PakupaiInstitutional-grade research, for everyone
← Thu Sep 10·Friday, September 11, 2026·filed ·Sat Sep 12 →revision 4 of 4
Bitcoin$77,200▲ +0.8% 1d
Funding 8h0.0033%p36 of 90d
OKX OI$2.09B▼ -1.9% 7d
US 10Y4.97%▲ +21 bp 1w
Fed funds3.63%0 bp 1w
Dollar (DXY)99.100% 1d
Brent$104.42▼ -3% 1d
Gold$4,390▲ +0.6% 1d
S&P 5007,657▲ +0.9% 1d

→ Superseded by the brief filed Sep 12, 12:59 AM ET. Read it.

01 · The call · revision 4, changed since 6:34 PM ET

Bitcoin stays inside the $76,500 to $78,300 range it failed to break on Sep 11, with leverage draining (OKX funding 0.0033% per 8h, the 36th percentile of 90 days; OKX open interest $2.09B, -1.9% on the week) and nothing resolving before the Sep 16 Fed decision. Moderate confidence.

Revision 4 overturns revision 3's funding claim: the close print is 0.0033% per 8h at the 36th percentile, cheaper than the 0.0046% two hours earlier, so longs were not paying up into weakness. Everything else stands: $77,199.73 inside the $76,500 to $78,300 range, with the Sep 16 Fed the next catalyst.

Statusunchanged →
Confidencemoderate
Revision4 of 4 · end of day
Fileddeepseek/deepseek-v4.1-flash · $0.10
What to do about it

Hold core spot, no perp leverage into the Fed, ladder bids at $74,000 and $72,000, alerts at $78,300 and $76,500.

The call

what changed since r3 (6:34 PM ET)compare with Thu Sep 10 EODr1r2

The thesis is now called unchanged; the posture changed.

r3 · 6:34 PM ET

The pre-CPI stop-run reclaim failed at $78,300 and Bitcoin closed $77,175, so the day is a lower-high rejection inside a $76,500 to $78,300 range, not the start of a trend. Leverage is leaving rather than building (funding 0.0046% per 8h, 48th percentile; OKX open interest $2.10B, -1.9% on the week), and nothing resolves before the Sep 16 Fed decision. Moderate confidence.

Revision 3 grades the morning's $78,300 reclaim test: it failed. Bitcoin closed $77,175, $848 below the 8:53 AM read and $1,124 under the pivot, with OKX open interest down to $2.10B, which undoes the morning's 'new longs entering' read. The day resolves as a failed reclaim inside a $76,500 to $78,300 range, with the Sep 16 Fed decision as the next real catalyst.

Statusweakened
Confidencemoderate
What to do about it

Hold core spot, no perp leverage into the Fed, keep ladder bids at $74,000 and $72,000, and treat $78,300 as the line that flips the range back to upside.

r4 · 7:50 PM ET · this version

Bitcoin stays inside the $76,500 to $78,300 range it failed to break on Sep 11, with leverage draining (OKX funding 0.0033% per 8h, the 36th percentile of 90 days; OKX open interest $2.09B, -1.9% on the week) and nothing resolving before the Sep 16 Fed decision. Moderate confidence.

Revision 4 overturns revision 3's funding claim: the close print is 0.0033% per 8h at the 36th percentile, cheaper than the 0.0046% two hours earlier, so longs were not paying up into weakness. Everything else stands: $77,199.73 inside the $76,500 to $78,300 range, with the Sep 16 Fed the next catalyst.

Statusunchanged →
Confidencemoderate
What to do about it

Hold core spot, no perp leverage into the Fed, ladder bids at $74,000 and $72,000, alerts at $78,300 and $76,500.

The steps

  1. Hold core spot; carry no perp leverage into the Fed.
  2. Alerts at $78,300 and $76,500: above flips the range up, below resumes defensive cuts.
  3. Ladder bids at $74,000 first, $72,000 second; fills only on a floor break.
  4. Long expression is spot or a defined-risk put spread (pays if price falls, loss known up front); shorting a cheap-funding range into a binary Fed is not the trade.

Execution riskElevated: a binary Fed five days out on a range-bound tape, the setup where leverage gets chopped out or liquidated before the range resolves, even if the thesis is right.

The steps

what changed since r3 (6:34 PM ET)compare with Thu Sep 10 EODr1r2

The position read was rewritten.

r3 · 6:34 PM ET

  1. Hold core spot and carry no perp leverage into the Fed.
  2. Alerts at $78,300 and $76,500; a daily close below $76,500 resumes defensive cuts.
  3. Keep ladder bids at $74,000 (first tranche) and $72,000 (second).no longer here
  4. Any long expression should be spot or a defined-risk put spread, which pays if price falls with the loss known up front; shorting strength is not the trade into a binary Fed.

Execution riskElevated into a binary Fed within five days. The deck is clean and funding cheap, but a range-bound tape chops leverage with stop-outs before the decision resolves, so the thesis can be right while the position gets liquidated first.

r4 · 7:50 PM ET · this version

  1. Hold core spot; carry no perp leverage into the Fed.
  2. Alerts at $78,300 and $76,500: above flips the range up, below resumes defensive cuts.
  3. Ladder bids at $74,000 first, $72,000 second; fills only on a floor break.
  4. Long expression is spot or a defined-risk put spread (pays if price falls, loss known up front); shorting a cheap-funding range into a binary Fed is not the trade.

Execution riskElevated: a binary Fed five days out on a range-bound tape, the setup where leverage gets chopped out or liquidated before the range resolves, even if the thesis is right.

In thirty seconds
Read the call and stop.
The sentence above and the box under it are the whole position. Everything that follows is the working.
In three minutes
Add 02 and 04.
What changed since the last revision, and where the market is actually positioned. That is the part that dates fastest.
The whole thing
Ten sections, about eight minutes.
Written in the same order every day, so the sections stay in the same place and the rest can be skipped.
[OBSERVED]measured directly[DERIVED]computed from measurements[INFERRED]an interpretation[HYPOTHESIS]not yet confirmed[UNKNOWN]looked for and not foundThe mark after a claim says how it is known. The glossary has the words.
Bitcoin · 4h · Sep 4 – Sep 11 · the levels that matterlast $77,200 +0.8% 1d
Bitcoin, 48 four-hour candles from Sep 4 UTC to Sep 11 UTC, between $76,047 and $81,428; 5 levels drawn as dashed lines and the price now, $77,200, as the solid line; dates in UTC75,00076,50078,00079,50081,00082,500WEEKLY LOWER-HIGH CEILING +3.6%FAILED RECLAIM PIVOT +1.4%NOW $77,200RANGE FLOOR -0.9%▼ FIRST LADDER RUNG -4.1%▼ SECOND RUNG -6.7%Sep 5 UTCSep 6 UTCSep 7 UTCSep 8 UTCSep 9 UTCSep 10 UTCSep 11 UTC
What to notice: Failed reclaim: close about $1,100 under the $78,300 pivot; range $76,500 to $78,300 [DERIVED]

What changed since 6:34 PM ET

The overnight in order: what printed, then what it did to the call.

  1. REVERSEDThe close print kills the re-crowding read: OKX funding 0.0033% per 8h at the 36th percentile of 90 days, cheaper than the 0.0046% read two hours earlier [OBSERVED].Re-crowding only starts above 0.01% per 8h on a bounce; until then a cheap deck is not a warning.
  2. STRENGTHENEDOpen interest keeps draining: OKX $2.09B, -1.9% on the week and flat versus $2.10B two hours earlier, with price up on the day [OBSERVED].Keep ladder bids queued rather than buying the range; no re-leverage until price and open interest rise together.
  3. NEWThe intraday fade is now sourced: Binance spot printed $78,308 at 12:27 UTC, about $1,100 above the close [OBSERVED].Alert $78,300; do not chase a weekend push through it.
How the previous calls turned out
confirmed
said Sep 11, 6:34 PM ET the failed reclaim left a $76,500 to $78,300 range, not a trendClose $77,199.73, still inside the range, no break either way [OBSERVED].
rejected
said Sep 11, 6:34 PM ET funding ticked up into a falling close, so some longs were still paying to stay longThe close print is 0.0033% per 8h, 36th percentile, cheaper not richer [OBSERVED].
How the call has moved today
The call by revision: r1 replaced, low confidence, 8:34 AM ET; r2 strengthened, low confidence, 8:53 AM ET; r3 weakened, moderate confidence, 6:34 PM ET; r4 unchanged, moderate confidence, 7:50 PM ET, this revisionREPLACEDr18:34 AMSTRENGTHENEDr28:53 AMWEAKENEDr36:34 PMUNCHANGEDr47:50 PM
Taller is more confident. The colour is the status each revision filed itself under.

What changed since 6:34 PM ET

what changed since r3 (6:34 PM ET)compare with Thu Sep 10 EODr1r2

3 new items in what changed.

r3 · 6:34 PM ET

  1. REVERSEDThe morning's $78,300 reclaim test failed: Bitcoin closed $77,175, $1,124 below the pivot and $848 under the 8:53 AM ET read [OBSERVED].Alert at $78,300; a daily close above reopens $80,000, otherwise the tape is a $76,500 to $78,300 range.no longer here
  2. REVERSEDOKX open interest (the total leveraged bets currently open) fell to $2.10B from $2.23B in the morning, -1.9% on the week, with price down [OBSERVED].No re-leverage until OI stabilizes with price; funding above 0.01% per 8h on any bounce is the re-crowding warning.no longer here
  3. NEWThe post-CPI equity close is now observed risk-on: S&P +0.9% to 7,657, Nasdaq +1.0%, while Bitcoin closed +0.8% and under its pivot [OBSERVED].Equity strength is not confirmation; require a Bitcoin close above $78,300 before treating the backdrop as supportive.no longer here
  4. WEAKENEDThe 10-year yield is 4.97%, +21bp on the week, while the effective funds rate is flat at 3.63%: market rates repriced, policy did not [OBSERVED].5.00% on the 10-year is the de-risk trigger for any long.no longer here
  5. WEAKENEDThe actual CPI prints remain unverified at the close, the third revision today without them [UNKNOWN].Treat today's read as price and flow evidence only, not confirmation of a soft inflation path.no longer here
How the previous calls turned out
partial
said Fri Sep 11, 8:53 AM ET the overnight break of $76,500 was a pre-CPI stop-run and the reclaim extends to $78,300 only with a 4h or daily close above itthe floor held, close $675 above $76,500, but the bounce never closed above the pivot ($77,175, $1,124 short).
rejected
said Fri Sep 11, 8:53 AM ET OKX open interest +$50M to $2.23B on the bounce is new longs entering, not just short coveringOI closed $2.10B, about $130M lower, with price down; the additions came back out.
unresolved
said Fri Sep 11, 8:53 AM ET funding above 0.01% per 8h on a bounce signals re-crowding, stand downfunding closed 0.0046%, 48th percentile of 90 days; the warning never triggered.

r4 · 7:50 PM ET · this version

  1. REVERSEDThe close print kills the re-crowding read: OKX funding 0.0033% per 8h at the 36th percentile of 90 days, cheaper than the 0.0046% read two hours earlier [OBSERVED].Re-crowding only starts above 0.01% per 8h on a bounce; until then a cheap deck is not a warning.
  2. STRENGTHENEDOpen interest keeps draining: OKX $2.09B, -1.9% on the week and flat versus $2.10B two hours earlier, with price up on the day [OBSERVED].Keep ladder bids queued rather than buying the range; no re-leverage until price and open interest rise together.
  3. NEWThe intraday fade is now sourced: Binance spot printed $78,308 at 12:27 UTC, about $1,100 above the close [OBSERVED].Alert $78,300; do not chase a weekend push through it.
How the previous calls turned out
confirmed
said Sep 11, 6:34 PM ET the failed reclaim left a $76,500 to $78,300 range, not a trendClose $77,199.73, still inside the range, no break either way [OBSERVED].
rejected
said Sep 11, 6:34 PM ET funding ticked up into a falling close, so some longs were still paying to stay longThe close print is 0.0033% per 8h, 36th percentile, cheaper not richer [OBSERVED].

