Foreign holdings of U.S. Treasuries saw two consecutive declines in the first quarter of this year
The latest monthly international capital flow report released by the U.S. Treasury Department shows that as of the end of March this year, foreign holdings of U.S. debt totaled $7.03 trillion, down $70.3 billion from the end of last month, the second consecutive month of net reductions in U.S. debt holdings, a one-year high reduction.
Last March, hit by the global pandemic of the new crown epidemic and economic shutdown, U.S. stocks melted down four times in ten days, international financial turmoil, and all realizable assets, including Treasuries and gold, were being sold off on a large scale. In that month, foreign investors reduced their net holdings of U.S. debt by $276.5 billion, the highest since March 2000. The following month, with the liquidity crunch alarm lifted and market fears eased, foreign investors re-increased their holdings of U.S. bonds. As of March this year, foreign holdings of U.S. debt peaked at $7.12 trillion at the end of January this year, but still did not exceed the previous high of $7.23 trillion at the end of February last year. At the end of March this year, foreign holdings of U.S. debt accounted for 25.0% of the U.S. Treasury balance, down 0.4 percentage points from the end of last month and 4.4 percentage points year-on-year, the lowest percentage since October 2007.
Since March last year to date, the U.S. government has introduced six rounds of fiscal stimulus bills, with a total size of $5.68 trillion. The U.S. federal deficit has soared, reaching 15.0% last year, a jump of 10.4 percentage points from the previous year. To make up for the huge fiscal deficit, the U.S. Treasury bond issue swelled significantly, with the net increase of $4.45 trillion in the U.S. bond balance from April last year to March this year, up 19%. The Fed has become the main force in the purchase of U.S. debt by offering the “zero interest rate + unlimited quantitative easing” king bomb. During the same period, the Fed increased its net holdings of U.S. debt by 1.96 trillion U.S. dollars, accounting for 44% of the increase in U.S. debt, far exceeding the net increase in foreign holdings of U.S. debt by $789 billion. By the end of March this year, the Fed held 17.6% of U.S. debt, up 5.0 percentage points year-on-year, roughly the same magnitude as the 4+ percentage point decline in foreign holdings of U.S. debt.
As can be seen above, the attractiveness of U.S. debt to foreign investors has declined, or foreign investors have contributed less to financing the U.S. government deficit, but the Fed’s asset purchases have made up for the gap. Interestingly, in March this year, the spike in 10-year U.S. bond yields once exacerbated market tightening expectations and triggered a short-term jolt in U.S. stocks. But since April, U.S. bond yields have turned down, falling from 1.70% to near 1.60%. This is not only related to the Fed’s market communication to downplay tightening expectations, but also closely related to its monetary policy operations. early April to May 19, the Fed’s net holdings of U.S. debt 125.4 billion U.S. dollars, equivalent to 1.53 times the amount of new U.S. debt over the same period, much higher than the first quarter of the Fed’s net holdings equivalent to 66% of the level of new U.S. debt.
Foreign holdings of U.S. debt are not necessarily related to the strength of the U.S. index
A correlation between the size of foreign holdings of U.S. debt and the monthly average Intercontinental (ICE) dollar index after removing the natural logarithm shows a low negative correlation of 0.365 between March 2000 and March 2021, i.e., over a longer period of time, net foreign holdings of U.S. debt increased and the U.S. index weakened. However, the picture is not quite the same when viewed in stages. For example, from January to December 2014, the two are highly positively correlated at 0.795, meaning that foreign investors increase their holdings of U.S. debt and the U.S. index moves higher; from July 2015 to December 2017, the two are highly negatively correlated at 0.801; and from March 2020 to March 2021, the two are also highly negatively correlated at 0.815.
In March last year, foreign investors reduced their holdings of U.S. debt in a single month to a record high, but the U.S. index also rose above 100 that month to a new high in more than three years amid panic market selling of all realizable assets, including U.S. debt, and a flight to dollar liquidity. In January this year, foreign investors increased their holdings of U.S. debt, but the monthly average U.S. index fell 0.3% YoY; in February and March, foreign investors reduced their holdings of U.S. debt, and the monthly average U.S. dollar index continued to move higher YoY, accumulating a 2.0% increase.
