Chinese Exchange Rate Pass-Through

Jian Wang had an interesting article on the Chinese rebalancing issue, and how renminbi revaluation would fit in. One point he raised pertained to exchange rate pass through. That inspired me to check the literature on this subject.

Figure 1 shows the yuan/dollar exchange rate (number of USD per CNY [fixed 1/28, 7:40am] CNY per USD), and the price of Chinese goods imports into the US.


Figure 1: Log nominal bilateral USD/CNY exchange rate (blue) [up is CNY appreciation], and log price of Chinese goods imports into the US, both rebased to zero at 2003M12. NBER defined recession dates shaded gray. Source: St. Louis Fed FREDII, BLS, NBER and author’s calculations.

It is interesting the price of Chinese imports kept on decreasing even after the CNY began appreciating against the dollar. This graph points out the hazards of bivariate comparisons; the omitted variable (or at least one) is the cost of production. In principle, one would want unit labor costs for Chinese exports to the US as an additional variable (page 19 IMF Article IV on China has a graph of the Chinese real exchange rate deflated by unit labor costs). (More on pass-through in general: [1], [2] [3]).

What can be gleaned from the data at hand? A very quick and dirty regression of the first difference of log price of imports on first differences of the exchange rate, the output gap, and Chinese CPI (lags of 0 to 3, 0 to 7, and 1 to 7) and monthly dummies yields an exchange rate pass through coefficient of 0.52. The adjusted R2 = 0.62, and SER = 0.0019, for a sample of 2005M07-2010M12. Here, I’ve used the deviation from HP trend of log US industrial production as a proxy for domestic demand conditions, and the Chinese CPI as a proxy for Chinese production costs.

Since the regression I ran is only a first cut, it pays to look for more careful analyses. I found two interesting studies on the subject, both out of the HKMA. Regarding Chinese exports, Cui, Shu, and Chang (2009) conclude:

… a 10-per cent appreciation in the NEER will lead to a 5 per cent fall in export prices denominated in domestic prices, while the other 5 per cent adjustment will take place in the foreign-currency price. On average, most of the adjustments will be on the exporters’ profit margins, which would be compressed as the exporters cannot pass on the full impact of the exchange rate to overseas buyers. A modest proportion of the adjustment (around 0.6 percentage points out of the five-percentage-point adjustment) reflects the decline in marginal cost due to the exchange-rate changes. The estimated coefficient for the marginal cost variable is positive and highly significant.

The conclusion is based on data over the January 2005-March 2009 period, and a pricing to market equation estimated in first differences.

Shu, Su, and Chow (2008) write about imports into mainland China:

With persistent inflationary pressures since the second half of 2007 in Mainland China, renminbi appreciation has been increasingly advocated as a means to curb inflation. The effectiveness of appreciation in controlling inflation depends on the impact of exchange rate movements on import and domestic prices. Our analysis finds fairly large and speedy exchange rate pass-through (ERPT) to import prices, with 50% of an exchange rate change passed onto import prices immediately, and 60% over a year. The degree of ERPT decreases along the price chain from upstream to downstream prices. ERPT for consumer prices — the most downstream prices — is much milder and has substantial lags. A 10% rise in the nominal effective exchange rate will dampen consumer prices by 1.1% within a year with limited pass-through in the first half year, and 2.0% over two years.

These estimates are based on an exchange rate pass through equation, with import prices on the left side, and lagged import prices, the exchange rate, foreign prices, commodity prices, and domestic demand, all in first differences, starting monthly from 2005 for import prices.

Taken together, these results suggest to me that it Chinese exchange rate pass through coefficients are higher than those reported for the US. This conclusion is consistent with my findings of relative high price elsaticities for at least Chinese exports, recounted here.

12 thoughts on “Chinese Exchange Rate Pass-Through

  1. KevinM

    Makes sense.
    Closer you work to the raw material, the more your product depends on raw material costs?

