US FED / TREASURY LIQUIDITY — WEEKLY MEMO
Week Ending 2026-09-11
The biggest liquidity injection of the year and the hawkish CPI that will take it straight back: net liquidity rose +$112.2B as the TGA drained -$100.7B to $843.7B with no coupon settlement in the window, pushing bank reserves +$107.2B to $3,036.5B and back above $3 trillion for the first time since June -- but the August CPI's core overshoot on Friday lifted September hike odds to roughly 86%, and the $119B refunding settlement plus the corporate tax date both land on September 15.
URLI Score
+40.50
Bullish liquidity backdrop
Net Liquidity
+$112.2B
Huge injection
Bank Reserves
+$107.2B
Reserve build
TGA Change
-$100.7B
Cash injection
RRP Change
-$8.1B
Cash released
Fed Regime
Hold
3.50%-3.75%
Treasury Outlook
TGA rebuild
Liquidity risk
Market Bias
RISK ON
Supportive
Executive Conclusion
Supportive
- Net liquidity rose +$112.2B, the largest weekly increase of 2026: the -$100.7B TGA drawdown and a -$8.1B RRP decline both added reserves while the Fed balance sheet was essentially flat (+$3.4B).
- Bank reserves rose +$107.2B to $3,036.5B, back above the $3 trillion threshold for the first time since June and scoring +100 on the reserve rubric.
- Funding markets eased rather than tightened as the cash arrived: SOFR-IORB averaged -0.80bp over the five days to September 10 and closed the window at -3bp, with SOFR printing below the administered IORB floor.
- The broad dollar softened on the confirmed prints, from 118.75 on August 28 to 118.07 on September 4 (-0.6%), a mild offset to the rates repricing.
- Core CPI at 2.4% year over year is the lowest core print of the cycle in level terms, even though the monthly figure overshot.
- Equities recovered on CPI day itself -- the S&P 500 rose +0.9% and the Nasdaq +1.0% on September 11 after three down days -- and the HY spread held at 2.70%, up only 2bp on the week.
Restrictive
- The August CPI core rose +0.3% month over month against a +0.2% consensus (headline +0.4% m/m, 3.4% y/y), and CME FedWatch odds of a 25bp hike on September 16 jumped to roughly 86% from about 60% a week earlier; TD Securities moved to forecasting three hikes in the cycle.
- The entire TGA drawdown is a settlement-calendar artefact: roughly $119B of September refunding settles Tuesday September 15, the same day quarterly corporate estimated taxes are due, so the cash comes straight back out of reserves next week.
- The Treasury outlook score moves from -30 to -50 (TGA rebuild expected): at $843.7B the balance is more than $100B below the QRA's ~$950B end-September target and ~$200B below its late-October ~$1.05T peak guidance.
- Yields repriced sharply higher across the curve: the 2-year +19bp to 4.56%, the 10-year +17bp to 4.95% and the 30-year +13bp to 5.37% between September 4 and September 10, scoring the dollar/yield component at -30 versus 0 last week.
- The forward projection for the week ending September 18 is -$90.7B of net liquidity and is flagged needs_manual_review, because the tax date, the quarter-end-adjacent window and the FOMC can each dominate the mechanical estimate.
- Bitcoin fell -3.0% to $77,276 and gold eased -0.7% (LBMA PM $4,386.25) even as reserves surged, a reminder that the rate path rather than the weekly plumbing is setting risk-asset direction.
Main Warning
- Do not read this week's URLI as a turn. Every dollar of the +$112.2B net-liquidity gain came from a Treasury cash balance that fell only because the September refunding had not settled yet, and it settles Tuesday September 15 -- roughly $119B of 3/10/30-year supply on the same day quarterly corporate estimated taxes are collected. The calendar-mechanical model already projects that reversal at -$90.7B of net liquidity for the week ending September 18, and it flags the estimate for manual review because a tax date, a quarter-end-adjacent window and an FOMC decision all fall inside it, any one of which can dominate the mechanical assumption. The policy side hardened at the same time: the August CPI's core overshoot (+0.3% m/m against +0.2% expected) moved September hike odds to roughly 86%, and the rate-path score is held at -50 only on a literal reading of the rubric, since the Committee has not actually raised rates yet. If it hikes on September 16 that score drops to -80 and the two negatives compound -- a Fed tightening into a mechanical TGA rebuild, with reserves handed back the $107B they just gained. The genuinely reassuring signal is the funding market: SOFR traded three basis points below IORB at the end of the window and reserves are back above $3 trillion, so the system has the cushion to absorb next week's drain without plumbing stress. The risk here is the rate path and the cash rebuild, not a funding accident.
