Some days are marked by a surplus of good material. The TIC data. The second quarter current account data. DeLong's magnus opus on what the great and good talked about on the margins of Jackson hole. A series of IMF papers (including an excellent paper from Laxton and Milesi-Ferretti). The Economist on the one of the IMF papers (good). The Economist on Asian reserves (not so good). Evidence of convergence between Stephen Roach and Richard Berner. Stephen Jen remains a holdout. I share the Roach/ Berner sense that the US current account deficit is set to rise a bit further from current levels, and that financing it may be a bit more difficult in 2006 than in 2005. But then again, that always has been my sense.
Turning to the Q2 current account data, five points:
First, $196 billion is a large deficit, even if it is a tad smaller than the (revised) $199 billion Q1 deficit. I feel pretty confident in my $820 billion 2005 current account deficit forecast. If the August trade deficit is a record, as Calculated Risk expects based on oil prices and West Coast port data, I might even revise that estimate up. The current account deficit fell in q2 because the deficit in goods and services trade stopped rising. I think that is temporary - even if non-oil imports stay flat, oil imports are rising. And transfers fell off their q1 peak levels.
Second, the income line in the current account deficit finally turned negative in the second quarter, if only by a small amount. A couple of additional comments here: First, the interest paid by the US government on Treasuries held abroad rose from 42.16 b in the first half of 2004 to $53.85 b in the first half of 2005 - an increase of 28%. Expect that to continue; Second - and this is something that I wish proponents of the "US is a great place to invest, that's why we have a current account deficit" view would wrestle with seriously -- the US continues to earn far more on its direct investment abroad ($120 b in the first half of 2004) than foreigners earn on their direct investment in the us ($62.5 b). The US has a bit more direct investment abroad and foreigner have investment in the US, but not two times as much. The US just gets more income from investing abroad than foreigners get (or report?) on their investment in the US. I don't quite see how the low realized return on direct investment in the US offsets the exchange rate risk associated with investing in a major deficit country ...
Third. In the second quarter, US direct investment abroad once again exceeded foreign direct investment in the US. But the pace of US FDI abroad, $61 b in the first half of 2005, continues to be well off the 2004 pace ($252 b for the full year). That reduces the amount of debt the US has to place abroad.
Fourth, official inflows picked up substantially in the second quarter - jumping from $25 b in q1 to $82 b in q2. For the last four quarters, official inflows have totaled $278 b even without Japanese reserve accumulation. And I think that total understates the full contribution foreign central banks are making to the financing of the US deficit. It leaves out offshore dollar deposits by central banks that help finance private inflows, and it is not clear to me that the US data is picking up all central bank inflows either. Reports of the end of central bank support have been exaggerated. In q2, foreign central banks build up their on shore bank accounts (up by almost $34b), so they provided over $80 b in total financing despite only buying a bit less than $23 b of Treasuries.
A small rant: the TIC "stock" data shows total central bank holdings of treasuries (bills and notes) falling by $5.3b in q1, and rising by $5.9b in q2. The BEA flow data shows a much larger inflow -- $14.3 b in q1, $22.7 b in q2. Their "securities" data does not match the TIC data for inflows into Treasuries in q2 (notes purchases of $37.4b according to the July TIC release), so it must be including the q2 fall off in bills (-30.6b). But for the life of me, I cannot figure out how these numbers match up. Help appreciated.
Fifth, both overall inflows and overall outflows were way up in q2 - largely because of a surge in claims on US banks which were offset by a surge in US bank claims on the rest of the world. Those flows usually offset, so I think they (and claims reported by non-banks) should be netted out for most purposes. Making that adjustment, foreigners invested $224.6 b in the US, and Americans invested $70.2 b abroad. That left $154-155 b to finance the US current account deficit. Not enough in other words. The gap is even bigger if you include all the bank flows: the "gap" (or statistical discrepancy) between the amount needed to finance the current account deficit and US outflows and recorded inflows was $54b. The gap in the first two quarters of this year exceeds the gap from all of last year -
I don't know what that means. But it is puzzling.
Turning briefly to the July TIC data:
1) China bought $12.6 b of long-term US securities in July -- $6.65-6.7 b of Treasuries, $4 b of Agencies, $1.95 b of US corporate debt. It also bought $1.3 b of "foreign" bonds from US residents - perhaps world bank bonds, perhaps something else. It financed those purchases in part by running down its holdings of Treasury bills - its T-bill holdings fell by $8.25 billion. All told, the net recorded inflow from China was $5.6-$5.7 b. Compare that with reserve accumulation of probably $20 b in July. For the year, Chinese purchases of t-bills and long-term debt securities (including Chinese purchase of "foreign bonds" recorded in the US data) provided $54.3 b in financing to the US. That is well under 50% of my estimate for the increase in the China's reserves (including reserves transferred to the banks).
2) Japanese demand for US debt was reasonably strong in July - Japanese investors bought $9.8 b of Treasuries, Agencies and corporate debt. $7.9b of that total came from Agency purchase.
3) The UK (meaning god only knows) continues to be the biggest source of demand for Treasury bonds ...