Markets weren’t panicking yet, Kelly said at the time. The deterioration is real, but gradual enough that investors have been able to look away. This week, Kelly was back with a bit of a progress report, and it’s not a reassuring check-up.

In a new analysis, Kelly maps out five distinct scenarios for where America’s debt trajectory will lead over the next decade — a more structured and expansive attempt to answer the question he said he gets asked more than any other in October: “When will the federal debt collapse?”

The answer, as before, is that it probably won’t collapse on a fixed schedule. But even Kelly’s most optimistic scenario ends with the federal debt-to-GDP ratio hitting 115% by 2036, up from roughly 101% today. His baseline is 130%. And the worst case — a full-blown fiscal crisis — he describes as “somewhat more likely” than any serious attempt to fix the problem.

Kelly isn’t the only JPMorgan voice sounding the alarm. His boss, Jamie Dimon, has been escalating his warnings in parallel. In January, Dimon warned the $39 trillion national debt was going to “bite.” By late April, he had hardened the prediction: “There will be a bond crisis,” Dimon said at a Norway sovereign wealth fund conference, “and then we’ll have to deal with it.”

Kelly’s analysis lands at a moment of mounting institutional alarm beyond Wall Street. The IMF warned in April that America’s debt problem isn’t a domestic anomaly, and the entire world has caught the American affliction of going broke slowly. The U.S. isn’t an outlier, the fund concluded. “It’s just the most visible symptom of a global disease.” IMF Fiscal Affairs Director Rodrigo Valdés was unsparing in his message: “This cannot wait forever.”

What Kelly’s new note contributes is the analytical architecture behind those headlines — not just whether a crisis is coming, but how it might unfold, through what mechanism, and what investors should do in the meantime.

How we got here

The numbers Kelly opens with are worth sitting with. Federal debt has surged from 31% of GDP in 2001 to 101% today — a generational accumulation driven by “unfunded tax cuts, stimulus checks and wars rather than prolonged economic underperformance.” The fiscal 2026 deficit is on track to come in at roughly $1.89 trillion, the gap between $7.4 trillion in spending and $5.5 trillion in revenues. Interest payments alone will consume more than $1 trillion this year, one of the starkest single data points in Kelly’s analysis.

When Kelly published his “going broke slowly” note last fall, the debt-to-GDP ratio was sitting at 99.9% and he projected it would cross 100% within a year. Now it has, with federal debt held by the public on track to hit $32.2 trillion, or 100.4% of GDP, by the end of this fiscal year. That number will almost certainly be higher by 2036. The only question is how much higher — and what it does to markets along the way.

Scenario 1: Steadily rising debt with rising borrowing costs (the baseline)

In its February outlook, the Congressional Budget Office projected federal debt rising to 120% of GDP by 2036. But Kelly argues that the forecast was built on assumptions that have already been overtaken by events — the CBO assumed tariff revenue would run at $403 billion annually and that the tax breaks in the One Big Beautiful Budget Act would expire on schedule. Strip those out, assume the tax cuts become permanent, and tariff revenue comes in lower, and Kelly argued that debt will hit 127.7% of GDP by 2036. Factor in at least one recession and one bout of inflation over the coming decade — both historically normal — and 130% is a reasonable working assumption.

The IMF largely agreed with Kelly’s diagnosis, with Fiscal Affairs Director Valdés arguing that stabilizing the trajectory would require fiscal tightening of roughly 4 percentage points of GDP, which would rank among the largest peacetime fiscal adjustments in modern American history. Already, he warned, bond markets are sending signals: “These are signs that markets are not as sanguine — as forgiving — as they were in the past.”

The bond market implications are significant. Kelly cites recent Dallas Fed research finding that each one-percentage-point increase in the debt-to-GDP ratio pushes the 5-year-ahead, 5-year Treasury yield up by 3 basis points. A 30-point rise in the ratio would therefore push that benchmark up by 90 basis points — implying 10-year Treasury yields climbing from today’s 4.56% to around 5.46% by 2036.

The same Dallas Fed researchers note that if debt increases as currently projected, long-term interest rates could rise more than 1.5 percentage points over the next 30 years — a longer-horizon estimate that makes Kelly’s decade-long projection look conservative.

Scenario 2: Slow deterioration with little market reaction (the best case)

Kelly’s most optimistic scenario still involves deterioration — just slower, and without a bond market revolt. The ingredients: AI delivers a stronger-than-expected productivity boost, immigration restrictions ease, allowing faster labor force growth, and a prolonged period of divided government prevents either party from piling on more unfunded stimulus. Under that combination, debt might stabilize around 115% of GDP by 2036. The CBO also acknowledged the potential of AI in its baseline, projecting faster growth “as generative artificial intelligence is more widely adopted” as a partial offset to rising debt burdens.

The IMF offers a more complicated read on AI’s fiscal potential. While the fund agrees the technology could “fundamentally reshape the way governments do their business” — boosting productivity, tightening tax administration, and improving delivery of public services — IMF Fiscal Monitor lead Era Dabla-Norris warned that AI also concentrates wealth and disrupts labor markets, and that it could hollow out the very income and payroll tax bases that fund government services. “Are our current tax systems — are our current social protection systems — fit for purpose?” she asked. It’s a question that cuts directly against Kelly’s optimism: the same force he’s counting on to slow the debt’s rise could simultaneously erode the revenue side of the ledger.

The result, Kelly writes, is the best investors can realistically hope for: “a slow deterioration with little market reaction,” with the federal finances still deteriorating, just at a slower pace than recently.

