*Issue 18 · 2026-07-29 · the sections run in the same order every day, so the delta takes a minute to extract.* ## 1 · The reading The composite reads **45 out of 100, STRAIN**, +0.2 against the last published reading, on 100% coverage. The pipes are working harder for the same result. This is the regime where surprises stop being cheap. The heaviest hand on the dial is **the calendar squeeze (weather)** at a score of 100, worth +11.0 points of the total. Pinned at the ceiling today: the calendar squeeze (weather), the cash buffers, RRP and TGA (buffers). Those components are carrying no new information from day to day, and the letter names every one of them rather than only the loudest. When one component carries the reading, watch that component, not the headline. Change gets attributed here, not just level: against the last letter, the tail law (tails) +0.2, +0.2 points net across the components. ## 2 · What moved The tape moved on its latest prints, none of them from last night: **Total stablecoin circulation** printed 304.84 $B (level z -0.8, change z -3.8, as of 2026-07-29); **2y Treasury constant maturity yield** printed 4.31 % (level z +3.4, change z -0.5, as of 2026-07-27). One qualification on that list: the 2y Treasury constant maturity yield flag is on LEVEL, not on change. The series sits somewhere unusual in its own history and did not travel to get there today, which is worth knowing and is not the same claim as a move. Still flagged from prints already covered in an earlier letter: **Central bank liquidity swaps outstanding (H.4.1)** (|z| 16.5, as of 2026-07-22) [de minimis]. A standing flag is context, not news; when a fresh print lands it goes back in the movers line. The flags marked de minimis (Central bank liquidity swaps outstanding (H.4.1)) sit at large sigma on series whose levels are a rounding error against their own history: the standardizer is reacting to a near-zero baseline, not to a large flow, and the letter would rather say that every day than let a big number do work the dollars do not support. SOFR's 99th percentile cleared the SRF ceiling on 1 of the last 20 sessions for 5 bp days of leak while SRF take-up peaked at $0.1B against a $500B facility, stigma score 10 of 100. ## 3 · The Tell The Tell, the gap between what the pipes measure and what the screens price, reads **+12.8**: plumbing indicators at the 54.0th percentile of their own history, market indicators at the 41.0th, a gap of +13.0. The gap is modest. Modest gaps are what most days look like, and saying so is part of the record. ## 4 · Reserve scarcity The desk fits the reserve demand curve continuously, a public-data approximation of the NY Fed's monthly Reserve Demand Elasticity (the Afonso, Giannone, La Spada and Williams lineage). Theirs is estimated on confidential bank-level fed funds transactions; this one is a hinge fit on SOFR minus IORB against reserves over GDP, which is a different and coarser instrument reaching for the same quantity, and the comparison below is the check on whether it gets there. Today's fit puts the kink near **$3,634B** of reserves; the system holds $3,062B, $572B below the estimate. Fit honesty: R² 0.62 against a flat curve; the observed SOFR minus IORB spread of -3.7bp versus the fit's -2.1bp gives a consistency of 0.87. A modest fit is a lens, not a verdict, and the number is printed so you can discount it yourself. Reserves are through the estimated kink, which is exactly where the spread should start answering to reserve changes. From here the board watches the slope, not the distance. Against that, the tape disagrees: SOFR is printing 3.7bp BELOW IORB, which is the abundance signature, not the scarce one. The honest reading is that the fitted kink is an estimate with a wide band and the observed spread is a fact, so scarcity here is a hypothesis the tape has not yet confirmed. Reserves fell $80.6B on the week to $3,062B, and the ledger says where from: the TGA rebuilt $73.4B, the residual absorbed $14.5B, the balance sheet added $4.3B, six legs summing to the reserve change to the dollar with $32.0B of the balance sheet, 0.47% of assets, still sitting in the unnamed residual. External check: the NY Fed's latest official RDE print (2026-07-06) reads -0.27bp per one percent of reserves; the desk's continuous fit implies -0.32bp, inside their 68% band, direction agrees. The desk's fit runs 16 days ahead of their release cycle, which is the whole point of fitting it continuously. Where the two diverge, one of us is wrong, and the scorecard keeps count. The running record, not just today: across 18 of their releases the desk's walk-forward fit landed inside their 68% band 8 times and agreed on direction 10 times, mean absolute gap 0.27bp. That is short of the two-in-three a calibrated nowcast should manage, so the fit is directionally useful and not yet precise. One matching print is an anecdote; this line is the claim, and it is graded against their current-vintage history rather than the print that stood on each cutoff date, which flatters neither side in a way the desk can measure. ## 5 · The official sector Foreign officials cut their Fed custody Treasuries by $106B over 13 weeks while foreign RRP parking rose by $27.7B, only a partial rotation because the combined official footprint still fell $77.9B to $2,975B, 74% of the custody drop leaving the official sector rather than moving inside it. ## 6 · The dates that matter The next date that matters is **2026-07-31**: calendar pressure date while reserves sit below the estimated kink, worst case reserves near $3,016B after the drain. The turn model puts 2026-07-31 (month_end) at +4.8bp with a band of [+0.9, +8.8], severity 2/5. FOMC decides today, and the letter cares in three named places: the IORB and ON RRP settings, which are the corridor every spread on this board is priced against; the runoff pace, which sets the drain rate behind the reserve path; and any change to the SRF, the ceiling whose quiet take-up the tripwires watch. The next issue grades this stanza against the statement. The corporate tax date lands 2026-09-15, 48 days out; tax dates drain reserves on a schedule everyone can read. ## 7 · What the board is honest about Faults on the board today: fedtext, gdelt. The affected inputs are degraded or dead and the composite's coverage says so. A dashboard that hides its broken gauges is lying with a straight face. The record is in PROOF: event recall 62% (95% interval 36% to 82%, which is wide because the sample is 23 alert runs, not thousands) against a base rate of 4%, run precision 17%, and the median lead of 60 trading days is censored: 7 of 8 episodes were still flagged at the edge of the evaluation window, so the true lead is at least that and the number is a floor rather than a central estimate. The misses sit next to the hits; read those before weighting today's letter.