Two radars in a row saw the whales' short tilt meltingon the same 139 addresses it grew both days.
Case #20260926 · 26.09.2026 06:47 UTC · BTC$83,908 / ETH$2,684 · analysis: Opus 5.5 by Anthropic
📅YESTERDAY
Friday's 25.09 quarterly settled at 08:00 UTC: BTC$83,931, ETH$2,671. The market-maker break-even corridor held in both assets for a third day running. Max pain, fixed a day before expiry, lay far below — $78,000 in BTC and $2,300 in ETH: 7% and 14% under the settlement price.
The registry closed 73 quarterly cases at the official settlement price: −$13.62M for the whales in total, 20 cases in profit, 53 at a loss. Another 37 cases have legs on later expiries and stay open.
⚡THE GIST
• The quarterly cost the whales $13.62M across 73 cases. The costliest were summer positions on a fall, opened when bitcoin stood near $60–65k. And the 13 spreads of the week that we called financing operations yesterday gave only −$41.5k together — they really weren't a position on a fall.
• The whales' short tilt in bitcoin grew over the day: on a constant set of 139 addresses — from $299M to $334M. The main reason is that the 35× long, which yesterday stood less than 2% from liquidation, closed on its own, without a liquidation. In ether the tilt on the same set grew too — from $270M to $312M.
• A correction to the last two radars: "the short tilt is melting" was an effect of our sample — big wallets were entering and leaving it. On a constant set of addresses the tilt grew yesterday as well, from $329M to $355M. The sum over the whole registry shows $280M today only because a large 40× long, held since 24.09, entered it for the first time.
• The funds bought on all five days of the week: $2.39B went into the BTC ETFs over 21–25.09, the 97th percentile of our measurements. But the daily inflow shrank every day — from $999M on Monday to $134M on Friday.
• Verdict — break ⚡: for two days now the whales' short tilt hasn't been melting — on a constant set of addresses it is growing.
• 📊 Our Polygon — Pendulum: the best of four variants, +$120 on the day, +$9,611 in total since 11.05.
• 📊 Our Polygon — signals: best — Dust Strategy V2 +$83,969 at a 4% win rate · worst — Wheel Trail V8 −$9,583. Table at the bottom.
🐋WHALES: WHAT CHANGED — three arenas
Arena one — Hyperliquid futures. Today we measure the change on a constant set of addresses. We track 248 wallets, but the list shifts a little every day: someone drops out of the snapshot, someone lands in it for the first time. So we count the day's change only across the 139 addresses present in both morning snapshots — at 06:16 UTC yesterday and today (how we track whale positions). In BTC on this set the long side slimmed from $286M to $220M — a 35× long closed — and the short side from $585M to $553M. The net short tilt is $334M against $299M: up by $35M. Yesterday, on the same basis, it grew too — from $329M to $355M. In ETH the short grew from $707M to $765M, the long from $437M to $454M; net $312M against $270M. Across the whole registry the sum today is: BTC long $280M against a short of $560M, ETH$435M against $793M. Overleveraged whales number 66 against 63 yesterday, $1.15B of positions together, 67% of them longs against 75%. Among large positions over $10M only one is still near its liquidation line — a 40× short: overnight it cut its position by 60% and still stands about 2% above the market.
Arena two — Deribit options. The quarterly is gone, and the options market is lighter. In bitcoin 351 thousand contracts remain open; the two heaviest dates now are 30.10 and 25.12, with 117–119 thousand contracts each. The registry closed 73 quarterly cases and took in not a single new one over the day. Quiet accumulation outside the blocks is the same for a sixth day (what a block is): October calls above the market. In ether, over the week, both calls above the market and puts below it were gathered — spread across eleven to twelve expiries, with no single main date.
Arena three — money at the exchange border. The snapshot is fresh — up to this morning. Over seven days $155.7M was brought into the exchange and $100.5M taken out. The bulk is USDC cash via the bridge: $123.4M in against $56.9M out. Bitcoin and ether, as yesterday, were taken out: BTC$19.4M out against $6.9M in, ETH$13.0M out against $5.5M in. More than four fifths of that outflow went to addresses of a single player.
This is positioning, not a direction forecast.
🧭CROSS-CHECK — five camps of witnesses
What the funds are doing. Spot ETFs are the only camp that measures purchases of the coin itself, not a leveraged position. On Friday $134M went into the BTC funds and $87M into the ETH funds. The funds are closed until Monday, so Friday 25.09 was the week's last session. The calendar week of 21–25.09 in bitcoin closed at +$2.39B against +$6M the week before, the 97th percentile of our measurements; in ether +$690M against −$140M, the 95th. Meanwhile the daily inflow into the BTC funds shrank every day: $999M, $715M, $347M, $191M, $134M. This camp stands against the whales: the funds bought on every day of the week, while the whales added to their short tilt over the day. Still, the daily inflow fell more than sevenfold during the week.
What the options market says. This camp measures one thing: what insurance costs on Deribit. BTCskew−0.8 against +0.2 yesterday: calls are now slightly dearer than puts (what skew is). In ETH−0.2 against +0.1. Insurance got cheaper still: BTC two-day vol sits in the 1st percentile since February — cheaper than in 99% of hours — against the 14th yesterday, DVOL34.6 against 36.0 (what implied vol is). The premium over the actual move on the monthly window is −2.0. Our fear-and-greed index stands at 79 out of 100 against 72 in yesterday's radar. This camp stands against the whales: the market no longer pays extra for protection against a fall, and calls have moved ahead — while the whales' short tilt keeps growing.
What the futures say. This camp measures who pays whom to hold a position. BTC funding is +0.005% on Hyperliquid in eight-hour terms — below the venue's base rate of +0.010%; on Binance +0.005%, same as yesterday. In ETH+0.010% and +0.009% (what funding is). This camp is silent: funding sits near its neutral base, and no demand for leveraged longs is visible.
What the crowd is doing. This camp measures how retail accounts stand against the whales. On Bybit perps 57% of BTC accounts stand long against 56% yesterday; in ETH67%, same as yesterday. On the constant set of addresses the whales' long share in BTC fell from 33% to 28%, so the gap between the crowd and the whales widened. The sum over the whole registry shows the opposite — only because of the 40× long that entered the sample for the first time. This camp is silent: the crowd moved by a point; it is the whales who are widening the gap.
Any signs of stress. This camp measures whether anyone has already been forced out of the market. Among the top Hyperliquid whales there was not a single forced closure over the day: the 35× long left on its own. On OKX swaps $421k was liquidated, a little over half of it shorts; that is 6% of the week's volume. This camp is silent: no cascade.
Again: this is positioning, not a direction forecast.
🌡ALTSEASON — the classic and our barometer
The classic altseason index stands at 50.6 out of 100 against 34.8 in yesterday's radar — but the coin list has changed since: only pairs with more than $1M of daily turnover enter the count, and today there are 257 such alts against 319 yesterday morning. On today's list yesterday's value is 51.8, so over the day the index didn't rise but slipped a little. This is the transition zone: altseason starts at 75, bitcoin season sits below 25.
Our barometer watches money, not prices: open interest, funding and retail come from Bybit perps, the whales from the Hyperliquid registry. Alts' share of open interest is 46.0% — $5.93B of $12.89B, against 43.4% yesterday.
Alt contracts with funding hotter than bitcoin's make up 75% of open interest. The bar here is BTC's own funding, and on Bybit today it is almost zero — 0.006% a day, so even moderate alts come out hotter than it.
Retail in the top-10 coins stands 67.7% long; the hottest are DOGE and XRP at 76–77%.
Whales in alts are structurally short: 30.9% long across $1.72B of positions, the biggest of them HYPE. That gap itself is permanent, the signal is in its CHANGE — over the day the whales' long share fell by a point.
Full barometer: indiciadesk.com/en/altseason
⚖️VERDICT OF THE DAY
break ⚡ — the whales' short tilt in bitcoin hasn't been melting for two days: on a constant set of addresses it grew from $329M to $355M yesterday and from $299M to $334M today. What we twice called melting was big wallets entering and leaving our sample. Witnesses: the funds and options stand against the whales; futures, the crowd and stress are silent.
