Michael Pettis

1/20
The views of Maurice Obstfeld (and other American economists) on the relationship between the internal and external accounts of the US are finally starting to evolve. 

https://www.brookings.edu/wp-content/uploads/2025/03/3_Obstfeld.pdf

2/20
From insisting that the the US external account is wholly driven by domestic imbalances, and that the only sustainable way the US can reduce its trade deficit is by reducing the fiscal deficit, he now accepts that the reality is "more nuanced".

3/20
But in getting there he now argues that people like me are wrong to insist that only external factors matter. Of course neither I nor anyone else has ever made that argument. What I've long argued is that the US plays a special role in absorbing global savings imbalances.

4/20
Of course he is right to say internal factors matter, but the fact remain that the US external balance and its internal balance must always be consistent, and that the US external balance must always be consistent with the external balance of the rest of the world.

5/20
Once you begin to accept that foreigners, too, have agency, and that large economies with restricted capital and trade accounts, very large state players, and highly controlled domestic banking systems can control their internal balances, then it becomes...

6/20
much easier to recognize that they also control their external balances, and by controlling their external balances, they effectively export the consequences of their internal balances to the rest of the world.

7/20
These consequences can include not just persistent trade deficits, but also a re-ordering of domestic economies. When large, relatively closed economies structure their internal imbalances, for example, around policies that boost manufacturing at the expense of...

8/20
consumption, a necessary corollary is that the expansion in their share of global manufacturing must be matched by a contraction in the share of global manufacturing of those of their trade partners that absorb their imbalances.

9/20
Obstfeld doesn't address this, of course, but perhaps he thinks it is just coincidence that the manufacturing share of advanced economies that have long run persistent deficits – like the US,  France, the UK, Canada, etc. – is well below...

10/20
the global average while the manufacturing share of advanced economies that have long run persistent surpluses – like Japan, South Korea, Germany, Taiwan, etc. – is well above the global average. 
https://www.phenomenalworld.org/analysis/trade-and-the-manufacturing-share/

11/20
But it's not coincidence. In a globalized world, industrial policies implemented in more closed and regulated economies can effectively become the "industrial policies" (in obverse) of more open and less regulated economies.
https://carnegieendowment.org/china-financial-markets/2024/07/which-country-should-design-us-industrial-policy?lang=en

12/20
As part of his broader dismissal of foreign agency, Obstfeld asks why, if major reserve currencies lead to trade deficits, doesn't the EU run a trade deficit. But he is looking at the wrong currency.

13/20
The more interesting question should be why, if US economic policies are wholly to blame for net foreign inflows, do other countries with similarly deep, flexible, well-governed and completely open financial markets – like...

14/20
the UK, Canada, and Australia (until recently, before soaring iron prices put it temporarily into surplus) – also run persistent net foreign inflows, trade deficits and manufacturing shares below the global average.

15/20
At any rate, the good news is that American economists are slowly beginning to remember what they used to know, which is that countries can choose between more global integration or more control over their domestic economies.

16/20
Countries that are determined to control their domestic economies will also control their external balances. Countries that are open to external imbalances have less control over their domestic economies. This creates a tension in a globalized world economy, as Joan...

17/20
Robinson explained decades ago. When some large economies choose the former and others the latter, because the former can externalize the costs of domestic imbalances through what she called "beggar thy neighbor" policies, in the end the global trading system...

18/20
must break down as the latter are eventually forced to become like the former.  Globalization requires that no major economies try to exert control over their external imbalances. When some do and others don't, globalization falls apart.

19/20
That's why Keynes proposed a global trading system in which countries were limited in their abilities to run external imbalances. These external imbalances, he explained, were just a way to force their trade partners to absorb their own weak domestic demand.

20/20
I am pretty sure that eventually everyone will rediscover the mutual links between the domestic and external balances, and the tension between global integration and national control. Until then, I recommend this book. 

1/9
The FT notes that China recorded its third consecutive year of population decline in 2024, but it seems to me that much of the discussion surrounding the population decline may be a little overdone. 

https://www.ft.com/content/516f9c37-b9a8-44b1-8e64-10e582f7d5ae via @ft

2/9
The real medium- and long-term challenge for China lies in the way it rebalances the domestic distribution of income. If it can raise household income fast enough and in a non-disruptive way, a declining working population is still consistent with...

3/9
rising living standards, even if that means (as it most likely does) a slowdown in the GDP growth rate. If it cannot , a declining working population will be the least of its problems.

4/9
For me the real demographic problem China faces is not the decline in its working population but rather the unevenness of that decline. That's because the working population will continue to rise at a healthy pace in the faster-growing, richer, eastern parts of China.

5/9
The rest of China, however, with its low income levels, weak growth, and unmanageable debt burdens, will absorb more than 100% of the decline in China's working population. What is worse, it will be their best-educated and most entrepreneurial young people who will migrate.

6/9
With this area already comprising a disproportionate share of non-productive investment in property and infrastructure, a declining working population means that the economic value of this investment will fall even faster, worsening the debt burden.

7/9
If this forces the poorer parts to cut back on educational spending and social services, it can further entrench the disparities.

