🎯 Arb Desk Report
Forty-four discrete arbitrage prints crossed the desk today, and the tape was unusually one-directional: almost every clean spread ran long Binance and short Coinbase, with a handful of second-tier venues — Bitunix, OKX, KuCoin — filling in the gaps at the smaller-cap end of the board. The headline number is T, which didn't just spread once but printed three consecutive snapshots — 29.45%, then 25.38%, then 24.20% — as the gap between Binance's spot book and Coinbase's thinner order book slowly closed. A 29.45% spread on a large-cap-adjacent name like T is not a typical Tuesday; it's the kind of print that either means real dislocation or a stale quote somewhere in the pipe, and both possibilities matter for how you size into it. Below that, ESPORTS and JASMY both cleared 14%, APE printed 12.75%, and a tail of mid-single-digit spreads (LRC, LAB, OPN twice) rounded out the top ten. Volume figures on this feed came back at $0.0M across pump, dump, buy-pressure, and sell-pressure aggregates — treat that as a reporting gap, not as evidence these books were empty, and lean on live order-book depth before committing size to any of the names below.
🏆 Top 5 Arbitrage Opportunities
- T — 29.45% spread, buy Binance $0.004540 / sell Coinbase $0.005680. This is the standout print of the session and it came with a tell: two follow-on snapshots at 25.38% (buy Binance $0.004610, sell Coinbase $0.005780) and 24.20% (buy Binance $0.004710, sell Coinbase $0.005850) show the gap compressing as Binance's ask walked up roughly 3.7% per print while Coinbase's bid drifted higher only slightly. That pattern reads as real capital chasing the arb and closing it in real time, not a single bad tick. Available volume wasn't reported by this feed, which is the single biggest caveat here — T trades with reasonable depth on Binance but Coinbase's book for it is materially thinner, so the executable size was almost certainly capped by the sell leg, not the buy leg. Withdrawal risk is moderate: T settles on multiple networks with sub-5-minute confirmation windows in normal conditions, but if network congestion hit during this window the arb could have closed before funds landed. Our take: executable in principle for anyone with pre-funded balances on both venues (skipping the withdrawal step entirely), but chasing it via on-chain transfer between exchanges was a race against a visibly decaying spread — the 24.20%-to-25.38%-to-29.45% sequence tells you this window was shrinking fast, likely down to single digits within minutes of the last print we captured.
- ESPORTS — 14.09% spread, buy Binance Futures $0.020536 / sell Bitunix $0.023430. This one is structurally different from the T trade because it crosses from a futures venue into a spot-listed pair on a tier-two exchange, which means basis risk stacks on top of the arb — you're not just capturing a price gap, you're also exposed to funding-rate drift on the Binance Futures leg until the position is closed out or hedged. Bitunix is a thinner, less battle-tested venue than Coinbase or OKX, so the sell-side fill quality here is the real question mark; a 14% headline spread on a name pricing at two cents can evaporate against a handful of large market orders on a shallow book. No volume data came through on this print either. Executable take: this is a spread for traders already holding inventory on both venues or running a futures-to-spot basis book, not a cold-start arb — the cross-instrument nature and Bitunix's liquidity profile make it a harder trade to size confidently than the pure spot-to-spot prints.
- JASMY — 14.03% spread, buy Coinbase $0.004420 / sell Coinbase $0.005040. Both legs sit on the same exchange, which is the important flag here: this is very likely a temporal spread — two snapshots of Coinbase's own book taken at different moments — rather than a simultaneous cross-venue dislocation you could actually execute as a hedged pair trade. That doesn't make it worthless information (a 14% intra-exchange move in minutes is a real signal of volatility and possible news flow), but it changes the trade entirely: there's no buy-here-sell-there mechanic, just a fast directional move you'd have needed to already be positioned for. Treat this print as a volatility alert on JASMY rather than a bankable arbitrage opportunity, and check the raw timestamps on your own feed before sizing anything against it.
