China, by population, is larger than Japan by ten times. So not sure how appropriate to compare the two economies. Similarly, China is larger than the US by nearly four times. So not quite fair to expect a fair competition.
One soccer team had 10 players, the opposing team just one player. Is that a fair competition? Likewise, Four workers competing against one worker is not a fair comparison. Nigeria is a basket case, no argument there.
I agree that China faces some genuinely Japan-like risks, especially weak domestic demand, demographics, balance-sheet adjustment and declining returns on parts of the capital stock even with the expansion of the new technology sectors. But I am less convinced that the capital-output ratio itself tells us very much about the eventual outcome.
Strictly speaking, capital deepening refers to rising capital per worker, K/L, while the chart here measures K/Y. A rising K/Y means declining average capital productivity, which is certainly worth watching, but it does not by itself imply Japan-style stagnation. South Korea’s capital-output ratio also rose to roughly 5.1 by 2023, close to Japan and China, without producing three lost decades.
The more important comparison, in my view, is therefore China and Japan’s productivity regimes, not simply their capital-output ratios. Japan entered the 1990s already rich, highly urbanized and close to the technological frontier. China still has substantial catch-up potential in services, automation, regional productivity and capital allocation.
So the key question is whether China’s current investment in AI, robotics, electrification and advanced manufacturing merely adds more low-return capital, or whether it eventually raises economy-wide TFP. That distinction will tell us far more about China’s “Japan problem” than K/Y alone.
I think that your assessment of Chinese investment is too negative, Stephen.
It is notoriously difficult to build capital stocks. A simpler and cleaner alternative is to look at the marginal product of capital (MPK) -- the inverse of the ICOR. The MPK can easily be calculated as the real GDP growth rate divided by the investment-to-GDP ratio. All the data needed are available in the IMF's WEO database.
In 2025, China's GDP grew by 5 percent, and the investment-to-GDP ratio was 0.388, so its MPK was 13 percent.
Averaging over the period 2020-25, China's MPK was 12 percent, 6th highest among the G20 and slightly better than the US at 11 percent.
China's MPK has fallen from 17 percent over 2010-19. This is to be expected as development progresses, especially as investment-intensive development as China undertakes. However, by this metric it is hard to argue that China's investment is inefficient.
There is no growth in China. The country is still struggling in deflation after the property market collapsed. With all these “ghost residential compounds” or incomplete construction sites, all local govt are in deep red.
This is one driver for the recent change in tax law to recoup 20% of income tax on all profits from overseas assets held by Chinese nationals. This crawl back to 2000, 25 years of unpaid taxes.
Secondly, the new law to prohibit Chinese nationals from leaving the country if you work in tech, AI, EV, high end tech industry deemed national security related,
Excellent analysis, Prof: Roach. The capital-output ratio comparison is a powerful and somewhat under-appreciated lens on China’s sustainability challenges.One additional distinction worth noting is how the two economies have managed their external capital positions. China has actively reduced its US Treasury holdings over the past decade, increased gold reserves, and maintained a managed yuan—moves consistent with a broader de-dollarization strategy and an effort to contain the domestic debt burden. In what I deem capital markets wars. Japan, by contrast, has maintained a more symbiotic capital-market relationship with the United States, most recently illustrated by coordinated yen intervention.These differing external strategies do not negate the internal capital-deepening risks you highlight, but they may shape how each economy absorbs or transmits future shocks.
China, by population, is larger than Japan by ten times. So not sure how appropriate to compare the two economies. Similarly, China is larger than the US by nearly four times. So not quite fair to expect a fair competition.
Nigeria is 3 times the population of Japan and comparable to the US in population. So?
One soccer team had 10 players, the opposing team just one player. Is that a fair competition? Likewise, Four workers competing against one worker is not a fair comparison. Nigeria is a basket case, no argument there.
I agree that China faces some genuinely Japan-like risks, especially weak domestic demand, demographics, balance-sheet adjustment and declining returns on parts of the capital stock even with the expansion of the new technology sectors. But I am less convinced that the capital-output ratio itself tells us very much about the eventual outcome.
Strictly speaking, capital deepening refers to rising capital per worker, K/L, while the chart here measures K/Y. A rising K/Y means declining average capital productivity, which is certainly worth watching, but it does not by itself imply Japan-style stagnation. South Korea’s capital-output ratio also rose to roughly 5.1 by 2023, close to Japan and China, without producing three lost decades.
The more important comparison, in my view, is therefore China and Japan’s productivity regimes, not simply their capital-output ratios. Japan entered the 1990s already rich, highly urbanized and close to the technological frontier. China still has substantial catch-up potential in services, automation, regional productivity and capital allocation.
So the key question is whether China’s current investment in AI, robotics, electrification and advanced manufacturing merely adds more low-return capital, or whether it eventually raises economy-wide TFP. That distinction will tell us far more about China’s “Japan problem” than K/Y alone.
Chat GPT strikes again!
I think that your assessment of Chinese investment is too negative, Stephen.
It is notoriously difficult to build capital stocks. A simpler and cleaner alternative is to look at the marginal product of capital (MPK) -- the inverse of the ICOR. The MPK can easily be calculated as the real GDP growth rate divided by the investment-to-GDP ratio. All the data needed are available in the IMF's WEO database.
In 2025, China's GDP grew by 5 percent, and the investment-to-GDP ratio was 0.388, so its MPK was 13 percent.
Averaging over the period 2020-25, China's MPK was 12 percent, 6th highest among the G20 and slightly better than the US at 11 percent.
China's MPK has fallen from 17 percent over 2010-19. This is to be expected as development progresses, especially as investment-intensive development as China undertakes. However, by this metric it is hard to argue that China's investment is inefficient.
Look at velocity of money, China is now 0.41. Explains it all.
There is no growth in China. The country is still struggling in deflation after the property market collapsed. With all these “ghost residential compounds” or incomplete construction sites, all local govt are in deep red.
This is one driver for the recent change in tax law to recoup 20% of income tax on all profits from overseas assets held by Chinese nationals. This crawl back to 2000, 25 years of unpaid taxes.
Secondly, the new law to prohibit Chinese nationals from leaving the country if you work in tech, AI, EV, high end tech industry deemed national security related,
This will start in 15/9/2026
Excellent analysis, Prof: Roach. The capital-output ratio comparison is a powerful and somewhat under-appreciated lens on China’s sustainability challenges.One additional distinction worth noting is how the two economies have managed their external capital positions. China has actively reduced its US Treasury holdings over the past decade, increased gold reserves, and maintained a managed yuan—moves consistent with a broader de-dollarization strategy and an effort to contain the domestic debt burden. In what I deem capital markets wars. Japan, by contrast, has maintained a more symbiotic capital-market relationship with the United States, most recently illustrated by coordinated yen intervention.These differing external strategies do not negate the internal capital-deepening risks you highlight, but they may shape how each economy absorbs or transmits future shocks.