PROTOTYPE. Answering one question: if the weekly stays and a monthly is added, do they differ or are they ~the same report? W21 is real, audited data (matches the weekly prototype). W18–W20 are illustrative — constructed into a plausible May arc, anchored to the real print dates (US CPI jump dated 2026-05-12; Japan cooling 2026-05-22). Built to show format differentiation, not to assert those weeks' numbers.
Macro Intelligence · Cadence design · May 2026

One month, five reports — what's shared and what's not

Four weeklies (W18 → W21) and the May monthly, laid side by side.

Why they don't collapse into the same report

Macro hard-data prints monthly or quarterly; only yields and FX move daily. So if each report just re-narrated the regime object, four weeklies plus a monthly would be ~80% identical — your concern, and it's correct. The fix isn't faked variation. It's giving each cadence a different job:

The weekly

= the derivative (rate of change)
  • Job: what changed this week + what's due next week.
  • Driven by: the event calendar. A quiet week is honestly short; an event week is heavy.
  • Disposable: read it, act on the delta, discard. Comparable week-to-week because the skeleton repeats.
  • Never carries: trajectory calls, deep-dives, scenarios.

The monthly

= the integral (level + trajectory) + a theme
  • Job: where are we in the cycle, and has the regime itself turned?
  • Driven by: the synthesis of 4 weeks into one arc — a shape no single week could see.
  • Substantial always: even a quiet month has a trajectory and a special-focus chapter.
  • Only place for: regime-shift calls, one deep-dive, a month-ahead scenario tree.
The honest overlap: about half of every report is shared — the 4-bloc standing-regime backbone, the data table, the concept library. Those are legitimately slow-moving, so they're shown collapsed/carried-over, not re-narrated each time. The differentiation lives in the lead content: the weekly leads with the delta, the monthly leads with the trajectory and a theme. See the matrix below — it makes "shared vs. unique" literal.

The four weeklies of May, at a glancethe deltas

Same skeleton every week (that's the feature — comparability). What differs is weight and lead content, pulled by what actually printed.

W18 ILLUS.
Apr 27 – May 3
Light
"Holding pattern — April regimes carry; nothing printed."
Firednothing — explicitly a quiet week
New vs lastStanding picture unchanged. Report is 3 paragraphs, not 12. The honest short weekly.
WatchMid-month US CPI; BOJ tone.
W19 ILLUS.
May 4 – 10
Light–med
"Calm before the print — positioning into a loaded calendar."
Firednothing new
New vs lastForward-leaning: flags CPI-print risk and that China's near-zero CPI leaves easing room (M11 dormant-watch).
WatchUS CPI (now days away); China LPR.
W20 ILLUS.
May 11 – 17
Heavy
"The inflation pop — US CPI jumps to 3.8%, soft-landing gets its first crack."
FiredUS upside-inflation surprise (headline-led)
New vs lastBig. Leads with the US print (real, as-of 2026-05-12). The eventful week — heavy report, full mechanism walkthrough.
WatchDoes core PCE confirm? Supply or demand?
W21 REAL DATA
May 18 – 24
Med–heavy
"Disinflation everywhere except the US."
FiredJapan M2 (disinflation); US M13 (factory trough)
New vs lastDigests the W20 shock; the divergence theme emerges. → full weekly prototype.
WatchJune ECB; yen at 159; China LPR.

Even the four weeklies aren't "heavily the same": W18 is a 3-paragraph quiet note, W20 is a heavy single-event report, W21 is a multi-mechanism divergence read. The skeleton repeats; the lead content and weight don't.

What carries over vs. what's uniquethe answer

★ new / signature content ✓ present, refreshed this issue ↻ carried over (slow / shared) — absent
Report elementW18W19W20W21MONTHLY
Standing-regime backbone4-bloc growth×inflation×policy ↻↻↻↻↻
Hard-data table (L3)values mostly static week-to-week ✓✓✓✓✓
Concept libraryevergreen, identical every issue ↻↻↻↻↻
Active change-mechanismthe week's lead — varies most ——★★✓
"What to watch" triggersthe weekly's signature ✓✓✓✓✓
Cross-currents synthesislight weekly / developed monthly ——✓✓★
Trajectory / regime-shift callhas the cycle turned? ————★
Special-focus chapterone deep-dive theme ————★
Month-ahead scenario treebase / risk paths with probabilities ————★

Read the bottom three rows: trajectory, special focus, and scenarios are monthly-only — that's the differentiation, and it's exactly the IMF-WEO / GS-quarterly content a weekly has no business carrying. The top three rows are the shared backbone (collapsed, not re-narrated). The middle rows are where weeklies differ from each other.

