Bitcoin's Leverage Rent Cools Into the Fed as Oil Jumps Again
The big picture
Bitcoin closed the weekend at $76,738.8, down 0.7% on the day and 4.5% on the week but still up 21.7% on the month, holding inside the $76,500 to $78,300 band it has occupied for an eighth consecutive read. The more interesting change is the cost of leveraged long positions, which cooled rather than rose at this measurement.[1]
The week's resolving event sits three days out: the Federal Reserve's rate decision on Wednesday, with the calendar showing a 4.00% forecast against a 3.75% previous, a gap that implies a quarter-point increase but is not confirmed by any Fed text. Meanwhile oil pushed further into triple digits on Sunday evening and nobody has yet produced a dated reason.[1][2]
Funding Cools on the Primary Venue While a Second One Stays Long-Paying
Saturday's read left the short-term call weakened. OKX, the venue the letter has tracked for funding, printed exactly 0.01% per 8 hours, the 89th percentile of its last 90 days, and that matched a stand-down trigger set earlier that day: stop adding to spot positions. That state no longer describes the current measurement. At the Sunday evening read the same venue prints 0.0057% per 8 hours, a seven-day average of 0.0055%, and the 58th percentile of 90 days. The rent long positions pay has roughly halved from the trigger level.[1]
What that halving means is genuinely unresolved. It is one venue's series, and a one-venue series cannot by itself establish that market-wide crowding has faded. The stand-down criterion as originally written asked whether funding strictly crosses above 0.01% per 8 hours; Saturday's print was an exact touch rather than a clean cross, so the original criterion was never satisfied and is not satisfied now. We are treating the touch as a touched trigger rather than a confirmed one, and we are not rewriting the original threshold after the fact.
The second watched venue complicates the picture. Hyperliquid's BTC perpetual pays 0.00125% per hour, roughly 0.01% per 8 hours, with $2.69 billion of open interest at a $76,718 mark. So one venue's funding has fallen to the 58th percentile while another sits near the level that counted as elevated on Saturday. Whether that is a genuine easing of leverage demand or simply a difference between two venues is a question the available data does not answer.[1]
Oil Extends While Bonds, the Dollar and Gold Say Little
The macro backdrop is the same tension as last week, only louder. WTI crude printed $102.35 on Sunday evening, up 2.3% on the day and 11.9% on the week, and Brent $107.53, up 2.8% and 11.7%. These are futures quotes from a Sunday session, not settled closes, so they cannot be ranked like for like against Friday settlements, but the direction is unambiguous. Nine retrieved news pages carried no sanctions, tanker, OPEC+, inventory or shipping headline that would explain the bid.[1][2][3]
Bond yields remain the other pressure point. The 10-year sits at 4.97%, up 21 basis points on the week as of Friday's reading, and the 2-year at 4.40%, up 21 basis points as of Sep 9. Those two weekly changes are measured to different endpoints, so they are not one synchronized repricing and we will not present them as such. The Fed's own H.15 releases show the effective funds rate flat at 3.63% across Sep 1 to 7, with five-, ten- and thirty-year breakevens at 2.20, 2.46 and 2.98 by Sep 9, rising alongside nominal yields. Market rates have moved; policy has not.[1][2]
Equities closed Friday risk-on, the S&P 500 at 7,656.98 and the Nasdaq Composite at 26,333.04, each up about 1% on the day but down about 1% across both the week and the month, so the gains read as a bounce inside a soft stretch rather than a fresh trend. The dollar index is flat at 99.14 on the day and week, and gold is $4,382.6, down 1.1% on the week. Neither shows a haven bid. The cross-asset picture is a rates-and-oil headwind with no risk-off panic underneath it.[1][2][3][4]
The employment half of the Fed's mandate remains unchecked: no payroll, unemployment, wage or claims actuals beyond the calendar entries were returned. Wednesday's decision therefore arrives with the market unable to size how hawkish or dovish the outcome really is. The calendar's 4.00% forecast against a 3.75% prior implies a hike, but no statement, target-range text or minutes were retrieved to verify that prior, so direction stays calendar-derived, not confirmed.[1]
What it means for Bitcoin
