MacroGuru
MacroGuru Decision Receipt · resolved

The Fed stays on hold — no July hike

The Fed leaves rates at 3.50%–3.75% on Wednesday. The hike risk is real later in 2026, not at this meeting.

Published Jul 24, 2026 Resolves Jul 29, 2026
94%published call

The frozen forecast

Polymarket prices roughly 93% no change, while June's FOMC projections and July's Monetary Policy Report still show inflation risk. The committee can stay hawkish without moving this week.

This is the immutable published call. It is not silently relabeled as a narrative-only, model-only, or market-only estimate.

Plain-language decision summary

Why it matters and what to watch

The probability is a frozen forecast, not a promise. The watchlist is conditional: it shows the assets most likely to react if the claim resolves as written.

Now what?

The decision is the easy call; the statement is the market mover. Watch whether Chair Warsh keeps a September hike live.

Track the public result → For wealth teams: check portfolio exposure →Open related scenario context →Context can support, oppose, or stress the forecast. It is not registered as the same claim.
Open the scientific audit trail Evidence timing, market comparison, model limits, divergence, hashes, and score contract
Evidenceretrospective passages verified not decision eligible
Narrativeabstained no prepublication directional passages
Marketcited point snapshot no executable interval
Modelnot independently measured
Divergenceobserved point gap not validated edge
Cascadelinked scenario and conditional watchlist
Scoreregistered

What changed—and what is actually evidenced

Evidence snapshot

retrospective passages verified not decision eligible

Short claim-to-span citations are frozen and lineage-clustered, but captured_at is after the forecast lock. They cannot be treated as evidence used by the original decision.

3 source references; verified retrospective passages: 3; original-decision effective independent sources: 0.

Evidence hash: 5cc1ad36c9fdbe13e5de9eaea957407ac8cd95f7763ebfd30ac733b384fe3e6e

Atomic claim and resolver

registered

TRUE if the July 28–29 FOMC statement leaves the target federal-funds range unchanged at 3.50%–3.75%; FALSE if the range is raised or lowered.

Resolver: Did the FOMC leave the target federal-funds range unchanged at 3.50%–3.75% in its July 29, 2026 policy statement?

Claim-to-span citations added after the lock

These claims are useful context today, but the stage receipt prevents them from being backdated into the original forecast.

  • Before the July meeting, the Fed reported that its 3.50%-3.75% target range had remained in place since the start of 2026.verified retrospective span · passage:fed-policy-range-maintained

Open the deterministic stage receipt → · Read the evidence-spine method →

Four lanes, no silent pooling

Narrative agreement

abstained no prepublication directional passages
Abstained
3 verified retrospective spans; 0 admitted before the forecast lock.

No passage-level agreement score is substituted from later evidence, headlines, source counts, market prices, or the published call.

The current source spans were frozen after the forecast lock, so they are published as retrospective citations but cannot influence or score the original decision. No narrative probability is inferred.

Statistical model

not independently measured
not measured

The published forecast is not relabeled as a statistical-model output. A separate frozen model receipt is unavailable for this claim.

The market

cited point snapshot no executable interval
93.0%
Frozen market input at publication · point only, not an executable quote.

Latest observed: 76.2%–76.4% at 2026-07-24T10:21:47.664000Z; 2 frozen observation(s). This later quote was not available to the original decision.

Open market history → · Compare all forecasts →

This is the point-in-time crowd probability cited by the frozen call. No bid/ask, depth, fee, or contract-correlation receipt is attached.

Combined estimate

withheld no performance learned weight
not measured
0 sensitivity points; none promoted as a learned production estimate.

No combined estimate is published until weights are learned on prior matched, prequential resolutions.

Where the estimates diverge

observed point gap not validated edge

The point difference is descriptive. It is not a learned trading edge and cannot establish disagreement after costs or shared evidence.

Published minus market: 1.0 probability points.

Latest frozen market comparison: 17.7 probability points; edge status: withheld shared input and unverified costs.

Open this shadow history → · Open the Forecast Divergence Lab →

Why it matters—and now what

Research decision

research monitor only

Now what? The decision is the easy call; the statement is the market mover. Watch whether Chair Warsh keeps a September hike live.

Activation: not registered. Invalidation: bound to false resolution not separately registered.

Scenario-conditional research framing only; not personalized investment advice.

How this forecast will be graded

Score contract

registered

Primary: brier. (published_probability - binary_outcome)^2

Benchmarks: no_skill_0.50, frozen_crowd_probability.

Open the public scorecard →