The sustainability of the current account deficit, once again

The market seems to have concluded that there is no need to put any pressure on the US to reduce its large external deficits.

The dollar is up, at least against the euro. So is the renminbi for that matter, at least against the euro. Treasury bond yields are down. No bond market vigilantes = no political pressure to reign in fiscal deficits. April retail sales were relatively strong. That suggests a likely pick up in non-oil import growth once the lagged impact of the Chinese Lunar New Year stops impacting the trade data. Remember: even if non-oil import growth stays around 8% and export growth stays around 9-10% going forward, the trade deficit will continue to expand over time.

It sort of feels like the markets have been listening to the Fed. After all, the Fed, with some notable exceptions, has taken something of the lead in arguing that US current account deficits of the current magnitude do not pose a serious policy concern - or at least not the kind of concern that requires any action on the part of the Fed. And for some in the Fed, they don’t really demand much action from the White House: If Bernanke’s savings glut thesis is taken to the extreme, recent large budget deficits arguably have served the usual purpose of preventing the global savings glut from driving real US rates even lower and thus have helped to ward off an even bigger housing boom (or bubble).

The latest don’t worry (very much) argument came from Michael Kouparitsas of the Chicago Fed. Suffice to say that I was less impressed by the arguments put forward by Dr. Kouparitsas than Dave Altig.

I am more in Paul Volcker’s camp. Or more in Robert Rubin’s camp. Probably not a surprise, I know. For a relatively short and non-technical discussion of the reasons why, I recommend this article that Nouriel and I wrote for CES info in Germany.

(much more follows)The funny thing is that Dr. Kouparitsas is using the same basic analytical framework that Nouriel and I used -- he jazzes up the basic debt sustainability model to provide a more complex picture of asset returns, but fundamentally, it is the same model. Nouriel and I also noted in our more technical paper that so long as US real growth rates exceed the real interest rate the US pays on US net external debt, small trade deficits are consistent with a stable external debt to GDP ratio.

The Kouparitsas paper argues that the past offers a reasonable guide to the future, and thus that historical returns on US assets and liabilities offer a plausible basis for estimating future returns - at least when it is not emphasizing how valuation gains on US assets abroad brought the US deficit down to close to sustainable levels in 2003.

Nouriel and I also looked at the 2003 numbers, including the large gap between US real growth rates and the real interest rate the US is paying on its external debt. We concluded that it would be a mistake to assume that similar conditions would hold going forward. We also concluded that would it be a mistake to bank on valuation gains like those experienced in 2003 going forward. After all US policy interest rates were exceptionally low in 2003 and 2004, and the dollar is not going to fall v. the euro every year (Don’t forget, a rising dollar currently implies large valuation losses for the US).

A standard rule of thumb for these kinds of debt sustainability exercises is to assume that the real interest rate will equal the real growth rate (it is often far higher in emerging economies with large net external debts, but the US is not an emerging economy). If that is true, the trade and transfers deficit has to fall to zero to stabilize the external debt (or more accurately, the NIIP) to GDP ratio. If the monthly trade balance averages $60 b during the course of the year, the US trade and transfers balance would be around 6.5% of GDP. That is a long way from zero.

It is certainly possible to argue that the US will be able to pay a relatively low overall interest rate on its debt going forward even as net US external debt rises. Even so, the size of the trade and transfers deficit that is consistent with a stable external debt to GDP ratio is still a low lower than the current deficits. 6% of GDP is not 1% of GDP.

The standard argument for why the US will continue to pay a relatively little on its (net) external debt is that the US holds lots of equity abroad while our creditors hold lots of debt, and equity typically earns more than debt. That argument doesn’t entirely work. Foreigners hold a decent amount of US equity as well as a ton of debt, offseeting most US equity investment abroad. In recent years, the US has earned more on its overseas assets than it has paid on its external liabilities in large part because the returns (at least the reported returns) on foreign equity investment in the US have been extremely, extremely low -- about 2% (nominal) between 2001-2003 by my calculations (using the current account data for payments and NIIP data for stocks). Put it this way: based on reported payments, foreign direct investors in the US would have been far better off holding Treasuries, which paid out at a higher rate.

Some very stylized facts (back of the envelope style): In 2003, the US had around $2.7 trillion in FDI abroad, and foreigners had $2.4 trillion in FDI here in the US. By the end of 2004, US direct investment abroad probably rose to around $3.3 trillion, and foreign direct investment in the US rose to around maybe $2.6 trillion (net FDI outflow plus valuation gains). The 2004 NIIP will probably show net equity (FDI +portfolio equity) holdings of bit over $1 trillion, offsetting net debt of close to $4 trillion, for an overall net international investment position of a bit under $3 trillion. To make the Kouparitsas math work, US equity investment abroad has to earn much more than foreign equity investment in the US.

It seems to me that the different parts of the Federal Reserve system often end up making slightly different argument, as do different people outside the Fed system. For example, Vice-Chairman Ferguson has made the standard argument that the US attracts substantial inflows because the US is an attractive investment destination. Attractive investment destinations usually are attractive because they offer relatively high returns, or at least relatively high risk-adjusted returns.

The problem: as I mentioned above, actual returns on foreign investments in the US have been terrible. No wonder that US investors are investing more in foreign equities (and foreign plant and equipment) than foreigners are investing in the US.

The Koupartitsas paper effectively makes the opposite argument: the current account deficit is more sustainable than most think precisely because the US is really is not all that attractive an investment destination. Foreigners will earn less on their investments in the US (in dollar terms) than the US will earn on its investments abroad, and that differential will help offset the impact of rising net US external debt ...

To quote: "This suggests that the US current account deficit to GDP ratio is actually close to a sustainable level, with the relatively high net exports deficit being offset by a relatively high net factor income surplus (NPI)." Koupartitsas concludes in a way that leaves a strong impression that he expects the surplus to perist in some form.

My problem with that argument? More or less the same as PGL’s problem over at Angry Bear: Why should foreigners be willing accept such poor returns going forward? They can see the returns US investors are getting abroad. Yet financing the current account deficit requires that foreign investors keep adding to their relatively poor-performing portfolio of US assets.

Part of the answer is that the dollar is a reserve currency, and that generates a natural inflow of dollars seeking safety. True, as far as it goes. But current rates of reserve accumulation are far above any normal standard. The dollar has been the reserve currency for a long time; the US historically hasn’t attracted anything like current levels of reserve financing. That goes off into another discussion though.

I expect one other argument will be revived soon as well: the current account deficit is sustainable it reflect high levels of investment. Investment tanked in 2001, but it has recovered significantly. In 2004 and 2005, the expansion of the current account deficit will be driven more by an increase in investment (and a fall in household savings) than by a surge in the fiscal deficit.

I don’t take much comfort in that fact, though. It is increasingly clear that much of that investment is in residential property -- and exporting houses remains hard (the US is quite good at exporting the payment stream on a mortgage though). The composition of investment right now certainly feels different than in the .com era.

I probably should stop here, but I want to conclude with one last set of thoughts. I think many of the people who are most worried about the US current account deficit -- Volcker and Rubin included -- are people with some first-hand experience managing crises in emerging economies. Emerging market crises certainly inform my thinking. I would distill four key lessons from that experience:

1) It is dangerous to have large amounts of debt denominated in someone else’s currency. No worries for the US here: our debt is in dollars.

2) Current account deficits of over 5 or 6% of GDP generally don’t last for long. Obviously, the US has plenty to worry about here. So far, the US has not had to pay a high interest rate to attract this kind of financing. To optimists, that suggests sustainability; to pessimists, that suggests the potential for dangerous dynamics should US interest rates ever rise.

3) Large deficits are particularly worrisome if funded by short-term debt. Sort of applies to the US. After all, the US has shortened the maturity of the Treasury stock significantly, and lots of short-term Treasuries are held abroad. But since many of those bonds are held by central banks, not private investors, it is not clear that they carry quite the same risk. This is far from my biggest worry.

4) External adjustment is more difficult if a country has a small export base relative to its GDP. Plenty to worry about here. A 6% trade and transfers deficit v a 10% export base is pretty extreme.

Of course, we don’t really know the dynamics of adjustment for such large economy as the US -- there just isn’t that much experience. The US trade deficit, relative to GDP, was much smaller in the 1980s. The Fed staff has tried to get as this question by looking at the experience of other industrial economies. However, it is not entirely clear that this experience applies (nor is it clear that it does not). Not only are all the countries in the sample set far smaller that the US, but most of these economies went into their period of current account adjustment with relatively high interest rates. They typically were trying to defend an exchange rate peg, and because of their appreciated exchange rate were running current account deficits despite the economic drag created by high rates. Dropping the peg let them drop interest rates, and exports expanded faster than imports. The current account deficit fell, the economy grew. Think of the UK after George Soros (and others) forced it out of the ERM in the early 90s.

That may describe the future US adjustment path; one can certainly hope so. But it is no sure thing. The US certainly doesn’t look entirely like an emerging economy. To me, though, it doesn’t fit the industrial country pattern well either.

(apologies for the length of this post)