Business Cycle Transmission from BRICS to Developing Countries, Some New Evidence

Abstract

Our motivation of running this research is the increasing importance of BRICS as a dynamic and emerging power with a growing role in global affairs. The topic of this research in recent years is extremely important given the progress of the international economy and the growing role of the BRICS into the world economical scene, particularly for the less developed countries. The BRICS members are all developing or newly industrialized countries but are distinguished (at least the original four) by their large, fast-growing economies and, more recently, by their significant influence on regional and global affairs.

This paper examines the role of emerging economies of Brazil, China, India, and Russia, as the new regional economic drivers for the less developed countries. To our knowledge, the case of the role of BRICS as dynamic emerging economies has not been entirely explored, so the target of our research is to contribute to this field.

We employ a Global Vector Autoregressive (GVAR) model to investigate the extent of business cycle transmission from BRICS to LDCs. Our research follows Samake and Yang (2011) work with a different sample of countries and different time span. Our sample includes Brazil, Russia, China, India, 10 EU countries, the USA, and 49 emerging and developing economies from Asia, Africa, Latin America, and Commonwealth of Independent States, covering the period 2000–2014.

Summarizing the results, we can notice that the BRICS doesn’t seem to play any significant role as economic leaders for the less income countries of that region, neither do they seem to have strong links among them.

The low bilateral trade weights between the core economies and their counterparts in Asia and Africa, LA (Latin America), and CIS (Commonwealth of Independent States) could be a possible explanation for the insignificant impact of a positive RGDP shock in the core economies. That is, the transmission via trade is trivial. The trade links are not strong enough to trigger transmission shock to the developing countries of Asia and Africa, Latin America, and CIS.

Appendices

Appendices

38.1.1 Appendix A: Data Description

The quarterly data set used for estimation in this paper cover the period 2000Q1–2014Q4.

The main data sources are CEIC, Direction of Trade-IMF, and the United Nations COMTRADE database. Whenever seasonally unadjusted data are collected from the source, the X-12-ARIMA seasonal adjustment in EViews package was used.

• RGDP: Quarterly GDP volume series were taken from CEIC database. For seasonal adjustment, we used Census X12 method.

• PPP-RGDP: Annual series were taken from IMF-WEO 2015.

• Import-export: Quarterly series, in US$, were taken from IFS database.

• Bilateral trade flows: Annual series were taken from COMTRADE database.

• Consumer Price Index: Quarterly series, CEIC database, IMF data.

• Exchange rate: Quarterly series were taken from BIS IMF database.

Definition of variables included in the GVAR model:

• 1. y _it = Real output: Log quarterly data seasonally adjusted, deflated by Consumer Price Index, CPI, in terms of US$ (y _it = ln(GDP _it /CPI _it ). For deflating of GDP, we used CPI instead of GDP deflator, because GDP deflator series were not available for all countries and the sample had to be shrink which will cause estimation problems in our model.

• 2. Dp _it = Inflation Rate: DP _it = % change in CPI, {DP _it =ln (CPI _it ) – ln (CPI_it-1)}.

• 3. RER _it = Real Exchange Rate: log (ER/CPI _t ), where ER is the nominal exchange rate against US$.

• 4. Imports: imp _it = Log quarterly series, US$.

• 5. Exports: exp _it = Log quarterly series, US$.

• 6. P_oil =Log of spot price for crude oil (Brent) in US$ per barrel.

• 7. P_com = Log of international nonenergy commodity prices.

38.1.1.1 Weight Matrices

We use two types of weight matrices in our analysis:

• 1. Trade weight matrix captures the importance of country j for county i in terms of trade dependence (it is used for the structure of county-specific foreign variables). We use bilateral annual trade data between all countries of the sample (annual basis). Baxter and Kouparitsas (2004) in studying the determinants of business cycle comovements conclude that bilateral trade is the most important source of intercountry business cycle linkages. Imbs (2004) also provides further evidence on the effect of trade on business cycle synchronization. He concludes that while specialization patterns have a sizeable effect on business cycles, trade continues to play an important role in this process. He also notes that economic regions with strong financial links are significantly more synchronized. Focusing on global linkages in financial markets, Forbes and Chinn (2004) also show that direct trade appears to be one of the most important determinants of cross-country linkages. The trade shares capture the importance of country i for country j in terms of trade. So they show the degree of trade integration between countries. The country-specific models are connected then, because the foreign variables are entering the equations.

