Public Kempax report

Weekly Dollar Liquidity Report · 2026-10-24

Kempax Research

Loop report · 2026-10-24

Weekly Dollar Liquidity Report — Week of Oct. 19–23, 2026

As of: 2026-10-25 UTC date-level cutoff · Latest complete market close: Oct. 23, 2026

Liquidity stance: Watch — mixed with a modest tightening bias. Treasury cash rebuilt during the measured week and the H.4.1 weekly-average reserve balance fell $59.533B while the Fed asset side was broadly stable. Marketable Treasury issuance exceeded redemptions by $81.906B in the measured Oct. 15–21 window. ON RRP accepted flows were very small, SOFR and EFFR remained within the observed policy range, and standing-repo usage was negligible in the latest checked operations. Credit and market signals were mixed: HY OAS widened 9bp , while VIX fell 4.8% , $SPY declined 0.43% , and $HYG $LQD rose modestly. The evidence supports selective caution around Treasury cash rebuilding and credit repricing, not a confirmed broad dollar-funding break or a synthetic net-liquidity estimate.

Executive summary

The Fed balance sheet was broadly stable, but reserve averages declined. Total assets fell $2.035B on the latest Wednesday observation, reserve-bank credit fell $0.789B on a weekly-average basis, and securities held outright rose $0.492B . Reserve balances averaged $2,962.607B , down $59.533B week over week H.4.1 Table 1, Fed H.4.1 .
Treasury cash rebuilding was the clearest measured drag. The H.4.1 TGA average rose $68.629B to $952.002B . DTS closing cash rose from $885.823B on Oct. 14 to $945.153B on Oct. 21 and $961.759B on Oct. 22 ; the source does not attribute the move to a particular receipt, auction, redemption, or outlay Treasury DTS .
Reverse-repo usage was low, but stocks and flows remain separate measures. Aggregate H.4.1 reverse-repo liabilities fell $4.953B to $327.238B , while New York Fed accepted amounts averaged only $269M per business day on Oct. 19–23 versus $444.25M in the prior inspected window. The latest accepted amount was $200M on Oct. 23 ; the liability stock remains dominated by foreign official international accounts New York Fed propositions, RRPONTSYD .
Treasury financing was net positive in the measured cash window. Marketable issuance was $720.336B , redemptions were $638.430B , and net marketable issuance was +$81.906B from the Oct. 14 to Oct. 21 MTD differences. Debt held by the public rose $82.635B over the same dates; this does not close the reserve effect Treasury public-debt transactions, Debt to the Penny .
Overnight funding remained orderly in the inspected observations. SOFR was 3.90% with $2.945T volume on Oct. 22; EFFR was 3.88% with $119B volume and remained inside the 3.75%–4.00% target range. SOFR ranged from 3.87% to 3.90% on Oct. 19–22, with no observed standing-repo pressure in the latest checked operations NY Fed SOFR, NY Fed EFFR .
Credit and volatility gave mixed signals. HY OAS widened from 3.31% to 3.40% between Oct. 16 and Oct. 22, while Chicago Fed NFCI remained negative at −0.496 for the week ended Oct. 16 and VIX fell from 19.99 to 19.03 . The combination is consistent with modest credit-specific repricing, not a broad deterioration in the measured financial-conditions indexes HY OAS, NFCI, VIX .
Global-dollar evidence is constructive but lagged. BIS USD credit to non-bank borrowers outside the United States rose $494B from 2025-Q4 to 2026-Q1, or about 3.5% q q and 7.3% y y ; BIS cross-border bank USD credit rose $793B in Q1 2026, up 12% y y . No current-week cross-currency-basis observation was verified, so these quarterly measures do not establish current funding conditions BIS global-liquidity table, BIS Q1 release .
Cross-market transmission was selective rather than disorderly. Treasury yields fell 4bp, 5bp, 9bp, and 10bp across the 3-month, 2-year, 10-year, and 30-year tenors from Oct. 16 to Oct. 23; 2s10s narrowed 4bp to 31bp . $SPY fell 0.43% , $UUP rose 0.14% , and $HYG $LQD rose 0.25% $0.44% . These prices are transmission checks, not proof of a single liquidity cause Treasury daily rates, market-history references .

