Ray Dalio says America is going broke. The same debt arithmetic is running in Beijing — but the two sit on opposite balance sheets, and they will pay in opposite currencies: one in inflation, the other in deflation. The number the whole comparison turns on is the one most readers never look at.
Two of the world’s largest economies are running the same debt cycle. They will not end it the same way — and the reason is a single number most people never check.
Ray Dalio’s latest essay, “How Countries Go Broke,” lays out a template for how a heavily indebted government eventually reaches a reckoning. He applies it to the United States. But he notes, almost in passing, that the UK, the EU, Japan and China are on the same path. That aside is the interesting part — because when you run China through the very same template, it fits, and then it diverges in a way that changes the whole conclusion.
Dalio’s framework, stripped to its bones, is three questions you can ask of any indebted government. One: how large is debt service relative to revenue — how much of what comes in is already spoken for? Two: how much debt must be sold relative to the demand to buy it — are there willing lenders? Three: is the central bank printing money to buy the debt the market won’t? When all three worsen together, he argues, the endgame arrives — and it is paid in the currency and in inflation.
On his own figures, the US is well into it: total federal debt around $40 trillion, an interest bill near $1 trillion (about a fifth of revenue), and once you add the principal that must be rolled over, debt service approaching 200% of the money coming in. His prescription is a “3% solution” — shrink the deficit to 3% of GDP through spending cuts, revenue and lower rates together — and his portfolio steer is to underweight bonds and hold a strategic slice of gold. We take his US numbers as his; our interest is in what happens when the same three questions are put to Beijing.
China fits the template on the numbers. Its headline government bonds are ~60% of GDP, but that understates the load: the IMF’s “augmented” measure — which counts the off-budget local-government financing vehicles (LGFVs) — puts the true burden at roughly 124–127% of GDP, and rising toward ~154% by 2030. The augmented deficit is around 14% of GDP. The revenue side is broken where it matters most: land sales, once ~30% of local-government income, have fallen by roughly half from their 2021 peak as the property market deflated. And the broadest price gauge has been falling: China’s GDP deflator was −1.0% in 2025, and while consumer inflation has since crept back to just +0.5% (July 2026, NBS), the deflator is still projected negative into 2026.
So far, a mirror. Same three questions, same worsening answers. The difference is in how the questions are answered — and that is where China stops looking like America and starts looking like something else entirely.
Take Dalio’s second question — debt sold versus demand. In America, debt is sold into an open, global market with real bond vigilantes; foreigners hold roughly a third of it and can vote with their feet, and a bad auction is a live signal. In China, it is sold into a closed, captive system. The capital account is largely shut, and a state-directed banking system absorbs the issuance — banks’ holdings of government bonds have climbed from ~4% of their assets to ~15% (by early 2025). Foreigners own only about 6% of Chinese government bonds (mid-2025, and trimming since). There are, in effect, no vigilantes: yields fall even as issuance surges (the 10-year is around 1.68%, as at 24 Aug), because the buyers are told to buy.
And the third question — money-printing — runs backwards in China. The central bank only began buying government bonds in 2024, then cut back because yields were falling too fast. The debt is absorbed by the banking system, not the central bank’s printing press. The much-touted ¥10 trillion debt swap tells the same story. Announced in November 2024, it is a five-year programme — an expanded ¥6 trillion local-borrowing quota plus ¥4 trillion of special bonds — that lets local governments move their expensive, off-the-books LGFV debt onto the official ledger as cheaper government bonds, shrinking the debt earmarked for the swap from ¥14.3tn toward ¥2.3tn by 2028. It saves a cumulative few hundred billion yuan in interest — genuine relief — but it re-labels and defers; it does not repay, and it does nothing to rebuild the collapsed land-revenue base that caused the squeeze in the first place.
Here is the number the whole comparison turns on, and the one most people skip. Rank these giants by gross government debt and the order is Japan (~230% of GDP), then China (~124% augmented), then the US (~130% including intragovernmental holdings). But gross debt is only the burden. The buffer is the net international investment position — what a country owns abroad minus what foreigners own of it — and there the ranking violently inverts.
| Country | Gross govt debt | Net external position (NIIP) | Reserves |
|---|---|---|---|
| China | ~124–127% GDP (augmented) | +$4.1tn net creditor (world #2) | $3.4tn |
| Japan | ~230% GDP (gross) | +$3.5tn net creditor (world #3) | — |
| United States | ~130% GDP | −$21.3tn — largest net debtor on earth | ~$0.9tn |
FX reserves are the liquid, state-held core within each net position — a component of it, not additional to it. China NIIP & reserves: China SAFE (end-2025); US NIIP: US BEA (Q1 2026).
