A 37-opportunity session led by an 11.29% BULLA dislocation between Binance Futures and Gate Futures, with COTI showing up four times in the top 10 as its KuCoin-Bybit and Bitget-Bitunix gaps decayed from 9.28% down to 6.07% over the session.
Thirty-seven arbitrage events crossed the scanner today, and the top of the board was loud: BULLA printed an 11.29% spread between Binance Futures ($0.014614 buy side) and Gate Futures ($0.016264 sell side), the kind of dislocation that either means a fat, fast payday or a liquidity trap dressed up as one. Below it, COTI showed up four separate times in the top 10 alone, a concentration that tells its own story about where the real, repeatable action was sitting today.
This is not a retail sentiment report. There's no pump volume, no dump volume, no buy/sell pressure to lean on today, so the entire read is built off spread structure, exchange pairing, and price precision. That's actually fine for this desk — arb doesn't care about narrative, it cares about the gap between two order books and whether you can close it before someone faster does. Let's get into it.
The clearest pattern today is KuCoin as the cheap side against Bybit as the expensive side, specifically on COTI, showing up three separate times (8.39%, 7.87%, and again at 6.07% further down the board). That's not noise — that's a persistent liquidity or flow imbalance between the two venues on this specific asset, and it's the single most repeatable corridor in today's data. Binance Futures shows up twice as the buy-side venue (BULLA, AKE) and once as the sell-side venue (COAI vs Bybit), which tells you Binance Futures order books on smaller-cap perps can be dislocated in either direction depending on which way retail flow is leaning that hour — it's not consistently cheap or consistently expensive, it's just frequently out of sync with peers on thin names.
Bitget shows up as the buy-side venue twice (COTI at 9.28%, BE at 6.17%), both times against a different counterpart (Bitunix, then Bybit), suggesting Bitget's altcoin books are simply running a touch behind the pack rather than reflecting a single structural relationship with one other exchange. The Binance-vs-Coinbase print on COTI (7.07%) is worth flagging on its own: Coinbase's retail fee structure and generally wider spreads on lower-cap alts make it a chronic laggard against Binance pricing, and that gap is one arb traders have been able to lean on repeatedly across many assets, not just today's COTI print. OKX Spot vs Binance on ZIL (6.22%) fits the same mold — OKX's spot book for lower-volume names tends to trail Binance's price discovery by a beat, which is exactly the kind of mechanical, repeatable seam this desk should keep on a watchlist rather than treat as a one-off.
The COTI data set today is a live case study in spread decay. Reading the four prints together — 9.28%, 8.39%, 7.87%, and 7.07%, before settling further down to 6.07% — the imbalance is clearly being arbed down in real time as capital rotates in to close it. That's the core tension of this business: the earliest mover into a 9%+ spread captures the fattest fill, but by the time a slower trader with a live withdrawal in flight gets there, the gap they're actually able to realize might be half that. Speed wins on the first bite; size only matters on what's left after the fast money has already taken its cut.
Position sizing should scale inversely with how many hops the trade needs. On a futures basis trade like BULLA, where no asset ever leaves the exchange, you can and should size up close to what the thinner leg's order book can actually absorb without moving price — Gate Futures depth is the constraint, not Binance's. On a spot-transfer trade like the COTI or ZIL prints, size should be capped by what you can execute using capital already parked on both exchanges; anything requiring a live withdrawal mid-trade should be sized small enough that slippage on entry and exit, plus the spread decay during the transfer window, still leaves meaningful profit. As a rule of thumb: quick, thin spreads (sub-2%, futures basis) reward speed and moderate size; wide, structural spreads on low-liquidity spot pairs (COTI, AKE) reward pre-positioned capital and conservative sizing over speed, because the withdrawal clock — not the competition — is the real risk.
Take the BULLA basis trade as the clean example, since it's futures-to-futures and skips withdrawal risk entirely. On a $10,000 notional per leg: buy on Binance Futures at $0.014614, sell (short) on Gate Futures at $0.016264, an 11.29% gross spread. Assuming typical futures taker fees of roughly 0.04% on Binance and 0.05% on Gate, round-trip fees run about $9 on a $10,000+ notional pair of legs — trivial against a spread this size. Net result: roughly 11.0-11.1% after fees, or close to $1,110 on $10,000 deployed, with the main residual risk being funding-rate cost if the position is held across a funding interval rather than closed same-session.
Now take the COTI spot trade at 9.28% (Bitget buy $0.014330, Bitunix sell $0.015248) on the same $10,000 notional: gross profit before costs is about $928. Assume standard tier-one spot taker fees of roughly 0.1% per side (about $10 on the buy, roughly $11 on the larger sell proceeds), plus a COTI network withdrawal fee that's typically a few dollars in token terms depending on chain congestion — call it $2 to $8. Net profit lands around $900-$905, or roughly 9.0% net on $10,000, assuming the transfer completes before the spread compresses further. Given today's COTI data shows this exact corridor decaying toward 6-7% within the session, a realistic execution — accounting for a 5-20 minute transfer window — might only capture 6-7% rather than the full 9.28% printed at the snapshot.
The floor for what's worth chasing depends entirely on trade type. For futures basis trades with no withdrawal step, fees alone rarely exceed 0.2-0.3% round trip, so almost anything above roughly 0.5% is worth a look if margin and funding cost are managed. For spot trades requiring a live cross-exchange transfer, the realistic floor is much higher — figure 0.2% in trading fees, a fixed withdrawal cost that matters more on smaller tickets, plus 0.2-0.5% of slippage and decay risk during the transfer window. That puts the practical minimum at roughly 1.5-2% for pre-funded spot capital, and closer to 3%+ if a live withdrawal has to happen mid-trade. Anything below that on a spot pair is fee and timing risk dressed up as opportunity.
COTI earns top billing again — the KuCoin-cheap, Bybit-expensive corridor showed up three times today and looks structural rather than incidental, so pre-funding COTI on KuCoon and USDT on Bybit (and vice versa) ahead of the next session is the highest-probability setup on the board. Binance Futures deserves a standing watch on both sides of the book, since it showed up as both the cheap venue (BULLA, AKE) and the expensive venue (COAI) today — that inconsistency itself is the signal that its smaller-cap perp books are prone to dislocation regardless of direction. Keep an eye on the Binance-vs-Coinbase and OKX-vs-Binance corridors too; both reflect structural fee and liquidity gaps rather than one-off noise, and both are likely to resurface on whichever mid-cap alt sees the next volume spike. Low-liquidity overnight hours, when major-exchange order books thin out unevenly, remain the highest-probability window for the next double-digit print.
Thirty-seven cracks in the order books today, one fat 11.29% basis trade, and a COTI corridor that told on itself four times before it closed. Watch the seams, not the noise. Arbitrage Hunter — July 29, 2026.