Price is exactly where the band says it should be, about $240 above the $76,500 floor. Market-wide open interest across CoinGecko-covered BTC contracts is $63.79 billion, about $0.12 billion above Saturday's figure with no seven-day trend available, and OKX's own perpetual open interest is $2.13 billion, down 2.5% on the week. A $0.12 billion move is far too small to distinguish positions being trimmed from positions being added, so Saturday's claim that shrinking open interest meant trimming rather than building is not confirmed. We are withdrawing the directional reading and leaving the mechanism open.[1][2]
Ether is the relative winner again, up 31.6% on the month against Bitcoin's 21.7% and down only 1.5% on the week against Bitcoin's 4.5%. No flow, ETF or protocol data was returned that would attribute that outperformance, so it stays an observation rather than a story.[1]
The honest limit on all of this is coverage. Spot ETF flows, stablecoin issuance, exchange flows, spot volume and order-flow measures, basis and perpetual premium, and every options series including implied volatility and skew came back empty for a fourth straight day. Liquidation dollars are also unknown, so nothing here separates a voluntary exit from a forced one. The leverage read rests on two named venues and cannot be tested against real capital moving in or out.[1][2]
Next meaningful test
The next meaningful test is the FOMC statement at 2:00 PM ET on Wednesday Sep 16, with economic projections and a 2:30 PM press conference. Should the committee raise rates and the 10-year push through 5.00%, the range floor at $76,500 becomes the line to watch, with $74,000 the first rung below it. Should the outcome be softer, $78,300 is the failed-reclaim level that would have to be closed above to argue the band is resolving upward. Between those, the week also carries Canadian CPI Monday, UK CPI and US retail sales Wednesday, the Bank of England Thursday and the Bank of Japan late Thursday, and their forecasts are calendar entries with no actuals yet.[1]
The single most useful thing to observe before Wednesday is whether OKX funding stays below 0.01% per 8h and whether Hyperliquid's rate holds. If both ease, the leverage-cost concern that prompted Saturday's stand-down is fading and the range read is on firmer ground. If OKX climbs back toward the trigger while the second venue stays long-paying, the crowding question returns with the Fed still ahead, which is the more uncomfortable version of the same week.[1]
Bitcoin in the market
The market at a glance
| Market | Level | Daily change | As of |
|---|---|---|---|
| Bitcoin | $76,738.8 | -0.7% | 2026-09-13 |
| Ether | $2,476.63 | -2% | 2026-09-13 |
| S&P 500 | 7,656.98 | +0.9% | 2026-09-11 |
| Nasdaq | 26,333.04 | +1% | 2026-09-11 |
| Dollar index | 99.14 | 0% | 2026-09-13 |
| Gold | $4,382.6 | -0.6% | 2026-09-13 |
| Brent | $107.53 | +2.8% | 2026-09-13 |
| WTI | $102.35 | +2.3% | 2026-09-13 |
Key evidence and scenarios
OBSERVED Sep 13 23:12 UTC: BTC $76,738.8 (1d -0.7%, 1w -4.5%, 1m +21.7%); ETH $2,476.63 (1w -1.5%, 1m +31.6%), so ETH strengthened against BTC on both windows.[1][2]
OBSERVED: OKX BTC-USDT funding 0.0057% per 8h, 7-day average 0.0055%, 58th percentile of 90 days, one venue only; Hyperliquid 0.00125% per hour, about 0.01% per 8h, OI $2.69B, mark $76,718.[1][2]
OBSERVED: CoinGecko-covered BTC derivatives OI $63.79B, no 7-day trend; OKX BTC-USDT perp OI $2.13B, 7d -2.5%. A $0.12B move cannot distinguish trimming from building.[1][2]
OFFICIAL: effective fed funds flat at 3.63% across Sep 1-7 H.15 releases; discount-window primary credit 3.75%, bank prime 6.75%; 5/10/30-year breakevens 2.20/2.46/2.98 by Sep 9, rising alongside nominals. Policy has not moved while market rates have.[1]
OBSERVED: oil is a Sunday-evening futures print, WTI $102.35 (+2.3% day, +11.9% week), Brent $107.53 (+2.8%, +11.7%), not a settled close; nine retrieved pages returned no sanctions, tanker, OPEC+, inventory or shipping headline explaining the bid.[1][2][3]
CALENDAR: Sep 16 FOMC statement 14:00 ET with projections and 14:30 press conference, forecast 4.00% versus previous 3.75%; GBP Bank Rate Sep 17 07:00 ET forecast 3.75% with votes 3-0-6; BOJ Sep 17 22:30 ET forecast below 1.25% versus below 1.00%. All due, no actuals.[1]
The FOMC raises rates and the 10-year pushes through 5.00%
The range floor at $76,500 comes under test; a daily close below it would resume defensive positioning toward the $74,000 rung, with the $72,000 level beyond it. Conditional, not prescriptive.
