A billion of fund buying landed in one dayand the biggest short added instead of cutting.
Case #20260922 · 22.09.2026 05:36 UTC · BTC$85,106 / ETH$2,720 · analysis: Opus 5 by Anthropic
📅YESTERDAY
Monday's 21.09 expiry settled at 08:00 UTC. BTC diverged from max pain by 1.3% and the market-maker break-even corridor held it. ETH diverged by 3.5% and broke the corridor — 1.2% above its top edge. The registry measures the levels at settlement time, so they do not match the morning levels we published yesterday.
Case registry: five new whale cases over the day and one closed — the same one we printed in the paid zone yesterday, and it closed against the whale.
⚡THE GIST
• The funds came back with the month's biggest purchase: $999M went into the BTC ETFs on Monday and $270M into the ETH funds. The week of 15–21.09 in bitcoin closed at +$845M against −$303M the week before.
• Whales in ether turned: the long side grew $94M over the day and the net short tilt slimmed from $444M to $367M. In bitcoin the tilt barely moved — $386M.
• The biggest short house did not capitulate: over the day it added to its short in both assets and wired in about $88M to its accounts.
• The options side is building above the market for October: over the week 22,073 calls were quietly accumulated at a level roughly 12% above spot, and 95% of them sit in the October expiry; one level below carries another 13,394.
• The block market woke up sixfold: 145 bitcoin structures over two days against 23 before.
• Verdict — continuation ⟳: the whales' short tilt holds, and the addresses holding it are the same ones adding margin.
• 📊 Our Polygon — Pendulum: standing position, $0 on the day, +$10,014 in total since 11.05.
• 📊 Our Polygon — signals: best — Dust Strategy V2 +$86,219 at a 5% win rate · worst — Wheel Trail V8 −$8,787. Table at the bottom.
🐋WHALES: WHAT CHANGED — three arenas
Arena one — Hyperliquid futures. A 4.7% rally split the two assets. In BTC the short side grew from $701M to $754M and the long side from $322M to $368M, so the net tilt is $386M against $379M yesterday (how we track whale positions). Inside the day the short reached $859M and slimmed by about a hundred million overnight: part of those positions was closed right at the price peak. In ETH it went the other way: the long side jumped from $355M to $449M and the short tilt slimmed from $444M to $367M — the smallest in a week. Among the top whales there was one forced closure of $3.3M over the day, and it was a long. Overleveraged whales number 61 against 63 yesterday, but two thirds of them are now longs against half yesterday.
Arena two — Deribit options. The blocks woke up, and all of the interest sits above the market. Over two days the snapshot recognised 145 structures in BTC against 23 before, and 38 in ETH (what a block is). The largest is a butterfly on calls, 20,000 contracts in all with its centre roughly 12% above the market: that is not a position on direction but a pinpoint position on a level. Alongside it went an aggressively bullish structure of 3,600 contracts and a diagonal of 4,528 that caps the rise. Quiet accumulation in small lots outside the blocks is almost all call-side this week and almost all in the October expiry: 22,073 calls at one level, 13,394 and 11,300 at two neighbouring ones. The registry took in five new cases, all in bitcoin.
Arena three — money at the exchange border. Over the week ether was brought in and bitcoin was taken out. Over seven days $103.9M of ether and $60.5M of USDC cash came into the exchange, while $97.3M of BTC left against $4.7M brought in. With all coins together, $182.9M came in against $169.8M out — the flow has almost levelled. The border snapshot refreshes once a day and stops on the morning of 21.09, so yesterday's rally is not in it yet.
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 Monday $999M went into the BTC funds — the biggest daily inflow in a month; $270M went into the ETH funds. The week of 15–21.09 closed at +$845M in bitcoin against −$303M the week before — but almost all of that is a single Monday: the four days before it added up to a minus. In ether +$9M against +$318M. This contradicts the whales: in a single day the funds bought almost a billion of the coin itself, while the whales hold a net short of $386M in bitcoin.
