Boring Boris breaks down July 3's arbitrage tape: 93 tracked spreads, led by a 34.22% CHILLGUY gap between Bitunix and Hyperliquid and a 29.54% ZEREBRO spread on Hyperliquid vs Binance Futures. Full profit math, exchange-pair patterns, and risk alerts for arb desks.
Ninety-three arbitrage opportunities crossed the tape today, and the desk's top line is a genuine head-turner: CHILLGUY printed a 34.22% spread between Bitunix ($0.009389 bid-side buy) and Hyperliquid ($0.012602 sell-side), the kind of number that makes you double-check your data feed before you double-check your bankroll. Right behind it, ZEREBRO posted a 29.54% gap between Hyperliquid ($0.032370) and Binance Futures ($0.041933) — two low-cap, high-volatility perpetuals doing what low-cap perpetuals do best: disagreeing violently about their own price.
Below those two headline spreads, the bulk of today's action clustered in the mid-teens: GUA alone contributed five separate spread events ranging from 11.19% to 19.63%, almost all of them routed through Bitunix, KuCoin, and Gate Futures — three venues that clearly aren't talking to each other on pricing today. BARD, ZKP, and THE rounded out the top tier with spreads between 16% and 19.3%, spanning a mix of spot and futures venues including OKX, Coinbase, Binance, and Bitunix again. If there's a theme to today's tape, it's this: the exchanges showing up repeatedly — Bitunix, Gate Futures, Hyperliquid — are the ones with either thinner order books, newer listings, or perpetual-vs-spot basis quirks that keep reopening the same gaps.
For arb traders, the headline percentages are bait until you check depth and withdrawal rails. A 34% spread on a token with $200 of resting liquidity on one side is not a trade, it's a screenshot. Let's go through what's actually workable.
Today's data shows a clear pattern: Bitunix appears on one side or the other of four of today's ten listed spreads (CHILLGUY, GUA x2, THE), consistently on the higher-price side. That's a signature of an exchange with thinner books and/or a retail-heavy, momentum-chasing user base — prices there tend to run ahead of the broader market on both pumps and dumps, which is exactly the environment that produces recurring arb opportunities.
Gate Futures vs Binance Futures and Gate Futures vs KuCoin also showed up repeatedly through the GUA cluster (13.00%, 12.86%, 11.19%) — a smaller, tighter band of spreads that reads more like normal cross-exchange futures basis noise than a dislocation. Hyperliquid's two appearances (CHILLGUY, ZEREBRO) are both against major CEX derivatives desks (implicitly Binance-adjacent via the futures leg), reinforcing that Hyperliquid's perp pricing on lower-liquidity names can drift meaningfully from centralized futures venues — a pattern worth setting a standing monitor for if you trade meme-perp basis regularly. OKX-to-Coinbase (BARD) and Binance-to-Coinbase (ZKP) both show major-exchange-to-major-exchange spreads in the high teens, which is unusual for top-tier venues and suggests a temporary liquidity or listing-related dislocation on one side rather than a persistent structural gap — these tend to close faster than the Bitunix/Gate Futures/Hyperliquid patterns.
The fundamental arb tradeoff is on full display today: the biggest spreads (CHILLGUY 34.22%, ZEREBRO 29.54%) are on the thinnest, most volatile pairs, meaning your realistic fill size before slippage eats the edge is small — probably low four figures in USD notional before you start moving the price against yourself on either leg. The smaller spreads in the 11-13% range (the Gate Futures/GUA cluster, Gate Futures/Binance Futures cluster) are on futures venues with generally deeper books, meaning you can push more size through before slippage bites, even though the headline percentage is a third of the top spread.
Position sizing rule of thumb for today's tape: for anything under $0.05 in price with a >25% spread (CHILLGUY, ZEREBRO), cap notional per leg at what you'd be comfortable losing entirely if the basis moves against you before convergence — these are momentum-driven micro-caps, not stable arb machinery. For the BARD/ZKP/THE tier (16-19%, established spot pairs), you can reasonably size up 3-5x versus the micro-cap tier given deeper books on OKX, Coinbase, and Binance. For the GUA/Gate Futures cluster (11-19%), treat each of the five listed spreads as fragments of one ongoing repricing event rather than five independent trades — sizing into all five simultaneously is just concentrating GUA exposure five times over, not diversifying.
Let's run the numbers on the BARD trade since it's the cleanest transfer-based setup: buy OKX Spot at $0.142300, sell Coinbase at $0.150000, gross spread 19.30%. Trading fees: OKX taker ~0.10% on the buy, Coinbase taker ~0.60% on the sell (Coinbase's advanced trade fees run higher than most competitors) — combined fee drag of roughly 0.70% of notional. Assume a withdrawal fee of a flat token amount, call it the equivalent of ~1.5% of a modest $1,000 position (withdrawal fees hit smaller trades disproportionately harder — this is why arb desks avoid sub-$500 trades). Net calculation: 19.30% gross - 0.70% trading fees - 1.50% withdrawal fee = roughly 17.10% net spread remaining. On a $1,000 position that's about $171 in theoretical profit, assuming the spread holds through the transfer window and doesn't compress while your BARD is in transit.
Now the same math on a thinner spread — take THE at 16.08% (Binance Futures $0.065590 vs Bitunix $0.076140). Futures venues typically charge lower taker fees (~0.04-0.05% each side on major futures desks, though Bitunix's futures fee schedule tends to run a bit higher for a newer exchange, so budget 0.10-0.15% there) — combined maybe 0.20% trading fee drag. But this is a futures-to-spot-style spread, so if executed as a basis trade rather than a physical transfer, there's no withdrawal fee at all — just funding rate exposure while the position is open. Net: roughly 15.8-15.9% available if you can hold the basis position long enough to converge, which is a strong net figure for a hedged trade.
General rule for this desk: anything under about 3-4% gross spread on a standard CEX-to-CEX spot transfer isn't worth chasing once you account for both-side taker fees (typically 0.2-0.7% combined) plus a fixed withdrawal fee that disproportionately punishes small size. For basis trades between futures venues or perp DEXs like Hyperliquid, the bar drops closer to 1-2% gross since there's no withdrawal cost, but funding rate risk needs to be underwritten instead. Everything in today's top 10 clears both bars comfortably on a gross basis — the real filter is liquidity and transfer-time risk, not fee erosion.
Watch GUA first thing tomorrow — five separate spread events across four venues in a single session usually means the token isn't done repricing, and a fresh dislocation between Gate Futures and either KuCoin or Binance Futures is a reasonable bet in the first few hours of trading when liquidity is thinnest. CHILLGUY and ZEREBRO both warrant a standing Hyperliquid-vs-major-futures monitor — low-cap perp basis trades like these tend to recur in clusters once a token starts showing volatility, especially around any funding rate reset windows. For the more 'normal' spot pairs, keep an eye on OKX-vs-Coinbase and Binance-vs-Coinbase specifically — both showed unusually wide gaps for major-tier venues today, and if that's driven by a temporary listing or liquidity event rather than noise, it could reopen around the next Asia-session open when OKX volume typically picks up first. Best general watch windows: the first hour after Asia-session open (thin overnight books repricing) and the hour around US futures market open, when Binance Futures and CME-adjacent flow tends to reprice fastest relative to spot.
Ninety-three spreads, one 34% headline that turned out to be a basis trade in disguise, and a handful of GUA repricings that say more about thin liquidity than genuine opportunity. Check your withdrawal times before you check your excitement level. Stay boring, stay solvent.
Arbitrage Hunter — July 3, 2026