The question is how Beijing will respond to these rising differences.

8/9
It would have to choose between diverting resources to richer China, which would be better for the economy, or to poor China, which will be better for social and political stability. I am not sure how Beijing will respond, but I tell my clients that we should...

9/9
increasingly consider not just how the Chinese economy will choose, or be forced, to adjust, but perhaps even more importantly, how unevenly this adjustment, including the demographic adjustment, will be distributed throughout the economy.

1/7
Very good FT article on how China achieved its dominance in manufacturing, along with the cost both to China and to its trading partners: "Using a unique combination of industrial policy, subsidies and other state-support coupled with private sector...
https://www.ft.com/content/724431ad-26db-4f6d-acab-ccb3cad11daa

2/7
entrepreneurialism and ferocious competition in China’s vast market, the country was able to sharply increase the share of Chinese producers domestically and internationally in many of the sectors, in some cases matching or exceeding foreign competitors’ technology."

3/7
Because the subsidies and other state support mostly came in the form of direct and indirect transfers from the household sector, the huge expansion in China's manufacturing sector was also the flip side of the huge contraction in the consumption share of GDP.

4/7
This matters, because the result—China's share of global manufacturing grew much more quickly than its share of global GDP, while its share of consumption grew much more slowly—had to be mirrored in the economies of its trade partners.

5/7
They collective saw their share of global manufacturing grow more slowly than their share of global GDP, while their share of consumption grew more quickly. This outcome was necessary as long as China controlled of its external account and some of its trade partners didn't.

6/7
This article reinforces the idea that not only did Chinese industrial policy massively restructure its own economy but also, through its impact on its external account and that of its trade partners, it effectively restructured the economies of its trade partners.

7/7
In a hyperglobalized world, in other words, countries that maintain open trade and capital accounts must absorb the reverse industrial policies of those implemented in countries that do control their trade and capital accounts.
https://carnegieendowment.org/china-financial-markets/2024/07/which-country-should-design-us-industrial-policy?lang=en

1/6
Emmanuel Macron: "Today, we are caught between the US and China and it is a matter of life or death for the European industry. We have become the adjustment market and this is the worst-case scenario."

He is absolutely right.

2/6
This is the point I have been making again and again over the years. The global economy is a closed system, and it must balance. This means that domestic imbalances created by countries that control their external accounts must...

3/6
necessarily be exported to and absorbed by those of their trade partners that chose not to control their external accounts. It also means that the latter must end up with domestic imbalances that accommodate the domestic imbalances of the former.

4/6
If China's strategy is to continue expanding its share of global manufacturing, and the US strategy is no longer to accommodate that expansion but instead try to reverse the decline in its share of global manufacturing, some other country must accommodate these policies by...

5/6
giving up manufacturing share. This is just arithmetic.

As Joan Robinson explained, when some major economies control their external accounts and others don't, the resulting beggar-thy-neighbor imbalances will eventually force the latter into protectionism.

6/6
That's why the only stable trading systems involve either
a)all countries choosing global integration over economic sovereignty,
b)penalties against countries that use trade to export their domestic imbalances, or
c)all countries controlling their external accounts.

Perhaps, but it did this by passing most of the adjustment costs onto workers and that politically is unsustainable. We need a flexible money system along with a trading regime that penalizes beggar-thy-neighbor trade imbalances.

1/17
I'm all in favor of myth-busting but this article seems to miss the point. To my knowledge, and contrary to Sharma's claim, no one has ever argued that the consumption share of China's GDP is low because consumption has grown slowly.
https://www.ft.com/content/bf1e8755-de42-4ef3-97bf-1215d8bf894e

2/17
The argument has always been that consumption and household income have grown much more slowly than GDP because of a series of explicit and implicit transfers from households that subsidized growth in investment, especially investment in manufacturing and infrastructure.

3/17
For example, the biggest decline in the consumption share of GDP occurred in the first decade of this century largely because of extremely low interest rates – highly negative in real terms – that were engineered to resolve the cleaning up of the banking system.

4/17
At that time, the GDP deflator was 8-10%, and nominal GDP growth was 16-20%, but lending rates were 7-8% and deposit rates were 3-4%. This allowed the AMCs that had purchased bad debts from the banking system to resolve these bad debts over 10-20 years with minimum losses.

5/17
But these low rates represented transfers from household savers to banks and borrowers that equalled roughly 5% or more of GDP. Not surprisingly, during this time household consumption dropped from a very low 48% of GDP to an astonishingly low 34% of GDP.

6/17
Because of what was effectively a huge implicit tax on household savings, along with an equally large subsidy to banks and borrowers (almost none of whom were households), the inevitable result was a sharp drop in the household share of GDP along with a surge in investment.

7/17
This, in turn, explains most of the decline in the consumption share of GDP during that period. Add to this the fact that the hukou system prevented the social safety network from keeping pace with GDP growth even as it prevented wages from keeping...

8/17
pace with productivity growth, and it is easy to see why the household savings rate remained high even as the household share of GDP declined consistently, both of which pushed up the national savings rate and pushed down the household consumption share.

