from Follow the Money

China: $154b in q1 reserve growth

April 11, 2008

Blog Post
Blog posts represent the views of CFR fellows and staff and not those of CFR, which takes no institutional positions.

More on:


Neither the fact that China added $154b to its reserves in the first quarter ($153.9b) nor the fact that the RMB (briefly) broke through 7 was exactly a surprise. China’s reserves were heading up quite nicely in January and February, and the RMB was steadily marching up (at least against the dollar) as well.

Breaking 7 though is to my mind a non-story. Yes, the pace of RMB appreciation against the dollar has picked up. But the RMB still isn’t appreciating fast enough against the dollar to offset the dollar’s depreciation against other currencies. The RMB hasn’t appreciated in nominal terms against a basket of G-3 currencies. It depreciated against the euro in the first quarter.

The US Treasury has gone out of its way to praise the increase in China’s pace of appreciation against the dollar. I think that is a mistake. The G-7 recently has called for broad-based RMB appreciation not just appreciation v the dollar. That broad-based appreciation has yet to happen. Right now the US should be supporting Europe in putting pressure on China to end the RMB’s depreciation against the euro. So long as the RMB doesn’t appreciate in nominal terms against a broad basket of currencies, all of its real appreciation will have to come from high levels of inflation.

The $154b in reserve growth in q1 is both a little less and a little more than meets the eye. I would estimate that $34b of the increase reflects valuation gains. That could well be an under-estimate of China’s valuation gains, as it is based on an assumption that China has 70% of its reserves in dollars. Netting out $34b leaves an underlying increase of around $120b.

The trade surplus in q1 was $41.4b, FDI inflows were $27.4 and a conservative estimate of China’s interest income would add another $16b (assuming $1600b in reserves and an average interest rate of 4%, or 1% a quarter). That works out to roughly $85b in reserve growth, leaving about $35b of the overall increase unexplained. Bigger valuation gains or more interest income would reduced the “unexplained” portion of China’s reserve growth - -and thus reduce the estimate of hot money inflows.

However, there is also good reason to think that China’s reported reserve growth understates the overall increase in China’s foreign assets. After adjusting for valuation, China added $53.9b to its reserves in January (when the trade surplus was large), $47.7b in February (when the trade surplus was small) and $15.8b in March (when the trade surplus was moderate).

The increase in January if not all that large relative to the underlying flows (FDI of $11b, a trade surplus of close to $20b and $5-6b of interest income). The increase in February is quite substantial -- large enough that it raises question about whether the January data reflects all the underlying flow. By contrast, the $15.8b increase in March (a level that implies large hot money outflows) is suspiciously small.

The key question is what might be left out of the reserve data, or more precisely who else inside China might be adding to their foreign portfolio to take pressure off the central bank.

There are two potential candidates: the state banks and the CIC.

It is widely thought that the state banks added something like $18-20b to their foreign portfolios in January (See Michael Pettis, among others), when they likely met the increase in the required reserves by increasing their foreign holdings. They may also have met the March increase in their reserve requirement by increasing their foreign holdings. That implies that China’s foreign assets could have increased by $36-40b more than reported reserves. $120b in reserve growth (after adjusting for valuation changes) might be $160b of reserve growth, counting the “fx reserves” the banks have been ‘encouraged” to hold to keep the fx off the central banks books.

And then there is the CIC. The Ministry of Finance issued a large bond to finance the CIC in late December. If the Ministry of Finance used the proceeds of that bond to buy fx from the central bank in q1, it could have taken $100b off the books of central bank. The unusually low March reserve growth number reminds me of the unusually low numbers in September and October of last year – months when funds were clearly shifted to the CIC. If all of the big December bond was converted into foreign exchange in q1, that could have subtracted something like $100b from China’s foreign exchange reserves, and added $100b to the overall increase in China’s foreign assets.

If valuation gains were larger than my current estimate, China’s “true” reserve increase might have been as low as $110b. If the banks increased their foreign holdings by $35b and the CIC increased its foreign holdings by $100b, the “true” increase in China’s foreign assets in q1 could have been as big as $245b. Michael Pettis has a similar estimate.

