Brad Setser

Follow the Money

Cross border flows, with a bit of macroeconomics

Print Print Cite Cite
Style: MLA APA Chicago Close


Good news for the fourth of July

by Brad Setser
July 4, 2008

The US economy has absorbed its share of blows over the last year:

Jobs have disappeared;
Oil prices have soared;
Consumer confidence is way down;
An important broker-dealer collapsed, prompting the Fed to facilitate the take-over of a very-troubled financial institution by a less-troubled financial institution;
And more financial institutions could have faced trouble absent access to Fed liquidity support.

There isn’t a lot for the US economy to celebrate this fourth of July.

But one thing hasn’t gone wrong.

The US net international investment position — if US direct investment abroad is valued at its market value — actually improved, absolutely and relative to US GDP.


Source data for the graph

The red and green bars reflect the improvement in the US international investment position from exchange rate moves (the dollar’s fall increases the dollar value of US investment abroad) and market moves (if foreign equity markets rise by more than the US equity market, the value of US investment abroad rises more rapidly than the value of foreign investment in the US) relative to what would be expected from simply summing up financial flows.

Adding in valuation gains doesn’t eliminate the gap between cumulative financial flows and the NIIP. The remaining gap comes from so called “other changes” — basically a statistical adjustment that reflects the gap between what the assets abroad the US identifies through the survey and cumulative flows and estimated valuation gains. Some call this statistical manna from heaven. Others suspect that the US just isn’t able to count all the US debt the rest of the world holds.

Historically, it has been an important reason why the US international investment position hasn’t deteriorated as much as might be expected. It has a big impact on the 2005 data too. But it hasn’t driven the data since then. The recent improvement reflects exchange rate gains and, above all, the better relative performance of foreign equity markets than the US equity market. This shows up more cleanly if valuation gains are plotted over time in billions of dollars.


Most US debt (and the foreign debt Americans hold) is denominated in dollars. The falls in the dollar don’t generate big valuation changes on the debt side. At least not for the US. The value of US debt has been falling when expressed in the currencies of some of the United States creditors. That is the flip side of US valuation gains. Valuation changes by contrast have pushed up the value of US equity investment abroad relative to foreign equity investment in the US.

A plot of the US net debt position and the US net equity position shows this. The net debt of the US (US borrowing from the world minus US lending to the world) is rising quickly, as the US has financed its current account deficit by placing debt with investors around the world rather than by selling equities to the rest of the world. And the US net equity position has improved — almost entirely as a result of valuation. Net equity flows have balanced themselves out over time (inflows in the 90s were matched by outflows after the .com crash; more recently the 2005 Homeland investment Act generated inflows that have offset some post 2005 outflows)


Back in 2004, Nouriel Roubini and I anticipated a large deterioration in the net debt position of the US. We didn’t anticipate the offsetting improvement in the net equity position. We didn’t expect the US equity market to perform so much worse than global markets. And we expected more dollar depreciation against Asia (where the US has less foreign investment) and less against Europe. Above all, we didn’t imagine a world where the US could attract large inflows on a sustained basis despite offering financial assets to the rest of the world that significantly underperformed global markets. Money usually flows to the place where it can get the highest return, not the place that produces the lowest return.

So who financed the US deficit even though the US dollar fell, US bonds paid low rates and the US market underperformed global markets? Well, above all, the official sector. Or rather, the official sector, once the large short-term flows associated with London’s emergence as a major offshore financial center for the US are netted out. The change in the official sector’s holdings over time offers one measure of capital inflows. Using the change in stocks to infer flows risks overcounting, because the rise in the stock will be influenced by changes in equity markets. But the official sector’s holdings of equity markets are small enough that this isn’t a major driver.


Three points are worth making here.

First, the change in official holdings in NIIP in 2007 is a bit over $500 billion, while official inflows in the 2007 balance of payments data are a bit under $400 billion. This isn’t a reflection of valuation gains on official equity investment. Rather it is a reflection of the impact of the 2007 survey on the balance of payments data. The balance of payments data for q2 likely will include significant revisions to the 2007 data.