The formula of bilateral trade weights is:

• Wij= exports from country i to j + import from country j to county i/total trade of country i.

• Note: Trade weights are computed as shares of exports and imports displayed in rows by region such that a row, but not a column, sums to one, the complete trade matrix used in the GVAR model. Because of the big size of the table, we couldn’t present it here. The table is available upon request. Source of the data: UN COMTRADE annual data.

• In the table below, we report the bilateral trade weights we used in the regional analysis.

PPP-GDP weight matrix captures a country’s output share contribution to its own regional GDP (used for construction of regional variables). The original data come from IMF World of Economic Outlook 2015 PPP-GDP series measured in billions of current international dollar.

In our model, we also use regional variables since we evaluate the interdependence of the regions and the shock transmissions between them. The regional variables are constructed as weighted averages of the individual countries’ endogenous variables. The weight here is the PPP-GDP shares, that is, the sum of average of each country’s GDP to its regional GDP. These weights capture each country’s contribution to its regional GDP.

(because of the big size of the table, we couldn’t present it here. The table is available upon request)

38.1.2 Appendix B

Table 38.10 BRICS_bilateral_trade_weights

38.1.3 Appendix C: GVARX* Estimation Tests

Table 38.11 WS unit root tests for the domestic variables at the 5% significance level (Based on AIC order selection criterion)

Table 38.12 WS unit root tests unit root tests for the foreign variables at the 5% significance level (Based on AIC order selection criterion)

Table 38.13 WS unit root tests for the global variables at the 5% significance level (Based on AIC order selection criterion)

Table 38.14 VARX* orders and cointegrating relationship in the country-specific models for the major trading blocs

Table 38.15 Test for weak exogeneity at the 5% significance level

Table 38.16 F-statistics for the serial correlation test of the VECMX* Residuals

Table 38.17 Contemporaneous effects of foreign variables on domestic counterparts (t-ratios were computed using White’s heteroskedasticity consistent variance estimator)

Chart 38.6

Persistence profiles

38.1.4 Appendix D: Dynamic Analysis Results-Generalized Variance Decomposition Results

Order of countries in graphs (Africa, Asia, Brazil, China, Cominstates, EU, India, Latin America, Russian Federation, the USA, BRICS)

The bars represent the bootstrapped mean values of the GIRF across the sample, while the 90% bootstrapped confidence intervals are represented by the thinner lines.

The bars represent the bootstrapped mean values of the GIRF across the sample, while the 90% bootstrapped confidence intervals are represented by the thinner lines.

Graph 38.3

One s.e. positive shock to Brazil’s RGDP – generalized impulse response functions for other regions’ exports

Graph 38.4

One s.e. positive shock to Brazil’s RGDP – generalized impulse response functions for other regions’ RGDP

Graph 38.5

One s.e. positive shock to China’s RGDP – generalized impulse response functions for other regions’ exports

Graph 38.6

One s.e. positive shock to China’s RGDP – generalized impulse response functions for other regions’ RGDP

Graph 38.7

One s.e. positive shock to India’s RGDP – generalized impulse response functions for other regions’ exports

Graph 38.8

One s.e. positive shock to India’s RGDP – generalized impulse response functions for other regions’ RGDP

Graph 38.9

One s.e. positive shock to Russia’s RGDP – generalized impulse response functions for other regions’ exports

Graph 38.10

One s.e. positive shock to Russia’s RGDP – generalized impulse response functions for other regions’ RGDP

Graph 38.11

One s.e. positive shock to the USA’s output – generalized impulse response functions for other regions’ exports

Graph 38.12

One s.e. positive shock to the USA’s output – generalized impulse response functions for other regions’ RGDP

Graph 38.13

One s.e. positive shock to BRIC’s RGDP – generalized impulse response functions for other regions’ exports

Graph 38.14

One s.e. positive shock to BRIC’s RGDP – generalized impulse response functions for other regions’ RGDP

38.1.4.1 Dynamic Analysis-Generalized Variance Decomposition Results

Table 38.18 GFEVDs of Africa, explained by the RGDP shocks from the selected economies

Table 38.19 GFEVDs of Asia, explained by the RGDP shocks from the selected economies

Table 38.20 GFEVDs of Latin America, explained by the RGDP shocks from the selected economies

Table 38.21 GFEVDs of CIS, explained by the RGDP shocks from the selected economies

Table 38.22 GFEVDs of BRICS explained by the RGDP shocks from the selected economies