What the options market says. This camp measures one thing: what insurance costs on Deribit. BTCskew+1.4, almost as yesterday: the extra charge for downside protection is holding (what skew is). In ETH the tilt is −1.0, calls slightly dearer. Insurance got dearer along with the move: BTCDVOL37.9 against 35.4 yesterday, a premium over the actual move of +1.9. Two-day vol sits in the 47th percentile of its own history since February, ten-day in the 39th (what implied vol is). Our fear-and-greed index stands at 78 out of 100 against 79 in yesterday's radar. This camp is silent: the price of downside protection is holding, while the week's largest footprint in open interest is calls for October above the market. Two layers pull in different directions.
What the futures say. This camp measures who pays whom to hold a position. BTC funding on Hyperliquid is +0.010% in eight-hour terms and +0.009% on Binance; in ETH the same values (what funding is). Longs have been paying shorts for a fourth day. This contradicts the whales: a leveraged long still costs money to hold, even though the whales hold short.
What the crowd is doing. This camp measures how retail accounts stand against the whales. On Bybit perps 51% of BTC accounts stand long against 52% yesterday, and 66% in ETH, as yesterday. The gap between the crowd and the whales in BTC is 18.6 points against 19.8 yesterday — a fourth day of compression, the 3rd percentile of the month. In ETH the gap is 30.1 against 35.5: over the day it compressed by five points, because the whales took on the long side. This confirms the whales: both gaps are compressing from the whales' side, not the crowd's.
Any signs of stress. This camp measures whether anyone has already been forced out of the market. Among the top Hyperliquid whales there was one forced closure of $3.3M over the day — the first in a week, and it was a long. On OKX swaps $2.6M was liquidated, 96% of it shorts; that is 26% of the whole week's volume. This contradicts the whales: the positions forced shut were mostly shorts, and the pace has grown for a second day.
Again: this is positioning, not a direction forecast.
🌡ALTSEASON — the classic and our barometer
The classic altseason index stands at 36.4 out of 100 against 39.5 in yesterday's radar — bitcoin territory: the rally lifted BTC itself, not the alts.
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 43.3% — $5.77B of $13.33B, almost unchanged, even though open interest itself grew by a billion over the day.
Alt contracts with funding hotter than bitcoin's make up only 4.5%. The bar here is BTC's own funding, and it jumps every day, so this scale makes sense only alongside funding itself.
Retail in the top-10 coins stands 68.1% long — the most in a week; the hottest are 1000PEPE and XRP.
Whales in alts are structurally short: 29.6% long across $1.59B of positions, the biggest of them HYPE. That gap itself is permanent, the signal is in its CHANGE — over the day the long share fell by 1.6 points.
Full barometer: indiciadesk.com/en/altseason
⚖️VERDICT OF THE DAY
continuation ⟳ — despite a 4.7% rally and a billion from the funds, the whales' short tilt in bitcoin did not shrink: $386M against $379M yesterday. The same two addresses that carry the market's largest short added to it in both assets over the day and wired in about $88M of margin. The break came only in ether, and it was made not by them but by new long addresses. Witnesses: the crowd sides with the whales; the funds, funding and liquidations go against them; the options side is silent.
What this does NOT mean. This is not a claim that the market turns down. Adding margin shows an intention to hold, not knowledge. And while one house tops up collateral, the options side quietly accumulates calls for October above the market — two groups of smart money stand against each other, and only price settles it.
🔒In today's full RADAR:
— what happened to the two addresses we showed yesterday: the 40× short with its liquidation 4% above the market and the biggest bitcoin long — neither position exists any more, and you can see the accounts they left with;
— how many coins the biggest short house bought into its short on the rally, how far its average entry rose and how much margin it wired in.