Second, the composition of official inflows has changed over time, with much larger flows going into long-term Agencies now than in the past. Official demand for equities is also rising. But the rise in official demand for equities in 2007 wasn’t as large as the rise in official demand for short-term bank deposits and similar claims. A few sovereign funds invested in risky financial institutions. But far more official players decided not to take any chances and piled into the safest stuff around. The official flight away from risk could have had a much of an impact on the market as capital injections in leveraged financial institutions; both flows need to be included in an analysis of how the official sector impacted the market.

Third, the 2007 data almost certainly understates true official demand for US debt. The survey only captures the change in stocks through the middle of 2007. Comparable data for the last bit of 2008 won’t be available until the spring of 2009. Official investors increasingly make use of private fund managers to try to get better returns — and those funds aren’t captured in the survey.

Why I am I fairly confident of this? Well, recorded official inflows, large as they are, are still well below the known increase in global reserves in 2007. And while the IMF’s COFER data has big gaps, the available data doesn’t suggest a shift out of the dollar.


Combine the big rise in reserves and an assumption that central banks who don’t report detailed data to the IMF acted like the central banks who did report data to the IMF, and implied dollar flows are way higher than recorded official inflows to the US. This is also story that helps to explain how the US has been able to attract big inflows despite offering subpar returns. The official sector’s dollar accumulation hasn’t been motivated by purely commercial (or least not purely financial) considerations.

Rather than focus on the resulting risks, though, today I’ll just highlight the result: the United States net international investment position is in far better shape than anyone could realistically have hoped after ten years of large current account deficits.


  • Posted by Howard Richman


    That is good news. It means that the United States internal debt is not increasing. The only problem is the reason: The United States is such a terrible place for financial investment at present that anybody who has their money abroad gets a better return on their financial investment.


  • Posted by Howard Richman

    Whoops I meant “external debt,” not “internal debt.”

  • Posted by Charles


    Great post. Looking through the inflows data you don’t mention the pickup in bank liabilities and other foreign assets. How would you explain this?

    By taking the change in the IIP you introduce valuation effects which is not necessarily appropriate for understanding the fund of the CA deficit. Looking at the capital inflows data I notice a decline in US portfolio debt inflows while lending and FDI inflows continue to be robust.

    Finally, how have you seen the crisis in the shadow banking system manifest itself in the data?

    Thanks – interesting work as always.

  • Posted by Rien Huizer

    How reliable are the NIIP data compared to the debt data. As said earlier, the US external debt as a percentage of GDP, or financial system size seems manageable. The NIIP seems be generating returns that would offset the interst burden on the debt. However, it all seems a bit woolly.

    The NIIP goes up when foreign equity markets go up and the USD goes down. The former seems to have not been the case during the first half of 2008 and perhaps, if (not very likely but possible) the USD would go up from here (except against the JPY of course, that is a clear case of misalignment) because of FED tightening (and that tightening on top of the downward momentum consumption appears to have in the US), share markets, including the foreign ones might even trade much lower. In that case all the good work by frugal US consumers for the BOP and the national balance sheet would be negated by a still rising debt, higher interest on the debt (large component short term), and declining (both exchange and share price) NIIP… You cannot win, it seems.

  • Posted by Michael McKinlay

    LOL …

    Good news as of December 31, 2007.

    Look at the foreign markets since the beginning of the year. Down, down, down … China minus 50% , India minus 30% etc, etc, etc …

    And while investments are being pummeled the interest rate on our net debt is rising, in some cases dramatically.
    As Bill Gross at PIMCO says, the last 30 years have been good for debt, the next 30 may not.

    How Brad can ballyhoo such a nasty turn of events is beyond me.

  • Posted by anon

    “The change in the official sector’s holdings over time offers one measure of capital inflows … recorded official inflows, large as they are, are still well below the known increase in global reserves in 2007”

    This area is a bit tricky, isn’t it? Foreign central bank reserves that offset dollar trade and capital flows with non-US countries won’t show up in US current account or NIIP, will they?

  • Posted by Chris White

    Thanks for the great post and compilation. How about looking at what you present through the eyes of the host countries rather than the US? Couldn’t this profile the substantiation of a potential argument for exchange and capital controls to offset losses on dollar denominated lending to the US?

  • Posted by bsetser

    I rarely yet accused of being overly optimistic about the state of the US balance of payments, but I guess there is a first for everything. The interest rate on our debt is, incidentally, falling — as a lot of US external lending is indexed to short-term rates and long-term Treasury rates are down from their levels a year ago. Right now the fall in the rate is overwhelming the rise in the stock, so net payments are also falling. Combine that with falling returns on foreign investment in the US (the recession) and the US income balance is actually improving.

