Open interest, funding & liquidations

The stats strip carries the positioning context for the pair you’re watching. Price tells you what the market did; these four numbers tell you who is positioned where, and how much of it is forced to act if price moves. That’s the question the chart alone can’t answer.

Open interest

Total outstanding contracts - every open long matched by an open short. OI doesn’t move when positions change hands; it moves when positions are created or destroyed, which is why it separates conviction from churn. Read it against price:

• Price up, OI up - new longs driving the move. Fresh conviction, but also fresh liability: every one of those positions has a liquidation price below.
• Price up, OI down - shorts covering. The rally is positions closing, which means it can run hard and stop dead: covering exhausts itself.
• Price down, OI up - new shorts pressing. Same conviction/liability trade as the first case, mirrored.
• Price down, OI down - longs bailing or being liquidated. If it’s fast and disorderly, you’re watching forced flow, not decisions.

OI building steadily while price grinds at a high is the raw material of a cascade: a stack of leveraged longs accumulating, all of whose breaking points sit below, in exactly the zones the liquidation heatmap draws. On liquid pairs, real OI is also one of the inputs that turns heatmap estimates into confirmed Liq Levels.

Funding + countdown

The periodic payment between longs and shorts that tethers the perp to the index, with time to the next exchange funding event. The sign tells you which side is crowded and paying for the privilege:

• Persistently positive - longs paying shorts to stay in. A crowded, impatient side.
• Persistently negative - shorts paying. Fuel for squeezes: a heavily negative pair that starts moving up forces shorts to cover into their own crowd.

Level matters less than persistence and change. Ordinary funding wobbles constantly; funding that stays pinned at an extreme through the countdown, reset after reset, is a market leaning hard on one side of the boat.

Long/short ratio

The split of positioned accounts (e.g. 52 / 48). It counts accounts, not notional - a thousand small longs and one whale short can read as extremely long-skewed. That bias is exactly what makes it useful: it approximates what the retail crowd is doing, which is frequently the side that ends up as fuel.

24h liquidations

Total forced closures on each side (e.g. L $109.5M · S $55.4M). A lopsided number tells you who has already been hurt - spent fuel, not standing fuel. Heavy long liquidations mean the leveraged longs are largely gone; the map below price has been consumed, and continued downside has less forced flow to feed on. The heatmap shows the fuel still standing; this number is the burn record.

Reading them together

A worked example: price grinding upward · OI climbing · funding pinned positive · ratio long-heavy · long liquidations near zero. Each number alone is unremarkable. Together: a crowded, leveraged, unburned long structure - a market where the interesting risk is below, at the clusters the heatmap will be drawing brighter by the hour. None of these predicts anything alone; stacked, they tell you which side of the market is brittle.

All four are recorded and replayed with everything else, so you can rewind any archived event and watch how positioning built before it broke.

Searching positioning across the record

The terminal gives you the latest positioning state. Research makes that state searchable across every eligible minute using five frozen features:

• OI velocity percentile rank asks whether open interest was changing unusually fast relative to that market’s own recent history. Missing or stale 5-minute references stay absent rather than becoming a false zero.
• Liquidated notional over the trailing hour is the direct USD sum of confirmed forced liquidations. It answers questions such as “when was at least $5 million liquidated in one hour?”
• Top-trader long ratio is the share of top-trader accounts that were long.
• Global long ratio is the equivalent account share across the wider market.
• Top-versus-global long skew subtracts the global ratio from the top-trader ratio. Positive means top traders were more long than the crowd; negative means the crowd was more long.

These are account ratios, not notional exposure, and divergence is not proof that either group is right. The value is in making a positioning hypothesis explicit, counting every occurrence, inspecting the full outcome distribution, and opening the matched moments in replay.