Report archive · 12.08.2026
Case No. 20260812 · 12.08.2026
Case No. 20260812 · Aug 12, 2026, 13:00 UTC · BTC $64,102 / ETH $1,910
analysis: Fable 5 by Anthropic, run on our own archive
Yesterday we called the whales' one-day lurch into BTC shorts the main change on the board. It didn't get a second day — and this morning's wick lower took out three of our scenarios at their cancellation boundaries in one stroke. We print that as loudly as we print the good days; the full review is below.
Futures (Hyperliquid). The snapshot is still bearish — top whales skewed $243M short in BTC and $241M in ETH — but the snapshot is the least interesting part. The pace is the real story:
• BTC — $158M shifted toward shorts over the week; the last day returned $6M to longs. Yesterday's "one-day decision" got no follow-through.
• ETH — $73M into longs over the week, $28M of it in the last day: the bearish thesis there is being cut faster and faster.
• 61 over-leveraged whales versus 65 a day ago — slightly less fuel on the board, and only one sits close to a margin call.
Options (Deribit). Two separate positions, as always.
• Quiet accumulation — it's there, and it's two-tiered: over 7 days the BTC $65,500 calls added 4,543 contracts (92% on the Aug 15 expiry); above that, another 3,728 calls at $70,000 with the core in September; on ETH the biggest add is 29,583 calls at the $2,000 strike. It's being built in small lots outside blocks: there's no trace on the tape — you only see it in the book.
• Mirror image below — protection is being gathered just as quietly: 12,637 ETH puts at $1,750, most of them expiring this Friday. That's not a contradiction — it's a range being built from both edges.
• Off-block structures — the detector confirmed 23 in 7 days, and not one of them is purely bullish: 15 range · 5 bearish · 3 volatility. The largest — a condor on 3,000 ETH. Legs and dates are in the paid section.
Money (the Hyperliquid border). Two channels, and the week's picture differs between them.
• Cash (the bridge, USDC only): $20.7M came in, $23.7M went out — the net is close to zero. What matters isn't the net, it's the addresses: $17.3M of the outflow landed on Coinbase deposit addresses in five tranches, a smaller slice on OKX.
• Coins (the Unit spot border): $32.5M of BTC withdrawn against $8.5M deposited — and nearly the entire outflow ties back to market maker Wintermute (how much exactly — in the paid section). SOL, by contrast, keeps arriving on a conveyor: +$8M net, every deposit from Binance; ETH slightly positive.
The classic index prints 49.5 out of 100: mid-transition zone, and it's crept higher three days running. The classic looks 90 days back; our barometer looks at where the money is going right now:
• Alts hold 41% of open interest ($3.59B of $8.76B across 240 Bybit perps) — money has entered alts, but most of it still sits in BTC and ETH.
• Hot funding has cooled: alt longs cost more than bitcoin longs in only 46% of coins — a day ago it was 79%.
• Retail in the top-10 alts runs 68.4% of accounts long (XRP 79.4% · DOGE 76.1% · SOL 70.5%), while Hyperliquid whales hold $575M in alts with only 23.9% of that money long. A gap of roughly 45 points — heads versus money.
• What the whales actually hold: HYPE $242M (23% long) · SOL $58M (6%) · ZEC $34M (47%) · XRP $34M (28%) — not a single top alt where they lean long.
Whales on Hyperliquid are structurally short, so the signal isn't the short itself — it's the CHANGE in this gap, and we log it daily.
Continuation ⟳. The week's frame is intact: perp whales are bearish by snapshot, options are quietly stacking both calls above the market and protection below it, and the volatility witnesses keep saying no move is priced. What's new isn't direction — it's the stall: yesterday's BTC short lurch didn't get a second day, and now the last 24h leans long on both coins while the week's off-block structures don't. That's a tug-of-war with no shared thesis; whichever side confirms with volume first resolves it.
The week's forecast counter. Six verdicts scored by code: 1 confirmed · 2 not confirmed · 3 invalidated at the cancellation boundary; two scenarios are still open — details in the review below.
What this is NOT. It's not an invitation to buy or sell. The fuel above the market is terrain, not timing: it only fires if price itself reaches the liquidation zones. Cheap protection doesn't make a move free and doesn't promise one. And the whales' short skew is no prophecy: the market's biggest ETH short sits on an eight-figure paper loss and is still standing.
