$3.95M on one strike. A day later, +$2M. Four days later, zero

Bitcoin sits at 79,400. In the afternoon a block prints in the Deribit feed: 1,000 calls at the 81,000 strike expiring 9 September sold, 1,000 of the same expiring 11 September bought. Same second, one trade. 37 seconds later, another identical block. Within the hour, two more, less than a minute apart.

That's a calendar spread: the short leg pays for part of the long one. The sold calls brought in 7.0 BTC, the bought ones cost 26.0 BTC. Net 19.0 BTC, about $1.51M.

The bet is simple: the price stays under 81,000 through Wednesday and gets there by Friday.

In the morning the first trade of the day prints on the freshly listed 12 September expiry: 1,250 calls at 81,000, bought at market. Five minutes later, another 1,250. Then almost three hours of silence. Just before noon, two more clips of 1,250, a minute apart.

5,000 contracts in all, 31.1 BTC, about $2.44M. The volatility he paid crept higher with every clip: 40.6, 42.2, 43.9, 44.1. After each pair, small trades were already printing at 38-40. So he was paying for speed.

That day 18,422 bitcoin options contracts traded across all of Deribit. 5,304 of them were this one instrument. On the Saturday expiry this strike grew to 60.7% of all open interest. Our radar flagged the wall at 81,000 three mornings in a row.

Overnight bitcoin climbs. At 08:00 UTC it's at 79,688, and in that very minute the sold leg expires worthless: the 9 September expiry settles at 79,224, below the strike. For a calendar that's the best possible outcome.

At that moment the two bought legs together are worth almost $6M at the exchange's mark. Until noon they hold around $5.5M. So on paper he's up $1.5-2M in a day.

No big sales in the feed: not one trade above 200 contracts on these expiries. Open interest at the strike moves from 5,060 to 5,047 over the day.

By the evening of 9 September price is back at 78,300 and the bet is worth $2.5M: it's no longer worth what was paid for it. On 10 September bitcoin is at 77,155 and the bet is worth $0.6M.

Two days to expiry, 5% to the strike. The options market prices the move over those two days at about 3.3%. Open interest at the strike: 5,032. In three days just 28 contracts of the 5,060 left.

The last expiry settles at 77,269, 4.8% short of the strike. So 5,000 calls expire worthless.

The structure's result, if held to the end: 50.1 BTC paid, zero back. About $3.95M. A bought option isn't liquidated and never needs more margin: everything he could lose was known the minute he bought.

A big pile of calls at one strike usually gets read as a ceiling: price hits it and bounces. That read quietly assumes a client sold the calls and a market maker holds them.

Here the trades point the other way. The client bought at market, so the market maker sold and was left short calls. A market maker in that spot has to buy bitcoin as price moves toward the strike, which pushes price up instead of holding it back.

Open interest alone can't tell these two cases apart. You need the direction of each trade, and it sits in the public feed. The 81,000 wall was somebody's bet on a breakout. It never paid off.