The case for and against

Both columns are kept honestly. If the right column ever gets longer than the left, the call is in trouble.

For the call

  • The $76,500 floor held; the close sits about $700 above it, so the overnight break was a shakeout, not a breakdown.$76,500 · Sep 11 close
  • Funding 0.0033% per 8h, 36th percentile of 90 days: cheap, so neither side is crowded.0.0033% · Sep 11 close
  • Hyperliquid BTC perp funding is -0.00005% per hour, meaning shorts still pay longs, so longs are not paying up there either.-0.00005%/hr · Sep 11 close

Against the call

  • Equities closed risk-on, S&P +0.9% and Nasdaq +1.0%, while Bitcoin could not clear $78,300.
  • The 10-year at 4.97%, +21bp on the week, is a live headwind under an asset that is +21.6% on the month.
  • No ETF flow, spot volume or options data exists in the record, so the deleveraging read rests on leverage prints alone.

The contradiction the call cannot resolveEquities say risk-on; Bitcoin under $78,300 says not in crypto. With no spot or options data, that drag read stays a hypothesis the Sep 16 Fed has to settle.

The case for and against

what changed since r3 (6:34 PM ET)compare with Thu Sep 10 EODr1r2

Nothing measured moved here; the wording changed.

r3 · 6:34 PM ET

For the call
  • The reclaimed $76,500 floor held: the close sits $675 above it, so the overnight break reads as a shakeout, not a breakdown.$76,500 · Sep 11 close
  • Funding closed 0.0046% per 8h, 48th percentile of 90 days: cheap and mid-range, neither side crowded.0.0046% · Sep 11, 22:26 UTC
  • Hyperliquid BTC perp funding is negative at -0.0002% per hour, meaning shorts pay longs, so longs are not paying up on that venue.-0.0002%/hr · Sep 11, 22:26 UTC
  • The monthly divergence persists: Bitcoin +21.6% against S&P -1.2% and Nasdaq -1.0%.no longer here+21.6% · Sep 11 close
Against the call
  • Equities closed risk-on while Bitcoin could not reclaim $78,300: either Bitcoin lags a genuine easing impulse or the equity rally trades something else.no longer here
  • Funding ticked up to 0.0046% from 0.0024% into a falling close, so some leverage was still paying to stay long into weakness.no longer here
  • The 10-year at 4.97%, +21bp on the week, is a live headwind under an asset that is +21.6% on the month.

The contradiction the call cannot resolveThe equity close says risk-on; Bitcoin's failed reclaim under $78,300 says not for crypto. One of those reads is wrong about the week, and the Sep 16 Fed decision is what settles it.

r4 · 7:50 PM ET · this version

For the call
  • The $76,500 floor held; the close sits about $700 above it, so the overnight break was a shakeout, not a breakdown.$76,500 · Sep 11 close
  • Funding 0.0033% per 8h, 36th percentile of 90 days: cheap, so neither side is crowded.0.0033% · Sep 11 close
  • Hyperliquid BTC perp funding is -0.00005% per hour, meaning shorts still pay longs, so longs are not paying up there either.-0.00005%/hr · Sep 11 close
Against the call
  • Equities closed risk-on, S&P +0.9% and Nasdaq +1.0%, while Bitcoin could not clear $78,300.
  • The 10-year at 4.97%, +21bp on the week, is a live headwind under an asset that is +21.6% on the month.
  • No ETF flow, spot volume or options data exists in the record, so the deleveraging read rests on leverage prints alone.

The contradiction the call cannot resolveEquities say risk-on; Bitcoin under $78,300 says not in crypto. With no spot or options data, that drag read stays a hypothesis the Sep 16 Fed has to settle.

Where the market is positioned

The state of play across the assets the tape reads, then Bitcoin in detail, where positioning is measured.

AssetLevelChange
S&P 5007,657+0.9% 1d
Nasdaq26,333+1% 1d
Dollar (DXY)99.100% 1d
Gold$4,390+0.6% 1d
Brent$104.42−3% 1d
US 10Y4.97%+21 bp 1w
US 2Y4.40%+21 bp 1w
Fed funds3.63%0 bp 1w

Partial divergence: equities risk-on, dollar flat, gold firm, Bitcoin lagging. Monthly, BTC +21.6% against S&P -1.2%, so the drag reads crypto-specific rather than macro [INFERRED].

Bitcoin in detail
Price
$77,199.73, +0.8% on the day, -3.1% on the week, +21.6% on the month [OBSERVED]
Higher timeframe
Monthly uptrend intact but decelerating; the $80,000 weekly lower-high ceiling is unbroken [DERIVED]
Daily
Failed reclaim: close about $1,100 under the $78,300 pivot; range $76,500 to $78,300 [DERIVED]
Funding
OKX 0.0033% per 8h, 36th percentile, cheaper than 0.0046% two hours earlier [OBSERVED]
Open interest
OKX $2.09B, -1.9% on the week; Hyperliquid $2.79B [OBSERVED]
Spot vs leverage
Price up with open interest down and funding cheaper is long unwind and deleveraging, not new shorting; spot flows UNKNOWN [DERIVED]
Liquidations
Dollar figures UNKNOWN; the open interest drift implies voluntary exits, not a cascade [INFERRED]
Support
$76,500, then $74,000
Pivot
$76,500, with the close about $700 above it
Resistance
$80,000 weekly lower-high ceiling
Crowded side
Neither side crowded: funding 36th percentile and Hyperliquid funding slightly negative [DERIVED]
What leverage costs · last 33 daysnow 0.0033% · p36 of 90d
Funding, 100 eight-hour prints from Aug 9 to Sep 11, between -0.0027% and 0.0100%; the top decile of the window starts at 0.0100%; the last print 0.0060%0.0100%-0.0027%00.0060%Aug 9Aug 16Aug 23Aug 30Sep 6Sep 11
What to notice: OKX 0.0033% per 8h, 36th percentile, cheaper than 0.0046% two hours earlier [OBSERVED]
The levels, and what to do at each

Written before the move, so no one is deciding under pressure.

  1. $80,000 Weekly lower-high ceiling +3.6%De-risk spot adds here unless funding stays under 0.01% per 8h.
  2. $78,300 Failed reclaim pivot +1.4%A daily close above kills the range thesis; switch to spot adds toward $80,000.
  3. $77,200now
  4. $76,500 Range floor -0.9%A daily close below resumes defensive cuts and opens the $74,000 rung.
  5. $74,000 First ladder rung -4.1%Keep the bid queued; fills only if the floor fails.
  6. $72,000 Second rung -6.7%Second tranche if the range breaks to the downside.

Where the market is positioned

what changed since r3 (6:34 PM ET)compare with Thu Sep 10 EODr1r2

Price now $77,200, from $77,176; most of the structure table was rewritten.

r3 · 6:34 PM ET

Partial divergence: equities risk-on, dollar flat, gold firm, Bitcoin lagging. If equities stay strong and Bitcoin still cannot clear $78,300, the drag is crypto-specific rather than macro [INFERRED].

Bitcoin in detail
Price
$77,175.61, +0.8% on the day, -3.1% on the week, +21.6% on the month [OBSERVED]
Higher timeframe
Monthly uptrend intact but decelerating; the weekly lower-high ceiling at $80,000 is unbroken [OBSERVED]
Daily
Failed reclaim: close $1,124 below the $78,300 pivot; range $76,500 to $78,300 [DERIVED]
Funding
OKX 0.0046% per 8h, 48th percentile of 90 days, up from 0.0024% in the morning but nowhere near crowded [OBSERVED]
Open interest
OKX $2.10B, -1.9% on the week and about $130M below the morning; Hyperliquid $2.79B, about $90M lower [OBSERVED]
Spot vs leverage
Price down plus OI down on both observed venues is long unwind and deleveraging, not new shorting; no spot-flow data to confirm [DERIVED]
Liquidations
Dollar figures UNKNOWN; the OI drawdown implies voluntary exits more than a cascade [INFERRED]
Support
$76,500, then $74,000
Pivot
$78,300, failed today and still the line
Resistance
$80,000 weekly lower-high ceiling
Crowded side
Neither side crowded: funding mid-range at the 48th percentile and Hyperliquid funding negative [DERIVED]
The levels, and what to do at each

Written before the move, so no one is deciding under pressure.

  1. $80,000 Weekly lower-high ceiling +3.7%De-risk spot adds here unless funding stays under 0.01% per 8h.
  2. $78,300 Failed reclaim pivot +1.5%A daily close above kills the range thesis; switch to spot adds targeting $80,000.
  3. $77,176now
  4. $76,500 Reclaimed floor -0.9%A daily close below resumes defensive cuts and opens the $74,000 rung.
  5. $74,000 First ladder rung -4.1%Keep the bid queued; fills only if the floor fails.
  6. $72,000 Second rung -6.7%Second tranche if the range breaks to the downside.

r4 · 7:50 PM ET · this version

Partial divergence: equities risk-on, dollar flat, gold firm, Bitcoin lagging. Monthly, BTC +21.6% against S&P -1.2%, so the drag reads crypto-specific rather than macro [INFERRED].

Bitcoin in detail
Price
$77,199.73, +0.8% on the day, -3.1% on the week, +21.6% on the month [OBSERVED]
Higher timeframe
Monthly uptrend intact but decelerating; the $80,000 weekly lower-high ceiling is unbroken [DERIVED]
Daily
Failed reclaim: close about $1,100 under the $78,300 pivot; range $76,500 to $78,300 [DERIVED]
Funding
OKX 0.0033% per 8h, 36th percentile, cheaper than 0.0046% two hours earlier [OBSERVED]
Open interest
OKX $2.09B, -1.9% on the week; Hyperliquid $2.79B [OBSERVED]
Spot vs leverage
Price up with open interest down and funding cheaper is long unwind and deleveraging, not new shorting; spot flows UNKNOWN [DERIVED]
Liquidations
Dollar figures UNKNOWN; the open interest drift implies voluntary exits, not a cascade [INFERRED]
Support
$76,500, then $74,000
Pivot
$76,500, with the close about $700 above it
Resistance
$80,000 weekly lower-high ceiling
Crowded side
Neither side crowded: funding 36th percentile and Hyperliquid funding slightly negative [DERIVED]
The levels, and what to do at each

Written before the move, so no one is deciding under pressure.

  1. $80,000 Weekly lower-high ceiling +3.6%De-risk spot adds here unless funding stays under 0.01% per 8h.
  2. $78,300 Failed reclaim pivot +1.4%A daily close above kills the range thesis; switch to spot adds toward $80,000.
  3. $77,200now
  4. $76,500 Range floor -0.9%A daily close below resumes defensive cuts and opens the $74,000 rung.
  5. $74,000 First ladder rung -4.1%Keep the bid queued; fills only if the floor fails.
  6. $72,000 Second rung -6.7%Second tranche if the range breaks to the downside.

How it transmits

The path from the news to the position, step by step. If one step breaks, the call breaks with it.