What this does NOT mean. This is not a claim that the market will go down. The tilt grew mainly because the 35× long closed on its own, not because someone opened a big short. The market's biggest short — a pair of addresses of one house carrying a carry trade — again barely moved and makes up 69% of the bitcoin short side. And the large 40× long stands above dense long fuel: if price slips 2–4%, it is longs that will be knocked out first.
🔒In today's full RADAR:
— the quarterly case by case: the whales' biggest loss and biggest gain — strikes, size and entry date;
— the 35× long that left without a liquidation, and the 40× long our sample didn't see for two days: size, account, liquidation line.
ANALYST access → indiciadesk.com/en/agent
💎DEEP
🔒WHALES IN DETAIL 💠
Portraits from the Hyperliquid registry, snapshot 06:16 UTC at BTC$83,874 and ETH$2,685; the liquidation lines of the two 40× longs were checked live at 06:36, the trade history via the exchange's public API. Nicknames aren't treated as identification.
The market's biggest loss.ETH SHORT $280.7M, leverage 5×, entry $2,304, uPnL −$39.9M, liquidation at $4,218 — 57% above spot. Over the day — almost no change in coins. The same address in BTC. SHORT $237.2M, entry $75,949, uPnL −$22.4M, liquidation at $140,996 (+68%), account $156.7M. A second address with the same pattern.BTC SHORT $146.9M and ETH SHORT $208.3M, both at 5×, uPnL −$16.5M and −$26.9M, account $122.1M. Together the two addresses carry $105.7M of unrealised loss against $107.1M yesterday; margin across the accounts is $278.8M against $263.7M yesterday — $15.1M more, although the loss shrank by only $1.4M. Their $384M is 69% of the bitcoin short side across the whole registry. The 35× long — how it left. Yesterday it was LONG $88.1M from $84,178 with a $2.4M account. During 25.09 the address partly cut bitcoin four times, closed it fully three times and reopened it twice; the last exit came around 21:15 UTC. Among its trades over the day there is not a single liquidation: every closure was its own order. The address now holds only NEAR worth $5.6M, account $0.39M — and no deposit or withdrawal since 25.09. So over the day the account slimmed by ~$2M on trading, including the open NEAR position. The 40× long we didn't see.BTC LONG $98.1M from $83,823, liquidation at $80,729 — less than 4% below spot, account $5.8M. The position isn't new: by its trade history the address began buying on 24.09 around 10:00 UTC and built the full size by 01:10 UTC on 25.09. The wallet entered our sample only today at 00:46 UTC — and it alone produced the $98M of "new long" in the registry sum. The 40× short stepped back. Overnight it cut its position in two steps from 500 to 200BTC: now SHORT $16.8M from $83,135, liquidation at $85,605 (+2%), account $0.55M against $1.2M yesterday. The 40× long trimmed.BTC LONG $37.7M from $84,373, liquidation at $80,113 (−4%), account $2.5M: 450BTC against 475 in yesterday's radar, although during 25.09 it reached 725. Who added short in ether. One address flipped from a $17.1M long to a $16.0M short at 20×; two more topped up: to $56.5M at 25× and to $52.2M at 15×. Another $26.9M of short in the registry sum comes from an address that entered the sample for the first time; in fact it had been cutting its short. The ether longs. The market's biggest gain is ETH LONG $81.3M at 20× from $2,134, uPnL +$16.7M. A 10× long from $2,694 at $110.7M is $355k in the red, liquidation 6% below spot. The 25× long — $88.7M from $2,667, uPnL +$600k, liquidation at $2,536 (−6%). Fragility.66 overleveraged whales against 63 yesterday, $1.15B of positions together, 67% of them longs against 75%. Formally nobody is near a margin call.