Bottom line: the current evidence supports a watch mixed stance with a modest tightening bias. A higher TGA and lower reserve average are headwinds; lower ON RRP take-up, stable Fed holdings, and orderly overnight pricing are offsets. Wider HY OAS warrants monitoring, but negative NFCI, lower VIX, and modestly higher liquid-credit ETF prices do not confirm broad funding stress. Human review is required before any investment decision.

Charts

Current TGA daily-path chart unavailable; prior observed path retained for context

Current overnight daily-path chart unavailable; prior observed path retained for context

Observed $SPY, $HYG, and $LQD returns were modest; other bars are historical

Detailed findings

1. Federal Reserve balance sheet H.4.1 release Oct. 22 — week ending Oct. 21

Observation: The latest Wednesday total-assets stock was $6,753.604B , down $2.035B from $6,755.639B . H.4.1 week-average reserve-bank credit was $6,706.861B , down $0.789B ; securities held outright were $6,470.860B , up $0.492B . Within securities, Treasury holdings rose $1.337B and agency MBS fell $0.845B . Fed loans averaged $8.968B , down $2.382B H.4.1 Table 1, Fed total-assets preview .

Inference: The asset side was close to flat over the reported comparison, so the lower reserve average is not explained by a large change in securities held outright. The accepted evidence does not identify transaction-level causation, and no synthetic net-liquidity total is calculated.

2. Treasury General Account

Observation: The H.4.1 weekly-average TGA rose $68.629B to $952.002B from $883.373B . DTS closing cash was $885.823B on Oct. 14 , $945.153B on Oct. 21 , and $961.759B on Oct. 22 . The Oct. 15–21 DTS window recorded $862.067B of deposits and $802.738B of withdrawals, for a +$59.329B net cash change; Oct. 22 alone recorded a +$16.606B net change TGA closing-balance response%20Closing%20Balance,record date:gte:2026-10-08&fields=record date,open today bal&sort=record date&page%5Bsize%5D=20 , Treasury DTS dataset .

Inference: A higher TGA balance is mechanically consistent with reserve absorption, making Treasury cash rebuilding the clearest tightening pressure in the current plumbing observations. The DTS arithmetic does not by itself identify how auctions, tax receipts, redemptions, or outlays generated the move.

3. Reverse repo facility

Observation: Aggregate H.4.1 reverse-repo liabilities averaged $327.238B , down $4.953B from $332.191B . The latest week was $326.910B in foreign official international accounts and $0.328B in other accounts. New York Fed accepted amounts were $330M, $227M, $237M, $351M, and $200M on Oct. 19–23, averaging $269M , versus $444.25M in the Oct. 13–16 inspected window. The latest FRED RRP observation was $0.200B on Oct. 23 . Standing-repo results checked $0 on Oct. 23 and $1M in the Oct. 22 afternoon operation current propositions, prior propositions, repo results .

Inference: Low accepted ON RRP flows and a slightly lower liability stock remove one potential source of reserve pressure, but accepted transactions and the H.4.1 liability stock are different measures. The foreign-official composition limits a simple domestic-liquidity interpretation.

4. Bank reserves

Observation: Reserve balances averaged $2,962.607B in the week ended Oct. 21, down $59.533B or 1.97% from $3,022.140B . This is a weekly-average measure and should not be compared one-for-one with a daily TGA close or an accepted facility flow reserve-balance series, H.4.1 current release .

Inference: The reserve-average decline reinforces the TGA-drag signal, but the evidence does not close a one-week balance-sheet reconciliation across Treasury cash, debt settlement, ON RRP, and Fed operations. The report therefore treats the direction as a watch signal rather than a precision estimate of available reserves.