There are many factors affecting the market exchange rate trend, and the two waves of US index strength from the beginning of last year to the present have been dominated by different factors.
In March last year, the financial turmoil triggered a rise in market risk aversion, with the U.S. S&P 500 Volatility VIX Index growing 1.94 times month-over-month, driving the U.S. index strongly above 100. After that, market risk aversion eased and the positive yield spread on U.S. bonds narrowed relative to major Treasuries, guiding the U.S. index lower by degrees. Throughout last year, the average daily 10-year U.S. bond yield spreads relative to Japanese, German and U.K. Treasuries narrowed by 136, 99 and 69 basis points, respectively, compared to the previous year (see Figure 5).
In March, the daily average 10-year U.S. bond positive yield spreads versus German and U.K. Treasuries widened by 41 and 15 basis points, respectively, compared with December last year. From early May to May 20, the daily average 10-year U.S. bond positive yield spreads relative to German and U.K. Treasuries decreased by 13 and 5 basis points, respectively, compared to March.
Of course, differences in vaccination and outbreak control among major countries have also been important influencing factors in the movements of major currency exchange rates so far this year.
It is inappropriate to use changes in foreign holdings of U.S. debt to study U.S. index movements
An important presupposition of this approach is to equate changes in foreign holdings of U.S. debt with overall U.S. cross-border capital flow conditions. This is not the case.
Foreign holdings of U.S. debt by the U.S. Treasury include both official and private holdings, so foreign holdings of U.S. debt are not equivalent to foreign central bank holdings of dollar-denominated foreign exchange reserves. Looking at the share of foreign holdings of U.S. debt equivalent to global dollar foreign exchange reserves, the share has fallen from 110%-150% from early 2009 to the end of 2016 to near 100% now. Clearly, for at least eight of the twelve years since the onset of the financial crisis in 2008, global dollar reserve holdings have made it difficult to accurately explain the movement in foreign holdings of U.S. debt.
International Monetary Fund (IMF) statistics show that the global share of dollar foreign exchange reserves was 59.0% at the end of last year, down 1.7 percentage points from the end of the previous year, and the share of dollar reserves was the lowest since 1999. But this is more a reflection of the results of the weakening U.S. index last year than necessarily its cause. In fact, last year, the IMF disclosed that the global share of seven non-dollar reserve currencies have increased compared with the previous year, including: the yen and the Swiss franc in terms of the original currency of the reserves decreased compared with the end of the previous year, its share rose solely as a result of the appreciation of the exchange rate against the U.S. dollar; the Australian dollar, the euro and the yuan in terms of the original currency of the reserves increased less than the increase in the exchange rate against the U.S. dollar, so the exchange gain contributed more to its share rise.
Correlation analysis of the global dollar reserve share and the quarter-end dollar index after taking the natural logarithm shows a moderate positive correlation of 0.694 from the first quarter of 1993 to the fourth quarter of 2020, with a high positive correlation of 0.831 from the third quarter of 2018 to the fourth quarter of 2020, i.e., the dollar reserve share rises as the dollar index strengthens. Last year, the dollar index fell by 6.7% for the year and the dollar reserve share retreated by more than 1 percentage point; in 2017, the dollar index fell by 9.9% and the dollar reserve share retreated by 4.0 percentage points. The dollar index is expected to rebound strongly in the first quarter of this year, and it is likely that the global dollar reserve share will rebound in that quarter.
Furthermore, changes in foreign holdings of U.S. debt do not fully reflect the cross-border capital flow situation in the United States. For example, in 2020, U.S. net international capital inflows were $537 billion, up 6.92 times year-on-year, of which $658 billion was net private sector inflows and $121.1 billion was net official sector outflows; in the first quarter of 2021, U.S. net international capital inflows were $325.8 billion, up 1.69 times year-on-year, of which $245.7 billion was net private sector inflows and US$80 billion.
Last year, the U.S. experienced large inflows of international capital and a depreciation of the dollar, and in the first quarter of this year, U.S. net capital inflows continued to be large in the chain, but with an appreciation of the dollar. It is thus clear that even if one could use the movement of foreign holdings of U.S. debt to observe U.S. cross-border capital flows, it would be difficult to accurately predict or explain the movement of the dollar exchange rate.