39 cross-exchange arbitrage opportunities detected on May 24, 2026. Top spread: MEW at 17.49% between Binance Futures and Hyperliquid. Full breakdown with profit calculations, risk flags, and tomorrow's setup for professional arb desks.
May 24, 2026 delivered a notably active session for cross-exchange arbitrageurs. The scanner closed out the day with 39 confirmed spread events — a number that sits well above the quiet baseline most desks see on a typical mid-week session. If you were at your terminal and your alerts were live, you had opportunities firing across nearly every tier of asset: micro-caps like MEW printing absurd double-digit spreads, mid-tier names like ARKM offering a controlled but very executable 16%-plus gap, and a handful of smaller DeFi tokens filling the middle of the board.
The headline number is MEW's 17.49% spread between Binance Futures and Hyperliquid — a gap wide enough to make any serious arb trader stop what they're doing and stare. ARKM followed at 16.34% between Coinbase and Binance, which is the kind of spread that's rare on two Tier-1 venues simultaneously. AGT showed up twice on the board — once at 10.67% and again at 6.12%, suggesting persistent structural inefficiency in how that token is priced across Gate, Binance Futures, and KuCoin. In total, 10 opportunities exceeded the 6% threshold, which is generally the floor where net-of-fees profit becomes meaningful for anything but the most capital-efficient operations.
The session had a clear geographic footprint: US-regulated venues (Coinbase) versus offshore derivatives desks (Binance Futures, OKX, Hyperliquid, Gate) accounted for the majority of the spread pairs. This is consistent with the broader 2026 trend where onshore-offshore price fragmentation continues to widen during periods of thin intraday liquidity. For professional arb desks — those with pre-funded accounts on both sides and low-latency execution — today was a gift. For retail participants trying to manually execute cross-exchange arb, most of these windows were too fast or too fee-sensitive to capture cleanly. Let's go through what actually mattered.
Buy side: Binance Futures at $0.000536. Sell side: Hyperliquid at $0.000630. Gross spread: $0.000094 per token, or 17.49% of the buy price. MEW — the cat-themed micro-cap meme token — has been a persistent arb target in 2026 precisely because it trades in vastly different liquidity environments depending on which venue you're on. Binance Futures sees deep perpetual funding flows that tend to suppress the mark price during low-volume periods, while Hyperliquid's order book on MEW is thinner and more reactive to spot momentum on Solana-adjacent liquidity. The combination creates these outsized gaps.
Risk factors are significant here. MEW's absolute price ($0.000536) means you're moving enormous token quantities to generate any meaningful dollar notional. A $10,000 position requires approximately 18.7 million MEW tokens on the buy side — and Hyperliquid's order book depth on MEW at the $0.000630 level is almost certainly not absorbing that without meaningful slippage. Realistically, the executable size was probably in the $2,000–$5,000 notional range before the spread began compressing. Window duration on a spread this wide in a token this illiquid is typically 2–8 minutes before the gap closes or the ask runs away. Verdict: highly executable for small size ($1K–$3K notional) with pre-funded accounts on both venues. Not scalable to institutional size.
Buy side: Coinbase at $0.123000. Sell side: Binance at $0.143100. Gross spread: $0.020100 per token, or 16.34%. This is the most interesting opportunity on the board today, and arguably the most anomalous. Coinbase and Binance are both deep-liquidity Tier-1 venues, and a 16%+ spread on a token with ARKM's market cap and trading history is not something you expect to persist for long. ARKM — the Arkham Intelligence token — has seen elevated volatility around data releases and token unlock schedules in 2026, and it's possible this spread materialized around one such event where Coinbase's orderbook lagged Binance's price discovery.
From an execution standpoint, this is the cleanest opportunity on the board. Both venues have deep liquidity, withdrawal infrastructure is well-tested, and the token is KYC-gated on Coinbase which means transfer times are predictable. The main risk here is transfer time if you don't have pre-funded accounts: Coinbase ACH/crypto withdrawals to Binance can take 10–30 minutes depending on network congestion, and a 16% spread on liquid venues does not survive 30 minutes. With pre-funded accounts on both sides and a delta-neutral setup (long spot Coinbase, short perp Binance), this was executable in seconds and the risk-adjusted return was exceptional. Window estimate: 4–12 minutes based on historical ARKM spread compression patterns. One of the best legitimate arb setups seen this week.
Buy side: Gate Futures at $0.019946. Sell side: Binance Futures at $0.020875. Gross spread: $0.000929 per token, or 10.67%. AGT appeared twice in today's scan — here on the Gate/Binance Futures pair, and later at 6.12% between Binance Futures and KuCoin. The fact that the same token shows persistent multi-venue spreads across the day is a structural signal worth flagging. This typically indicates either fragmented market maker coverage (different MMs quoting different venues with minimal cross-venue inventory management) or a token with insufficient unified liquidity to keep prices sticky across exchanges.
Gate Futures as the buy venue introduces specific risks: Gate's futures infrastructure has historically had wider bid-ask spreads on low-cap assets, meaning your actual fill on a market or aggressive limit order may be worse than the quoted mid. The 10.67% gross spread gives you room to absorb that slippage, but you need to model it carefully. Withdrawal from Gate to Binance (or vice versa if you're hedging on-chain) adds another 15–30 minute delay risk. Best execution here was probably futures-to-futures basis trading — short the Binance Futures contract simultaneously with the Gate buy — which eliminates transfer risk entirely. Verdict: executable for delta-neutral basis traders with accounts on both derivatives venues.
Buy side: Binance at $0.029800. Sell side: Coinbase at $0.032490. Gross spread: $0.002690 per token, or 9.50%. The AI token (ticker: AI) is another asset that has benefited from narrative-driven retail flows in 2026's AI-sector crypto meta, and retail-driven tokens are textbook arb targets because retail doesn't arb — they chase price. Coinbase's premium here over Binance ($0.032490 vs $0.029800) is consistent with US retail demand spikes, where Coinbase orderbooks temporarily front-run global price discovery.
This is a spot-to-spot arb between two top-tier venues, which means transfer times become the critical variable. Sending AI tokens from Binance to Coinbase for same-session selling requires: initiating withdrawal on Binance, waiting for network confirmations (typically 2–5 minutes for a fast L1/L2, longer on congested networks), and receiving on Coinbase before the spread collapses. Realistically, you need pre-funded Coinbase balances to sell into the premium immediately while simultaneously buying on Binance to replenish your position. With that setup, this was very clean. Without it, the window was probably too short. Coinbase has strong liquidity depth on mid-cap tokens like AI, so slippage was manageable up to $20,000–$50,000 notional.
Buy side: Bitunix at $0.026613. Sell side: Gate Futures at $0.028580. Gross spread: $0.001967 per token, or 7.39%. DYM (Dymension) has been a modular blockchain narrative play throughout 2026, and Bitunix as the buy venue is the main story here. Bitunix is a smaller-tier derivatives exchange with lower liquidity, which is precisely why the price lagged Gate Futures by 7.39%. The Bitunix order book was likely sitting on stale quotes — market makers on smaller venues refresh more slowly, and fast-moving markets leave those stale bids and asks behind.
The risk profile on this one is elevated. Bitunix carries counterparty risk that Binance and Coinbase do not. Withdrawal reliability, KYC infrastructure, and operational risk are all factors you'd need to stress-test before deploying capital there. For desks that have already vetted Bitunix and have active accounts, the 7.39% spread was absolutely worth pursuing. For desks encountering Bitunix for the first time, the onboarding time alone would have killed the opportunity. This is a good reminder that arb preparedness — having funded accounts across a wide venue universe — is as important as the detection algorithm itself.
Today's data reveals several recurring exchange-pair patterns that professional desks should be monitoring on a structural basis, not just as one-off alerts.
