11–26 June, the Deribit trade feed, 76,000+ trades under the microscope. We were looking for an insider who “knew about the news”. Instead we found smart money leaning on something more reliable than news. This person would have lost everything had they flinched one night. They did not flinch.
Every number below is a direct measurement from the public trade feed. Where it is an estimate, we write so: (estimate). This is a journal of decisions, not signals.
Open the case ↓Act 1 · The night of 21→22 June
Sunday. The quietest day of the month on the option board: over the whole day only $2.75M was put down here — the market is asleep. At 20:00 UTC bitcoin starts sliding. By 23:50 it is down 1.44%, the low is $63,250. And the feed shows fear waking up:
A classic picture: frightened at the low — bought protection — the market turned up. Everyone who “traded the news” was late in both directions. And now — the fate of those 273 insurance contracts bought in the darkest minute of the night:
−100% · a miss by $455
Final settlement 26.06: $60,455. The bought insurance started paying below $60,000. The market fell 11% over the week — and the insurance still burned to zero: the payout was short by four hundred and fifty-five dollars. The right direction. The right week. A dead position.
If even calling the fall right can cost you everything, then who actually makes money here?
Act 2 · Nine days earlier
We wound the feed back a week to look for someone who “knew about the news”. Nobody like that was found: not a single bet on a pump before the news (this matters, and we say it plainly). But from 12 June somebody was assembling a position at prices far from the market — where no chart looks. Methodically. In batches. Here is their handwriting:
In one block: sell 500 $75k calls (bets on a rise — to someone who believes the fairy tale) and buy 250 $50k puts. Eighty-nine minutes later, at 09:38:39 — exactly the same pack once more. A mirror structure, the same size. This is not a mood. This is a plan.
The behavioural cluster of “one player” — from timestamps and structure (assumption)
A series of December $120k call sales melting like an iceberg: 365 → 182 → 109 → 73 lots. Someone slices a large order so as not to show up. Proceeds — ~$302k from those who buy the dream of $120k.
Four flat packs of 100: 400× $55k puts bought (~$155k). The same day — 1,000× $73k calls sold, in four packs of 250. The conveyor works its last day.
The feed goes quiet — the player has finished and vanished. The position is assembled: over the week ~5,300 bets on a rise sold (proceeds $656k), ~2,100 bets on a fall bought ($1.41M put in). The net bet against the market: ~$750,000. Now — wait.
Note what he did: the optimists themselves funded half of his bet against themselves. They sold them lottery tickets on a rally — and with the proceeds bought tickets on a fall, with time to spare into September and December.
In the textbooks this is called smart money. In the feed it looks simpler: while everyone watches the chart, somebody unhurriedly assembles a position at prices almost nobody believes in — and where nobody looks.
Act 3 · The test
The night of 21→22 June. Ceasefire news. The market jumps to $65,471. The July $55k puts are now priced by the market at twice as cheap than he was buying them: minus 50% on this part of the position. Half a million dollars of the bet melting on the screen. The feed of 22 June shows their actions with measurement precision:
Nothing. Zero reversal trades.
Telegram shouts “to the moon”, the chart is green. What do you do?
We do not know what you would actually have done — and neither do you, until you have stood at that point with real money. We know what they did: nothing. Because his bet was not against the news — against the market’s structure. And the structure did not change because of one post.
Act 4 · The outcome
Four days after the pump the market does what the structure was prepared for: minus 11%, to $58,054. The puts open up. The sold calls deflate into dust — even at the peak of the pump not one of them came close to the money: his “risk” was not risk for a single minute. The account of the case:
| Part of the bet | Put in | Estimate as of 26.06 | Result |
|---|---|---|---|
| Far puts bought (Jul–Dec) | $1,356k | $2,472k | +$1,116k |
| Far calls sold ($75k–$120k) | $656k collected | buyback $330k | +$326k |
| Cheap 50k “lottery tickets” in a pack | $53k | $0 · burned | −$53k |
| Total | ~$751k net | ~$2,142k | ≈ +$1.39M |
(estimated at the instruments’ latest market prices as of 26.06; this is the aggregate flow of buyers/sellers at prices far from the market, brought together — “one player” inside the clusters is established by behaviour, not by wallet ID; the participants’ full portfolios are unknown)
−100%
It bought protection at the low of the candle, an hour before the reversal. It called the direction of the week right — and still lost everything: a miss by $455.
+185%
It assembled the position over a week, survived −50% on the news without a single trade — and took ≈$1.39M on a move that had been in preparation for weeks (estimate).
They did not know about the post. They knew about gravity: where the market falls when the noise ends.
Act 5 · The limits of the case
This investigation survived three verification passes — including the fact that we burned our own first version of this story when the data did not support it. Here is what we know for sure — and where the limit is:
Smart money builds structures like this constantly — and they are visible in public data if you know where to look. We look there every morning at 08:31: whales on Hyperliquid, the options footprints of smart money on Deribit — who is collecting protection, where the walls are growing, where the liquidation fuel is shifting.
Today’s radar — free → Access levelsThis case took three analysis passes and 76,000 trades. For subscribers this is a daily routine.