Brad Setser

Brad Setser: Follow the Money

Some things I got (more or less) right

by Brad Setser Saturday, December 31, 2005

Unfortunately, this list is a bit shorter than the list of things I got wrong. 

Still, I think my crystal ball got a few things right.

  1. The rise in China's current account surplus.  I – influenced by Morris Goldstein  – thought China's surplus was poised to expand dramatically in 2005.  It did.  No one can now argue that China's surplus with the US (and Europe) is offset by its deficit with the rest of the world.  It is not.  China runs a current account surplus that is roughly comparable in size (when assessed relative to China's GDP) to the deficit of the United States.  China's surplus will be above 6% of its (revised) GDP in 2005, and the US deficit will be above 6% of US GDP as well.  
  2. China would emerge as the true anchor of Bretton Woods 2, and would be forced to provide ever-more financing to the US.    That still seems right.  The pace of China's reserve accumulation increased in 2005.  Adjusting for roughly $30 billion in estimated valuation losses and the transfer of $15 billion in reserves to one of China's state banks, I suspect China will have added close to $250 billion to its reserves in 2005.  Maybe more if the small November reserve increase reflects the PBoC's currency swap, not a big fall off in capital inflows. $250 billion is up from the $185 billion or so China added to its reserves in 2004 (assuming that valuation gains accounted for about $15 billion of the $200 billion "headline" increase in China's reserves). Chinanow accounts for roughly ½ of global reserve accumulation, up from maybe 1/3 in 2004 and a bit over a quarter in 2003. 
  3. At least some in China don't think it makes sense for China to keep on adding to its dollar reserves at its current pace.  They worry that the United States' large trade deficit suggests the dollar is likely to fall in value over time.   Given that a large share of China's national savings is now invested in dollars, a fall in the dollar relative to the RMB (or the euro relative to the RMB) would reduce Chinese wealth.  China's central bank understands this, probably better than anyone else. Professor Yu Yongding has made that clear (hat tip: Steve Hsu).  The PBoC has good reason to fret: they have issued RMB debt (and RMB) cash against China's dollar and euro reserves.   Chinese savers want RMB, not dollars.  So do lots of other folk.  
  4. Petrodollars are really petrodollars.  I am going out on a bit of a limb here, since we still don't really know.   The core message of the most recent BIS quarterly report is not that the oil states are showing a stronger preference for dollars, but rather that the data collected by the BIS from the world's banks only tells us the currency composition of a tiny fraction of the oil states' external assets.  But I think most people do believe that the oil sheiks are providing some financing of the US, though in ways that do not show up in the US data.  When the final tally comes in, I suspect that the oil states will be the source of 40% of the world's current account surplus in 2005, and the US something like 80% of the world's deficit.   So I'll put it this way: the roughly $400 billion current account surplus of the world's oil exporters exceeds the combined current account deficit of the countries outside the US that have deficits (treating the euro area as a single unit), so the global current account only adds up if the oil states are financing the US in some way, directly or indirectly. 
  5. The US trade and current account deficit did not peak in early 2005.  Alan Greenspan thought he detected incipient market pressures for adjustment back in March.  The Maestro's call now looks just a bit off.   Even then, the market pressures for adjustment were limited – they were present in the euro/dollar, but not in many other currency pairs or in the bond market.  And almost as soon as Greenspan detected these pressures for adjustment, they started to disappear.  The q4 2005 US current account deficit is likely to come in at close to 7% of US GDP, a new record.    The overall 2005 current account deficit will come in a bit over 6.5% of US GDP – up from 5.7% in 2004.

The large q4 current account deficit is one reason I am fairly confident that the US current account deficit will continue to expand in 2006.   The q4  2005 current account deficit is likely to come in at $225 billion, if not a bit more.  Annualized, that implies a $900 billion current account deficit.  And I expect the 2006 US current account deficit will be a bit bigger than that. 

Remember, rising interest payments will push the current account deficit up even if the trade deficit stabilizes.   I would be shocked if (net) interest payments do not add more than $50 billion to the US current account deficit in 2006.  And I also suspect the trade deficit has not peaked either – at least not in dollar terms.   On this, I disagree with Calculated Risk.  

