PakupaiAI-generated Bitcoin research, with sources
Sat Sep 12·Sunday, September 13, 2026·filed ·record-only
Sep 13, 2026, 12:48 PM ET Edition history

The Funding Touch Fades, the Bond Market Does Not

The big picture

New spot additions were stopped at Saturday's read because OKX funding had printed exactly 0.01% per eight hours, the 89th percentile, a fourth consecutive rise at flat price. The next read was set to decide whether that touch held or faded. It faded.[1]

Bitcoin trades $77,314, flat on the day, down 3.8% on the week and up 22.6% on the month, still inside the $76,500 to $78,300 band that has held since Sep 11, an eighth consecutive read. What changed is that the leverage pressure inside the range relaxed while the macro pressure outside it did not.[1]

A Bond Market Pressing at 5%, and Oil Nobody Can Explain

The most consequential development of the past week is not in crypto. The US 10-year Treasury yield touched 4.9708% in Asian trading on Sep 11, a three-year high, and Reuters reported that 5% is the level traders describe as disruptive for equities. Two measurements of that yield sit on the table and should not be blended: the Yahoo quote of 4.97% is dated Sep 11, while the Federal Reserve's official par-yield series shows 4.83% for Sep 10. Fourteen basis points apart, different instruments and days, same direction.[1][2][3]

What matters for inflation is that the move is not purely nominal. The 10-year real yield rose to 2.46% on Sep 10 from 2.42% on Sep 3, a fourth consecutive weekly increase. The Treasury also widened its buyback of 10- to 20-year paper to as much as $6bn from a $2bn maximum, and Reuters described investors as unconsoled; a $39bn 10-year auction drew the strongest demand since 2019 without stopping the rise.[1][2][3]

Oil adds to it. WTI closed Friday at $100.05 and Brent at $104.61, up 9.6% and 9.5% on the week and roughly 20% on the month, both still in triple digits. Reuters frames the bond selloff as oil-driven inflation risk. That is a proposed transmission, not a demonstrated cause: no dated supply, inventory, OPEC+ or shipping event was found. The rate move and the oil move are dated observations; the chain linking them is a hypothesis.[1][2][3]

Equities Shrug While the Policy Path Stays Unreadable

Equities closed Friday risk-on, the S&P 500 up 0.9% and the Nasdaq up 1.0% on the day, each down about 1% on the week. The dollar index is flat at 99.1 and gold is down 1.8% on the week, so there is no haven bid and no dollar confirmation of a fresh risk-off impulse. Rates at multi-year highs and oil at $100 with equities still buying is a peculiar mix.[1][2][3][4]

The policy picture is genuinely unsettled. The effective fed funds rate has been 3.63% for the week ended Sep 10, unchanged, while the calendar lists the previous policy rate as 3.75%, roughly twelve basis points higher, and Reuters reported on Sep 9 that futures priced about 60% odds of a hike. Those cannot all describe one stance, and no primary FedWatch or dot-plot source was returned. The expected path is unknown until Wednesday, which argues for keeping decision risk small.[1][2][3]

What it means for Bitcoin

OKX prints 0.008% per eight hours now, the 78th percentile of 90 days, and Hyperliquid 0.0004% per hour against Saturday's 0.00055%, so the touch was released rather than confirmed. The seven-day average ticked up to 0.0055% from 0.005%: the drift is real but far from the 0.05% mean this letter treats as genuine crowding. The stop on new additions nonetheless stands, on macro rather than positioning grounds.[1][2]

Open interest is where measurements genuinely disagree. CoinGecko-covered BTC derivatives show $64.34 billion against $63.67 billion on Saturday, about a 1% move with no seven-day trend and no confirmation that both reads cover identical venues. One researcher reads that as re-leverage, another as a qualified bounce. A single-provider 1% move cannot establish a reversed direction, so the prior claim stays unresolved rather than confirmed or contradicted.[1]

Price structure is unchanged and uninformative. The $76,500 floor and the $78,300 failed-reclaim level have contained every print since Sep 11, and the $80,000 weekly lower-high ceiling is unbroken. None of this can be tested against real capital: spot ETF flows, stablecoin issuance, spot volume, cumulative volume delta, basis, options volatility and liquidation dollars are all unobserved, and an attempt to widen the funding read returned stale aggregator values on Sep 11.[1][2]

Outlook: weakened. Confidence: moderate.

