Macroeconomics in 2026
Prompt: Write an explainer on the macroeconomics of the world in 2026. Cover growth, inflation, the central-bank easing cycle, the AI capex boom, debt and deficits, trade fragmentation, demographics and energy, with diagrams and tables. Concise and engaging, UK English.
Inflation is mostly tamed, central banks are cutting, the United States is borrowing on a scale that used to be reserved for wars, China is exporting deflation, and a handful of companies are spending more on AI hardware than entire countries spend on defence. A guided tour of the forces shaping the 2026 economy.
- The 2026 snapshot
- From spike to soft landing
- The easing cycle, unevenly
- The AI capex boom
- Debt, deficits and the term premium
- Fragmentation: tariffs, friend-shoring, reshoring
- China: deflation and the property hangover
- Demographics: the slow squeeze
- Energy: cheap electrons, expensive molecules
- What could break it
1. The 2026 snapshot
The world economy in 2026 is calmer than the one we had in 2022 or 2023, but it is not normal. Inflation is back near target in most advanced economies. Central banks have started cutting rates, though more slowly than markets hoped a year ago. Growth is positive almost everywhere, but the United States is pulling away from Europe again, China is still working through a property bust, and a single industrial story, the build-out of AI compute, is moving the dials on global capex, electricity demand, and trade balances.
| Region | Real GDP growth | CPI inflation | Policy rate | Story in one line |
|---|---|---|---|---|
| United States | ~2.1% | ~2.6% | 3.75–4.00% | AI capex and big deficits keep demand hot |
| Euro area | ~1.1% | ~2.0% | 2.00% | Cutting faster than the Fed, growth still soft |
| United Kingdom | ~1.2% | ~2.4% | 3.75% | Services inflation sticky, fiscal headroom thin |
| Japan | ~0.9% | ~2.1% | 0.75% | Tightening, finally, as wage growth holds |
| China | ~4.5% | ~0.6% | 1-yr LPR ~2.9% | Industrial overcapacity vs. weak consumer |
| India | ~6.6% | ~4.2% | 5.50% | The standout. Capex, demographics, services exports |
Order-of-magnitude figures for the start of 2026. The point is the pattern, not the decimal places.
2. From spike to soft landing
The 2021–23 inflation shock was the largest in advanced economies since the 1970s and ended much faster than the 1970s analogue. Three things did most of the work: goods supply chains unsnarled, energy prices fell back, and central banks raised rates fast enough to anchor expectations without triggering a deep recession. By early 2026 headline CPI in most rich economies sits within a percentage point of the 2% target. The piece that remains sticky is services inflation, driven by wages.
Why the soft landing actually happened
- Goods disinflation arrived on time. Once container freight, semiconductor lead times and shipping insurance normalised, the price level of tradeable goods fell back quickly.
- Energy reversed. Oil and natural gas spiked on the 2022 Russia–Ukraine shock and unwound as Europe rerouted gas supply and built LNG capacity.
- Expectations stayed anchored. Survey and market-based measures of long-run inflation never properly de-anchored, which is what the 1970s lesson said to watch for.
- Real wage catch-up was slow but not explosive. Workers got back to roughly pre-pandemic real wages by 2025 without a 1970s-style wage-price spiral.
The sticky bit: services
Goods inflation is near zero in most economies. Energy is flat. The remaining gap to target is almost entirely core services excluding housing, which tracks wages closely. This is why central banks remain a little cautious about cutting too fast, even though headline inflation looks tame.
3. The easing cycle, unevenly
Central banks have begun cutting, but the pace and the destination differ. The European Central Bank has cut further and faster than the Federal Reserve because euro-area growth is weaker. The Bank of England sits in between. The Bank of Japan is, unusually, going the other way: gentle hikes to normalise after decades at zero. Real rates (nominal rates minus expected inflation) are still positive almost everywhere, which is why monetary policy is described as still restrictive even after the cuts.
What is the neutral rate?
The new question, occupying most macro debate, is whether the neutral rate, the level of policy rates that neither stimulates nor restrains the economy, has risen since the 2010s. Two reasons it might have:
- Higher fiscal deficits mean governments are competing harder for savings, pushing real rates up.
- The AI capex boom is absorbing enormous quantities of capital, which raises the marginal return on investment.
If neutral is indeed 1 to 1.5 percentage points higher than it was pre-pandemic, then "back to normal" monetary policy is materially tighter in absolute terms than the 2010s baseline. Bond markets seem to be pricing about half of this view.