The honest bottom line: so far only the calmest scenario keeps getting confirmed, and both fresh "corridors" were taken out not by price at the edge of the range but by cancellation boundaries we set too tight. That's a mark against our own calibration — and we print it instead of rewriting it after the fact.
signal trades win last total since Dust Speck Strat 33 6% −$1,125 +$97,469 Mar21 Dust Speck Strat 27 11% −$1,125 +$41,531 Mar20 Vol Convergence 75 31% −$678 +$2,409 May16 Flat Wings 3 67% +$850 +$943 May11 Loaded Spring 3 100% +$263 +$636 May11
In the red: Counter-Current · upper −$171 · Counter-Current · lower −$194 · Skew 2.0 −$260. All three share one cause, and it's the theme of this whole report: they're counter-trend structures, and the market hasn't given them any amplitude for weeks — the same reason vol sits in its bottom percentiles.
The most profitable signal — Dust Speck — wins only 6% of its trades: rare big winners cover dozens of small losers. Counter-Current · lower wins 62% of its trades and still sits in the red. Win rate isn't money — that's why we print both columns.
🔒 In today's full RADAR (positions — Hyperliquid, transfers — the bridge and the Unit spot border, structures — Deribit): who holds the $99M short with a liquidation line practically at the current price · the fund that added both margin and short in a single day · how much bitcoin Wintermute moved out and what it was doing on perps meanwhile · the desk's one big winner and why it's winning.
ANALYST access → indiciadesk.com/agent · 3 days free
💎DEEP
A duel at maximum leverage — BTC.
◦ the desk's most fragile short — a wallet with no public label: entry $63,999, the position didn't exist a week ago, $1.3M sits in the account beneath it, and it cut 250 BTC in the last day — not adding, carefully trimming.
◦ "DoshiAtoll" — a $101.1M short at 40×, entry $64,039, liquidation $64,581, paper loss $194k; a week ago it stood on the opposite side, and it added 124 BTC in the last day.
◦ "WelcomingTatsuya" — a $77.0M short at 20×, entry $64,418, liquidation $66,034, paper gain $308k — the only one of the three whose line isn't pressed right against the market.
◦ against them, "bowen1476" — a $64.2M long at 40×, liquidation $61,896, up $210k; it flipped sides within the week and added 300 BTC in a day.
This isn't a one-sided bearish desk — it's a head-on collision of maximum leverage: the liquidation lines of the two biggest shorts live less than one percent above the market. If price gets there, the most fragile one burns first — and that's exactly where the squeeze scenario starts.
The market's biggest pain has a name. An ETH short of $95.6M at 3× — paper loss $10.6M, the worst on the desk. The owner neither trims nor adds: liquidation is +14% away, and time works for his patience, not for a margin call. By our saved Arkham snapshot, it's pension-usdt.eth.
A fund added both margin and position. Fasanara Capital added 4,785 ETH to its short in a day (now $48.7M, paper loss $877k) — and wired $3.3M of cash across the bridge in six tranches over the week. A whale from our registry trucking collateral in under a losing thesis: conviction paid for in money — though no guarantee of being right.
The bearish ETH camp runs wider.
◦ Abraxas Capital Mgmt (Heka Funds) — a $44.2M short at 5×, paper loss $830k; liquidation far away, +54% from the market.
◦ wallet "Bennett" — a $36.4M short at 3×, down $2.35M.
◦ against them all — a fresh "AllegraSeam" long of $38.2M at 4×, up $519k: the only big long in the ETH top.
Wintermute: two hands. The market maker added 6,412 ETH to its $35.8M short in a day — and over the week moved roughly $28.5M of BTC out across the spot border against $4.2M in: that's nearly the entire coin outflow we flagged in the free section. For an MM this is inventory rotation, not a view on price; we show it because a move like this is easy to mistake for a whale fleeing the desk.
Who's making money. The desk's biggest gain — an $8.6M ETH short entered at $2,122: paper +$941k at 20× (self-styled "BirchCapricious"). It's not the biggest short that earns — it's the one that entered high: entry decided more than size did.
Structures with dates (Deribit).
◦ BTC, Sep 25 — a put spread, 1,000 contracts: bought 500 puts at $55,000, sold 500 puts at $60,000 — paid protection against a moderate drop: it earns between the levels, deeper the payout is capped.
◦ BTC, Aug 13 — a call spread, 200: sold 100 calls at $65,500, bought 100 at $66,000 — money against a squeeze by tomorrow, standing directly opposite the quiet accumulation on that very strike.
◦ BTC — rolling $62,000 puts from Aug 14 to Aug 28 (240 contracts): the protection wasn't removed, its deadline was pushed back.