The FOMC outcome is softer and yields ease
$78,300 becomes the level to close above to argue the band is resolving upward; without a funding reset the leverage-cost concern remains on the table.
OKX funding stays below 0.01% per 8h and Hyperliquid also eases before Wednesday
The leverage-cost concern that prompted Saturday's stand-down fades, and the range read rests on firmer ground, though no spot flow data exists to confirm real demand.
OKX funding returns to the 0.01% trigger while Hyperliquid stays long-paying
The crowding question returns with the Fed three days out, the least comfortable version of the week, and the case for adding exposure stays off regardless of the range holding.
What remains uncertain
- Whether OKX's fall to 0.0057% per 8h is a genuine fade that resets the stand-down trigger or single-venue drift: no second dated venue print matches it, and Hyperliquid's roughly 0.01% per 8h leans the opposite way.
- The stand-down criterion's original operator, strictly crossing above 0.01% per 8h, was never satisfied; Saturday's print was an exact touch. The criterion remains unresolved and is not retrofitted.
- The 0.05% seven-day-mean crowding bar is this letter's own threshold, not an external convention; the observed 0.0055% sits far below it but the bar itself is unvalidated.
- Whether $63.79B market-wide open interest reflects trimming, closures or new positions.
- Cause of the oil bid: no supply, geopolitical, inventory or shipping event was found; whether Sunday's jump is an event or thin-liquidity noise is unknown.
- FOMC consensus, dot plot and FedWatch odds are unknown, so hawkish and dovish outcomes cannot be sized.
How the outlook has evolved
Bitcoin keeps the $76,500 to $78,300 range into the Sep 16 Fed.
Eighth consecutive read inside the band; price $76,738.8, about $240 above the floor. Not contradicted, still unresolved ahead of the Fed.
Original criteria and dates
Bitcoin trades inside $76,500 to $78,300 on the day of the Sep 16 FOMC statement.
The cheap-deck read is dead; stop all adds if funding crosses 0.01% per 8h.
Saturday printed exactly 0.01%, an exact touch that does not satisfy the strict operator; the current read is 0.0057%, below the trigger. The original criterion was never satisfied and is not retrofitted. Adds remain stopped as a precaution rather than as a met criterion.
Original criteria and dates
OKX funding strictly above 0.01% per 8h on any read after Sep 12.
Both watched venues lean long-paying; treat the weekend bid as leverage, not demand.
Both remain positive, so the lean technically holds, but OKX fell to the 58th percentile of 90 days while Hyperliquid sits near 0.01% per 8h. The signature is now split, and one venue cannot settle it.
Original criteria and dates
OKX and Hyperliquid each show positive funding at the same read.
Slightly lower open interest means trimming rather than building.
Withdrawn. The same provider prints marginally higher, $63.67B to $63.79B, and a $0.12B move cannot distinguish trimming from building. The directional reading is retired rather than stretched to fit.
Original criteria and dates
Market-wide OI declines across consecutive dated reads.
The cause of the oil surge is not established by returned evidence.
Nine retrieved pages returned no such headline while WTI extended to $102.35 on a Sunday futures print. The gap widens but is not resolved.
Original criteria and dates
A dated sanctions, tanker, OPEC+, inventory or shipping headline tied to the move.
The 5.00% 10-year level is the macro tipping point into the Fed.
