By Michael Pettis
-Originally published at China Financial Markets
(Note: This is a follow-up to an article here from May 16 and an updated version of a previously published January 2018 blog post. The original version can be accessed here. )
Suggestions by some Chinese officials that they may reduce their purchases of U.S. Treasury bonds show just how poorly the world understands the balance of payments. Here is what a recent Financial Times article had to say:
It was an unnerving piece of data for investors last week, buried halfway down an esoteric spreadsheet released by the US government that tracks how many Treasuries foreign investors buy and sell. China, the largest foreign creditor to the US government with total Treasury holdings in excess of $1.2tn, sold $20bn of securities with a maturity exceeding one year in March, according to US government data. The sales amounted to China’s largest retreat from the market in more than two years.
The article then goes on to suggest that China’s reduced holdings of U.S. Treasury bonds may reflect a strategic response to the escalating trade conflict between Beijing and Washington:
The data reignited fears that Beijing may weaponise its holdings as part of the trade war, wreaking havoc with the biggest bond market in the world, pushing interest rates higher and increasing the US government’s cost of borrowing.
“If China starts dumping its Treasuries, it would cause huge financial instability,” said Mark Sobel, a former Treasury department official who spent nearly four decades at the agency, adding that he considered this an unlikely scenario.
In January 2018, I explained on this blog why China cannot “weaponize” its holdings of U.S. government bonds. It is not because, as many observers seem to think, that selling off the bonds would cause havoc in the market and in doing so would undermine the value of China’s own holdings. This is very unlikely. First of all, the Federal Reserve could easily act to overcome any temporary volatility. Second, as another article in the same issue of the Financial Times points out, rising uncertainty is causing investors to increase their purchases of U.S. government bonds:
US Treasury yields plunged to their lowest level since 2017 and shares fell more than 1 per cent on Thursday as the deepening trade dispute between the US and China raised concerns about global economic growth.
The rush to the relative safety of government debt pushed the yield on 10-year US Treasury bonds to roughly the same level as when the Federal Reserve began raising interest rates in 2015. Longer-term rates fell below shorter-term ones, a yield curve inversion that is seen by many traders as an indication of an impending economic downturn.
I thought it would make sense to revisit and update my January 2018 post. As I explained in that entry, the real reason China cannot sell off its holdings of U.S. government bonds is because Chinese purchases were not made to accommodate U.S. needs. Rather, China made these purchases to accommodate a domestic demand deficiency in China: Chinese capital exports are simply the flip side of the country’s current account surplus, and without the former, they could not hold down the currency enough to permit the latter.
To see why any Chinese threat to retaliate against U.S. trade intervention would actually undermine China’s own position in the trade negotiations, consider all the ways in which Beijing can reduce its purchases of U.S. government bonds:
- Beijing could buy fewer U.S. government bonds and more other U.S. assets, so that net capital flows from China to the United States would remain unchanged.
- Beijing could buy fewer U.S. government and other U.S. assets, but other Chinese entities could then in turn buy more U.S. assets, so that net capital flows from China to the United States would stay unchanged.
- Beijing and other Chinese entities could buy fewer U.S. assets and replace them with an equivalently larger amount of assets from other developed countries, so that net capital flows from China to the United States would be reduced, and net capital flows from China to other developed countries would increase by the same amount.
- Beijing and other Chinese entities could buy fewer U.S. assets and replace them with an equivalently larger amount of assets from other developing countries, so that net capital flows from China to the United States would be reduced, and net capital flows from China to other developing countries would increase by the same amount.
- Beijing and other Chinese entities could buy fewer U.S. assets and not replace them by purchasing an equivalently larger amount of assets from other countries, so that net capital flows from China to the United States and to the world would be reduced.
These five paths cover every possible way Beijing can reduce official purchases of U.S. government bonds: China can buy other U.S. assets, other developed-country assets, other developing-country assets, or domestic assets. No other option is possible.
The first two ways would change nothing for either China or the United States. The second two ways would change nothing for China but would cause the U.S. trade deficit to decline, either in ways that would reduce U.S. unemployment or in ways that would reduce U.S. debt. Finally, the fifth way would also cause the U.S. trade deficit to decline in ways that would likely either reduce U.S. unemployment or reduce U.S. debt; but this would come at the expense of causing the Chinese trade surplus to decline in ways that would either increase Chinese unemployment or increase Chinese debt.
By purchasing fewer U.S. government bonds, in other words, Beijing would leave the United States either unchanged or better off, while doing so would also leave China either unchanged or worse off. This doesn’t strike me as a policy Beijing is likely to pursue hotly, and Washington would certainly not be opposed to it. Let’s consider each possibility in turn.
1) Beijing could buy fewer U.S. government bonds and more other U.S. assets, so that net capital flows from China to the United States would remain unchanged.
This would be a non-event. Beijing would in effect simply redirect its purchases from U.S. government bonds to other U.S. assets. Of course, the seller of those other assets would then be forced to deploy the proceeds of the sales elsewhere, so that directly or eventually the proceeds would be used to buy the U.S. government bonds that Beijing sold. The only thing that would change, in this case, is that Beijing would have swapped riskless U.S. assets for risky U.S. assets.
In that case, there would be no net impact on overall U.S. interest rates and a very small impact on relative interest rates. Because this outcome represents nothing more than a swap by Beijing out of lower-risk assets into higher-risk assets, with no net change in demand for U.S. assets, the result might be at most a small rise in yields on riskless assets matched by an equivalent tightening of credit spreads.
There would be no change in overall U.S. investment except to the extent that tightening credit spreads would cause a small rise in risky U.S. investments. What is more, Beijing’s decision would leave the U.S. capital account surplus unchanged, so it could not have an impact on the U.S. current account or trade deficits. Finally, Beijing’s decision would leave the Chinese capital account deficit unchanged, so it could not have an impact on the Chinese current account or trade surpluses.
2) Beijing could buy fewer U.S. government and other U.S. assets, but other Chinese entities could then in turn buy more U.S. assets, so that net capital flows from China to the United States would stay unchanged.
Again, this would largely be a non-event. The volume of Chinese capital flows to the United States would be unaffected, but there would be minor changes in the composition of assets to which the flows are directed. As in the previous case, there would be no net impact on overall U.S. interest rates and a very small impact on relative interest rates. Again, the result might be at most a small rise in yields on riskless assets matched by an equivalent tightening of credit spreads.
Again, as in the previous case, there would be no change in overall U.S. investment, except to the extent that tightening credit spreads cause a small rise in risky U.S. investments. Beijing’s decision would also leave the U.S. capital account surplus unchanged, so it could not have any impact on the U.S. current account or trade deficits. Finally, Beijing’s decision would leave the Chinese capital account deficit unchanged, so it could not have any impact on the Chinese current account or trade surpluses.
3) Beijing and other Chinese entities could buy fewer U.S. assets and replace them with an equivalently larger amount of assets from other developed countries, so that net capital flows from China to the United States would be reduced, and net capital flows from China to other developed countries would increase by the same amount.
In this case, China’s overall capital account deficit and current account surplus would remain unchanged, but there would be a reduction in its bilateral capital account deficit and current account surplus with the United States, and an increase in its capital account deficits and current account surpluses with the rest of the developed world. The reduction in the U.S. current account deficit would mean a reduction in the excess of U.S. investment over U.S. savings. If U.S. investment were constrained by an inability to access savings, this reduction would occur in the form of lower U.S. investment. Because this is not the case. Given that U.S. businesses have easy access to as much capital as they need to fund investment, the adjustment would occur in the form of higher U.S. savings.
Savings can be forced up in many different ways, almost always involving either less debt or lower unemployment. For example, a reduction in capital inflows can deflate asset bubbles and so discourage consumption through wealth effects, such a reduction can lower consumption by raising interest rates on consumer credit, or this reduction could even take place by encouraging stronger consumer lending standards. A reduction in capital inflows can also increase savings by reducing unemployment. One way or another, in economies like the United States that do not suffer from weak access to capital, a reduction in foreign capital inflows will automatically increase domestic savings.
It may be harder than we think for China to redirect capital flows from the United States to other developed economies. Continental Europe, Japan, and the UK are the only developed economies large enough to absorb a significant change in the volume of capital inflows, but none of them are eager to absorb the current account implications. Some economists, misunderstanding the nature of the account identity that ties net capital inflows to the gap between investment and savings, will undoubtedly argue that these inflows would cause investment in Europe, Japan, and the UK to rise, but this is wrong. It would only be true if investment in these economies had previously been constrained by scarce savings, but because this is clearly not the case in today’s environment, the impact of higher capital inflows into developed economies could only be to reduce domestic savings.
For developed economies, in other words, significantly higher capital inflows from abroad would either cause savings to decline as the inflows strengthen their currencies and reduce exports—causing either unemployment or consumption to rise—or, if their central banks act to sterilize the inflows, to increase imports by increasing consumer debt. If continental Europe, Japan, and the UK are unwilling to accept higher unemployment or higher debt, they would be unwilling to allow unlimited Chinese access to domestic investment and may quickly take steps either to retaliate or to redirect the flows to the United States.