    Rien, as you note, a recovery in the US could easily “undo” some of the recent improvement in the US NIIP. US markets would rise, the dollar would recover, etc. That is why I don’t put much stock in the niip data as a predictor. Right now, the US data portrays the NIIP as better than it really is — if one expects, over time, that some of the existing valuation gains will prove to be cyclical rather than structural. That said, some of the underperformance of US assets relates to a $ fall that may not be reversed, which could generate positive one-off gains.

    That said, the fact that the NIIP gets better when the US does worse is a positive — it is far better than the opposite case, where the niip gets worse when things go bad. Emerging markets with $ denominated debt faced this problem, and it wasn’t pretty. The US would be in deep trouble if its external debts were denominated in euros.

    One thing tho does worry me greatly. The net debt position of the US was negative 6 trillion at the end of 07. On current trends, the net debt position will be negative 8 trillion at the end of 09. If US rates eventually normalize, and say the average int. rate on US external debt rises from around 3% to around 5% (not a high rate, all things considered), net interest payments would rise by $160b. That is a recently strong headwind working against sustained current account improvement.

    So if you dig down, I do think there are some troubling components in the NIIP data. Nonethless, if you compare the NIIP today with the trajectory Nouriel and I mapped out in 04, the US is in far better shape than one would have expected.

  • Posted by bsetser

    Anon– I am not quite sure what you are getting at. If a government holds dollars as part of its reserves, it is holding a claim on the US (unless it is holding the dollar denominated debt issued by another government). So the dollars various countries that don’t trade much with the US hold as part of their reserves are still financially claims on the US. Over time, — after waiting for the revisions to the US data and adding in offshore central bank dollar deposits — overall dollar holdings usually rise along with the overall rise in reserves.

    Charles — there BoP flow data rather than the NIIP data tells the story of the shadow banking system best. Basically, there was a big fall in gross outflows from the US and gross inflows to the US after the shadow banks collapsed in August. The overall data for 07 doesn’t show this cleanly b/c inflows/ outflows (including lots of “other investment” outflows, i.e. bank flows) in h1 were strong. But there was a very sharp fall.

  • Posted by anon

    “I am not quite sure what you are getting at”

    E.g. UK pensioner buys stuff from China; UK pension fund sell US treasuries to PBOC to finance pension payments and outflow (with FX transactions that net to 0); Chinese reserves up; UK NIIP deteriorates; US NIIP unchanged?

    Net effect is that the liability mix of US NIIP changes from private to official but remains unchanged in total?

    Can’t some of the official reserve increases partly reflect such changes on balance sheets outside the US, in addition to net financing back to the US? And if so, how can you tell what the mix is?

  • Posted by anon

    2nd paragraph above, I mean foreign ownership of US liabilties switches from private to official.

  • Posted by bsetser

    Anon — that is happening. that kind of transaction doesn’t register at all in the TIC data or the balance of payments data. It would theoretically show up in the NIIP in the way you described — as a change in the ownership of the existing stock. And then consider the case where a UK pension fund sells US equities to a new UK investment fund that has a mandate to manage money for the CIC. The second transaction wouldn’t show up at all in the US data.

    I cannot tell what the mix is — i.e. to what extent existing US assets have been sold to China. Total official purchases are clearly larger than recorded purchases, so some of the flow should be reallocated to the official sector. But total sales of Agencies and Treasuries to foreigners in 07 were around $600b (this comes from memory), which struck me as a bit on the low side even if all these sales ultimately were to foreign central banks. I was expecting a bigger rise in the BIS $ deposits that showed up at the end of the year. An alternative would be the process you described, i.e. a net sale of existing foreign holdings of US debt to official actors.

    the strongest evidence of the “EU sells its US holdings to China” flow would come if the rise in official stocks in the survey exceeded the sum of private and official purchases in the flow data (which is used to produce the BoP data until it is revised to reflect implied flows in the survey).

  • Posted by anon


    To restate my point – with the enormous buildup in official dollar reserves globally, much of it obviously shows up as new net or incremental gross flows into the US, but some might show up as a shift in ownership of previous flows (i.e. “stocks”) – i.e. expansion of official dollar reserves includes incremental flows into the US, net and gross, plus an increasing share of liabilities already outstanding in the US IIP.

  • Posted by Howard Richman


    Your estimate that the United States could be soon paying a net $160 billion a year in interest on our external debt suggests one of our recommendations in Trading Away Our Future — that we should charge a withholding tax on our interest payments to foreigners, whether private foreigners or governments. (Currently we don’t charge any income tax on interest earned by foreigners, though we do charge up to 35% income tax on interest earned by Americans.)

    Let’s say that we charged a 30% withholding tax on the $160 billion paid to foreign governments, as we did on interest earned by private foreigners until 1984 (except when tax treaties reduced that rate). Then that would raise $48 billion in additional tax revenue each year.

    Furthermore, taxing foreigners could have the good effect of raising US interest rates making it more likely that Americans would start saving more.


  • Posted by Howard Richman

    Whoops I didn’t mean to say “the $160 billion paid to foreign governments”, I meant to say “net interest paid to foreigners in general”.

    Howard Richman

  • Posted by Rien Huizer


    This is interesting material. Given the rather loose correlation between exchange rates, interest differentils and relative equity returns, if one knew the asset class and currency composition (roughly would probably enough) and the rpricing structure of the debt, one could simulate at least eight groups of scenarios, not all eqully plausible of course but useful in modeling the unthinkable. Your 160 bn drag for instance.

    Also, given these loose correlations again (USD could keep going down despite massive tightening by the FED, look at the late 70s) Stockmarkets do not go up when monetary policy is eased, etc.) the “equity portion”, ie NIIP minus debt would probably be highly chaotic.

    Finally: how a re privately held investments (eg US MNC’c claims on subsidiaries) accounted for?

  • Posted by anon

    Where is residential real estate held by foreigners accounted for? (small but interesting).

  • Posted by Howard Richman


    I suspect that you may have made one minor mistake regarding the motivation of foreign central banks when you wrote: “Official investors increasingly make use of private fund managers to try to get better returns — and those funds aren’t captured in the survey.”

    I suspect that the growing use of private fund managers could be coming from countries such as Iran, Venezuela, and China, whose foreign policies are opposed to the United States. By keeping the ownership of their funds hidden, they are taking less risk that the US government could freeze their funds in case of foreign policy conflict, as President Carter once did to Iranian funds.


  • Posted by Howard Richman


    For discussion of central bank motivation in keeping their dollar deposits offshore see Robert McCauley’s article in the September 2006 BIS Quarterly Review ( )


  • Posted by Dave Chiang

    Here are a few things to reflect upon:

    While the DOW is down 20% from its high, Saudi Arabia’s stock market is up 21% since the start of the year, and Brazil’s is up 14.7%. I recommended Brazil Petrobras a year ago to everyone on Brad’s blog. Anyone listening would be up 150% in the past year.

    The BRIC countries which consume raw materials (India and China) have had the worst performing markets this year; those which produce commodities have done better.

    Commodities in a half year outperformed all stocks by 2 ½ times, with a gain of 26.8%. Oil and gas commodities are up even more, just under 30%.

    There was once a time when the US stopped Russia from building missiles in Cuba, now Russia and China are jointly financing a new oil drilling program offshore Cuba in the Caribbean. LOL 🙂

  • Posted by bsetser

    Howard — it is likely that China still keeps huge sums at the new york fed, which is a bit at odds with the notion that china is using london fund managers to hide.

    Rien — I think that MNCs foreign subsidiaries are valued by looking at comparable listed equities (or at book, if there isn’t a good comparator). But i would need to look at the survey of current business article that goes through the BEA’s methodology to know for sure.

  • Posted by Howard Richman


    Hmmm. Clearly they are not hiding even close to all of their funds.

    But I would never expect them to hide their funds in London. If they were hiding their funds they would likely use Singapore and/or Hong Kong banks to do so, not London banks.

    By the way, I believe that I read somewhere within the last few weeks that Iran just moved its funds to Singapore to put them out of reach of U.S. freezing of funds.


