If US investors start buying foreign securities like they buy foreign cars, watch out

There is no way to put a positive gloss on the December TIC report.  Total inflows -- counting short and long-term flows -- were negative.   The foreign securities bought by US private investors abroad exceeded the US securities bought by private investors abroad; net private were very negative.    Long-term private outflows (net) were around $8.4b, and short-term private outflows (net) were $34.1b.    Absent substantial net official inflows ($24b of long-term purchases, $31.5b total) the overall totals would have been far worse. 

Isabelle Lindenmayer  of the Wall Street Journal put it well.

"The TIC report suggests U.S. investors are starting to find confidence and value in foreign securities much the same way U.S. consumers have long found reliability and thrift from Toyotas and Hondas. Just as worrisome, foreign investors, too, may be starting to lose their appetite for U.S. securities.

... Though the TIC data are an imprecise gauge, they don't bode well for the U.S. ability to finance its huge trade and current account deficits in the long term, said analysts."

Our unnamed analysts are stating the obvious.  To finance a roughly $850-900b current account deficit, you need need over $70b of net inflows a month.   Not zero.  And certainly not a net outflow of $11b.

The disaggregated data continues to frustrate.  Overall, according to the TIC data, official investors invested $128.7b in the US in 2006 - up from $87.3b in 2005 but down from $341.6b in 2004.  Net long-term official inflows were a bit higher -- $185.6b -- but still below 2004 levels ($235.6b).   The big fall in short-term holdings clearly reflects the $38.5b fall in Russian short-term holdings (Russia moved a lot of dollars offshore in 2006. 

Relative to 2004, there was a clear shift toward agencies.   

Only the last point really fits other data sources.   Based on the COFER data, along with trends in the BIS offshore deposits, Christian Menegatti and I (RGE subscription required, sorry) estimated that official institutions added about $550b to their dollar holdings.   A lot of that will be held in banks offshore.  But I still suspect the US data is dramatically understating total foreign official purchases.

A couple of examples.   China added $78b to to its reserves in q4, $68b after after adjusting for valuation changes.   Recorded Chinese purchases of long-term US securities in q4 though were only $20.8b.  China reduced its short-term holdings by $10.6b in q4, so net inflows from China in q4 were only $10.2 (assuming I did the  math right).

That is way too low for a country that added $65-70b to its reserves -- especially when Chinese banks and Chinese pension funds were also likely buying dollars (remember, Chinese q4 reserve growth was well below the estimated q4 current account surplus + net FDI inflows).

The "Asian oil exporters" -- TIC slang for the Gulf -- didn't do much in q4 either: $2.7b in purchases of long-term debt were offset by a $2.7b fall in their short-term holdings.  Oil prices did fall in q4, but, well, the Saudis and the others in Gulf were still clearly adding to their foreign assets.

Voldemort, indeed.  (More here)

Relative to $200b plus in q4 reserve growth, the $34b in net official inflows in the q4 data is very, very low.  It is possible central banks jsut stopped buying dollars.   But all the available evidence suggests that central banks tend to buy dollars when others don't want to.  Others clearly didn't want to in December.   And if that pattern holds, total central bank dollar purchases were way, way up in q4 2006.   Christian and I estimated that central banks needed to add around $150-155b to their dollar portfolios in q4 to hold the dollar share of their reserves constant.  Either we are way off, or most of those purchases didn't show up in the TIC data.