Cases in the registry — the quarterly is closed.73 cases (65 in bitcoin, 8 in ether) closed at the official price: −$13.62M for the whales, 20 in profit and 53 at a loss. The heaviest losses came from bearish call spreads — 27 cases, −$6.18M, mostly June and July ones with calls sold at $70–79k — and from risk reversals: 12 cases, −$4.84M. Case #500 · BTC risk reversal: 250 puts at $55,000 bought, 250 calls at $72,000 sold, expiry 25.09, entered 03.07 at $62,498, net premium ≈2.6BTC paid (≈$162k). The settlement came in $11,931 above the sold call — −$3.15M, the quarterly's biggest loss.
Case #452 · BTC risk reversal: 100 puts at $45,000 sold, 100 calls at $75,000 bought, expiry 25.09, entered 01.07 at $60,295, net premium ≈0.22BTC received (≈$13k). The mirror position, on a rise — +$0.91M, the quarterly's biggest gain.
Профіль виплат на експірацію 2026-09-25. Кит отримав $13k премії. Беззбитковість: $44 867.
Case #1120 · BTC bear call spread: 190.5 calls at $68,000 sold and 190.5 calls at $80,000 bought, expiry 25.09, premium ≈27.0BTC received (≈$2.27M) — almost the full width of the spread. Result −$15k. Together the 13 such spreads since 23.09 gave −$41.5k, while #1121, which stood out of the money, made +$24.4k: the money flowed through these trades rather than being made on them.
Профіль виплат на експірацію 2026-09-25. Кит отримав $2.3M премії. Беззбитковість: $79 922.
Case #948 · ETH condor: 2,750 puts at $1,700 bought, 2,750 puts at $1,900 sold, 3,000 calls at $2,800 sold, 3,000 calls at $3,000 bought, expiry 25.09, net premium ≈36.8ETH received. The $2,671 settlement landed between the sold strikes — the whale kept the premium in full, +$90.6k, on day 31.
Профіль виплат на експірацію 2026-09-25. Кит отримав $91k премії. Беззбитковість: $1 867 і $2 830.
Case #1080 · BTC risk reversal: 100 calls at $70,000 sold and 100 puts at $70,000 bought, expiry Friday 30.10.2026 — a synthetic short from $70,000, premium ≈8.22BTC received (≈$624k), day eleven. From the $75,860 entry spot has moved 10.6% against this position.
Профіль виплат на експірацію 2026-10-30. Кит отримав $624k премії. Беззбитковість: $76 236.
ΔOI over the week:BTC is calls only and almost all in the 30.10 expiry: +23,062 at $95,000 (94% of it in 30.10), +19,535 at $90,000, +12,058 at $100,000. ETH: +16,822 calls at $2,800 and +15,997 at $3,000 against +15,893 puts at $2,500.
🔬Who exactly
Both of the biggest shorts are Abraxas Capital Mgmt (Heka Funds): over the day the house's positions barely changed; unrealised loss $105.7M, margin $278.8M.
Machi Big Brother — the 25× ether long and a separate BTC LONG $34.6M at 40× with its liquidation line 14% below spot.
An address labelled Fasanara Capital holds ETH SHORT $59.1M at 25× (uPnL −$2.5M) and flipped bitcoin over the day: from a $4.3M long to a $20.8M short.
The 40× long that entered the sample for the first time, the 35× long that left, the 40× long that trimmed and the 40× short are addresses without a public label.
Money by address. Over seven days $16.9M of bitcoin out of $19.4M of total BTC outflow went to addresses of Wintermute, and $10.3M of ether out of $13.0M.
🔒THE WEEK'S DYNAMICS 🔬
Perps on a constant set of addresses.BTC: from 23 to 24.09 the short tilt fell from $621M to $357M — that time the shorts really did close, by $266M; from 24 to 25.09 it grew from $329M to $355M; from 25 to 26.09 — from $299M to $334M. ETH: from $374M to $318M, then from $318M to $277M, now from $270M to $312M. We don't give the usual weekly and 24-hour shifts today: they are computed across the whole sample, whose makeup has changed.
Gamma flip.BTC across all expiries $71,752 against $76,664 yesterday: a third of bitcoin's options open interest left with the expiry. Spot is 17% above it (what the flip is).
The price of fear. On the weekly window the premium is zero (IV34.7 against the actual weekly HV34.7) against +1.2 yesterday; on the monthly window −2.0.
