Fifty-three arbitrage opportunities hit the board today, headlined by a 20.26% MAGMA spread between Bitunix and Bitget. EVAA's repeated Gate Futures discount emerges as the session's clearest structural pattern, while AGLD flips the usual Coinbase-is-tightest assumption on its head.
Fifty-three arbitrage signals crossed the desk today, July 13, 2026, and the spread action was concentrated in the mid-cap and micro-cap corner of the market — exactly where liquidity is thin enough for real dislocations to survive more than a few ticks. Top of the board: MAGMA, printing a 20.26% gap between Bitunix ($0.370990 on the buy side) and Bitget ($0.385720 on the sell side). That's not a rounding-error spread — that's a venue-pricing failure, the kind that either means one order book is badly out of sync with the rest of the market, or there's a liquidity crunch on one side that a fast mover can exploit before the books converge.
Behind MAGMA, LUMIA posted a 15.02% Bitget-to-Binance-Futures gap, and EVAA showed up three separate times in the top ten — 12.47%, 12.39%, and 10.53% — all variations on the same theme: Gate Futures pricing EVAA meaningfully below where Bitget and Bitunix have it marked. When one symbol dominates the board like that, it's not noise, it's a structural basis problem worth watching past today. Round out the top five with AGLD's 10.96% OKX-to-Coinbase spread — notable because Coinbase is usually the deep, efficient side of any pair, not the expensive one.
This is a report built for people who move fast and don't hold bags overnight. Treat every number below as a snapshot, not a standing quote — CEX order books this size close spreads like this in minutes when bots are watching, and stay open for hours when they aren't.
MAGMA — 20.26% spread. Buy on Bitunix at $0.370990, sell on Bitget at $0.385720. This is the standout of the day, and a print this size on a token like MAGMA usually means one of two things: either Bitunix's ask-side book got picked off and hasn't refreshed, or Bitget's bid moved on a broader market update that Bitunix hasn't caught up to. No per-pair volume feed was available for this scan, so size discipline matters more than usual — MAGMA isn't a top-100 asset, and Bitunix in particular is a mid-tier venue where books can be a few thousand dollars deep before slippage eats the edge. Withdrawal risk is the other variable: if capturing this requires moving MAGMA between venues rather than trading from pre-funded balances on both sides, network confirmation time alone can erase the edge before the transfer lands. My take: executable for accounts already holding inventory on both Bitunix and Bitget — chasing this cold, with a live withdrawal in the middle, is a coin flip against the clock.
LUMIA — 15.02% spread. Buy on Bitget at $0.077090, sell on Binance Futures at $0.080624. A spot-to-futures spread this size usually reflects funding or basis dislocation rather than a simple pricing error, and LUMIA's low absolute price (under $0.08) means slippage shows up fast in percentage terms even on modest size. Binance Futures is deep, so the sell leg isn't the constraint — the constraint is Bitget's spot book absorbing a buy order without walking the price up and compressing the spread being chased. This is a spot-buy/futures-sell structure, so executing it cleanly needs either a delta-neutral hedge already in place or a plan for what happens to the futures leg on settlement. Executable for traders running cross-venue basis books; not a plug-and-play spot arb for someone without a futures account already funded.
EVAA — 12.47% spread. Buy on Gate Futures at $1.493200, sell on Bitget at $1.578500. EVAA is the story of the day — it shows up three times in the top ten, always with Gate Futures on the cheap side. That consistency is the tell: this isn't a one-off mispricing, it's Gate Futures running a persistent discount to Bitget/Bitunix pricing on this name, which smells like a funding-rate or index-price lag rather than random noise. That's actually good news for arb desks — a repeating dislocation is easier to plan around than a one-shot spike. The risk is standard futures-arb risk: margin requirements, funding payments if the position is held through a funding window, and Gate's futures liquidity running noticeably thinner than Bitget's spot book. Executable, and worth building a standing watch on this pair specifically rather than treating it as a one-off.
EVAA — 12.39% spread. Buy on Gate Futures at $1.375100, sell on Bitunix at $1.479600. Same asset, same cheap-side venue, different session — this is the second EVAA/Gate-Futures print of the day and it confirms the pattern above rather than adding new information. The price level here ($1.3751) is meaningfully lower than the other EVAA entries ($1.4932, $1.0049, $0.9020), which says EVAA had real intraday volatility today, and Gate Futures consistently lagged whichever spot venue was leading the move. Risk profile mirrors the first EVAA entry — futures margin, funding exposure, thinner book on the buy side. My take: this is the clearest recurring setup in the whole dataset, and it's the one worth a standing alert rather than a one-off chase.
AGLD — 10.96% spread. Buy on OKX Spot at $0.192700, sell on Coinbase at $0.203034. This one's structurally different from the rest of the top five because Coinbase is normally the tightest, most efficient book in any pair it's in — seeing it as the expensive side of an 11% gap against OKX is unusual and worth a second look before sizing up. Possible explanations: a U.S.-specific demand spike (Coinbase's user base skews retail-US) or a temporary Coinbase liquidity gap on a name that isn't heavily traded there. Moving AGLD onto Coinbase from OKX is straightforward on-chain since AGLD is an ERC-20, so network cost is the main friction rather than availability — but Coinbase deposit processing can add confirmation delay that a fast-moving OKX price won't wait for. Executable for accounts with standing Coinbase balances; more marginal if it requires a live cross-chain transfer mid-trade.