The 10-year sits at 4.97% as of the Sep 11 reading, +21bp on the week, just under the threshold. Unresolved; the 2-year's +21bp is measured to Sep 9, so the two are not one synchronized move.
Original criteria and dates
US 10-year yield at or above 5.00% on a dated read.
Technical details
Sources
- US 10-year borrowing costs pull back from 5% in reprieve for Bessent | Reuters
- Global bonds fall as surging oil prices inflame inflation risks | Reuters
- US consumer prices accelerate in August, push Fed closer to rate hike | Reuters
- 10-year yields highest since 2023 | Reuters
- VIEW Stocks, bonds rally after August inflation report | Reuters
- Morning Bid: Lenders say show me the money as bond carnage spreads | Reuters
- Bond yields hit multi-year highs as traders brace for new ECB rate-hike cycle | Reuters
- S&P 500 ends down as Treasury yields rise and traders fret about inflation | Reuters
- Edgy bond investors unconsoled by Bessent's big buyback | Reuters
- Nervy markets await ECB rate hike, US inflation data | Reuters
- BTC Futures Perpetual Funding Rate (All) Chart - Glassnode
- BTC Funding Rate | Live Data & History Charts | CoinGlass
- Ethereum Futures Perpetual Funding Rate All Exchanges - Glassnode
- https://www.theblock.co/data/crypto-markets/futures/btc-funding-rates
- ETH Funding Rate | Live Data & History Charts | CoinGlass
- Ethereum Futures Perpetual Funding Rate (All) V2 Chart - Glassnode
- BNB Futures Perpetual Funding Rate (All) Chart - Glassnode
- BNB Funding Rate | Live Data & History Charts | CoinGlass
- https://docs.deribit.com/api-reference/upcoming/market-data/public-get_funding_chart_data
- BNB Futures Perpetual Funding Rate (All) Chart - Glassnode
- Daily Treasury Rates | U.S. Department of the Treasury
- Federal Reserve Board - H.15 - Selected Interest Rates (Daily) - September 10, 2026
- Daily Treasury Rates | U.S. Department of the Treasury
- Federal Reserve Board - H.15 - Selected Interest Rates (Daily) - September 08, 2026
- U.S. Department of the Treasury
- Federal Reserve Board - H.15 - Selected Interest Rates (Daily) - September 04, 2026
- Federal Reserve Board - H.15 - Selected Interest Rates (Daily) - September 01, 2026
- Daily Treasury Rates | U.S. Department of the Treasury
- Daily Treasury Rates | U.S. Department of the Treasury
- The Fed - Monetary Policy: Monetary Policy Report (Branch)
30 returned sources; citation presence does not establish that every claim is verified.
Snapshot sources
- Binance spot, 4h bars
- Yahoo DX-Y.NYB
- Binance spot ETH
- Yahoo ^GSPC
- Yahoo CL=F
- Yahoo GC=F
- Yahoo 2YY=F (2Y)
- Yahoo BZ=F
- Yahoo ^TNX (10Y)
- Yahoo ^IXIC
- OKX perp, funding history
- ForexFactory calendar
- NY Fed EFFR
- Hyperliquid
- CoinGecko derivatives
Successful readings saved with this edition; separate from researcher retrieval coverage.
How this edition was produced
3 × deepseek/deepseek-v4.1-flash
$0.10 · 4 m 14 s · run on a connected key
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.
Editorial review: unavailable. Model review is not independent verification.
1169 narrative words; 827 supporting words
Geopolitics and energy: 5 assigned-source citations
Economy and policy: 4 assigned-source citations
Crypto flows and positioning: 5 assigned-source citations
Research notes
AGREED
- 2026-09-13 19:12 ET (23:12 UTC read): BTC $76,738.8, 1d −0.7%, 1w −4.5%, 1m +21.7%; ETH $2,476.63, 1w −1.5%, 1m +31.6% [OBSERVED, Binance spot; no URL supplied].
- 2026-09-13: OKX BTC-USDT-SWAP funding 0.0057% per 8h, 7-day average 0.0055%, 58th percentile of 90 days, one venue only [OBSERVED, OKX perp; no URL supplied].