Changing of the guard. Over the week the whales assembled 355 confirmed structures: 126 bullish, 135 bearish, 25 on volatility, 69 on range. Bearish outnumber bullish for a third day, and part of them came from the wave of financing spreads under the quarterly.
🔒CROSS-CHECK IN DETAIL 💠
BTC levels. Everything below comes from Deribit options open interest unless stated otherwise. The nearest expiry is today's Saturday one, only 6,096 contracts: the market-maker break-even corridor is $83,500–$86,500, spot near its lower edge (what the corridor is); max pain$85,000 on Deribit and $84,500 on the combined two-exchange open interest (what max pain is); put $83,000 (829 contracts), call $86,000 (781) (what the walls are).
Further down the calendar. Sunday 27.09 — 4,208 contracts. The first notable expiry is Friday 02.10, 25,702. Friday 30.10 — 117,030 contracts, max pain$74,000 and a call wall at $95,000 on 23,300; Friday 25.12 — 118,791.
ETH levels. Today's expiry: corridor $2,660–$2,740, max pain$2,700, put $2,660 (2,807), call $2,800 (6,560).
Fuel. Under BTC$162.7M of long positions within 5% of the market ($79,736–83,933), but in the nearest 2% only $10.8M against $125.7M yesterday: the 35× long left, and the 40× long has its line deeper. Above the market $24.4M of shorts within 5%, $23.1M of them in the nearest 2%, and most of that is the same 40× short. In ETH$25.0M below the market against $2.8M above it. This is about the RANGE of a move for the same push, not about a higher probability of that move (what the fuel map is).
Gamma vacuum. In BTC it is empty below at $77,000–$78,000 and above at $92,000–$94,000.
Stress test ±2%. On a 2% fall BTC runs into the densest put wall — the move is damped. On a 2% rise it stays under the call wall, and the ceiling holds.
OI flow.BTC put/call across all expiries: month 0.59 · week 0.56 · now 0.52 against 0.60 yesterday — mostly because puts burned off together with the quarterly. ETH0.56 · 0.52 · 0.53 (what put/call is).
Two options venues. On the same expiries up to 30 days BTC put/call is 1.12 on Deribit against 0.87 on Bybit — yesterday it was 0.93 against 1.26. The quarterly has left the window, so different expiries are being compared now; we won't treat this as a turn in sentiment. ETH0.91 against 0.90.
IV by tenor.BTC 2d P19/C19 · 6d P31/C29 · 34d P35/C34, percentiles since February 1st, 6th and 10th. ETH 2d P26/C26 · 6d P40/C40 · 34d P48/C48, percentiles 1st, 4th and 13th. BTCDVOL34.6 against 30-day realised 36.6 — premium −2.0; ETHDVOL48.3 with HV 49.7, −1.4 (what the premium is).
💎SCENARIOS IN DETAIL 💠
The code's verdicts. Two scenarios from 23.09 reached their deadline: the corridor was not confirmed — price stayed inside only 12% of the time; the slide into the long fuel was confirmed — a minimum of $83,197 against the $84,750 trigger. Three from 24.09 are closed too: the corridor was confirmed — price stayed inside the whole time; the slide was not confirmed — a minimum of $83,197 against $82,150; the squeeze up was not confirmed — a maximum of $84,766 against $85,500. Three scenarios from 25.09 run until tomorrow, Sunday 27.09, 06:34 UTC.
The all-time tally (N counts only those that reached a verdict): squeeze up through the short fuel — 12 of 27, 9 more cancelled at their level; corridor — 12 of 27 plus 6 partial, 5 more cancelled; slide into the long fuel — 7 of 15, 10 more cancelled. The directional types are still no better than a coin toss.
The corridor holds — 50%. ⏱ 2 days, to Monday 28.09. BTC stays between the dense band of long fuel and the 40× short's liquidation line: $82,250–$85,600. Cancellation level: a touch of $79,750 — the bottom edge of the long fuel.
