to withdraw capital and when to redeposit based on block speed. The improved system can predict the optimal threshold for withdrawal, such as the threshold B2, which marks the point where
the gains from withdrawing offset the fees forgone.
-
More accurate estimation of pool liquidity dynamics by predicting how block speed affects total liquidity depth. The improved system can predict that "Depth has a unique global maximum at an intermediate block rate Bd < B2," allowing it to forecast when faster blocks will cause depth to fall and liquidity to concentrate.
-
Improved modeling of active LP behavior via cream-skimming by tracking the shift in exposure from toxic flow. The improved system can quantify how
Faster blocks give the active LP’s withdrawal a greater chance to execute before the arbitrage trade
and predict that hertoxic-exposure share falls from 62.9% to 47.5%
on Base after Flashblocks. -
Improved market quality prediction by forecasting price efficiency and effective spread changes under different block speeds. The improved system can predict that
Price efficiency improves unambiguously,
while simultaneously forecasting thatthe effective spread widens by 0.07 bps on Base relative to Arbitrum
at higher block speeds, even as pool prices track the external reference price more closely.
Sources
Related papers
- AIDA: An AI Outage Registry with Evidence from On-Chain MEV
- MLP-Initialized Reinforcement Learning for Exit Timing in SPX Short-Straddle Portfolios
- Beyond Anonymous Order Flow: Wallet Identity as a Feature for Short-Horizon Return Prediction
- Combinatorial Arbitrage in Prediction Markets
- TradeMech: A Method to Multilaterally Net Trades Without Altering Counterparty Exposure
- Why financial prices exhibit Brownian motion despite predictable order flow: An exactly solvable nonlinear price impact model