ANALYST access → indiciadesk.com/en/agent
💎DEEP
🔒WHALES IN DETAIL 💠
Portraits from the Hyperliquid registry, snapshot 03:46 UTC at BTC$85,563 and ETH$2,737. Nicknames aren't treated as identification.
The market's biggest loss.ETH SHORT $285.8M, leverage 5×, uPnL −$45.3M against −$38.1M yesterday, liquidation at $4,007 — 46% above spot. The coin count grew for the first time in four days: about 104,400 against 102,900, and the average entry rose from $2,297 to $2,304. The same address in BTC. SHORT $222.9M against $171.9M yesterday, leverage 5×: about 2,600 coins against 2,110, average entry from $72,731 to $75,083. uPnL −$27.3M, liquidation at $136,877 (+60%). The account grew from $88.1M to $141.1M — it was topped up. A second address with the same pattern.ETH SHORT $212.3M and BTC SHORT $136.9M: it added on both sides too, with average entries up to $2,338 and $73,362. uPnL −$30.9M and −$19.5M, account from $73.6M to $109.2M. Together the two addresses carry $123.0M of unrealised loss against $96.2M yesterday, while margin across the two accounts grew from $161.7M to $250.3M — about $88M wired in over the day. The biggest bitcoin long. LONG $124.9M against $87.2M yesterday at the same address — about 1,460 coins at 20× from $81,765, uPnL +$5.5M, liquidation at $69,820 (−18%), account $12.7M. Yesterday's biggest long closed at the peak. The address that held 1,333 BTC at 3× from $78,057 still had +$8.8M on paper at 10:16, and by midday the position was gone: the account fell from $108M to $14M — the money was withdrawn. In the evening the same address opened a small 500BTC short from $85,994. Yesterday's 40× short did not survive the rally. The position we showed yesterday with its liquidation 4% above the market had grown to 863 coins from $80,181. As price went through that level it was cut in stages: 390 coins in the morning, 145 by midday, and then the address disappeared from the registry. Its account fell over that time from $3.8M to $0.8M. Whether it was closed by force the snapshot does not show. A new long at 35× leverage.BTC LONG $90.5M from $81,875, uPnL +$3.9M, liquidation at $81,515 — only 5% below spot on an account of $5.5M. A week ago the position did not exist. The short at 25×.BTC SHORT $103.3M, average entry up to $80,145, uPnL −$6.5M, liquidation at $101,256 — 18% above spot, account $23.7M. The same address holds ETH SHORT $79.5M from $2,597, with its liquidation 24% above spot. The ether longs. A new ETH LONG $112.9M at 10× from $2,694, liquidation at $2,513 (−8%), account $11.3M. The risky 25× long now holds about 31,400 coins ($85.9M) against about 34,800 yesterday, its average entry up from $2,563 to $2,618: they added higher and shed part of it. Liquidation at $2,477 — 10% below spot. The market's biggest gain is ETH LONG $82.9M at 20× from $2,134, uPnL +$18.3M. The third big bitcoin short.BTC SHORT $60.9M at 30× leverage from $80,851, uPnL −$3.4M, liquidation at $157,327 (+84%); over the day it cut 41 coins. Fragility.61 overleveraged whales against 63 yesterday, $1.16B of positions together, but 66% of them are now longs against 52% yesterday. Not one sits near a margin call.
Cases in the registry — five new over the day, one closed with a score. Case #1096 · BTC call ratio spread: 20 calls at $82,000 bought and 40 calls at $83,000 sold with expiry today — 60 contracts, entered at $81,301. Yesterday we printed it with the explanation that the whale starts losing if the market clears $83,000 with room to spare. That is exactly what happened: overnight price reached $87,023. The registry has already closed the case at about −$30k, computed at the current price; the final figure comes from the 08:00 UTC settlement, and at the high it would have been about twice as large. What broke it was the sold half — for every call bought there were two sold.
Профіль виплат на експірацію 2026-09-22. Кит заплатив $650 премії. Беззбитковість: $82 029 і $83 967.
