Skip to main content
Log in

A provincial view of global imbalances: regional capital flows in China

  • Original Paper
  • Published:
Review of World Economics Aims and scope Submit manuscript

Abstract

We model capital flows among Chinese provinces. A version of the present value model of the current account (PVMCA) with non-tradable goods and a savings wedge accounts for around 80% of the variation in inter-provincial capital flows over the 1986–2010 period. The PVMCA also allows us to identify the channels of external adjustment in capital flows at the province-level: variation in intertemporal prices (domestic and international interest rates, the provincial real exchange rate) and intertemporal variation in quantities (output, investment and government spending). We find that cross-province variation in the importance of these channels is correlated with the importance of private and state-owned enterprises and demographic factors. We discuss implications of our results for global imbalances in capital flows.

This is a preview of subscription content, log in via an institution to check access.

Access this article

Price excludes VAT (USA)
Tax calculation will be finalised during checkout.

Instant access to the full article PDF.

Institutional subscriptions

Fig. 1
Fig. 2
Fig. 3
Fig. 4
Fig. 5
Fig. 6

Similar content being viewed by others

Notes

  1. Another justification for introducing a savings friction is that, using the methodology developed in Gourinchas and Jeanne (2013), Cudre (2014) identified them as the key driver of provincial external balances (as opposed to investment wedges).

  2. To see the formal similarity with a savings wedge in the Gourinchas–Jeanne setup, write \((1+r_{t}^{T,k})=(1+i_{t}^{W})(1-\tau _{t}^{k})/(1+\pi _{t}^{k})\), where \(\tau _{t}^{k}\) is a province-specific wedge and \(i_{t}^{W}\) is the nominal world rate of interest. In our setup, we assume \(\tau _{t}^{k}=\delta ^{k}\tau _{t}\) (i.e. the province-level wedge is the product of a province-level degree of financial integration and a China-wide wedge vis-à-vis the rest of the world). Taking logs then gives the representation above. The assumption implicit in this formulation is that time variation in savings wedges is common across provinces, whereas the relative degree of access of provinces to the global capital market is unchanged over time. Since province-level interest rates are not directly observable, this approach allows us to calibrate \(\tau _{t}\) directly from observables while estimating \(\delta ^{k}\) as a parameter of the model.

  3. Kano (2008) obtained an expression for the CA/NO ratio. As no income flows data among regions are available, we use the approximation \(\widetilde{\frac{NX}{NO_{t}}}^{k}=\widetilde{\frac{CA}{NO_{t}}}^{k}-b\widetilde{r}_{t}^{T,k}\), where b is the steady-state value of out-of-province assets.

  4. Even though c and \(\kappa\) will vary by province, for the sake of readability, we will generally drop the index k for these and other province-specific constants whenever it is unambiguous to do so.

  5. This follows Bergin and Sheffrin (2000) and Bouakez and Kano (2008). However, these models do not feature a savings wedge.

  6. Thereafter, we assume \(0<\phi <1\), which is fulfilled for values of risk aversion (\(\gamma\)) higher than one and most empirical values of the consumption ratio (c).

  7. Even though we do not fully spell out the investment side of the model, we note the following: as in all neoclassical models, the household here is assumed to choose consumption by taking the path of investment and output as given. This includes, in particular, the possibility that investment itself could be subject to frictions. For example, if lack of access to external funds forces firms to finance future investment from current retained earnings, saving surpluses will predict future increases in investment which, ceteris paribus, would imply an expected decline in net output. In such a setting, a \(\mathrm {\scriptstyle NXO}\) surplus might predict low net output growth because corporate savings predicts increasing investment and not because households save to cushion future declines in their incomes. Unfortunately, our data are not sufficiently detailed to allow us to distinguish between savings by firms or households at the province-level.

  8. In the same spirit, Hoffmann (2013) and Kano (2008) exploit common variation interest rates to identify global shocks.

  9. The next subsection describes in detail how we estimate the decomposition (4) from province-level data. Here we just note that by including interest rates for the country as a whole in our empirical models, we implicitly account for much of the cross-province correlations in the determinants of intra-national capital flows.

  10. Note that these \(\beta _{x}^{k}\)s are not to be confused with the discount parameter of the utility function (\(\beta\)).

  11. The autonomous regions are Tibet, Xinjiang, Guangxi, Inner Mongolia and Ningxia. The cities of Beijing, Tianjin, Shanghai as well as the region of Chongqing are municipalities. In this paper, we use the term ’province’ as a general qualifier for provinces, autonomous regions and municipalities.

  12. http://chinadataonline.org/. The CDC reports values as soon as they are published in the corresponding yearbook. Although data have sometimes been subject to official revisions in later years, the CDC did not systematically adapt past values.

  13. For example, over 1985–2010, the correlation of the first difference of national net exports with cumulated net exports is 0.80. It rises to 0.87 for 2000–2010, the period in which global imbalances arose.

  14. All VAR systems appear stationary with exception of Shandong where we find one eigenvalue of the companion matrix to be outside of unit circle. For formal stationarity tests, see Table A.1 in ESM appendix.

  15. Since Eq. (6) holds for all \(Z_{t}\) we have the non-linear set of cross-equation restrictions \(\mathbf {e}_{nx}^{\prime }=\left[ -\mathbf {e}_{\Delta no}^{\prime }+\phi \mathbf {e}_{\Delta q}^{\prime }+(1-\phi )(\mathbf {e}_{r}^{\prime }+\delta ^{k}\mathbf {e}_{\tau }^{\prime })\right] \kappa \mathbf {A}\left[ I-\kappa \mathbf {A}\right] ^{-1}\). Based on Monte-Carlo evidence, Bouakez and Kano (2009) show that the standard non-linear Wald test of the PVMCA is biased against the null in small samples and advocate the use of the linear version instead. Multiplying with \(\left[ I-\kappa \mathbf {A}\right]\) and rearranging yields the linear set of restrictions \(\mathbf {R}(\phi ,\delta ^{k},\kappa )=\mathbf {e}_{nx}^{\prime }+\left[ -\mathbf {e}_{nx}^{\prime }+\mathbf {e}_{\Delta no}^{\prime }-\phi \mathbf {e}_{\Delta q}^{\prime }-(1-\phi )(\mathbf {e}_{r}^{\prime }+\delta ^{k}\mathbf {e}_{\tau }^{\prime })\right] \kappa \mathbf {A}=0\). We choose \(\phi\), \(\delta\) and \(\kappa\) to minimize \(\mathcal{W}^{l}=\left[ \mathbf {R}(\phi ,\delta ^{k},\kappa )\right] \left[ \frac{\partial \mathbf {R(\mathbf {A)}}}{\partial \mathbf {A}}\widehat{\Sigma }\frac{\partial \mathbf {R(A)}}{\partial \mathbf {A}}^{\prime }\right] ^{-1}\left[ \mathbf {R}(\phi ,{\delta }^{k},\kappa )\right] ^{\prime }\) where \(\widehat{\Sigma }\) is the covariance matrix of the VAR-parameters. \(\mathcal{W}^{l}\) has a \(\chi ^{2}\)-distribution with mp degrees of freedom, where m is the dimension of the companion matrix and p the number of estimated parameters (\(p=3\) in our case).

  16. The model is estimated on the 1986–2010 period with the exception of five provinces for which we have a shorter sample: Guangxi, Yunnan, Shaanxi, Shanxi and Ningxia. For all provinces, the consumption ratio c is estimated over the sample length using the same deflator as for output and government consumption.

  17. All estimates are obtained using the pooled common-correlated effects (CCEP) estimator. Pesaran (2006) shows that the pooled estimator has slightly better convergence in small samples than the mean group estimator.

  18. Forming groups of provinces based on the median is more robust to outliers in the measurement of the characteristics than would be a direct regression of the province-specific \(\beta _{x}^{k}\) on these characteristics. This matters because both the individual \(\beta _{x}^{k}\) as well as a lot of province-characteristics (for which in many cases we only have a few years of observations) are likely to be noisy. For completeness, we report the province-specific adjustment patterns in Table A.4 in the ESM appendix.

