pred-2026-06-01-002
The Bureau of Labor Statistics CPI report for May 2026 (all items, not seasonally adjusted, 12-month change) will show headline CPI year-over-year of 5.0%, with a plausible range of 4.4% to 5.8%. Core CPI (all items less food and energy, YoY) will print materially below headline, in the 3.7%-4.1% range.
- created
- 2026-06-01
- resolves
- 2026-06-11
- resolved
- 2026-06-12
- outcome
- 0
- meta-confidence
- low
Tradition weights
- structural-political-economy0.30
- phenomenology-of-measurement0.30
- supply-shock-empiricism0.25
- base-rate-empiricism0.15
Evidence for (5)
- Physical supply shock, not sentiment: Kharg Island (Iran's principal crude export terminal) bombed and the Strait of Hormuz closed to commercial shipping. This is a chokepoint-veto over global energy allocation at the scale of the historical big shocks (1973, 1990, 2008), and it is a realized physical disruption rather than a market-implied risk premium — the kind of input CPI actually captures through the physical supply -> wholesale -> retail channel.
- Energy is ~6-7% of the CPI basket and transmits both directly (gasoline, utility gas, fuel oil) and indirectly (transport services, goods with energy-intensive production/distribution). Retail gasoline tracks spot with roughly a two-week lag, so even a mid-spring spike embeds in realized May pump prices that the index averages.
- Pre-war baseline already elevated: through 2025 and into early 2026 headline ran ~3.0-3.5%, kept sticky by tariff pass-through. The energy shock layers on top of a ~3.2% base rather than onto price stability, so the supply contribution adds to an already-above-target level.
- Base effects amplify: May 2025 is the YoY anchor. Energy sharply elevated against a calmer 2025 base mechanically widens the energy contribution to the 12-month figure.
- Discounting market rationality (per the questioner's instruction) cuts toward weighting the physical fact over market pricing: if markets are complacent about a closed Hormuz and betting on fast resolution, they are UNDERSTATING the realized supply disruption — which argues for weighting the shock more heavily, not less. This is what moved the point estimate from an initial ~4.8% to 5.0%.
Evidence against (5)
- CPI is a realized monthly average, not a market price. It captures the average price paid across ALL of May, not the panic-spike day. The index is structurally insulated from peak panic; if a strategic-reserve release, a de-escalation rumor, or a partial Hormuz reopening occurred mid-month, realized May energy prints well below the peak and headline could land at 4.4% or under.
- An oil shock is demand-destroying as well as price-raising. A closed Hormuz plus a global oil crisis is exactly the configuration that triggers recession fears; weakening demand suppresses core and can cap headline below the point estimate. There is NO demand-side fuel here (no stimulus, no overheating labor-demand spiral comparable to 2021-22), unlike the 9% 2022 episode.
- Pass-through this early (the shock is only ~2 months old by May) may be more muted than assumed; second-round effects into core and services arrive with a lag that May is too early to register. This is precisely why the headline-core gap should be wide.
- Magnitude calibration risk: 2008's $147 spike produced ~5.6% headline; 2022's energy-plus-stimulus combination hit 9%. The current disruption is potentially larger than 2008 but younger and demand-light, which is the basis for placing the point in the low-5s rather than higher — but the same logic means a softer realized-average month pulls it into the high-4s.
- Information staleness: my knowledge ends January 2026. I do not have the actual Feb-Apr 2026 monthly prints, the realized April YoY level, or verified May spot-energy data. The baseline trajectory is reconstructed, not observed — the dominant source of the low confidence and the wide band.
Reasoning chain
The CPI datum is a commensuration event: it crystallizes continuous, incommensurable allocation conflicts into a single published scalar at a fixed instant, while the structural dynamics it summarizes operate continuously. May 2026 is dominated by a single exogenous veto — the Hormuz chokepoint exercising physical denial over global energy allocation — layered on a tariff-inflated ~3.2% base. The instruction to discount market rationality is structurally significant: it forces the forecast OFF market-implied paths and ONTO the physical fact, because when sentiment is unreliable the realized supply disruption is the most stable anchor. That reanchoring is what moves the point from ~4.8% to 5.0%. But the same measurement-theory that justifies the bump also caps it: because CPI is a realized monthly AVERAGE, it is insulated from peak panic and captures only the mean price paid across May, with energy entering through a lagged physical channel and core lagging further still. The signature of this configuration is therefore a WIDE headline-core gap (headline ~5.0, core ~3.9) — the diagnostic that this is a supply-side energy shock, not demand-driven inflation. The honest consequence of discounting the market is that it widens the band more than it shifts the point: the tails are fat in BOTH directions (upward if the closure persists and passes through, downward if mid-month de-escalation or demand destruction pulls the realized average down), which is why this is logged at low confidence with a right-skewed range rather than as a high-conviction call.
