Perpetual versus spot volume: leverage heat check

Why the perp-to-spot volume ratio matters.

The perpetual-futures to spot volume ratio is the cleanest single read on whether a market move is leveraged or organic. A ratio of 5x means perpetuals are trading at five times the spot volume. What counts as "high", though, depends on the asset and the era: on Binance, BTC's daily ratio has had a median of about 8.4x since January 2024, so a fixed 5x alarm would be ringing most days.

The ratio is computed from each venue's ticker quoteVolume — the USD-equivalent traded value over a rolling 24-hour window. Binance, OKX, Bybit and Bitget all publish this via public APIs.

Historical ranges — top, bottom, sideways.

Measured on Binance BTCUSDT (USDS-M perpetual daily quoteVolume divided by spot daily quoteVolume), there is no fixed "top" or "bottom" level. Tops have printed at rising levels each cycle, and the one long stretch below 2x — July 2022 to March 2023 — coincides exactly with Binance's zero-fee BTC spot trading, which inflated spot volume: the ratio fell from about 8.2x on 7 July 2022 to 2.1x on 8 July, and jumped from 1.6x on 21 March 2023 to 4.8x on 22 March.

Reading bands used on this site.

Instead of a fixed 5x line, readings are banded by each asset's own percentiles. Method: Binance USDS-M perpetual daily quoteVolume divided by spot daily quoteVolume, 1 January 2024 to 25 September 2026 (999 daily readings per asset), cut at the 20th, 50th and 80th percentiles. For BTC that gives: below 7.0x cool, 7.0x-8.4x normal, 8.4x-9.8x active, above 9.8x overheated. The same cuts for the other majors: ETH 6.9 / 9.1 / 12.3, SOL 4.4 / 6.1 / 7.9, BNB 2.3 / 2.9 / 3.8, DOGE 4.2 / 5.4 / 7.0. Data from July 2022 to March 2023, when Binance ran zero-fee BTC spot trading, is not comparable and is left out.

Four practical reading techniques.

1. Anchor on the absolute level first.

Before reading trend, place the current value against the asset's own recent history. A reading of 4x means different things from a 2x base versus an 8x base — rank it against the asset's own percentile bands above rather than a fixed line.

2. Read trend, not snapshot.

A ratio of 5x with a 30-day downtrend is bullish (deleveraging into spot); a ratio of 5x with a 30-day uptrend is bearish (leverage building). The point value alone is uninformative.

3. Watch for divergence with price.

Price making new highs while the ratio falls indicates spot-led buying — bullish. Price flat while ratio rises indicates leveraged positioning building without spot confirmation — fragile.

4. Always read alongside funding rates.

High ratio plus high positive funding is the most fragile configuration: leverage is crowded and expensive. High ratio with neutral funding is less stressed. Low ratio with negative funding is the cleanest accumulation profile.

Common misreads.

1. Treating instant high ratio as sell signal.

A single-day jump into the overheated band is usually a one-off event (CPI print, FOMC, single liquidation cascade). Sell signals come from sustained elevation, not snapshots.

2. Assuming uniform "healthy zones" across assets.

Over the year to 25 September 2026, Binance daily ratios had medians of about 8.6x for BTC, 12.4x for ETH, 7.8x for SOL, 6.8x for DOGE and 3.8x for BNB. The spread between assets is wide and does not follow a simple "smaller coin, higher ratio" rule, so compare each asset with its own history.

3. Mistaking Binance-only data for the whole market.

Binance is one of the largest venues, but OKX, Bybit, Bitget, CME and others carry substantial volume too. A Binance-only ratio can diverge from the global picture during venue-specific events.

4. Confusing rolling 24h windows.

Different venues compute the 24h window with different start times (some midnight UTC, some rolling). Apparent ratio divergences between venues sometimes just reflect window timing.

5. Underweighting holiday and US-market effects.

Perp trading is more sensitive to US market open and to Asian holiday cycles than spot. Around major holidays, check whether a move in the ratio simply reflects thinner participation before treating it as a signal.

6. Mistaking quoteVolume for actual dollars exchanged.

quoteVolume sums the USD value of every trade. Internal washing and wash-trading inflate the number on some smaller venues. Binance's ratio is more reliable than aggregate cross-venue ratios.

From signal to decision — three usable patterns.

Pattern A — De-risk warning light.

