China exchange rate regime.
The current China exchange rate system can be said to be an ongoing work. In the year 2005, July after a whole year of discussion the government it took to revaluing the China RMB by 2.1, changing to a basket from the dollar peg, giving the RMB chance to float more freely. It expanded the forward market in august by permitting the foreign and local banks with licenses to to take part in foreign exchange market interbank to transact RMB with interbank market and clients, the banks had the power to independently determine the forward rates. By September it did away with inconsistency existing between the rules for dollar fluctuations and basket fluctuations. In the new found Chinese exchange rate regime in July, daily dollar fluctuations were to be restricted to 0.3 per day, while the basket daily fluctuations restricted to 1.5 per day. The major problem in this system was that if the dollar had advantage and moved 2 againstother currencies, event that is not precedented, it may be almost impossible following the rules.1 Due to this, in 23rd September PBOC (Peoples Bank of China) changed RMBs fluctuation by a 3 per day rate against the major currencies non-dollar, Yen and Euro currencies. On 25th November, the regime system regulators chose to continue with market makers system to trade RMB for other currencies. Parallel to this, PBOC improved banks ability on foreign currency hedging through the first ever conducted currency swap domestically, it sold 6 billion to about ten domestic banks. In January 2006, PBOC allowed over counter foreign exchange transactions, running concurrently was the centralized APMC (Automatic Price Matching System), this was basically intended to establish more market oriented price finding .
To assess china s sensitivity in imports and exports to changes in renminbi real exchange rate, I estimate the standard export and import equations. We are interested in the long run equation so we will use the co integration techniques. On top of that I use form import export equations to eliminate bias in equations. Anyway to eliminate omitted variables problems we include the determinants of demand.2
and supply in the reduced equation form. The estimation equations are
Xt o 1 REERt Yt n i3, i controlst t
Xt o 1 REERt Yt n i3, i controlst t
Where Xt is China export volume
Mt is China import volume
REERt is real effective exchange rate of the reniminb
Yt is foreign demand
YtWWb is china s domestic demand.2
The parameters estimated are 1 export exchange elasticity, 2 income export elasticity, 1 imports exchange rate elasticity, 2 import income elasticity.
The Chinese processing sector is very important for the economy we therefore estimate the ordinary exports and the processing sectors separately, the same is applied in ordinary imports and imports for processing. The Chinese data is difficult to work with evidently because no export and import price indexes are available at the aggregate levels. Hence the use of proxy price data. We use China s CPI (Consumer Price Index) as export prices proxy basically because China s National Bureau of Statistics doesn t give producer price index prices. Import prices proxy, we calculate China s weighted index 25 most significant trade partner s export prices and deflate imports index of China. As test of robustness, Hong Kong SAR was used as China s proxy prices of export and results maintained.
Where N is the currencies involved in the index, which is the with currency weight and reri, is the bilateral real exchange rate going against every China s trade partners. We use REER, BIS constructed as the robustness test there was a change in the results.3
The exports exchange rate elasticity is expected to be negative, as the products from China come to the global market. In Chinese case, the imports exchange elasticity is not clear as expected. If gained power of purchasing is stronger than the decreased demand following the related exports fall then imports should be fostered by a real appreciation. The import structure will also affect the reaction. Price elasticity are expected to be positive if domestic production are mainly substituted by imports.4 However, if the imports generally goods investments taken to the export sector, which is the common factor in China, then appreciation will negatively affect them.
Foreign Chinese exports demand is measured through global imports (don t include imports to China) and deflated through the world price index of imports. It s obvious that some production based tactics measures were used, though the data for every month don t exist. For China s local demand for these ordinary imports, industrial production volume is used. GDP is always used to measure the entire nation s output, in the wake of 2005 statistical reform in China, it is about to produce the GDP statistics for the period 1994-2005.5 And for the imports for process sing, processed exports are use as the demand factor for the long run purposes. We expect positive income elasticity for both imports and exports.
Additional controls are therefore placed in the import and export equations on their relevance basis in the literature of trade and in the Chinese case. Relevance of rebates of VAT (Value Added Tax) are tested for the case of exports used in China as policies to either discourage or encourage exports depending on the cycle of the business. A positive VAT sign rebates is expected. Capacity utilization measure is used to introduce considerations supply in the reduced form equation. The a priori should be that high utilization capacity should indicate towards the significant supply constraints, which might hinder growth of export. Capacity utilization can therefore be defined in brief as the difference between industrial production the trend in it, Hodrick-Prescott filter is used to calculate the latter.6
Finally, inward FDI (Foreign Direct Investment) real stock is the last control variable that is in the export equation. While the export and trade relations are well drawn in literature, it could be very relevant particularly to China now that large capacity of FDI is being directed towards the export industry. Although someone may think that FDI stock increase should foster the exports of China, the production chain complicated structures may bring complications in such an a priori.
Let s now move to the import equation, we have to include import tariffs because they have reduced substantially, particularly since China joined the WTO. The FDI stock is the second control. We expect, by principle, to get a positive coefficient on the stock of FDI because the foreign firms are more likely to prefer using imported machinery, parts and components in their production processes than the Chinese firms.7 However as the companies from the foreign countries start to direct their whole chains of production to China, import needs could be reduced and accompanied with FDI stock increase.
Finally, both the imports and exports equations include a deterministic trend when it is significant statistically. The variable of the trend should be vital in trapping improvements in
production and the reforms that are ongoing in the China economy, which cannot be measured easily anyway.
All the other variables in the model except the import tariffs and VAT rebates measured as value share of imports and exports are all in logarithms. The standard seasonal pattern may not be exactly followed by the Chinese data use of an adjustable series is preferred as we introduce dummies for Chinese December and the New Year.
We choose to use every month data for this period 1994 to 2005. Starting the analysis before the year 1994 would have definitely made little sense because there was China s market reforms breakthrough in the same year. Some of these reforms about China market were relevant to all the questions we were asking, i.e. there was unification in the 2 exchange rate systems, a compulsory imports planning was done away with and a reduction in the quotas and licensing requirements. And lastly reforms on prices were pushed forward, renminbi could be converted on the current account and development of private sector derived some benefits from the company laws initiated.
We now test for integration order of the variables involved in the analysis as a preliminary test. We use ADF (Augmented Dickey-Fuller) tests for unit root existence. Almost all the variables are found to be non-stationary in the levels but are stationary in the initial differences. Co integration vectors existence is then tested using the Johansen procedure. We find vector co integration (at least one) for every variable group.9 Co integration vectors presence gives us the chance to estimate lagged determinants regression and their differences using the approach of non-linear least squares as proposed by Loretan and Phillips. Approaches like these yield consistent and unbiased estimates of both the short run and long run parameters.
Apart from regressions on import and export equations for our full samples shown earlier (1994-2005), we must also run regression like those in the shorter periods let s say from the year 2000 to 2005 concentrating on the WTO influence period. It is important to consider to differentiate between ordinary trade and processed, running equations for each separately in both import and export case. Three was the short term lag maximum number introduced ultimately we only include the statistically significant ones.
As expected the China s exports exchange rate elasticity for the long term both ordinary and processed are negative and significant in both the full sample and the one after WTO entry. When transformed appropriately, the long run impact estimated for real exchange rate is about -1.3 for the exports processed in both periods. And for the ordinary exports it reduces from -2.3 for the whole term to -1.6 for more current period. (These results are similar to results that have been founded for U.S, U.K among other OECD states.
For both the processed and ordinary exports, the positive effect of the long run is very small and is not significant statistically in the full sample, though it does become statistically significant after acquiring the WTO membership. To add to that, for most of the current sample, the Chinese export income elasticity is around 1 as is expected.
On the control variables, capacity utilization has significant impact over the exports only with a delay of one month or contemporaneously. The capacity utilization sign is negative in conjunction with the idea that proposing that large production shares stays in high growth times in the domestic market. VAT rebates statistically insignificant in all of the specifications we therefore don t include them in the final estimations. The FDI stock data begins in 1997 which is the explanatory variable but only at the more current sub period. A FDI stock surprisingly does not affect the exports of Chinese an extent that is statistically significant. The trend goes on positive for all the equations while the exports are seen to be reducing at the Chinese New year and reduce during December. If the trends are not included in the estimation both the FDI stock and World demand coefficients would definitely become positive and statistically significant. However the exchange rate elasticity results would remain unchanged.
Finally, the imports exchange rate elasticity always come out negative and significant with an exception of imports for processing usually for the latter sub period, where the exchange rate negative coefficient is significant at only the level
Categories:
0 comments:
Post a Comment