9/17
Other policies also matter. China's massive overinvestment in infrastructure since 2008-10 transferred resources from the country as a whole (again, mainly through low interest rates) to the businesses that benefitted from the consequent reduction in logistical costs.

10/17
More recently, the surge in manufacturing investment as China externalized the cost of the property crisis was another similar transfer, and explains why even as Chinese growth slowed sharply, the growth in consumption slowed even more.

11/17
Sharma says that "calls to free the Chinese consumer persist alongside mounting evidence of the steady growth in their spending", but this isn't true. The consumption share is still lagging. Last year for example, consumption accounted for a miserly 45% of...

12/17
GDP growth which – excluding 2020, when consumption contracted – is the lowest recorded in decades. It is more typical for consumption to account for 60% of China's GDP growth, which is already low in a world in which consumption typically accounts for 75% of GDP growth.

13/17
We will almost certainly see consumption lag again this year. For all the talk, in other words, the consumption share continues to decline even as growing subsidies boost manufacturing capacity well above the ability of Chinese consumption to keep pace.

14/17
Sharma argues that the problem is not too little consumption but too much investment, but these are simply two sides of the same coin. China overinvests for the same reason that it underconsumes – because of the extent of transfers from households that subsidize...

15/17
spending on infrastructure and, increasingly,  manufacturing. To reverse these transfers will be extremely difficult because, as Sharma correctly notes, China has locked itself into GDP growth targets that are way too high.

16/17
But it is the only way to do either. Until Chinese households are able to see their total disposable income rise in line with their productivity, consumption must continue to lag growth, even as investment continues to lead growth.

17/17
I agree with him that a sustainable GDP growth rate is a lot closer to 2.5% than to 5%, but to separate the problem of underconsumption from that of overinvestment fails to recognize that both are a consequence of the same set of distortions.
https://carnegieendowment.org/china-financial-markets/2023/04/can-chinas-long-term-growth-rate-exceed-2-3-percent?lang=en

1/15
Kenneth Rogoff says: "There is, for example, a terrific chapter in which Ray Dalio brutally critiques Japanese policymakers for failing to force debt writedowns after the country’s early 1990s financial crisis.'

https://www.ft.com/content/e1b99ad7-0f01-4f95-a3d8-489f6a7c51a9 via @ft

2/15
"Instead," he continues, "they allowed debt overhang to hamstring the financial system and sap two decades of growth."

I haven't read Dalio's book, but this is an extremely important point, and one that Beijing should note.

3/15
Beijing has an enormous of debt that is ostensibly backed by the book value of associated assets (most Chinese debt was used to fund investment), but the economic value of these assets are not worth nearly as much as their book value.
https://www.ft.com/content/630f828c-ce4b-4f41-a867-9593bfaf0528

4/15
We've already seen this in the property market, where the value of the assets plunged relative to the value of the debt, but this is an even bigger problem in infrastructure investment and has become a rapidly-growing problem in manufacturing investment.

5/15
One way to understand this is to recognize that there are huge losses on China's collective balance sheet that result from the difference between the asset's book value and its real economic value.

6/15
As long as debt and a large portion of the debt-servicing cost can be continuously rolled over, Beijing can pretend that these losses don't exist.

But these losses are real, and ultimately there is no way the associated debt can be serviced.

7/15
The problem is that these losses distort both economic behavior and the functioning of the financial system in ways that will increasingly undermine growth. In finance theory we refer to this process as "financial distress costs".

8/15
Until the losses are recognized and written down, which means until they are allocated, China will find it impossible to maintain acceptable GDP growth rates without rapidly-increasing amounts of government debt to fund more problematic investment.

9/15
This means that the longer China postpones recognizing and allocating the losses, the more debt it takes to generate each unit of growth, and the more debt it takes, the worse the associated financial distress costs. It is a vicious circle which, at some point, must stop.

10/15
We are already seeing this process, and by the way, this is why China's growing trade surplus is so important for stability in the near term. It is one of the ways China can externalize a part of the financial distress costs.

https://carnegieendowment.org/posts/2025/02/the-relationship-between-chinese-debt-and-chinas-trade-surplus?lang=en

11/15
Dalio seems to be saying that Tokyo made a huge mistake in its failure to write down the bad debt quickly, and the consequence was decades of slow growth. In the early 1990s, Japan comprised 17% of global GDP. Two decades later, it was less than 7%.

12/15
Dalio is almost certainly right, and Japan is far from being the only example of a country whose economy suffered from the failure to recognize growing amounts of bad debt. There are a huge number of historical precedents.

13/15
So then why not write it down quickly? The problem is that writing down bad debt is just another way of saying that the associated losses must be allocated to some sector or other of the economy—households, businesses or governments.

14/15
But this is politically hard to do. While it makes sense economically to write it down quickly and efficiently, doing it quickly can be economically and socially very disruptive. That is why governments typically try to postpone the resolution as long as possible.

15/15
But while "extend and pretend" may be good for short term stability, it also causes the costs of bad debt to rise until they can no longer be hidden or postponed, and the longer it takes, the worse the longer term impact on the economy.