Crypto squeeze risk and evidence today

Current source-backed snapshot

This page ranks markets where leverage-stress evidence identifies a crowded side: hourly-normalized funding, non-negative one-hour open-interest change, adverse perpetual price change, observed OKX same-side liquidations and opposing Binance Spot CVD when available. Every directional row publishes both scores and the state of each component. Because rising OI adds pressure to both scores but does not reveal direction, score ties and OI-only buildup are separated as direction-undetermined instead of being forced into a long- or short-side label. The score is a review priority—not a squeeze probability.

Input documentation: Hyperliquid public API · OKX public API · MEXC Futures API · Coinbase International Exchange API · Binance Spot klines · CoinNudge methodology

Observation window: Aligned one-hour leverage and price context plus rolling 15-minute liquidations

Calculation cadence: Auto-refresh about every 300 seconds

Calculation version: squeeze-component-score-2.1

Download current dataset: JSON · CSV Raw values retain the dataset's published precision. Free fair-use limit: 60 requests per minute per IP, shared across all Research JSON and CSV endpoints.

Leverage buildup without directional evidence

Rising OI can raise pressure on both sides, but it cannot identify who is crowded. Visible score ties and rows without side-specific evidence remain unassigned.

Current source health

• research coinbase derivatives: live; last success 2026-09-13 15:39:37 UTC
• research okx derivatives: live; last success 2026-09-13 15:39:43 UTC
• research mexc derivatives: live; last success 2026-09-13 15:39:40 UTC
• hyperliquid: live; last success 2026-09-13 15:39:31 UTC
• binance volume candles: live; last success 2026-09-13 15:35:21 UTC
• okx liquidations: live; last success 2026-09-13 15:39:52 UTC

The score ranks direction-qualified evidence while OI is stable or building. OI buildup is shared pressure and cannot select a side alone; rows tied at the visible 0.1-point precision or lacking side-specific support are separated as direction-undetermined. Falling-OI rows are excluded as active deleveraging. This is not a liquidation forecast or probability, and venue coverage differs.

How to read this page

• Read long_score, short_score and the component states before accepting a selected side.
• Rising OI adds the same base pressure to both scores; it cannot identify long or short crowding by itself.
• Visible score ties and rows without funding, price, liquidation or CVD direction are separated instead of being forced into a side.
• Falling OI is reported as active deleveraging; OKX liquidations and Binance Spot CVD retain their named venue boundaries.

What can this page tell you quickly?

Best for

Finding leverage buildup where direction-specific evidence agrees.

Primary window

One-hour OI, funding, price and CVD plus 15-minute observed liquidations.

Ranking

Direction-qualified support count first; undetermined buildup remains in a separate table.

Do not infer

A score or direction label is not a squeeze probability, target or entry signal.

What creates long-squeeze or short-squeeze risk?

A long squeeze becomes more plausible when long positioning is expensive, exposure has grown and price starts moving against it. A short squeeze is the mirror image: negative funding can indicate crowded shorts while rising price creates adverse pressure. None of those fields is sufficient alone.

Falling OI after a violent move may show leverage already leaving rather than new crowding. That is why the table keeps price, OI, funding, liquidations and spot aggression separate instead of reducing the market to a single unqualified arrow.

How can the squeeze risk score be reproduced?

Long and short scores are the sums of their five published component points, not probabilities. Each component begins at its threshold, contributes half its cap there, scales linearly and stops at the cap once the reading reaches twice the threshold. OI buildup is deliberately added to both sides because it measures exposure growth but not which side owns that exposure.

The API publishes raw and one-decimal scores, direction-specific support counts and every component's long and short points. A direction is published only when the visible scores differ and the leading side has funding, adverse price, same-side liquidation or opposing CVD support. Exact ties, differences hidden by one-decimal display precision and OI-only rows have risk_side=null and appear in the separate undetermined table. The $1 million OI value is an eligibility floor; absolute OI size does not add score points.

What should you do after a market ranks highly?

First verify that at least three components are marked supporting rather than merely available. Then inspect whether price is still accelerating, whether OI continues building and whether order-book depth can absorb forced orders. A squeeze-risk alert without liquidity context can arrive after the easiest part of the move has already occurred.

Finally compare the current market regime and volatility. A leverage imbalance inside broad market stress behaves differently from an isolated candidate during a calm session.

Why does CoinNudge avoid a squeeze probability?

Public data does not reveal every trader's leverage, liquidation price, collateral or cross-margin exposure. Venue OI definitions and coverage differ, and liquidations observed on OKX cannot be expanded into a whole-market total.

Calling the score a probability would create false precision. CoinNudge publishes a transparent ranking of observable pressure so the user can decide what to investigate.

How should undetermined squeeze direction enter a research sample?

Keep it as its own state. Shared OI pressure can raise both scores without providing directional evidence, so assigning the row to a long or short squeeze would create a label unsupported by the inputs.

The component score is not a probability. A score threshold alone cannot establish directional accuracy or a tradable return.

• Store both scores, risk_side, component states and supporting-evidence count.
• Separate qualified directions, undetermined pressure and excluded falling-OI observations.
• Define later stress outcomes before testing and retain the original scoring version.

Verify: Hyperliquid public API · OKX public API · Binance Spot klines · CoinNudge methodology

What is measured, and what is not?

Method and data boundary

CoinNudge begins with perpetual markets carrying at least $1 million of tracked OI and excludes falling-OI rows from this forward-risk ranking because they describe active deleveraging rather than leverage buildup. For each side, five components sum to at most 100 points: funding 20, rising OI 15, adverse price 25, same-side OKX liquidations 25 and opposing Binance Spot CVD 15. A component starts scoring only at 0.01% hourly funding, 3% hourly OI growth, 1% adverse hourly price, $100,000 liquidations or 5% CVD share; its score reaches half its cap at that threshold and is capped at twice the threshold. OI points are shared by both sides and cannot determine direction alone. A side is assigned only when its score remains higher at the published one-decimal precision and at least one side-specific component—funding, price, liquidation or CVD—supports it. Otherwise risk_side is null and the row moves to Leverage buildup without directional evidence. Evidence count includes only components marked supporting for a direction-qualified side; opposite readings are conflicting, sub-threshold readings neutral and unavailable readings missing.

The score ranks direction-qualified leverage evidence while OI is stable or building. OI buildup is shared pressure, so ties at the visible 0.1-point precision and rows without side-specific support are labeled direction-undetermined. Falling-OI markets are excluded as active deleveraging. This is not a liquidation forecast or probability; venue coverage differs and a crowded position can persist.

Read the complete CoinNudge methodology

Sources and verification

• Hyperliquid public APIPerpetual price, funding and OI observations.
• OKX public APIPerpetual OI, funding and observed liquidation context.
• MEXC Futures APIPerpetual price, funding and OI observations.
• Coinbase International Exchange APIPerpetual price, funding and OI observations.
• Binance Spot klinesCompleted taker-buy quote volume used for CVD.
• CoinNudge methodologyAlignment, coverage, thresholds and missing-data rules.

Frequently asked questions

Receive leverage-stress signals in Telegram

Get the crowd side and supporting evidence without treating one funding or OI number as a prediction.