A 64-event arbitrage scan on September 1, 2026 turned up back-to-back spreads on SKR, BTR, and ESPORTS, led by an 8.80% SKR gap between Hyperliquid and Bybit. Here's the desk breakdown — best plays, exchange corridors, and the fee math that separates paper profit from real fills.
Sixty-four arbitrage events crossed the scanner today, and the board was dominated by two tickers most of you have never traded size in: SKR and BTR. That's the first thing worth saying out loud before we get into the individual plays — this wasn't a BTC/ETH basis day, it was a low-cap, thin-book day, and that changes every assumption you'd normally bring to an arb desk.
The best spread of the session belonged to SKR, printing an 8.80% gap between Hyperliquid ($0.025549 bid-side buy) and Bybit ($0.026522 ask-side sell). That's a wide spread by any CEX-to-CEX standard, let alone CEX-to-perp-DEX. A second SKR opportunity followed almost immediately at 8.76% on a different venue pair, which tells you the SKR mispricing wasn't a one-off blip — it was a structural lag that stuck around long enough to fire twice.
BTR was the volume leader in terms of event count, showing up six times in the top tier with spreads ranging from 6.49% up to 8.69%, across five different exchange pairs. When one asset dominates the opportunity feed like that, it's rarely random — it usually means a listing imbalance, a funding dislocation, or one venue's price feed genuinely lagging the rest of the market. We'll dig into which of those it looks like in the exchange-pattern section below.
One flag before the breakdown: the volume and pressure totals fed into this scan all came back at $0.0M. That means we have spread and price data, but no confirmation of how much size was actually sitting on either side of the book when these prints fired. Treat every 'executable' call below as a read on the price data alone — size verification has to happen live, on your own book depth, before you route a single order.
The Gate Futures → Bitunix corridor is the standout pattern of the session. It fired on BTR at 8.69% (Gate $0.133650 → Bitunix $0.139288) and again on ESPORTS at 7.30% (Gate $0.016430 → Bitunix $0.017630), plus a related Gate Futures → Exchange51 leg on BTR at 7.09% and 6.78%. Two different assets, same buy-side venue, tight timing — that's a feed-lag signature on Gate Futures' futures pricing relative to faster-updating venues, not asset-specific news. If you're building a monitoring bot, this is the corridor to weight heaviest tomorrow.
Bybit shows up as the sell-side landing spot in three of the top prints (both SKR opportunities, plus BTR at 7.37% against Binance Futures and 6.49% against KuCoin). Bybit repeatedly being the higher-priced destination on smaller-cap tickers suggests either slower price discovery on thin BTR/SKR books there, or simply that Bybit users are paying up for access to names that aren't deeply listed elsewhere yet. Either way, Bybit is functioning as the consistent 'sell into' venue today, not the 'buy from' venue — worth remembering for order routing.
Exchange24 and Exchange51 (the newer venues in the rotation) both appear on the BTR list — Exchange24 as a buy-side venue at 7.21% against Exchange51, and Exchange51 as a sell-side destination twice more against Gate Futures. New or lower-volume venues showing up repeatedly on the same ticker is a liquidity-thinness tell: these are exactly the conditions that produce real percentage spreads on paper that evaporate the moment you try to push five figures of size through them. Treat any Exchange24/Exchange51 leg as small-size-only until you've personally verified book depth.
Hyperliquid appeared once, on the day's largest spread. A single DEX-perp appearance against a wide field of CEX pairs isn't enough to call a pattern yet, but it's consistent with what we usually see: DEX perp funding/basis dislocations against CEX spot tend to be real and structural (driven by funding rate mechanics) rather than a stale price feed — which is one reason opportunity #1 is genuinely interesting despite the settlement friction.
Every spread on today's board falls into the same bucket: small-cap, thin-liquidity names with double-digit-cent-to-sub-dollar pricing. That means the classic arb tradeoff — chase the wide, fast-decaying spread with small size, or wait for a deeper, slower-moving spread you can push real size into — tilts hard toward the 'small size, fast execution' side of the ledger today. There is no BTC/ETH-scale basis trade in this dataset where you could responsibly run six figures.
On SKR and BTR specifically, expect slippage to eat into your realized spread faster than the fee math alone would suggest. A 0.025-0.03 price-per-unit token with a market cap thin enough to produce 7-9% cross-exchange dislocations is, by definition, a market where a $5,000-$10,000 order can move the book 1-2% against you on entry alone. Position sizing recommendation: cap single-leg size at whatever your own book-depth check shows will fill within 0.3-0.5% slippage, and don't scale up just because the headline spread looks fat — the headline spread is a snapshot, not a guarantee of fill quality.
The repeating Gate Futures → Bitunix and Gate Futures → Exchange51 corridors are the better candidates for slightly larger, more patient size, precisely because they showed up multiple times rather than once. A spread that repeats across two assets and three separate prints in one session is more likely to be a structural feed lag you can trade repeatedly at modest size than a single flash spike you have to catch perfectly. Speed still matters, but it matters less than being positioned on both venues in advance.
Let's walk the math on the day's headline play — SKR at 8.80%, buying Hyperliquid at $0.025549 and selling Bybit at $0.026522 — using a $10,000 notional per leg as the example size.
Now the Gate Futures → Bitunix BTR play at 8.69% ($0.133650 buy → $0.139288 sell), same $10,000 notional: gross profit $869.00. Futures taker fees run lighter (~0.05% Gate, ~0.10% Bitunix) for -$15.00 in trading costs. Because both legs are futures-style venues, you can often net the position without moving spot balances between exchanges at all — running it as a matched long/short book instead of a physical buy-transfer-sell — which removes the withdrawal-fee line item entirely if you're pre-funded on margin on both sides. Net profit: roughly $854, or ~8.5%.
The rule of thumb for this desk: on a same-exchange-type, pre-funded corridor (perp vs perp, no withdrawal needed), a spread above roughly 1.0-1.5% is worth chasing once fees are netted out. On a genuine cross-exchange, capital-must-move corridor like the SKR Hyperliquid→Bybit play, you need at least 3-4% gross spread before withdrawal fees and bridge delay risk leave you with a profit that justifies tying up capital and taking settlement risk. Anything below that on a low-cap token is fee-and-slippage bait, not a real opportunity — no matter how good the headline percentage looks.
Watch the Gate Futures → Bitunix and Gate Futures → Exchange51 corridors first — they fired on two different assets today (BTR, ESPORTS) with consistent directionality, which is the strongest signal in the dataset that a feed-lag or liquidity gap on Gate Futures pricing is structural rather than one-off. If it's still there tomorrow, it's tradeable with pre-positioned capital rather than something you have to catch live.
Keep an eye on SKR and BTR specifically for continuation — both had multiple independent prints today, which typically means the underlying liquidity imbalance (a new listing, thin market maker coverage, or a recent unlock) hasn't resolved yet. ESPORTS and ZORA are secondary watches; ZORA's single 6.13% print (KuCoin buy at $0.009863, Binance Futures sell at $0.010252) is worth checking again for repeats before committing size.
Best times to watch: low-cap tokens like these tend to show their widest cross-exchange dislocations during Asia-session low-liquidity hours and immediately around US futures funding resets, when smaller venues' price feeds lag the majors the most. If you're running an automated scanner, weight the Gate Futures / Bitunix / Bybit / Exchange51 venue set highest for tomorrow based on today's repeat-fire pattern.
Sixty-four prints, one clean recurring corridor, and a reminder that on low-cap names the spread on your screen is never the spread in your wallet — fees, withdrawal time, and unverified book depth all take a bite before you do. Trade the repeats, not the flashes, and always check your own fills before you check the headline number.
Arbitrage Hunter — September 1, 2026