Next meaningful test

The next meaningful test is the FOMC decision on Wednesday Sep 16 at 2:00 PM ET, with economic projections then and a press conference at 2:30 PM. The calendar carries no consensus and FedWatch odds were not retrieved, so hawkish and dovish cannot be sized in advance. The bond market, not the crypto book, is the transmission: if the 10-year holds above 5.00%, the plausible path is a test of $76,500 and then $74,000; if it retreats, the path of least resistance is a retest of $78,300 and then $80,000. Neither is a probability this letter can assign.[1][2]

The smaller test arrives first: whether OKX funding stays under 0.01% per eight hours through Monday and Tuesday. Holding below it leaves the leverage deck unremarkable and the decision entirely with the Fed. Re-crossing it would restore the crowded-longs condition with less room for a hawkish surprise. The Bank of England on Thursday and the Bank of Japan on Thursday and Friday extend the same question, how far the global repricing goes, but Wednesday answers it first.[1][2]

The cost of leverage

Snapshot taken Sep 13, 2026, 12:42 PM ET.

Funding, 100 eight-hour prints from Aug 11 to Sep 13, between -0.0027% and 0.0100%; the top decile of the window starts at 0.0100%; the last print 0.0068%0.0100%-0.0027%00.0068%Aug 11Aug 18Aug 25Sep 1Sep 8Sep 13
OKX BTC-USDT perpetual funding, 8h intervals, percent per interval

The market at a glance

Measurements from Sep 13, 2026, 12:42 PM ET.

MarketLevelDaily changeAs of
Bitcoin$77,314.010%2026-09-13
Ether$2,503.58-0.9%2026-09-13
S&P 5007,656.98+0.9%2026-09-11
Nasdaq26,333.04+1%2026-09-11
Dollar index99.10%2026-09-11
Gold$4,408.90%2026-09-11
Brent$104.61-2.8%2026-09-11
WTI$100.05-2.4%2026-09-11

Key evidence and scenarios

OBSERVED Sep 13 12:42 PM ET: OKX BTC-USDT perp funding 0.008% per 8h, 7-day average 0.0055%, 78th percentile of 90 days, down from Saturday's exact 0.01% and 89th percentile; Hyperliquid 0.0004% per hour, down from 0.00055%. Both watched venues faded, so the Sep 12 touch did not persist.[1][2]

OBSERVED Sep 13: Binance BTC $77,314.01 (1d 0%, 1w -3.8%, 1m +22.6%); ETH $2,503.58 (1d -0.9%, 1m +33%). CoinGecko-covered BTC derivatives open interest $64.34B against $63.67B on Sep 12, about 1%, 7-day trend n/a; coverage limits prevent reading it as a direction change.[1][2][3]

Official Sep 10 par yields: 2-year 4.43%, 10-year 4.83%, 30-year 5.28%; 10-year TIPS real yield 2.46% versus 2.42% on Sep 3, a fourth weekly rise. EFFR 3.63%, unchanged, per H.15 and the NY Fed. The Yahoo 10-year quote of 4.97% is a different instrument on Sep 11 and is not averaged with the par close.[1][2][3]

Reuters Sep 9-11: 10-year touched a three-year high 4.9708% in Asia with 5% flagged as disruptive for equities; Treasury buyback of 10- to 20-year paper widened to as much as $6bn from $2bn; a $39bn 10-year auction drew the strongest demand since 2019; fed funds futures reported near 60% odds of a hike, secondary and unreconciled with a 3.63% EFFR and a blank calendar consensus.[1][2][3]

