China's March trade surplus topped $11b - and its goods exports were a very impressive $78 billion. China February surplus was about $2.5 billion, and its exports only 54.1 billion. China had an impressive surplus in January, when its exports were $65 billion ...
I mention these numbers because I suspect the big fall in Chinese exports in February holds clues to interpreting the US February trade data.
February is a short month, and the US tends to import more in the fall than in the winter. The holiday shopping season. That is why seasonal adjustment matters.
But the US data - even with seasonal adjustment - sometimes shows a bit of month to month volatility. Particularly in the first quarter. Often because of the Chinese new year.
Last year, non-oil imports fell by about $6 billion in March (compared to February). Chalk it up largely to the impact of the Chinese new year - though there was also a slowdown in non-oil import growth/ an inventory correction going last winter and spring.
This year, the Chinese new year showed up in the February data. Non-oil imports fell by about $4.6 billion, driving a $4.2 billion fall in overall imports. That fall explains the fall in the United States overall deficit, which shrank from a very large $68.6b to an almost as large $65.7b.
Nouriel identifies a couple of reasons why high oil prices might be a bigger drag going forward -
- The current run-up in oil prices seems to be driven at least in part by concerns about supply. Iran. Nigeria's delta. And, more generally, the folks that have oil are far more interested in renegotiating the terms of their existing contracts with big oil companies than in reaching agreement with the big oil companies on a new round of investment.
- And the ever-resilient US consumer may not prove to be quite as resilient. Home prices may still be rising, but sales aren't. Since there hasn't been any shortage of investment in residential real estate in the US, that should eventually have an impact on prices. In 2003, 2004 and 2005, consumers tapped into their rising home equity (or just saved less as their homes appreciated) and kept on spending even as oil prices rose. Nouriel thinks that this process won't continue, whether because the Fed doesn't want it to continue or because US consumers will burn out, and lose their desire to take on mor debt.
That is still a rather condensed version of Nouriel's argument. I certainly wouldn't rule it the scenario he describes. But I have been very struck by one thing that Nouriel doesn't put a lot of emphasis on. Call it the oil savings glut.
There is no doubt that the spare savings of the world's oil exporters - savings in excess of their investment - is now enormous. And with oil at $70, it will only get bigger.
And there is no doubt that the United States' need to borrow savings is enormous. And also it too is getting bigger.
I suspect one of the reasons why oil didn't exert more of a drag on the world economy is that the US had - by that time - entered into a cycle of expansion fueled by a surge in residential investment and rising consumption spurred by consumers' ability to borrow against rising home values.
The oil exporters spare savings stepped into the breach left by the reduction in the pace of Asian central bank intervention. Non-Chinese Asian central bank intervention that is. China is a special case: its current account surplus grew even as its oil import bill grew.
By holding US real interest rates down, the oil exporters reinforced a process that got started with the Fed cut rates, and got further fuel from Asia's unwillingness to allow their currencies to appreciate against the dollar from 2002 on.
There is a certain lovely symmetry:
The countries with the highest propensity to save - China and the oil exporters - financed the country with the highest propensity to borrow in order to spend.
Everything worked out.
If the US had been less willing to spend, all the oil savings - about ½ the oil windfall has been saved, according to Alessandro Rebucci and Nikola Spatafora of the IMF (try their podcast too) and in some key gulf countries, about three-quarters of the windfall has been saved - would have implied a shortage of aggregate demand.
And if the oil exporters had been more willing to spend, that too would have created problems. The oil exporters don't buy American. Their spending would have benefited Europe and Asia, not the US. The US - Guy de Jonquieres not withstanding - doesn't have an economy geared around the export of manufactured goods. The US is really good at producing debt - and splicing and dicing that debt into collateralized debt obligations, collateralized mortgage obligations, synthetic CDOs and the like - and then selling that debt to foreigners. But not so good at producing goods the oil exporters want to buy.