  2. Jian

    Menzie, thanks for citing my EL article and the interesting discussions on the ERPT in China. Just a minor comment. The exchange rate in the chart seems to be dollar per yuan, not yuan per dollar.

  3. ppcm

    Interesting theory “If a country is expected to grow more rapidly in the future, running a current account deficit is an optimal behavior”Is it holding when the assets, prices growth, are predicated on housing,household consumption,everything being supported by consumer credit.
    Econbrowser has covered these topics (Representative Ryan Requests,Richard Clarida’s retrospective on the financial crisis,Evidence on Financial globalization and crisis Global imbalances …)
    Borrowing,leverages are not against nature as long as not defying gravity,and driven towards production, productivity.These charts below are exemplifying the dream of Icarus:
    Gross Private Domestic Investment (GPDI)
    Debt Outstanding Domestic Nonfinancial Sectors – Household, Consumer Credit Sector (HCCSDODNS)
    The same in Europe please see ECB data warehouse.Few are still wondering whether this greater depression can be explained.
    Puzzling are the last minutes of the Fed meeting Release Date: January 26, 2011
    Consistent with its statutory mandate, the Committee seeks to foster maximum employment and price stability.
    Federal Reserve Act 1913
    “The Board of Governors of the Federal Reserve System and the Federal Open Market Committee shall maintain long run growth of the monetary and credit aggregates commensurate with the economy’s long run potential to increase production, so as to promote effectively the goals of maximum employment, stable prices, and moderate long-term interest rates.”
    Same Federal act
    “The Federal Reserve is responsible for conducting the nation’s monetary policy, supervising the banking system, and managing the production and distribution of U.S. currency”

  4. Ricardo

    This article by Hossein Askari and Noureddine Krichene in the Asian Times is an important read to understand how US monetary policy is harming the rest of the world. Too often we only see the issue through a US perspective.
    Because the dollar is the world’s reserve currency, the US is no Greece or Ireland; it faces no limit to fiscal and monetary expansion. For years, it has been running external (current account) deficits without tears, that is, without facing any balance of payments constraint. Such is not the case for Burundi, for example, which cannot expand deficits without running into foreign exchange constraints.
    The dollar depreciation has hurt a number of economies as it essentially exports inflation to these economies; the rates of inflation have exceeded tolerable levels in China, India, and in a number of other countries. US policy makers have little concern for soaring gold, food and energy prices. They are narrowly pre-occupied with domestic political and economic agendas, including coping with a massive unemployment rate of about 9.4%.
    …Consumers in many poor countries are being heavily taxed through the depreciation of the dollar. Creditors, such as China, sovereign wealth funds, World Bank, and other financial institutions, would lose real value of their loans. Currency wars would lead to trade wars and impede trade and growth. As recent events have shown, reserve currency governments will not be dissuaded from excessive fiscal spending, nor will reserve central banks be persuaded to renounce massive liquidity injections and near-zero interest rates.
    Tensions between China and US have escalated with the US wanting China to appreciate its currency so that the US can inflate freely, depreciate its own currency, and expand exports. Industrial countries are plagued with mass-unemployment, slow economic growth, and a bankrupt banking system. The European Central Bank has undertaken massive purchase of its bankrupt members’ bonds. There is turmoil all around.
    As in Genoa in 1922 and Bretton-Woods in 1944, a reform of the international would require a conference under the United Nations that would evaluate alternative plans and establish a new monetary system that establishes symmetry, stability, and independence from the monetary and fiscal policies of one or two member countries.
    It has to be a system where neither Hu nor Sarkozy would have to beg Obama to restrain dollar liquidity; it should be a system where the US should face the same balance of payments constraints as Burundi, instead of Burundi transferring real resources to the US.