POSITIVE URLI +40.50 — Bullish liquidity backdrop.
WEEKLY LIQUIDITY MAP
Liquidity Waterline
Every node is a water tank: the solid fill is this week's level, the dashed line is last week, and the faint line is the 3-month average. Funding sources feed net liquidity, which flows through the risk gate to crypto markets. Fill colour marks liquidity effect, not raw level.
Funding Sources
TGA-$100.7BADD
Fed B/S+$3.4BNEUTRAL
Reserves+$107.2BADD
RRP-$8.1BNEUTRAL
Net Liquidity
US Net Cash+112.2BADD
Risk Gate
Risk AssetsURLI +40.5ADD
Crypto Markets
Crypto Beta$2.61TNEUTRAL
Meme Beta$32BADD
AddDrainWatchNeutralLast week3-month avg
Liquidity tanks update weekly (Fed H.4.1 / FRED · as of Sep 11, 2026). Crypto & meme market cap and read use the weekly report snapshot (CoinGecko · as of Sep 13, 2026). Homepage and latest weekly memo use the same Waterline snapshot. AI-readable: JSON · Markdown.
What Changed This Week
| Item | Previous | Latest | Change | Impact |
|---|---|---|---|---|
| Fed total assets | $6,737.2B | $6,740.6B | +$3.4B (+0.1%) | POSITIVE |
| Bank reserves | $2,929.3B | $3,036.5B | +$107.2B (+3.7%) | POSITIVE |
| TGA | $944.4B | $843.7B | -$100.7B (-10.7%) | POSITIVE |
| RRP | $357.7B | $349.7B | -$8.1B (-2.3%) | POSITIVE |
| Fed rate path | — | hold with hike risk, now close to certain -- the August CPI released Friday September 11 came in hot on the margin that matters: core CPI rose +0.3% month over month against a +0.2% consensus, with headline +0.4% month over month and 3.4% year over year. Core is still only 2.4% year over year in level terms, but the monthly overshoot five days before the decision removed the last reason for the Committee to wait, and CME FedWatch odds of a 25bp hike at the September 15-16 meeting jumped to roughly 86% from about 60% a week earlier. The target range is still 3.50%-3.75% and the Committee has not yet acted, so the rubric score stays at -50 rather than -80 | 3.50%-3.75% | NEGATIVE |
| Dollar / yields | — | negative -- the scored 10-year move was +17bp (4.78% on September 4 to 4.95% on September 10), well outside the +/-8bp flat band, as the curve repriced a September hike; the broad dollar scores flat only because the H.10 index's publication lag leaves September 4 as both the latest confirmed print and the latest print seven days before the report date. On the confirmed prints the broad dollar actually SOFTENED, from 118.75 on August 28 to 118.07 on September 4 (-0.6%), so the dollar leg is a mild offset to a clearly negative rates leg | Broad USD 118.07 (FRED/H.10 DTWEXBGS, 2026-09-04 -- the confirmed print advanced one week from last report's 2026-08-28, and the H.10 index's multi-day publication lag means no confirmed print past 2026-09-04 as of this report date). Against the prior confirmed weekly print (118.75 on 2026-08-28) the broad dollar eased -0.6%, giving back the post-payrolls firming; the scored change is nonetheless 0.000% because the rubric compares the latest confirmed print with the latest print seven days before the report date, and both resolve to 2026-09-04. | NEGATIVE |
Previous week: 2026-09-04 (H.4.1 weekly levels).
