World Debt Is About to Cross $118 Trillion
Global government debt stands at approximately $117.49 trillion as of September 6, 2026, and on this site's model it crosses $118 trillion around September 27, 2026. It is rising by about $24.2 billion a day β roughly $8.83 trillion a year. Every figure below is computed from the same dataset that drives the live tracker, so all of it can be checked against the raw file.
What a trillion-dollar line actually costs
World government debt grows at $279,686 per second in this dataset. That is about $24.2 billion a day and roughly $8.83 trillion a year, which means each new trillion takes about 41 days to arrive. The recent lines landed on exactly that cadence:
| Line | Crossed |
|---|---|
| $115 trillion | May 25, 2026 |
| $116 trillion | July 6, 2026 |
| $117 trillion | August 16, 2026 |
| $118 trillion | September 27, 2026 |
That regularity is worth being suspicious of, and we come back to it below. It is a property of the model, not a discovery about the world.
Who owes it
Four blocs carry most of the total. The United States, China, the European Union aggregate and Japan together account for about 73.7% of all government debt in the dataset:
| Bloc | Debt | Share of world |
|---|---|---|
| πΊπΈ United States | $40.10 trillion | 34.1% |
| π¨π³ China | $21.39 trillion | 18.2% |
| πͺπΊ European Union | $16.11 trillion | 13.7% |
| π―π΅ Japan | $9.02 trillion | 7.7% |
Those same four contribute about 57.1% of the per-second growth β a smaller share than their 73.7% of the stock, because faster-growing borrowers elsewhere now add more each second than the size of their existing debt would suggest. Japan is the clearest illustration: its dollar figure is currently falling, not because it is repaying, but because the yen is weakening against the dollar faster than the debt is growing. We cover why Japan's position holds despite the highest ratio in the developed world.
Why the date is softer than it looks
This is the part most likely to be misread, so we will be blunt about it. The projected crossing date above is a modeled estimate, not an announcement. No institution publishes a real-time global debt total; nobody rings a bell. Our figures are educational estimates anchored to IMF World Economic Outlook (April 2026) general government gross debt baselines and extrapolated forward at a constant per-second rate, reviewed quarterly.
That constant rate is exactly why the trillion lines above arrive every 41 days like clockwork. Real sovereign borrowing does nothing of the kind β it moves in discrete steps as treasuries settle auctions, redeem maturing securities and revise their own accounts, and it is seasonal besides. A tidy 41-day rhythm is the signature of a straight line drawn through real endpoints, and it should be read as "roughly this often", not as a schedule.
Definitions matter too. Gross versus net, general government versus central government, market versus face value β these produce materially different totals for the same country, which is why our figure and any given headline figure will rarely match. Our explainer on how the IMF measures government debt covers the gaps, and the full extrapolation method is on the methodology page.
Is $118 trillion the right thing to worry about?
Not on its own. A nominal total has no denominator, and the economy that services it grows in nominal terms too. The ratio economists actually watch is debt-to-GDP, which this dataset puts at about 94% for the world. 21 of the 137 sovereigns tracked here carry debt above 100% of GDP, and the range runs from Somalia at 9% to Venezuela at 240%.
There is also no level at which something mechanically breaks. The widely cited 90% threshold came from a study undone by a spreadsheet error, and what actually determines sustainability β who holds the debt, in what currency, at what maturity and what interest rate β is country-specific. We take that apart in full. The more informative thing to watch than the trillion count is the interest bill, which is the fastest-growing line in most budgets.
And a low ratio is not automatically good news: several of the lowest-debt governments in the dataset are there because markets will not lend to them, not because they chose restraint.
What comes next
All projections from the current baseline and rate β not scheduled events. If borrowing accelerates or slows, they move.
- World debt reaches $119 trillion β around November 7, 2026.
- China passes $22 trillion β around November 24, 2026.
- World debt reaches $120 trillion β around December 18, 2026.
- US debt passes $41 trillion β around February 25, 2027.
Before most of those, the dataset reaches its October 2026 baseline review, when the extrapolation is reconciled against the IMF's autumn World Economic Outlook. That review is where reality gets to disagree with the model: if global borrowing ran hotter or cooler than the April baseline implied, the base figure and the per-second rate are restated and every projection here is revised with them β including, potentially, a retroactive shift in when the $118 trillion line was really crossed. We publish the reconciliation either way, and we say plainly when the model was off. Past revisions are listed in the changelog.
Sources
- Global Debt Clock dataset (debt.json) β every figure on this page is computed from this file.
- IMF World Economic Outlook, April 2026 β general government gross debt baselines.
- IMF Fiscal Monitor β deficit and debt projections underlying the growth rates.