  5. endorendil

    How is the “log price of Chinese imports” calculated? If it is an average over a representative basket that’s stable over the period over the graph, I can understand the piece. But if it is just an average price over all imports, it could just as well indicate that the rise in the yuan removed some higher-priced imports from the market, leaving a qualitatively different import basket in its wake. The devil is usually in the details…

  6. Menzie Chinn

    Jian: Thanks! My mistake — now fixed.

    endorendil: The construction of BLS import and export price indices is described here (essentially modified Laspeyres). They are not simple unit value indices.

    In general, descriptions of BEA, BLS and other US government statistics can be found with one Google search.

  7. don

    “The effectiveness of appreciation in controlling inflation depends on the impact of exchange rate movements on import and domestic prices.”
    This seems to me to be a very shallow analysis. Do the authors assume the yuan appreciation comes about with no reduction in CCB purchases of foreign exchange (and so no effect on China’s monetary base? Or do they assume that the CCB completely and successfully sterilizes any effects of these purchases on China’s monetary base? Or do they simply ignore this issue?

  8. don

    “This is a partial equilibrium exercise, as recounted in textbooks; hence, it invokes ceteris paribus.”
    So, doesn’t this mean we shouldn’t use the result as an indication of what night actually happen if China stopped intervening in currency markets and allowed the yuan to appreciate?

  9. endorendil

    Hi Menzie,
    Thank you very much for the pointer. I quote from that resource:
    “Answer: The International Price Program (IPP) selects sample establishments based upon their relative trade value in imports and exports during the course of a year. After an establishment is selected for inclusion, a BLS field economist visits the establishment to enlist cooperation and to select the exact items that will be priced on a monthly basis. All information provided by the establishment is protected under BLS confidentiality rules.”
    That means that the average price is based on what is actually traded during a year. Within this year, price decreases reflect actual decreases in the price of a particular product. From year to year, however, price changes also reflect changes in the kind of goods procured. This means that drawing conclusions is not as easy as just saying that “prices kept going down after the revaluation, so the Chinese economy just got more efficient”. There is also the uncorrected influence of a changing basket of goods.
    To be sure, I think it is likely that the Chinese economy is getting more efficient, but I’m not discounting the idea that re-valuatations of the yuan change the type of products exported to the US. At least not based on these data.

  10. Menzie Chinn

    endorendil: You can see a much more detailed discussion of the BLS import price indices here. Of course, if you think about it, your characterization of what’s going on applies to many, many price indices, including those in the NIPA, and with greater force, since they are chain weighted. Even the CPI with fixed weights for broad categories has at the sub-category level chain-weighted type indices.

  11. endorendil

    hi Menzie,
    Thanks for another interesting article. You’re entirely correct this is a common problem with price indices (and, frankly, almost anything related to macro-economic product mixes as they are invariably variable).
    Let me may apply the model used in the article. Menu costs explain why adjustment to price changes is slow, i.e. why you see that bulge in prices after the yuan appreciation. The initial response is for companies to eat increased costs by reducing margin. The estimate that only a third of a small cost increase gets through to the customer explains why the price bulge is about a third of the yuan rise – in line with normal price elasticity. After the initial shock is absorbed, further recalibration is by menu changes – the product mix changes. I don’t think this is necessarily attributable to price elasticity.
    So I think that the graph shows that the initial price shock leads to the usual response – margins decrease, after which product mixes change. To me, the most interesting thing is what happens in the year after the price bulge: nothing. In the previous years, import prices dropped precipitously each year in an almost linear fashion (on this log scale), reflecting very large downward pressure on import prices. Having the import prices go flat is a stunning change, and it could mean that the Chinese economy no longer can increase its productivity. It could also mean that after the product mix was adjusted to remove overpriced products, they started to be put back in as prices in the US started to rise or as productivity gains in China started to make them viable for export to the US again. It could also mean that Chinese exporters no longer see themselves as bottom feeders, and are starting to increase margins when productivity increases instead of dropping prices. I assume all of these things happen at the same time, and it would be useful to know their relative importance. I can’t determine that, though, at least not from this graph. Has anyone tried to pry these causes apart for this period in time (Q2 2009 – Q3 2010)?

Comments are closed.