  19. Besides the relative role of state and private enterprise in the local economy, this indicator also takes account of various other aspects such as province-level competition in the banking sector, regional differences in the legal framework, or the extent to which a province is subject to restrictions on labour migration and interregional flows in goods and services. It therefore overlaps to some extent with the indicators on trade and financial openness.

  20. We thank an anonymous referee for suggesting this second mechanism.

References

  • Aguiar, M., & Amador, M. (2011). Growth in the shadow of expropriation. The Quarterly Journal of Economics, 126(2), 651–697.

    Article  Google Scholar 

  • Bergin, P. R., & Sheffrin, S. M. (2000). Interest rates, exchange rates and present value models of the current account. Economic Journal, 110(463), 535–58.

    Article  Google Scholar 

  • Bernanke, B. (2007). Global imbalances: Recent developments and prospects. Bundesbank Lecture speech.

  • Bouakez, H., & Kano, T. (2008). Terms of trade and current account fluctuations: The Harberger–Laursen–Metzler effect revisited. Journal of Macroeconomics, 30(1), 260–281.

    Article  Google Scholar 

  • Bouakez, H., & Kano, T. (2009). Tests of the present-value model of the current account: A note. Applied Economics Letters, 16(12), 1215–1219.

    Article  Google Scholar 

  • Brandt, L., Tombe, T., & Zhu, X. (2012). Factor market distortions across time, space and sectors in China. Review of Economic Dynamics, 16 (1), 39–58.

  • Brandt, L., & Zhu, X. (2010). Accounting for China’s growth. Institute for the Study of Labor (IZA) (Discussion Paper).

  • Caballero, R. J., Farhi, E., & Gourinchas, P.-O. (2008). An equilibrium model of “global imbalances” and low interest rates. The American Economic Review, 98(1), 358–393.

    Article  Google Scholar 

  • Chamon, M., Liu, K., & Prasad, E. (2013). Income uncertainty and household savings in China. Journal of Development Economics, 105, 164–177.

    Article  Google Scholar 

  • Chamon, M. D., & Prasad, E. S. (2010). Why are saving rates of urban households in china rising? American Economic Journal: Macroeconomics, 2, 93–130.

    Google Scholar 

  • Chan, K. (2010). The household registration system and migrant labor in China: Notes on a debate. Population and Development Review, 36(2), 357–364.

    Article  Google Scholar 

  • Chan, K., & Wang, M. (2008). Remapping China’s regional inequalities, 1990–2006: A new assessment of de facto and de jure population data. Eurasian Geography and Economics, 49(1), 21–55.

    Article  Google Scholar 

  • Chan, K. W. (2013). The encyclopedia of global human migration (China, internal migration). Hoboken: Blackwell.

    Google Scholar 

  • Cudré, S. (2014). Regional external imbalances in China: Data and stylized facts. First chapter of PhD thesis "A provincial view of China's external imbalances", University of Zurich.

  • Cudré, S. (2014). Capital’s long march west: Investment and saving frictions in Chinese regions. University of Zurich Department of Economics Working Paper Series, no. 161.

  • Fan, G., Wang, X., & Zhang, L. (2001). Marketization Index for China’s Provinces, China and World Economy, Institute of World Economic and Politics, 9(5), 3–8.

  • Gourinchas, P.-O., & Jeanne, O. (2013). Capital flows to developing countries: The allocation puzzle. The Review of Economic Studies, 80(4), 1484–1515.

    Article  Google Scholar 

  • Gourinchas, P.-O., & Rey, H. (2007). International financial adjustment. Journal of Political Economy, 115(4), 665–703.

    Article  Google Scholar 

  • Hoffmann, M. (2013). What drives China’s current account? Journal of International Money and Finance, 32, 856–883.

    Article  Google Scholar 

  • Holz, C. (2008). China’s 2004 economic census and 2006 benchmark revision of GDP statistics: More questions than answers? China Quarterly, London, 193, 150.