When the BTC perp/spot ratio climbs into the top decile of its own past year (about 10.5x or higher over the year to September 2026) while funding is also rising, reduce leverage on any open long positions. Treat it as a risk-budget cue, not a top call: around the April 2021 top the ratio was only 3.3x-4.9x.

Pattern B — Accumulation window indicator.

When the ratio drops into the bottom of its own recent range during a price drawdown, spot may be absorbing the selling without leverage. Check for data artefacts first: the sub-2x readings around the November 2022 FTX bottom fell inside Binance's zero-fee spot period, and during the August 2024 crash the daily ratio only dipped from about 10.6x to 7.9x.

Pattern C — Event verifier.

Pair the dashboard with news flow. When a positive catalyst (ETF approval, regulatory clarity, macro pivot) hits while ratio stays low, the move is high-confidence spot-led. When a positive catalyst hits while ratio is already elevated, the move is more easily faded.

Placing this metric in the indicator map.

Perp/spot ratio sits at the intersection of three other layers. Funding rates tell you the cost of the perp positioning; open interest tells you the size; this ratio tells you the share. All three together describe leverage condition in one read. For this site's free read, the three free dashboards on /en/articles/funding-rate-guide.html, /en/articles/open-interest-guide.html and this page give equivalent coverage.

Four historical case windows worth memorizing.

Case 1 — April 2021 first ATH (top textbook).

Binance's BTC perp/spot ratio was about 3.5x on 12 April 2021 and 3.9x on 14 April, the day of the $64,854 top; its highest daily reading that month was about 5.4x. The clearer crowding signal was funding: 16 of April's 90 BTCUSDT settlements were at or above 0.1%.

Case 2 — November 2022 FTX collapse (bottom textbook).

During the FTX crash the ratio actually rose, from about 1.2x on 6 November 2022 to 2.7-2.8x on 8-9 November, as perpetual volume surged faster than spot; through the third week of November it sat around 1.6x-2.1x. All of this was inside Binance's zero-fee spot period, so the low absolute level says little about leverage leaving the market.

Case 3 — March 2024 first new ATH (ETF era).

The March 2024 ATH was reached with a ratio of about 7.2x (5.7x-7.9x across the month) — higher than around either 2021 top. ETF spot demand happened on US ETF and Coinbase rails, not in Binance's spot volume, so a Binance-only ratio does not capture it.

Case 4 — August 5, 2024 flash crash (cleanout textbook).

On a daily basis the ratio fell, from an average of about 10.6x over 1-4 August 2024 to about 7.9x on 5 August, because panic spot volume rose even faster than perpetual volume. By 10 August it was back near 11.6x. Daily data cannot show intraday swings; use hourly candles if you need them.

FAQ — frequently asked questions.

Why use quoteVolume instead of volume?

quoteVolume is USD-equivalent; raw volume is asset-count. Comparing across assets requires the dollar number.

Why do Coinglass and Velo show different numbers?

Different aggregator coverage of venues. The direction agreement is typically high; absolute level differences reflect methodology choice.

Why not include BTC-margined perps (BTCUSD_PERP)?

Most retail users on Binance trade the USDT-margined contract, so the ratio sticks to that one. Adding coin-margined volume is possible, but it changes the level and would break comparison with the history quoted above.

Why does the ratio not match what's on Binance's own page?

Binance does not publish a ready-made perp/spot ratio. You compute it yourself from the public endpoints with the formula in the methodology section below.

Is there an "absolute top" signal level?

No. Since January 2024 the BTC ratio has been above 8x on about 57% of days, so a fixed level like that is routine rather than a warning. Use percentiles against the asset's own history, together with funding and open interest.

Why no 24h delta on the ratio itself?

Ratio of ratios is unstable on a daily timeframe. A 30-day series of daily readings shows direction better than a 24h percent change.

Companion tools and further reading.

• Funding rate guide — perp pricing pressure.
• Open interest guide — perp positioning size.
• Long-short ratio — perp positioning side.
• BTC volatility — realized volatility regime.

Data methodology.

This English page is a reading guide without a live widget. The live dashboard on the Chinese version polls Binance Spot ticker /api/v3/ticker/24hr and Binance USDS-M Futures ticker /fapi/v1/ticker/24hr every 60 seconds for BTC, ETH, SOL, BNB and DOGE, and builds its 30-day series from the daily kline endpoints. The historical figures on this page use the same formula on daily klines: perp quoteVolume divided by spot quoteVolume, same UTC date.