◦ ETH, Aug 21 — sold 1,000 calls at $2,050: a position against upside — someone's fixing a ceiling under the call wall.
◦ ETH, Aug 13 — a straddle, 750: bought 375 puts and 375 calls at the $1,900 strike — positioning for a strong one-day move, entered near the bottom edge of our volatility history since February.
◦ ETH, an Aug 14 / Oct 30 combo at the $1,850 strike (700 contracts) — financed protection in four legs: the bought protection is paid for by the sold legs.
The tape didn't disclose net premiums on these blocks: we publish legs, sizes and dates — no invented figures.
The case ledger — with the denominator. No case resolved in the last day: the last two resolutions (both winners) we printed yesterday. The two biggest are in progress, both in ETH:
◦ No. 546 — a strangle $1,000/$4,000 on 7,500 contracts, expiring all the way out on Jun 25, 2027;
◦ No. 898 — a call diagonal $2,400/$2,800 on 4,500 contracts, expiring Sep 25, 2026.
Fragility in detail. $914.9M rides at 8× leverage or higher, and two-thirds of that is short. Cross it with volatility and you get the day's main point: the fuel is there and vol isn't pricing it — that's what a market that doesn't believe in its own sharp move looks like.
Gamma flip (Deribit book). BTC — $62,949, drifted $674 lower over three days; ETH — $1,847, essentially in place. When the flip slides down away from price, the zone where dealers dampen a decline widens: the floor under BTC is deeper than a week ago, but it's still within one bad day.
Dealer stress test (±2%). On BTC, a two-percent step down ($62,900) punches through the flip: the regime turns negative and dealers start amplifying the fall instead of damping it. The same two percent up runs into the band where the week's freshest calls are stacked — the ceiling holds. On ETH both edges are still in the damping zone: dips get bought.
Levels and walls (aggregate across all expiries).
◦ BTC: MM break-even corridor $62,500–65,000 · max pain $64,000 · put wall $61,000 (3,347 contracts, support) · call wall $64,500 (858, resistance)
◦ ETH: corridor $1,840–1,940 · max pain $1,900 · put wall $1,750 (37,239 contracts, support) · call wall $2,000 (73,226, resistance)
The BTC call wall is light next to the put side: the top right now is held not by an old wall but by freshly accumulated calls; below sits a classic heavy put wall, far under the market.
The calendar. Tomorrow's Aug 13 BTC expiry — max pain $64,000 with walls of a few hundred contracts: it decides nothing. The real money lives higher and later:
◦ BTC — September holds 32% of all open interest (core $70,000), December 29% (core $80,000);
◦ ETH — December 31% (core $3,200), September 30% (core $2,000).
That's why a one-day lull doesn't contradict the book's distant bullish positioning.
OI flow (Put/Call: month · week · now).
◦ BTC: 0.54 · 0.52 · 0.56 — calls consistently outnumber puts in the book, no drift
◦ ETH: 0.55 · 0.53 · 0.52 — fewer and fewer puts bought per call
The price of fear (VRP). The premium is +7.9 points (implied vol 35.9 against realized 28.0), and it's grown 5.3 over the week: implied pulled away from realized faster than the market itself moved. The vol seller's usual regime, not an anomaly.
1. Corridor holds. ⏱ 2 days, to Aug 14 · weighed at 45%. BTC stays inside the $63,000–65,400 band: vol sellers eat the clock, the stress test shows damping for the nearest two percent either way, and range is the most common thesis among the week's structures. After two cancellations in a row we're setting the boundary lower — under the morning wick. Cancellation boundary: a touch of the gamma flip at $62,949.
2. Squeeze up into the short fuel. ⏱ 2 days, to Aug 14 · 30%. Price touches $65,445 — that's where the main short-fuel band begins, and the liquidation lines of the two biggest 40× shorts sit even closer to the market: the first touch can start a chain. Cancellation boundary: a touch of $63,000.
3. Slide under the gamma flip. ⏱ 2 days, to Aug 14 · 25%. Price touches $62,949 and the market makers' regime changes first; below that there's little fuel down to the put wall, so this is a slow slide rather than a crash. Cancellation boundary: a touch of $65,400.
A system's decision journal. Not investment advice. Trading derivatives carries high risk of loss.
Radar is assembled automatically from live Deribit / Hyperliquid / OKX data · time-stamped snapshot. A journal of the system's decisions, not investment advice. © 2026 INDICIA DESK.