Fed Balance Sheet Detail
| Indicator | Latest | Weekly Change | Read |
|---|---|---|---|
| Fed total assets | $6,740.6B | +$3.4B (+0.1%) | POSITIVE |
| Securities held outright | $6,468.3B | -$0.0B (-0.0%) | NEUTRAL |
| Treasury securities | $4,552.3B | -$0.0B (-0.0%) | NEUTRAL |
| Bank reserves | $3,036.5B | +$107.2B (+3.7%) | POSITIVE |
| Discount window | $5.9B | +$0.6B (+10.3%) | WATCH |
H.4.1 Wednesday levels in millions of dollars for September 9, 2026, from the September 10 release. WALCL (total assets 6,740,619, +3,415), WRBWFRBL (bank reserves 3,036,508, +107,223), WDTGAL (TGA 843,705, -100,659) and WLRRAL (RRP 349,663, -8,079) verified against FRED. Detail lines are read from the LAST column of H.4.1 Table 1 ('Wednesday Sep 9, 2026'), not the 'Averages of daily figures' columns that come first: Reserve Bank credit Wednesday level 6,693,205 (the week-average is 6,691,811, which is what FRED's WRESCRT carries -- the average series must not be used for this field); Securities held outright 6,468,269 = U.S. Treasury 4,552,337 + agency debt 2,347 + MBS 1,913,585 (components sum exactly); Loans (discount_window) Wednesday level 5,907, of which primary credit 5,838, seasonal credit 54 and PPPLF 16; other_emergency_facilities = net portfolio holdings of MS Facilities 2020 LLC, Wednesday level 895. Securities held outright, Treasury securities, MBS and Loans were each cross-checked against the corresponding FRED Wednesday-level series (WSHOSHO 6,468,269; TREAST 4,552,337; WSHOMCB 1,913,585; WLCFLL 5,907; WSHOFADSL 2,347). Each _change is computed Wednesday-over-Wednesday against the September 2 Wednesday levels recorded in last week's file (total assets 6,737,204; RBC 6,689,880; securities held outright 6,468,278; Treasuries 4,552,347; MBS 1,913,585; Loans 5,355; MS Facilities 895): securities held outright -9, entirely a -10 mark on Treasury holdings with MBS flat for a second week, Loans +552, MS Facilities unchanged, Reserve Bank credit +3,325. The release's own Table 1 change column (+4,366 for Reserve Bank credit) is a week-average-over-week-average change and is deliberately not used here. Internal consistency check on the Wednesday column: Reserve Bank credit 6,693,205 + foreign currency denominated assets 19,567 + gold stock 11,041 + SDR certificates 15,200 + Treasury currency outstanding 53,381 = 6,792,394 = total factors supplying reserve funds, as printed.
Net Liquidity Calculation
Net Liquidity Change = Fed Balance Sheet Change - TGA Change - RRP Change
| Component | Weekly Change | Effect | Notes |
|---|---|---|---|
| Fed balance sheet | +$3.4B | POSITIVE | Balance-sheet growth adds liquidity. |
| TGA | -$100.7B | POSITIVE | TGA up pulls cash into Treasury; TGA down injects it. |
| RRP | -$8.1B | POSITIVE | RRP up parks cash at the Fed; RRP down releases it. |
| Net liquidity | +$112.2B | HUGE INJECTION | Sum of the three flows above. |
URLI — US Risk Liquidity Index
