Brad Setser

Follow the Money

Cross border flows, with a bit of macroeconomics

Worry about a fall in China’s demand for the world’s goods, not a fall in China’s demand for Treasuries

by Brad Setser Sunday, January 18, 2009

Last week, Sam Jones of FT’s Alphaville wrote:

If the Chinese economy collapses, or even slows dramatically, then the raison d’etre for the country’s huge FX reserves – as a sterilisation measure to dampen domestic inflation – will evaporate. With that, so will China’s US Treasury holdings.

Or alternatively the Chinese could devalue the yuan. Either way, the US will be in trouble.

I like a good China scare as much as any one. But the first concern is, I think, off. A slump in China doesn’t mean an end to Chinese financing of the world, or even necessarily a fall in China’s reserves. Let me see if I can explain why.

Suppose China’s economy slows sharply — a not-impossible development given the rather starling fall in the OECD’s leading indicators for China. How would that impact China’s balance of payments?

The first impact is rather obvious. China would import less. It would buy less. And since the rise in Chinese demand helped push the price of various commodities up, it stands to reason that a fall in Chinese demand would push prices down. It probably already has. That implies a big fall in China’s import bill, and a larger trade surplus. A slowing global economy would hurt China’s exports, but in this scenario China would slow more than the world. That means China’s imports would fall more than its exports. China’s trade surplus would rise.

But, you might say, the current account surplus is determined by the gap between savings and investment. Why would that change in a slowdown? Simple. China’s slowdown reflects a fall in investment (especially in new buildings and the like). Less investment and the same level of savings means a bigger current account surplus. In practice, though, savings would also likely fall a bit — as a slowdown would cut into business profits and thus business savings. It possible that China’s households would reduce their saving rate to keep consumption up as their income fell. But it is also possible that Chinese households might worry more about the future and save more. My best guess though is that the fall in investment would exceed the fall in savings, freeing up more of China’s savings to lend to the world. That surplus savings has gone into Treasuries and Agencies in the past.

The second impact is harder to assess. A big fall in output might lead China’s savers to lose confidence in China, or rather to lose confidence in the RMB bank accounts as a stable store of value. The big fall in output might, for example, create expectations that China would devalue the RMB v the dollar — and Chinese households, anticipating the devaluation, would have a strong incentive to hold dollars. That likely means a rise in capital outflows.

Since reserve growth is a function of both the current account balance and capital outflows, it is possible that the rise in capital outflows could overwhelm the rise in the current account surplus. That seems to be what happened in q4.

Read more »

A few quick words on the November TIC data

by Brad Setser Friday, January 16, 2009

China sold $9.2 billion of long-term Treasuries. But it also bought $38.2b of short-term Treasuries. China’s total Treasury holdings are up by $29.1b. By contrast it sold $3.1b of long-term Agencies and also reduced its short-term holdings by about $5 billion. China reallocated its US portfolio, but it hasn’t cut back on its dollar purchases.

The following graph, prepared with help from Arpana Pandey, plots the average increase in China’s reserves (defined broadly, to include hidden reserves) over the last 3 months v my best estimate (taking flows through London into account*) of China’s Treasury and Agency purchases. It speaks for itself.

The same story applies to the official sector as a whole. Central banks sold $26.2b of long-term Treasuries, but added $66.6b to their short-term Treasury holdings. Net central bank holdings of Treasuries — judging from the TIC data — rose by $40.4b. That is consistent with the $49.1b rise in the Fed’s custodial holdings of Treasuries. Central banks by contrast are reducing their holdings of short-term and long-term Agencies. They sold $14.3b of long-term Agencies, and their short-term holdings likely fell by a comparable amount.**

The countries that are really running down their Treasury portfolio — Korea and Brazil — are countries whose reserves are falling and need the cash. Russia is running down its Agency portfolio for a similar reason. It really needs the cash. Its short-term Agency holdings are down to $13.7b. They were close to $100b — 496.8b — in December of 2007.

The big stories in the TIC data, it seems to me, are:

— The ongoing reallocation of central bank portfolios toward short-term Treasuries. That reallocation has been huge. Central banks held $$276.8b of t-bills at the end of September. They hold $427.2b at the end of December. And judging from the Fed’s custodial data there is every reason to think that total rose in December. Foreign central bank demand for safe and liquid assets rose at the same time as private demand rose.

