And it isn’t hard to see how the q3 and q4 current account deficits could be large enough to bring the annual total up to around $900b.
The July trade deficit was $68b. Suppose the monthly trade deficit stays constant – the US oil bill is likely to fall a bit in October, if not before. But I am not yet totally convinced the non-oil balance has stabilized (the pace of deterioration clearly has slowed). That works out to quarterly trade deficit of $204b.
The deficit in income payments and transfers was around $25b in q2. Suppose it rises to $30b in q3 and q4. Personally, I think it will be bigger in q4 … more debt and higher interest rates usually mean more interest payments.
That would imply a current account deficit of around $235b in q3 and q4, which, combined with a $213b deficit in q1 and a $218b deficit in q2 would push the overall deficit up to around $900b.
The current account data release doesn’t tell us much about the trade deficit. But it does tell as a bit about the income balance – and the sources of financing that have allowed the US to run large ongoing deficits. A few things caught my eye ….
Second, the income balance continues to deteriorate more slowly than I would have expected. What is going on?
First, the gap between what the US earns on its direct investment abroad and what the US pays on direct investment in the US continues to grow. It was $36b in q2 – a $144b annual pace. That compares with $134b in 2005. If you dig into the data the BEA now releases electronically, it seems that dividend payments on foreign direct investment in the US remained at an extremely low level in q2 (US firms made four times the dividend payments of foreign firms, and trust me, the gap comes because dividend payments on investment on FDI in the US are really, really low – they are running at about ½ their 2005 level in 2006). Combine a growing gap on dividends – largely because foreign firms operating in the US don’t seem to be making money – and the persistent gap in reinvested earnings and, well, you get the data that appears in the BEA release. Call it dark matter. Or call it data that doesn’t make much sense. Take your pick.
Second, interest payments on US external debt are rising rapidly. The sum of “other private payments” and “US government payments” offers a decent proxy for US interest payments. Other payments include dividends on foreign holdings of US stocks, but those are small. This measure of interest payments rose from $79b in q2 2005 to $119b in q2 2006 – an increase of around 50% (or $40b). Higher short-term rates are having an impact.
But the US lends as well as borrows. It doesn’t lend as much as it borrows, but it does lend. Other private receipts is one proxy for US interest income. It rose from $50.7b in q2 2005 to $79.8b in q2 2006 – an increase of nearly $30b. The percentage increase (57%) was actually higher than the percentage increase (50%) on US borrowing. That – I think – reflects the fact that most US lending is denominated in dollars and is very short-term. It has “repriced” faster than the US debt, slowing the pace of deterioration in the US income balance.
Still, net interest payments (using my proxy) rose by over $10b ($40b annualized) between q2 2005 and q2 2006 … that should continue.How did the US finance its roughly $430b ($860b annualized) current account deficit in the first half -
It is a bit of a mystery actually. Net purchases of US long-term debt have been trending down. See Jay Bryson of Wachovia.
Net FDI flows were slightly negative – US direct investment abroad (counting reinvested earnings) exceeded foreign direct investment in the US (counting reinvested earnings) by about $15b. That is probably off, as the earnings on FDI in the US seem to low to be fully believable (reinvested earnings are probably higher than reported), but let’s use it.
Americans also bought more foreign stocks than foreigners bought stock (data here), but only just. Total outflows (net) in the first half were around $6b. US equity outflows slowed in q2 – but so did foreign purchases of US equities.
In aggregate though, the US once again financed its deficit entirely by the sale of debt.
OK, not entirely.
The US needed $450b or so in net purchases of US debt to cover a deficit of around $430b and net equity outflows of $20b. It only got $335b or so though.
Errors and omissions (around $110b) made up the difference. The sums don’t quite add up because of rounding, and because some of the data series aren't totally harmonized.
Net private purchases of US debt securities (overwhelming corporate and agency debt, there were virtually no net purchases of Treasuries) provided around $210b in debt financing (data here, but that data is totally harmonized with the data in the main release, which is consistent with more like $195 in net debt inflows).
Foreign central banks provided another $150b (data here). About $40b of that came in through the banking system (in q2), $55b came from purchases of agencies, $30-35b came from treasury purchases and maybe $20b from purchases of corporate debt (other official assets).
Add in a $20-25b net outflow from the financial system (other claims reported by banks and non-banks) and everything should more or less sum up.
I suspect that the Treasury rally in August and September coincides with some pickup in foreign demand for Treasuries. But the fall in off in overall demand for Treasuries – whether from central banks or private investors – in the first half of 2006 continues to be striking.Sorry for the unstylish data reporting. It is more fun to talk about dark matter. But it takes time to turn data in dark matter and dark anti-matter …