Mandel v. Setser. Round two. More on intangible exports and dark matter
More on:
I was rather busy last week, so I was not able to respond immediately to Michael Mandel's response to my response to his Business Week cover story - and now Micheal Mandel is shooting for slightly bigger prey.
But there is nothing like a few snow days to make up for lost time. Mandel makes two points in response to my argument. Both are worth a bit of further discussion.
First, he challenges the conventional wisdom that the United States has taken out external debt to finance a surge in consumption (or, more accurately, a surge in consumption relative to income) and a surge in investment in residential real estate.
Rather than borrowing from China to build suburban houses (or to bid up the price of the existing housing stock in "zoned land") or finance government deficits, we in the US are borrowing from China to invest in ourselves. We are borrowing to invest in our human capital. College education isn't cheap, but, in aggregate, it generates solid returns.
Mandel thinks pessimists are wrong on a second count as well. American are not just investing in assets that yield a purely domestic return. Think houses and licenses to sell residential real estate. We are investing in assets that generate external revenues - revenue that can be used to pay the US import bill once we stop exporting so much debt. And we sure do export a lot of debt now. Treasury just released the end-2005 data, and the US sold over a trillion in US debt to foreigners -- the US sold $1049b of its long-term debt to foreigners, and only bought $16b of the rest of the world's long-term debt.
No worries, according to Mandel.
Starbucks China may pay rent to the Chinese communist party (and its well-connected friends), and employ Chinese labor to operate Italian expresso machines. But the (predominantly) American owners get most of the profits, and those future profits in China can pay for lots of low-end electronics assembly. The returns that Intel captures by exporting its know-how to plants abroad will pay for importing the actual chips.
Is Mandel right? Are we in the US taking on external debt to ramp up our investment in ourselves, not just to spend more than ever? Are we investing in intangible assets - human capital, management training, R&D, brands - that will generate enough dark matter and intangible exports to allow the US pay for lots of imported Asian assembly (and more than a bit of oil)?
One warning -- the word count on the remainder of this post is almost as high as some of the snow drifts in Central Park on Sunday.
Mandel thinks the worry warts (i.e. the gloom and doom caucus, trade deficit division) pay too much attention to the United States' lack of "tangible" exports. That is the kind of stuff that China does. Not the US. The US comparative advantage lies in intangibles.
More broadly, US managerial genius ekes out higher returns on US investment abroad than foreigners get on their investment in the US. In other words, it doesn't matter that the US had more offshore liabilities than assets - we can make enough on our assets to pay the interest on our liabilities AND to pay our import bill.
Is he right?
Let me turn to Mandel's first argument first:
Should our image of the US economy shift from a home equity line financed by the Chinese central bank -- OK, really, the Chinese central bank, the Saudi monetary authority, the Russian central bank, the Abu Dhabi investment authority and a few less visible central banks in the emerging world -- to a college loan and corporate bond issue to finance US firms R and D, with the bulk of the financing still coming from the Chinese central bank and a few others?
I originally argued that counting education and R and D as "investment" and money set aside from current earnings - whether corporate earnings or labor income - to pay for R and D and education as savings would just increase the amount of savings and investment in the US economy, it would change the fact that savings has fallen relative to investment (or the fact that investment has shifted toward residential real estate).
Mandel asks how do I know? To be honest, I don't know whether spending on education has risen relative to GDP since say 2000. I suspect it has, but not by much. But I don't know. Mandel himself seems though to suggest that there hasn't been much of a recent increase.
What about corporate investment? Has it risen since 2000, if one redefines R and D as investment? I rather doubt it. Why: all the cash on corporate balance sheets. Cold hard cash. Cash left over after spending on investment of all kinds (including all the current expenditures that Mandel would like to redefine as investment). I suspect the better image of the US corporate sector is Exxon Mobil spending more on stock buybacks than on R and D and investment, not Apple engineers and designers thinking up the next Ipod. Obviously, both are going on. But I suspect that since 2000, the balance has shifted toward the cash hoards, not investing in new products. That is certainly what the aggregate data showing a surge in corporate savings suggests.
UPDATE: Mandel says the top 10 US companies have increased their R&D by $11 b since 2000. But not all of that is spent in the US (I suspect), the $11 b seems small relative to say the rise in Exxon Mobil's profits and the increase in firms cash balances, and, above all, small relative to the $500b plus shift in the US savings and investment balance between 2000 and 2006.
Mandel wants to compare the US today with the US in the 1950s, and argue that since the 1950s, investment in intangibles is way up, even if investment in tangibles is way down. I want to compare the US in 2006, when it will run a 7% of GDP plus current account deficit with the US of 2000, when it ran a 4% of GDP deficit. Actually, I would argue the right baseline is not 2000, when the US really did have an investment boom, but rather a current account deficit of closer to 2% of GDP.
My baseline world is one where the US current account deficit expanded during the US investment boom (and Asian bust) of 98-00, but then comes down. Again I digress.