URLI = 0.35 x Net Liquidity + 0.20 x Bank Reserves + 0.15 x Fed Rate Path + 0.10 x Treasury Outlook + 0.10 x Funding Stress + 0.10 x Dollar/Yield Pressure
| Component | Weight | Score | Contribution |
|---|---|---|---|
| Net Liquidity | 35% | +100 | +35.00 |
| Bank Reserves | 20% | +100 | +20.00 |
| Fed Rate Path | 15% | -50 | -7.50 |
| Treasury Outlook | 10% | -50 | -5.00 |
| Funding Stress | 10% | +10 | +1.00 |
| Dollar / Yield Pressure | 10% | -30 | -3.00 |
| Metric | Value |
|---|---|
| Weekly URLI | +40.50 |
| 4-week moving average | -2.00 |
| 13-week moving average | -9.44 |
Observed URLI history
| Week | URLI | Net Liquidity | Reserves | TGA | RRP | Read |
|---|---|---|---|---|---|---|
| 2026-09-11 | +40.50 | +$112.2B | +$107.2B | -$100.7B | -$8.1B | BULLISH LIQUIDITY BACKDROP |
| 2026-09-04 | -3.50 | +$19.8B | +$12.5B | -$15.1B | +$1.6B | NEUTRAL |
| 2026-08-28 | -29.50 | -$20.3B | -$14.0B | +$23.0B | -$17.5B | NEGATIVE |
| 2026-08-21 | -15.50 | -$6.8B | -$16.8B | -$23.0B | +$15.6B | NEGATIVE |
| 2026-08-14 | -39.25 | -$57.4B | -$55.1B | +$30.1B | +$38.7B | NEGATIVE |
Show 60 earlier entries
| Week | URLI | Net Liquidity | Reserves | TGA | RRP | Read |
|---|---|---|---|---|---|---|
| 2026-08-07 | +41.25 | +$69.2B | +$58.2B | -$41.1B | -$17.7B | POSITIVE |
| 2026-07-31 | -64.50 | -$128.6B | -$120.4B | +$135.0B | -$15.7B | NEGATIVE |
| 2026-07-24 | -41.50 | -$33.7B | -$35.6B | +$39.4B | -$1.4B | NEGATIVE |
| 2026-07-17 | -38.50 | -$44.9B | -$36.9B | +$46.7B | +$5.6B | NEGATIVE |
| 2026-07-10 | +31.75 | +$59.1B | +$60.4B | -$58.1B | +$10.0B | POSITIVE |
| 2026-07-03 | +36.75 | +$81.5B | +$122.6B | -$94.5B | +$1.9B | POSITIVE |
| 2026-06-26 | +25.75 | +$53.0B | +$18.1B | -$54.7B | +$0.9B | POSITIVE |
| 2026-06-19 | -66.50 | -$162.7B | -$175.1B | +$155.4B | +$18.3B | NEGATIVE |
| 2026-06-12 | +29.75 | +$67.0B | +$65.8B | -$44.6B | -$8.5B | POSITIVE |
| 2026-06-05 | -33.50 | -$20.9B | -$21.4B | +$3.1B | +$24.9B | NEGATIVE |
| 2026-05-29 | -41.50 | -$28.2B | -$39.7B | +$60.7B | -$41.7B | NEGATIVE |
| 2026-05-22 | -23.50 | -$5.7B | -$10.7B | -$25.4B | +$16.3B | NEGATIVE |
| 2026-05-15 | +31.25 | +$70.7B | +$65.9B | -$55.3B | +$3.7B | POSITIVE |
| 2026-05-08 | +58.50 | +$134.9B | +$132.3B | -$125.3B | -$0.0B | POSITIVE |
| 2026-05-01 | -2.00 | +$14.0B | +$4.6B | -$19.1B | -$2.4B | NEUTRAL |
| 2026-04-24 | -39.25 | -$66.3B | -$65.6B | +$82.7B | -$14.7B | NEGATIVE |
| 2026-04-17 | -51.50 | -$210.5B | -$203.3B | +$227.4B | -$5.0B | NEGATIVE |
| 2026-04-10 | +58.50 | +$119.7B | +$119.1B | -$106.5B | +$5.3B | POSITIVE |
| 2026-04-03 | +21.00 | +$46.6B | +$28.4B | -$33.9B | +$5.5B | POSITIVE |
| 2026-03-27 | +29.00 | +$36.9B | +$37.3B | -$38.4B | +$2.8B | POSITIVE |
| 2026-03-20 | -42.25 | -$66.3B | -$74.3B | +$70.0B | +$5.9B | NEGATIVE |
| 2026-03-13 | +38.25 | +$52.7B | +$59.4B | -$41.2B | +$6.0B | POSITIVE |
| 2026-03-06 | -5.00 | +$16.8B | +$9.3B | +$8.0B | -$9.7B | NEUTRAL |
| 2026-02-27 | +41.25 | +$50.4B | +$44.5B | -$49.9B | -$0.1B | POSITIVE |
| 2026-02-20 | +0.50 | +$4.0B | +$1.5B | -$21.7B | +$8.7B | NEUTRAL |
