The idea that China could sell its vast horde of U.S. Treasuries in retaliation for the Trump administration’s tariffs has been called the ‘nuclear option’ in the trade war. But new data suggests it could be more of a bottle-rocket.
An article that ran in Politico and the South China Morning Post last week explained the fear. “China could fire back by dumping its vast holdings of U.S. government debt. Flooding the market with treasuries would push down US bond prices and cause the yields to spike,” it claimed.
Data released this week shows that China cut its Treasury holdings by nearly $20.5 billion in March, the fastest pace of selling in more than two years. China has reduced its holdings all but three months since last June–when President Donald Trump slapped a 25 percent tariff on $50 billion of Chinese imports.
But far from hurting the ability of the U.S. government to issue debt or driving down the price of Treasuries, China’s selling has been accompanied by Treasury prices moving up and yields moving down. The yield on 10-year Treasuries was 3.09 on November first. On Wednesday it was 2.37.
In March, the 10-year yield fell from 2.76 to 2.41. Yields move in the opposite direction of prices, so falling yields indicate bond prices rising.
The bond sales represent just a small portion of China’s $1.12 trillion Treasury holdings. But taken together, the sales over the 12-month period ending in March, amount to a 5.6 percent reduction. That is a significant change in China’s holdings with no impact on bond prices.
We’ve been through this before. In 2016, China sold about $188 billion of Treasuries, 15% of its portfolio. Treasury yields hit record lows.
If China were to attempt to sell its entire portfolio all at once that probably would have at least a temporary effect on yields. That much supply coming into the market all at once would set traders scrambling and push down bond prices until an orderly market could be restored and prices would recover. But in the meantime, it would inflict significant financial losses on China, the largest foreign holder of Treasuries, depleting its foreign currency reserves.
Most forecasters expect China would take a gradual approach to shrinking its holdings. According to Jeff Cox at CNBC, UBS estimates that a big reduction in China’s holdings would result in the 10-year yield rising just 0.4 percent. That would still leave yields below where they stood last summer–hardly the equivalent of dropping an atomic bomb on the U.S. economy.
Ironically, heavy Chinese selling could actually send yields downward, at least temporarily, if investors saw it as a sign of an escalation in the trade-war.
“To the extent that China’s Treasury sales could be either the cause or the effect of a more risk-averse global environment, the positive impact on Treasury yields could be smaller than estimated if private investors step up their Treasury purchases,” UBS strategist Chirag Mirani and others said in a note to clients, according to CNBC.
That’s what happened last week. On the Friday before Trump announced the tariff hike, the 10-year yield was 2.54. It fell to 2.40 on the first trading day after the higher levies took effect.
China’s significance in the Treasury market has declined in recent years. At its 2012 peak, China’s holdings amounted to 12 percent of U.S. debt outstanding. Now it is just 7 percent.
It’s also helpful to remember that an average of $500 billion or so of Treasuries trade hands every day. So China’s entire holdings amounts to under two-and-a-half days worth of volume,
Most likely, the recent selling has not been an attempt to drive up the borrowing costs of the U.S. government but to prop up its own currency. The yuan has depreciated 7.3 against the dollar over the past year. The makes its exports cheaper, offsetting at least some of the effect of the U.S. tariffs, but it also reduces the buying power of Chinese consumers and businesses and could prompt capital flight from the country.
Foreign accounts sold a total of $12.6 billion in March, the 11th consecutive month of selling. But this has had little, if any, effect on yields, which bounced around a bit last year and have been falling ever since the Fed changed its policy from gradual hikes to patience.
The lack of response in bond yields to foreign sales should not be surprising. In general, the long-term yield of U.S. government bonds reflects the expected path of short-term rates, which in turn is based on where investors think the Fed will set the overnight rate. Think of a long-term bond as being composed of a series of short-term bonds.
Inflation expectations may play a role because investors presumably would demand higher returns if they expect inflation to rise. But since the Fed targets inflation, it’s not clear that inflation expectations do anything that expectations for short-term rates are not already doing.
There’s also an odd sort of risk premium built into the price of Treasuries, according to economists. Investors cannot be certain their predictions about short-term rates are correct, especially five years or so into the future. So economists think bond prices include a “term premia,” a little extra-return on longer term bonds that would cushion the cost of getting future rates wrong.
For example, right now the yield on the 10-year is around 2.4. Suppose that the interest rate on the 1-year U.S. Treasury is expected to be around 2.2 percent over the next 10 years. Then the term premium on the 10-year U.S. Treasury note would be about 0.2 percent. Note that since investor expectations cannot be directly observed, calculating the term premia is a lot more complicated than that. In fact, economists cannot even agree with each other on how it should be calculated.
Not only do economists not agree on how to calculate the term premia, some are skeptical that it even exists, arguing that long rates are fully determined by short rates expectations. In this view, the term premia is an illusion created by the mismeasurement of expectations. Others say that it exists but is caused by something other than uncertainty. Perhaps imbalances between supply and demand, tax distortions, or some types of investors preferring certain durations over others for idiosyncratic reasons.
There exist at least a few bond traders who maintain that the term premia is 100 percent reflexivity: the term premia is created solely by expectations of the term premia itself. That is, the term premia rises and falls in accordance with investor expectations that future investors will demand a term premia.
None of this leaves much room for foreign investors attacking the U.S. by selling government bonds. China cannot force bond yields up on anything more than a very short-term basis, if at all, by selling Treasuries. If China no longer wanted to accumulate American dollars and dollar-denominated assets, it would need to reverse the enormous trade surplus it runs with the U.S. That is extremely unlikely.
China’s bond holdings are not so much a nuclear arsenal it can wield against the U.S. as they are fireworks that might garner some attention but not do much harm.