- APE — 12.75% spread, buy Coinbase $0.149000 / sell Coinbase $0.168000. Same caveat as JASMY: both legs are Coinbase, so this reads as a same-venue price move captured across two ticks rather than a true two-sided arb. APE at 14.9-16.8 cents implies enough absolute price (versus the sub-cent names above) that slippage per dollar of size is more forgiving, which is the one point in its favor if this does turn out to reflect a genuine short-lived book dislocation rather than pure directional drift. Our take: flag for monitoring, not for execution, until the feed confirms two distinct venues were involved.
- LRC — 9.70% spread, buy OKX $0.012738 / sell KuCoin $0.013460. This is the cleanest cross-venue setup in the top five — two genuinely distinct, reasonably liquid exchanges, a sub-10% spread that's large enough to survive typical taker fees on both legs, and a price level that keeps per-share slippage manageable. OKX and KuCoin both support LRC withdrawals on standard networks with typically sub-10-minute processing in normal conditions, which matters if you're not pre-funded on both sides. No volume was reported, but LRC generally carries adequate depth on both venues for mid-size clips. Our take: of everything on this board, this is the one closest to a textbook, executable spot arbitrage — smaller headline number than T, but far less ambiguous about whether it's a real, tradeable two-sided gap.
📊 Exchange Spread Patterns
The dominant corridor today was Binance-to-Coinbase, accounting for five of the ten opportunities logged — all three T prints plus both OPN prints (7.73% buying at $0.060900 and selling at $0.065500, then 7.68% buying at $0.061200 and selling at $0.065900). That's a consistent structural pattern, not noise: Coinbase tends to run a premium on mid-cap and small-cap names relative to Binance's deeper, more arbitrage-efficient spot book, and that premium shows up repeatedly across unrelated tickers on the same day. The OPN pair is a useful confirmation of the T pattern — nearly identical spread size (7.73% vs 7.68%) across two consecutive snapshots, both on the same buy/sell venue pair, both showing Binance's ask and Coinbase's bid drifting up together. That's the signature of a persistent, slow-closing structural gap rather than a one-off glitch. Bitunix showed up on both sides of the book today — as the sell venue for ESPORTS (14.09%) and as the buy venue for LAB (8.12%, against OKX as the sell side at $0.183700) — which is worth flagging on its own: a smaller venue appearing as both the expensive and the cheap leg across different pairs in the same session usually means its book is quoting somewhat independently of the majors, which is exactly the kind of inefficiency arb desks want, but also exactly the kind of venue where fills can be worse than the quoted price. OKX and KuCoin (LRC) round out the cleaner cross-venue setups. No Bybit or Hyperliquid prints came through in today's feed, so no read on those pairs this session.
⚡ Speed vs Size Analysis
The board splits cleanly into two trade profiles. The high-percentage, sub-cent names — T, ESPORTS, OPN — offer eye-catching spread percentages but sit on thin books where the visible top-of-book price is only good for a small clip; push size through and the effective spread you capture degrades fast as you walk the book. These are speed trades: get in with modest size while the gap is fresh, because the 29.45%-to-24.20% decay on T shows exactly how quickly compression happens once other participants notice. The larger-absolute-price names — APE at $0.149-$0.168, LAB at $0.1725-$0.1837 — carry smaller headline percentages but generally tolerate more size before slippage eats meaningfully into the edge, simply because a one-tick move represents a smaller percentage of the total price. Position sizing recommendation: on the sub-cent, high-percentage prints, size to a fraction you'd be comfortable losing entirely to a bad fill or a closed window — think in the low single-digit percentage of your arb allocation per trade — and scale up only on the names where you can independently verify order-book depth beyond the single quoted price. On the LRC-style cross-venue setups with moderate, more durable spreads, larger clips are reasonable precisely because the edge is less likely to be a fleeting artifact.
