Advanced level
Derivatives dashboards show how traders are positioned, not where price will go. Open interest counts contracts that are still open, funding shows which side is paying to hold its position, liquidation totals count forced closes at notional value, and options data shows what protection and upside cost. Each number has a measurement quirk, and each reads best alongside price and the others, as context and never as a signal on its own.
Start here: every long has a short
A derivatives contract is created when one trader goes long and another goes short. There is never more long exposure than short exposure; the two always match, contract for contract. So no dashboard can tell you that "everyone is long". What the data can hint at is who has been aggressive, crossing the spread to get in, who is paying to stay in, and who is under pressure because the price has moved against them.
Keep three habits throughout. Check the unit before the number. Normalize before comparing venues. And treat every reading as a probability shift in a story you can test, not as a verdict. None of what follows is a trading signal, and none of it is advice.
Open interest, and the unit trap
Open interest is the number of contracts that are open: one long matched with one short counts once. It rises when new positions open on both sides and falls when positions close, whether voluntarily or by liquidation. Volume, by contrast, counts every trade, including ones that simply pass a position from one trader to another.
The trap is that most dashboards quote open interest in dollars. Dollar open interest moves with price even when nobody opens anything:
| Open interest in BTC | Price | Open interest in dollars | |
|---|---|---|---|
| Monday | 100,000 | 60,000 | 6.0 billion |
| Friday, no new positions | 100,000 | 66,000 | 6.6 billion |
A 10% rise in dollar open interest here is purely the price. Always look at open interest in coin terms, or divide the dollar figure by the price, before concluding that leverage is building. Aggregated charts that add up venues also add up different contract sizes, margin currencies and user bases, so treat the total as a rough gauge.
Price and open interest together
Open interest means little on its own. Paired with price it suggests which side has been pushing, with the emphasis on suggests.
| Price | Open interest | Often consistent with | Also possible |
|---|---|---|---|
| Rising | Rising | New longs opening aggressively; new shorts absorbing them | Short sellers adding into strength, which can fuel a squeeze later |
| Rising | Falling | Shorts closing or being liquidated | Longs taking profit into the rally at the same time |
| Falling | Rising | New shorts opening aggressively | Longs averaging down, which leaves more to liquidate |
| Falling | Falling | Longs closing or being liquidated | A general deleveraging where both sides step back |
A rally on falling open interest is often called a short squeeze, and a sell off on falling open interest a long flush. Both labels are guesses about the aggressor. The same pattern can come from basis traders unwinding a hedge, which moves open interest without any directional view at all. Liquidation data, below, is the best cross check.
Funding as a positioning gauge
Funding is the cleanest positioning data there is, because it is a price actually paid. When longs are crowded, the perp trades above the index and longs pay shorts to hold their positions; when shorts are crowded, the reverse. The mechanics are in the perps guide. Three rules for reading it:
- Normalize the interval. Binance and Bybit usually quote per 8 hours; Hyperliquid and dYdX pay hourly. Multiply by intervals per year to annualize: 1,095 at 8 hours, 8,760 at one hour.
- Know the neutral level. On venues using the standard clamp, a balanced market settles at 0.01% per 8 hours, about 10.95% a year paid by longs. That is the resting state, not a bullish reading.
- Weight by size. A small venue at an extreme rate matters less than a large one slightly above neutral. Open interest weighted funding is the fairer summary.
| Quoted rate | Interval | Annualized | Reads as |
|---|---|---|---|
| 0.0100% | 8 hours | 10.95% | Neutral on clamp based venues |
| 0.0013% | 1 hour | 11.39% | About the same neutral level, hourly |
| 0.0500% | 8 hours | 54.75% | Longs paying heavily to stay in |
| 0.0100% | 1 hour | 87.60% | Very crowded long, easy to misread as normal |
| −0.0300% | 8 hours | −32.85% | Shorts paying, often after a sharp fall |
Persistently high funding says leverage is paying up to stay long, which leaves the market with more positions that must be sold if price drops. Deeply negative funding says the same about shorts. Funding also caps out and lags: it is an average over an interval, and on fast days price moves first.
Liquidation data: totals, feeds and heatmaps
Liquidation totals are reported at notional, the full size of the positions closed, not the money lost. An isolated 10x position with 1 million dollars of notional that is liquidated cost its owner at most about 100,000 dollars of margin. That is why the headline of October 10, 2025, more than 19 billion dollars liquidated in 24 hours, the largest day on record, overstates what traders actually lost by a wide margin, even though it was a real and brutal deleveraging: total perp open interest fell about 43% in the selloff, from about 217 billion to 123 billion dollars, according to CoinDesk research.
The feeds are also incomplete by design. Binance’s public liquidation streams push, for each symbol, only the latest liquidation order within each 1,000 millisecond window, as its API documentation states. In a cascade, dozens of liquidations a second on one symbol reach the public as one message. Aggregators that build totals from these feeds undercount the busiest moments, and different aggregators fill the gap differently, which is why their totals disagree.
Liquidation heatmaps go further and estimate where future liquidations sit. A typical model takes changes in open interest at each price, assumes a spread of leverage levels, and projects liquidation prices from there.