The May 2026 monthly — in fullthe integral

Macro Intelligence · Monthly outlook · May 2026

From "synchronized soft landing" to a US that's decoupling

May began with the April consensus intact — a world gliding toward a synchronized soft landing, nobody tightening. It did not end there. Mid-month, US headline inflation re-accelerated to 3.8%; by month-end, disinflation had broadened across the rest of the world (Japan's core CPI to 1.1%, China pinned at 0.0%) while the US stayed hot. The month's signature is not any single print — it's the divergence that opened up between a late-cycle US and a disinflating everywhere-else, and what it does to the dollar, the yen carry, and the Fed's room to hold.

The month's arc — what the four weeklies added up to

W18
Quiet. April regimes hold. Consensus: soft landing intact.
W19
Still calm; positioning into a loaded calendar.
W20
US CPI 3.8%. The first crack. Headline-led, core unconfirmed.
W21
Japan & China disinflate further. The divergence is now the story.

No single weekly could call this — each saw one delta. Stacked, they trace a regime under strain: the soft-landing thesis required inflation to keep falling everywhere; instead it fell everywhere except the one economy that anchors global rates.

Regime-shift call (the monthly-only judgment):
synchronized disinflation / soft landing → US late-cycle decoupling watch

Not yet confirmed — it hinges on whether the US re-acceleration is supply (reverts) or demand (sticks). That decomposition is this month's deep-dive.

★ Special focus

Supply or demand? Decomposing the US inflation re-acceleration

The whole regime call rests on one question a weekly can't properly host: is the jump from 3.3% to 3.8% the kind the Fed reacts to, or the kind it looks through? Three pieces of evidence point to supply-side:

  • The news backdrop: Bloomberg tied sagging global factory activity to a third month of a "war-induced energy crunch" — an energy/supply shock lifts headline more than core.
  • Core lagging headline: headline jumped 0.5pp but core PCE ticked only 2.8%→2.9%. A demand-driven overheating shows up in core; this is concentrated in the volatile components.
  • The labour market isn't tightening: unemployment edged up to 4.2% — not the wage-price spiral a demand re-acceleration would bring.

The counter-case: GDP at 3.3% is above trend, and services PMI is strong (52.0) — genuine demand that could keep inflation sticky even after energy fades.

Why this single distinction decides the Fed's June path
A supply shock raises prices and dampens activity — tightening into it would deepen the slowdown, so the Fed prefers to look through it. A demand-driven overheating is what rate hikes are for. Same 3.8% print, opposite policy response. This is why "is it supply or demand" — not the headline number — is the variable that matters, and why the monthly spends a chapter on it while the weekly just flags "watch core PCE."

Month-ahead scenarios★ monthly-only

Base — supply, reverts

~55%

Energy-driven CPI eases as the crunch fades; core stays contained; Fed holds at 4.5%. The global disinflation tailwind (Japan/China) dominates.

durationgrowth equityUSD fades

Inflation sticks — demand

~25%

Core PCE follows headline up; the wage/services side confirms. Fed back in play; "higher for longer" re-prices. Late-cycle US trade dominates.

long durationEM / carryUSD, value, banks

Stagflation-lite

~20%

The W21 factory trough (M13) is a head-fake — PMI rolls back under and falls (→ M6) while inflation stays sticky. The worst quadrant: weak growth, no easing room.

defensivesgoldcyclicals, small-caps

What to watch — month-ahead

  • USNext core PCE is the swing data — confirms or kills the demand case. Plus: does mfg PMI hold the trough or roll back (M13→M6)?
  • EuroJune ECB meeting — Stournaras's credibility-hike talk becomes the world's only tightening impulse if acted on.
  • ChinaAny LPR cut fires M11 (credit impulse) — the cleanest tailwind for the EM/commodity complex; CPI at 0.0% keeps deflation-watch live.
  • JapanYen at 159 + further disinflation — slower BOJ plus higher-for-longer Fed keeps carry-unwind / intervention (M12) risk elevated.

Below this would sit the same standing-regime backbone, data table, and concept library as the weeklies — carried over, collapsed by default. Omitted here to keep the contrast clean.