  1. 10-year 4.97%, +21bp on the week, while the effective funds rate sits flat at 3.63%: market rates repriced, policy did not.
  2. Dollar index 99.1, flat on the week, and gold $4,390, +0.6% on the day: no dollar or haven confirmation of a fresh risk-off impulse.
  3. S&P +0.9% and Nasdaq +1.0% closed risk-on, yet Bitcoin closed under its pivot, so the bid is not reaching crypto.
  4. Implication: Bitcoin ranges into the Sep 16 Fed unless the 10-year tops 5.00% (headwind) or spot demand returns (tailwind).

How it transmits

what changed since r3 (6:34 PM ET)compare with Thu Sep 10 EODr1r2

The transmission chain was rewritten.

r3 · 6:34 PM ET

  1. 10-year 4.97%, +21bp on the week, while the effective funds rate sits flat at 3.63%: market rates repriced, policy did not.
  2. Dollar index 99.1, unchanged on the week, and gold $4,390, +0.6% on the day: no dollar or haven confirmation of a fresh risk-off impulse.
  3. S&P +0.9% and Nasdaq +1.0% on the day: equities read the week's data as benign.no longer here
  4. Bitcoin closed +0.8% but under its pivot, so the risk-on bid is not reaching crypto.
  5. Implication: Bitcoin ranges into the Sep 16 Fed unless the 10-year tops 5.00% (headwind) or spot demand returns (tailwind).

r4 · 7:50 PM ET · this version

  1. 10-year 4.97%, +21bp on the week, while the effective funds rate sits flat at 3.63%: market rates repriced, policy did not.
  2. Dollar index 99.1, flat on the week, and gold $4,390, +0.6% on the day: no dollar or haven confirmation of a fresh risk-off impulse.
  3. S&P +0.9% and Nasdaq +1.0% closed risk-on, yet Bitcoin closed under its pivot, so the bid is not reaching crypto.
  4. Implication: Bitcoin ranges into the Sep 16 Fed unless the 10-year tops 5.00% (headwind) or spot demand returns (tailwind).

The one story

If you tell one person one thing about today, tell them this.

Bitcoin closed $77,199.73, down 3.1% on the week and inside the $76,500 to $78,300 range, and the session's real information is in leverage, not price: funding at the close is 0.0033% per 8h, cheaper than the 0.0046% two hours earlier, with open interest flat at $2.09B. Positions are leaving, not building.

Equities closed risk-on, S&P +0.9% and Nasdaq +1.0%, and Bitcoin did not follow through above $78,300. With no ETF flow or options data in the record, whether that lag has a crypto-specific cause stays a hypothesis, and the Sep 16 Fed settles it.

The one story

what changed since r3 (6:34 PM ET)compare with Thu Sep 10 EODr1r2

The one story was rewritten.

r3 · 6:34 PM ET

CPI day resolved sideways for Bitcoin: an overnight flush below $76,500 was bought, the bounce stalled under $78,300, and the close at $77,175 left the floor intact and the upside unclaimed while equities closed higher.no longer here

The plumbing matters more than the print: open interest fell on both observed venues and funding stayed mid-range, so this was exit flow, not a new bearish bet. The range is the market saying it has no edge before the Sep 16 Fed decision.no longer here

r4 · 7:50 PM ET · this version

Bitcoin closed $77,199.73, down 3.1% on the week and inside the $76,500 to $78,300 range, and the session's real information is in leverage, not price: funding at the close is 0.0033% per 8h, cheaper than the 0.0046% two hours earlier, with open interest flat at $2.09B. Positions are leaving, not building.

Equities closed risk-on, S&P +0.9% and Nasdaq +1.0%, and Bitcoin did not follow through above $78,300. With no ETF flow or options data in the record, whether that lag has a crypto-specific cause stays a hypothesis, and the Sep 16 Fed settles it.

The week ahead

What to do before each event, and how to read it afterwards.

EventWhen (ET)ConsensusPriorBefore it Hot / hawkish Soft / dovish
Weekend perp tape, no US dataSat Sep 12, 12:00 AM ETn/an/aSize down; thin books swing funding with no cash market to anchor it.A weekend push through $78,300 is not a breakout; do not chase.A drift to $76,500 is the range working, not a new signal.
FOMC rate decisionWed Sep 16, 2:00 PM ETn/an/aSpot only, ladder bids at $74,000 and $72,000, no directional perps into the decision.A hike or hawkish tone tests $76,500 and fills the $74,000 rung.A pause or dovish tone opens $78,300 and $80,000, spot only.
Weekend perp tape, no US dataSat Sep 12, 12:00 AM ET
before itSize down; thin books swing funding with no cash market to anchor it.
hot ▲A weekend push through $78,300 is not a breakout; do not chase.
soft ▼A drift to $76,500 is the range working, not a new signal.
FOMC rate decisionWed Sep 16, 2:00 PM ET
before itSpot only, ladder bids at $74,000 and $72,000, no directional perps into the decision.
hot ▲A hike or hawkish tone tests $76,500 and fills the $74,000 rung.
soft ▼A pause or dovish tone opens $78,300 and $80,000, spot only.

The week ahead

what changed since r3 (6:34 PM ET)compare with Thu Sep 10 EODr1r2

Nothing measured moved here; the wording changed.

r3 · 6:34 PM ET

Weekend perp tape, no US dataSat Sep 12, 12:00 AM ET
before itSize down; funding can swing on thin books with no cash market to anchor it.
hot ▲A push through $78,300 on weekend volume is not a breakout; do not chase.
soft ▼A drift toward $76,500 is the range working, not a new signal.
FOMC rate decisionWed Sep 16, 2:00 PM ET
before itHold spot, keep ladder bids at $74,000 and $72,000, no directional futures into the decision.
hot ▲A hike or hawkish guidance tests $76,500, then the $74,000 rung fills.
soft ▼A pause or dovish tone opens $78,300 and $80,000, with spot only.

r4 · 7:50 PM ET · this version

Weekend perp tape, no US dataSat Sep 12, 12:00 AM ET
before itSize down; thin books swing funding with no cash market to anchor it.
hot ▲A weekend push through $78,300 is not a breakout; do not chase.
soft ▼A drift to $76,500 is the range working, not a new signal.
FOMC rate decisionWed Sep 16, 2:00 PM ET
before itSpot only, ladder bids at $74,000 and $72,000, no directional perps into the decision.
hot ▲A hike or hawkish tone tests $76,500 and fills the $74,000 rung.
soft ▼A pause or dovish tone opens $78,300 and $80,000, spot only.

What would change the call

Written now, so the goalposts cannot move later.

  • A daily close above $78,300The range resolves up; switch to spot adds targeting $80,000, still no leverage.
  • A daily close below $76,500 with OKX open interest still shrinkingThe floor failed for real; resume cuts and let the $74,000 rung fill.
  • 10-year above 5.00% or OKX funding above 0.01% per 8h on a bounceStand down; treat any strength as de-risk, not trend.

What would change the call

what changed since r3 (6:34 PM ET)compare with Thu Sep 10 EODr1r2

The invalidation conditions changed.

r3 · 6:34 PM ET

  • A daily close above $78,300The range resolves up; switch to spot adds targeting $80,000, still no leverage.
  • A daily close below $76,500 with OI still fallingThe floor failed for real; resume cuts and let the $74,000 rung fill.
  • 10-year above 5.00% or funding above 0.01% per 8h on a bounceStand down; treat any strength as de-risk, not trend.

r4 · 7:50 PM ET · this version

  • A daily close above $78,300The range resolves up; switch to spot adds targeting $80,000, still no leverage.
  • A daily close below $76,500 with OKX open interest still shrinkingThe floor failed for real; resume cuts and let the $74,000 rung fill.
  • 10-year above 5.00% or OKX funding above 0.01% per 8h on a bounceStand down; treat any strength as de-risk, not trend.

What to watch

Conditional triggers, not predictions.

  • IF Bitcoin closes above $78,300 AND funding stays under 0.01% per 8hthe range resolves up; add spot toward $80,000, no leverage.
  • IF a close falls below $76,500 AND OKX open interest keeps shrinkingthe floor failed; resume cuts, let $74,000 fill, lower the ladder to $72,000.
  • IF equities stay risk-on AND Bitcoin fails $78,300 for two straight sessionsthe drag is crypto-specific; sell rallies into $78,300 and keep ladder bids queued.
  • IF the 10-year tops 5.00% OR OKX funding jumps above 0.01% per 8hmacro headwind or re-crowded longs; stand down and wait for the Fed.

What to watch

what changed since r3 (6:34 PM ET)compare with Thu Sep 10 EODr1r2

2 watch lines were replaced.

r3 · 6:34 PM ET

  • IF Bitcoin closes above $78,300 AND funding stays under 0.01% per 8hthe range resolves up; add spot toward $80,000 and keep no leverage.
  • IF a close falls below $76,500 AND OKX open interest keeps shrinkingthe floor failed; resume cuts, let $74,000 fill, and lower the ladder to $72,000.
  • IF equities stay risk-on AND Bitcoin fails $78,300 for two straight sessionsthe drag is crypto-specific; sell rallies into $78,300 and keep ladder bids queued.
  • IF the 10-year tops 5.00% OR OKX funding jumps above 0.01% per 8hmacro headwind or re-crowded longs; stand down and wait for the Fed.

r4 · 7:50 PM ET · this version

  • IF Bitcoin closes above $78,300 AND funding stays under 0.01% per 8hthe range resolves up; add spot toward $80,000, no leverage.
  • IF a close falls below $76,500 AND OKX open interest keeps shrinkingthe floor failed; resume cuts, let $74,000 fill, lower the ladder to $72,000.
  • IF equities stay risk-on AND Bitcoin fails $78,300 for two straight sessionsthe drag is crypto-specific; sell rallies into $78,300 and keep ladder bids queued.
  • IF the 10-year tops 5.00% OR OKX funding jumps above 0.01% per 8hmacro headwind or re-crowded longs; stand down and wait for the Fed.

What could not be verified

Named so the call can be weighed properly. A brief that never lists gaps is not being honest about its inputs.

  • Sep 11 CPI, PPI, claims, UoM and ECB actuals: UNKNOWN for a fourth consecutive revision.
  • Spot bitcoin ETF flows, stablecoin issuance and exchange flows: UNKNOWN, so whether real capital is entering cannot be tested.
  • All options data, including any Sep 16 event-vol pricing: UNKNOWN.
  • CME FedWatch odds and FOMC consensus: UNKNOWN.
  • Market-wide open interest and liquidation dollars, ETH, dominance, Treasury issuance, Fed balance sheet and reserves: UNKNOWN.

What could not be verified

what changed since r3 (6:34 PM ET)compare with Thu Sep 10 EODr1r2

Nothing measured moved here; the wording changed.

r3 · 6:34 PM ET

  • Actual Sep 11 CPI prints (core and headline, m/m and y/y): unavailable for a third consecutive revision.no longer here
  • UoM sentiment and 5-year inflation expectations actuals: unavailable.no longer here
  • Spot bitcoin ETF flows, stablecoin issuance, and exchange flows: unavailable, so whether real capital is entering is UNKNOWN.
  • All options data: implied vol, skew, strike concentration, and Sep 16 event pricing are UNKNOWN.no longer here
  • CME FedWatch odds and FOMC consensus: unverified.
  • Cross-venue aggregate open interest, market-wide liquidation dollars, ETH price, and dominance: UNKNOWN.no longer here
  • Treasury issuance, refunding, auction demand, Fed balance sheet, reserves, RRP, TGA: UNKNOWN.no longer here
  • Calendar beyond the Sep 16 FOMC (PCE date, auctions, large expiries): UNKNOWN.no longer here

r4 · 7:50 PM ET · this version

  • Sep 11 CPI, PPI, claims, UoM and ECB actuals: UNKNOWN for a fourth consecutive revision.
  • Spot bitcoin ETF flows, stablecoin issuance and exchange flows: UNKNOWN, so whether real capital is entering cannot be tested.
  • All options data, including any Sep 16 event-vol pricing: UNKNOWN.
  • CME FedWatch odds and FOMC consensus: UNKNOWN.
  • Market-wide open interest and liquidation dollars, ETH, dominance, Treasury issuance, Fed balance sheet and reserves: UNKNOWN.
If you remember one thing

Hold core spot, no perp leverage into the Fed, ladder bids at $74,000 and $72,000, alerts at $78,300 and $76,500.

Contribute the next revision

Connect an OpenRouter key and press run. The key stays in your browser; three researchers, one agreement, one write, and what results is published here for everyone, with its sources and its checks on the page.

3 × deepseek/deepseek-v4.1-flash·$0.10·5 m 03 s·12 web sources·no silent reads·run on a connected key

Information for the reader's own decisions, not financial advice.

sources: 12 answered · 0 silent
silent (keyless reads that did not answer)

none: every read answered

128823 in / 37664 out tokens

the agreed research the brief was written from
# AGREED

**Macro plumbing**
- 10y Treasury 4.97%, +21bp w/w; 2y 4.40%, +21bp w/w, stamped 2026-09-09; EFFR 3.63%, 0bp w/w as of 2026-09-10 [OBSERVED, finance.yahoo.com ^TNX and 2YY=F; newyorkfed.org EFFR; all three].
- 2y stamp lags the 10y by two days, so the curve comparison mixes dates [OBSERVED, finance.yahoo.com; held by A and C].
- FOMC decision Wed 2026-09-16, 14:00 ET, carried forward; consensus and pricing unverified [calendar feed; A and C].
- Fed balance sheet, reserves, RRP, TGA; Treasury issuance, refunding, auctions, dealer and foreign demand: UNKNOWN, no source in any set [A and C].
- CME FedWatch odds, FOMC consensus, dot-plot expectations, latest Fed communication: UNKNOWN [A and C].
- HY OAS, credit spreads, financial conditions, dollar liquidity, collateral: UNKNOWN [A and C].

**Cross-asset tape**
- S&P 7,656.98, +0.9% d, -1.2% w, -1.2% m; Nasdaq 26,333.04, +1.0% d, -0.9% w, -1.0% m [OBSERVED, finance.yahoo.com, close 2026-09-11; all three].
- DXY 99.1, 0.0% d, +0.1% w, -0.9% m; gold $4,390, +0.6% d, -2.3% w [OBSERVED, finance.yahoo.com, close 2026-09-11; all three].
- WTI $99.99, -2.4% d, +9.5% w, +20.1% m; Brent $104.42, -3.0% d, +9.3% w, +17.4% m [OBSERVED, finance.yahoo.com, close 2026-09-11; all three].
- Monthly divergence live: BTC +21.6% against S&P -1.2% and Nasdaq -1.0% [OBSERVED, finance.yahoo.com, 2026-09-11; all three].
- ETH price, BTC dominance, alt breadth, crypto equities: UNKNOWN [A and C].

**Bitcoin structure**
- BTC $77,199.73, +0.8% d, -3.1% w, +21.6% m [OBSERVED, Binance BTCUSDT, close 2026-09-11; all three].
- Close is $24 above r3's $77,175.61: immaterial, same structure [DERIVED from OBSERVED; B and C].
- Failed $78,300 reclaim, $76,500 floor intact, range $76,500 to $78,300 [DERIVED from OBSERVED; all three]. The $78,300 pivot is a DERIVED structural level, not a sourced print.
- PerpFinder Binance spot $78,308 at 12:27 UTC sits above the close, consistent with an intraday fade [perpfinder.com/asset/BTC; A, B, C].
- "Weekly lower-high ceiling at $80,000" is carried from r3 with no search source in any set; DERIVED, not observed [all three].

**Derivatives and positioning**
- OKX BTC-USDT-SWAP funding 0.0033% per 8h, 7d avg 0.004%, 36th percentile of 90d [OBSERVED, okx.com; all three]. Direction of travel: 0.0046%/48th (r3) → 0.0033%/36th, cheaper into a green close.
- OKX BTC-USDT perp OI $2.09B, -1.9% w/w, versus r3's $2.10B [OBSERVED, okx.com; all three]. Same series two hours apart; not a contradiction beyond the OBSERVED print.
- Hyperliquid BTC perp funding -0.00005% per hour, OI $2.79B, mark $77,158 [OBSERVED, hyperliquid.xyz; all three]. r3 print was -0.0002%/hr, so shorts still pay longs, less so.
- Read: price flat-to-up with OI flat-to-down on both observed venues and funding cheaper is deleveraging, not re-crowding [DERIVED; A and C].
- Hyperliquid settles funding hourly, so its rate is not comparable to 8h venues [hyperliquid.gitbook.io/Hyperliquid-docs/trading/funding, undated docs; perp.wiki/funding-rates; A and C].
- Cross-venue OI map, undated: Binance $8.47B (22.9%), Gate.io $4.51B, Bybit $4.31B, MEXC $3.74B, Hyperliquid $2.80B (7.6%), Bitget $2.62B, HTX $2.32B, OKX $2.11B, Deribit $863M [sharpe.ai/futures/bitcoin, undated; A, B, C]. Shape of market only; corroborates the OBSERVED OKX $2.09B.
- Bybit funding +0.0051%/8h, Hyperliquid +0.0013%/8h, OI $3.74B across three venues [perpsbtc.com, undated; B]. Consistent with mild positive long premium.
- Market-wide OI, market-wide liquidation dollars, perp basis, perp premium, futures and spot CVD, realized vol: UNKNOWN [A and C].

**Capital flows**
- Spot BTC ETF flows, stablecoin issuance, exchange in/outflows, spot volume trend, institutional proxies: UNKNOWN, no source in any set [A and C]. Real-capital test cannot be run.

**Options**
- All options data: IV, IV vs realized, term structure, skew, put/call, strike concentration, Sep 16 event vol, expected move: UNKNOWN [A and C].

**Calendar / catalysts**
- Sep 11 Core CPI m/m forecast 0.2%, prior 0.2%; Core CPI y/y forecast 2.4%, prior 2.5%; CPI m/m forecast 0.4%, prior 0.1%; CPI y/y forecast 3.4%, prior 3.4%; UoM sentiment forecast 51.0, prior 51.0; UoM inflation expectations prior 4.3% [calendar feed, 08:30 and 10:00 ET 2026-09-11; all three].
- Sep 10 Core PPI m/m forecast 0.3%, PPI m/m forecast 0.4%, claims forecast 205K; ECB Main Refinancing Rate forecast 2.65% vs prior 2.40%; GBP GDP m/m forecast 0.0% vs prior 0.3% [calendar feed; all three].
- Actual prints for CPI (core and headline, m/m and y/y), PPI, claims, UoM, ECB, GBP GDP: UNKNOWN for a fourth revision [A and C].
- PCE date, NFP, auctions, large expiries in the next 7 days: UNKNOWN [A and C].

**The one story**
- No dominant single driver surfaced in any set [A and C]. The equity risk-on close against the BTC lag, and whether it has a crypto-specific cause, is HYPOTHESIS only.

**Methodology / discard list**
- Stale or price-inconsistent prints to discard: MarginPad $64,259 (OI $28.96B, bands $62,050/$65,650); ARX $64,459 (build 2026-07-29); loris.tools $79,791 (updated 2026-09-06) and its Hyperliquid 24h liquidation $1.29M [all three].
- Quantority OI internally inconsistent ($16.33B headline vs $18.36B total) and undated; funding APR Binance +6.04%, Bybit +7.51%, Bitget +8.87% [quantority.com/open-interest/BTC; held by C only, A and B did not find it].

# RESOLVED

- PerpFinder Hyperliquid annualized +4.40% vs OBSERVED -0.00005%/hr: OBSERVED wins; different snapshot or oracle convention. Losing claim held by A, B, C.
- PerpFinder Binance annualized +10.95% vs OBSERVED OKX 0.0033%/8h: not the same window; the annualized figure cannot characterize current funding. Losing claim held by A, B, C.
- PerpFinder $78,308 (12:27 UTC) and perpsbtc.com $78,373.6 as current price: both predate the 23:45 UTC OBSERVED $77,199.73 and are superseded. OBSERVED wins. Losing claim held by A, B, C.
- perpsbtc.com price and its Bybit/Hyperliquid funding are undated and price-inconsistent with the close; not usable as current. Losing claim held by A, B, C.
- perp.wiki Hyperliquid -0.0017%/hr vs OBSERVED -0.00005%/hr: OBSERVED is the measured read. Losing claim held by B.
- sharpe.ai OI sum: A computed ~$36.4B, C ~$37.0B. Accept the line items, reject both sums; no timestamp.
- OKX OI $2.10B (r3) vs $2.09B (OBSERVED): same series two hours apart, no contradiction.
- "Weekly lower-high ceiling at $80,000": keep only as DERIVED, not a sourced level. Losing claim held by all three.
- r3's conflict line that funding rose to 0.0046% into a falling close and implied re-crowding: overturned by OBSERVED 0.0033%/36th. Losing claim held by all three.
- Close $77,199.73 vs r3 $77,175.61: same state, not a revision.

# UNRESOLVED

- Actual Sep 11 CPI prints, PPI, claims, UoM actuals, ECB decision, GBP GDP: UNKNOWN despite a fourth revision without them.
- CME FedWatch odds, FOMC consensus and dot-plot path: UNKNOWN.
- Spot BTC ETF flows, stablecoin issuance, exchange in/outflows, spot volume trend: UNKNOWN. Whether real capital or only leverage is moving this market; largest missing piece.
- All options data, including any Sep 16 event-vol pricing: UNKNOWN.
- Market-wide OI beyond the undated tables, market-wide liquidation dollars for Sep 11, perp basis, perp premium, spot and futures CVD, realized vol: UNKNOWN.
- ETH price, BTC dominance, alt breadth, HY OAS, credit spreads, financial conditions: UNKNOWN.
- Fed balance sheet, reserves, RRP, TGA; Treasury issuance, refunding, auctions, dealer and foreign demand: UNKNOWN.
- Calendar beyond Sep 16 (PCE, NFP, auctions, expiries): UNKNOWN.
- The dominant single driver of the Sep 11 session; whether the equity risk-on close and the BTC lag have a crypto-specific cause: HYPOTHESIS only.
- Quantority OI figures and funding APRs: UNKNOWN, internally inconsistent and undated; held by C only, A and B not finding it is a gap, not a contradiction.
- Researcher B's position is missing, not empty; B's independent findings are UNKNOWN.
how the three researchers agreed

3 of 3 researchers returned notes; 2 took a position on the others' notes; deepseek/deepseek-v4.1-flash consolidated them into the one agreed document the brief was written from.

the notes, one researcher each
A · deepseek/deepseek-v4.1-flash · 1,558 words
# Research notes, revision 4. Fri Sep 11, 2026, 7:45 PM ET

All OBSERVED rows come from the page's own keyless reads taken 2026-09-11T23:45:19Z. Everything else is dated and attributed. Where no current search result existed, the item is UNKNOWN.