Let's limit ourselves to the expansion in the US current account deficit since the end of the last US high-tech investment boom. Does the 3-3.5% of GDP increase in the US current account deficit reflect an increase in intangible investment since 2000, or a fall in savings (out of current income) as the US government swung from structural balance to a structural deficit and US consumers used housing as an ATM to make up for weak wage growth, and, more generally, spent more of their current income as rising housing prices pushed up household net worth and let Americans to spend more of their current income?
I think the overall answer is pretty clear. Corporate savings has risen relative to corporate investment (what JP Morgan calls the Corporate savings glut). Redefining a lot of spending as investment in intangibles doesn't change the balance between corporate savings and investment unless firms are investing a lot in intangibles than at the peak of the .com bubble. That's unlikely. A lot of folks spent a lot of money investing in competing online pet stores back then. And I suspect that detailed data on the US household sector would show that household borrowing is up (and savings is down - see the Levy institute, among others) to make up for stalled wage income growth, not to pay for an upgrade in America's human capital.
Moreover, even if Mandel is right, the United States lack of domestic savings still has a cost. Borrowing from China to finance US education implies that China gets some of returns on the US investment. Right now, the interest rate China charges is low, but that might change. Similarly, US investment abroad is not financed out of US savings, but out of foreign savings. It is a management contract so to speak - the world lends money to Starbucks to invest in China. And it expects to get paid interest on its loan.
What of Mandel's second argument - one that parallels the argument of Hausmann and Sturzenegger. Will US investment in new intangible assets (the Starbucks brand, The Ipod's software, design and brand, and so on) allow the US to pay the bill for all its tangible imports?
And, for that matter, bill for the intangibles that the US imports. As I noted in my conversation with Mark Gongloff of the Wall Street Journal, Americans kind of like Japanese and German automotive design. Some parts of America seems to import a fair amount of French and Italian design -- tho not yet Chinese/ Swiss luxury brands. The Spanish and the Swedes also seem to be doing alright in the global market for cheap chic - another intangible.
Not to mention that special Irish intangible the US likes to import: corporate tax minimization. When Pepsi and Coke make concentrate in Ireland, it doesn't have much impact on employment in the US. Most of the jobs come from marketing, bottling and distribution. Not making the Pepsi and Coke concentrate. Shifting the manufacture of Pepsi concentrate and Liptor (Pfizer's blockbuster) to Ireland primarily shifts revenues from the US government (Paul O'Neill wasn't able to abolish the US corporate income tax) to Pepsi and Pfizer's shareholders. Its overall impact on the US balance of payments is probably smaller than its impact on US government revenues ...
I digress.
There is no doubt that the US does better exporting intangibles than exporting tangibles. But that is a low bar for a country that imports twice as many tangible goods as it exports.
The only question is does it all add up.
I suspect not.
A point of clarification. The export of "intangibles" can come from foreign direct investment. Starbucks had the genius to marry Italian café culture with American drive-through (and walk-through) habits. The profit margins are higher if you take your tall skinny latte to go. In Mandel's vision, the profits from exporting the Starbucks experience -- the return on Starbucks investment in China -- will pay the US import bill. Coffee is a commodity. As is the labor required to make a tall skinny latte. Starbucks is an experience -- and it will get most of the profits on China's future demand for tall skinny lattes to go.
The US can also earn intangible revenue abroad from the exports of services - the screening of a US film in Beijing counts as a service export, as do the royalties the world pays Fox for Simpson's reruns and Microsoft for the use of its software. Or from the export of a good. Boeing planes that leave Seattle for Beijing require lots of design and engineering.
What matters in each case is the net external revenue the US gets. The profits on Starbucks Beijing. The royalties on Microsoft's software (net of Microsoft's imports of software from the rest of the world). Boeing's external sales (net of its imports of Japanese, European and other components).
US intangible exports are not limited to Starbucks, McDonalds and Microsoft. The US, for example, could also export its expertise at fund management to China, or even Saudi Arabia. 2 and 20 pays a bit better than assembly work. That may require some changes though. Wall Street still tends to make its money - at least right now - managing America's wealth, structuring transactions that let American investors avoid US taxes, helping hedge funds trade and selling US debt to the world. Not by using US based labor to manage the world's wealth. There is a reason why the Wall Street Journal dropped its international page from its US edition.
US will have a bit of competition here. The Swiss were on to this a long time ago. The Brits too. Singapore now wants a bit of the action too.
i digress.
The question is not whether or not the US exports "intangibles." But rather whether it can export enough intangibles to pay for all the tangible goods that enter US ports every month. If you sum up the United States future external income from exporting intangibles (and tangibles), those revenues need to be large enough to pay for the United States future import bill. Standard debt sustainability analysis says you can borrow to pay interest on your existing debt without running up your debt to GDP ratio, but not also borrow to run ongoing trade deficits.
Assembly work may be a commodity, but the US does import a lot of it. Oil too. We did not follow Japan's example and radically reduce the energy intensity of the US economy. Rather, we took advantage of cheap oil in the 1990s to supersize our automobile fleet. And isn't cheap any more. Microsoft's margins are good. But not Saudi good. Some Saudi oil comes out of the ground for about $3 a barrel. It sells for $60.
