🌐

MONETARY TRANSMISSION

How the policy rate actually reaches asset prices, and where the chain jams

bzhmacro bzhmacro.com
Funding markets · 2019 plumbing, 2026 reality

Mapping monetary transmission through funding markets — and where the chain jams.

Transmission is a relay race between balance sheets. Each arrow below is a balance sheet somebody has to volunteer; each red badge is a reason they might not. Built from Pozsar's Global Money Notes #20–26 (Credit Suisse, 2019) and updated to the June 2026 institutional landscape, then taken to data.

9
chains & frictions
F1–F9 live frictions on the map
7
estimations
taken to a 3,076-obs daily panel
11.0%
of GDP
fitted reserve-abundance elbow
+16bp
ceiling leak
SOFR above SRF on days it is tapped
00

The argument, and what to read

Three artefacts: the interactive map on this page, the written note behind it, and the estimation results that test it. Each note is downloadable as a PDF.

ChartsTen funding-market charts and the seven estimation exhibits.FX intervention mapInteractive participants, T-account chain and funding-route calculator.Shadow banking appendicesThe 2010 Pozsar maps: eight appendix graphs, T-account reconstructions, the consolidated system map.Source registerEvery series, release and paper behind the map, with dates.
F1–F9 · live frictions new or rebuilt since 2019 rate-setting arrows

Diagrams are wide — scroll them sideways on a narrow screen.

01

The grand transmission map (2026)

How the policy rate actually reaches asset prices: administered rates → overnight money markets → the front end → term & global markets → end investors. In Pozsar's terms, each arrow is a balance sheet someone must volunteer; each red badge is a reason they might not.

100% zoom in, then drag sideways to pan
POLICY & FACILITIES O/N MONEY MARKETS FRONT END / COLLATERAL TERM & GLOBAL END DEMAND FOMC target range meeting-to-meeting · macro horizon IORB banks' reserve remuneration ON RRP — the floor MMF outside option · drained ~$2.5T 23–25 SRF — the ceiling launched 2021 · morning op Jun-25 cap removed Dec-25 (full allotment) NEW 2021+ FIMA repo foreign officials monetize USTs at Fed NEW Discount window stigma uncured · reform agenda F8 Balance sheet: RMPs QT ended Dec-25 · ~$40bn/mo bills RDE ampleness telemetry (2024) NEW Unsecured: EFFR / OBFR FHLBs lend → foreign-bank arb the "$65bn pond" — lagging gauge Secured: SOFR complex ~$12.6T daily exposures (Q3-25) tri-party (MMF cash) · $3.1T FICC cleared + sponsored · $4.4T uncleared bilateral (NCCBR) · $5.0T mandatory clearing: cash Dec-26 · repo Jun-27 RESHAPED F5 Dealers' matched books primary dealers underwrite all supply SLR · G-SIB score · daily vs Q-end reporting F2 Bank HQLA toggle reserves ⇄ repo when GC > IORB bounded by internal liquidity stress tests F1 Intraday plumbing BNY tri-party · Fedwire · daylight credit >50% of value settles after 3pm F4 T-bills MMFs $8.3T · stablecoins ~$300bn foreign repo pool ~$320bn bill share above 15–20% guideline (ATI) +STABLES F6 TGA debt-ceiling drawdowns / rebuilds ±$500bn reserve swings, "shadow QT" F7 Treasury supply coupons vs bills mix · refunding cycle buybacks since 2024 ($4bn/wk liquidity) chunky settlement days >$25/50bn Collateral velocity specials ↔ GC · netting at FICC CCP margin = new state variable FX swaps / hedging cost 3m cost = OIS + XCCY basis turn premium ≥50bp from Oct 1 Pozsar's 3FXs/10s inversion gauge Cash–futures basis HF short futures >$1T notional Cayman UST $1.85T · ~18:1 leverage funded in sponsored repo F3 Term UST curve slope vs funding cost decides who the marginal buyer is SOFR-indexed credit LIBOR dead Jun-23: loans now price off collateralized repo, not bank credit NEW FX-hedged real money Japan lifers: hedge ratios <30% (14y low) Taiwan: ~47% hedged · May-25 panic 3m roll horizon · home-bogey comparison F9 Leveraged funds now the marginal UST financier — replaced 2019's hedged foreigners MMFs & cash investors daily allocation: RRP vs bills vs repo 37% of MMF repo now FICC-cleared Borrowers / real economy floating-rate debt resets off SOFR; repo squeezes pass into loan pricing corridor SOFR sets bill yields hedged carry THE FRICTIONS — what can break each link F1 Reserve scarcity & distribution. No true "excess" reserves: LCR, internal stress tests, resolution liquidity lock them down. Late-2025: SOFR > IORB at ~$2.9T reserves; LCLoR higher than believed (Waller: 8–9% GDP). F2 Dealer balance sheet. SLR + G-SIB score cap matched books; eSLR recalibrated Apr-26 but no UST/reserve carve-out; year-end window dressing structurally intact (G-SIB fix only proposed Mar-26). F3 No basis-trade backstop. ~$1T+ leveraged position with no dedicated facility; Kashyap–Stein basis-purchase facility is academic only. Forced unwind = March 2020 redux; April 2025 was the warning shot. F4 Intraday payment timing. Copeland–Duffie–Yang channel: low reserves → delayed payments → repo distortions. Morning SRF op (Jun-25) helps; Fedwire hours expansion raises reserve demand. F5 CCP margin procyclicality. Mandatory clearing (cash Dec-26, repo Jun-27) economizes balance sheet via netting but concentrates risk at FICC/CME/ICE and adds margin spirals as a stress channel. F6 Stablecoin run risk. GENIUS Act money holds bills/repo with no Fed access — an untested runnable liability at the very front of the curve. F7 Fiscal–monetary entanglement. ATI bill-heavy issuance + Fed RMP bill buying + stablecoin bill demand crowd one instrument; TGA/debt-ceiling swings remain the largest single reserve shock. F8 Discount-window stigma uncured (SVB went to FHLBs first); SRF works as the stigma-free window only for repo collateral, only for counterparties. F9 Hedging panics. Marginal foreign holders increasingly unhedged; re-hedging is now an episodic shock (TWD +8% in days, May-25) rather than a smooth 3m roll.
02