  • Posted by Howard Richman

    I just did a google search but can’t find the source of the story that Iran was moving its funds from Europe to Singapore, but here is a segment from another article that discusses the story ( ):

    “An Iranian weekly previously said Iran has withdrawn $75 billion from Europe
    to prevent the assets from being blocked under threatened new sanctions over
    its disputed atomic plans…. The Shahrvand-e Emrouz weekly quoted Mohsen Talai, deputy foreign minister in charge of economic affairs, as saying that a part of Iran’s assets in European banks had been converted to gold and shares and another part
    transferred to Asian banks….”


  • Posted by anon

    Well, what do you know?

    “In the title of this post, I mentioned “What’s Missing”. One issue Frank identifies, which was news to me, was that real estate purchases are not included in the balance of payments data, and hence not in the international investment position estimates.

    In principle, cross-border transactions and holdings of residential real estate should be included as part of direct investment, as is currently the case for commercial real estate. In practice, individual homeowners are not surveyed and hence these data are omitted from the recorded DI figures. To the extent that foreign activity in the U.S. residential real estate market is of the same magnitude as the level of activity of U.S. residents in the foreign real estate markets, there is no net impact on the international transactions accounts. However, recent surveys conducted by the National Association of Realtors (NAR) suggest that this may not be the case.”

  • Posted by john

    Sell Italian bonds. Italian public debt has reached a record high at 1646,7 billion euros.It is worse than 1992 when the country went very near to declare default(insolvency)

  • Posted by Rien Huizer


    Found the BEA release. FDI at current value (essentially historical cost adjusted for cumulative inflation). I doubt that one could get meaningful numbers re country (and currency?) composition.

    The two stocks both contain some USD 4-5 tr in working capital items that would better be netted when discussing relative size etc. Striking difference in liquidity and obligor composition between the two portfolios. US has about 7 tr in marketable assets and 9.5 tr in marketable” liabilities” (incl stocks). That gap can easily double in 3-4 years (barring a recession in the US while the rest of the world keeps growing, combined with a weak USD) By then it would be a sizeable proportion of GDP.

  • Posted by Rien Huizer


    Are you offering a discount? Italian bonds are mainly owned by Italians (but perhaps also in a few offshore accounts owned by Italians). Incidentally, one of the problems of the EUR is that it is hard to let countries default and hard to keep them solvent. Perhaps the Italians should resurrect Mussolini. He made the trains run on time. Italy will have a hard time negotiating concessions (and every country has a wish list) in Brussels this is setting a bad example for zero-sum cowboys like the Poles, etc. Some day, like in any exclusive club, a misbehaving member will be suspended or thrown out.

  • Posted by RebelEconomist


    While I think that for the US to resist official capital inflows is like looking a gift horse in the mouth, I do agree that if the US is not happy about other countries dollar exchange rate policies and those countries have a closed capital account that prevents the US from counter-intervention, the most appropriate measure is to restrict or penalise their capital inflows rather than their imports into the US. That avoids disturbing the relative competitiveness of various imports and rewarding lobbying by particular US industries.

  • Posted by bsetser

    Rien — to get country data (and infer currencies), you would need to look at the detailed FDI article/ data published every year (in July I think) in the Survey of Current Business. It has a lot more detail than the NIIP release.

  • Posted by don

    RE: “While I think that for the US to resist official capital inflows is like looking a gift horse in the mouth, I do agree that if the US is not happy about other countries dollar exchange rate policies and those countries have a closed capital account that prevents the US from counter-intervention, the most appropriate measure is to restrict or penalise their capital inflows rather than their imports into the US.”
    How might that be done? Would it be better than a uniform tariff on all the offender’s imports?

  • Posted by RebelEconomist


    It would be more appropriate to restrain the asset purchases because this where the offending country is distorting the market while using capital controls to prevent the US from responding in kind. Presumably, to the extent that the proceeds of import sales are used to purchase US goods and service exports, America is content to receive imports, so applying a tariff to all the imports from the offending country would represent overkill. Also, in practice I doubt that any tariff would be uniform; it would probably vary according to the strength of lobbying from domestic producers and importers.

  • Posted by Esmeralda

    Interestingly was, but there is someone who does not quite agree with the author?

  • Posted by romonoeroetoko

    Hm that sounds good but I would like to know more details.

  • Posted by romonoeroetoko

    Your news is a cool stuff man, keep it going.

  • Posted by amenodimeno

    Good story for me but please more details.

  • Posted by queroeropoo

    Good information to me.