There is a marginal note in an old edition that says it all. In the IMF’s July World Economic Outlook update, the headline number for 2026 global growth is 3.0 percent — stable, steady, almost boringly reassuring. But flip a few pages forward in the same document and you find the forecast for global inflation being revised up from 4.1 percent in 2025 to 4.7 percent in 2026. The number that got the headlines, curiously, is the one the institution itself seems least sure about; the number buried deeper is the one doing the work.
Let me tell you the story in order. On September 2, at the G20 meetings, IMF Managing Director Kristalina Georgieva said the outlook for global growth in 2026 has become more solid, around 3 percent, and that is where the forecast expects it to stay. Reassuring words, delivered in the measured tone that the Fund’s leadership is paid to project. But she added the sentence that the markets should have underlined: the risk of an energy shock from a closure of the Strait of Hormuz remains persistently high. That sentence is the marginal note. Everything else in the press conference was the official text.
The numbers and the reasoning in the IMF’s own July update are worth reading carefully, because they frame what Georgieva actually meant. Global growth is forecast at 3.0 percent for 2026 — below the 3.5 percent average of 2024-25 — before recovering to 3.4 percent in 2027. Inflation, meanwhile, is expected to rise from 4.1 percent in 2025 to 4.7 percent in 2026, easing back to 3.9 percent in 2027. Read those two trajectories together and a quiet tension emerges: growth easing while inflation climbs is the exact combination that makes central banks uncomfortable, because the two forces pull policy in opposite directions.
Here is the detail that fascinates me as a reader of forecasts. The Fund lists re-escalation of conflict in the Middle East as the largest downside risk to its baseline. The same document, in the same breath, acknowledges that the impact of the war was partially offset by AI-driven global technology investment. So the baseline forecast of 3 percent is not a prediction about the world as it is; it is a prediction about the world as it would be if a specific large risk — the one the Fund itself names — keeps not happening. Curiously, that conditional quality is rarely printed in the news summary.
The energy channel deserves particular attention, and this is where I want to digress briefly, in the manner of a bookseller recommending a side shelf. The Strait of Hormuz carries a meaningful share of the world’s seaborne oil. A closure is not a scenario the forecast models as a footnote; it is the kind of event that rewrites global inflation in a single quarter. What makes the current situation distinctive is that the risk has been raised, repeatedly, by the institution itself — not by a commodity trader’s newsletter. When the IMF names a strait, the inflation forecast that follows is already leaning on the conservative side.
The World Bank, for its part, puts 2026 global growth at 2.5 to 2.6 percent, noticeably below the IMF’s 3.0. The two institutions rarely disagree by half a point without meaning something. The gap is not an arithmetic disagreement; it is a difference in how much weight each gives to the downside risks. Curiously, both reports point at the same fault line — energy and conflict — while disagreeing on the arithmetic.
The gap between the two institutions is worth pausing on, because it is not a footnote for economists. The IMF sees the AI investment offset as strong enough to hold growth at 3.0; the World Bank weights the energy and conflict channels more heavily and lands at 2.5 to 2.6. When two of the most careful forecasting shops in the world disagree by half a percentage point, the honest reading is not that one is wrong and one is right — it is that the range of plausible outcomes has widened, and the widest points of that range are defined by events that no econometric model can assign a stable probability to. A strait is not a variable in a spreadsheet.
Now let me correct something in my own reading, because I initially wanted to present the IMF’s 3 percent as the story. It is not. The story is the inflation revision. A forecast that moves inflation from 4.1 to 4.7 percent while growth drifts down is the Fund’s quiet way of saying that the largest variable in the next two years is not growth policy but the price of energy. Central bankers reading that page understand it instantly: the constraint on their rate decisions is no longer demand management, it is supply shocks they cannot vote on. Picture the analyst’s desk on the morning the update lands: the inflation column revised up, the growth column revised down, and a sticky note on the monitor with a single word — Hormuz.
For central bankers, the revision from 4.1 to 4.7 percent inflation is the uncomfortable part of this story. A central bank can respond to demand-led inflation with policy; it has almost no instrument for supply-side energy shocks, except to tighten and hope the shock does not spiral into expectations. With growth cooling at the same time, the classic dilemma sharpens: too much tightening and the 3.0 percent forecast becomes optimistic; too little, and the 4.7 percent inflation number starts to look like a floor. That is why the Fund’s inflation revision is the more consequential line in the document — it is the number that will constrain every rate decision for the next two years.
What makes this moment unusual in the history of these forecasts is the presence of a genuine offset. AI-driven technology investment is large enough to be visible in the global aggregate — the Fund cites it as a partial counterweight to the conflict shock. That is a rare admission, and a revealing one. It means the global economy is now being pulled by two forces moving in opposite directions: an energy shock from the Middle East and an investment boom from AI. The 3 percent number is the midpoint of those two forces. The interesting question is which one wins the next two years.
For the reader who wants the signal rather than the arithmetic, the practical translation is straightforward. Inflation at 4.7 percent in 2026 means real incomes stay under pressure for another year in most economies. Growth at 3.0 percent, below the 3.5 percent trend of the prior two years, means the labour market recovery slows. And a central bank that must defend its inflation target while growth cools has, as a practical matter, less room to cut rates than the markets would like. That is the message encoded in the revision from 4.1 to 4.7.
For an ordinary reader, the practical translation is a household one. Inflation at 4.7 percent does not arrive as a headline; it arrives as the monthly electricity bill, the supermarket basket, the cost of filling a car. Growth at 3.0 percent is a global average, which means some regions will do much better and others will feel stagnant. The honest advice a reader can take from this forecast is defensive: the global economy is being pulled by two large forces, and the balance between them is genuinely uncertain. Budgets built on stable prices and steady incomes should carry a margin for the scenario the Fund keeps naming.
I keep returning to the marginal note, because it fits the subject too well. Official forecasts are printed in neat columns; the interesting judgment is always in the margin, where someone has written the thing the headline cannot absorb. Georgieva’s sentence about Hormuz is exactly that kind of margin note. The baseline of 3 percent is the official text. The risk that keeps it honest is the strait. Marginalia of this kind is how the Fund has always left its true opinions — in the footnotes, the briefing remarks, the quiet asides that a headline writer would skip.
The uncomfortable thing about a forecast that rests on a single strait is how narrow the margin of safety has become. The world economy has been through supply shocks before, but usually with a thicker cushion of inventories, spare capacity and political alternatives. What the IMF is describing is a world in which the baseline itself depends on something not happening — and the thing is neither cyclical nor policy-controlled. That is a different category of risk from the usual forecast caveats. It is the difference between saying the path is bumpy and saying the path may not exist.
The forecast to watch, then, is not the growth line. It is the inflation line. If energy prices stay contained, the Fund’s 3 percent holds and 2027 brings the mild recovery the table suggests. If the strait scenario materialises, the 4.7 percent inflation number becomes an early draft of a worse table. Curiously, both outcomes are already printed in the same document — the Fund has, in effect, given us both versions of the future and asked us to choose which one we believe. The margin note tells you which one the authors believe.
Let me also admit the limit of this exercise. Forecasts are not facts; they are the current best judgment of institutions paid to be careful, written at a moment when the two largest variables — a war and an investment boom — are both in motion. The IMF could be wrong in either direction, and its own document quietly acknowledges as much by printing both the baseline and the risks. What the update is genuinely useful for is calibration: if you have a view on the next two years, this is the professional centre of gravity around which the arguments arrange themselves. That is more than most documents offer.
Booksellers have a habit that fits here: we keep the first editions not because they are scarce but because they are honest. The first edition of a forecast is the version written before the world argued with it, before the revisions and the caveats and the polite rewordings. The IMF’s July update is a first edition in that sense — an honest, dated snapshot of how the institution saw the world in the middle of 2026, with the risks printed in the same ink as the baseline. Years from now, when the outcome is known, this document will be read the way old margin notes are read: not for what it predicted, but for what it chose to worry about. It chose a strait.
Let me end where I began, with the bookseller’s instinct. A good first edition tells you how the author thought before the world agreed with him; a good forecast tells you the same thing about an institution. The IMF’s July update is not an exercise in optimism or pessimism — it is an exercise in honesty about uncertainty. Growth at 3.0, inflation rising to 4.7, a named downside risk, a named offset. That is a balanced account, and balanced accounts are rare in this business. The line I would underline is not the growth rate. It is the strait.