Players × time horizon — who acts when

The same market clears at five speeds. Stress propagates downward (strategic decisions set the stage) and erupts upward (intraday failures force strategic responses — Sept 2019, Dec 2025).

100% zoom in, then drag sideways to pan
INTRADAY minutes–hours BNY (sole tri-party agent): settles every dealer; daylight credit. Dealers: 9am auction takedown → 3:30pm refill; clearing accounts must end flat. Banks: time payments strategically — >half of Fedwire value after 3pm when reserves are tight. Fed desk: SRF ops 8:15am & 1:30pm. Foreign repo pool: cash returned 8:30am — official accounts hold intraday liquidity others must pre-fund. Binding constraint: sunset repayment of daylight overdrafts. DAILY o/n decisions MMFs ($8.3T): allocate across ON RRP / bills / tri-party / FICC sponsored — the marginal cash lender. Banks: HQLA toggle (lend reserves into GC if GC > IORB). Hedge funds: roll repo on basis books daily; sponsored repo via BNY/State Street/JPM. FHLBs: fed funds supply before 9am. Stablecoins: mint/redeem = bill flows. Fed desk: RMP bill purchases (~$40bn/mo since Dec-25); SRF as standing ceiling. Watch: SOFR–IORB, EFFR–IORB, RDE, SRF take-up. MONTH / QUARTER reporting dates European/Japanese dealers: leverage is snapshot at quarter-end → matched books shrink, repo & FX swap spreads pop (US banks report daily averages — they stay). Corporates: tax dates (Apr/Jun/Sep/Dec 15) drain MMF/bank cash into TGA. Treasury: quarterly refunding (Feb/May/Aug/Nov) sets coupon sizes; TBAC guidance. FX-hedged investors: 3m hedge rolls — hedged carry vs home bogey decides participation (Pozsar's 3FXs/10s). YEAR-END the G-SIB window US G-SIBs: surcharge score snapshot Dec 31 → shed repo, FX swaps, market-making from Q4 (preference: repo < USTs < FX swaps in score cost); equity upgrades w/ Canadian pensions. From Oct 1: year-end turn enters the 3m FX-hedging window — hedging costs +≥50bp mechanically. Foreign banks withdraw; non-Basel III dealers (Cantor, Jefferies…) arbitrage. Fed: pre-announced multi-day SRF ops; record take-up Dec-25 ($74.6bn) — ceiling held, but SOFR still printed through the SRF rate (redistribution friction F1/F2). STRATEGIC quarters–years FOMC: framework, QT/RMP path, facility design (multi-year lead times — Pozsar's "race against time"). Regulators: eSLR (Apr-26), Basel endgame & G-SIB reproposal (Mar-26), clearing mandate (Dec-26/Jun-27). Treasury: bills-vs-coupons mix (ATI), buybacks, debt-ceiling brinkmanship. Congress: GENIUS Act, discount-window bills, debt ceiling. Japanese/Taiwanese lifers: hedge-ratio policy, home-market rotation (BOJ exit 2024 → JGBs dominate hedged USTs). BOJ/ECB: local curves set the dollar-hedging calculus.
03

The o/n rate hierarchy — Pozsar's repo map, 2026 edition

After GMN #25's "hierarchy of o/n repo rates": segments matched through dealer balance sheets, ~5bp apart in calm markets, policed by the ON RRP below and — since 2021 — the SRF above. The 2019 problem (no ceiling) is fixed; the 2026 problem is whether ceiling cash can reach non-counterparties (no netting, no all-to-all).

100% zoom in, then drag sideways to pan
rate ↑ Discount window (primary credit) = top of range all banks · stigmatized → rarely polices anything (F8) unsecured, collateral pre-positioned SRF minimum bid = top of range — THE CEILING (since Jul-21) primary dealers + banks · full allotment since Dec-25 · 8:15am + 1:30pm ops · trades don't net (F1) Oct 31-25: SOFR fixed 4.22% vs SRF 4.00% — ceiling leaked 22bp on month-end (redistribution) Uncleared bilateral (NCCBR) ~$5.0T — hedge funds borrow here & in sponsored price by wallet share: "big fish pay less" · migrates to FICC by Jun-27 FICC cleared GC (DVP/GCF) + sponsored ~$4.4T — the growth segment nets → cheapest in balance-sheet terms · sponsors: BNY, State Street, JPM · 37% of MMF repo SOFR — volume-weighted median of the secured complex; post-LIBOR it indexes the entire credit stack Tri-party GC (TGCR) — captive MMF cash, settles on BNY platform "no point paying up for money-fund cash" — bottom of the private stack ON RRP = bottom of range — THE FLOOR MMFs + GSEs · $2.55T peak Dec-22 → ~$0 by mid-25: the QT buffer is gone (F1 now binds directly) Specials — trade through the floor (collateral premium, fails cap at 3% charge) barometer of collateral scarcity, shorts & basis demand SPREADS THAT CARRY THE SIGNAL (estimable) SOFR − IORB: scarcity gauge — structurally negative when abundant, positive when scarce EFFR − IORB: bank-side scarcity (lags repo; the 2019 "low-beta" lesson) SOFR 99th pct − median: tail stress on settlement days (chunky supply >$25/50bn) SOFR − SRF rate: ceiling integrity / redistribution friction GCF/cleared − tri-party: dealer balance-sheet rent across segments (~5bp calm) Bills − OIS / bills − RRP: front-end glut vs shortage (−30bp norm pre-2018) 3m FX-hedged 10y UST − home bogey (3FXs/10s): foreign participation switch t/n FX swap implied − o/n GC: the most peripheral rate (~50bp+, blows out first) Futures − cash basis & swap spreads: leveraged-fund positioning stress Turn premia term structure (Dec/Mar IMM FX points): reporting-date frictions priced ON RRP + SRF usage jointly: cash & collateral stuck on opposite sides of Fed B/S RDE (NY Fed, monthly): official reserve-demand-elasticity telemetry, negative = scarce Episodes that identify the curve: Sep 17 2019 — SOFR 5.25%, EFFR breaks band: no ceiling, free float ≈ 0 (GMN #22 called the mechanism) Mar 2020 — dealer capacity overwhelmed; Mar 2023 — FIMA $60bn draw; Apr 2025 — basis tremor, no Fed action Oct–Dec 2025 — scarcity 2.0: SOFR > IORB > SRF leaks; QT stopped, cap removed, RMPs started
04

The intraday clock — GMN #22's settlement mechanics, 2026 edition

Pozsar's core insight: reserves are scarcest at specific hours, not on average. Treasury settlement mornings destroy dealer cash at 9am that only returns via repo at 3:30pm — the gap is bridged by daylight credit that must die by sunset.

100% zoom in, then drag sideways to pan
7:00 8:15 8:30 9:00 9:30 13:30 15:00 15:30 ~18:30 8:15–8:30 Morning SRF op (new Jun-25 — intraday fix) 8:30 Foreign repo pool cash returned (bilateral, not 3:30!) 9:00 — settlement dealers draw clearing accts to take bonds; TGA fills; daylight overdraft peak 9:30 Treasury redeems maturing Fed holdings (runoff days shred reserves) 13:30–13:45 Afternoon SRF op (the original one) 15:00+ — payment timing game banks settle >50% of Fedwire value after 3pm; delays worsen as reserves shrink (CDY channel) 15:30 tri-party settlement: dealers refill clearing accts by selling or repoing Sunset daylight overdrafts at Fed must be repaid; unpaid → DW loan + disclosure ("repo literally happens at night") Why this map matters for estimation: · Reserve demand is an intraday object: regress payment-delay measures & repo dispersion on reserve levels and settlement-day size (Copeland–Duffie–Yang, QJE 2025). · Settlement-day dummies (net new supply >$25bn / >$50bn) shift SOFR's right tail — Pozsar's "chunky day" rule is directly testable on FRBNY percentile data. · The morning SRF op is a natural experiment (Jun-25): did 8:15am availability compress morning repo dispersion and payment delays? · T+1 (May-24) compressed FX legs for foreign buyers; CLS cutoffs unchanged — check t/n swap pricing around US holidays.
05

The friction calendar

Regulatory windows and fiscal dates make funding stress seasonal and partly predictable — the basis of any turn-trading or stress-anticipation framework.

100% zoom in, then drag sideways to pan
JanFebMarApr MayJunJulAug SepOctNovDec quarter-ends: EU/JP dealer leverage snapshots → repo/FX-swap turn premia (Mar, Jun, Sep) Dec 31: G-SIB score snapshot + year-end turn — the big one corporate tax dates (15th of Apr · Jun · Sep · Dec): cash → TGA, reserves drain — the Sept-2019 detonator when paired with a settlement day quarterly refunding (Feb · May · Aug · Nov): coupon sizes, buyback schedule, bill-share guidance — the fiscal supply calendar Oct 1: year-end turn enters the 3m hedging window — FX hedging costs jump ≥50bp; inversion vs funding deepens mechanically Episodic overlays: debt-ceiling X-dates (TGA drawdown → artificial reserve abundance → violent rebuild after resolution) · month-ends (window dressing lite: Oct 31-25 SOFR through SRF) · auction settlement days (chunky >$25/50bn) · MMF/stablecoin redemption shocks. Estimation: turn premia from FX forward points and SOFR futures vs realized — a priced, recurring friction whose compression/widening tracks regulatory reform in real time (eSLR Apr-26 is the live experiment).
06

2019 → 2026: what got fixed, what's still open, what's new

Scorecard of Pozsar's 2019 worry-list against the June 2026 institutional landscape.