Case #1099 · BTC call diagonal: 100 calls at $85,000 bought for Friday 25.09.2026, 100 calls at $87,000 sold for Friday 30.10.2026 — 200 contracts, entered at $85,485, new. Net premium about 2.29BTC received (≈$196k): the near leg is in the money, the far one out of it and a month further away.
Профіль виплат на експірацію 2026-09-25. Кит отримав $196k премії. Беззбитковості в показаному діапазоні немає.
Case #1100 · BTC put ratio spread: 100 puts at $80,000 sold for Friday 30.10.2026 and 50 puts at $86,000 bought for Friday 25.09.2026 — 150 contracts, entered at $85,485, new. Net premium about 1.17BTC received (≈$100k): the whale sells far-dated protection in twice the size it buys the near-dated one.
Профіль виплат на експірацію 2026-09-25. Кит отримав $100k премії. Беззбитковості в показаному діапазоні немає.
Case #948 · ETH condor for the quarterly: 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 Friday 25.09.2026 — 11,500 contracts, entered at $2,462, day 28. Net premium about 36.8ETH received (≈$91k). Spot $2,720 stands 3% below the sold call strike — the winning zone still holds, three days from expiry.
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 about 8.22BTC received (≈$624k), seventh day. From the $75,860 entry spot has moved 12.2% against this position.
Профіль виплат на експірацію 2026-10-30. Кит отримав $624k премії. Беззбитковість: $76 236.
Also new: #1098 — a diagonal on calls at $88,000/$90,000 across 360 contracts; #1101 and #1102 — two small structures on puts with expiries on 23 and 25.09. ΔOI over the week, both ways:BTC is calls only and almost all in the 30.10 expiry: +22,073 at $95,000, +13,394 at $90,000, +11,300 at $100,000. ETH: +19,730 calls at $2,800 and +12,763 at $3,000 against +13,956 puts at $2,500.
🔬Who exactly
Both of the biggest shorts are Abraxas Capital Mgmt (Heka Funds): after four days without motion the firm added about 500 coins to its bitcoin short and fifteen hundred to its ether one over the day, raised both average entries and wired about $88M into the two accounts. The firm's unrealised loss is $123.0M against $96.2M yesterday.
The 25× BTC short and the ether short of the same address are self-labelled VBVIT; nicknames aren't treated as identification.
The third big bitcoin short, which cut 41 coins over the day, is Fasanara Capital.
The risky 25× ether long, its average entry up to $2,618, belongs to Machi Big Brother.
The new 35× bitcoin long with its line 5% below spot, the biggest bitcoin long, the new $112.9M ether long and the market's biggest gain are addresses without a public label.
Money by address. In the weekly snapshot the coin channel sent out $97.3M of BTC, the largest tranches to Wintermute; ether, by contrast, was brought in — $103.9M.
🔒THE WEEK'S DYNAMICS 🔬
Perps.BTC: net $386M short (long $368M against short $754M), $5M toward short on the 24-hour window (from yesterday's snapshot, $7M), $104M over seven days — the pace is fading. ETH: net $367M short (long $449M against $816M), $83M toward long on the 24-hour window (from yesterday's snapshot, $77M) — a turn inside a weekly window that still shows $24M toward short.
Gamma flip.BTC across all expiries $71,367, spot 19% above it (what the flip is). ETH$2,237, spot 22% above. Over the week the call wall rose by $10,000 and the put wall by $9,500: market makers carried their protection up behind price in full.
The price of fear. On the weekly window the premium is +1.7 (DVOL38.0 against the actual weekly HV36.3), and over seven days it rose by 4.5. A second day running the market charges a mark-up for fear instead of a discount.
Changing of the guard. Over the week the whales assembled 272 confirmed structures: 86 bullish, 78 bearish, 31 on volatility, 77 on range — more bullish than bearish for the first time in five days.
