What is the long/short ratio?

What the ratio counts

In any futures market, the total size of long positions always equals the total size of short positions, because every contract has a buyer and a seller. So a long/short ratio cannot be "more money long than short". What exchanges publish instead is a count of accounts: of all the users with a position in a contract, what share is net long and what share is net short.

The ratio is simply the long share divided by the short share. If 60% of accounts are net long and 40% net short, the ratio is 1.5. A ratio of 1.00 means an even split, 2.00 means two long accounts for every short one. This site shows both the share (easier to read) and the ratio (what most charts quote).

The series we track

• Binance, all accounts (globalLongShortAccountRatio): the share of all Binance futures accounts with a position in the symbol that are net long.
• Binance top traders, accounts (topLongShortAccountRatio): among the top 20% of accounts by margin balance, the share of accounts that are net long. Each account counts equally.
• Binance top traders, positions (topLongShortPositionRatio): among the top 20% of accounts by margin balance, the share of their combined position that is long. This one weights by size, not by heads, so a few large accounts can move it.
• Bybit, all accounts (account-ratio): Bybit's buyRatio, the share of users holding a long position.
• Gate, all accounts (contract_stats): the share of accounts with a position that are long, counted by accounts rather than position size.
• Gate top traders, accounts and positions (also contract_stats): two separate series for Gate's top traders. The account share counts heads; the position share weights their combined positions by size.
• Bitget, all accounts (account-long-short): the share of accounts with a position that are long, counted by accounts rather than position size.
• HTX top traders, accounts and positions (swap_elite_account_ratio and swap_elite_position_ratio): account-count and position-weighted shares for HTX's top traders. HTX excludes accounts that are both long and short (its "locked" share); we show buy divided by buy plus sell. These shares are not directly comparable with the other venues' top-trader shares.
• Binance taker buys (takerlongshortRatio): the share of 5-minute taker volume that was buying. This measures trading flow, not accounts or outstanding positions; see CVD and taker flow.

Exchanges publish these through their public market-data APIs in 5-minute steps. Binance serves only the last 30 days of history. This site polls every 5 minutes and keeps the history. Binance and Bybit cover the top 100 coins ranked by combined Binance and Bybit USDT-perp volume; Gate, Bitget and HTX cover only the first 50 of those coins, where they list the USDT perp. The selection refreshes hourly. See the long/short ratio page and each coin's page, for example BTC long/short ratio.

Why the crowd is often long

On major coins the all-accounts ratio sits above 1 most of the time, often well above it. That does not mean the market is about to rise. Retail traders tend to buy dips and hold small longs, while the other side of their contracts is often held by fewer, larger accounts: market makers, funds hedging spot holdings, and basis traders who are long spot and short the perp. Many small longs against a few large shorts gives a high account ratio with perfectly balanced contracts.

That is why many traders read the all-accounts ratio as a contrarian signal: when it is unusually high, retail is heavily long and the larger players are on the other side. The top-trader ratio is often read the other way, as what better-capitalised accounts are doing. Neither reading is reliable on its own.

How to read it

Changes beat levels. Each coin has its own normal range. BTC with 55% of accounts long may be ordinary; the same reading on a small coin that usually sits at 70% may be unusually bearish. The live box above shows the 24-hour change in percentage points next to each reading for that reason.

Divergences. When price rises but the share of accounts long falls, shorts are being added into the rally, or longs are taking profit. When price falls and the long share climbs, traders are buying the dip with leverage. Both set up squeezes if the move continues against the crowd.

Combine it. The ratio is most useful next to funding and open interest. A high long share with high funding and rising OI is a crowded long. A high long share with negative funding is a warning that the account count and the money disagree.

Limits

• Account series count accounts, not traders or money. One person can have several accounts, and one account can hold a hedged book. Position-weighted series measure size within the selected top-trader group instead.
• Each exchange measures only its own users and defines its own top-trader group, so ratios differ and should not be averaged.
• It says nothing about leverage or entry prices, so it cannot tell you where liquidations sit.
• Hyperliquid does not publish an equivalent exchange-wide account series. The series here cover Binance, Bybit, Gate, Bitget and HTX, with the coverage and measurement differences described above.

Treat the long/short ratio as a sentiment gauge for exchange users: useful context, especially when it reaches an extreme for that coin, but not a forecast.

Checked against the exchanges' own documentation on 21 October 2026: Binance API: long/short ratio · Binance API: top trader long/short position ratio · Bybit API: long/short ratio. Not financial advice.