In the latter case, of course, it would again be a non-event. To the extent that developed countries do not redirect Chinese capital inflows to the United States, however, Chinese sales of U.S. government bonds would affect the U.S. economy, but largely in positive ways. First of all, and contrary to popular perception, a reduction of Chinese capital flows to the United States would not cause U.S. interest rates to rise except to the extent that it would cause U.S. economic growth to pick up. Because the reduction of the U.S. capital account surplus would result in an increase in U.S. savings, this would fully match the reduction in Chinese savings that had previously been imported by the United States. This is just the logical consequence of the balance of payments constraints.
There would be no direct change in overall U.S. investment, and there would be an increase in U.S. savings, driven by either lower unemployment or a reduction in consumer debt. There might be an indirect change in U.S. investment eventually as the American trade deficit declines. Remember that because Beijing’s decision would reduce the overall U.S. capital account surplus, it would also automatically reduce the U.S. current account and trade deficits, for reasons that I discuss in an earlier blog entry. Finally, because Beijing’s decision would leave the Chinese capital account deficit unchanged, it would have no impact on the Chinese current account or trade surpluses.
4) Beijing and other Chinese entities could buy fewer U.S. assets and replace them with an equivalently larger amount of assets from other developing countries, so that net capital flows from China to the United States would be reduced, and net capital flows from China to other developing countries would increase by the same amount.
In this case, as in the previous, China’s overall capital account deficit and current account surplus would remain unchanged, but there would be a reduction in its bilateral capital account deficit and current account surplus with the United States, and an increase in its capital account deficits and current account surpluses with the developing world. As explained above, the reduction in the U.S. current account deficit would occur through an increase in U.S. savings.
There is no difference between this case and the previous one as far as its impact on the United States or on China. Interest rates in either country would remain unchanged, the U.S. trade deficit would decline, and China’s trade surplus would remain unchanged.
There is one important difference to the global economy, however. Because investment in developing countries is often constrained by difficulty accessing global savings, a redirection of Chinese capital from the United States to developing countries would boost investment in those countries. This would increase global growth and would benefit both developed economies and developing economies, including the United States. But the reason this is unlikely to happen to any large extent is that China has had a very bad experience with its investments in developing countries and may not be eager to raise them significantly more than it has already planned.
5) Beijing and other Chinese entities could buy fewer U.S. assets and not replace them by purchasing an equivalently larger amount of assets from other countries, so that net capital flows from China to the United States and to the world would be reduced.
Finally, China could reduce its overall capital account deficit by reducing the amount of capital directed to the United States and not replacing it with capital directed elsewhere. China, in other words, would export less capital abroad. This would mean, by definition, that China must either reduce domestic savings or increase domestic investment. This would also mean, of course, that Beijing must run lower current account and trade surpluses.
One way savings can decline quickly is if a drop in exports causes unemployment to rise. The only other way is if there is a surge in consumer debt. For investment to rise quickly, there almost certainly has to be either a rise in unsold inventory as exports drop or a rise in nonproductive investment in infrastructure. In either case, this would mean a rising debt burden.
As in the previous two cases, there would be no direct change in overall U.S. investment, and there would be an increase in U.S. savings, the latter driven either by lower unemployment or a reduction in consumer debt. There might be an indirect change in U.S. investment eventually as the American trade deficit declines. This is because as Beijing’s decision reduces the overall U.S. capital account surplus, it also would automatically reduce the U.S. current account and trade deficits. Most importantly for China, Beijing’s decision would reduce the Chinese capital account deficit and so it would necessarily also result in a reduction in the Chinese current account or trade surpluses.
Conclusion
Even if Beijing forced institutions like the People’s Bank of China to purchase fewer U.S. government bonds, such a step cannot credibly be seen as meaningful retaliation against rising trade protectionism in the United States. As I have showed, Beijing’s decision would have no impact at all on the U.S. balance of payments, or it would have a positive impact. It would have almost no impact on U.S. interest rates, except to the extent perhaps of a slight narrowing of credit spreads to balance a slight increase in riskless rates.
It would also have no impact on the Chinese balance of payments in the case that it leaves the U.S. balance of payments unaffected. To the extent that it would result in a narrower U.S. trade deficit, there are only three possible ways this might affect the Chinese balance.
First, China could export more capital to developed countries, in which case the decision would have no immediate impact on China’s overall balance of payments, but it would run the risk of angering its trade partners and inviting retaliation. Second, China could export more capital to developing countries, in which case the decision would have no immediate impact on China’s overall balance of payments, but it would run the very high risk of increasing its investment losses abroad. Or third, China could simply reduce its capital exports abroad, in which case it would be forced into running a lower trade surplus, which could only be countered, in China’s case, with higher unemployment or a much faster increase in debt.