5. Treasury issuance, auctions, and the buyback program

Observation: Marketable Treasury issuance increased $720.336B and redemptions increased $638.430B from the Oct. 14 to Oct. 21 MTD readings, producing +$81.906B of net marketable issuance. Debt held by the public rose from $32.483712T on Oct. 14 to $32.566347T on Oct. 21 , or +$82.635B ; total public debt rose +$84.409B . On Oct. 22, debt held by the public was $32.597664T and total public debt was $40.444343T Oct. 14 issues, Oct. 21 issues, debt totals .

Inference: Net financing was positive during the measured window and is directionally consistent with Treasury cash rebuilding, but gross issuance, redemptions, DTS cash, and reserve balances are different measures. No auction-level attribution or buyback-specific observation is asserted from the accepted evidence.

6. Money markets

Observation: SOFR was 3.90% with $2.945T volume on Oct. 22, while EFFR was 3.88% with $119B volume. SOFR ranged from 3.87% to 3.90% on Oct. 19–22; EFFR was 3.88% on Oct. 16 and 3.88% on Oct. 22 . The stated target range was 3.75%–4.00% . The latest checked standing-repo operations showed negligible take-up NY Fed SOFR, NY Fed EFFR, repo results .

Inference: The inspected rates and facility usage show no obvious secured- or unsecured-funding dislocation. The daily panel is still partial relative to the Oct. 25 cutoff, so orderly pricing through Oct. 22 is not proof that no later pressure occurred.

7. Credit conditions

Observation: ICE BofA U.S. High Yield OAS widened from 3.31% on Oct. 16 to 3.40% on Oct. 22 , a 9bp increase. Chicago Fed NFCI was −0.496 and ANFCI −0.518 for the week ended Oct. 16; the NFCI was reported unchanged week over week at rounded −0.50 , with negative values indicating looser-than-average conditions. VIX fell from 19.99 on Oct. 16 to 19.03 on Oct. 22 , or −0.96 points −4.8% . No current CP or H.8 observation is asserted because it is not in the accepted evidence HY OAS, NFCI, VIX .

Inference: HY spread widening alongside negative NFCI and lower VIX is consistent with a modest credit-specific repricing rather than broad financial stress. The source set does not identify whether the spread move reflected supply, liquidity, or risk-premium changes.

8. Global dollar funding FX

Observation: The official nominal broad-dollar index was 121.4048 on Oct. 16 ; no Oct. 23 official observation was available. The $UUP proxy rose from 28.92 to 28.96 , or +0.14% , between Oct. 16 and Oct. 23. BIS USD credit to non-bank borrowers outside the United States rose from $14,254B in 2025-Q4 to $14,748B in 2026-Q1 , or +$494B about +3.5% q q and +7.3% y y . BIS cross-border bank USD credit rose $793B in Q1 2026 , or +12% y y broad-dollar series, $UUP history, BIS global-liquidity table, BIS Q1 release .

Inference: The lagged BIS observations show expansion in covered dollar credit, not a contemporaneous global-funding squeeze. No current-week cross-currency basis, USD JPY or GBP USD basis, offshore deposit flow, or official swap-use series was verified, so the global read remains incomplete and should not be broadened from the $UUP proxy.