100% zoom in, then drag sideways to pan
FIXED / BUILT SINCE 2019 ✓ SRF (Jul-21) — the o/n repo facility GMN #21/#25 demanded; priced top-of-range; bank counterparties (Pozsar's design); morning op Jun-25; cap removed Dec-25 ✓ FIMA repo (2020/21) — foreign officials no longer must sell; $60bn draw Mar-23 worked ✓ Foreign repo pool defanged — repriced toward ON RRP; "Sagittarius A*" closed ✓ Ample-reserve telemetry — RDE published monthly (Oct-24); flagged late-25 scarcity early ✓ QT stopped on the indicators (Dec-25) + RMP bill purchases — "mini-QE" framework exists ✓ eSLR recalibrated (eff. Apr-26) — G-SIB buffer 50% of Method-1 surcharge; frees dealer B/S ✓ LIBOR → SOFR (Jun-23) — credit stack now indexed to the secured rate the Fed can police ✓ FICC sponsored repo — MMF cash reaches borrowers with netting; 37% of MMF repo ✓ Treasury buybacks (2024) — off-the-run liquidity backstop, cash-management smoothing ✓ MMF reform (2023) — gates/fees delinked; institutional prime shrank further ✓ T+1 (May-24) — settled without FX-fails spike Verdict: the 2019 architecture gap — a floor with no ceiling — is closed. Sept-2019 (no ceiling, SOFR +300bp) vs Oct/Dec-2025 (ceiling, leak of ~20bp): same shock class, two orders of magnitude smaller price error. STILL OPEN / NEWLY CREATED ✗ Redistribution — SRF cash reaches counterparties only; trades don't net (FICC settlement discussed, not done); SOFR printed through the ceiling Oct–Dec 25 ✗ No UST/reserve SLR carve-out — rejected in the Nov-25 final rule; untested at scale ✗ G-SIB year-end window — Method-2 recalibration only proposed (Mar-26); turn intact ✗ Basis trade >$1T with no backstop — Kashyap–Stein facility is a paper; Apr-25 was a near-miss; CCP margin on it from Jun-27 with unknown net effect ✗ Discount-window stigma — uncured (SVB); reform bills pending; LCR credit unbuilt ✗ LCLoR uncertainty — ~$2.9T was already uncomfortable in late-25; Waller's 8–9% of GDP vs Logan's lean-reserves view — the Fed is steering by a noisy gauge ⚠ NEW: stablecoins — runnable bill-backed money, no Fed access, $300bn → $2T? ⚠ NEW: fiscal-monetary entanglement — ATI bill issuance + Fed RMPs + stablecoin demand jointly manage the front end; three hands on one steering wheel ⚠ NEW: ownership fragility — marginal UST holder went from FX-hedged lifers (slow, mandate-driven) to leveraged Cayman funds + unhedged Asian books (fast, fickle) ⚠ NEW: CCP concentration — FICC/CME/ICE become the system's choke points after 2027 The 2019 question was "can the Fed keep o/n rates in the band?" The 2026 question is "can ceiling liquidity reach a leveraged, cleared, fiscally-dominated market before positions unwind?"
07

Notes, results and data

The written companions to this map, as HTML and as PDF.

08

Sources

SRC Pozsar, Global Money Notes #20–26, Credit Suisse (Feb–Dec 2019) — the 2019 baseline for every panel above.
SRC FOMC and NY Fed operating policy statements: SRF launch (28 Jul 2021), morning operations (28 May 2025), QT end (29 Oct 2025), full allotment and reserve management purchases (10 Dec 2025).
SRC Speeches: Perli (Nov 2025 – May 2026), Waller (Jul 2025), Logan (Oct 2025), Williams (Sep 2025).
SRC Rules: eSLR final rule (Fed/OCC/FDIC, 25 Nov 2025, effective 1 Apr 2026); Basel III endgame and G-SIB reproposals (Mar 2026); SEC Treasury clearing rule (Dec 2023) and extension (25 Feb 2025 → cash Dec 2026, repo Jun 2027); SEC money-market fund reform (Jul 2023); GENIUS Act (18 Jul 2025).
SRC Research: Afonso et al., Market Events of Mid-September 2019 (SR918) and Reserve Demand Elasticity (monthly since Oct 2024); Kahn et al. (OFR 23-04); Copeland, Duffie & Yang, Reserves Were Not So Ample After All (QJE 2025); Duffie, Resilience Redux (Jackson Hole 2023) and SR1070; Kashyap, Stein, Wallen & Younger (BPEA 2025); OFR, Sizing the US Repo Market (Dec 2025).
SRC Data: FRED, the Federal Reserve H.4.1, the Daily Treasury Statement, NY Fed reference rates and operations, OFR repo and hedge-fund monitors, CFTC Traders in Financial Futures, SEC N-MFP, TreasuryDirect auction results.

The full register, with release dates and refresh cadence, is on the sources page.