Scenario 3: A full-blown fiscal crisis (the worst case)

Kelly doesn’t mince words about the odds. “A fiscal crisis scenario is somewhat more likely” than any serious attempt to reduce deficits through spending cuts or tax increases, he writes — a sentence from one of Wall Street’s more sober voices that is now standard, as alarm mounts over the debt situation.

The first outcome he predicted is a debt ceiling standoff. Congress raised the ceiling from $36.1 trillion to $41.1 trillion last July as part of the OBBBA, and the next crunch won’t come until at least summer 2027. But when it does, the familiar hostage-taking dynamic could return — and while markets have conditioned themselves to shrug off these threats, an actual default would be, in Kelly’s word, “catastrophic” for Treasuries and global financial markets alike.

The second threat is Fed independence. This administration has aggressively pressured the Fed to cut rates, including reported attempts to fire Fed Governor Lisa Cook and a Justice Department investigation into former Chairman Jerome Powell. That pressure has eased somewhat — the DOJ probe was dropped, Powell opted to stay on as governor — but Kelly warns that a Supreme Court ruling in favor of the President’s authority to fire Cook could reignite the crisis immediately. A Fed perceived as subservient to the White House would shatter investor confidence in the Treasury market, raising the specter of the central bank being conscripted to finance federal spending and abandoning its inflation mandate.

The IMF adds a global dimension to this threat. Valdés noted that real interest rates are now running some six percentage points above pre-pandemic levels, compounding the burden of every existing dollar of debt worldwide — meaning any loss of confidence in U.S. Treasuries wouldn’t stay contained to American markets. “The world economy is being tested again,” Valdés said, “and this is a world that has less degrees of freedom as public finances are more stretched in many, many countries.” Under stress scenarios representing the 95th percentile of plausible outcomes, global public debt could spike to 121% of world GDP within three years.

If any of these triggers fired and global investors lost confidence in U.S. Treasuries en masse, long-term rates would spike, the dollar would fall, and risk assets around the world would sell off sharply — though Kelly notes, with some irony, that some global assets might suffer even greater setbacks than Treasuries in the initial shock.

Scenario 4: Reining in debt through spending cuts

It is theoretically possible to slow the growth of debt through aggressive spending cuts. Kelly is blunt about the obstacles. The $1 trillion-plus annual interest bill can’t be reduced by pressuring the Fed to cut rates without risking an inflationary credibility crisis that would push long rates even higher. Social Security cuts are politically radioactive. Medicare and Medicaid face a double headwind: a demographic surge of aging Baby Boomers and a proliferation of costly new drugs and treatments. Defense spending reductions would require a level of global diplomatic cooperation that Kelly notes is “sadly, not very evident today.” In fact, the last military budget surged upward by half a trillion dollars to $1.5 trillion, and President Donald Trump was overheard at a White House event saying that day care, Medicare and Medicaid should be cut to pay for it.

What about everything else? Federal civilian employment has already been cut 11.5% over the past 15 months, falling to 2.665 million jobs — the lowest raw number since 1966 and the smallest share of total employment since at least 1939. There’s simply not much left to cut. If a serious spending-reduction effort were somehow mounted anyway, the result would likely be positive for bonds — but ambiguous for equities, since the economic drag could easily outweigh the benefit of lower interest rates.

Scenario 5: Reining in debt through tax increases

Higher taxes are the other lever. Kelly dismisses broad-based options — hiking payroll taxes or across-the-board income tax rates — as political non-starters. But he assigns slightly higher odds to targeted measures: increases in corporate taxes, personal taxes on upper-income households, capital gains taxes, or estate taxes. If enacted without being offset by other tax cuts or new spending, such measures could slow the debt-to-GDP climb and would likely benefit Treasuries. The impact on stocks is murkier. Lower rates would be a tailwind — but higher taxes on investment income would reduce after-tax returns on affected assets, and prices could fall accordingly.

The IMF’s diagnosis on tax reform rhymes with Kelly’s skepticism. The fund’s Fiscal Monitor flagged that AI and structural economic shifts are already quietly eroding tax bases — raising the prospect that even targeted tax increases might yield less revenue than projected. “There’s a lot of uncertainty in the way AI will play out,” Dabla-Norris said, “what actual impact it will have on labor markets, what actual impact it will have on inequality.” The challenge for governments, she said, is whether their systems are “adaptable” enough to meet risks that are still taking shape.

The political diagnosis

Overall, Kelly makes a strong argument about political architecture. He contends that the American electoral system is almost purpose-built to prevent fiscal responsibility. A first-past-the-post, low-turnout primary system pushes candidates to the extremes. Unlimited special-interest and private money in elections entrenches commitments to existing tax breaks and spending programs. And interminably long campaign cycles — where media coverage fixates on the horse race and almost never engages with policy — leave voters both disengaged and underinformed.

Kelly’s conclusion here goes further than most: “We can be reasonably sure that no serious attempt will be made to reduce deficits through tax increases and spending cuts over the next decade.”

The IMF frames the same dysfunction in starker global terms. “This is not just a cyclical problem,” Valdés said flatly. “It basically reflects policy choices — permanently higher spending and lower revenues.” Every year of delay, he warned, makes the eventual reckoning more severe.

Rising debt isn’t a reason to abandon long-term investing, Kelly argued. But it is a reason to stop assuming the status quo holds. The most probable path remains the baseline he’s been describing since last fall — debt grinding steadily higher, periodically goosed by crises or political irresponsibility, but partially offset by technological progress and labor force growth. Going broke slowly, in other words. Just now, with the range of outcomes mapped out in uncomfortable detail, and the best case no longer looking quite as reassuring as it once did.