9. Cross-market impact assessment

Close-to-close market and rate evidence for Oct. 16–23, 2026 shows:

Equities: $SPY fell from 763.25 to 759.97 , or −0.43% . A current $QQQ comparison was not supplied; no current growth-proxy return is inferred $SPY history, $QQQ history .
Bonds: Treasury par yields fell from 4.26% to 4.22% at 3 months, 4.74% to 4.69% at 2 years, 5.09% to 5.00% at 10 years, and 5.43% to 5.33% at 30 years. 2s10s narrowed from 35bp to 31bp . The current $TLT return was not supplied. Inference: the bull-flattening move is compatible with duration demand, policy-expectation, or term-premium repricing, but the evidence does not identify the cause Treasury daily rates .
U.S. dollar: $UUP rose 0.14% . This is an ETF proxy, and the official broad-dollar index has no Oct. 23 observation in the accepted evidence $UUP history .
Credit and risk appetite: $HYG rose from 76.86 to 77.05 , or +0.25% , and $LQD rose from 103.40 to 103.85 , or +0.44% . These are ETF prices rather than spread measures. VIX fell 4.8% , even as HY OAS widened 9bp . Inference: the tape supports a selective-credit watch rather than a broad volatility shock $HYG history, $LQD history, HY OAS .

The conditional implication is mixed: Treasury cash rebuilding and lower reserve averages are headwinds; lower ON RRP take-up and orderly overnight pricing are offsets; the higher HY OAS is a targeted caution. The evidence supports monitoring a Treasury-cash credit-repricing channel, but it does not establish one liquidity cause or authorize a trade.

Unresolved risks

Release-lag gap: H.4.1 is current through the week ended Oct. 21; DTS through Oct. 22; rates through Oct. 22; repo RRP checks through Oct. 23; HY OAS and VIX through Oct. 22; NFCI ANFCI through the week ended Oct. 16; the official broad-dollar index through Oct. 16; and BIS global-liquidity data through 2026-Q1. No later values are inferred.
Basis and reconciliation risk: H.4.1 weekly averages, Wednesday stocks, daily DTS cash, MTD debt differences, and facility flows are not interchangeable. The evidence does not close a transaction-level reserve reconciliation or support a synthetic net-liquidity number.
Treasury attribution risk: Net marketable issuance and higher TGA are observed together, but the accepted evidence does not isolate auction settlement, tax receipts, redemptions, outlays, holder type, or settlement timing as the cause of the reserve move.
Credit and global-funding coverage: Current CP, H.8, cross-currency basis, USD JPY or GBP USD basis, offshore dollar-flow, and official swap-use observations are absent. HY OAS is a spread index; $HYG $LQD are price ETFs; $UUP is a dollar proxy. They should not be merged into one stress measure.
Market-data gaps and revisions: Current $QQQ and $TLT comparisons are unavailable, and the preserved historical chart bars are explicitly not current returns. Treasury, Federal Reserve, BIS, and market-series observations may be revised. Human approval remains required before any investment decision.

Next-week watch

Refresh the next DTS release after Oct. 26 and the next H.4.1 observation; compare the TGA close, reserve average, and reserve stock without mixing their bases.
Watch SOFR, EFFR, ON RRP, and standing repo for a rate volume change that would confirm or reject the current orderly-funding read.
Refresh HY IG CCC OAS, NFCI ANFCI, VIX, and commercial-paper data as they become available; separate spread, volatility, and quantity signals.
Monitor $SPY, $HYG, $LQD, and $UUP , and seek current $QQQ $TLT observations and a dated cross-currency-basis measure before broadening the global-dollar interpretation.

Sources index primary

Federal Reserve: H.4.1 current release; H.4.1 Table 1 via FRED; total-assets DDP preview; reserve balances; TGA series.
Treasury: Daily Treasury Statement dataset; TGA closing-balance response%20Closing%20Balance,record date:gte:2026-10-08&fields=record date,open today bal&sort=record date&page%5Bsize%5D=20 ; TGA deposits response,record date:gte:2026-10-08&fields=record date,open today bal&sort=record date&page%5Bsize%5D=20 ; TGA withdrawals response%20 - ,record date:gte:2026-10-08&fields=record date,open today bal&sort=record date&page%5Bsize%5D=20 .
Treasury debt: public-debt transactions; Debt to the Penny; daily par yields.
New York Fed: SOFR; EFFR; current reverse-repo propositions; prior reverse-repo propositions; repo results.
Credit and financial conditions: HY OAS; NFCI; ANFCI; Chicago Fed current-data methodology; VIX.
Dollar and global liquidity: broad-dollar series; BIS USD global-liquidity table; BIS global-liquidity overview; BIS Q1 2026 statistical release.
Market-history references: $SPY · $QQQ · $TLT · $UUP · $HYG · $LQD · $VIXY.