The euro/ $ and yen/ $ are not the only factors that drive the US trade deficit – far from it.  One of my pet peeves to is to equate the dollar with the dollar/ euro.   Bringing the trade deficit down will require a far broader set of global adjustments: the currencies of emerging Asia need to appreciate against the dollar, consumption growth in the US needs to slow, consumption growth in China needs to pick up and, if oil prices do not fall, the oil states need to consume more too.  But the euro/$ and the yen/ $ do matter – US made goods compete with goods from Europe and Japan in a range of markets, not just in the US.  And the dollar's rise against the euro and yen in 2005 should start to slow US export growth in 2006.   

And even if Calculated Risk is right and a housing slowdown starts to drag down the pace of US consumption growth, I doubt it will drag non-oil import growth down enough to prevent the trade deficit from rising.   Right now, a number of signs (October US import data, October and November trade data from China and Korea) suggest that the US inventory correction is over, and non-oil imports are ticking up – that is what one would exect.  US imports recently have tended to grow about twice as fast as US consumption growth.   Above all, though, the math is just daunting.  Exports have to grow 60% faster than imports, more or less, just to keep the trade deficit stable.

The big wildcard: oil.  A big fall in the price of oil might lower the US deficit significantly — maybe even by enough to offset the expected rise in interest payments.  But a big fall would presumably need a big trigger.

Read more »

Things I got wrong in 2005

by Brad Setser Friday, December 30, 2005

Alas, this list is rather long.   There is a reason why Mike Dooley ended the first segment of our Econoblog debate by noting that he had been right (and I and other worry warts had been wrong) for 28 months.  Make that 30 now.

In the paper I did with Dr. Roubini critiquing the stability of the Bretton Woods two system of central bank financing of US deficits, we predicted that the system would show signs of increasing strain in 2005, and perhaps crack in 2006.   In reality, the system showed far more signs of stress at the end of 2004 than it did in most of 2005.  Our paper lagged rather than led.

Going out on a limb has risks.

So, here is a (no doubt partial) list of things that I got wrong:

  1. US interest rates.  I certainly did not expect the yield on the ten year bond to average 4.27 during the first 11 months of this year, almost exactly its average in 2004.   Nor did I expect the yield curve to invert.   At the end of 2004, the ten-year was somewhere between 4.2 and 4.3.  Now it is somewhere between 4.3 and 4.4.   But US short-term rates a bit higher now than back then.  I continue to be astonished by the willingness of foreigners to buy long-term dollar denominated US bonds at current US interest rates in the face of a large (and still rising) US current account deficit. In both the first part of this decade and the first part of the 1980s, the US trade and current account deficit expanded.  But right now, unlike in the 1980s, the US is not drawing money by offering a high interest rate on dollar bonds.  At least not a high absolute rate.
  2. An interest rate differential that supports the dollar need not come from a rise in US interest rates.  It also can come from falling rates abroad.  If I could change one thing in the Bretton Woods 2 paper, I would put a bit more emphasis on non-US interest rates.
  3. The dollar/ euro.  I did not expect the dollar to keep on falling v. the euro so much as for other countries to gradually catch up to the euro.   But I also did not expect the dollar to appreciate against the euro.  That reduced the gap between the dollar's depreciation against the euro since 2002 and the dollar's depreciation against key currencies in Asia and the emerging world.  But it did so mostly by reducing the dollar's cumulative depreciation against the euro.
  4. The dollar/ yen.  And the dollar/ Asia more generally.
  5. Risk spreads in emerging economies.  They fell.  Yet when Bailouts and Bail-ins came out, Nouriel and I argued that the combination of low interest rates, strong US and global growth and high commodity prices – ideal for emerging economies – was too good to last through 2005.  We were wrong.  Just look at the money that flooded into places like Turkey this year.   Or read Mohammed El-Erian's PIMCO swan song.   After the EMBI's (the index of dollar denominated emerging market bonds) great run in 2003-2004, I certainly did not expect the EMBI to deliver as strong a performance in 2005 as in 2004.   See figure one in El-Erian's presentation.
  6. The impact of $60 a barrel oil on the US economy. Or more precisely, the lack of a bigger impact.  See Martin Wolf.  Higher oil was not such a surprise per se, but the absence of a larger drag from higher oil prices certainly was.    I suspect the recycling of petro-profits into petro-dollars is a key factor supporting low US interest rates.  And low US interest rates support the high valuations of many assets, not the least housing.   Indirectly, then higher oil prices have helped to support the housing led US economy.  Who would have guessed.
  7. Oil exporters' willingness to finance the US — to join Bretton Woods 2, so to speak. Higher oil prices shifted much of the world's current account surplus from oil-importing Asia to oil exporting Russia and the Middle East.  That allowed the weakest links in the Bretton Woods two system – the smaller Asian central banks – to get out of the business of financing the US.   Korea, Thailand and India all saw their current account surpluses shrink and in some cases shift into deficit.  the pace of their reserve accumulation also slowed dramatically.    Conversely, the surpluses of the oil exporters absolutely exploded.  To date, the oil exporters seem to be willing to lend the dollars they earn selling oil to the (oil intensive) economies in Asia back to the US, helping the US finance both its own oil imports and its imports from Asia.
  8. China's exchange rate regime.   I read at least one set of tea leafs correctly, and did get the timing of China's shift to a basket peg more or less right.  But even if I got the timing almost right, I got the magnitude of the change wrong.  I expected a bigger move.  And it turns out that China really did not make much of a change in its regime either.  IDEA global calculated that China now pegs to a basket of 98% dollars, 1% yen and 1% euro with a small trend appreciation against the dollar.   For an economist, the irony in a basket composed of 98% dollars is obvious.  Let's agree to call it a  dollar peg once again.
  9. The ease with which China has sterilized a $20 billion monthly reserve increase – a reserve buildup of well over 10% of China revised GDP.  I thought that China would find it difficult to sterilize its (still very rapid) reserve buildup, and that these difficulties would spill over into faster than desired monetary growth, potential inflationary pressures and higher interest rates on sterilization bills.  I was wrong.   Strong deposit growth combined with administration curbs on lending left the Chinese banks very liquid, and they opted to buy PBoC bills for next to nothing rather than hold cash.   Right now, there is not much evidence of inflationary pressures inside China.  The  slowdown in bank financed investment engineered by the PBoC during 2004 and early 2005 – perhaps aided by controls on some energy prices – seems to have worked.  If anything, high saving/ low spending/ potentially over-built China may face the opposite set of pressures. 
  10. The sharp fall off in recorded central bank support for the US.   Think $200 billion in 2005 v $400 billion in 2004  (Or something like $300 billion in 2005 v. $500 billion in 2004, if you using the BIS's broader measure of central bank financing of the US)  But on this, I dispute the data.  I think the fall off in observed central bank support is more apparent than real.   Central bank flows through offshore centers are being registered as private flows, and offshore dollar deposits from central banks are supporting other private flows. 

Certainly, the world's emerging markets have not stopped adding to their reserves.   They just seem to have stopped investing those reserves in the US.  And I would emphasize the world "seem."

Once all is said and done, I suspect that China will end up adding around $250 billion to its reserves.  That includes the $15 billion transferred to a state bank, and an adjustment for the fall in the dollar value of China's euro and yen denominated reserves.   And I would bet far more than 40% of those reserves are invested in dollars, even if only 40% of China's reserves show up in the US data.

Read more »

China’s reserves. Why did reserve increase by less than China’s trade surplus in November?

by Brad Setser Thursday, December 29, 2005

Suppose that you had $800 billion at the end of September. OK, not $800 billion exactly. Let's say $560 billion in dollars, $160 billion in euros (euro 132.7 billion), and $80 billion in yen (yen 9063.2 billion). By the end of November, that portfolio would have shrunk to $792.2 billion. The euro fell by 2.22% against the dollar (mostly in November); the yen fell by 5.32% against the dollar.

China, of course, didn't quite have $800 billion in reserves. Try $769 billion. Though I would argue that the real number is $829 billion, counting the $60 billion that have been transferred, at least nominally, to three state banks. And I certainly don't know the currency composition of China's reserves. 70% dollars, 20% euros and 10% yen is just a guess – though it is one of the guesses made by the IMF's former mission chief to China and his current colleague in the IMF's research department.

But it does suggest one reason why the pace of increase in China's dollar reserves slowed significantly in October, and then slowed even more in November. China is sailing into a significant headwind, as valuation losses could easily have reduced the dollar value of its foreign exchange reserves by $7.8 billion, if not more, over the last two months.

China has leaked that its reserves totaled $784.9 billion at the end of October, and $794.2 billion at the end of November. That implies that China's reserves increased by $15.9 billion in October on the back of a $12.0 billion trade surplus, and then increased by another $9.3 billion in November. The $9.3 billion increase in November is quite low – it is smaller than China's $10.5 billion November trade surplus.

I think it is safe to assume that the dollar value of China's existing reserve portfolio fell substantially in November, generating valuation losses that slowed the increase in China's reserves.

Of course, China reserves generally have increased by more than the country's trade surplus – big net capital inflows have contributed significantly to China's growing reserves. In the third quarter, those capital inflows were in the range of $10 billion a month. Roughly $5 billion in FDI and $5 billion in hot money inflows. Add together a $10 billion monthly trade surplus and $10 billion in monthly inflows generates reserves growth of around $20 billion a month.

Had China's reserve growth continued at that pace in October and November – setting valuation gains/ losses aside – China's reserves would now be $809 billion, rather than $794.2 billion. Even if you add in $7.8 billion in valuation gains, that only brings the (adjusted) total up to 802 billion.

Read more »

Feldstein disagrees with Greenspan (politely)

by Brad Setser Thursday, December 22, 2005

I am a bit late here, John Berry reported on Martin Feldstein's paper on monetary policy in a world of more integrated global capital markets over a month ago.

But it is still interesting to read Feldstein's paper along side Greenspan's latest musing on US external deficits. 

It turns out Martin Feldstein thinks that central bank reserve accumulation has a thing or two to do with the fall in home bias that Alan Greenspan emphasized in his speech.  Greenspan portrays the fall in observed home bias as the result of capital market integration, the retreat of the state from the market and private investment decisions.  Feldstein seems to disagree:

"Although these government purchases of dollar bonds appear to be an increase in global capital market integration, it is really very different from the kind of capital market integration that standard textbook analysis envisions."

I, obviously, have a great deal of sympathy for Feldstein's view.   

One of the more interesting points Feldstein makes is that the fall in home bias among OECD countries – and the resulting gap between national savings and national investment – seems far more pronounced for small countries than or large countries.  Weight the data by GDP, and you get a different result. 

Feldstein notes that the fall in home bias among small OECD countries – what Greenspan would call increased dispersion between national savings and national investment – seems driven by two countries in particular.  Ireland, with a history of large current account deficits financed by large capital inflows from abroad, and Norway, with its large current account surplus.   Both Ireland's deficit and Norway's surplus are arguably the result of unusual circumstances.  Ireland emerged as one of the world's favorite tax havens – leading a whole host of US pharmaceutical firms to locate drug manufacturing there.   And Norway's large current account surplus reflects the Norwegian decision to use their oil windfall to build up their petroleum fund's external assets. 

Read more »

Memo to John Snow: Do housing bubbles count?

by Brad Setser Thursday, December 22, 2005

John Snow's latte-drinking budget mythmaking has gotten plenty of (fully warranted) criticism already, but I'll still pile on.

Snow argues that the large budget surplus at the tail end of the Clinton Administration was the product of the stock market bubble.  No doubt true to some degree.   But what counts is that the Clinton Administration saved rather than spent the windfall, building up ammunition for a counter-cyclical fiscal policy. 

Today, the .com bubble has been replaced by the .home bubble.   Or at least localized .home froth.   Housing prices in much of Kansas (and, for that matter, most of the depopulated, poor, rural Great Plains) have not kept up with housing prices on the coasts, or in other places experiencing rapid growth.    No matter though – the market cap of US housing is not driven by the market price of homes in shrinking small towns in the great plains, but rather by metropolitan New York, Washington, Boston, Los Angeles, San Francisco, Las Vegas, Miami and the like.

If the housing market starts to deflate and the housing ATM shuts down, that will have consequences for US interest rates (see Paul McCulley), for US equities (see Barry Ritholtz), and, I would suspect, for the budget. 