## Macro plumbing
- Fed funds effective 3.63%, flat on the week, as of 2026-09-10 [OBSERVED, NY Fed EFFR].
- US 10-year 4.97%, +21 bp on the week, as of 2026-09-11 close [OBSERVED, Yahoo ^TNX].
- US 2-year 4.40%, +21 bp on the week, as of 2026-09-09 [OBSERVED, Yahoo 2YY=F]; note the 2-year stamp is two days older than the 10-year stamp, so the curve read is a mixed-date comparison.
- DXY 99.1, 1d 0.0%, 1w +0.1%, 1m -0.9%, as of 2026-09-11 close [OBSERVED, Yahoo DX-Y.NYB].
- ECB Main Refinancing Rate was scheduled 2026-09-10 08:15 ET with forecast 2.65% vs prior 2.40%, a hike, per the page calendar; actual print UNKNOWN (no search result carried it).
- UK GDP m/m scheduled 2026-09-11 02:00 ET, forecast 0.0%, prior 0.3%; actual UNKNOWN.
- Fed balance sheet, reserves, RRP, TGA: UNKNOWN. No search result returned current levels.
- Treasury issuance, refunding schedule, bill-vs-coupon mix, auction demand, dealer absorption, foreign demand: UNKNOWN.
- CME FedWatch odds for the Sep 16 FOMC and the Fed's own latest communication: UNKNOWN. Nothing in the search set addressed it.
- HY OAS and financial conditions indexes: UNKNOWN.

## Cross-asset
- S&P 500 7,656.98, 1d +0.9%, 1w -1.2%, 1m -1.2%, 2026-09-11 close [OBSERVED, Yahoo ^GSPC].
- Nasdaq Composite 26,333.04, 1d +1.0%, 1w -0.9%, 1m -1.0%, 2026-09-11 close [OBSERVED, Yahoo ^IXIC].
- Gold COMEX front $4,390, 1d +0.6%, 1w -2.3%, 1m -0.4%, 2026-09-11 close [OBSERVED, Yahoo GC=F].
- WTI front $99.99, 1d -2.4%, 1w +9.5%, 1m +20.1%, 2026-09-11 close [OBSERVED, Yahoo CL=F]; Brent $104.42, 1d -3.0%, 1w +9.3%, 1m +17.4% [OBSERVED].
- Bitcoin 1m +21.6% against S&P 1m -1.2% and Nasdaq 1m -1.0%: the monthly divergence is still live [OBSERVED].
- ETH spot price: UNKNOWN in the observed set; last dated alternative is $2,493.3 on 2026-09-06, five days stale, [loris.tools](https://loris.tools/markets/exchanges/hyperliquid). BTC dominance and alt breadth: UNKNOWN.

## Bitcoin structure
- Bitcoin spot $77,199.73, 1d +0.8%, 1w -3.1%, 1m +21.6%, 2026-09-11 close [OBSERVED, Binance BTCUSDT].
- Conflicting same-day prints: $78,308 at 12:27 UTC [perpfinder.com](https://perpfinder.com/asset/BTC) and $78,373.6 undated [perpsbtc.com](https://perpsbtc.com/). Both sit above the observed close; resolve in favor of the observed 23:45 UTC read.
- Stale prints to discard: $79,791 last updated 2026-09-06T13:27:18Z [loris.tools](https://loris.tools/markets/exchanges/hyperliquid); $64,259 undated [marginpad.io](https://marginpad.io/coin/btc/); $64,459 with build snapshot 2026-07-29 07:54 UTC [arx.trade](https://arx.trade/markets/btc/); $66,445 undated [beacontrade.io](https://beacontrade.io/coins/btc/open-interest).
- HTF: 1m +21.6% and 1w -3.1% together describe a monthly uptrend decelerating into a weekly pullback. No search result independently characterized the weekly structure, so the $80,000 lower-high ceiling remains a [DERIVED] read from the observed series, not a sourced one.
- Volatility regime, realized vol, and volume data: UNKNOWN.

## Derivatives and positioning
- OKX BTC-USDT-SWAP funding 0.0033% per 8h now, 7d avg 0.004%, 36th percentile of 90 days [OBSERVED, OKX]. Down from 0.0046% and the 48th percentile at the 22:26 UTC revision: cheap and drifting cheaper.
- OKX BTC-USDT perp open interest $2.09B, 7d -1.9% [OBSERVED, OKX]. One venue only.
- Hyperliquid BTC perp funding -0.00005% per hour (negative, shorts paying nothing meaningful, longs not paying up), OI $2.79B, mark $77,158 [OBSERVED, Hyperliquid].
- Offsetting funding prints from other venues, all undated or loosely dated: Hyperliquid annualized +4.40% as the highest executable and dYdX -0.11% as the lowest, executable spread 4.51%, at 12:27 UTC [perpfinder.com](https://perpfinder.com/asset/BTC); Bybit +0.0051% per 8h and Hyperliquid +0.0013% per 8h with $3.74B OI [perpsbtc.com](https://perpsbtc.com/); Hyperliquid BTC funding +0.01% per 8h, 10.95% APY, 7d avg +0.009%, as of 2026-09-06 [loris.tools](https://loris.tools/markets/exchanges/hyperliquid).
- Conflict to preserve: the observed Hyperliquid read (negative overnight, ~-0.44% annualized equivalent) contradicts perpfinder's +4.40% annualized for the same venue. Different snapshot times and, possibly, different oracle/indexing conventions. The observed hourly figure is the one this page measured.
- Cross-venue open interest, undated, [sharpe.ai](https://www.sharpe.ai/futures/bitcoin): Binance $8.47B (22.9%), Gate.io $4.51B, Bybit $4.31B, MEXC $3.74B, Hyperliquid $2.80B (7.6%), Bitget $2.62B, HTX $2.32B, OKX $2.11B, WhiteBIT $1.59B, KuCoin $1.21B, Deribit $863.16M, Bitfinex $701.17M, Crypto.com $488.84M, BingX $417.92M, GRVT $200.99M. Sums to roughly $36.4B tracked. No timestamp shown, so treat as a shape-of-market map, not a dated level. Note OKX $2.11B there versus $2.09B observed: close enough to corroborate the venue read.
- Liquidations: 24h Hyperliquid BTC $1.29M as of 2026-09-06 [loris.tools](https://loris.tools/markets/exchanges/hyperliquid), five days stale. Market-wide liquidation dollars for today: UNKNOWN. MarginPad's clusters (long band near $62,050, short band near $65,650) sit in the stale $64K price regime and are unusable [marginpad.io](https://marginpad.io/coin/btc/).
- Basis and perp premium: Hyperliquid mark/oracle basis -4.76 bps, build 2026-07-29, stale [arx.trade](https://arx.trade/markets/btc/). Current basis: UNKNOWN.
- Platform mechanics worth one line: Hyperliquid settles funding hourly, not every 8 hours, so its rate is not directly comparable to Bybit, OKX, or Binance 8h rates [perp.wiki](https://perp.wiki/funding-rates); funding is the periodic payment between longs and shorts that keeps the perp anchored to spot [satoshimacro.com](https://satoshimacro.com/tools/crypto/derivatives/btc-funding-rate/).

## Capital flows
- Spot bitcoin ETF flows, daily and weekly: UNKNOWN. No search result returned them.
- Stablecoin issuance or contraction: UNKNOWN.
- Exchange in/outflows: UNKNOWN.
- Institutional proxies, spot volume trend, futures CVD: UNKNOWN.
- Consequence: whether real capital or only leverage is moving this market cannot be established from today's evidence. This is the single biggest gap in the current read.

## Options
- Implied vol, IV change, IV versus realized: UNKNOWN.
- Term structure, skew, put/call demand: UNKNOWN.
- Strike concentration and major OI strikes: UNKNOWN. Deribit venue OI of $863.16M appears only inside the undated sharpe.ai table [sharpe.ai](https://www.sharpe.ai/futures/bitcoin).
- Sep 16 event vol and expected move: UNKNOWN.
- No search result in this set touched options pricing at all. Direction and vol pricing cannot be separated today.

## Calendar, with consensus and prior
- 2026-09-11 08:30 ET USD Core CPI m/m, forecast 0.2%, prior 0.2%; Core CPI y/y, forecast 2.4%, prior 2.5%; CPI m/m, forecast 0.4%, prior 0.1%; CPI y/y, forecast 3.4%, prior 3.4%. Actuals UNKNOWN for a fourth consecutive revision.
- 2026-09-11 10:00 ET Prelim UoM Consumer Sentiment, forecast 51.0, prior 51.0; UoM Inflation Expectations, prior 4.3%. Actuals UNKNOWN.
- 2026-09-10 08:30 ET Core PPI m/m forecast 0.3%, prior 0.2%; PPI m/m forecast 0.4%, prior 0.0%; Unemployment Claims forecast 205K, prior 206K. Actuals UNKNOWN.
- 2026-09-10 08:15 ET ECB Main Refinancing Rate, forecast 2.65%, prior 2.40%, plus press conference 08:45 ET. Actuals UNKNOWN.
- 2026-09-11 02:00 ET GBP GDP m/m, forecast 0.0%, prior 0.3%. Actual UNKNOWN.
- 2026-09-09 21:15 ET President Trump speaks; no forecast or prior given.
- FOMC decision: Wed 2026-09-16 14:00 ET, carried forward from the prior revision. Consensus and FedWatch pricing: UNKNOWN.
- PCE date, Treasury auction schedule, large options expiries beyond Sep 16: UNKNOWN.

## The one story
- UNKNOWN. No search result in this set identified a dominant market driver for Sep 11. The only dated, sourced macro-style item is the stale MarginPad page, which is unusable. Nothing in the search set corroborates the day's CPI outcome or explains the equity close.

## Anomalies
- Price +0.8% on the day while funding falls from 0.0046% to 0.0033% and OKX OI slips $2.10B to $2.09B: an up-close with leverage leaving, not entering.
- Funding percentile 48 to 36 in under two hours while the daily close was green.
- Equities risk-on (S&P +0.9%, Nasdaq +1.0%) with Bitcoin unable to hold above $78,300 in any print observed today.
- WTI +20.1% and Brent +17.4% on the month alongside a 10-year up only 21 bp on the week and gold down 2.3% on the week: an oil shock that has not yet transmitted to either the long end or the haven bid, or has not been read that way.
- Bitcoin +21.6% on the month against both major equity indexes negative on the month.
- Binance-venue annualized funding quoted at +10.95% [perpfinder.com](https://perpfinder.com/asset/BTC), [loris.tools](https://loris.tools/markets/exchanges/hyperliquid) against an observed Hyperliquid read near -0.4% annualized equivalent: a real cross-venue dispersion that no single venue's number should be used to characterize as "market funding."

## What contradicts the previous thesis (revision 3)
- Revision 3's conflict bullet "funding ticked up to 0.0046% from 0.0024% into a falling close" does not survive revision 4: observed funding is 0.0033% per 8h, the 36th percentile, so the re-crowding warning it implied eases.
- Revision 3's "price down plus OI down is long unwind, not new shorting" still holds on the observed set: OKX OI $2.09B, -1.9% on the week, unchanged in direction.
- Revision 3's failed-reclaim thesis is neither confirmed nor refuted by revision 4: the close moved $24, from $77,175.61 to $77,199.73, with $76,500 and $78,300 both untouched.
- Revision 3's central unknown, the CPI prints, remains unknown for a fourth revision. The gap list is not shrinking.
- Contrary evidence against revision 3's range framing: an undated sharpe.ai aggregate showing roughly $36.4B of tracked perp open interest and a Hyperliquid share at 7.6% [sharpe.ai](https://www.sharpe.ai/futures/bitcoin) suggests the observed two-venue OI drawdown may understate total market leverage, making the "leverage is leaving" read less certain than one venue's numbers imply.
- Biggest contradiction to the standing thesis: leverage is cheap, the deck is clean, and Bitcoin still closed under its pivot while equities closed green. Cheap funding with a failing pivot is not a bullish tell unless spot demand is shown to be the buyer, and no spot-flow data was found to show it.