China owns ~$4.1 trillion more of the world than the world owns of it (its net international investment position at end-2025) — and roughly $3.4 trillion of that sits in official FX reserves, the liquid core of the position, inside the $4.1tn rather than on top of it. It is a bigger net creditor, now, than Japan. The US is the mirror image: Americans own ~$43tn abroad, but foreigners own ~$65tn of America, leaving a net hole of $21.3 trillion — the largest of any nation, much of it foreigners owning US companies and shares, not merely lending. So surely China’s fortress reserves make its debt safe? The trap is twofold. First, scale: that ~$4tn net-creditor cushion is barely a fifth of China’s own ~$24tn of augmented public debt — a buffer, not a match for the burden. Second, and deeper, the buffer is the wrong tool — even if it were bigger, it could not easily reach the debt, which a simple picture makes plain.
Picture a family that owes a large sum to neighbours and local shops — all in local currency — and also owns a villa abroad and a portfolio of foreign shares. The overseas holdings are a genuine cushion. But they do not easily meet the debts at home.
To pay the neighbours with the villa, the family would have to sell it, bring the money home, and change it into local cash — and converting a large sum of foreign money tends to push the local currency up (unhelpful for a household whose local business competes on price) while running down the overseas nest-egg that keeps their foreign bank relaxed. They can do it — to fund an emergency, or prop up one struggling relative — but that buys time, not a cure: the local debts and the weak local income are still there.
So the villa is the family’s shield against a foreign lender, or a run on their standing abroad. It is the wrong first tool for owing too much at home. China’s ~$3.4 trillion of official reserves are that villa: a powerful defence against a currency run or capital flight, but not a free pool of domestic spending power. They can buy time and fund targeted rescues; they cannot, by themselves, repair local-government finances, the property sector, or falling prices. The buffer and the burden sit on different balance sheets.
Put the pieces together and the two adjustments fall out of the two balance sheets — and they are not the same adjustment. America has no net-asset buffer, but it has something arguably stronger for avoiding an external crisis: it borrows in the currency it prints, so it cannot be forced into a foreign-currency default the way an emerging market can. Its vulnerability is therefore not a sudden dollar collapse but a slower repricing — if the world’s appetite to fund US deficits fades, the strain shows up first in higher real yields and term premium, and only then, should policy credibility fray, in inflation and a softer dollar. The privilege is real; it is simply not unconditional — it rests on the very foreign confidence a rising debt path keeps testing.
China sits at the opposite corner. Captive funding and a closed border make an abrupt, emerging-market-style funding stop remote — but they do not make stress disappear. China has had developer defaults, sharp property-price falls and visible bank strain; what its system does is defer and distribute the losses rather than detonate them. With no market disciplinarian to force a single reckoning, the adjustment comes not as a crash but as a slow, administered balance-sheet repair — carried by savers (low real returns on deposits), banks (squeezed margins, which the IMF confirms are still weakening), and local governments (arrears and austerity), against a backdrop of falling prices. The signature is not a blow-up; it is the Japan playbook — deflation, financial repression, sub-trend growth.
For an investor, two takeaways. First, stop waiting for a Chinese “Lehman moment.” The structure that spares China a market disciplinarian also lets it defer the reckoning; the tell is not a headline default but the slow signals — CGB yields lower-for-longer, bank margins, a firm-but-managed yuan, and whether Beijing’s extend-and-pretend keeps pace with an augmented debt path the IMF sees passing ~154% of GDP by 2030. Second, a caution on the leap so often made here: this balance-sheet comparison explains why investors are drawn to hard, non-fiat stores of value when one giant’s adjustment runs through its currency and the other’s through repression — but it does not, on its own, make the case to own them. That case turns on valuation, volatility, liquidity and your own objectives; it is a suitability conversation, not a conclusion you can read off a debt table.
What to look for. The NPC Standing Committee (late August) for the next stimulus and local-debt quota signal; whether the GDP deflator climbs back above zero (the single cleanest sign the deflation is breaking); bank net-interest-margin disclosures (the clearest victim); and the managed path of the yuan, which is where any real pressure would leak out. On the US side, the term premium and foreign appetite at Treasury auctions — the tells Dalio’s template watches.
What to be wary of. Reading China’s $3.4tn of reserves as an offset to its domestic debt — it is not (different currency, different balance sheet); waiting for a China debt crisis the closed system is built to defer; and ranking these economies on gross debt alone, which hides the buffer that flips the picture. Deflation is not automatically bad for asset prices, but persistent deflation with a broken revenue base is a poor backdrop for domestic banks and property.
What we are not doing. This is general market commentary, not advice or a recommendation on any security. Countries, markets and instruments are described to illustrate a pattern, not as a trade. Anything that touches your own portfolio is a suitability conversation — one that needs a person, not a page. We are always glad to have it.
Net International Investment Position (NIIP · 净国际投资头寸) — a country’s balance sheet with the world: everything its residents own abroad, minus everything foreigners own of it. Positive = a net creditor (the world owes it on balance); negative = a net debtor (it owes the world).
Augmented debt (广义／增广债务) — the IMF’s wider measure of Chinese government debt that adds off-budget local-government (LGFV) borrowing to the official bonds — the honest “true burden” figure.
The three deficit measures (官方／广义／IMF 增广赤字) — one government, three perimeters. The official headline deficit (~3–4% of GDP) counts only the general public budget; the broad deficit (~9%) adds the land-reliant Government Funds Budget; the IMF “augmented” deficit (~14%) also consolidates off-the-books LGFV and policy-bank borrowing — the truest gauge of the fiscal stance. The same ladder scales the debt: ~60% explicit bonds → ~99% general government → ~124–127% augmented.
LGFV — Local Government Financing Vehicle (地方政府融资平台) — companies Chinese local governments set up to borrow and build infrastructure, whose debt sits off the official government books.
Bond vigilantes (债券义警) — investors who discipline a government by selling its bonds — pushing yields up — when they doubt its finances. A closed market like China’s has none.
GDP deflator (GDP 平减指数) — the broadest gauge of economy-wide prices. Below zero means deflation — prices falling across the whole economy, not just one basket of goods.
Financial repression (金融抑制) — policies that quietly move wealth from savers to borrowers and the state, chiefly by holding deposit and bond rates below where a free market would set them.
Debt swap (化债／债务置换) — replacing expensive or hidden debt with cheaper official debt. It changes the label and the interest cost — not the amount owed.
Intragovernmental holdings (政府内部持有) — debt one arm of a government owes another (e.g. US Social Security trust funds holding Treasuries), rather than debt owed to outside investors.
Reserve-currency privilege (储备货币特权) — the edge the US enjoys because the world needs dollars: it can borrow cheaply, and in the very currency it itself prints. Sometimes called “exorbitant privilege.”
General commentary only — not investment advice. By contacting us you consent to Touchstone using your personal data to respond to your enquiry, in accordance with the PDPA. We do not share it with third parties.
瑞·达利欧（Ray Dalio）在《国家如何破产》中，为高负债政府的最终清算给出了一套模板，并用它诊断美国：联邦债务约 40 万亿美元、利息支出近 1 万亿美元（约占财政收入两成），叠加到期本金后，还本付息已逼近收入的 200%；其结局，由货币与通胀来偿付。他顺带提到，中国也在同一条路上——把同一套模板套到北京，起初高度吻合，随后却在偿付方式上彻底分道。
同一道算术，不同的门牌 · 为何中国没有「时刻」
中国的广义（含 LGFV 隐性债务）政府债务约为 GDP 的 124–127%、广义赤字约 14%，物价实际在下跌（2025 年 GDP 平减指数 −1.0%），而昔日占地方收入约三成的土地出让金已较 2021 年峰值腰斩。但答案的方式不同：中国的债务卖给一个封闭、被动的体系——资本账户基本关闭、国有银行被引导承接（银行持有的政府债券占资产比重由约 4% 升至约 15%），外资仅持有约 6% 的政府债券；因此没有「债券义警」，收益率在发行激增之际反而走低（10 年期约 1.68%）。央行只是浅尝购债（2024 年），旋即因收益率下行过快而收手。10 万亿元化债只是「以低息官债置换高息隐性债」——再标注、往后推，而非偿还，更未修复坍塌的土地财政。
被误读的「缓冲」 · 两种结局
关键的一笔，是多数人忽略的净国际投资头寸（NIIP）——一国对外资产减对外负债。按总债务排序是日本＞中国＞美国；但按净头寸则完全反转：中国为 +4.1 万亿美元净债权国（全球第二）、日本 +3.5 万亿（第三），而美国是 −21.3 万亿美元、全球最大净债务国（外资持有的美国资产约 65 万亿，远超美国持有的海外资产约 43 万亿）。该 +4.1 万亿美元净债权中，约 3.4 万亿美元为官方外储（在其内，而非额外叠加）——净债权规模如今已超过日本。但这份「缓冲」是错配的工具：外储与净外币资产只能抵御货币／国际收支危机，并非政府的财政资产，对本币计价的国内债务与通缩几乎无能为力——「缓冲」与「负担」根本不在同一本账、同一种货币上（详见英文版「海外别墅」的比喻）。于是两种结局由两张资产负债表决定：美国无缓冲，只能靠货币——通胀与美元走软，把损失输往全球美元持有者；中国有堡垒却困于内部——把损失内部化给储户、银行与地方政府，表现为通缩、金融抑制与低于趋势的增长，即「日本化」，是慢性寒冬，而非一次性崩塌。
对投资者的启示
其一，不必再等中国的「雷曼时刻」：让中国失去市场纪律的结构，同样使其难以崩盘；要看的是慢变量——CGB 收益率长期偏低、银行净息差、被管理的人民币，以及「化债—展期」能否追上迈向 ~154%/GDP 的债务路径。其二，正如我们近期关于美元与黄金的评论：两种截然不同的财政结局，仍指向同一类「保险」——当一方贬值、另一方抑制，既不听命于前者、也不听命于后者的硬资产／非法定货币价值储存，因相反的理由而各得其所。这是组合构建的问题，而非追涨。
本摘要为节选，非全文翻译。本刊为一般性市场评论，不构成投资建议或任何证券推荐；文中所涉国家、市场与工具均作举例说明。如与英文版本存在任何歧义，概以英文版本为准。