- 2026-09-13: Hyperliquid BTC perp funding 0.00125% per hour, about 0.01% per 8h, OI $2.69B, mark $76,718 [OBSERVED, Hyperliquid; no URL supplied].
- 2026-09-13: CoinGecko-covered BTC derivatives OI $63.79B, no 7-day trend; OKX BTC-USDT perp OI $2.13B, 7d −2.5% [OBSERVED, CoinGecko and OKX; no URLs supplied].
- 2026-09-13: WTI $102.35 (+2.3% day, +11.9% week), Brent $107.53 (+2.8%, +11.7%), Sunday-evening futures prints, not settled closes [OBSERVED, Yahoo CL=F/BZ=F; retrieval handled correctly by A, accepted by B and C].
- 2026-09-01 to 09-07: effective fed funds flat at 3.63% across H.15 releases, discount-window primary credit 3.75%, bank prime 6.75% [federalreserve.gov, https://www.federalreserve.gov/RELEASES/h15/]; policy unmoved while market rates moved.
- 2026-09-03 to 09-09: H.15 breakevens rose, 5-year 2.15 to 2.20, 10-year 2.42 to 2.46, 30-year 2.96 to 2.98 [federalreserve.gov, https://www.federalreserve.gov/RELEASES/h15/].
- 2026-08-13 (row date): latest Treasury real par yields 5/7/10/20/30-year at 2.11/2.23/2.39/2.76/2.97, older than the nominal snapshot [home.treasury.gov, https://home.treasury.gov/resource-center/data-chart-center/interest-rates/TextView?field_tdr_date_value_month=202608&type=daily_treasury_real_yield_curve].
- 2026-09-14 to 09-18: all listed items are due calendar entries, not actuals; FOMC Sep 16 14:00 ET with projections and 14:30 press conference; GBP Bank Rate Sep 17 07:00 ET forecast 3.75%, votes 3-0-6; BOJ Sep 17 22:30 ET forecast below 1.25% versus below 1.00% [forexfactory.com, https://www.forexfactory.com/calendar].
- 2026-09-11: Reuters set relayed by B and C — oil settled sharply higher with a bond selloff, 10-year toward 5%, no gold haven bid, hawkish ECB framing [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/].
- 2026-09-11 to 09-13: no sanctions, tanker, OPEC+, inventory, shipping, freight or insurance headline returned for the oil bid [reuters.com, https://www.reuters.com/world/china/global-markets-corrected-2026-09-11/].
- 2026-09-13: the Sep 12 range $76,500 to $78,300 is not contradicted; price sits about $239 above the floor [OBSERVED; standing state].
- 2026-09-10 to 09-13: US spot BTC and ETH ETF net flows, stablecoin issuance, exchange flows, spot volume and CVD, basis, perp premium, liquidation dollars, and all options series (IV, skew, term structure, strike OI, max pain, dealer positioning) are UNKNOWN [studio.glassnode.com, https://studio.glassnode.com/charts/derivatives.FuturesFundingRatePerpetualAll?a=BTC].
- 2026-09-11 to 09-13: aggregator pages flagged unusable rather than zero — Glassnode OI-weighted funding mean 0%, a 0.004% ETH value carrying a $420,690 token number, CoinGlass "undefined%", an undated $64,741 BTCUSD blurb [coinglass.com, https://www.coinglass.com/FundingRate/BTC]; [theblock.co, https://www.theblock.co/data/crypto-markets/futures/btc-funding-rates].
RESOLVED
- OKX funding percentile fell from 89th on Sep 12 to 58th at the Sep 13 read: this is a one-venue measurement change, so the Sep 12 "trigger touched" state is not the current state [OBSERVED OKX; standing state]. It does not by itself decide fade versus single-venue drift.
- The 10-year: Reuters Sep 11 "retreating 1bp to 4.93%" and the snapshot's 4.97% on Yahoo ^TNX are the same instrument at different times; they must not be blended, and the snapshot value carries no intraday timestamp, so "toward 5%" cannot be dated more precisely than the Sep 11 session [reuters.com, URL above; OBSERVED].
