China is running an enormous trade surplus but there is a clear puzzle that despite the trillions of US dollars flowing, we can find nearly no evidence of the US dollars staying in China and specifically under the control of Chinese state owned banks. In Parts I and II we showed very clearly that Chinese state owned banks show no evidence of rapid growth in US dollar assets. In Part III we showed that inflows and outflows of foreign currency actually come very close to balancing.
There was one very clear outlier and that is Chinese banks recording net inflows in transaction terms of nearly $2 trillion US dollars in recent years but no evidence the banks continue to hold those US dollars. So if total transaction flows into and out of China come close to balancing but with a $1.9 trillion net inflow of US dollars and no evidence of those dollars sloshing around China, we are left with the question: where are the US dollars from China’s massive trade surplus?
Just because a dollar does not exist in a Chinese bank does not mean it does not exist in some other form or some other institution. The question before is can we locate the other forms those USD might take if not in Chinese bank deposits.
Again, China’s State Administration of Foreign Exchange (SAFE) helps provide evidence of what is actually happening and as we have demonstrated the US dollars are not going to Chinese state owned banks. SAFE publishes a dataset called China’s International Investment Position. It breaks down different flows and types of assets and liabilities held by Chinese corporations and individuals. What it reveals is that Chinese holdings of foreign assets have increased, very much in line with the broader trade surplus but no evidence that the foreign currency holdings are being held as US deposits at Chinese state owned banks.
From the beginning 2022 to the end of 2025, China’s net international investment position, growth in assets less growth in liabilities, grew by $1.83 trillion USD. If that number sounds familiar, it should. As noted in the last piece unpacking the flows of foreign currency into and out of China, we noted, that SAFE found a net inflow of US dollars totaling $1.9 trillion USD with no corresponding out flow of USD even though total cross border currency flows roughly balanced. In other words, the large Chinese current account surplus is showing up not on the balance sheet of Chinese state owned banks but in the rapid growth of Chinese investment assets outside of China.
However, SAFE gives us much more insight beyond where China is putting its hard currency. China’s international investment holdings of deposits only grew by $103 billion US dollars and financial sector holdings of debt instruments grew by only $4 billion. However, China’s direct investment grew by $809 billion and portfolio investment grew by $995 billion. Put another way, even if we exclude the financial sector share growth in direct investment assuming it is all held by Chinese state owned banks, non-financial sector growth in direct and portfolio investment is responsible for 92.3% of the increase in China’s net international investment position. In other words, the US dollars China is getting from its massive trade surplus are not going to Chinese state owned banks but into international investment assets.
Breaking it down even further, the large amount of non-financial sector direct investment is in equity holdings with nearly 68% of direct investments being equity holdings and the remaining 32% being debt. The $995 billion in portfolio holdings show a similar pattern with $621 billion or 62.4% and the remaining in debt portfolio holdings. So unless Chinese state banks have become the asset manager of off balance sheet special purpose vehicles, even if we incorporate all of the debt security assets from portfolio investment, a very unlikely assumption for many reasons, we can say with a high degree of certainty that Chinese state owned banks simply are not the primary asset manager of China’s non-RMB asset holdings.
There are a few final notes on this. First, Chinese data should always be examined skeptically, however, we must start with the assumption the official stated data is correct and proceed to work around the official data to prove or disprove the data. One should never start from the assumption the data is false and then seek to prove this assumption. What has been demonstrated is that multiple Chinese banks and official regulatory data of different types show broadly consistent data on Chinese state owned banks holdings of US dollars and non-RMB asset from multiple different datasets. This does not absolutely prove the Chinese banks data is true but it makes it much harder to argue the data is incorrect or false.
Second, the data on asset holdings closely match the current current account surplus data published by China and aggregated via other countries data. We do not yet have detailed granular data on where the investment is going or who is holding it, something we will turn to in the next piece, however, we can say with high confidence where the current account surplus dollars are not being held and at least some organizations who are not managing those US dollars.
The real question now becomes not how much the Chinese state owned banks can buffer the Chinese RMB policy, not very much, but a potentially more dangerous question where are all the Chinese US dollars going and who controls them?