💰 Profit Calculations
Walking through the OPN print (7.73% gross, buy Binance $0.060900 / sell Coinbase $0.065500) on a hypothetical $10,000 notional: buying on Binance at $0.060900 with a 0.10% taker fee gets you roughly 164,203 OPN for $10,010 all-in. Selling that stack on Coinbase at $0.065500 before fees generates $10,755.30; Coinbase's taker fee (assume 0.60%, typical for a standard retail tier) removes about $64.53, leaving $10,690.77 in proceeds. If you had to withdraw OPN from Binance to Coinbase rather than trading from pre-funded balances, subtract a flat network withdrawal fee — call it roughly $5 equivalent for a token at this price level. Net result: about $10,675.77 back against $10,010 deployed, for a net profit near $665.77, or roughly 6.7% net after both trading fees and withdrawal costs — versus the 7.73% gross headline. That haircut, from 7.73% gross to about 6.7% net, is the realistic tax on this kind of trade: two taker fees (0.10% + 0.60% = 0.70% combined) plus a withdrawal fee that matters more on smaller notional and less on larger. Minimum spread worth chasing: with typical combined taker fees running 0.7%-1.3% round trip depending on the venue pair, and withdrawal/slippage adding another 0.3%-1% depending on size and token, anything under roughly 2% gross spread is marginal-to-not-worth-it for a withdrawal-based arb. Above 3-4% gross, you're comfortably in profitable territory even accounting for a partial fill or minor slippage. Traders running pre-funded balances on both exchanges (no withdrawal leg, no wait) can profitably work spreads as low as 0.5%-1% at high frequency, which is a fundamentally different strategy than the one-shot withdrawal arbs implied by most of today's board.
⚠️ Risk Alerts
Volume data across the entire session came back at $0.0M for pump, dump, buy-pressure, and sell-pressure — that's a feed gap, and it means none of today's spreads can be sized against real reported liquidity. Do not assume a quoted top-of-book price on these sub-cent names is good for meaningful size without checking live depth yourself. Bitunix appeared on both sides of the book today (buy leg for LAB, sell leg for ESPORTS), and as a smaller-tier venue it's more prone to withdrawal delays, wider effective spreads once you go past the top quote, and occasional listing-specific liquidity gaps — treat any Bitunix-involved spread as needing an extra liquidity check before committing capital. Coinbase, despite being the larger venue in most of today's pairs, tends to run slower withdrawal processing than Binance during high-traffic periods, which matters directly for the T and OPN trades where Coinbase sits on the sell side — a delayed transfer into Coinbase risks the spread closing before your inventory arrives. JASMY and APE both showed same-exchange buy/sell legs (Coinbase-to-Coinbase) — these are very likely temporal price snapshots rather than true simultaneous cross-venue arbs, and treating them as executable spreads would be a mistake; verify timestamps independently before sizing into either. Finally, the rapid three-print decay on T (29.45% → 25.38% → 24.20%) is itself a risk signal: whatever caused the initial dislocation is actively being arbitraged away by other participants, so late entrants should expect meaningfully worse pricing than the headline number suggests.
🔮 Tomorrow's Setup
The Binance-Coinbase corridor was the clear structural theme today across T and OPN, and that kind of persistent premium on mid/small-cap listings tends to recur across sessions rather than being a one-off — worth setting standing alerts on both tickers, plus scanning other names that share Coinbase's smaller-cap listing profile for the same pattern. Bitunix's appearance on both sides of the book (LAB, ESPORTS) suggests its quotes are drifting somewhat independently of the majors right now; keep it in the rotation against OKX and Binance Futures specifically, since that's where today's clearest Bitunix-involved gaps showed up. LRC's OKX-KuCoin spread is worth rechecking early in tomorrow's session — it was the cleanest genuine two-venue setup on the board and those relationships often persist for a few sessions once established. Best windows to watch: liquidity-thin stretches — late US evening into early Asia session, and the low-volume hours around major exchange maintenance windows — are historically when cross-venue quote sync lags the most, which is exactly when spreads like today's T print tend to open up. Prioritize confirming live order-book depth before the next session opens, since the $0.0M volume gap in today's data means sizing decisions for tomorrow still need to lean on your own real-time book checks rather than this feed's aggregate figures.
Sign Off
Forty-four prints, one dominant corridor, and a T trade that told you its own story as it closed in real time — that's a good session for anyone paying attention to the tape instead of just the headline percentage. Check depth, respect the withdrawal clock, and don't chase a spread that's already visibly decaying. Arbitrage Hunter — July 17, 2026.
◈ tags
#analysis#crypto#market#arbitrage#spreads#trading