| 2026-02-13 | +12.00 | +$19.9B | +$21.0B | +$2.8B | -$6.2B | POSITIVE |
| 2026-02-06 | +38.25 | +$63.6B | +$54.6B | -$45.0B | -$0.3B | POSITIVE |
| 2026-01-30 | -36.75 | -$82.3B | -$73.6B | +$81.5B | +$3.8B | NEGATIVE |
| 2026-01-23 | -45.25 | -$96.9B | -$105.2B | +$94.1B | +$5.6B | NEGATIVE |
| 2026-01-16 | +29.00 | +$25.2B | +$38.8B | -$6.5B | -$10.5B | POSITIVE |
| 2026-01-09 | +53.00 | +$156.2B | +$169.6B | -$89.3B | -$133.9B | POSITIVE |
| 2026-01-02 | -61.00 | -$152.6B | -$127.1B | +$71.3B | +$140.7B | NEGATIVE |
| 2025-12-26 | +43.75 | +$95.0B | +$46.9B | -$59.9B | -$10.7B | POSITIVE |
| 2025-12-19 | -30.00 | -$38.7B | -$40.1B | +$55.6B | +$0.7B | NEGATIVE |
| 2025-12-12 | +56.00 | +$107.2B | +$95.6B | -$102.7B | -$1.0B | POSITIVE |
| 2025-12-05 | -26.00 | -$25.7B | -$19.8B | +$8.8B | +$0.2B | NEGATIVE |
| 2025-11-28 | -5.50 | -$6.9B | -$19.5B | -$1.2B | +$5.1B | NEUTRAL |
| 2025-11-21 | +26.00 | +$32.2B | +$34.2B | -$42.3B | -$15.1B | POSITIVE |
| 2025-11-14 | +31.00 | +$37.4B | +$31.4B | +$0.4B | -$30.1B | POSITIVE |
| 2025-11-07 | +7.00 | +$21.8B | +$24.0B | -$41.2B | +$5.1B | NEUTRAL |
| 2025-10-31 | -55.25 | -$94.3B | -$101.7B | +$78.8B | +$13.0B | NEGATIVE |
| 2025-10-24 | -42.25 | -$66.6B | -$58.6B | +$53.1B | +$6.5B | NEGATIVE |
| 2025-10-17 | -24.50 | -$44.3B | -$45.7B | +$57.9B | -$7.9B | NEGATIVE |
| 2025-10-10 | +38.25 | +$58.0B | +$54.3B | -$25.3B | -$28.9B | POSITIVE |
| 2025-10-03 | -34.25 | -$58.2B | -$20.1B | +$61.4B | -$24.5B | NEGATIVE |
| 2025-09-26 | -12.00 | +$11.2B | -$20.7B | -$49.2B | +$37.7B | NEGATIVE |
| 2025-09-19 | -51.50 | -$125.5B | -$130.6B | +$139.6B | -$11.4B | NEGATIVE |
| 2025-09-12 | -2.50 | -$16.0B | -$17.0B | +$5.6B | +$14.2B | NEUTRAL |
| 2025-09-05 | -26.00 | -$43.0B | -$48.9B | +$66.1B | -$24.4B | NEGATIVE |
| 2025-08-29 | -44.25 | -$78.9B | -$80.8B | +$69.7B | -$5.9B | NEGATIVE |
| 2025-08-22 | -29.00 | -$32.7B | -$30.7B | +$10.6B | -$3.1B | NEGATIVE |
| 2025-08-15 | +2.00 | -$4.7B | -$2.0B | +$51.2B | -$43.7B | NEUTRAL |
| 2025-08-08 | +35.00 | +$33.2B | +$30.7B | +$44.9B | -$79.8B | POSITIVE |
| 2025-08-01 | -41.25 | -$59.7B | -$58.7B | +$85.9B | -$41.3B | NEGATIVE |
| 2025-07-25 | -1.50 | -$15.3B | -$16.7B | +$21.5B | -$7.7B | NEUTRAL |
| 2025-07-18 | +23.00 | +$31.4B | +$33.0B | +$1.0B | -$35.1B | POSITIVE |
| 2025-07-11 | +48.25 | +$84.9B | +$85.5B | -$61.2B | -$21.4B | POSITIVE |
| 2025-07-04 | -42.75 | -$87.3B | -$90.8B | +$37.7B | +$47.1B | NEGATIVE |
| 2025-06-27 | +24.50 | +$23.6B | +$24.3B | -$49.3B | +$6.9B | POSITIVE |
| 2025-06-20 | -53.00 | -$104.1B | -$106.7B | +$106.8B | +$1.1B | NEGATIVE |
Observed URLI history covers completed weekly runs. Historical percentile ranking uses URLI-Core: the four data-derived components (Net Liquidity, Bank Reserves, Funding Stress, Dollar/Yield Pressure), representing 75% of URLI weight, ranked against frozen weekly FRED history since 2020 and shown in the distribution gauge above. The two policy-judgment components and 13-week live average are outside this percentile lens.