Long short ratios and their bias
Since total long and short exposure are always equal, a "long short ratio" has to measure something else. Most count accounts: the share of a venue’s accounts that are net long versus net short. Some exchanges also publish a ratio for their largest traders, by accounts or by position size.
- Accounts are not money. Ten thousand small accounts long and one large account short is a ratio of 10,000 to 1 with balanced exposure.
- Each venue has its own crowd. A ratio describes that exchange’s users, not the whole market.
- Hedges look like bets. A fund short perps against spot shows up as a bearish account.
Used carefully, a ratio that moves sharply away from its own history on one venue is worth a note. As an absolute reading of market sentiment, it is weak.
Options data: put call ratio, skew, term structure and max pain
Options data shows what traders are paying for protection and for upside. The options guide explains the pricing behind each measure.
Put call ratio
Put open interest divided by call open interest, or the same with volume. With 7,000 puts and 9,000 calls open the ratio is 0.78. A high ratio is not necessarily bearish: many puts are insurance bought by holders who are long spot, and many calls are sold by holders running covered calls. The ratio says nothing about who bought and who sold.
Skew
The 25 delta risk reversal is the implied volatility of the 25 delta call minus that of the 25 delta put. At 48% for the call and 55% for the put it is −7 points: the market is paying more for downside protection than for upside. Its direction and its change over time carry more information than any single reading.
Term structure
Implied volatility by expiry. An upward slope, where later expiries cost more, is the usual state in calm markets. When the front end jumps above the back, the market is paying for near term movement, often around stress or a scheduled event. Deribit’s DVOL index condenses 30 day implied volatility into one number; divide it by √365 for the implied daily move.
Max pain, and why it is weak
Max pain is the settlement price at which option buyers, as a group, would collect the least. Using a toy set of open interest in 1 BTC contracts:
| Strike | Call OI | Put OI | Total paid to buyers if settled here |
|---|---|---|---|
| 55,000 | 500 | 3,500 | 23.5 million |
| 60,000 | 1,500 | 2,500 | 8.5 million (max pain) |
| 65,000 | 3,000 | 800 | 13.5 million |
| 70,000 | 4,000 | 200 | 37.5 million |
The theory says sellers will push price toward 60,000 into expiry. The flaws are serious. It assumes all sellers are one coordinated group with the means and the motive to move spot, when many option sellers are market makers who hedge their delta or holders writing covered calls against coins they own. It ignores futures and spot positions held against the options. And open interest changes all the way to expiry. Pinning near a heavy strike can happen when dealers are positioned to hedge toward it, but max pain is a description of open interest, not a force. Deribit’s large expiries settle at 08:00 UTC on Fridays, which is when this debate peaks.
Basis as a sentiment reading
The annualized basis on dated futures, and the premium of perps over the index, are both prices for leverage. CF Benchmarks found in October 2025 that the CME bitcoin basis correlated with momentum indicators at about 0.50 to 0.54, and concluded it is shaped more by momentum and sentiment than by the cost of carry. It ran near 25% annualized in February 2024 and briefly dipped below zero in March 2025. Read it the same way as funding: a rich basis says demand for leveraged longs is high and the market holds more positions that need rising prices; a basis near or below Treasury bill yields says that demand has drained away. Since spot ETFs arrived, part of that demand is met by hedged funds, so read the basis beside ETF flows. The basis guide covers the trade on the other side of it.
A routine that keeps you honest
Fix the units
Open interest in coins, funding annualized on a common interval, liquidations as notional and, where you can, as margin.
Read the series together
Price, open interest, funding and liquidations tell a coherent story or they do not. A rally with rising open interest, rising funding and short liquidations hangs together; one metric alone rarely does.
Compare with its own history
A reading that is extreme relative to the last few months on the same venue says more than a comparison with another venue’s number.
Write down what would change your mind
Positioning data is best used to ask what would hurt the crowded side, and whether that is already happening.
Set it beside the wider picture
Sentiment gauges such as the fear and greed meter and chart based indicators give different angles on the same crowd.
Common questions
Does rising open interest mean the price will go up?
No. Rising open interest means more contracts are open, with a long and a short on each. Paired with a rising price it is often consistent with aggressive new longs, but the same pattern can come from shorts adding into strength or from basis traders hedging. It describes leverage building, not direction.
Why do liquidation totals differ between data sites?
Because public feeds are incomplete and each aggregator fills the gaps its own way. Binance, for example, publishes only the latest liquidation per symbol in each one second window, so bursts are undercounted. Sites also differ in which venues they cover and how they value notional.
What funding rate counts as neutral?
On venues that use the standard clamp, 0.01% per 8 hours, which is about 10.95% a year paid by longs to shorts. Hourly venues are neutral near 0.00125% an hour. Always normalize the interval before calling a rate high or low.
Is max pain where the price will settle?
Not reliably. It is the strike where option buyers would collect least, which assumes option sellers are a single group willing and able to move spot. Most sellers hedge their exposure or own the underlying, so max pain describes open interest rather than predicting settlement.
Is a high put call ratio bearish?
Not by itself. Puts are often bought as insurance by holders who are long the asset, and calls are often sold against holdings. Skew, which shows what traders pay for each side, usually says more.
Where to go next
Next: Oracles: how prices get onchain and how they fail
Guide 9 of 16 in the Advanced level. Your place is kept on this device, so you can pick up exactly here next time.