August 24, 2026 delivered 40 arbitrage opportunities across the board, headlined by five VELVET spreads between Gate Futures, Bitunix, and Binance Futures — including a monster 35.72% gap. TRUMP, TAC, PROM, and STORJ rounded out a session that rewarded traders who could move fast and had capital pre-positioned across venues.
Forty arbitrage flags crossed the desk today, and the top of the board was dominated by a single ticker throwing off spreads that, on paper, look almost too good to be true. VELVET posted five separate opportunities in the double digits, capped by a 35.72% gap between Gate Futures ($0.773370) and Binance Futures ($0.811988). That is not a typo. A near-36% spread on a listed perpetual pair is the kind of number that either means a liquidity air pocket, a funding-rate dislocation, or a pricing feed hiccup — and arb traders need to know which before they touch size.
Below VELVET, the board had real texture: TAC threw an 11.53% KuCoin-to-Gate Futures spread, TRUMP printed twice (an 11.08% futures-futures gap and a cleaner 7.39% spot-spot gap between OKX and Bybit), PROM opened a 5.93% Binance Futures-to-KuCoin window, and STORJ closed out the top ten with a textbook 5.80% Binance-to-OKX spot spread. Forty total events in a single session is an active day — this wasn't one exchange having a bad print, it was a market where pricing consensus broke down in multiple corners at once.
The theme for today's arb hunters: the biggest headline numbers were concentrated in futures-versus-futures or futures-versus-spot legs on a low-cap, thin-liquidity name (VELVET), while the more conservative, genuinely executable spreads sat in spot-versus-spot pairs on majors and mid-caps (STORJ, TRUMP's OKX/Bybit leg). Know which bucket you're trading before you size the position.
Gate Futures showed up as the buy-side leg in four of today's top opportunities (all four smaller VELVET prints plus the TRUMP 11.08% gap where it was the sell side), consistently pricing below Binance Futures on the VELVET pair. That's a meaningful pattern — if Gate Futures is structurally lagging Binance Futures on illiquid listings, it's worth watching that pair specifically for repeat dislocations rather than treating today as a one-off.
Binance showed up on both sides of the ledger today: Binance Futures was the sell-side leg on four VELVET/TRUMP prints, while spot Binance was the buy-side leg on the STORJ opportunity. That dual role — cheap on spot, rich on futures relative to peers — lines up with a market where Binance's futures order book absorbed more aggressive buying pressure than its competitors during the session.
The spot-to-spot pairs told a cleaner story. OKX Spot appeared on both sides of the board — buy side for TRUMP (undervalued relative to Bybit) and sell side for STORJ (rich relative to Binance). Bybit Spot and Bitunix each appeared once as the pricier venue. None of Hyperliquid, Bitget, or KuCoin (aside from KuCoin's TAC and PROM legs) showed up in the top-tier prints today, suggesting liquidity on those venues stayed tighter to consensus pricing while Gate, Binance Futures, and Bitunix carried the day's dislocation risk.
The VELVET cluster is a five-alarm reminder of the speed-versus-size tradeoff. A 35.72% spread is enormous on paper, but the token's price bouncing across five distinct bands in one session ($0.23 to $0.81) tells you the underlying liquidity was too thin to support real size without blowing through the quoted price. On a name like this, the correct move is small, fast clips — probe-sized fills that test whether the book actually holds at the quoted level — not backing up the truck on the theoretical spread.
Contrast that with STORJ's 5.80% Binance-to-OKX gap and TRUMP's 7.39% OKX-to-Bybit gap. These are both spot-spot, majors-adjacent pairs where order books are deeper and slippage on moderate size is far more predictable. These are the spreads worth sizing up on, precisely because the percentage is smaller — the market is telling you the mispricing is real and not just an artifact of a hollow book.
Position sizing recommendation: treat headline spread size as inversely correlated with safe position size on unfamiliar low-cap tickers. On VELVET-type prints, cap size to what you'd be comfortable losing entirely if the spread was a stale-quote illusion. On STORJ/TRUMP-type spot-spot prints with established liquidity, size can scale toward normal risk limits, capped by the venue's actual order book depth at the quoted price — always check the book before committing, never trade off the headline number alone.