🔒CROSS-CHECK IN DETAIL 💠
BTC levels. Everything below comes from Deribit options open interest. The market-maker break-even corridor is $80,000–$85,500 against $79,000–$82,000 yesterday — it rose behind spot, and spot stands near its top edge; by our study across 396 expiries price stays inside such a corridor in roughly 82% of cases (what the corridor is). Max pain is $82,500 on Deribit, and $83,000 on the combined two-exchange open interest for today's expiry (what max pain is).
Today, Tuesday 22.09.BTC 4,334 contracts — a thin expiry: max pain$83,000 on the combined open interest, put wall $82,000 on 404, call wall $83,000 on 290, both walls below spot (what the walls are). ETH 44,986 contracts, max pain$2,700 on the combined open interest and $2,680 on Deribit; put wall $2,640 on 5,452, call wall $2,900 on 3,566.
Tomorrow, Wednesday 23.09.BTC 2,558 contracts, ETH 26,922 — thin again.
ETH levels. The corridor is $2,620–$2,740, spot in its upper part; max pain$2,680.
The quarterly, Friday 25.09.BTC: 185,516 contracts, max pain$75,000 both on Deribit and on the combined open interest — 12% below spot; densest are the $70,000 put (8,154) and the $70,000 call (9,006), with the $90,000 call (8,013) alongside. ETH: 772,446 contracts, max pain$2,300 on the combined open interest — 15% below spot; put $2,100 on 35,903, call $3,000 on 34,067.
The October expiry, Friday 30.10. That is exactly where the week's accumulation went: 101,429 contracts in BTC, and the densest level of the whole chain is the $95,000 call on 23,322 contracts, with $90,000 on 15,665 beside it.
Fuel. Under BTC$110.7M of long positions in the $81,285–83,852 band against $28.6M of shorts above the market in the $87,274–89,841 band — over the day the tilt flipped: four times more below. In ETH both sides are thin: $10.3M under the market against $3.8M 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 above at $92,000–$94,000 and below at $78,000–$79,000. In ETH it is empty above at $2,900–$3,000 and below at $2,500–$2,550.
Stress test ±2%. On a 2% fall BTC runs into its densest put wall — the move is damped. On a 2% rise it stays under the $90,000 call wall, so the ceiling still holds.
OI flow.BTC put/call across all expiries: month 0.59 · week 0.55 · now 0.57; ETH0.56 · 0.51 · 0.53 — both stable (what put/call is).
Two options venues. On the same expiries up to 30 days BTC put/call is 0.68 on Deribit against 1.46 on Bybit; ETH0.65 against 1.02 — over the day both turned more put-heavy. We don't know who owns the Bybit positions, so this is neither "whales" nor "retail" — just a second venue with a different lean.
IV by tenor.BTC 2d P38/C37 · 10d P38/C38 · 38d P38/C37, percentiles since February 47th, 39th and 34th. ETH 2d P50/C55 · 10d P51/C52 · 38d P52/C53, percentiles 52nd, 39th and 40th. BTCDVOL37.9 against 30-day realised 36.0 — premium +1.9; ETHDVOL52.8 with HV 50.8, premium +2.0 (what the premium is).
💎SCENARIOS IN DETAIL 💠
The code's verdicts for yesterday. From 19.09: the corridor came in — price stayed inside the band the whole time; the squeeze up did not come in (a maximum of $81,840 against the $82,650 trigger); the slide did not either. The "slide into the long fuel" scenarios from 20.09 and 21.09 were cancelled at their cancellation levels — a maximum of $87,023. Two corridor scenarios and two squeeze scenarios are still open: price has already passed both squeeze triggers, but the code will score them at their horizon. The week since 15.09: 21 scenarios — 6 confirmed, 5 not, 6 cancelled at their cancellation levels, 4 still running. All-time: squeeze up 10 of 31, corridor 10 of 27 plus 6 partial, slide into the long fuel 5 of 22.
