Skip to content
Research noteBT-2026-0236

Faster blocks paid arbitrageurs and thinned the pools

A study of Base's tenfold block-speed upgrade finds lower priority-fee bids, higher arbitrage profit, faster LPs and shallower pools.

2 minAlgorithmic & AI TradingFresh · 11 Nov

A paper posted to arXiv on 10 November by Ruizhe Jia and Shihao Yu asks how fast a blockchain's blocks should be for the trading venues that run on it. On a chain, trades do not execute as they arrive; they wait for the next block, and within each block the sequencer usually sells priority through fees. Block time is therefore a design choice that sets the execution clock of every pool. The authors argue this matters more now that the SEC's conditional exemptions of September let US stocks trade as tokens in automated market maker pools on public chains.

Their model makes three predictions. Faster blocks weaken the priority-fee race between arbitrageurs and shift rents from the block producer to them. An active liquidity provider can rebalance and avoid toxic flow sooner, cream-skimming the trades that do not hurt it, while passive capital loses more per unit of time. And pool depth peaks at an intermediate block speed, falling once blocks get fast enough.

They test this on Base's Flashblocks upgrade of 16 July 2025, which cut the effective block interval from two seconds to 200 milliseconds while keeping the priority gas auction, with Arbitrum, whose block time did not change, as the comparison chain. Gas bids fell most on large arbitrage trades, by up to 73% in the top size decile. Gross profit extracted from pools fell by about 29%, yet arbitrageurs' net profit rose by 40% because the gas saving outweighed it. The active provider's rebalancing time fell by about 66%, and its share of exposure to toxic trades fell 8.93 percentage points more than the pool's.

Pools ended up shallower and more concentrated. The largest provider's share of liquidity rose by 4.20 percentage points, and the effective spread on trades of about 1 ETH widened by 0.07 basis points, a 47% increase, with larger rises for bigger trades, even though pool prices tracked outside prices more closely. In the authors' reading, faster blocks moved rents away from the sequencer toward arbitrageurs and the active provider, at the expense of passive liquidity.

Retold from arXiv. This is a summary in our own words; follow the link for the original reporting.

Questions

What happened to arbitrageurs after Base Flashblocks?
According to the paper, gas bids fell, most on large trades, by up to 73% in the top size decile. Gross profit taken from pools fell by about 29% on Base, but net profit rose by 40% because the gas saving outweighed it.
Are faster blocks always better for DEX liquidity?
Not in this model. Pool depth peaks at an intermediate block speed; beyond a threshold an active liquidity provider cream-skims and passive capital shrinks, so total liquidity falls.

Read next

Across the network

Desks that share a zone with this one on the BITBRIEF coverage map.

Terms defined