Slide into the long fuel — 30%. ⏱ 2 days, to Monday 28.09. Trigger: a touch of $82,250 — entry into the band holding $151.9M of long liquidations. Cancellation level: a touch of $85,600.
Squeeze up through the short fuel — 20%. ⏱ 2 days, to Monday 28.09. Trigger: a touch of $85,600 — the 40× short's liquidation line; above it there is almost no short fuel, so little is left to burn there. Cancellation level: a touch of $82,250.
This is how we weigh the odds — not a promise; the code will compute the verdict from spot history.
🔭WHAT TO WATCH
• The large 40× long and the dense long fuel beneath it at $79,750–$82,250 (🔬).
• Monday 28.09: the funds reopen after a week in which the daily inflow shrank every day (💠).
• Friday 02.10: the first notable expiry after the quarterly (💠).
🆓TAIL
📊POLYGON
(our own signals at real prices, not a backtest)
⚙️Pendulum · Wheel(the best of four variants, a live curve at real prices)
+$120 on the day · +$9,611 in total · since 11.05.2026
⭐ Top-5 signals($1,000 per signal; "total" is the sum of all trades, not the return on a single thousand):
signal trades win last total since
Dust Strategy V2 454%−$1,125+$83,969 21.03
Dust Strategy V1 368%−$1,125+$31,406 20.03
Volatility Convergence V3 3142%+$915+$13,355 26.05
Skew 2.0 V2 4436%−$780+$9,682 26.05
Fear Flash V2 2330%−$590+$2,432 16.05
Deepest in the red: Wheel Trail V8 −$9,583 · Wheel Trail V6 −$8,481 · Wheel Trail V7 −$6,983 — three variants of one signal that hunts a reversal and pays for every attempt while the trend runs. Over the day five trades closed in the top five, and all of them in the red: two in each Dust Strategy variant, −$1,125 apiece, and one in Skew 2.0 V2 for −$780. The most profitable signal wins only 4% of its 45 trades. Win rate isn't money.
📅PUBLIC LEDGER
(computed by code at the settlement price)
The market-maker corridor. Across 404 expiries since February the market-maker break-even corridor held the settlement price 329 times — 81%. The quarterly held in both assets. Decision: we lean on the corridor knowing that roughly one expiry in five it doesn't hold.
Max pain. An exact match with the settlement in 35% of those same 404 expiries. On the quarterly max pain lay 7.1% below the settlement in BTC and 13.9% in ETH. Decision: we don't chase max pain.
The walls. The call wall held the settlement in 69 of 145 expiries where price reached it — 48%; the put wall in 41 of 98, 42%. On the quarterly price went straight through the BTC call wall. Decision: we don't publish a wall without an expiry and a contract count, nor its percentage without a base.
Whale structures. The registry: 858 closed cases (5 of them from Bybit), −$19.32M in total for the whales; the quarterly alone added −$13.62M. Decision: win rate isn't money, and we don't copy whale structures — we track where they stand.
Whale positions on perps. Twice running we called melting of the tilt what was in fact big wallets entering and leaving our sample. Decision: from today we count the day's change only on a constant set of addresses, and show the sum over the whole registry as a level, not as a move.
Polygon. Pendulum +$9,611 since 11.05 with +$120 on the day — both numbers from one series, the live curve of the best of four variants. Decision: we publish the tally daily with the same denominator and with the variant's name.
📏Scale calibrator
• What protection costs. Our fear-and-greed index — 79 out of 100 against 72 yesterday. We compute it from the price of volatility and the option skew on Deribit, not from news, which is why it diverges from the well-known index.
• Cheap or expensive for this market itself.BTC two-day vol — 1st percentile of the hourly history since February, six-day — 6th, thirty-four days out — 10th: short-term insurance is cheaper than almost at any time since February.
• Which way protection tilts.BTCskew−0.8: calls are slightly dearer than puts — yesterday puts were dearer by 0.2. Terms: indiciadesk.com/en/glossary/
Not investment advice and not a recommendation to trade. Derivatives trading carries a high risk of loss; INDICIA DESK is a research publisher, not an adviser licensed by MAS, the SEC or the FCA — decisions and their consequences are yours.