Walking the STORJ opportunity as the cleanest, fully-transferable spot-to-spot example: buy on Binance at $0.046600, sell on OKX Spot at $0.048920, gross spread 5.80%. Trading fees: assume 0.10% taker on the Binance buy leg and 0.10% taker on the OKX sell leg, for 0.20% total in trading costs. Withdrawal: moving STORJ off Binance to OKX carries a flat network withdrawal fee, which on a modest trade size can eat anywhere from 0.10% to 0.30% of notional depending on trade size — smaller trades get hit harder proportionally. Netting it out: 5.80% gross, minus roughly 0.20% trading fees, minus roughly 0.15-0.30% withdrawal drag, leaves a net spread in the 5.3-5.5% range — still a strong, tradeable edge if you can execute both legs before the gap closes.
The TRUMP OKX-to-Bybit leg (buy $2.648401, sell $2.844000, 7.39% gross) runs the same math: roughly 0.20% combined trading fees, plus withdrawal drag that's proportionally smaller on a higher-priced token like TRUMP since the flat network fee represents a smaller percentage of a larger per-unit value. Net spread lands closer to 7.0-7.1% — one of the cleanest opportunities on today's board.
The VELVET, TAC, and PROM prints that cross a futures leg don't fit this framework at all — there's no withdrawal step because futures contracts aren't transferable assets. Instead, the real cost is capital lockup: margin sitting idle on two exchanges simultaneously, funding-rate exposure while the position is open, and the execution risk of both legs filling at quoted prices before the basis reconverges. Treat those as a different asset class of trade entirely, with profit determined more by funding-rate math and reconvergence speed than by a simple gross-minus-fees calculation.
Minimum spread worth chasing: for clean spot-to-spot pairs, anything under roughly 1.5-2.0% gross is not worth the round-trip once fees and withdrawal drag are counted, especially on lower-priced tokens where flat withdrawal fees bite hardest. For futures-basis trades, the bar is different — funding costs and capital opportunity cost mean you generally want a wider buffer, north of 3-4%, before committing dual-exchange margin to catch reconvergence.
VELVET's five-print, five-price-band session is the biggest flag on the board today. A token repricing this violently across this many venues in one session is not a stable arbitrage environment — it's a sign of either a liquidity event, a listing/delisting catalyst, or a market maker stepping away from one or more books. Treat every VELVET quote today as suspect until confirmed live in the order book; stale or thin quotes can show a spread that simply isn't fillable.
Withdrawal timing is the silent killer on spot-spot arb. STORJ and TRUMP's spot legs both depend on network confirmation times between exchanges — if a withdrawal gets stuck in a congested mempool or hits an exchange's manual review queue for a larger transfer, the spread can fully close or reverse before funds land, turning a profitable trade into a loss. Always check current network conditions and each exchange's live withdrawal status before initiating the transfer leg, not after.
Futures-futures and spot-futures legs (VELVET's five prints, TAC, PROM, TRUMP's futures pair) all require capital already resident on both exchanges — don't attempt to fund a new account mid-opportunity, the window will be long gone before a deposit clears and gets margin-eligible. Keep standing balances on your core arb venues if you want to actually capture these when they appear.
Watch VELVET first thing tomorrow — a session with five distinct dislocations rarely resolves cleanly overnight, and there's a real chance the Gate-Binance and Bitunix-Binance legs are still mispriced at the open, especially if today's catalyst (likely a liquidity or listing event) hasn't fully played out. If VELVET stabilizes, watch whether Gate Futures continues trading persistently below Binance Futures on other low-cap listings — today's pattern suggests it might be a venue-level lag rather than a one-token fluke.
For clean spot-spot setups, keep an eye on OKX Spot versus both Bybit and Binance — it showed up on both sides of the ledger today, meaning its pricing is moving independently enough from peers to keep throwing spreads. TRUMP and STORJ both look like names where volatility could reopen similar gaps if broader market conditions stay choppy. Best windows to watch: the first hour after major exchange funding resets and any period of elevated overall crypto volatility, since that's when order books thin out fastest and pricing consensus breaks down across venues.
Forty opportunities, five of them on a single chaotic ticker, and the clearest lesson of the day: the biggest number on the screen isn't always the best trade — the cleanest edge today was a boring 5.80% spot spread on STORJ, not the flashy 35.72% VELVET headline. Trade the book, not the ticker. — Arbitrage Hunter — August 24, 2026.