I wish I knew of a good estimate of the impact of rising house prices on both the federal budget and local government revenues (think property taxes).  But with the federal deficit now heading back up to the $400 billion (it would be higher but for the social security surplus), it seems pretty clear that the Snow Treasury has not built up the kind of fiscal buffer in (relatively) good times that would make running counter-cyclical fiscal policy easy should .home froth dissipate.

The .com capital gains windfall (and the income tax windfall from folks exercising stock options) was not the prime reason for the fiscal improvement in the Clinton years either.   Revenues rose for a host of reasons, and expenditures were kept under control — without, generally speaking, a boost from very low policy rates that lowered the government's interest bill.   In 2005, I would bet, the combination of relatively low US interest rates, a one-off surge in corporate tax revenue and the direct and indirect effects of housing froth probably made the budget look to be in better shape than it really is, given the administration's tax cuts and its other policy initiatives.   Prescription drugs and keeping a big chunk of the US Army in Iraq do cost money.   Time will tell.

Just what are the world’s central banks doing with their (growing) reserves

by Brad Setser Wednesday, December 21, 2005

The IMF just released new quarterly data on global reserve accumulation (Hat tip, Menzie Chinn).   Just don't look to it to find out if the world's central banks are adding to their dollar or euro reserves.  Of the $294 billion increase in global reserves in the first three quarters of this year, $209 billion of it has come from countries that do not provide data on the currency composition of their reserves to the IMF.

Consider the third quarter.

  • The increase in global reserves: $95.3 billion.  
  • The increase in dollar reserves among reporting countries, according to the IMF data:  $21.1 billion.

So over $74 billion of the increase in global reserves in the third quarter came from countries that don't report the currency composition of their reserves to the IMF.   Yep, China doesn't report.

We actually know a bit more than can be divided from the IMF data alone. Official inflows to the US in the third quarter, according to the US data, were $38.4 billion.  But even if we use the US data for the total increase in dollar reserves, there is still a big gap — $56.9 billion – between the recorded official inflows to the US and known increase in global reserves.   Like Martin Feldstein and a host of others, I suspect that the US data signficantly understates central bank flows.

And there is plenty more detective work than can be done using the IMF's data, and a few other sources …

Read more »

The Chinese conundrum: External (financial) strength, domestic (financial) weakness

by Brad Setser Monday, December 19, 2005

I disagree with Michael Mandel on the US current account deficit – I think it is a concern, he doesn't.  We also disagree on the US fiscal deficit.   He doesn't think it is a big problem; I disagree.  If US households don't save, the US government cannot run a fiscal deficit without adding to the overall national savings shortage, and a worrying large current account deficit.

But we agree that the rapid expansion of credit inside China over the past few years poses a serious risk to the long-term health of its financial system. I just finished a paper that looks at China's financial vulnerabilities and tries to relate them to macroeconomic developments inside China over the past few years – soaring savings, soaring investment, soaring bank lending, soaring foreign currency reserves, soaring exports and the like.  

That paper, alas, is behind the RGE firewall: I have to make a living.

But the basic argument is fairly straightforward:

Most severe banking crisis in emerging economies – at least recently – have coincided with a currency crisis.    Many such crises, particularly in Asia, had their roots in an expansion of bank credit which drove up domestic demand, often generated a real estate boom and usually generated significant current account deficits.   In some cases, those deficits were financed by short-term debts denominated in foreign currency.   When external creditors – or domestic residents — lost confidence in the country, its currency collapsed.    The falling currency increased the real value of foreign currency denominated debts.   Firms with weaker balance sheets could no longer get access to credit – if they could pay at all.   Banks faced bankruptcy as a result of rising non-performing loans (NPLs) without a costly government bailout.  The crisis countries typically raised domestic interest rates to defend the currency, further adding to the distress of the financial system.    

China clearly does not fit the stylized facts that emerged out of the "Asian crisis" perfectly.  Broadly speaking, China shares most of the domestic vulnerabilities of the Asian crisis countries on the eve of their crises, but not their external vulnerabilities. 

Read more »

Tim Adams frets …

by Brad Setser Monday, December 19, 2005

Treasury Under Secretary Adams worries about "which developing country is ready to go bust, what hedge fund is heading toward a meltdown, which currency is teetering on the edge of collapse, what protectionist legislation Congress is yearning to pass."

My leading candidate for the currency teetering on the edge of collapse: the dollar.  Obviously, the markets disagree. 

But a rally in the currency of a country with a 7% of GDP current account deficit (and in all probability a growing income deficit) looks like a recipe for future trouble.   I am what Stephen Jen would call a structural dollar pessimist.   7% of GDP trade and transfer deficits and a 10% of GDP export base is not a good long-term combination for the dollar, particularly if net interest payments on the US external debt are set to rise significantly.   Deficits did not matter to the market in 2005, but they may start to matter in 2006.   Even Stephen Jen is tempering his dollar bullishness a bit.

That is why I applaud Treasury Undersecretary Adam's attempt to "urge everyone to prepare for a set of conditions that are less benevolent than the current ones."

My list of things of trouble spots would not be quite the same as Adams.   He left the now-set-to-expand US fiscal deficit, and the structural gap between government spending and tax revenues off his list.  The 2006 fiscal deficit may return to the $400 billion range.  The problem is not that the US fiscal deficit is so large absolutely relative to US GDP, but rather that if the US fiscal deficit stays large or even grows, the only way the current account deficit can adjust is through a large fall in private investment, or a sharp increase in private savings.  

Adams beats up on continental Europe a bit more than I would; the current account surpluses that finance the US deficit are not found in Europe, and, relative to the consensus, I put far less emphasis on a growth rebound in Europe.  I don't expect a European current account deficit financed by emerging economies to replace a US current account deficit financed by emerging economies.  I would be happy if Europe can just keep on growing should the euro once again start to appreciate.   I don't think US protectionism is the only thing China needs to worry about either; it also shoudl worry about a sharp fall in investment.  That would lead to a rising Chinese current account surplus, a China that is even more dependent on exports and thus less willing to change its dollar peg, and a likely confrontation with the US in a (Congressional) election year — or, if the investment slowdown doesn't materialize, in the Presidential election cycle. 

I would fret about excessive savings, large budget surpluses and dollar pegs in oil exporters – not just froth in their local equity markets.   Oil exporters have to play a role in the resolution of global imbalances one way or another.  Either oil prices fall, or consumption in oil states needs to rise … 

Read more »

First Brazil, now Argentina

by Brad Setser Sunday, December 18, 2005

Both Brazil and Argentina now want to repay the IMF before the end of the year.  Call it contagion of a different sort.  Turkey may not be far behind, given how much private money is now flooding into Turkey.

Kirchner announced on Thursday that Argentina would pay its entire $9.8 billion debt to the International Monetary Fund by year's end. Two days earlier, Brazil pledged to pay its $15.5 billion to the lender this month.  "Let's be sincere. If Brazil hadn't taken the first step, Argentina wouldn't have been able to advance on its own," Kirchner is quoted as saying to leading newspaper Clarin.  The Peronist leader said wiping out the IMF debt was the "most important" thing he had done since taking office in 2003, adding that it had "vast internal and external consequences."

Paying the IMF offers obvious political benefits for both Brazil's Lula and Argentina's Kirchner.  The IMF is not terribly popular in Brazil, let alone Argentina — even though Brazil would almost certainly joined Argentina in default without the IMF.  

The whole idea behind IMF lending to countries facing financial pressure is to provide a bridge to the return of financial stability.  Brazil got a ton of IMF money in 2002 and 2003.  It should be paying it back now.  By the end of 1997, Mexico had repaid most of the money it received in 1995.  In Bailouts and Bail-ins, Dr. Roubini and I used repayment of the IMF as a sign of success.   The IMF lends in bad times, and should get paid back in (relatively) good times.   The problem comes when the IMF doesn't get paid, not when it does!

Argentina got a ton of money in 2001, and after it defaulted, many wondered when it would be in a position to repay the Fund.   But with $27 billion in the bank, Argentina is now in a position where it can pay.

That too is a success of sorts.  It also means that the IMF will not be in a position to shape the Argentina's future policies.  That will disappoint some – particularly those who were hoping that IMF pressure would force Argentina to reopen its exchange and give those still holding Argentina's defaulted bonds a second chance to go into the deal.

On the other hand, the IMF has not played much of a role shaping Argentina's policies over the past few years in any case.  

Read more »