So long as the oil exporters are willing to buy US debt, though, everything works out. The oil shock didn't preclude the US from continuing to do what it is good at (producing debt). It just meant that the US sold its debt to a new group of buyers.
Moreover, as Rebucci and Spatafora note, a permanent increase in the price of oil increases the wealth of the oil exporters, not just their current income. So equilibrium requires not just that the US borrow against the rising value of their homes but also that the oil exporters not borrow against the rising value of their oil reserves, and instead finance the US.
And to be honest, that is what I think happened.
But there is one part of this story that is deeply unsatisfying: the absence of any evidence showing that oil exporters are buying long-term US debt.
The flow of financing from Japan to the US from 2002 thourgh 2004 was very visible, the flow of financing from China to the US is pretty visible too, though I suspect China provides more financing than shows up in the US data. The flow of financing from the Gulf to the US, by contrast, is invisible.
The formal reserves of the big Gulf states aren't growing.
And according to the US data, they aren't buying US securities.
Yet David Lubin of HSBC convincingly has argued that the Gulf states are building up their deposits in the international banking system either. Higher deposits in the international banking system from the Gulf have been roughly offset by higher loans from the international banking system to borrowers in the Gulf.
He thinks the Gulf states' surplus has been invested in securities - the banks have been disintermediated. The IMF's data tells the same story. In aggregate, the oil exporters are buying more securities this time around, and putting far less into the international banking system.
It just doesn't show up in the US data.
However, if you look carefully, you can find traces of this flow.
Look, for example, at the debt the US sold to "UK" investors. If the Saudis buy US debt through a custodial account at a UK bank, it shows up as the purchase of US debt by a private buyer in the UK.
Or look at the growth in non-reserve assets of the Saudi Monetary authority, including their holdings of international securities. They just report this data in riyal ...
Or look at the surge in Russian holdings of short-term claims on the US. That doesn't quite fit the story though - since Russia seems not to have bought much long-term debt.
The fact this flow is invisible - or least well-hidden -- means that it hasn't gotten the attention it deserves. You cann't go to the Treasury web page and get the data needed for a story on petro-sheik financing of the US.
But I am pretty sure this flow is big and important.
That makes me a bit more cautious than Nouriel. Higher oil prices means that the oil windfall will keep on growing faster than oil states can spend it. Which means more oil savings. And more demand for international debt.
And that plays into the United States competitive advantage.
Nothing precludes - or so it seems to me - a repeat of 2005. The US keeps on churning out debt that the Gulf states (and the People's Bank of China) buy. They buy it through London, just to confuse everyone. And to allow a few analysts who haven't figured this all out to argue that the US really finances its deficits by selling to private investors.
Interestingly, work by Laura Kodres and Frank Warnock (embedded in the IMF report) suggests Asian buying has more impact than (indirect) Gulf buying. But they concede that the data problems make any conclusion tentative.
Still, there is little doubt that foreign demand for US debt - including demand from the oil exporters -- keeps US long-term rates relatively low. And housing prices relatively high. That supports consumption growth. And demand growth.
But even though I cannot convince myself that this pattern will break, and I do worry that it might. Nothing guarantees a repeat of 2004 and 2005.
The Fed may push up short-rates to the point where the oil states are happy to hold short-term claims. And that will push long-rates up.
Even if real rates remain relatively low, they may not fall. And without the fall, housing prices may not rise. And the 2004-2005 pattern hinges not just on high housing prices, but on rising housing prices.
Or the petro-sheiks may decide that they don't like US debt anymore than they like US goods. Has anyone done a chart showing the evolution of US treasury yields after the US said no to Dubai Ports World?
All this speculation has resulted in a long post.
So I should probably do an abridged Setser.
The oil savings glut supported housing and debt-centric growth in the US in 2004 and 2005. The associated financial flows were well-hidden, but they still left a few traces. The oil savings glut looks set to get bigger in 2006. And it may - or may not - support continued strong US growth.
I am now gun-shy. I have been wrong too many times.