The clearest structural signal in today's data is Gate Futures running persistently cheap against Bitget and Bitunix on EVAA — three separate prints, three different price levels, same direction. That's not coincidence; that's a venue whose futures pricing lags the market leader on this specific name, likely tied to funding-rate mechanics or a slower index-price feed. Any systematic EVAA arb strategy should treat Gate Futures as the structural buy side and Bitget/Bitunix as the structural sell side until the pattern breaks.
The second pattern worth flagging: Bitget shows up on both sides of the ledger today — cheap side for LUMIA (the buy leg), expensive side for MAGMA and EVAA (the sell leg). That tells you Bitget isn't systematically mispriced in one direction; its book just moves independently enough from the majors that it throws off spreads in both directions depending on the asset. Treat Bitget less as 'the cheap venue' or 'the expensive venue' and more as 'the venue most likely to be out of sync,' which is exactly what deserves continuous monitoring rather than a fixed assumption.
OKX and Coinbase showed up once each, on opposite sides of the same AGLD trade — OKX cheap, Coinbase expensive. That's the one pattern in today's data running against the usual expectation, since Coinbase is typically the efficient, tight-spread venue, which makes it worth monitoring for a repeat rather than trading in isolation. Binance Futures and KuCoin each appeared once, both on the expensive/sell side — too thin a sample to call a pattern yet, but filed for tomorrow's comparison.
Every spread on this board falls into one of two buckets, and mixing up which bucket a trade is in is how arb traders bleed out on fees and slippage. Bucket one: small, fast, CEX-to-CEX spot spreads like TAC's 8.35% (KuCoin to Binance Futures) or PIPPIN's 8.74% (OKX to Bitget) — these close in minutes once bots notice them, so the play is small size, immediate execution, and no attempt to scale in. Trying to put six figures through a sub-9% spread on a low-cap token like TAC or PIPPIN guarantees walking your own buy order up and your own sell order down before the fills complete, compressing the very edge being chased.
Bucket two: the structural, recurring dislocations — EVAA's repeated Gate Futures discount, MAGMA's outsized 20% print. These can absorb more size because they're not single-order-book noise, but they carry more risk per dollar (futures margin on EVAA, thin secondary-venue liquidity on MAGMA). The right sizing discipline is to test with a small clip first, confirm the spread holds through your own fill, then scale — never assume the quoted spread survives full position size.
As a rule of thumb for this dataset: anything under 10% spread on a sub-$0.10 token gets treated as a scalp — cap position size at whatever your own book-depth check shows won't move the price more than 1-2% on entry. Anything over 12% that's recurring, like EVAA, is worth automating as a standing strategy rather than trading as a one-off, sized to the futures venue's margin requirements rather than the spot venue's book depth.
Gross spread is never net profit — here's the walk-through using today's two ends of the spectrum, assuming a $10,000 notional and standard taker-fee tiers.
MAGMA (20.26% gross spread):
TAC (8.35% gross spread):
The minimum spread worth chasing on CEX-to-CEX spot pairs, assuming pre-funded balances on both sides with no live withdrawal in the loop, is roughly 0.5-1% after the standard 0.1-0.2% round-trip taker-fee load — below that, it's trading for pennies against slippage risk. The moment a live cross-exchange transfer enters the trade, that floor jumps to 3-5% minimum, because network confirmation time is a window where the spread can close against you with zero recourse. Every spread in today's top ten clears both bars comfortably — the question was never 'is it profitable on paper,' it's 'can you actually get filled at these prices before the book moves.'
The standing watch for tomorrow is EVAA against Gate Futures — three prints in one day against Bitget and Bitunix is a pattern, not a coincidence, and it's the setup here most worth rechecking before the US session opens, since that's typically when Binance/Bitget spot liquidity thickens and any lagging futures venue gets pulled back into line fastest. Keep MAGMA on a secondary watch: a 20%+ print this size on a mid-cap token often isn't a one-time event if the underlying cause is a liquidity gap on Bitunix rather than a data error — worth checking Bitunix's MAGMA book again around the same time tomorrow.
For pair-watching generally, prioritize Gate Futures vs. Bitget/Bitunix spot on lower-cap names — that's where today's board showed the most repeatable structure. Keep a secondary eye on Bitget broadly, since it showed up on both the cheap and expensive side of different trades today, suggesting its book runs independently enough from the majors to keep throwing off spreads in either direction. Best timing window based on today's pattern: early US morning / late Asia session, when EVAA's Gate Futures gap appeared widest across the three prints.
Fifty-three signals, one clear structural play in EVAA/Gate Futures, and a reminder that the biggest number on the board isn't always the best trade — MAGMA's 20.26% looks great until you check whether Bitunix's book can actually fill you. Do the fee math before the fill, not after.
Arbitrage Hunter — July 13, 2026