$1.97M paid up front on Deribit: now the seller loses while bitcoin stands still
On 19 October someone on Deribit sold 500 March bitcoin calls at the 88,000 strike and bought 500 November calls at 83,000, in a single block. The difference, about $1.97M, landed in their account right away.
Three days later the position is down $148K on paper, even though bitcoin has barely moved. It hurts most if bitcoin sits near 83,000 on 6 November: then the loss is about $1.2M.
And on 21 October another trader bet on the exact opposite for the same morning.
- Source: our hourly archive of every Deribit trade with the aggressor side and block number, since 11 June 2026
- Cross-check: the November 83,000 call was listed on 15 October, and the archive holds all 105 of its trades. Open interest: exchange snapshot of 22 October
- Data frozen 22.10.2026, 07:28 UTC
On 19 October sold 500 March 88,000 calls and bought 500 November 83,000 calls in one block. At least 420 of the 500 pairs are still open.
One block, $1.97M at once
On Monday morning one two-leg trade went through on Deribit. Sold: 500 bitcoin calls at the 88,000 strike expiring 26 March 2027, at 0.0793 BTC each. Bought: 500 calls at 83,000 expiring 6 November, at 0.0314 BTC.
The March calls cost more because they have five months left, the November ones two weeks. That is 0.0479 BTC per pair, 23.95 BTC in total. With bitcoin at 82,420 that is about $1.97M, and the seller got it on the spot.
A trade that buys and sells calls on different expiries and different strikes is called a diagonal. Our register has 234 of them on bitcoin since 11 June, and the next biggest by cash taken in brought 11.4 BTC. So this seller was paid twice as much as anyone before.
Not a bet on direction
Right now both calls react to price almost the same way: each moves about 47 cents per dollar of bitcoin. One is bought, the other sold, so a thousand-dollar move barely changes the seller's account today.
Time does. The November call expires in two weeks and loses value faster every day than the March one. The bought leg sheds about $40K a day, the sold leg brings in about $13K, and the gap, $27K, leaves the account every day bitcoin stands still.
So three days in, at almost the same price, the position is down $148K on paper out of the $1.97M taken in.
Line: the 23.95 BTC taken in plus the value of both legs, $K. Between trades on the legs we use a model with each leg's last known volatility; the last point is at exchange marks on 22.10, 07:28 UTC.
Is it still open
The November 83,000 call was listed on 15 October. Since then 763.5 contracts have traded on it, the block included, and every one of those trades is in our archive. On 22 October its open interest was 700.
To close a contract you have to sell it or buy it back, which means another trade. So out of 763.5 traded, at most 63.5 could have been closes, and neither side of the block could have exited most of its 500.
On the March 88,000 call only 80 contracts traded after the block. So at least 420 of the 500 pairs are still held by whoever made the trade on 19 October.
What it bets on until 6 November
On the morning of 6 November the November calls settle. Every thousand dollars bitcoin sits above 83,000 at that point pays the seller $500K. Below 83,000 they expire worthless.
The March calls stay open and rise with bitcoin. By our model, if settlement lands exactly on 83,000, the November calls expire at zero while the March ones are still expensive: a loss of about $1.2M. If bitcoin is where it is now, the loss is about $1M.
The position turns positive only far from 83,000: below roughly 76,650 or above 87,600, which means a 7% move either way. So the seller is betting neither on a rise nor on a fall, but on bitcoin moving within two weeks.
Blue line: Deribit bitcoin index, hourly from our archive. Dashed lines to 6 November: the 83,000 strike where the seller loses most, and the lines beyond which the position turns positive, by our model.
What is left after 6 November
After 6 November the seller is left with only the sold March leg, for another 140 days. From then on the account depends on direction: at today's prices every $1,000 of bitcoin upside would cost about $237K.
If bitcoin is below 88,000 on 26 March and the seller holds to the end, the March calls expire worthless and all 23.95 BTC stays with them. Above 88,000 they pay out of every thousand.
So the November call covers them only for the first two weeks. After that, until spring, the seller faces bitcoin's upside alone.
Someone bets the other way
On Wednesday morning two blocks a minute apart carried 150 identical structures for the same 6 November morning. Someone sold 83,000 calls and 82,000 puts and covered against a big move by buying 88,000 calls and 77,000 puts. They took in 5.61 BTC, about $464K.
This structure is called a condor. It earns the most if bitcoin settles between 82,000 and 83,000 and loses beyond roughly 78,900 and 86,100. In other words, where the March call seller loses the most, the condor earns the most.
Bets on a rise sit on the same morning too: at least 2,500 call spreads at 90,000 and 95,000, about $1M in total. So at 08:00 UTC on 6 November at least three large bets get their answer at once.
The case is open. We will come back on 6 November after the 08:00 UTC settlement: what the November calls paid and what happened to the March leg. Sooner, if open interest on the November 83,000 call falls below 500 or a large block prints on the March 88,000.
You'll see the next 500 BTC trade on the day it prints
The radar every morning: which blocks went through overnight, where open interest moved and who is betting on what. Free in Telegram, unsubscribe in one click.
Send me the radarSet a beacon on a strike →Next
Evidence and limits
Trades: our hourly Deribit archive since 11 June 2026 with block number and aggressor side. BLOCK-303262, 19.10, 04:01:26 UTC: BTC-26MAR27-88000-C sold 500 at 0.0793 BTC, BTC-6NOV26-83000-C bought 500 at 0.0314 BTC. Neighbours on 6 November: BLOCK-303386 and BLOCK-303387 (21.10), BLOCK-303354 to 303357 (20.10), BLOCK-303373 (21.10).
Cross-check with the exchange: the archive holds all 105 trades of BTC-6NOV26-83000-C from its listing on 15.10 with no gaps, 763.5 contracts in total. Open interest on 22.10, 07:28 UTC: 700.0 on the November and 891.3 on the March call. The exchange's public trade history now returns only the last day or two, so the 19.10 block no longer shows there; trade numbers run in sequence.
Exact accounting: (0.0793 minus 0.0314) × 500 = 23.95 BTC × 82,419.8 = $1.974M. State on 22.10: 500 × (0.026 minus 0.0775) = minus 25.75 BTC, together with the cash taken in minus 1.80 BTC, or minus $147.9K. "420 of 500" = 500 minus the 80 March contracts traded after the block.
Model for 6 November: Black formula, March call volatility held flat (37.4% annualised on 22.10), forward from the March future basis; on 22.10 the model prices the March call 2% above the exchange mark. The 76,650 and 87,600 lines and the $1.18M loss at 83,000 shift if volatility changes: each point of it is worth about $76K to the position.
Limits: the exchange does not reveal identity; we follow the side that, by the exchange's flag, sold March and bought November. We cannot see that side's spot bitcoin or futures: if it holds bitcoin, the sold March leg is covered. We do not link the condor or the call spreads to this trade.
A decision journal, not investment advice. Reconstructed from public exchange records.