Published
Our live dashboard pulls three separate positioning feeds from Binance Futures for 24 major perp pairs every 30 minutes: the long/short account ratio for "top traders" (Binance's own classification of its largest accounts by margin balance), the same ratio for all accounts globally, and the taker buy/sell volume split -- how much of recent trading volume was aggressive buying versus aggressive selling. Today's snapshot (8 November, 10:00 UTC) has an odd pair of extremes sitting on the same coin.
The most long-skewed top-trader ratio of the 24
UNI's top-trader long/short ratio is 3.50 -- 77.8% of large accounts are net long, 22.2% net short. Scanning all 24 pairs we track, that's the single highest top-trader ratio in the set, ahead of APT (2.81), DOGE (2.89) and HYPE (2.59). By this measure, UNI is the pair where Binance's largest accounts are, as a group, most lopsidedly positioned long right now.
The most sell-skewed taker flow of the same 24
Over the same window, UNI's taker sell volume was 173,930 contracts against taker buy volume of 103,131 -- a buy/sell ratio of 0.59. That means for every dollar of aggressive buying hitting the order book, $1.69 of aggressive selling hit it too. Scanning the same 24 pairs for taker ratio, UNI is also the single most sell-skewed of the set -- lower than UNI's own global-account ratio (0.65 would be "balanced"; TRX at 0.82 and LTC at 0.67 are the next-closest sell-skewed pairs, both well above UNI's 0.59).
| Metric | UNI value | Rank of 24 |
|---|---|---|
| Top-trader long/short ratio | 3.50 (77.8% long) | 1st most long |
| Global-account long/short ratio | 1.83 (64.7% long) | mid-pack |
| Taker buy/sell volume ratio | 0.59 (62.8% of flow is selling) | 1st most sold |
What the rest of the tape says
Three more feeds line up with the sell side, not the long side. UNI's open interest is down -1.97% over 24 hours per our open interest tracker -- positions closing, not new ones opening. Funding is negative on six of the seven exchanges in our cross-exchange feed (Bybit -0.0132%, Binance -0.0058%, Gate -0.0045%, MEXC -0.0058%, OKX -0.0030%, all per 8h; only Bitget and Hyperliquid sit slightly positive), meaning shorts are the side getting paid on most venues right now. And UNI's perp is trading at a -0.073% discount to its spot index price on Binance -- a small but real backwardation, the opposite of the premium you'd expect if leveraged demand were skewed long.
So which number is "right"?
Neither, because they're not measuring the same thing. The long/short account ratio is a headcount: it tells you what fraction of accounts (by number, on Binance's top-trader classification -- not dollar-weighted exposure) are sitting net long versus net short at a point in time. It says nothing about size, and nothing about direction of change. Taker volume is a flow: it tells you which side initiated more of the trading that just happened, weighted by size, over the measurement window. A large account can sit net long for weeks while smaller, faster money does the actual buying and selling around it -- the headcount barely moves while the flow swings hard.
What today's UNI snapshot actually shows, stacked honestly: the account-count metric says large traders are, in aggregate, still positioned long. Every flow-based metric we have -- taker volume, open interest direction, funding sign, basis -- is leaning the other way. That's a real disagreement between "who's positioned where" and "what's trading right now," not a data error, and not something we're going to call a top or a bottom. Positioning data describes the crowd. It doesn't predict price.
Where to check this yourself
This snapshot updates every 30 minutes and will not look the same by the time you read this. Our long/short ratio page and live dashboard read the same feed used for this post, and the UNI coin page carries all four metrics -- top-trader ratio, global ratio, taker flow and open interest -- side by side for this specific pair. Check the numbers before acting on anything in this article; by definition, they've already moved.