World GDP 2026F3.0%World CPI 2026F4.7%Brent$88/bblFed funds3.50-3.75%ECB deposit2.25%Data centre load485 TWhReadings as at 1 Aug 2026
QuantiaQuantitative economics
Insight · Macro

Disinflation has stalled: implications for the second half

The IMF has revised 2026 global inflation up three times running, from 3.8 to 4.4 to 4.7 percent. Three revisions in one direction is a signal about the model, not the shock.

Published 20 July 2026Quantia Economics

The IMF's July update puts global headline inflation at 4.7 percent for 2026. In January the same forecast was 3.8 percent. In April it was 4.4. Three revisions in the same direction within seven months is not a statement about the shock. It is a statement about the model.

ChartThree revisions, one directionIMF projection for world headline consumer price inflation in 2026, by publication round.

Sources: IMF World Economic Outlook Update, July 2026; reporting on the July update and earlier vintages.

Interpreting successive revisions

A forecast that is revised once has met new information. A forecast revised three times in one direction has a structural feature that keeps producing errors of the same sign. In this case there are two plausible candidates and they have different implications.

The first is that the shock itself kept getting larger. That is partly true: the energy disruption that began in late February was not in the January round at all, and its persistence was not fully in the April round.

The second is that the passthrough coefficient in use was too low. If a model consistently underestimates how much of a given input cost reaches consumer prices, it will produce exactly this pattern regardless of how the shock evolves. Distinguishing these matters, because the first implies the errors stop when the shock stabilises and the second implies they do not.

The core inflation signal

Core inflation forecasts have been broadly unchanged across the same revisions. That combination, headline moving and core stable, is the signature of a shock passing through mechanically rather than propagating into wage and price setting. The IMF's own language is that the world economy has weathered the shock better than feared, with limited evidence of second-round effects so far.

If core starts being revised up alongside headline, the picture changes materially and quickly. That is the single series we would watch through the second half.

Composition of the growth outturn

Global growth is projected at 3.0 percent for 2026 and 3.4 percent for 2027, against a 3.5 percent average across 2024 and 2025. The IMF describes the path as V-shaped and broadly unchanged cumulatively from April.

The composition is the uncomfortable part. The reason the aggregate holds up is that a technology-led investment cycle is offsetting an energy-led drag. Those are two independent processes that happen to be netting out at the world level. They do not net out at the country level: Korea was revised up to 2.6 percent on AI-related exports, while the euro area was cut to 0.9 percent because it captures little of the technology upside and carries most of the energy exposure.

An aggregate that is stable because two large opposing forces are offsetting is more fragile than an aggregate that is stable because nothing much is happening. Either force changing direction moves the total substantially.

Determinants of the second half

  • Whether the Hormuz reopening holds. The July forecast assumes a gradual reopening from mid-July. Oil has already round-tripped once. If the risk premium re-establishes above 90 dollars into the fourth quarter, the 2027 inflation path of 3.9 percent is too low.
  • Whether the capex cycle keeps accelerating. The technology offset works through the change in investment, not the level. A plateau removes the offset without any decline in spending.
  • Whether central banks tolerate an overshoot they cannot fix. The Fed held in July with three dissents preferring a hike, the first time since 2016 that three members dissented in the same direction. The ECB has already raised once. Neither has a costless option, and a committee that is visibly split is a committee whose reaction function is harder to forecast.

Our base case is that the mechanical passthrough completes without opening the second-round channel, and that headline inflation falls back through 2027 roughly as projected. The risk to that view is not the oil price. It is the possibility that the technology investment offset fades before the energy drag does, leaving weak growth and sticky prices at the same time.

Sources

All insights