Read more »

How badly were the Gulf’s sovereign funds hurt by the 2008 crisis?

by Brad Setser Thursday, January 15, 2009

It takes a bit of courage to put out a paper that is sure to get a few things wrong. But when it comes to the Gulf’s sovereign funds, the alternative to getting a few things wrong is not writing much at all. The funds cloud themselves in secrecy. Educated guesses have to substitute for analysis based on hard data.

The large Gulf sovereign funds are financed out of oil revenue, so the amount of new money they have to invest abroad is presumably linked to size of the fiscal and current account surpluses of key Gulf states. If – and it is a challenge – the path of spending and investment can be estimated, the size of that surplus will be largely a function of the price of oil. Production volume matter too, but in most circumstances production changes more slowly than price. And the Gulf funds are known to be large investors in the world’s equity markets, so their performance is likely to track, at least in broad terms, movements in major equity indexes. Sure, they have invested in “alternatives” – London real estate, private equity, hedge funds – too. But most “alternative” investments also have performed poorly over the past year.

Rachel Ziemba of RGEMonitor and I used these basic insights to built a model of the Gulf funds that allows us to estimate – very roughly – the trajectory of the various Gulf funds over the past few years. That model is only going to be as good as our assumptions – and while we made every effort to calibrate the model using available data it no doubt is going to be somewhat off. That probably isn’t the best advertisement for the Setser/ Ziemba paper on the Gulf’s large sovereign funds. But sometimes caveats are important. This is a paper with a lot of known unknowns.

Among other things, Rachel and I argue:

— The capital losses on the Gulf’s existing portfolio overwhelmed large inflows from high oil prices in 2008. Close to $300 billion flowed into the big Gulf funds — the Abu Dhabi Investment Authority/ Abu Dhabi Investment Council, the Kuwait Investment Authority, the Qatar Investment Authority and the Saudi Arabian Monetary Agency’s foreign assets. But the market value of their Gulf’s foreign portfolio fell by an estimated $350 billion over the course of 2008. Throw in a roughly $30b fall in the Gulf’s reserves as hot money betting on a revaluation left and the total value of the Gulf’s external assets likely went down over the course of 2008.

Read more »

Secrets of SAFE: A sharp slowdown in reserve growth and large “hot” outflows in q4….

by Brad Setser Tuesday, January 13, 2009

China’s formal foreign exchange reserves rose by about $40 billion in the fourth quarter, rising from $1906 billion to $1946 billion. Adjusted for valuation changes, that works to a $55 billion or so increase. But appearances can be deceiving.

In the past, China “hid” the pace of increase in its reserves by forcing the banks to hold dollars as part of their required reserves. Those dollars showed up on the PBoC’s balance sheet as “other foreign assets” but weren’t counted as part of China’s formal reserves. In the fourth quarter, China’s reported reserves overstate its true reserve growth, as the banks reserve requirement was cut. That led to a $26 billion fall in the PBoC’s other foreign assets in October and, I assume, a comparable fall in December. Given the size of the reduction in the reserve requirement in December, that is conservative.*

That means that the PBoC’s “true” reserves didn’t grow at all in the fourth quarter, best I can tell.

Valuation changes had a fairly modest impact on the data for the quarter as a whole, but a huge impact on the data for individual months. They pulled reserves down in October and pushed reserves up in December. Both effects were quite significant – in the $50 billion range. That has one big implication: China was adding to its reserves (if reserves are defined broadly to capture the PBoC’s other foreign assets) in October but not in December. My best guess is that China added about $15-20 billion to its reserves in October, another $10 billion in November and lost $20 billion in December. My numbers are a bit different than those of Morgan Stanley’s Wang Qing. He believs capital outflows peaked in October. I would put the peak in December — which incidentally implies that the small RMB devaluation that China tried in early December had a big impact on capital flows.

I wouldn’t put too much emphasis on the monthly estimates though. Combine China’s huge reserves and large moves in the currency markets and you get large valuation changes. If my estimate of the currency composition of China’s reserves is off, my monthly estimates will be a bit off too. The trend though is clear. Chinese reserve growth looks to have peaked earlier this year.* On a rolling 3m basis (adjusting for valuation changes and likely changes in the PBoC’s other foreign assets), Chinese reserve growth has essentially stopped.