The corridor holds — 45%. ⏱ 2 days, to Thursday 24.09. BTC stays between the fuel bands, $83,850–$87,250 (levels rounded to the band edges). Cancellation level: a touch of $81,250 — the bottom of the long-fuel band.
Squeeze up through the short fuel — 30%. ⏱ 2 days, to Thursday 24.09. Trigger: a touch of $87,250 — the edge of the band of short positions. Cancellation level: a touch of $83,850.
Slide into the long fuel — 25%. ⏱ 2 days, to Thursday 24.09. Trigger: a touch of $83,850 — the edge of the band of long positions. Cancellation level: a touch of $87,250.
This is how we weigh the odds — not a promise; the code will compute the verdict from spot history.
🔭WHAT TO WATCH
• Friday 25.09: the quarterly — 185,516 contracts in BTC and 772,446 in ETH, with max pain on both 12–15% below the market (💠).
• The margin of the biggest short house: $250.3M across two accounts against $161.7M yesterday — whether it keeps topping up or starts cutting (🔬).
• The October expiry on 30.10: the $95,000 call on 23,322 contracts — the densest level of the whole bitcoin chain (💠).
🆓TAIL
📊POLYGON
(our own signals at real prices, not a backtest)
⚙️Pendulum · Wheel(standing position, real prices)
$0 on the day · +$10,014 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 435%−$1,125+$86,219 21.03
Dust Strategy V1 349%−$1,125+$33,656 20.03
Volatility Convergence V3 3142%+$915+$13,355 26.05
Skew 2.0 V2 4137%−$775+$10,606 26.05
Fear Flash V2 2232%−$418+$3,023 16.05
Deepest in the red: Wheel Trail V8 −$8,787 · Wheel Trail V6 −$7,718 · Wheel Trail V7 −$6,179 — three variants of one signal that hunts a reversal and pays for every attempt while the trend runs. One trade closed in the top five over the day: Volatility Convergence V3 in the green by $915, and its win rate rose to 42%. The most profitable signal wins only 5% of its 43 trades. Win rate isn't money.
📅PUBLIC LEDGER
(computed by code at the settlement price)
The market-maker corridor. Across 396 expiries since February the market-maker break-even corridor held the settlement price 323 times — 82%. Monday's expiry held in bitcoin and did not hold in ether. 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 36% of those same 396 expiries. Yesterday BTC diverged from it by 1.3% and ETH by 3.5% — both outside tolerance. Decision: we don't chase max pain; on the quarterly it stands 12–15% below the market, and that is not a promise of a move there.
The walls. The call wall held the settlement in 49% of expiries, the put wall in 43%: roughly a coin toss. Today's bitcoin expiry is thin, and both of its walls stayed below spot after yesterday's rally. Decision: a wall without an expiry and a contract count isn't published.
Whale structures by type (our Deribit registry, 727 closed cases). Call ratio spread — the type of the case that closed against the whale: 30 closed, not one came in on its strikes, −$1.07M in total. Call diagonal — the type of two of the new cases: 51 closed, −$1.55M in total. Condor — the type of the big quarterly case: 36 closed, 69% came in, −$126k in total. All closed cases together: −$5.84M. Decision: win rate isn't money, and we don't copy whale structures — we track where they stand.
Polygon. Pendulum +$10,014 since 11.05: the running tally rose by $372 over the day although the daily line shows zero — the table writes the result in after the position closes. One trade closed in the top five, in the green. Decision: we publish both tallies daily with the same denominator, including the days when nothing happened.
📏Scale calibrator
• What protection costs. Our fear-and-greed index — 78 out of 100 against 79 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 — 47th percentile of the hourly history since February, ten-day — 39th, thirty-eight days out — 34th: the middle of its own range.
• Which way protection tilts.BTCskew+1.4: puts cost more than calls, almost as yesterday. 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.