Forward View
Projected URLI — 2026-09-18 estimate
| Item | Estimate | Bias |
|---|---|---|
| Mechanical midpoint (review only) | -60.75 | PRELIMINARY |
| Backtested URLI range | -69.50 to +40.50 | RANGE |
| Uncertainty band: net liquidity | -$321.5B to $140.1B | NEUTRAL |
| Range note | Low confidence — tax-date / TGA-rebuild week: big swing possible, direction uncertain. Use net liquidity -$321.5B to $140.1B, not the mechanical midpoint, as the primary read. | WATCH |
| Confidence | Model backtest cleared: RMSE $115.4B vs naive $142.9B; hit rate 52.7%. Low confidence: tax-date / TGA-rebuild timing widens the uncertainty band to 2.0x RMSE; manual review required. | WATCH |
| Backtest gate | RMSE improvement 19.25% | MODEL |
| Component | Projected Change | Method |
|---|---|---|
| Fed balance sheet | -4.8B | Trailing 4-week mean |
| TGA | +88.0B | Coupon net + assumed bill roll |
| RRP | -2.1B | Trailing 4-week mean |
| Net liquidity | -90.7B | Fed BS - TGA - RRP |
Projection track record — current estimator
| Metric | Value | Meaning |
|---|---|---|
| Scored weeks (N) | 7 | Scored since the 2026-07-21 model change |
| Directional hit rate | 4/7 | Projected regime sign matched realized |
| Range containment | 7/7 | Realized URLI inside the displayed ±band |
| Avg absolute error | 38.6 pts | Mean |projected − realized| |
| Point bias | -7.4 pts | Mean projected − realized (− = bearish skew) |
The projection model changed materially on 2026-07-21 (maturing bills are now assumed to roll rather than counted as Treasury cash outflows). The 10 earlier scored weeks are excluded above because they were produced by the previous estimator.
Model track record: last 12 completed projections hit 7/12; mean absolute URLI error 39.96.
Assumptions
- TGA projection separates coupon cash flow (+88.0B) from bills. Bills are assumed roll (+0.0B); gross settlements and maturities remain $625.0B and $584.0B.
- Fed balance sheet change carries forward the trailing 4-week mean: -4.8B.
- RRP change carries forward the trailing 4-week mean: -2.1B.
- Projected bank reserves use net liquidity as a reserve-identity approximation; actual reserves can differ.
- Fed rate path, Treasury outlook, funding stress, and dollar/yield scores are held at the latest actual week.
- 2026-09-14 quarterly tax/FOMC-adjacent window; manual review required.
- 2026-09-15 tax-date watch: potential Treasury cash inflow / liquidity drain.
- 2026-09-15 quarterly tax/FOMC-adjacent window; manual review required.
- 2026-09-16 quarterly tax/FOMC-adjacent window; manual review required.
- 2026-09-17 quarterly tax/FOMC-adjacent window; manual review required.
- 2026-09-16 scheduled FOMC decision falls in this projection window; rate-path and balance-sheet guidance can dominate the mechanical estimate.
Projection is a model-derived scenario with a measured historical error band, not an observed URLI value. It is replaced by actual H.4.1 / TGA / RRP data in the next weekly run.
