from Follow the Money

The IMF on China

More on:


The IMF’s Article IV on China.

It suffers from one obvious problem.  It is dated.  The IMF completed this report in early July, and it is now almost November.   China’s current account surplus simply is not going to be $180b.  Q3's trade surplus showed  that.   It will be far bigger.   If the IMF thinks the ability to its analysis to stimulate public debate is central to the impact of its assessment, particularly for countries that are not likely to borrow from the IMF, it needs to shorten the lag between finishing the report and releasing the report.   Yes, Board debate is important – but, well, could this still have come out a bit earlier? 

I am eagerly looking forward to the staff report (mentioned in a footnote) which will explore the reasons why China’s current account surplus exploded in 2005.    Among other things, the IMF Article IV indicates that if you adjust for reserves shifted to ICBC and for swaps with the banking system, China’s 2005 reserve growth was $236b, not $208.   That is useful information for reserve-obsessed people like me – among other things, it implies I actually was slightly underestimating China's 2005 reserve growth (the swaps were larger than I had estimated), and thus slight underestimating global reserve growth.

The IMF also notes that China's July revaluation against the dollar has had virtually no impact on China's real exchange rate: "on a real effective exchange rate bases, the renminbi by May 2006 had returned to roughly its June 2005 level despite the revaluation."   Why?  The change in the RMB/ $ was offset by changes in the $/ euro -- and by higher inflation in the US.  Updating the calculation through the end of October wouldn't change much.   The RMB would still be about 15% below its early 2002 levels.

What else jumped out at me as I skimmed the report? 
  • The surge in China’s national savings since 2001.  Savings is estimated to have increased from 34% of GDP in 2001 to 51% of GDP in 2006.  That is extraordinary.   Household savings is high, but the IMF – like the World Bank – says that the real drivers of the surge in savings have been the government and businesses, especially business.    A lot of the debate over China comes down to why folks think business savings surged.    One explanation is rising profits in the export sector linked the RMB’s big real depreciation since 2002 (the IMF, as usual, has great charts).   Another is a set of reforms that allowed state firms to shed their social responsibilities – increasing their profitability --- without introducing mechanisms for the distribution of their profits.  And rather than deposit their profits in the banks ay 2% and change (nominal), firms had a strong incentive to invest ….
  • Investment is expected to rise from 34% of GDP in 2001 to 44% of GDP in 2006.  That kind of investment surge usually leads to a current account deficit … the puzzle of China is that it was overwhelmed by an even bigger surge in savings.
  • Consumption has fallen dramatially as a share of China's GDP since 2002 (the Economist should take note)
  • And the chart comparing bank lending growth and m2 growth shows that a ton of liquidity has been bottled up inside the banking system since early 2004.   A ton.   If the banks were ever allowed to lend all their deposits out, investment could soar even higher …  as the banks now have the ability to lend out more than they take in for quite some time.
In Martin Wolf’s analysis, China’s ability to bottle up these funds in the banking system is a key reason why the RMB has depreciated in real terms, not just nominal terms.   Only by constraining bank lending could China hold inflation down.   Turning a nominal depreciation into a sustained real depreciation required that China adopt a set of policies that raised national savings ….

That is some sense is the nub of the policy debate on China. 

Some, like Roach, McKinnon and Buiter, argue that China’s high savings rate effectively requires a weak exchange rate – China wants to save so much more than it can reasonably invest that it was to export some of its spare savings, and that means running a large current account surplus.   The corollary is that exchange rate adjustment wouldn’t work – savings would still be high, and China would still run a large surplus.

Others, like Wolf, argue that China’s weak exchange rate has required a high government savings rate – he counts state owned firms and their share of business savings as part of government savings.   The corollary is that exchange rate adjustment would work.  A stronger exchange rate would allow China’s government to reduce its savings rate. 

I tend to side with Wolf.  More later.

A quick update.   The paragraph in the IMF Article IV on the RMB was a bit stronger than I expected, given the rather tepid PIN (Press information notice) that the IMF issued after the board discussion.  I think it was appropriately hard hitting.

After noting that the renminbi has not appreciated in real terms (through May) despite the nominal revaluation, the IMF noted (paragraph 24/ p.17):

"In fact, movements in the renminbi's real value over a considerable period of time have not been in line with most fundamental facotrs that are generally considered to be important in determining the exchange rate's real value.  In particular, since the previous peak in the renminbi's real effective value in early 2002, the currency has depreciated, while such factors as a substantial net foreign assets accumulation and sharp rise in China's productivity relative to partner countries would be expected to have contributed to a real appreciation of the currency.   The current account surplus also has risen from 1.3% of GDP in 2001 to 7.2% in 2005, with the surplus doubling as a ratio of GDP in 2005 alone.  

With investment as a ratio of GDP continuing to rise during this period, the surge in the current account surplus represents a widening of China's savings-investment balance that is explained by an increase in government savings ... and a substnantial rise in corporate savings that is increasingly difficult to explain in terms of fundamental determinants of savings behavior.  It is especially difficult to pinpoint a change in the fundamental determinations that would explain the doubling of the surplus in relation to GDP in 2005 and that suggest that the surplus at its present level could be considered to be a new "normal" level of the savings-investment balance for China. ....  All of  these developments point to the currency as being undervalued and that this undervaluation has increased further since last year's Article IV." (emphasis added) 

That seems right to me.  It is tough, but appropriately so. The IMF's job is to identify exchange rate misalignments that impede the orderly adjustment of the global balance of payments. 

Given that China's 2006 current account surplus looks set to exceed the IMF's forecast in the face of continued strong investment growth, I don't think the IMF's basic view has changed.  It is hard to explain why a country like China should be running a sustained 9-10% of GDP current account surplus.

I also liked the illustrative medium-term scenario in Table 8 (p. 38), largely because it didn't forecast the problem away.   It forecasts that China will add roughly $250b a year to its reserves over the next four years (a bit less in 2007, a bit more in 2010), bringin China's total reserves to $2 trillion by the end of 2010.   And limiting China's reserve increase to $250b a year or so requires assuming both that China's current account surplus falls back to 6% of GDP ... and rather herculean assumptions in the capital account -- net FDI inflows fall significantly as a share of GDP, to the point where net inflows of FDI are offset by a rise in Chinese private bank accounts abroad.    That is a bit of a stretch, but given the difficulties with these kinds of exercises, I don't want to be too critical. 

I'll just say that I think there is a real risk that China's reserves will approach $2 trillion by 2009 -- not by 2010 -- if China isn't willing to take more aggressive policy steps.

More on: