I think this piece pushes the evidence a bit too far. There is an important distinction between declining willingness to hold long-duration US government debt and erosion of the international power of the dollar.
If investors demand a higher yield to hold 10- or 30-year Treasuries, that may mean the US is losing some of the extraordinary borrowing privilege it has historically enjoyed. Given persistent deficits, inflation risk and enormous duration supply, that would hardly be surprising. But it does not follow that the dollar itself is losing its central role.
The latest BIS survey makes the distinction particularly stark. The dollar was on one side of 89.2% of all FX transactions in 2025, and every one of the ten most-traded currency pairs involved USD. That is dollar power in a fairly literal sense. Even when neither party ultimately wants to hold dollars, routing the transaction through the dollar is generally the most efficient path because of the depth and liquidity of dollar markets.
A Treasury is an asset denominated in dollars; it is not the dollar. Selling a Treasury initially leaves the investor with dollars, after which there is a separate decision about where to invest them. And the currency in which we choose to save is itself distinct from the currency through which the world transacts.
There may be a good argument that the US is losing some of its privilege to issue long-term debt unusually cheaply. That is economically important. But it seems a considerable leap from there to evidence of declining dollar power. The evidence presented here strikes me as much stronger for the former than the latter.
This is the main terrain within markets at the moment, so Professor Tooze is spot on to feature two days running. As a practitioner, not an academic, of 35 years standing, my point to analysts, portfolio managers and allocators has always been that there is not necessarily a natural correlation between bond yields and equity returns, nor necessarily a tag-in, tag-out reliable behaviour. The fracture in that dubious framework explains the sub-cutaneous unease in tradable markets.
However, it's important to highlight that every technological innovation in finance this century is favoring dedollarization (blockchains, stablecoins etc.).
It seems the Dollar Standard and the US's status as the financial superpower will end very slowly, by a thousand cuts, and not by a sudden cataclysmic event. Either way, the US is certainly getting relatively poorer, less developed, less capable of dragging the rest of the world through progress, and certainly less resourceful.
The fix is higher taxes or lower spending. Lowering spending during a war footing is impossible, when the communication to the public is that all the cows in the healthcare and old age benefit pens are sacred. I did not see any comparison of this analysis with MMT theories. The logical solution is for the upper brackets for personal income and for the corporate side to accept a 1 to 2 percent increase in taxes. Certainly, in the upper brackets, that money will not be missed. It may be merely symbolic, but the increased revenue can't hurt.
Thanks for putting out chartbook - its a must read daily for me. And I know that you are knowledgeable about MMT and how the plumbing works. You make genuinely interesting observations about Treasury market mechanics but reach the wrong conclusion, I think. Rising Treasury yields and shifting stock-bond correlations reflect investor portfolio repositioning and political uncertainty, not evidence that the US government is running out of money. The US creates its own currency; it cannot be forced into default on dollar-denominated debt any more than a scoreboard can run out of points. The real question is always whether spending outpaces the economy's productive capacity and generates inflation, not what the 30-year yield is doing. The article's 'financial repression or austerity' binary ignores a third option: productive public investment that expands real capacity. And the comparison to past yield-cap inflation episodes ignores that Japan has run yield curve control for years without catastrophe. Sophisticated financial market analysis dressed in technical language can still rest on the fundamental error of treating a currency-issuing government as if it were a currency-using household.
Fascinating read and still getting my head around it. However as Paul Krugman Jared Bernstein discussion US debt to GDP would be more like 47% if Bush and Trump tax cuts which largely went to the rich and a bit of the middle class were reversed so there is a remedy at hand if desired by a new administration. On the other hand we read nearly half of US Treasuries are held by hedge funds now and this is highly leveraged we are told. Given what is going on worldwide I do feel all the holes in a Swiss cheese are lining up for a financial contagion to unfold. Highway to hell? Beyond a reasonable doubt.
I am disappointed by the lack of finance knowledge in this post. Diversification is not a function of correlation. It is a function of covariance. I won't go into the weeds, but simply know that covariance is a function of both correlation and variance. Bond returns are generally far less volatile than equity returns - especially at shorter maturities - and so adding bonds to an equity portfolio will reduce its volatility (risk) even if the correction is positive.
Are Treasuries unattractive assets now? It depends on your liabilities and how you view the future. Could we have a crisis? Sure, but more likely we will continue to muddle through. Is the current situation worse than the inflation of the 1970s or Nixon taking the Dollar off gold?
Come to think of it, the dollar premium could also have included payment for the American military expenditure that guaranteed security for its partners, but that think is probably not prevalent with current US administration to put in a Southern (British) English way
So come a serious attack on confidence the Fed will tout treasuries as before, but plenty of other folk will give different advice. Not what one needs to weather a crisis of confidence, I shouldn't think. Default on interest payment on US$ 40 trillion debt I wonder.
The loop is the part I am marking, policymakers read a fiscal selloff as plumbing, intervene, and muffle the signal, then bills-heavy issuance raises rollover risk until the next auction really does look like plumbing.
Lustig argues Treasury bonds are unsafe because the term premium accounts for the stock-bond correlation switch in 2022, not growth or inflation expectations as implied by a supply shock, and, as Lustig tells us, “The post-2021 period was unusually supply-shock heavy.”
Lustig uses the stock-bond correlation to infer the nature of shocks, from which he infers which component of the bond return “should” adjust. This is an inference too far. The stock-bond correlation informs whether growth or inflation expectations have changed. A positive correlation means inflation expectations have changed; a negative correlation means growth expectations have changed. Liberation Day raised inflation expectations.
Dollar hegemony was never the currency. It was the world stuffing savings into Treasuries because the brochure said "safe." Cognitive dissonance is the buyers noticing the issuer is now the bid.
The part that should keep people up at night isn't the correlation — it's who's actually holding the paper. Count it properly and the largest offshore owner of America's long debt turns out to be the Cayman Islands. That isn't island money. It's leveraged relative-value funds, financed overnight, running the basis trade.
Call that a plumbing problem if you like. Plumbing doesn't get margin-called, and that bid only holds while repo stays cheap.
You're right that the officials are misdiagnosing this. They're not misreading a price. They're misreading their own buyer.
I think this piece pushes the evidence a bit too far. There is an important distinction between declining willingness to hold long-duration US government debt and erosion of the international power of the dollar.
If investors demand a higher yield to hold 10- or 30-year Treasuries, that may mean the US is losing some of the extraordinary borrowing privilege it has historically enjoyed. Given persistent deficits, inflation risk and enormous duration supply, that would hardly be surprising. But it does not follow that the dollar itself is losing its central role.
The latest BIS survey makes the distinction particularly stark. The dollar was on one side of 89.2% of all FX transactions in 2025, and every one of the ten most-traded currency pairs involved USD. That is dollar power in a fairly literal sense. Even when neither party ultimately wants to hold dollars, routing the transaction through the dollar is generally the most efficient path because of the depth and liquidity of dollar markets.
A Treasury is an asset denominated in dollars; it is not the dollar. Selling a Treasury initially leaves the investor with dollars, after which there is a separate decision about where to invest them. And the currency in which we choose to save is itself distinct from the currency through which the world transacts.
There may be a good argument that the US is losing some of its privilege to issue long-term debt unusually cheaply. That is economically important. But it seems a considerable leap from there to evidence of declining dollar power. The evidence presented here strikes me as much stronger for the former than the latter.
https://www.reuters.com/business/finance/deutsche-bank-named-clearing-bank-chinas-renminbi-2026-08-10/
This is the main terrain within markets at the moment, so Professor Tooze is spot on to feature two days running. As a practitioner, not an academic, of 35 years standing, my point to analysts, portfolio managers and allocators has always been that there is not necessarily a natural correlation between bond yields and equity returns, nor necessarily a tag-in, tag-out reliable behaviour. The fracture in that dubious framework explains the sub-cutaneous unease in tradable markets.
I mean, maybe, the juri's still out.
However, it's important to highlight that every technological innovation in finance this century is favoring dedollarization (blockchains, stablecoins etc.).
It seems the Dollar Standard and the US's status as the financial superpower will end very slowly, by a thousand cuts, and not by a sudden cataclysmic event. Either way, the US is certainly getting relatively poorer, less developed, less capable of dragging the rest of the world through progress, and certainly less resourceful.
The fix is higher taxes or lower spending. Lowering spending during a war footing is impossible, when the communication to the public is that all the cows in the healthcare and old age benefit pens are sacred. I did not see any comparison of this analysis with MMT theories. The logical solution is for the upper brackets for personal income and for the corporate side to accept a 1 to 2 percent increase in taxes. Certainly, in the upper brackets, that money will not be missed. It may be merely symbolic, but the increased revenue can't hurt.
Tax increases certainly seem likely once Trump is gone. Before he's gone a Democratic Congress could conceivably shoot for cuts to DoD and DHS.
Thanks for putting out chartbook - its a must read daily for me. And I know that you are knowledgeable about MMT and how the plumbing works. You make genuinely interesting observations about Treasury market mechanics but reach the wrong conclusion, I think. Rising Treasury yields and shifting stock-bond correlations reflect investor portfolio repositioning and political uncertainty, not evidence that the US government is running out of money. The US creates its own currency; it cannot be forced into default on dollar-denominated debt any more than a scoreboard can run out of points. The real question is always whether spending outpaces the economy's productive capacity and generates inflation, not what the 30-year yield is doing. The article's 'financial repression or austerity' binary ignores a third option: productive public investment that expands real capacity. And the comparison to past yield-cap inflation episodes ignores that Japan has run yield curve control for years without catastrophe. Sophisticated financial market analysis dressed in technical language can still rest on the fundamental error of treating a currency-issuing government as if it were a currency-using household.
Fascinating read and still getting my head around it. However as Paul Krugman Jared Bernstein discussion US debt to GDP would be more like 47% if Bush and Trump tax cuts which largely went to the rich and a bit of the middle class were reversed so there is a remedy at hand if desired by a new administration. On the other hand we read nearly half of US Treasuries are held by hedge funds now and this is highly leveraged we are told. Given what is going on worldwide I do feel all the holes in a Swiss cheese are lining up for a financial contagion to unfold. Highway to hell? Beyond a reasonable doubt.
I am disappointed by the lack of finance knowledge in this post. Diversification is not a function of correlation. It is a function of covariance. I won't go into the weeds, but simply know that covariance is a function of both correlation and variance. Bond returns are generally far less volatile than equity returns - especially at shorter maturities - and so adding bonds to an equity portfolio will reduce its volatility (risk) even if the correction is positive.
Are Treasuries unattractive assets now? It depends on your liabilities and how you view the future. Could we have a crisis? Sure, but more likely we will continue to muddle through. Is the current situation worse than the inflation of the 1970s or Nixon taking the Dollar off gold?
Inflation rate implied by spread of bonds over TIPS is still around 2 per cent. But it's the obvious way for US to fix its fiscal problems
Come to think of it, the dollar premium could also have included payment for the American military expenditure that guaranteed security for its partners, but that think is probably not prevalent with current US administration to put in a Southern (British) English way
So come a serious attack on confidence the Fed will tout treasuries as before, but plenty of other folk will give different advice. Not what one needs to weather a crisis of confidence, I shouldn't think. Default on interest payment on US$ 40 trillion debt I wonder.
a terse definition of plumbing would not come amiss
Dishonesty increasingly judging by Adam's column?
The loop is the part I am marking, policymakers read a fiscal selloff as plumbing, intervene, and muffle the signal, then bills-heavy issuance raises rollover risk until the next auction really does look like plumbing.
Lustig argues Treasury bonds are unsafe because the term premium accounts for the stock-bond correlation switch in 2022, not growth or inflation expectations as implied by a supply shock, and, as Lustig tells us, “The post-2021 period was unusually supply-shock heavy.”
Lustig uses the stock-bond correlation to infer the nature of shocks, from which he infers which component of the bond return “should” adjust. This is an inference too far. The stock-bond correlation informs whether growth or inflation expectations have changed. A positive correlation means inflation expectations have changed; a negative correlation means growth expectations have changed. Liberation Day raised inflation expectations.
Dollar hegemony was never the currency. It was the world stuffing savings into Treasuries because the brochure said "safe." Cognitive dissonance is the buyers noticing the issuer is now the bid.
https://paretoinvestor.substack.com/p/treasury-collapse-2026-portfolio-defense
The part that should keep people up at night isn't the correlation — it's who's actually holding the paper. Count it properly and the largest offshore owner of America's long debt turns out to be the Cayman Islands. That isn't island money. It's leveraged relative-value funds, financed overnight, running the basis trade.
Call that a plumbing problem if you like. Plumbing doesn't get margin-called, and that bid only holds while repo stays cheap.
You're right that the officials are misdiagnosing this. They're not misreading a price. They're misreading their own buyer.