OBSERVED Sep 11 close: WTI $100.05 (+9.6% w/w), Brent $104.61 (+9.5% w/w); S&P 500 +0.9% and Nasdaq +1.0% on the day but about -1% on the week; DXY 99.1 flat; gold -1.8% on the week. No dated supply, OPEC+, sanctions or shipping driver for oil was returned, so the causal link between oil and the bond selloff stays a hypothesis.[1][2][3][4]

The 10-year holds above 5.00% through Wednesday's FOMC statement and press conference
The rates-and-oil headwind becomes binding; the plausible path is a test of the $76,500 floor, then the $74,000 rung. Holding core spot through it is defensible, new additions are not.

The projections and press conference read as less restrictive than the recent repricing implies and the 10-year retreats from 5%
The path of least resistance becomes a retest of $78,300 and then the $80,000 weekly ceiling; additions become defensible only in spot and only if funding stayed under 0.01% per 8h.

OKX funding re-crosses 0.01% per 8h on Monday or Tuesday with price still inside the range
The crowded-longs condition returns into a binary event with less capacity to absorb a hawkish surprise, restoring Saturday's stand-down on additions rather than merely holding it.

What remains uncertain

  • FOMC Sep 16 consensus, dot-plot expectations and FedWatch odds: unobserved, and the calendar consensus field is blank, so hawkish and dovish cannot be sized.
  • Whether the 3.75% calendar policy rate with a 3.63% EFFR means a cut was already delivered or the calendar row is stale: unresolved, so no target-rate assertion is supportable.
  • Cause of the oil surge: no dated supply, inventory, OPEC+, sanctions or shipping source returned; the oil-to-rates chain remains a proposed transmission.
  • Spot BTC ETF net flows, stablecoin supply, spot volume, CVD, basis and perp premium, all options data and liquidation dollars: unobserved, so spot-led versus leverage-led stays open.
  • Market-wide open interest 7-day trend, and whether the two CoinGecko reads cover identical venues: unknown, which is what limits the open-interest grading.

How the outlook has evolved

continuing

Bitcoin keeps the $76,500 to $78,300 range into the Sep 16 FOMC.

Eighth consecutive read inside: $77,314, about $800 above the floor. Unresolved; the Fed remains the resolving event.

Original criteria and dates

First recorded 2026-09-12. Horizon: 2026-09-16T14:00:00-04:00.

Daily closes remain inside $76,500 to $78,300 through the Sep 16 FOMC decision; a close above $78,300 or below $76,500 resolves it early.

resolved

Stop all spot adds if OKX funding crosses 0.01% per 8h.

The Sep 12 touch was exact, not a clean cross, and Sunday's 0.008% printed back below the threshold, so the follow-through fork resolves on the fade side and the trigger resets.

Original criteria and dates

First recorded 2026-09-12. Horizon: 2026-09-13T12:00:00-04:00.

A clean cross above 0.01% per 8h on the next funding read confirms a persistent crowding signal; a print back below it resets the trigger.

resolved

Both watched venues lean long-paying; treat the weekend bid as leverage, not demand.

Formally still true, both venues print positive funding, but at 0.008% per 8h and 0.0004% per hour against Saturday's 0.01% and 0.00055%: resolved as a fading lean, weaker than the claim implied.

Original criteria and dates

First recorded 2026-09-12.

Both OKX and Hyperliquid print positive funding at the read.

continuing

No re-leverage; open interest flattened rather than reversing.

Graded unresolved, not contradicted: $64.34B against $63.67B is about 1% on one provider with a 7-day trend of n/a, which cannot establish a reversed direction.

Original criteria and dates

First recorded 2026-09-12.