4. The AI capex boom
The single largest microeconomic story driving the 2026 macro picture is the build-out of AI compute. A handful of US hyperscalers, plus a long tail of model labs, sovereign data-centre projects and electricity utilities, are spending at a scale that registers in GDP. US capital expenditure on information processing equipment and intellectual property products has more than doubled since 2022, and AI now accounts for an outsized share of US business investment.
How AI capex shows up in macro data
| Channel | Mechanism | Net effect in 2026 |
|---|---|---|
| Business investment | Servers, GPUs, data-centre buildings count as fixed investment. | Adds roughly 0.4–0.6 pp to US GDP growth. |
| Trade balance | Most advanced chips are made in Taiwan and South Korea; the US runs a bigger goods deficit. | Larger US trade deficit, larger Asian surpluses. |
| Electricity demand | Hyperscaler data centres are now a material share of US electricity load growth. | Tighter power markets, rising industrial electricity prices. |
| Equity market concentration | A few firms own most of the upside; their capex is funded mostly from cash flow. | S&P returns and earnings increasingly dominated by AI infrastructure firms. |
| Productivity (eventually) | If models actually raise output per hour, the prize is permanently higher trend growth. | Visible in some sectors, not yet clearly in aggregate productivity data. |
Is this a bubble?
The honest answer in 2026 is "it depends what you mean." Three observations:
- Unlike the 1999 telecoms boom, the spending is funded mostly from current cash flow at firms with very high operating margins. Default risk inside the hyperscaler group is low.
- The cost-of-capital math nevertheless requires AI services to generate cash flows that exceed the depreciation of GPUs that lose roughly a quarter of their value per year as the next generation lands. That is a high bar, and so far the revenues are growing but smaller than the capex.
- If the bar is not cleared, the contraction would look more like the 2002 tech-capex bust (a sharp drop in investment, modest macro recession) than the 2008 financial crisis. There is no leveraged consumer or banking exposure in the way mortgage lending was in 2007.
5. Debt, deficits and the term premium
Public debt is the headline macro vulnerability almost everywhere. The pandemic added 15 to 20 percentage points of GDP in debt across the rich world, and the post-pandemic adjustment never properly happened. The United States is running deficits of around 6% of GDP at full employment, which is essentially unprecedented outside wartime. France is running close to 5%. The UK is running close to 4% with much less fiscal headroom.
Approximate IMF-style general government gross debt. China's number understates local-government financing vehicles; the de-facto figure is higher.
Why the term premium is back
For two decades, the extra yield investors demanded to hold a 10-year bond over rolled-over short paper, the "term premium", was unusually low and at times negative. In 2025 it climbed back into clearly positive territory and stayed there through 2026. Three reasons:
- Supply. Treasury issuance to fund the deficit has stayed at war-time levels.
- Less price-insensitive buying. The Fed is no longer growing its balance sheet, and the Bank of Japan is gradually stepping back from yield-curve control.
- Inflation memory. Holders of long bonds got hurt in 2022 and now want to be paid for inflation uncertainty in a way they did not in 2019.
A higher term premium means long rates can stay elevated even as central banks cut, which is exactly what has happened. Mortgage rates and corporate borrowing costs sit well above pre-pandemic norms.
6. Fragmentation: tariffs, friend-shoring, reshoring
The 1990–2008 era of ever-deeper trade integration is over. Tariffs, export controls and industrial subsidies are now the default policy stance in the US, the EU and China alike, and the political consensus to roll any of this back is absent. The terminology shifts (friend-shoring, derisking, reshoring) but the direction does not.
What changed in the 2024–26 wave
- US tariffs. The average effective US tariff on Chinese imports is now roughly 25–30%, with sector-specific rates much higher for EVs, batteries and solar. Tariffs on imports from third countries are smaller but no longer zero.
- EU trade defence. The EU has adopted countervailing duties on Chinese EVs and is investigating steel, wind turbines and medical devices.
- Export controls. The US and allies maintain sweeping restrictions on advanced semiconductor manufacturing equipment and frontier AI chips to China.
- Industrial subsidies. The US Inflation Reduction Act and CHIPS Act, the EU Chips Act and Net-Zero Industry Act, and equivalent programmes in Japan and Korea, are all rebuilding domestic manufacturing capacity.
The macro consequences
None of these effects is large enough on its own to derail growth. Together they nudge the global economy towards higher trend inflation, more government involvement in investment, and slower productivity growth than the 1995–2020 baseline. AI is the major offsetting force on productivity, which is why so much hinges on whether the AI productivity gains arrive on time.
7. China: deflation and the property hangover
China's growth in 2026 looks fine on the surface, around 4.5%, and dismal underneath. The 2021 property correction is still working through the system. Local governments, which historically funded themselves by selling land, are squeezed. Households, whose savings sit disproportionately in housing, feel poorer. Headline consumer prices have hovered close to zero for three years.
Beijing has responded the way it knows: pour credit into industrial investment and let exports take the strain. The result is record capacity in EVs, solar panels, batteries, drones, shipbuilding, robotics and machine tools, and a current account surplus that, measured in dollars, is the largest in history.
| Indicator | State in 2026 | Why it matters globally |
|---|---|---|
| Real GDP growth | ~4.5%, lower than the 2010s norm | Marginal demand for commodities, autos, capital goods is weaker. |
| CPI / PPI | CPI near zero; PPI mildly negative | China is exporting disinflation in tradeable goods. |
| Property investment | Still contracting in real terms | The biggest drag on growth and on construction-linked commodities. |
| Manufacturing capacity | Sharply expanded since 2022 | Triggers tariff responses in the US, EU, India, Brazil, Turkey. |
| Currency | Slightly weaker against the dollar | Amplifies the export-of-deflation story. |
8. Demographics: the slow squeeze
The slowest-moving macro force is also the most certain. Working-age population growth is now negative in China, Japan, Korea, Italy, Germany and most of Eastern Europe. It is still positive but slowing in the US, France and the UK. The growth engine of the global labour force has shifted decisively to India, Southeast Asia and sub-Saharan Africa.
Demographics is the reason every advanced-economy fiscal projection looks ugly. Fewer workers per retiree means more spending on pensions and healthcare, less revenue from labour income tax, and slower potential GDP growth. It is also the reason for the persistent political pressure on immigration, in both directions. AI enters the debate here too: if it raises output per worker enough, the demographic squeeze shrinks. If it does not, the squeeze keeps tightening through the 2030s.
9. Energy: cheap electrons, expensive molecules
The energy transition has stopped being a forecast and become a fact in the electricity sector. Solar and battery deployment in 2025 broke records again. China installed more solar in one year than the entire United States grid was at the end of the 1990s. The cost curve for batteries has fallen another 30% since 2023. The same is not true for liquid fuels or for industrial heat: oil demand is still growing slightly, and decarbonising steel, cement and chemicals remains slow and expensive.
The new energy macro picture
- Power demand is back. After two flat decades, US electricity load is growing 2–3% a year. Industrial electricity prices are rising in the regions hosting hyperscaler data centres.
- Oil is plateauing, not collapsing. Petrochemicals, aviation and trucking keep demand near 102–103 million barrels per day. The price floor is set by OPEC+ supply discipline, not by demand erosion.
- Critical minerals matter more than oil now. Lithium, copper, nickel and rare earths are where the geopolitics is most acute, and where the bottlenecks bite first.
- Nuclear is back on the table. Hyperscaler purchase agreements have given small-modular-reactor projects a credible first customer for the first time.
10. What could break it
The 2026 picture is fragile in a specific way: most of the risks are correlated through interest rates, AI valuations and trade policy.
| Risk | Trigger | How it transmits |
|---|---|---|
| AI capex bust | Revenue grows slower than depreciation forces capex cuts at hyperscalers | Sharp drop in US investment; equity wealth shock; weaker semiconductor exports from Asia |
| Bond market accident | Failed Treasury auction or sudden rise in term premium | Higher mortgage and corporate rates, fiscal scramble, dollar moves |
| Tariff escalation | Major new round of US–China or US–EU tariffs | Goods inflation up, supply chains rewired again, growth shaved by 0.5–1 pp |
| China stimulus surprise | Beijing turns to large household-side stimulus | Commodity reflation, weaker disinflationary impulse, stronger CNY |
| Energy shock | Middle East conflict, hurricane season, gas pipeline incident | Headline inflation up, real incomes squeezed, central banks delay cuts |
| Productivity surprise (upside) | AI delivers measured productivity gains earlier than the 2010s software lag | Higher trend growth, easier debt dynamics, room for further cuts |
The base case for 2026 is muddling through: growth a bit above stall speed, inflation a bit above target, rates coming down a bit, deficits stubbornly large, and AI capex carrying more of the cycle than is comfortable for anyone outside the firms doing the spending. The interesting question is which of the risks above is the one that ends up writing the 2027 story.
Written as a learning aid. Numbers are illustrative and rounded; for current data see the IMF World Economic Outlook, the BIS Annual Economic Report, and the central-bank statements behind each rate path.