Hyperliquid vs. CEX (Centralized Exchanges): The MEW opportunity at 17.49% is the most extreme example, but the broader pattern is clear — Hyperliquid's on-chain perpetuals frequently deviate from CEX pricing on low-cap and meme-adjacent tokens. Hyperliquid's price discovery mechanism relies on its own oracle and liquidity providers, and when CEX volumes spike or dump on a low-cap token, Hyperliquid's mark price lags. Desks that run continuous Hyperliquid monitoring against Binance Futures and OKX marks on the same tickers are mining a persistent structural edge in 2026.
Coinbase vs. Offshore Venues (Binance, OKX, Gate): Three of the top 10 opportunities today involved Coinbase as one side of the pair — ARKM (16.34%, Coinbase buy), AI (9.50%, Coinbase sell), ME (6.63%, Coinbase sell). The directional pattern is interesting: Coinbase was the premium venue on AI and ME (US retail bid) but the discount venue on ARKM (possible US regulatory overhang or retail panic). The Coinbase premium/discount flip-flop across assets on the same day reflects how sensitive US-listed assets are to domestic narrative flows. Monitoring the Coinbase mid against Binance spot for all shared listings is a high-signal overlay.
Gate Futures as a Persistent Discount Venue: Gate appeared as either buy or sell side on multiple opportunities today — DYM (sell side at premium) and AGT (buy side at discount). This duality isn't contradictory — Gate Futures is a mid-tier venue with inconsistent market-maker coverage across its listed pairs, meaning some tokens are overpriced and others underpriced relative to larger venues at any given moment. The consistent takeaway is that Gate Futures is worth scanning continuously for cross-venue divergence. It's not a structurally cheap or expensive venue — it's a structurally illiquid one, which creates noise that disciplined arb desks can harvest.
OKX as a Mid-Tier Reference: OKX appeared on BSB (sell side, 7.05%) and ME (buy side, 6.63%). OKX generally has tighter spreads and deeper books than Gate but lags Binance and Coinbase in liquidity depth on smaller caps. The BSB gap (Binance Futures buy, OKX sell) is particularly telling — when Binance Futures funding rates diverge, the perp mark can temporarily undershoot spot prices on other venues, and OKX's spot book catches the premium. Funding rate monitoring on Binance perpetuals is a useful leading indicator for these events.
The fundamental tension in cross-exchange arbitrage has always been: do you move fast on small size, or do you build position slowly at risk of spread compression? Today's opportunities illustrate both extremes.
Fast, small: MEW at 17.49% was a speed game. The spread was exceptional but the size ceiling was low — realistically $2,000–$5,000 before slippage ate your edge. The window was short. You needed sub-minute execution. If your system wasn't already watching MEW across Binance Futures and Hyperliquid with alerts configured, you missed it. The gross dollar profit on a $3,000 notional trade at 17.49% spread is approximately $525 before fees — meaningful for a 2-minute window, but you need infrastructure to catch it consistently.
Slow, larger: ARKM at 16.34% between Coinbase and Binance offered a wider execution window — both venues are liquid, price discovery is transparent, and the spread likely persisted for 5–15 minutes. A $25,000 notional position at 16.34% generates $4,085 gross before fees. But the risk is also higher: Coinbase and Binance are both efficient enough that a 16% spread suggests something unusual was happening in price discovery, which means the spread could snap back suddenly or one side of your book could gap against you if the catalyst resolves.
Slippage is the silent killer on both ends. On micro-cap tokens like MEW and AGT, market-impact slippage on position entry and exit can consume 30–50% of the stated gross spread on any meaningful size. On liquid tokens like ARKM and AI, slippage is lower but spread compression (other bots entering the same trade) closes the gap faster. The practical recommendation: size your position so that your market impact is less than 20% of the gross spread. If the spread is 10% and your expected slippage is 3–4% total (both sides combined), you have 6% left to work with before fees. Anything tighter than that and you're gambling on execution, not arbitraging.
Position sizing framework: For spreads above 10% — max $5,000 notional per trade (assume illiquid, fast-closing). For spreads 6–10% — max $15,000–$25,000 notional with pre-funded accounts and sub-30-second execution. For spreads below 6% — only viable for fully automated systems with API execution, pre-funded balances, and sub-5-second latency. Manual execution below 6% is almost never net profitable after fees and time cost.
Let's walk through three representative calculations using today's actual data. These assume standard retail/professional fee tiers — institutional negotiated rates would improve these further.
Minimum spread threshold for profitability: With standard Coinbase (0.60%) plus Binance (0.10%) taker fees, the fee drag alone is 0.70% per side = 1.40% round trip minimum before slippage and withdrawal costs. Add 1–2% for slippage on mid-liquidity assets and you need a minimum 3.5–4.5% gross spread just to break even. Below 5% spread, you're in noise territory for most retail-tier arb setups. Below 3%, only automated systems with maker rebates and zero-fee withdrawal routes are in the money. Today's board — with 10 opportunities all above 6% — was well above that threshold. Most of the top opportunities were genuinely executable with proper infrastructure.
Withdrawal delays are the single largest operational risk in cross-exchange arb, and today's venue mix amplifies that risk significantly. Several specific warnings for active traders:
Based on today's patterns, here's where to focus scanner attention for May 25, 2026:
AGT is the highest-priority watch. A token that shows persistent multi-venue spreads in a single session (10.67% and 6.12% today) will likely continue showing fragmented pricing tomorrow unless a major market maker steps in to unify the order book. Monitor AGT across Gate Futures, Binance Futures, and KuCoin continuously. Set alerts at 5% spread threshold to get early warning before the move peaks.
Hyperliquid vs. Binance Futures on meme-cap tokens: MEW printed a 17.49% spread today, which means the Hyperliquid oracle was meaningfully out of sync with Binance's mark price. That structural lag doesn't disappear overnight. Scan the full Hyperliquid perpetuals universe against Binance Futures equivalents tomorrow — any token with a Hyperliquid OI above $500K and Binance Futures OI is a candidate. Focus on tokens under $0.01 price where the spread in percentage terms amplifies faster.
Coinbase premium watch: AI and ME both traded at Coinbase premiums today (9.50% and 6.63%). If US retail sentiment remains elevated tomorrow — check social volume and Coinbase app ranking metrics — expect the Coinbase premium pattern to persist on similar mid-cap AI-narrative tokens. Scan Coinbase spot versus Binance spot for all shared listings, filter for tokens under $500M market cap, and flag anything above 4% spread.
Best monitoring windows: Cross-exchange arb spreads consistently widen during three time windows — (1) the 30-60 minutes after major US economic data releases (check the economic calendar for May 25), (2) the 15-minute window around Binance futures funding rate resets (every 8 hours at 00:00, 08:00, 16:00 UTC), and (3) low-liquidity overnight hours (02:00–06:00 UTC) when market makers reduce quote sizes. Set your scanner to maximum sensitivity during these windows and reduce polling frequency during the liquid US session midday.
Exchange pairs to add to your monitoring matrix if you haven't already: Hyperliquid vs. Binance Futures (any shared perp), Coinbase spot vs. OKX spot (shared mid-caps), Gate Futures vs. KuCoin spot (fragmented tier-2 assets), Bitunix vs. Gate Futures (stale-quote harvesting on micro-caps). Today's board gave you the map. Tomorrow, you follow it.
39 opportunities. Double-digit spreads on two separate assets. A rare dual Tier-1 gap on ARKM. For a mid-week session with thin volume across the board, that's a productive hunting ground. The edge today wasn't exotic — it was preparation. Pre-funded accounts, continuous cross-venue scanning, and the discipline to size correctly for the liquidity environment. The desks that had their infrastructure ready walked away with clean, compounding returns. The ones that were watching from the sidelines got a good lesson in why readiness isn't optional in this game. See you at the screens tomorrow.
Arbitrage Hunter — May 24, 2026