Chinese Trade Elasticities, Updated

The price and income elasticities of Chinese trade flows are key parameters in the debate regarding the importance of Renminbi revaluation in achieving rebalancing. [0][1] I was hoping to update my estimates to incorporate data spanning the recent crisis, but Shaghil Ahmed at the Fed beat me to the punch with a new working paper that includes data spanning the recent downturn in Chinese trade flows. From Are Chinese Exports Sensitive to Changes in the Exchange Rate?

This paper builds a model of two types of Chinese exports, those processed and assembled laregely from imported inputs (“processed” exports) and “non-processed” exports. Based on this model, the sensitivity of Chinese exports to exchange rate changes is empirically examined. Unlike previous work, the estimation period includes the net real appreciation of the renminbi that has occurred over the past three years. The results show that greater exchange rate appreciation dampens export growth, both for non-processed and processed exports, with the estimated cumulative price elasticity being substantially greater than unity. When the source of the increase in the Chinese real exchange rate is appreciations against the currencies of other emerging Asian trading partners, the e¤ect on processing exports is positive but insignificant, while the effect on non-processing exports is significantly negative. By contrast, when the source of the increase in the Chinese real exchange rate is appreciation against China’s advanced-economy trading partners, the effects on both types of exports are
negative. These results are consistent with the predictions of the theoretical model. Counterfactual simulations based on the estimated model strongly suggest that if the trade-weighted real renminbi had appreciated at an annual rate of 10 percent per quarter since mid-2005, Chinese real exports would have been roughly 30 percent lower today. Thus greater exchange rate flexibility could contribute to lowering China’s huge trade surplus through restraining growth
of exports.

The paper uses up-to-date data (through 2009Q2), and specifications analogous to those used in Cheung, Chinn and Fujii, but using cumulative FDI as the supply shift variable, and estimated in first differences (in Cheung et al., we use DOLS).

The results indicate that real exchange rate appreciations have contemporaneous
and lagged negative e¤ects on real export growth, while foreign consumption growth
has positive e¤ects. The growth of the FDI capital stock has first a positive e¤ect
and then a small, but significant, negative one later on export growth. The long-
run solution of the statistical model, also presented in table 1, shows that a one
percentage point increase in the annual rate of appreciation of the real exchange
rate would have a cumulative negative effect on real export growth of 1.8 percentage
points, which is statistically signi…cant. A one percentage point increase in foreign
consumption growth would increase export growth by 5.9 percentage points, which
is also statistically signi…cant, and appears to be an implausibly large effect. Also,
a 1 percentage point increase in the growth rate of the FDI capital stock raises
export growth by a cumulative and statistically significant 0.3 percentage points.
This suggests significant supply-side factors at work in the determination of the
equilibrium growth rate of exports. All the estimated e¤ects are in line with theory.
The estimated model also indicates a large and signi…cant e¤ect on export growth
associated with China’s entry into WTO.

Regressions based on disaggregated (processing, non-processing) trade are also estimated. Using some of these estimates, the author conducts a simulation, assuming a 10% annualized appreciation in the real effective rate since mid-2005 (to 2009Q2), instead of the actually observed 5.5% rate. This would have meant a RMB 20% stronger than actually observed as of December. Holding all else constant, exports then would take the counterfactual path illustrated in Figure 12.


Figure 12: from Ahmed, “Are Chinese Exports Sensitive to Changes in the Exchange Rate?” IFDP No. 987 (December 2009).

In contrast to the results in Cheung et al., there is a substantial impact on exports. According to Figure 12, the gap is about 100 billion (2004$), on a quarterly basis. We found a 50 billion (2000$) impact for a 20% revaluation on an annual basis. Using an alternative error correction model, we obtain a somewhat larger 90 billion dollar impact — but this is still substantially below the estimates Ahmed obtains.

8 thoughts on “Chinese Trade Elasticities, Updated

  1. Cedric Regula

    So that would indicate that on average, Chinese company profit margins are too low to be able to absorb RMB currency appreciation in the income statement and hold selling price (in dollars) constant. Most likely due to over capacity and too many Chinese firms competing with each other.
    That would also indicate an unwillingness on the part of the Party to continue with RMB appreciation. And actually I thought they announced a year or two ago that they would halt appreciation of the RMB.
    Also, how can this study separate out the effect of declining export demand due to a weaker global market vs currency valuation on prices and therefore export volume? That seems like quite a trick to me.

  2. Tom

    This analysis examines only the extent to which Chinese exports are boosted by the suppression of wages in hard-currency terms that is accomplished by manipulating yuan exhchange rates. There is obviously some boost, but nobody who understands trade and exchange rates could possibly imagine that it could be the main reason for China’s huge trade surplus.
    The bigger factors are the suppression of imports, which is also a result of manipulating yuan exchange rates, and the sum of various other restrictions on imports.
    China would still be a very successful exporter without exchange-rate manipulation. The direct negative impact on exports of freeing exchange rates would not be large, and over time, these would be greatly outweighed by the positive indirect impacts of a more open, more competitive domestic economy.

  3. RicardoZ

    Cedric wrote:
    Also, how can this study separate out the effect of declining export demand due to a weaker global market vs currency valuation on prices and therefore export volume? That seems like quite a trick to me.
    You stole my post! 🙂 Excellent observation!
    One can do amazing analysis with “all things equal” and “in the aggregate.” Of course the results only agree with reality by accident because macro effects are simply the sum of micro activity – oops, I forgot this is Keynesian analysis so the micro doesn’t matter.

  4. Shaghil Ahmed

    The effects of China’s trading partners’ GDP/consumption on Chinese exports are included in the regressions, so the effects of weaker global demand are in there, at least as measured through these standard variables.

  5. Cedric Regula

    OK, I saw a great big equation, and some words that sounded sort of to that effect, but I wasn’t sure.
    Also, I wonder how that can work (
    “standard” data and regressions) if China hasn’t had a bad economy since the 80s, but that’s another one that economists probably have an answer for.

  6. RicardoZ

    Kevin and Shag,
    Can you give me a count of the number of times you find “variable” or “assumed” mentioned to define the formulas? Also where is the “factor” or “variable” for the contraction of the US and other western economies? I see the factor for other Asian economies.
    The natural question then is how would I write the equations? My answer, I would not. The premise can only be proven theoretically based on the assumption that the premise itself is correct. It cannot be proven experientially. It is simply Sokudo in a different form.

  7. RicardoZ

    After writing this: It is simply Sudoku in a different form I realized how unfair I was. Sudoku actually has a mathematical solution. Sorry for the confusion.

Comments are closed.