Next-week liquidity calendar
| Date | Event | Expected Size | Liquidity Effect | Bias |
|---|---|---|---|---|
| 2026-09-14 | 13-week bill auction ($92B) and 26-week bill auction ($79B) | $92B 13-week, $79B 26-week, per the TreasuryDirect upcoming feed | Both settle Thursday September 17, inside the week ending September 18; assumed to roll against maturing bills in the calendar-mechanical model | NEUTRAL |
| 2026-09-15 | Settlement of the September refunding (3-year $58B, 10-year $39B, 30-year $22B) and quarterly corporate estimated-tax date | ~$119B combined 3/10/30-year settlement, plus quarterly corporate tax receipts | The single largest scheduled drain of the quarter to date: coupon settlement and the corporate tax date land on the same day, rebuilding the TGA from $843.7B toward the QRA's ~$950B end-September target and draining reserves | NEGATIVE |
| 2026-09-15 | 20-year bond reopening auction (19-year 11-month, $13B); 6-week bill auction ($75B) | $13B 20-year reopening; $75B 6-week bill | The 20-year settles September 18 and the 6-week bill September 17, both inside next week's window; a long-end demand test with the 30-year at 5.37% | NEGATIVE |
| 2026-09-16 | FOMC decision, 2:00 p.m. ET, with Summary of Economic Projections and Chair Warsh press conference (meeting September 15-16) | N/A | No direct plumbing effect in the window, but a hike would move the Fed rate-path score from -50 to -80 and the dot plot will set the path; markets price roughly 86% odds of a 25bp hike after the August CPI | WATCH |
| 2026-09-17 | 9-year 10-month note auction ($19B); 4-week and 8-week bill auctions | $19B note, settling September 30 | Settles September 30, outside next week's window; the bills settle September 22 | NEUTRAL |
| Week of 2026-09-14 | Projected net-liquidity swing: gross settlements ($625B) vs maturities ($584B), coupon cash flow +$88.0B, bills assumed to roll | Projected TGA build of +$88.0B and net liquidity -$90.7B for the week ending September 18 | The calendar-mechanical model flags this week for manual review: the September 15 tax date, the quarter-end-adjacent window and the September 16 FOMC can each dominate the mechanical bill-roll assumption, so the uncertainty band is widened to 2.0x RMSE and the point estimate is published as provisional rather than as a normal directional call | NEGATIVE |
Alerts & Warnings
| Alert | Status | Notes |
|---|---|---|
| Huge liquidity injection | POSITIVE | Net liquidity +112.2B breached the +$100B major threshold. |
| Reserve shock | POSITIVE | Bank reserves +107.2B, above the +$75B threshold. |
| TGA rebuild risk | WATCH | The TGA fell -$100.7B to $843.7B, the largest single-week drawdown since the spring, and that drawdown is almost entirely a calendar artefact rather than a change in fiscal stance. The September refunding was auctioned inside this week (3-year September 8, 10-year September 9, 30-year September 10) but every one of those settles Tuesday September 15, outside this data week, so the week carried benefit payments and other outlays with no offsetting coupon cash inflow. The result is a cash balance sitting more than $100B below Treasury's own ~$950B end-September target, which makes a rebuild next week close to mechanical rather than merely expected: roughly $119B of 3/10/30-year supply settles September 15, the September 15 quarterly corporate estimated-tax date falls the same day, and the $13B 19-year-11-month (20-year) reopening auctioned September 15 settles September 18. Bills auctioned September 14-15 ($92B 13-week, $79B 26-week, $75B 6-week) settle September 17, inside next week's window. The score therefore moves from -30 (bill/coupon issuance pressure) to -50 (TGA rebuild expected): the QRA still guides the balance to a late-October peak near $1.05 trillion, and the path from $843.7B to that peak is a sustained drain on reserves. Long-end yields backed up on the hawkish CPI repricing -- the 30-year reached 5.37% and the 10-year 4.95% on September 10, both the highs of the cycle window -- which raises the cost of that rebuild and is the main reason the 3/10/30-year auctions this week drew close attention. |
| RRP shock | NEUTRAL | RRP -8.1B; below the $50B shock threshold. |