Mandel's argument is that the US right now is investing in ways that will generate future external revenues - whether from the profits from US foreign direct investment, the United States skill at borrowing low and lending to the rest of the world high, service exports or future goods exports. So we shouldn't worry too much about the gap between what the US currently buys from the world and what it sells.
Why am I skeptical.
Fist, the US will need to sell an awful lot of intangibles to make up for the exceptionally large gap between its (large) tangible imports and (small) tangible exports. Moreover, in the future, it will have to export intangibles to pay for the interest on that current gap.
Second, exporting intangibles has problems of its own. Copyright enforcement and all. Those problems won't go away in a world of sovereign states. China has yet to conclude that it is in its interest to pay the full price for its imports from Seattle (Sorry, Microsoft) Los Angeles (Sorry, Hollywood), or Paris (sorry, Louis Vuitton). And, given how profitable counterfeiting is, some folks somewhere seem to be trying to get their a cut of US and European intangible exports. Those problems won't go away. It is not enough to have a comparative advantage in the production of the Louis Vuitton experience. You need to convince the rest of the world that it is in their interest to pay full price for that experience -
On the plus side, China doesn't charge the US full price for its assembly services either -- the artificially low RMB and all. It works both ways. Chinese counterfeiters make tons of money at the expense of US and European firms. US and European firms make tons of money employing cheap Chinese labor at the expense of Chinese taxpayers who will eventually have to bail out China's central bank for the cost of China's massive currency bet.
Third, I am not convinced that the US economy is gearing up to produce "intangible" goods that can readily be exported -- rather than just shifting into the production of domestic intangibles (where there is less competition from China). I don't see US firms doing a wonderful job building up the kind of intangible assets that easily translate into future exports. Like Calculated Risk, I see a lot of investment in housing, and lots of people training to be real estate brokers.
Fourth, I suspect that a lot of what Hausmann and Sturzenegger call dark matter - or what Mandel would call the gains from exporting "intangible" US expertise at financial and brand management - will turn out to be the product of exporting the Federal Reserves super-low US interest rates.
I would be the first to concede the world has been -- and still is -- wiling to loan the US money at very generous terms. My big concern is that those terms are too generous to last. The rate the US has to pay is rising as we speak -- though more because the Fed is raising rates than because foreign investors are demanding hiher rates in the market.
Isn't it a bit suspicious that US dark matter soared ("Hausmann and Sturzenegger estimate that dark matter grew by $559 billion a year between 2000 and 2004") just when US rates fell to historic lows?
Hausmann and Sturzenegger (in Mark Gongloff's Wall Street Journal article) say that there is no correlation historically between interest rates and dark matter. That is hard to believe. Try doing the dark matter calculations if the average return on foreign investment in the US stayed at 3.6% (its level in 2000) rather than falling to 2.4% in 2003 (it was 2.7% in 2004). Remember, dark matter is inferred from the gap between what the US pays on its liabilities and what it gets on its assets. The rate the US pays on its liabilities is tightly correlated with interest rates. So how can dark matter be entirely independent of interest rates?
Just to be clear, I recognize that some dark matter comes from the returns the US gets on a US FDI abroad (high returns) for foreign FDI in the US (low returns) swap ... I just don't think that fully explains the recent surge in dark matter.
Let me conclude by noting that in 2006, the US will export about one trillion less - counting both US exports of tangible goods and intangible services, including the service of financial intermediation - than it imports.
It will - if all goes well - maybe borrow $1.2 trillion from the rest of the world. About $900 billion of that will go to pay for US impost of goods and services, and about $70 billion will go to pay interest on past US imports of goods and services (US debt). That $70 billion would be a lot larger if US firms do not continue to earn more on their investment abroad than foreigners earn on their investment in the US. About $200 billion will be used to finance US foreign direct investment abroad and US portfolio equity investment.
Is there any plausible way that the $200 billion investment will generate sufficient investment income to pay for the interest on the $1 trillion the US takes on, let alone generate enough investment income in the future to let the US pay for its ongoing imports of tangible goods out of the returns on its past investment.
Try doing the math. It doesn't work. It is darn hard to generate enough investment income from say $11 trillion in external assets to pay the interest on $15 trillion in external liabilities (I am forecasting out to the end of 2006) and a $1 trillion trade deficit in current tangible and intangible exports. It is hard to grow your external assets rapidly when every external asset has to be financed by adding to your external debt.
5% on $15 trillion works out to $750 billion a year. That is what the US likely will be paying on its external debts -- assuming that US rates stay low and foreign firms will continue to struggle to eke out a profit in the US. To pay this and its import bill, the US would need to earn $1750 billion off its $11 trillion in external assets. That works out to an average return of over 15%.
And since a decent fraction of the United States external assets are low-yielding cross border bank lines, the returns on US FDI and portfolio equity investments (about $7 trillion of the total) would need to be very, very large indeed.
It ain't gonna happen. And if it did, the juicy profits the US earns by intermediating the world's savings into US brand equity would be a nice fat target - one that would attract new entrants. Why should the rest of the world lend to the US at 5% so it can get super-returns rather than try to make the kinds of investments that US firms make offshore in order to earn these super-returns?
More on: