Can small traders profit from crypto price gaps, or do fast robots take everything?

"Compared to counterfactual trades executed on V3, all CoWSwap trades exhibit significantly positive welfare, with trades smaller than 1 ETH exceeding 500 bps."

"which is expected to be realized over three years if the trade is held to maturity"

"When the underlying value of the assets changes, the first arbitrageur who trades with the CFAMM will bring its marginal price (i.e., the price of an arbitrarily small trade) in line with the equilibrium price."

"The average values for green and orange are 41.7 Gwei and 32.5 Gwei correspondingly."

"This adjustment reflects a delay in executing DeFi Alpha arbitrage opportunities as a result of the proliferation of digital asset treasury companies and the consolidation of digital asset prices in the latter half of 2025, which has temporarily reduced arbitrage activity and compressed available trading spreads."

"The total sum of DeFi Alpha trades for 2025 thus far is $27 million."

"sandwich attacks in which an attacker front-runs a victim’s swap with a swap in the same direction and then back-runs it with another swap in the opposite direction"

"With the proliferation of digital asset treasury companies in the latter half of 2025, the Company experienced delays in its pipeline of arbitrage opportunities and compressed arbitrage profits."

"The main result is that when trading happens in continuous time, arbitrageurs compete on speed. When traders are instead batched (that is, collected over a given time interval and then settled), arbitrageurs compete on price"

"Of all blocks since block 3,875,490 (the first block we observed with a pure revenue OO fee from decentralized exchange arbitrage), approximately 3.6% contain at least one pure revenue arbitrage transaction."

"Consequently, 46 out of 88 se-"

"Of these blocks, 17,897 blocks contain pure revenue OO profits that are < 1% of the total stealable fees, confirming our earlier conclusions that the most frequent pure revenue arbitrage opportunities are for small price differences in the course of normal trading across exchanges."

"yieldingasuccessrateof27.3%(9,955outof36,492)."

"Nonetheless, many blocks do contain substantial pure revenue fees. 36,860 observed blocks contain pure revenue fees that exceed 20% of total stealable fees"

"cyclicarbitrageursis21,360ETH(upperbound:25,742ETH),"

"a long-only, realistically-costed trader has no detectable edge that price information can supply"

"which has successfully generated approximately C$113.8 million (US$83.4 million) thus far in Q2 from low-risk arbitrage trades, marking a robust start for this new venture."

"generated a one-time return of approximately C$30.2 million (US$22 million) from an arbitrage trade."

"Price-based timing shows no detectable edge at any horizon (intraday, swing, cross-sectional, cross-chain): real data performs no better than, and usually worse than, its own shuffle"

"generated C$132.1 million (US$96.8 million) in revenue through systematic arbitrage trading"

"The strategies practitioners actually use fail too, net of cost (market-neutral hedging, launch timing, liquidity provision, cross-pool arbitrage, regime conditioning, portfolio aggregation, and order flow)"

"the median strategy loses even at zero cost"

"its predictability is concentrated in the lower tail (it flags which coins crash, not which rise)"