| Fed rate path | NEGATIVE | No FOMC meeting this week; the decision lands September 16 (meeting September 15-16, with a Summary of Economic Projections and dot plot). The August CPI on Friday September 11 was the hinge event flagged in last week's report, and it broke hawkish. Headline CPI rose +0.4% month over month and 3.4% year over year; core (all items less food and energy) rose +0.3% month over month and 2.4% year over year. Both the BLS release and FRED's CPIAUCSL/CPILFESL series were checked and agree (CPIAUCSL +0.40% m/m, +3.35% y/y on the seasonally adjusted index and +3.40% on the NSA index; CPILFESL +0.29% m/m, +2.45% y/y). The level of core inflation at 2.4% is not itself alarming -- it is the lowest core year-over-year print of the cycle -- but the market traded the monthly core overshoot (+0.3% vs a +0.2% consensus) rather than the level, and CME FedWatch odds of a 25bp September hike moved to roughly 86% from about 60% a week earlier, with prediction markets in the 65-80% range. At least one primary dealer (TD Securities) moved to forecasting three hikes in the cycle, with the next two in October and January. The score is held at -50 ('hold with hike risk') rather than moved to -80 ('active tightening') on a deliberately literal reading of the rubric: the fed funds target range is unchanged at 3.50%-3.75% and the Committee has not raised rates. That is a conservative choice and it understates the forward drag -- if the Committee hikes on September 16 this score moves to -80 next week, and the rate-path component of URLI will fall accordingly. Chair Warsh's Jackson Hole framing of the economy as 'at full employment' and the +162,000 August payroll print from September 4 remain the backdrop. |
| Funding stress | POSITIVE | FRED-derived: -0.80 bps avg SOFR-IORB (2026-09-03 to 2026-09-10, 5 business days): +1.00 bps Sep 3, +0.00 bps Sep 4, -1.00 bps Sep 8, -1.00 bps Sep 9, -3.00 bps Sep 10. IORB held at 3.65% throughout; SOFR closed the window at 3.62%, three basis points below the floor. The 5-day average sits in the -3 to +3 bps band, scoring +10. The steady drift below the floor through the week is the funding-market signature of the TGA drawdown adding reserves, not a stress signal. |
| Historical significance | POSITIVE | URLI-Core (4 data components, 75% weight) = +53.00 ranks in the 91st percentile of weeks since 2020 (N=337). |
Market Transmission
| Asset | Bias | Reason |
|---|---|---|
| SPX / QQQ | SUPPORTED | Bullish liquidity backdrop; index-level liquidity supports multiple expansion. |
| Small caps / breadth | NEUTRAL | Rate-sensitive; needs reserve support and lower yields. |
| BTC / ETH | SUPPORTED | Crypto tracks net liquidity; this week's flow was +$112.2B. |
| High-beta alts | WATCH | Need both a positive URLI trend and BTC leadership to outperform. |
POSITIVE Supportive.
Bull / Base / Bear Scenarios
| Scenario | Liquidity Setup | Market Impact | Confirms / Invalidates |
|---|---|---|---|
| BULL | TGA drawdown resumes; SOMA support continues; yields cool. | Risk assets bid. | Confirms: net liquidity > +$50B. Invalidates: TGA rebuild accelerates. |
| BASE | Flat net liquidity; Treasury supply offsets SOMA. | Choppy, range-bound. | Confirms: stable reserves, calm SOFR. Invalidates: auction stress or USD breakout. |
| BEAR | TGA rebuild + bill supply + higher yields. | Risk-off. | Confirms: TGA spikes with rising yields. Invalidates: TGA drawdown and falling yields. |
Sources & Data Definitions
| Field | Source | Type |
|---|---|---|
| Balance sheet / reserves / RRP / TGA | Federal Reserve H.4.1 | Official |
| Verification series | FRED WALCL / WRBWFRBL / WDTGAL / WLRRAL | Official |
| Fed policy / rate path | FOMC statement & minutes | Official |
| Treasury outlook | Quarterly Refunding Statement | Official |
| Funding stress score | FRED SOFR / IORB — 5-business-day avg SOFR−IORB spread (bps) | Official |
| Dollar/yield score | FRED DTWEXBGS / DGS10 — weekly Δ broad dollar (%) × Δ 10Y yield (bps) | Official |
| Metric | Definition |
|---|---|
| Fed balance sheet | H.4.1 Total assets, weekly point-in-time level. |
| Bank reserves | H.4.1 Table 5 'Other deposits held by depository institutions' (weekly reserve balances). |
| TGA | U.S. Treasury, General Account. TGA down is liquidity positive. |
| RRP | Reverse repurchase agreements. RRP up parks cash at the Fed. |
| Basis | Latest week 2026-09-11; previous week 2026-09-04; H.4.1 levels in $M, displayed in $B. |