Information for the reader's own decisions, not financial advice.
B · deepseek/deepseek-v4.1-flash · 1,404 words
# Research notes, Fri Sep 11 2026, 7:45 PM ET (revision 4). Findings only.

## Macro plumbing
- OBSERVED: US 10y 4.97%, +21bp w/w; 2y 4.40%, +21bp w/w (2y stamped 2026-09-09); EFFR 3.63%, 0bp w/w as of 2026-09-10 [OBSERVED].
- FOMC Sept 16, 2:00 PM ET, next scheduled decision: carried from the prior state only, not re-verified this revision [UNKNOWN].
- CME FedWatch odds, FOMC consensus, dot-plot expectations: UNKNOWN, no source found.
- Fed balance sheet, reserves, RRP, TGA balances: UNKNOWN.
- Treasury issuance, refunding calendar, bill vs coupon mix, auction demand, dealer absorption, foreign demand: UNKNOWN.
- Dollar liquidity, collateral conditions, credit spreads, HY OAS: UNKNOWN.
- No Fed speaker or QT headline surfaced in search [UNKNOWN].

## Cross-asset
- OBSERVED: S&P 7,656.98 +0.9% d/d, -1.2% w/w, -1.2% m/m; Nasdaq 26,333.04 +1.0% d/d, -0.9% w/w, -1.0% m/m [OBSERVED].
- OBSERVED: DXY 99.1, 0% d/d, +0.1% w/w, -0.9% m/m [OBSERVED].
- OBSERVED: gold $4,390 +0.6% d/d, -2.3% w/w, -0.4% m/m [OBSERVED].
- OBSERVED: WTI $99.99 -2.4% d/d but +9.5% w/w, +20.1% m/m; Brent $104.42 -3.0% d/d, +9.3% w/w [OBSERVED]. Oil above $100 with a 10y at 4.97% is the standing macro tension.
- Cross-asset classification: equities risk-on, dollar flat, gold firm, Bitcoin lagging. No independent cross-asset source found this revision [DERIVED].
- ETH price, alt breadth, BTC dominance: UNKNOWN. No source in results.
- Crypto equities, credit, stablecoin cross-asset proxies: UNKNOWN.

## Bitcoin structure
- OBSERVED: BTC $77,199.73, +0.8% d/d, -3.1% w/w, +21.6% m/m, close of 2026-09-11 [OBSERVED]. This is $24 above revision 3's $77,175.61 close; the change versus r3 is a marginally higher close, still below the $78,300 pivot.
- OBSERVED: Hyperliquid BTC mark $77,158, essentially at spot [OBSERVED].
- Search-source cross-check on spot: [perpfinder.com](https://perpfinder.com/asset/BTC) $78,308 from Binance at 12:27 UTC (7:27 AM ET); [perpsbtc.com](https://perpsbtc.com/) $78,373.6; both predate the US close and read above it, consistent with an intraday fade into the close.
- HTF: monthly uptrend intact, +21.6% on the month; weekly still lower-high under the $80,000 ceiling [DERIVED from OBSERVED].
- Daily: failed reclaim of $78,300; range $76,500 to $78,300 [DERIVED from OBSERVED].
- Volume, spot CVD, futures CVD, realized vol, momentum oscillators, VWAPs, major MAs: UNKNOWN.
- Stale source, do not use: [marginpad.io](https://marginpad.io/coin/btc/) shows BTC $64,259 with ATH $126,080 set in 2025 (undated but clearly stale); [arx.trade](https://arx.trade/markets/btc/) build snapshot Jul 29 2026 with BTC $64,459 (stale).

## Derivatives and positioning
- OBSERVED: OKX BTC-USDT-SWAP funding 0.0033% per 8h now, 7d avg 0.004%, 36th percentile of 90d. Versus r3's 0.0046% at 48th percentile: funding ticked down and the percentile fell 12 points [OBSERVED]. Cheap, mid-to-low range, longs not paying up.
- OBSERVED: OKX BTC-USDT perp OI $2.09B, -1.9% w/w (r3 had $2.10B) [OBSERVED].
- OBSERVED: Hyperliquid BTC perp funding -0.00005% per hour (slightly negative, shorts pay longs marginally), OI $2.79B (flat vs r3's $2.79B), mark $77,158 [OBSERVED].
- Conflict: [perp.wiki](https://perp.wiki/funding-rates) lists Hyperliquid BTC at -0.0017% hourly (more negative than OBSERVED); [perpfinder.com](https://perpfinder.com/asset/BTC) shows Hyperliquid annualized +4.40% (positive) and Binance CEX +10.95%, dYdX -0.11%, spread 4.51% annualized. The OBSERVED read is the one to trust; the discrepancy is sampling time and venue definition. Direction of travel: Hyperliquid funding near zero, not crowded long.
- [perpsbtc.com](https://perpsbtc.com/): Bybit funding +0.0051% per 8h (5.6% annualized), Hyperliquid +0.0013% per 8h (1.4%), OI $3.74B across three venues; undated but consistent with a mild positive long premium.
- Cross-venue OI, [sharpe.ai](https://www.sharpe.ai/futures/bitcoin) (undated): Binance $8.47B (22.9%), Gate.io $4.51B, Bybit $4.31B, MEXC $3.74B, Hyperliquid $2.80B (7.6%), Bitget $2.62B, HTX $2.32B, OKX $2.11B, Deribit $863.16M. Treated as approximate; no timestamp.
- [perpfinder.com](https://perpfinder.com/asset/BTC) total OI $7.02B across four snapshot venues; not market-wide.
- Liquidations: aggregate dollars UNKNOWN. Hyperliquid BTC 24h liquidation $1.29M as of 2026-09-06 per [loris.tools](https://loris.tools/markets/exchanges/hyperliquid) (stale); nothing current found.
- Basis, perp premium, futures volume, spot-vs-futures CVD: UNKNOWN.
- Hyperliquid funding mechanics for reference: paid hourly at 1/8 of the 8h rate, interest component 0.01% per 8h, cap 4%/hour, i.e. negative per-hour prints mean longs receive [hyperliquid.gitbook.io](https://hyperliquid.gitbook.io/Hyperliquid-docs/trading/funding).
- Relational read from OBSERVED: price roughly flat to slightly up, OI flat to slightly down on both venues, funding cheaper = positions leaving, not new longs entering. Classification: deleveraging, not re-crowding [DERIVED].
- [satoshimacro.com](https://satoshimacro.com/tools/crypto/derivatives/btc-funding-rate/) has a live aggregate funding chart but returned no current value in the results.

## Capital flows (incl. spot ETF flows)
- Spot bitcoin ETF daily flows, weekly trend, issuance/redemption: UNKNOWN this revision, no source found.
- Stablecoin issuance or contraction: UNKNOWN.
- Exchange in/outflows: UNKNOWN.
- Spot exchange volume, institutional proxies: UNKNOWN.
- Implication of the gap: whether real capital is entering or leaving cannot be established; the spot-led vs leverage-led test rests on price, OI and funding only [UNKNOWN].

## Options
- BTC implied vol, IV change, IV vs realized vol, term structure, 25-delta skew, put/call demand: UNKNOWN.
- Strike concentration, Sep 16 FOMC event pricing, expected move, major expiries: UNKNOWN.
- Deribit BTC OI $863.16M appears in [sharpe.ai](https://www.sharpe.ai/futures/bitcoin) but that figure is perp/derivatives OI, not options metrics, and is undated.
- Net: the options opportunity set cannot be graded this revision; direction and vol pricing cannot be separated [UNKNOWN].

## Calendar: consensus and prior
Provided calendar, Sep 11 (times ET as published):
- 08:30 Core CPI m/m, forecast 0.2%, prior 0.2%; Core CPI y/y, forecast 2.4%, prior 2.5%; CPI m/m, forecast 0.4%, prior 0.1%; CPI y/y, forecast 3.4%, prior 3.4%. Actuals: UNKNOWN (third revision without them).
- 10:00 Prelim UoM Consumer Sentiment, forecast 51.0, prior 51.0; UoM Inflation Expectations, forecast n/a, prior 4.3%. Actuals: UNKNOWN.
- Sep 10: ECB Main Refinancing Rate forecast 2.65% vs prior 2.40% (a tightening); ECB press conference; US Core PPI m/m forecast 0.3% vs prior 0.2%; PPI m/m forecast 0.4% vs prior 0.0%; claims forecast 205K vs prior 206K. Actuals: UNKNOWN.
- Sep 11 02:00 GBP GDP m/m, forecast 0.0%, prior 0.3%. Actual: UNKNOWN.
- Sep 16, 2:00 PM ET FOMC decision: from prior state, consensus and pricing UNKNOWN.
- Beyond Sep 16 (PCE, NFP, auctions, refunding, large expiries, other central banks): UNKNOWN, none found.

## The one story
- No new headline dominated the search results. The standing story is unchanged: Bitcoin closed +0.8% at $77,200 while the S&P rose 0.9% and the Nasdaq 1.0%, so a risk-on equity tape is not transmitting into crypto, and the $78,300 pivot held as a ceiling for a second session [OBSERVED].
- Mechanism: 10y at 4.97% (+21bp w/w) and WTI near $100 keep the discount-rate and inflation channels tight, so the equity bid that lifted on the CPI release did not reach an asset that trades as the long-duration, liquidity-sensitive end of the risk curve [DERIVED].

## Anomalies
- Bitcoin +21.6% on the month against S&P -1.2% and Nasdaq -1.0%: the monthly divergence persists through the weak week [OBSERVED].
- Oil +9.5% w/w and 10y +21bp w/w while equities close risk-on: if inflation expectations were the driver, the equity bid and the oil move should not coexist this calmly [DERIVED, unresolved].
- Funding percentile fell to 36 from 48 while price closed slightly higher: leverage leaving into flat-to-up price, the opposite of a squeeze setup [OBSERVED].
- Hyperliquid funding at -0.00005%/hr with OI flat at $2.79B: a venue that normally carries a positive long premium is pricing no long crowding at all [OBSERVED].
- Search coverage problem itself: every derivatives page returned either stale (Jul 29, Sep 6) or undated, so cross-venue confirmation is weak this revision.

## What contradicts the previous thesis (r3)
- r3 conflict "funding ticked up to 0.0046% into a falling close, some leverage paying to stay long" is WEAKENED: OBSERVED funding is now 0.0033% at the 36th percentile, and Hyperliquid hourly funding is marginally negative, so the re-crowding warning is not firing.
- r3's $76,500 floor holds and r3's failed-reclaim read holds: close $77,200, above the floor, below the $78,300 pivot. Thesis status unchanged, still weakened-not-invalidated.
- r3's core contradiction (equity risk-on vs crypto lagging) is STRENGTHENED: same pattern for a second day, S&P +0.9%, Nasdaq +1.0%, Bitcoin +0.8% and under the pivot.
- r3's macro chain holds and extends: 10y still 4.97%, so the 5.00% de-risk trigger is untested but close.
- r3's gaps on CPI actuals, ETF flows, options and Fed pricing remain open and are the reason confidence cannot rise above moderate.

Unverified items carried as UNKNOWN: CPI and PPI actuals, UoM actuals, ECB outcome, ETF and stablecoin flows, all options data, CME FedWatch, market-wide OI and liquidations, ETH and dominance, Treasury and Fed balance-sheet plumbing. Information for the reader's own decisions, not financial advice.
C · deepseek/deepseek-v4.1-flash · 1,160 words
# Research notes, Fri Sep 11, 2026, 7:45 PM ET

OBSERVED rows below are the page's own keyless reads taken 2026-09-11T23:45:19Z. Everything else is from the search results returned; each line carries its date (or a STALE/UNDATED flag) and source.

## Macro plumbing
- Fed funds effective 3.63%, flat on the week, as of 2026-09-10 (NY Fed EFFR, OBSERVED).
- US 10y 4.97%, +21bp on the week, close 2026-09-11 (Yahoo ^TNX, OBSERVED).
- US 2y 4.40%, +21bp on the week, as of 2026-09-09, a 2-day lag (Yahoo 2YY=F, OBSERVED).
- CME FedWatch odds for the Sep 16 FOMC, FOMC consensus, and dot-plot expectations: UNKNOWN, no source returned.
- Fed balance sheet, reserves, RRP, TGA: UNKNOWN.
- Treasury issuance, refunding, auction demand, dealer absorption, foreign demand: UNKNOWN.
- ECB Main Refinancing Rate forecast 2.65% vs prior 2.40%, Sep 10 08:15 ET (calendar feed). Actual decision and statement: UNKNOWN.
- WTI $99.99, -2.4% d, +9.5% w, +20.1% m; Brent $104.42, -3.0% d, +9.3% w, +17.4% m, close 2026-09-11 (Yahoo CL=F / BZ=F, OBSERVED). Energy is the live inflation-transmission channel.
- DXY 99.1, flat d, +0.1% w, -0.9% m, close 2026-09-11 (Yahoo DX-Y.NYB, OBSERVED).
- Gold $4,390, +0.6% d, -2.3% w, close 2026-09-11 (Yahoo GC=F, OBSERVED).

## Cross-asset
- S&P 7,656.98, +0.9% d, -1.2% w, -1.2% m, close 2026-09-11 (Yahoo ^GSPC, OBSERVED).
- Nasdaq 26,333.04, +1.0% d, -0.9% w, -1.0% m, close 2026-09-11 (Yahoo ^IXIC, OBSERVED).
- BTC +0.8% on the day vs S&P +0.9% and Nasdaq +1.0%: BTC lagged both (OBSERVED).
- One-month: BTC +21.6% vs S&P -1.2% and Nasdaq -1.0%: divergence persists (OBSERVED).
- ETH price, BTC dominance, alt breadth, HY OAS, credit spreads: UNKNOWN, no source returned.

## Bitcoin structure
- Close $77,199.73, +0.8% d, -3.1% w, +21.6% m, 2026-09-11 (Binance spot, OBSERVED).
- Prior revision (r3) recorded the close at $77,175.61: the observed close is $24 higher, immaterial, same read.
- $78,300 pivot failed today and $76,500 floor held: carried from r3, no search source contradicts either.
- Weekly lower-high ceiling at $80,000: carried from r3, no search source corroborates.
- Volume, momentum, realized vol, BTC weekly structure: UNKNOWN.

## Derivatives and positioning
- OKX BTC-USDT perp funding 0.0033% per 8h now, 7d avg 0.004%, 36th percentile of the last 90 days (OBSERVED). r3 had 0.0046% at the 48th percentile.
- OKX BTC-USDT perp OI $2.09B, 7d -1.9% (OBSERVED). r3 had $2.10B: unchanged, still deleveraging.
- Hyperliquid BTC perp funding -0.00005% per hour (roughly -0.44% annualized), OI $2.79B, mark $77,158 (OBSERVED). r3 had -0.0002%/hr.
- Cross-exchange OI table, UNDATED: Binance $8.47B (22.9%), Gate.io $4.51B, Bybit $4.31B, MEXC $3.74B, Hyperliquid $2.80B (7.6%), Bitget $2.62B, HTX $2.32B, OKX $2.11B, Deribit $863M; 25 venues listed, sum about $37.0B [sharpe.ai](https://www.sharpe.ai/futures/bitcoin).
- Conflicting cross-exchange OI, UNDATED and internally inconsistent: the same page shows both $16.33B headline and $18.36B total with 24h +4.0%; funding APR Binance +6.04%, Bybit +7.51%, Bitget +8.87%, OKX +8.28% [quantority.com](https://quantority.com/open-interest/BTC). Low confidence.
- PerpFinder: Binance spot $78,308, annualized funding Binance +10.95%, Hyperliquid +4.40%, Bybit +2.17%, dYdX -0.11%; $7.02B OI over 4 venues; generated 12:27 UTC, date not shown [perpfinder.com](https://perpfinder.com/asset/BTC). Funding far above the OKX/HL reads, so either stale or a different window; UNVERIFIED.
- perpsbtc.com: BTC $78,373.6, OI $3.74B, Bybit +0.0051%/8h, Hyperliquid +0.0013%/8h [perpsbtc.com](https://perpsbtc.com/). UNDATED, price inconsistent with the $77,200 close.
- MarginPad: BTC $64,259, OI $28.96B, funding +0.0032%, $25.9M liquidated in 24h (longs $21.6M), 60.5% of accounts long, long-liquidation band $62,050, short band $65,650, 7d $228.6M across 16,175 events [marginpad.io](https://marginpad.io/coin/btc/). STALE: price 17% below spot, discard.
- ARX: BTC $64,459, Hyperliquid OI $2.31B, hourly funding +0.0013%, build snapshot Jul 29 2026 07:54 UTC [arx.trade](https://arx.trade/markets/btc/). STALE by six weeks.
- Loris Tools: Hyperliquid BTC $79,791, OI $2.81B, funding +0.0100%/8h (7d avg +0.0090%), 24h liquidation $1.29M, last updated 2026-09-06T13:27:18Z [loris.tools](https://loris.tools/markets/exchanges/hyperliquid). Five days stale.
- Market-wide liquidation dollars for Sep 11, perp basis beyond the Hyperliquid mark-oracle read, futures CVD, spot CVD: UNKNOWN.
- Funding-rate mechanics (paid every 4 or 8 hours, or hourly on Hyperliquid) restated by [satoshimacro.com](https://satoshimacro.com/tools/crypto/derivatives/btc-funding-rate/) and [perp.wiki](https://perp.wiki/funding-rates); no data.

## Capital flows
- Spot BTC ETF flows for Sep 11: UNKNOWN.
- Stablecoin issuance or contraction: UNKNOWN.
- Exchange in/outflows, Coinbase premium, institutional proxies: UNKNOWN.
- No source in the result set addresses real capital entering or leaving.

## Options
- Deribit BTC implied vol, IV vs realized, term structure, skew, put/call demand, strike concentration, Sep 16 expiry pricing, expected move: UNKNOWN, no source returned.
- r3 flagged all options data unknown; nothing changed.

## Calendar
- Sep 11 08:30 ET: Core CPI m/m forecast 0.2%, prior 0.2%; Core CPI y/y forecast 2.4%, prior 2.5%; CPI m/m forecast 0.4%, prior 0.1%; CPI y/y forecast 3.4%, prior 3.4% (calendar feed). ACTUALS: UNKNOWN, fourth consecutive revision without them.
- Sep 11 10:00 ET: Prelim UoM sentiment forecast 51.0, prior 51.0; inflation expectations prior 4.3%. ACTUALS: UNKNOWN.
- Sep 10 08:30 ET: PPI m/m forecast 0.4%, prior 0.0%; Core PPI m/m forecast 0.3%, prior 0.2%; claims forecast 205K, prior 206K (calendar). ACTUALS: UNKNOWN.
- Sep 11 02:00 ET: GBP GDP m/m forecast 0.0%, prior 0.3% (calendar). ACTUAL: UNKNOWN.
- Sep 16 14:00 ET: FOMC decision. Consensus: UNKNOWN, no source.
- Sep 9 21:15 ET: President Trump speaks (calendar, medium impact). Content: UNKNOWN.
- PCE date, NFP, Treasury auctions, large expiries over the next 7 days: UNKNOWN.

## The one story
- No single dominant development surfaced in search. The tape item remains the failed $78,300 reclaim inside a $76,500 to $78,300 range into the Sep 16 FOMC, per r3 and the OBSERVED close. No search source corroborates or contradicts.

## Anomalies
- Funding percentile fell to 36 from r3's 48 while price closed +0.8%: leverage got cheaper into a rising close, not more crowded (OBSERVED).
- Equities and BTC all closed up (S&P +0.9%, Nasdaq +1.0%, BTC +0.8%) yet BTC stayed under its $78,300 pivot: no source explains the crypto-specific lag (OBSERVED).
- WTI fell 2.4% on the day after +9.5% on the week and +20.1% on the month, while the 10y sits +21bp on the week: a one-day disinflation impulse not yet in yields (OBSERVED).
- Cross-exchange OI pages disagree by more than 2x ($16.3B, $18.4B, about $37B) with no timestamps: unreconcilable [CONFLICTING].
- Funding reads across undated pages span +10.95% APR (PerpFinder/Binance) to +6.04% (Quantority/Binance) to 0.0033%/8h on OKX today: not measuring the same window [CONFLICTING].

## What contradicts r3
- r3 funding 0.0046%/48th percentile vs OBSERVED 0.0033%/36th: the "funding ticked up into a falling close" line no longer describes the current state, and the re-crowding warning is weaker, not stronger.
- r3 close $77,175.61 vs OBSERVED $77,199.73: +$24, immaterial.
- r3 OKX OI $2.10B vs OBSERVED $2.09B: unchanged, deleveraging read stands.
- r3 Hyperliquid funding -0.0002%/hr vs OBSERVED -0.00005%/hr: shorts still pay longs, less so; "longs are not paying up" stands.
- r3 equity risk-on read is confirmed by the OBSERVED close (S&P +0.9%, Nasdaq +1.0%).
- r3's Sep 11 CPI gap persists at the fourth revision: ACTUALS still UNKNOWN.

Information for the reader's own decisions, not financial advice.
the positions, each on the others' notes
A · deepseek/deepseek-v4.1-flash · 735 words
ACCEPT

- BTC spot $77,199.73, +0.8% d, -3.1% w, +21.6% m, close 2026-09-11 [OBSERVED, Binance BTCUSDT; held by all three].
- OKX BTC-USDT-SWAP funding 0.0033% per 8h, 7d avg 0.004%, 36th percentile of 90d [OBSERVED, OKX]. Down from r3's 0.0046%/48th; the r3 "funding ticked up into a falling close" line no longer describes the state (held by A, B, C).
- OKX BTC-USDT perp OI $2.09B, -1.9% w/w (r3 $2.10B), unchanged in direction [OBSERVED, OKX].
- Hyperliquid BTC perp funding -0.00005%/hr, OI $2.79B, mark $77,158 [OBSERVED, Hyperliquid]. Note the r3 print was -0.0002%/hr, so shorts pay longs but less.
- Hyperliquid settles funding hourly, not every 8h, so its rate is not comparable to OKX/Bybit 8h rates [hyperliquid.gitbook.io/Hyperliquid-docs/trading/funding, undated docs; perp.wiki/funding-rates].
- 10y 4.97%, +21bp w/w; 2y 4.40%, +21bp w/w, stamped 2026-09-09; EFFR 3.63%, 0bp w/w as of 2026-09-10 [OBSERVED].
- DXY 99.1, 0.0% d, +0.1% w, -0.9% m; gold $4,390, +0.6% d, -2.3% w; WTI $99.99, -2.4% d, +9.5% w, +20.1% m; Brent $104.42, -3.0% d, +9.3% w [OBSERVED, 2026-09-11 closes].
- S&P 7,656.98, +0.9% d, -1.2% w, -1.2% m; Nasdaq 26,333.04, +1.0% d, -0.9% w, -1.0% m [OBSERVED].
- Monthly divergence live: BTC +21.6% vs S&P -1.2% and Nasdaq -1.0% (held by all three).
- Sep 11 close is $24 above r3's $77,175.61, immaterial; $76,500 floor held, $78,300 pivot not reclaimed [DERIVED from OBSERVED; corroborated by A, B, C].
- Deleveraging, not re-crowding: price flat-to-up with OI flat-to-down on both observed venues and funding cheaper [DERIVED; A, B, C agree].
- Sep 11 CPI (core and headline), UoM sentiment and inflation expectations, Sep 10 PPI and claims, ECB decision, GBP GDP: forecasts only, ACTUALS UNKNOWN for a fourth revision (calendar feed; A, B, C).
- FOMC decision Wed 2026-09-16 14:00 ET carried forward; consensus and pricing unverified [B, C].
- Cross-venue OI map, UNDATED: Binance $8.47B, Gate.io $4.51B, Bybit $4.31B, MEXC $3.74B, Hyperliquid $2.80B, OKX $2.11B, Deribit $863M, ~$36-37B across venues [sharpe.ai/futures/bitcoin]. Shape of market only, corroborates OKX $2.09B observed [A, B, C].
- Bybit funding +0.0051%/8h, Hyperliquid +0.0013%/8h, OI $3.74B across three venues [perpsbtc.com], UNDATED; consistent with mild positive long premium [B].