The following prior-period links remain in the register for continuity; they are source references only and do not assert current-week observations.

Federal Reserve H.4.1 Sep. 24, 2026 · reserve balances · reverse-repo liabilities · TGA weekly series
Treasury Daily Treasury Statement · Oct. 1 operating-cash query · Oct. 2 gap query
Treasury securities auctions · upcoming auctions · August refunding statement
New York Fed reference rates · SOFR · ON RRP · standing repo
Federal Reserve H.15 · commercial paper · H.8 Sep. 25
ICE BofA credit spreads, Oct. 1 · Chicago Fed NFCI · St. Louis Fed Financial Stress
CME EUR USD basis · Fed central-bank liquidity swaps · BIS dollar credit
Treasury Sep. 18 curve · Treasury Sep. 25 curve · Cboe VIX
\$SPY price history · \$QQQ price history · \$TLT price history · \$UUP price history
\$HYG price history · \$LQD price history · \$VIXY history this week's observation unavailable
Federal Reserve H.4.1 Sep. 10, 2026 · H.4.1 Sep. 3, 2026 · reserve balances
Treasury Daily Treasury Statement · Treasury securities auctions · TreasuryDirect results
Treasury buyback operation, Sep. 10 · Treasury borrowing estimates · August refunding statement
New York Fed reference rates · repo operations · swap operations
FRED HY OAS · FRED RRPONTSYD · FRED financial stress · Chicago Fed NFCI
CME EUR USD basis · CME FX Swap Rate Monitor · NY Fed swap operations
Treasury daily curve rates · Sep. 11 curve display
Market-history references: $SPY · $QQQ · $TLT · $UUP · $HYG · $LQD · $VIXY
Federal Reserve H.4.1 Sep. 17, 2026 · H.4.1 Sep. 10, 2026 · reserve balances · reverse-repo liabilities
Treasury Daily Treasury Statement · Treasury securities auctions · public-debt transactions · buybacks
Treasury liquidity-support buyback policy · Treasury borrowing estimates · August refunding statement
New York Fed reference rates · SOFR · EFFR · ON RRP · repo operations · swap operations
FRED HY OAS · FRED IG OAS · FRED commercial paper · FRED financial stress · Chicago Fed NFCI
CME EUR USD basis · Fed SWPT · BIS global liquidity · BIS USD credit
Treasury Sep. 11 curve · Treasury Sep. 18 curve · Cboe VIX
Market-history references: $SPY · $QQQ · $TLT · $UUP · $HYG · $LQD · $VIXY this week's observation unavailable

<sub Report details — Coverage: Oct. 19–23, 2026 latest complete market close Oct. 23 · Data as of: 2026-10-25 UTC; H.4.1 week ended Oct. 21, Treasury DTS through Oct. 22, overnight rates through Oct. 22 with repo RRP checks through Oct. 23, HY OAS and VIX through Oct. 22, market closes through Oct. 23 where available, official broad-dollar index through Oct. 16, BIS through 2026-Q1 · Status: research complete; candidate for human review · Owner: research lead portfolio managers · Run: looprun-95d646c0-282c-4ccd-b9b4-994e91d547e6 · Session: loopsession-6cc8cfe2-c7ca-4b07-88ad-dd1168168d41 · Closure: report candidate written; no trading, scheduling, external communication, or publication action was taken · Next step: refresh the dated gaps and reassess persistence of the TGA drag, credit-spread move, and funding indicators before any investment decision.< sub