    Google Scholar 

  • Im, K. S., Pesaran, M. H., & Shin, Y. (2003). Testing for unit roots in heterogeneous panels. Journal of Econometrics, 115(1), 53–74.

    Article  Google Scholar 

  • Kano, T. (2008). A structural VAR approach to the intertemporal model of the current account. Journal of International Money and Finance, 27(5), 757–779.

    Article  Google Scholar 

  • Lettau, M., & Ludvigson, S. (2001). Consumption, aggregate wealth, and expected stock returns. The Journal of Finance, 56(3), 815–849.

    Article  Google Scholar 

  • Lijuan, D. (2010). An overview of China’s real estate price index. National Bureau of Statistics of China.

  • Mendoza, E. G., Quadrini, V., & Rios-Rull, J.-V. (2009). Financial integration, financial development, and global imbalances. Journal of Political Economy, 117(3), 371–416.

    Article  Google Scholar 

  • Modigliani, F., & Cao, S. (2004). The Chinese saving puzzle and the life-cycle hypothesis. Journal of Economic Literature, 42(1), 145–170.

    Article  Google Scholar 

  • Obstfeld, M., & Rogoff, K. (1996). Foundations of International Macroeconomics. Cambridge: MIT Press.

    Google Scholar 

  • Pesaran, M. H. (2006). Estimation and inference in large heterogeneous panels with a multifactor error structure. Econometrica, 74(4), 967–1012.

    Article  Google Scholar 

  • Pesaran, M. H. (2007). A simple panel unit root test in the presence of cross-section dependence. Journal of Applied Econometrics, 22(2), 265–312.

    Article  Google Scholar 

  • Sachs, J., Cooper, R., & Fischer, S. (1981). The current account and macroeconomic adjustment in the 1970s. Brookings Papers on Economic Activity, 1981(1), 201–282.

    Article  Google Scholar 

  • Scharping, T. (2001). Hide-and-seek: China’s elusive population data. China Economic Review, 12(4), 323–332.

    Article  Google Scholar 

  • Shen, J. (2006). Estimating urbanization levels in Chinese provinces in 1982–2000. International Statistical Review, 74(1), 89–107.

    Article  Google Scholar 

  • Song, Z., Storesletten, K., & Zilibotti, F. (2011). Growing like China. The American Economic Review, 101(1), 196–233.

    Article  Google Scholar 

  • Wei, S. J. & Zhang, X. (2009). The competitive saving motive: Evidence from rising sex ratios and savings rates in China. (NBER Working Paper).

  • Zhang, Z. (2008). Dark matter in china’s current account. Unpublished manuscript, International Monetary Fund.

Download references

Acknowledgements

The authors gratefully acknowledge support by the Swiss National Science Foundation in the framework of the ProDoc project Capital Flows among Heterogeneous Economies. The authors would also like to thank Paul Bergin (the editor), Cédric Tille, Ulrich Volz, Fabrizio Zilibotti, Aaron Mehrotra and seminar participants at the macro committe meeting of the German Economics Association, the Bundesbank East Asia Workshop, the Hong Kong City University-BOFIT Renminbi Conference, the Mainz Workshop in Trade and Macro, the HKIMR China Conference and at the University of Zurich for comments and suggestions. We are also grateful to two anonymous referees whose comments greatly helped us improve the paper.

Author information

Authors and Affiliations

Authors

Corresponding author

Correspondence to Mathias Hoffmann.

Additional information

This paper was part of Samuel Cudré's Ph.D. Thesis at University of Zurich.

Electronic supplementary material

Below is the link to the electronic supplementary material.

Supplementary material 1 (pdf 137 KB)

About this article

Check for updates. Verify currency and authenticity via CrossMark

Cite this article

Cudré, S., Hoffmann, M. A provincial view of global imbalances: regional capital flows in China. Rev World Econ 153, 573–599 (2017). https://doi.org/10.1007/s10290-017-0278-0

Download citation

  • Published:

  • Issue Date:

  • DOI: https://doi.org/10.1007/s10290-017-0278-0

Keywords

JEL classification

Navigation