DISPUTE

- PerpFinder $78,308 (12:27 UTC) and perpsbtc.com $78,373.6 as current price: both predate the 23:45 UTC OBSERVED read of $77,199.73 and are superseded. Held by A, B, C; OBSERVED wins.
- PerpFinder Hyperliquid annualized +4.40% against the OBSERVED negative Hyperliquid funding: OBSERVED venue read stands; the +4.40% is a different snapshot/window [perpfinder.com/asset/BTC]. Held by A, B, C — losing claim to OBSERVED.
- MarginPad ($64,259, OI $28.96B, liquidation bands $62,050/$65,650), ARX ($64,459, build Jul 29), loris.tools ($79,791, Sep 6), loris Hyperliquid 24h liquidation $1.29M (Sep 6): stale or price-inconsistent, unusable. Held by A and C; B already flagged them.
- Quantority OI figures ($16.33B headline vs $18.36B total on one page) and Binance/Bybit/Bitget funding APR +6.04/+7.51/+8.87%: internally inconsistent, undated [quantority.com/open-interest/BTC]. Held by C; discarded.
- "Weekly lower-high ceiling at $80,000": carried from r3 with no search source in any of the three sets. It is a DERIVED structural read, not a sourced level, and cannot be stated as observed. Held by all three; ruling: keep only as DERIVED.
- r3's conflict line that funding rose to 0.0046% into a falling close and implied re-crowding: overturned by the OBSERVED 0.0033%/36th. Held by all three as WEAKENED.
- OKX OI $2.10B (r3) vs $2.09B (OBSERVED): not a contradiction, same series two hours apart; no ruling needed beyond the OBSERVED print.

UNKNOWN

- Sep 11 CPI actuals (core/headline, m/m and y/y), UoM actuals, PPI, claims, ECB, GBP GDP.
- CME FedWatch odds, FOMC consensus, dot-plot expectations, latest Fed communication.
- Spot BTC ETF flows, stablecoin issuance, exchange in/outflows, institutional proxies, spot CVD, futures CVD.
- All options data: IV, IV vs realized, term structure, skew, put/call, strike concentration, Sep 16 event vol and expected move.
- Market-wide OI beyond the undated tables, market-wide liquidation dollars for Sep 11, perp basis, perp premium.
- ETH price, BTC dominance, alt breadth, HY OAS, credit spreads.
- Fed balance sheet, reserves, RRP, TGA; Treasury issuance, refunding, auctions, dealer and foreign demand.
- Calendar beyond Sep 16: PCE date, NFP, auctions, large expiries.
- The dominant single driver of the Sep 11 session; whether the equity risk-on close and the BTC lag have a crypto-specific cause. HYPOTHESIS only.
- Whether real capital or only leverage is moving this market: spot-flow gap is the largest missing piece.
B · deepseek/deepseek-v4.1-flash · silent: OpenRouter stopped streaming the position call before the answer finished (120 s cap); by then the model had streamed 5 255 characters of reasoning and 772 of the answer.
C · deepseek/deepseek-v4.1-flash · 623 words
## ACCEPT

- All OBSERVED rows, since the page measured them: BTC $77,199.73 (1d +0.8%, 1w -3.1%, 1m +21.6%); OKX funding 0.0033%/8h, 7d avg 0.004%, 36th pct; OKX OI $2.09B, -1.9% w; Hyperliquid funding -0.00005%/hr, OI $2.79B, mark $77,158; 10y 4.97% (+21bp w); 2y 4.40% (+21bp w, stamped 2026-09-09); EFFR 3.63% (flat); DXY 99.1; gold $4,390; WTI $99.99; Brent $104.42; S&P 7,656.98; Nasdaq 26,333.04. Held by all three.
- Funding fell 0.0046%→0.0033% and the percentile 48→36 from r3 to now, into a green close: cheap and drifting cheaper. Held by all three.
- OKX OI effectively unchanged, $2.10B→$2.09B, still -1.9% w: deleveraging, not re-crowding. Held by A and C, consistent with B.
- Hyperliquid perp funding negative, -0.00005%/hr, longs not paying up; OI flat at $2.79B. Held by all three.
- Close $77,199.73 is $24 above r3's $77,175.61: immaterial, same structure. Held by B and C.
- Failed $78,300 reclaim, $76,500 floor intact, range $76,500 to $78,300. Carried from r3; no source contradicts. Held by all three.
- PerpFinder Binance spot $78,308 at 12:27 UTC, above the close, consistent with an intraday fade. Held by A, B, C.
- 2y stamp lags the 10y by two days, so the curve read is a mixed-date comparison. Held by A and C (B did not flag).
- Stale prints to discard: MarginPad $64,259; ARX $64,459 (build Jul 29 2026); Loris $79,791 (updated 2026-09-06). Held by all three.
- Hyperliquid settles funding hourly, so its rate is not comparable to 8h venues. Held by A and B ([perp.wiki](https://perp.wiki/funding-rates); [hyperliquid.gitbook.io](https://hyperliquid.gitbook.io/Hyperliquid-docs/trading/funding)).
- Cross-venue OI line items as reported identically by A, B, C ([sharpe.ai](https://www.sharpe.ai/futures/bitcoin), undated): Binance $8.47B (22.9%), Gate.io $4.51B, Bybit $4.31B, MEXC $3.74B, Hyperliquid $2.80B (7.6%), Bitget $2.62B, HTX $2.32B, OKX $2.11B, Deribit $863M. Shape-of-market only.
- All calendar forecast/prior figures as printed; all actuals (CPI, PPI, UoM, ECB, GBP GDP) UNKNOWN. Held by all three.
- Sep 16, 14:00 ET FOMC decision carried from prior state; consensus UNKNOWN. Held by all three.

## DISPUTE

- PerpFinder Hyperliquid annualized funding +4.40% [perpfinder.com](https://perpfinder.com/asset/BTC) vs OBSERVED -0.00005%/hr (~-0.44% annualized): OBSERVED wins; different snapshot or oracle convention. Held by A, B, C.
- PerpFinder Binance +10.95% annualized vs OBSERVED OKX 0.0033%/8h: not the same window; the annualized figure cannot characterize current market funding. Held by A, B, C.
- perpsbtc.com $78,373.6 and its funding (Bybit +0.0051%/8h, Hyperliquid +0.0013%/8h) [perpsbtc.com](https://perpsbtc.com/): undated and price inconsistent with the close; not usable as current. Held by A, B, C.
- Quantority.com $16.33B headline vs $18.36B total, funding APR Binance +6.04% [quantority.com](https://quantority.com/open-interest/BTC): internally inconsistent, no timestamp; low confidence. Held by C only; A and B did not find it, which is a gap, not a contradiction.
- sharpe.ai OI sum: A computed ~$36.4B, C ~$37.0B. Accept the line items, reject both sums; no timestamp.
- [perp.wiki](https://perp.wiki/funding-rates) Hyperliquid -0.0017%/hr vs OBSERVED -0.00005%/hr: OBSERVED is the measured read. Held by B.

## UNKNOWN

- Spot BTC ETF flows, stablecoin issuance, exchange in/outflows, spot volume trend: all three, no source. Real-capital test cannot be run.
- All options data: IV, IV vs realized, term structure, skew, put/call, strike concentration, Sep 16 event vol, expected move. All three.
- CME FedWatch odds, FOMC consensus, dot-plot expectations. All three.
- Fed balance sheet, reserves, RRP, TGA; Treasury issuance, refunding, auctions, dealer and foreign demand. All three.
- HY OAS, credit spreads, financial conditions, collateral, dollar liquidity. All three.
- ETH price, BTC dominance, alt breadth; crypto equities. All three.
- Market-wide OI, market-wide liquidation dollars, basis, perp premium, futures and spot CVD, realized vol. All three.
- Actual Sep 11 CPI, PPI, UoM, ECB, GBP GDP prints; PCE date; NFP; large expiries next 7 days. All three.
- The one story: no dominant driver surfaced in any set. All three.
how to read this

One fixed method writes every brief: three researchers read the web side by side, each takes a position on the others' notes, the first consolidates one agreed research document, then one write call diffs it against the previous brief, updates the thesis and delivers in this order. The thesis is a call, not a fact; the sections after it are the evidence for it, the evidence against it, and what would prove it wrong. Bracketed marks say how each claim is known: OBSERVED was measured, DERIVED was computed, INFERRED is an interpretation, HYPOTHESIS is unconfirmed, UNKNOWN could not be found. The tape at the top is what the page itself read seconds before the model was called. Anyone can contribute the next revision by connecting an OpenRouter key and pressing run; the models and the time are printed above. When a day has more than one revision, an area that changed says so when you hover it; click it to compare the previous version with this one.

previous briefs
  1. Sat Sep 12, 20261 rev · weakenedBitcoin keeps the $76,500 to $78,300 range it has held all week and nothing resolves before the Sep 16 Fed, but the leverage deck is no longer cheap: OKX funding re-crowded overnight from 0.0033% to 0.0062% per 8h (64th percentile of 90 days) with price flat, and Hyperliquid flipped to longs paying shorts. The range read stands; the cheap-funding read does not.
  2. Fri Sep 11, 20264 revs · unchangedBitcoin stays inside the $76,500 to $78,300 range it failed to break on Sep 11, with leverage draining (OKX funding 0.0033% per 8h, the 36th percentile of 90 days; OKX open interest $2.09B, -1.9% on the week) and nothing resolving before the Sep 16 Fed decision. Moderate confidence.
  3. Thu Sep 10, 20266 revs · weakenedBitcoin is still boxed in a $76,500 to $78,300 range into Friday's 8:30 AM ET CPI print, and the range, not the trend, is what this tape is trading. The 9:04 PM ET view that leverage had de-crowded into the event fails on this read: funding ticked up to 0.01% and OKX open interest rebuilt, so longs are paying up again right into a binary number. The prior view is otherwise standing: no close below $76,500, accumulation planned at $74,000 and $72,000 if the floor goes.
  4. Wed Sep 9, 20261 rev · baselineBitcoin is in a cooling-off pullback after a +23% monthly rally, holding the $78K area ahead of the week's two inflation tests (PPI Thursday, CPI Friday). The base case is a range between roughly $76.5K and $80K that resolves directionally on the CPI print: soft CPI re-ignites the risk-on bid and sends BTC back toward the highs, while hot CPI accelerates the correction into the mid-$70Ks. The dollar weakening, gold rising alongside BTC, and oil pushing above $100 combine into a cross-current where Bitcoin behaves more like a real-asset store of value than a pure risk proxy.