Market-wide open interest does not reverse direction from its flat-to-lower trend.

continuing

The 10-year near 4.97% with WTI above $100 keeps a rates-and-oil headwind under an asset up more than 20% on the month; the cause of the oil surge is unverified.

The 10-year touched 4.9708% intraday on Sep 11 but has no close at or above 5.00%; oil remains triple digits. Criterion not met; the real 10-year yield also rose, to 2.46% from 2.42%.

Original criteria and dates

First recorded 2026-09-11. Horizon: 2026-09-16T14:00:00-04:00.

The 10-year at or above 5.00% on a close, or WTI holding above $100 into the FOMC, demonstrates a binding macro headwind.

Every edition and failed attempt
Technical details

Snapshot taken Sep 13, 2026, 12:42 PM ET. These measurements belong to this edition.

Bitcoin$77,3140% 1d
Funding 8h0.008%p78 of 90d
Open interest$64.34B 
US 10Y4.97%▲ +21 bp 1w
Fed funds3.63%0 bp 1w
Dollar (DXY)99.100% 1d
Brent$104.61▼ -2.8% 1d
Gold$4,4090% 1d
S&P 5007,657▲ +0.9% 1d
Next event3d 0hFederal Funds Rate
Bitcoin, 48 four-hour candles from Sep 5 UTC to Sep 13 UTC, between $76,047 and $80,560; 0 levels drawn as dashed lines and the price now, $77,314, as the solid line; dates in UTC76,00077,00078,00079,00080,00081,000NOW $77,314Sep 6 UTCSep 8 UTCSep 10 UTCSep 12 UTC
Additional measurements from this edition's recorded snapshot.
Sources

30 returned sources; citation presence does not establish that every claim is verified.

Snapshot sources

Successful readings saved with this edition; separate from researcher retrieval coverage.

How this edition was produced

3 × deepseek/deepseek-v4.1-flash

$0.10 · 5 m 44 s · run on a connected key

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

Editorial review: flagged. Model review is not independent verification.

903 narrative words; 833 supporting words

Geopolitics and energy: 7 assigned-source citations

Economy and policy: 4 assigned-source citations

Crypto flows and positioning: 3 assigned-source citations

Research notes
AGREED
- OBSERVED, Sep 13 12:42 PM ET: OKX BTC-USDT perp funding 0.008% per 8h, 7-day avg 0.0055%, 78th percentile of 90 days, and Hyperliquid 0.0004% per hour; both below the Sep 12 reads of 0.01% and 0.00055%, so the Sep 12 stand-down touch did not persist on either watched venue [OBSERVED snapshot, OKX perp; Hyperliquid; also ACCEPT in all three positions].
- OBSERVED, Sep 13 12:42 PM ET: Binance BTC $77,314.01 (1d 0%, 1w -3.8%, 1m +22.6%); ETH $2,503.58 (1w -0.4%, 1m +33%) [OBSERVED snapshot, Binance spot; Researcher A, C].
- OBSERVED, Sep 13 12:42 PM ET: CoinGecko-covered BTC derivatives open interest $64.34B, 7-day trend n/a [OBSERVED snapshot, CoinGecko; Researcher A, C].
- Sep 10, official H.15 daily series: effective fed funds 3.63%, unchanged week over week; NY Fed reference rate confirms [federalreserve.gov](https://www.federalreserve.gov/releases/h15/); [newyorkfed.org](https://www.newyorkfed.org/markets/reference-rates/effr); Researcher B, C.
- Sep 10, official par yields: 2-year 4.43%, 10-year 4.83%, 30-year 5.28%; 10-year TIPS real 2.46% versus 2.42% on Sep 3, a fourth weekly rise, so the move is not purely nominal [federalreserve.gov](https://federalreserve.gov/releases/H15/default.htm); [home.treasury.gov](https://home.treasury.gov/resource-center/data-chart-center/interest-rates/TextView?field_tdr_date_value_month=202608&type=daily_treasury_yield_curve); Researcher B, C.
- Sep 9, Reuters: 10-year yield highest since 2023 on the buyback announcement; strong $39bn 10-year sale at 4.834%, described as strongest demand since 2019 [reuters.com](https://www.reuters.com/business/10-year-yields-highest-since-2023-2026-09-09/); Researcher A, B.
- Sep 10, Reuters: Treasury buyback expanded to as much as $6bn of 10- to 20-year paper from a $2bn maximum; investors described as unconsoled [reuters.com](https://www.reuters.com/legal/transactional/edgy-bond-investors-unconsoled-by-bessents-big-buyback-2026-09-10/); Researcher A, B.
- Sep 11, Reuters: 10-year pushed toward 5%, touching a three-year high 4.9708% in Asia, with 5% flagged as the level that could disrupt stocks [reuters.com](https://www.reuters.com/world/china/global-markets-view-europe-2026-09-11/); Researcher A, B, C.
- Sep 11, Reuters wrap: 10-year off about 1bp at 4.93% after the inflation report, framed as relief that a hot print had not forced a faster Fed reaction [reuters.com](https://www.reuters.com/world/europe/global-bond-selloff-pushes-10-year-us-yield-toward-5-oil-rate-hike-fears-2026-09-11/); Researcher A, B.
- Sep 10, Reuters: hawkish ECB expectations and rising energy costs knocked euro-zone markets; euro-zone yields hit multi-year highs [reuters.com](https://www.reuters.com/business/bond-yields-hit-multi-year-highs-traders-brace-new-ecb-rate-hike-cycle-2026-09-10/); [reuters.com](https://www.reuters.com/world/china/global-markets-wrapup-1-2026-09-10/); Researcher A, B.
- Sep 11 close, OBSERVED: WTI $100.05, Brent $104.61, still triple digits; no dated supply, OPEC+, sanctions or shipping driver returned by any seat [OBSERVED snapshot; Researcher A, B, C].
- Sep 11 close, OBSERVED: S&P 500 +0.9%, Nasdaq +1.0% on the day but -1.2% and -0.9% on the week, so no broad risk-off tape forced the crypto drift [OBSERVED snapshot; Researcher A, C].
- Sep 13, supplied calendar: FOMC Sep 16 2:00 PM ET with projections and 2:30 PM press conference; BoE Sep 17 7:00 AM ET; BoJ Sep 17-18; UK CPI Sep 16; Canada CPI Sep 14; US retail sales Sep 16; claims and Philly Fed Sep 17 [forexfactory.com](https://www.forexfactory.com/calendar); Researcher B.
- Sep 11, Glassnode all-venue funding mean printed 0% and CoinGlass live funding field returned "undefined%": both stale or non-populating, unusable as fresh market-wide prints [studio.glassnode.com](https://studio.glassnode.com/charts/derivatives.FuturesFundingRatePerpetualAll?a=BTC); [coinglass.com](https://www.coinglass.com/FundingRate/BTC); Researcher C.

RESOLVED
- The Yahoo 10-year 4.97% (Sep 11 quote) and the Treasury official 10-year 4.83% (Sep 10 par close) are different instruments on different days, 14bp apart; the source comparison settles that they must not be averaged or used interchangeably, though the direction agrees. Researcher A and B both flagged this and neither value is retired [Yahoo ^TNX OBSERVED; federalreserve.gov H.15].
- The oil-to-rates causal chain: Reuters' "oil-driven inflation risk" framing is secondary commentary about a proposed transmission, and no dated supply, inventory, OPEC+, sanctions or shipping event was returned by Assignment A. The rate move and the oil move stand as dated observations; the causal link does not. This is settled against treating the chain as established cause [reuters.com, Sep 11 URLs above; Researcher A, B, C].
- The prior edition's fork "the next funding read decides hold or fade" is answered on the fade side: both watched venues printed below their Sep 12 levels (0.008% and 0.0004%/hour against 0.01% and 0.00055%). The touch did not hold; the Sep 12 trigger is not confirmed as a persistent crowding signal, though the 7-day mean ticked up to 0.0055% [OBSERVED snapshot; Researcher B, C].
- The dispute over whether market-wide OI "reversed" from $63.67B to $64.34B: the two values are single-provider CoinGecko figures about $0.27B-$0.67B apart, roughly 1%, with a 7-day trend still n/a. Source comparison settles the magnitude as within provider coverage limits; neither a reversed direction nor a strictly flat deck can be asserted from this pair. Researcher B's "qualified bounce, not a reversed direction" is the applicable reading, not Researcher C's "direction change" [OBSERVED snapshot, CoinGecko; Researcher B against C].
- Reuters' "about 60% odds of a hike" from fed funds futures (Sep 9) is secondary reporting of implied odds, and it sits awkwardly against an EFFR of 3.63%, roughly 12bp below the calendar's 3.75% prior policy rate, with a blank consensus field. It cannot be treated as established expectations without a primary FedWatch or dot-plot source [reuters.com Sep 9; federalreserve.gov H.15; Researcher B, C].
- The ECB items are background framing only: no ECB decision date or outcome was returned, so they are not a dated policy event [reuters.com Sep 10; Researcher A, B].

UNRESOLVED
- Cause of the oil surge (WTI +9.6% on the week, +20.2% on the month; Brent +9.5%): no dated supply, freight, inventory, refinery, sanctions or Hormuz/Red Sea transit source returned; UNKNOWN across all three seats [OBSERVED snapshot; Researcher A, B, C].
- FOMC Sep 16 consensus, dot-plot expectations and CME FedWatch odds: no source returned; the calendar consensus field is blank, so hawkish and dovish outcomes cannot be sized [supplied calendar; Researcher B, C].
- Whether the standing policy rate is a 3.75% target with a 3.63% EFFR (a cut already delivered) or the 3.75% calendar row is itself stale: unresolved by this pass; no target-rate assertion is supportable [federalreserve.gov H.15; supplied calendar; Researcher B, C].
- Spot BTC ETF net flows, stablecoin supply, spot volume, spot and futures CVD, basis/perp premium: all UNKNOWN; spot-led versus leverage-led stays open [Researcher A, B, C].
- All options data, including Sep 16 event vol, IV, 25-delta skew, term structure and max pain: UNKNOWN; no options structure may be named [Researcher C].
- Liquidation dollars, clusters and long/short split: UNKNOWN [Researcher C].
- Market-wide OI 7-day trend, and whether the Sep 13 CoinGecko figure covers the identical venues as Sep 12: UNKNOWN; this is what limits the OI grading above [Researcher B, C].
- Grading of the prior "no re-leverage; open interest flattened rather than reversing" claim remains unsettled: Researcher C calls it contradicted, Researcher B calls the move a qualified bounce, and no clean criterion covers a single-provider ~1% move [Researcher B against C].
- Whether the standing $76,500 to $78,300 range claim resolves before its Sep 16 FOMC horizon remains open; no premature grading [Researcher B].
- US CPI, PPI, payrolls, JOLTS and UoM actuals, and prior-week ECB, UoM and GBP GDP outcomes: no official release returned, an eighth consecutive pass without an actual; the supplied calendar carries no US CPI row inside the seven-day window [Researcher B, C].
- Fed balance sheet, reserves, RRP, TGA and Treasury auction demand: UNKNOWN; the plumbing cannot be checked [Researcher B].
- Cause of the divergence between the Yahoo 10-year 4.97% quote and the official 4.83% par close: the direction agrees, the 14bp gap's composition (timing, instrument, methodology) is unresolved [Researcher B, C].

Contribute the next chapter

Choose one researcher or three in Run. New work is checked before it replaces the standing letter; unsuccessful attempts remain available in the archive.

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