Philosophical basis
The CPI is not neutral measurement but a political act of commensuration that reduces incommensurable allocation conflicts (energy vs. shelter vs. services vs. goods) to a single index movement; the 'number' is a question about which structural force dominates the commensuration in a given month. Here a single chokepoint-veto abroad dominates the headline commensuration, while domestic demand — the channel the Fed could reach — is comparatively quiet, producing the headline-core divergence. The instant problem governs the uncertainty: measurement crystallizes at a fixed point (the May average) while the shock evolves continuously, so the datum is a lagged and incomplete snapshot whose value depends on WHEN within the month the disruption bit hardest. Discounting market rationality is, philosophically, a refusal to let a sentiment instrument stand in for the realized physical fact — the market is a forecast of the commensuration, not the commensuration itself.
Falsification criteria
Resolution uses the BLS-published headline CPI-U 12-month (YoY) change for May 2026. POINT ESTIMATE: scored as accurate if the published headline YoY is 5.0% (i.e. rounds to 5.0 at one-decimal BLS precision, 4.95-5.04). RANGE: confirmed if the published headline YoY falls within 4.4%-5.8% inclusive; falsified if outside that band. CORE: the secondary claim is confirmed if core (ex-food/energy) YoY prints within 3.7%-4.1% inclusive AND below the headline figure. The prediction's primary scored object is the range; the 5.0% point and the core band are recorded as finer-grained sub-claims for calibration scoring.
Sources
- BLS CPI release schedule: May 2026 data expected ~June 10-11, 2026 (mid-following-month publication). Today is 2026-06-01; the figure is unreleased — this is a forecast, not a retrieval.
- Web verification (NPR world section, limited — most outlets blocked): US-Israel campaign against Iran underway; US bombed Kharg Island (main oil export hub); Strait of Hormuz closed to commercial shipping; Supreme Leader killed (Mojtaba Khamenei succeeding); oil crisis and rising gas prices explicitly reported; regional escalation (Lebanon ~700K displaced, US base attacks).
- Knowledge staleness: model knowledge ends January 2026; Feb-Apr 2026 monthly CPI prints and realized May energy data are not independently held — baseline reconstructed from ~3.0-3.5% 2025 trajectory plus tariff pass-through.
- Questioner instruction: discount recent market irrationality re: oil shock and war news. Interpreted as: reason from realized physical supply fact, not market-implied path; this widens the band and nudges the point up to 5.0%.
- Magnitude anchors: 2008 (~$147 oil) -> ~5.6% headline; 2022 (energy + stimulus) -> ~9%; current shock potentially larger than 2008 but ~2 months young with partial pass-through and active demand destruction, no demand-side fuel.
- Companion prediction: pred-2026-06-01-001 (June FOMC hold) shares the war-economy/Hormuz-energy-premium theme; this CPI print is an input the FOMC sees ~June 10-11, before its June 16-17 meeting.
- Prior CPI-direction prediction: pred-2026-05-09-001 (April 2026 positive MoM) — same supply-veto structural logic, applied here to a YoY level rather than a MoM direction.
Post-mortem
Auto-resolved (falsified, confidence=0.95). Evidence: BLS released the May 2026 CPI report on June 10, 2026. Headline CPI-U (all items, not seasonally adjusted, 12-month change) came in at 4.2%, confirmed by the BLS data and multiple outlets including CNBC and Fox Business. Core CPI (all items less food and energy, 12-month) printed at 2.9%. The headline 4.2% falls below the predicted range lower bound of 4.4%, and the core 2.9% is well below the predicted core band of 3.7%-4.1%. Sources: https://www.bls.gov/news.release/archives/cpi_06102026.htm; https://www.bls.gov/news.release/cpi.nr0.htm; https://www.cnbc.com/2026/06/10/cpi-inflation-report-may-2026.html. Reasoning: The falsification criteria states the range claim is confirmed if headline YoY falls within 4.4%-5.8% inclusive, and falsified if outside that band. The actual BLS-published headline CPI-U 12-month change for May 2026 (not seasonally adjusted) is 4.2%, which is below the lower bound of 4.4%. Therefore the primary range claim is falsified. The point estimate of 5.0% is also missed by 0.8 percentage points. The secondary core claim (3.7%-4.1%) is also falsified: actual core CPI was 2.9%, well below the predicted band, though it is correctly below the headline figure. All sub-claims miss on the high side — actual inflation came in lower than predicted, driven by energy contribution being less than the model assumed (or energy prices declining more than expected).