It is hard to run $60b monthly trade deficits, and $65 b monthly current account deficits …

If foreign investors only buy $54 b of your long-term debt and equities every month.

Obviously, the US can draw on other sources of financing. FDI for one. Short-term bank flows for another. But in recent years, FDI has flowed out of the US, not into the US. And European banks are not in the business of taking in euro denominated deposits and lending unhedged in dollars. So the sale of long-term securities has been the key source of financing for the US external deficit.

Indeed, net foreign purchases of long-term US securities -- roughly $740b in 2003, $905 b in 2004 -- significantly exceeded the US current account deficit over the past two years. In effect, the US financed its current account deficit, its FDI abroad and its purchase of foreign equities by selling a ton of debt. In 2004, sales of long-term debt to foreigners were about $200 billion more than the US current account deficit ($875b v. @ $665b).

Back when the market was focused on the US external deficit, weak TIC data -- foreigners only bought 53.6 b of US long-term securities in April and net inflows only totaled $47.4b since the US bought 6.2b of foreign securities -- would have sent the dollar tumbling. Today’s data did not help the dollar, but it also didn’t put a major dent in the dollar’s rally either.

I am not sure that the data can be shrugged off quite so easily, however. True -- inflows from the sale of long-term securities were strong in January and February of this year. Consequently, annualizing this year’s four month total produces estimated 2005 inflows of $793b -- perhaps enough to finance this year’s expected current account deficit, though not enough to finance both a current account deficit and a net outflow of FDI. But if you annualize the weak March and April flows, you get a much smaller number: $528 billion. That is not even close to enough to finance this year’s current account deficit, particularly if Calculated Risk is right about the May import numbers. Money doesn’t grow on trees ...

It is possible, though, that foreigners have shifted from financing the US through the purchase of long-term US debt toward holding short-term dollar bank deposits. That is something that I want to explore.

The country level flow data presents no shortage of puzzles, not the least because it does not match the stock data . For example, China’s holdings of treasuries went up even though on a flow basis, it did not make large net purchases. Fair enough. A few months ago its flow purchases did not show up in the stock data. But this is the sort of thing that makes the TIC data frustrating.

China -- on a flow basis -- was on the sidelines, at least in the bond market. It bought 0.4b of Treasuries and 0.8b of corporate bonds, but sold $1.2b of agencies. That works out to about zero -- even though China’s reserves no doubt continued to increase. Its holdings of short-term Treasuries, however, soared -- as did "other" central bank claims. The PBOC’s stock of short-term Treasuries rose by $6.5b, and its stock of "other" claims rose by $13b (data here).

Fortunately, Japan stepped up its purchases -- buying 10.3 billion of Treasuries, agencies and corporate bonds. The Norwegians also reversed themselves. After selling $17b of Treasuries in March, they bought about $13 billion of Treasuries in April ...

Germany and Luxembourg also seem to be buying Treasuries (that interest rate differential), but not the OPEC countries, despite all their oil revenue. That remains one of the mysteries of the global market. If anyone knows where the OPEC countries are placing their surplus, do tell.