What if the danger in global debt is not collapse, but comfort? The modern fear is of a sudden break, a dramatic cascade, a debt trap with a loud ending. Yet the more unsettling possibility is duller and more durable: debt keeps working, then starts working less well, until societies mistake endurance for strength. That is the condition now suggested by the latest global figures. The world is still borrowing at immense scale, but the real question is no longer whether the system can add debt. It is whether it can still convert debt into growth, resilience, and political room to move.
The headline numbers look severe enough to satisfy the doom merchants. Global gross debt reached $365 trillion, or 311% of world GDP, in the first half of 2026, according to the Institute of International Finance. But the ratio tells a more interesting story than the absolute figure. Between 2018 and June 2026, the global debt-to-output ratio rose by less than 10 percentage points. It peaked after the pandemic at 337% in early 2021, then fell 27 points by the end of 2023, helped by inflation. That is not the pattern of an economy about to snap. It is the pattern of a structure that bends, then strains, then bends again.
The appearance of stability can be its own hazard. A bridge does not need to fall to be unsafe. It only needs to accumulate fatigue in the joints until every extra load matters more than the last. The IIF’s concern is not that debt has suddenly become unpayable everywhere, but that room for manoeuvre is shrinking in the places that matter most. It is more worried about high-income economies than emerging or developing ones, and more worried about public debt than private debt. That ranking matters. The mature states that built the global monetary order are now those with the least slack.
The most revealing number may be the one that sounds least dramatic: interest costs for mature-market governments rose 1.5 percentage points between December 2021 and the end of August 2026, reaching 3.3% of GDP. In the US, interest costs rose 1.6 percentage points of GDP between January 2021 and 2026. This is how debt turns political before it turns financial. It is not simply that governments owe more. It is that more of the budget is pre-committed before any new choice is made. A government with rising interest costs is like a ship with deeper ballast and less cargo room. It still floats, but it carries less of what voters think they are paying for.
The IIF put the point sharply: “Los gobiernos de mercados maduros ahora gastan más en intereses de lo que el mundo invierte en inteligencia artificial, defensa o energía.” Even allowing for translation, the implication is plain. Debt service is now competing with the spending categories politicians like to call strategic. That is the quiet theft of maturity: every year of high rates and heavy issuance forces a trade-off between old promises and new ambitions. The country does not default. It merely grows less free.
This is why simple debt ratios can mislead. A ratio is a photograph; interest costs are weather. A debtor can survive a large balance sheet if financing remains cheap, growth remains firm, and confidence remains intact. But once rates rise and growth slows, the same debt begins to behave like a fixed weight on a climber’s ankle. The danger is not only insolvency. It is policy sclerosis. Governments under pressure do not necessarily do less; they do more of the wrong things, faster, because they need relief before they need reform.
The IIF’s second warning is even more corrosive: “La deuda se ha convertido más en una cuestión política que macrofinanciera, creando un círculo vicioso entre elecciones y soluciones rápidas a corto plazo, y una vulnerabilidad a largo plazo a medida que disminuye la utilidad marginal de una mayor deuda.” That is a dense sentence, but the meaning is simple. Debt stops being judged by what it does to future capacity and starts being judged by what it can postpone until after the next vote. This is classic game theory in the real world: the short-term payoff goes to the politician who buys time, while the cost is spread across taxpayers, holders of currency, and future governments.
History offers little comfort. Rome did not fall because one emperor printed too many coins. It weakened through repeated acts of delay, each rational in isolation, each corrosive in aggregate. Late-stage empires do not usually collapse under the weight of a single blunder. They decay through many acts of arithmetic denial. Modern democracies are not ancient empires, but the logic is similar. The more debt becomes an instrument for pushing pain outward in time, the more future policy space is consumed by present impatience.
And yet the world has not reached the brink. That matters. Panic is often just a bad substitute for analysis. The data show that the debt burden, measured against output, has risen only modestly since 2018 and has already come down from its pandemic peak. Inflation did some of the work, which is hardly a noble solution, but it still reduced the ratio. That should remind investors and policymakers alike that debt problems are rarely solved by moral language. They are solved, when they are solved, by growth, inflation, restructuring, repression, or time. The order of those tools matters.
The corporate side adds another layer of irony. AI-linked corporate bond issuance is estimated at $541 billion in 2026, up from $90 billion three years earlier. The average maturity of AI-related corporate debt reached 13.4 years in 2026, up from 11.5 in 2024. By contrast, US Treasuries carry an average maturity of just 2.5 years, down from 6.1 in 2024. Those figures do not prove a bubble, and they do not prove prudence. They show something more subtle: capital markets are willing to finance long-dated bets on future transformation, while governments are increasingly forced to live inside shorter refinancing cycles.
That contrast says a great deal about investor psychology. Private capital can stretch duration when the story is compelling enough. It can pretend that a long asset life is the same as a long certainty. But debt is not magic; it is a schedule. The longer the maturity, the more confidence the lender is expressing in the borrower’s future cash flow or, in the public case, future tax base. If the future disappoints, duration becomes a slow trap rather than a cushion. The market’s enthusiasm for one sector does not remove the economy’s broader fragility. It merely relocates it.
The more debt is used to fund ambition, the more people confuse motion with strength. That is a recurring error in markets. A factory full of machinery is not automatically productive. A balance sheet full of liabilities is not automatically fragile either. The question is whether the underlying system can absorb shocks and still adapt. That is the difference between fragility and antifragility. Some systems gain from disorder. Most debt structures do not. They are engineered for continuity, not stress. When rates change, growth wobbles, or politics turns hostile, their weakness is revealed not by collapse but by constraint.
It is easy to mistake constraint for crisis. The FT summary of the IIF’s findings points to an economy that is not on the eve of disaster, but one in which manoeuvre is shrinking. That phrasing is important because it resists melodrama. The world is not drowning in debt in a single dramatic wave. It is standing on a seawall that must be raised again and again. Some years the tide recedes. Then inflation helps. Then rates rise. Then politicians lean harder on borrowing. The water level is not the whole story. The real story is erosion.
For investors, that means the old habit of treating sovereign debt as an unquestioned anchor deserves more scepticism. An anchor is useful only if the seabed holds. If the ground softens, the anchor drags. For governments, it means debt is no longer just a financing tool. It is a test of institutional discipline. The more markets and voters assume that borrowing can always postpone hard choices, the more likely it is that the hard choices arrive with less capital, less trust, and less time.
The next important test will not be a single dramatic date or event. The IIF’s next Global Debt Monitor would be the natural follow-up, but no date is confirmed. That uncertainty is fitting. Debt crises rarely announce themselves cleanly. They accumulate in the unglamorous places: in budgets, in refinancing needs, in political promises that compound faster than revenues. The world may not be on the edge of disaster. But it may be nearing the edge of convenience. And that is often the more dangerous place, because it feels survivable right up to the point where it is not.