It would be nice to know. So far, the creation of the CIC -- together with the likely rise in the foreign exchange holdings of the state banks -- has made it far harder to determine how heavily China has to intervene in the market to maintain its current exchange rate regime. As a result, the overall level of transparency in the international financial system has gone done.

A few graphs might help (they follow below the fold)

The following graph shows China’s trade surplus, FDI inflows and estimated interest income along with the increase in China’s reserves all on the rolling 12m basis. Reserve growth has been adjusted to strip out valuation gains.

I have also included an estimate -- roughly $20 billion -- of the funds shifted to the CIC in q4 2007. I have NOT included an estimate for the funds shifted to the CIC in q1 because I don’t have enough data to make a good estimate. I also have not included an estimate the funds accumulated by the state banks -- including the funds the CIC injected into the state banks in q4. I am not sure that I have found a data sources that captures all of the increase in 2007 -- and there is no data yet for 2008. The graph consequently shows the minimum level of intervention that China has done to support its exchange rate regime.


The gap between reserve growth and the sum of the underlying sources of reserve growth is – of course – one measure of hot money inflows. My personal view is that the overall increase in China’s foreign assets, counting the rise in the foreign assets of the state banks, over the past 12 months has been closer to $600b than to $450-500b. That implies much larger hot money inflows that suggested in the graph above. remember, this graph doesn’t capture the increase in the foreign assets of China’s banks or the q1 increase in the CIC’s foreign assets.

The next graph shows the same data – again without an estimate of the rise in the banks foreign assets over time or the CIC’s foreign assets in q1 – on a rolling 3m basis. Not that the pace of growth in China’s visible foreign assets has swung around wildly. The very low pace of increase in the assets of the CIC and SAFE together in q4 likely reflects an enormous accumulation of foreign exchange by the state banks, not a reversal in capital inflows.


My personal view is that hot money inflows were at least as strong in the first quarter of 2008 as in the first quarter of 2007 -- and that the overall increase in China’s foreign assets in q1 was over $150b, and perhaps substantially more than $150b.

And finally a comment on China’s trade data. The rise in China’s import bill shouldn’t be a surprise. It reflects higher oil and commodity prices above all else. The y/y increase in March exports was quite strong (30% y/y). But overall q1 export growth was around 20% -- something of a slowdown. However, in dollar terms, the y/y increase in China’s exports hasn’t actually changed much. It has been around $250b for several months.

That is why I find it hard to believe that China’s export sector has really been squeezed by the RMB’s appreciation. The RMB hasn’t appreciated against China’s largest trade partner. And the increase in China’s exports – measured in dollar billion – remains very very strong.

The following chart plots the y/y increase in China’s exports (in $ billion) and China’s trade surplus (as a rolling 12m sum).


China’s trade surplus has stabilized, largely because of higher commodity prices. Over time, the IMF still expects China’s trade surplus to continue to grow. Export growth has slowed in percentage terms, but has only stabilized in dollar terms.

Draw your own conclusions!

UPDATE: Michael Pettis has produced an excellent table showing how the various possible adjustments to reserve growth would increase China’s estimated intervention in the foreign exchange market in q1, and thus increase estimated hot money inflows. He also draws attention to the excellent work Logan Wright has done to try to pin down how the state banks have been used to limit formal reserve growth and to estimate the timing of the transfer of foreign exchange to China’s reserves. I suspect Pettis’ table is easier to follow than my words; do check it out. It hardly needs to be said that the upper end estimates for China’s foreign asset growth in q1 are truly extraordinary. Not so long ago we marveled that China had amassed more than $1000b in reserves; the upper end estimates now suggest that China could end up having to add $1000b to its foreign assets (reserves, state bank’s foreign assets and the CIC) in a single year to control the pace of RMB appreciation. Even low-end estimates would put Chinese intervention in the $600-700b range.

More on: