Published
Binance publishes a "global" long/short account ratio (a proxy for retail) and a "top trader" ratio limited to its biggest position holders, for every perpetual pair. Our dashboard pulls both every 30 minutes across the 24 pairs we track. Last week I checked whether the two groups agreed on direction. This time I subtracted one ratio from the other, pair by pair, to see who's leaning harder. On the 2026-10-01 23:00 UTC snapshot, 17 of the 24 pairs have the top-trader ratio sitting above retail β the bigger accounts more committed to long than the crowd. ETH is the sharpest pair that doesn't fit, and it isn't a small-cap curiosity. It's the second-largest coin we track.
The full spread, sorted by the gap
Top-trader ratio minus retail ratio, all 24 pairs. Negative means retail is more long-biased than the big accounts; positive means the opposite. Taker ratio is buy volume over sell volume on the same snapshot β below 1.00 means more aggressive selling than buying over the window.
| Coin | Retail ratio | Top-trader ratio | Gap | Taker ratio |
|---|---|---|---|---|
| ETH | 2.61 | 1.51 | -1.10 | 0.93 |
| BNB | 2.22 | 1.53 | -0.69 | 1.33 |
| XRP | 2.51 | 1.98 | -0.53 | 0.90 |
| TRX | 1.11 | 0.67 | -0.44 | 0.83 |
| SUI | 2.38 | 2.13 | -0.25 | 0.93 |
| ADA | 2.39 | 2.30 | -0.09 | 0.87 |
| HYPE | 1.87 | 1.85 | -0.03 | 0.49 |
| DOT | 1.96 | 2.03 | +0.07 | 0.68 |
| PEPE | 2.00 | 2.12 | +0.12 | 1.17 |
| AVAX | 2.57 | 2.70 | +0.13 | 0.61 |
| XLM | 1.63 | 1.85 | +0.22 | 0.84 |
| LINK | 2.08 | 2.32 | +0.24 | 0.70 |
| SOL | 1.95 | 2.22 | +0.27 | 0.90 |
| SHIB | 1.62 | 2.15 | +0.54 | 1.13 |
| NEAR | 1.59 | 2.21 | +0.62 | 0.73 |
| ARB | 1.18 | 1.96 | +0.78 | 0.85 |
| BCH | 1.64 | 2.50 | +0.86 | 0.91 |
| BTC | 1.04 | 1.92 | +0.88 | 0.92 |
| POL | 1.07 | 2.00 | +0.92 | 0.84 |
| DOGE | 2.83 | 3.98 | +1.15 | 1.08 |
| LTC | 2.13 | 3.85 | +1.72 | 0.97 |
| ATOM | 1.21 | 3.23 | +2.02 | 0.85 |
| UNI | 1.54 | 3.60 | +2.07 | 1.30 |
| APT | 1.23 | 3.53 | +2.31 | 1.01 |
Average across all 24: retail sits at 1.85 (about 65% long), top traders at 2.34 (about 70% long). That 0.49 gap is carried almost entirely by a handful of pairs at the extremes, not spread evenly.
ETH: the one top-10 coin where retail out-bulls the pros
ETH's retail accounts are 2.61:1 long β about 72% long, 28% short. Its top traders sit at 1.51:1 β about 60% long, 40% short. That 1.10 gap is the largest divergence of any of the 24 pairs, in either direction, and it's the only pair among the top 10 coins by market cap where retail leans harder long than the big accounts do. BNB and XRP are in the same column, but ETH's gap is 60% wider than BNB's and more than double XRP's. The taker ratio adds a second data point pointing the same way: 0.93, meaning more aggressive sell volume than buy volume crossed the tape in the same window retail's resting long position sat at a two-year-old high. Six of the seven pairs where retail outweighs top traders on the account ratio β ETH, XRP, TRX, SUI, ADA, HYPE β also show a sub-1.0 taker ratio. BNB is the one exception, at 1.33.
APT is the mirror image, and it's not close
At the other extreme, APT's top traders sit at 3.53:1 long against a retail crowd that's barely leaning at 1.23:1 β a 2.31 gap, the widest of the 24. Retail on APT is close to neutral, 55% long / 45% short. The big accounts are 78% long. ATOM (+2.02) and UNI (+2.07) sit nearly as wide. All three are mid-cap alts where the account count skews retail-heavy by nature, so a handful of large wallets can move the top-trader ratio further than on a deeper market like ETH or BTC β worth keeping in mind before reading too much conviction into the number itself.
BTC and ETH are pointed in opposite directions from each other
The detail I didn't expect going in: the two largest coins we track aren't just different in degree, they're on opposite sides of the line. BTC's gap is +0.88 β top traders notably more long-biased than a retail crowd that's barely leaning long at all (1.04:1, almost a coin flip). ETH's gap is -1.10 β the reverse setup entirely. If you're holding both on the assumption that "crypto majors" move as one positioning block, the account data says the two largest ones currently don't.
What I'm doing with this
This isn't a signal to short ETH or buy APT. Positioning gaps have sat wide before and resolved by the crowd being right, not the big accounts. What changes for me is sizing, the same way the lopsided-but-aligned positioning in last week's snapshot changed my sizing without changing my direction. A long on ETH right now sits on the side the smaller accounts are more committed to, with the bigger accounts and the recent taker flow leaning away from it β that's a reason to run the time to liquidation calculator before adding leverage, not a reason to flip. And if ETH is one of several positions you're counting as diversified, the math in five longs are really one bet is worth rereading with this specific pair in mind. I'm pulling the retail-vs-top gap alongside funding rate on our dashboard before sizing anything on ETH or APT this week β not as a trade trigger, just as one more number that's cheap to check and expensive to ignore.