Read more »

The fall in the US trade deficit in November

by Brad Setser Tuesday, January 13, 2009

I’ll be unusually brief. I have been working through the latest data on China — and, well, it is hard to top Calculated Risk.

The fall in the US trade deficit today was easy to anticipate. The average price of imported oil was sure to fall from $92 in October. It came down to $67 a barrel in November. The petrol deficit fell by about $13 billion.

That wasn’t all though. The non-petrol goods deficit also fell by about $4 billion. Y/y non-petrol goods imports were down 9.3% (they were down about 1% in October); y/y non petrol goods exports were down 4.4% in November. They were up 2.4% in November. So long as non-petrol imports fall faster than non-petrol exports, the US deficit will shrink.

This shows up in clearly in a plot — prepared by Arpana Pandey — of real goods imports and exports. In November the down turn in imports was sharper than the downturn in exports.

There isn’t anything good in this graph other than the fall in the trade deficit. Falls in exports and imports signal contracting global activity. And I am not even sure that the improvement in the trade balance will continue, as I suspect the combination of a stronger dollar and a broadening global slowdown will eventually have a major impact on exports — and the US fiscal stimulus will bleed into imports. But the US led the world down and for now, US imports are falling faster than US exports …

The US government has already proved it can raise over $1.5 trillion in a year ..

by Brad Setser Monday, January 12, 2009

The large deficit projected for fiscal 2009 stunned many. It is natural to wonder how such a huge deficit could be financed. But remember one thing: the US placed $1685 billion of Treasuries in the market in 2008 without pushing interest rates up.

Some facts.

In 2008, the stock of marketable public debt rose by $1257 billion.

The Fed’s holdings of Treasuries – counting the securities it has lent out to the market – fell by $427 billion.

That implies an absolutely huge increase in the stock of Treasury debt in the market. The outstanding stock of marketable Treasuries not held by the Fed rose from about $3795 billion to about $5480 billion.

All that debt had to be bought by the PBoC and other foreign central banks right?

Well, yes and no.

The Fed’s custodial holdings of Treasuries rose by $481 billion. That leaves $1.2 trillion of the $1.7 trillion increase in the stock of marketable Treasuries in the market in private hands. Foreign central banks don’t hold all their Treasuries at the Fed. Past data revisions – and a bit of extrapolation on my part – suggest that centrals bought $192 billion that doesn’t show up in the Fed’s custodial accounts. That implies a nearly $700 billion increase in central banks’ holdings of Treasuries.

Read more »

This really doesn’t look good

by Brad Setser Sunday, January 11, 2009

Words don’t really do justice to the brutality of recent downturn in Korean and Taiwanese exports.

These look a lot like charts of financial variables after a bubble bursts, not charts of the level of exports. That isn’t good.

Looking just as the monthly data risks being misleading. There is a lot of seasonality in Taiwan’s exports. They usually dip in February. It is a short month, it often corresponds with the Chinese new year and the data isn’t seasonally adjusted. A small dip in December after the end of the Western holiday season also isn’t unusual. But such a big dip in December is most unusual. Plotting the rolling 3m sum eliminates the big February dip. The current downturn is real.

The data do not look much better if plotted as a percent change. The y/y change in the 3m rolling sum isn’t quite as bad as it was at the depth of the tech bust. But give it time. The 40% fall in Taiwan’s December exports is worse than anything observed then.

Almost all of Taiwan’s exports seem to go to China for final assembly. Korea though exports to both China (electronic components, steel, no doubt other products) and the US (cars). And there has been a very sharp fall in both Korean shipments to both the US and China. Then again, it seems that exports are down across the board. Europe doesn’t look much different.*

Read more »

The global savings glut and the current crisis

by Brad Setser Friday, January 9, 2009

I have always believed that the debtor and the creditor tend to share responsibility for most financial crises. One borrows too much, the other lends too much. Wynne Godley (ideologically, no Hank Paulson), Dimitry Papadimitrriou and Genaaro Zezza seem to agree. They write:

“The process by which U.S. output was sustained through the long-period of growing imbalances could not have occurred if China and other Asian countries had not run huge current account surpluses , with an accompanying “savings glut” and a growing accumulation of foreign exchange reserves …. flooding the US market with dollars and thereby helping to finance the lending boom. Some economists have gone so far as to suggest that the growing imbalance problem was entirely the the consequence of the savings glut in Asian and other surplus countries. In our view, there was an interdependent process in which all parties played an active role. The United States could not have maintained growth unless it had been happy to sponsor, or at least permit, private sector (particularly personal sector) borrowing on such an unprecedented scale.”

Thanks to Martin Wolf for highlighting the Godley et al paper.

Their argument seems right to me. Absent a large savings surplus in Asia and the oil exporters, rising US rates would have choked off the housing bubble much earlier. High long-term rates aren’t conductive to rising home prices — and without rising home values it is hard to turn a home into an ATM. Felix Salmon writes:

it was the global liquidity glut, and the concomitant demand for a bit of extra yield on fixed-income assets, which encouraged the lax subprime underwriting which etc etc. If the world hadn’t been perfectly happy to throw trillions of dollars a year into America bottomless appetite for capital, the bubble would never have happened, and neither would the subsequent bust.

At the same time, US policy makers didn’t exactly try to reign in US borrowing or leverage in the financial sector. They were far more likely to cheer this process on than to try to get in the way.

Five additional observations:

There is little point denying that there was something of a “savings glut” in much of the emerging world. Just look at Table A16 of the statistical appendix of the IMF’s WEO. In 2006, 2007 and 2008 the developing world’s savings was around 33% of its GDP, well above its 24% average in the late 80s and 90s. That allowed emerging economies to both increase investment above their historic levels (think of all the construction in the Gulf and China) and at the same time lend unprecedented sums to the US and increasingly Europe. (Graph here)

Read more »

More bad trade data from Asia

by Brad Setser Friday, January 9, 2009

Taiwan’s December exports are down over 40% y/y — largely because of a huge fall in exports of electronic components to China. The fall topped expectations. It adds to the mounting evidence that the current global deceleration is quite sharp and quite deep.

Taiwan’s imports are also down by around 45%. Thank the big fall in commodity prices. Taiwan managed to post a trade surplus in December even with the enormous fall in its exports.

Korea’s exports were also down in December — though the 17.4% y/y fall in December wasn’t quite as large at the 19% fall in November. Imports though fell by more — over 21%. And Korea also posted a (small) trade surplus.

My guess — based on the fall in Taiwan’s exports to China and the fall in commodity prices — is that China’s December trade data will show a quite significant fall in imports, keeping China’s surplus large even as China’s exports fall sharply.

The large fall in Germany’s November exports highlights that the slump in trade is global. Japan’s exports fell by more (in percentage terms) than Germany’s exports in November. US exports are almost certain to start shrinking too.

Expectations have been revised down dramatically, but most data points still seem to be worse than expected. That worries me.

China hasn’t (yet) lost its appetite for US Treasuries …

by Brad Setser Thursday, January 8, 2009

Agencies, yes. But not Treasuries.

Keith Bradsher of the New York Times, citing Ben Simpfendorfer of RBS, argues that China’s government is likely to reduce its purchases of US debt.

“All the key drivers of China’s Treasury purchases are disappearing — there’s a waning appetite for dollars and a waning appetite for Treasuries, and that complicates the outlook for interest rates,” said Ben Simpfendorfer, an economist in the Hong Kong office of the Royal Bank of Scotland.

In some sense China’s purchases of US debt has to fall from its current level, as the current level of purchases is unsustainable in a context where China’s reserve growth seems to have slowed. The TIC data show a $44.4b increase in China’s US holdings in September and a $67.5b increase in October, with nearly all the increase coming from the rise in China’s short-term Treasury holdings.

That said, the available data from US suggest that China has yet to lose its appetite for either dollars or Treasuries, despite all the talk coming out of China.

We don’t have data for November or December, so the available US data points are by now a bit stale. But China’s $67.9b of purchases of Treasuries in October were exceptionally high ($43.5b in September isn’t shabby either). That level of Treasury purchases suggests, if anything, that China was shifting funds into dollars, as China’s recorded US purchases almost certainly exceeded China’s October reserve growth. I suspect that China wasn’t shifting into the dollar so much as holding more dollars in ways that register in the US data, so I would discount this data point a bit. Still, the raw October data doesn’t indicate a shift away from either the dollar or Treasuries. Rather the opposite.

Over the last 12 months, China’s recorded US Treasury purchases have topped $190 billion — a record. Most of the rise has come in the past few months of data. The US survey of foreign portfolio investment has tended to revise China’s purchases of Treasuries up, so $190 billion should be considered a minimum.

Read more »