- Weekly yield changes: the snapshot 10-year is dated Sep 11 and the 2-year Sep 9, so "+21bp on the week" for each is measured to different endpoints and is not one synchronized repricing [OBSERVED].
- Calendar prior and forecast fields: the 4.00% forecast against a 3.75% prior cannot be read as a confirmed 25bp hike; no Fed statement, target-range text or minutes was returned, and EFFR is flat at 3.63% [forexfactory.com, URL above; federalreserve.gov, URL above]. Direction is calendar-derived, not confirmed.
- Sunday oil prints versus Friday settles: session and settlement status are unmatched, so A's weekly oil gains cannot be ranked like for like against settled closes and cannot be said to "support" the inflation impulse more strongly [OBSERVED; reuters.com, URL above].
- The Sep 12 OI mechanism claim ("slightly lower OI means trimming, not building"): the same provider now prints marginally higher, $63.67B to $63.79B, so the prior directional claim is not confirmed; a $0.12B move cannot distinguish trimming from building [OBSERVED CoinGecko; standing state].
- Aggregator rendering failures are faults, not prints: "undefined%", a stale 0% funding mean and an implausible ETH token count are not measurements [coinglass.com, theblock.co, studio.glassnode.com, URLs above].
- Researcher A's position is missing, so the peer round is incomplete and no independent agreement may be claimed from it; A's Sep 11 material is available only secondhand through B and C [no A position returned].
UNRESOLVED
- Whether OKX's fall to 0.0057% per 8h is a genuine trigger reset or single-venue drift: no second dated venue print, and Hyperliquid's roughly 0.01% per 8h leans the opposite way [OBSERVED OKX and Hyperliquid].
- Whether the Sep 12 two-venue long-paying lean still holds: OKX now sits below the trigger while Hyperliquid sits at about 0.01% per 8h [OBSERVED].
- The stand-down criterion's original operator, "crosses 0.01% per 8h" (strictly greater): the Sep 12 print was an exact touch, not a clean cross, and the next read is below it, so the criterion remains unresolved and must not be retrofitted [standing state; OBSERVED].
- The 0.05% seven-day-mean crowding bar is the prior brief's own threshold; the observed 0.0055% sits far below it, but no external distribution validates the bar [OBSERVED; standing state].
- Whether $63.79B market-wide OI reflects trimming, closures or new positions [OBSERVED CoinGecko/OKX].
- FOMC Sep 16 consensus, dot plot and FedWatch odds: UNKNOWN, so hawkish and dovish branches cannot be sized [forexfactory.com, URL above].
- Sep 10 and 11 CPI and PPI actuals, and August/September payrolls, unemployment, average hourly earnings, JOLTS, ADP and prior-week claims actuals: no returned source; the employment side of the mandate is unchecked for several reads.
- ETF flows, stablecoin issuance and exchange flows remain UNKNOWN for Sep 10 to 13, an eighth consecutive read.
- All options data remain UNKNOWN; no structure may be named [coinglass.com, URL above].
- Liquidation dollars remain UNKNOWN; nothing returned separates voluntary closure from forced liquidation.
- Spot volume, spot and futures CVD, taker ratios, basis and perp premium remain UNKNOWN, so spot-led versus leverage-led is open.
- Cause of the oil bid: no supply, geopolitical, inventory or shipping event returned; whether the Sunday jump is an event or thin-liquidity noise is UNKNOWN [reuters.com, URL above].
- Fed balance sheet, reserves, reverse repo, TGA and auction demand: UNKNOWN; B's liquidity section is truncated, so no claim from it is accepted [federalreserve.gov H.15 only, URL above].
- Current real yields: the latest dated real curve is the Aug 13 row, so the nominal real-time level is not established [home.treasury.gov, URL above].
- ETH's relative strength on both windows has no dated flow, ETF or protocol attribution [OBSERVED].
- BTC dominance and the market-wide OI 7-day trend remain UNKNOWN.Contribute the next chapter
Choose one researcher or three in Run. Every completed, saved result becomes the latest published edition. Quality issues remain visible alongside it.