تعیین کنترل سرمایه: کدام نقش را رژیم نرخ ارز بازی می کند ؟

This paper investigates the role of exchange rate regime choices in the determination of capital controls in transition economies. We first use a simultaneous equations model to allow direct interactions between decisions on capital controls and on exchange rate regimes. We find that exchange rate regime choices strongly influence the imposition or removal of capital controls, but the feed-back effect is weak. We further estimate a single equation model for capital controls with exchange rate regime choices as independent variables, and we find that there is a hump-shaped relationship between exchange rate regime flexibility and capital control intensity.

The turbulence in the international financial markets and the reoccurrence of currency crises in the 1990s have once again sparked a debate over the merits of increased international capital mobility. The fact that international capital movements played an important role in recent financial and currency crises has lead many academic researchers and policy advisors to reassess the implications of capital mobility, especially for the viability of various exchange rate regimes. Some authors argue for a “bi-polar” solution for the choice of exchange rate regimes and recommend adjustment of exchange arrangements to the new environment of heightened capital mobility (Eichengreen, 1994; Fischer, 2001). Other authors warn against the excessive volatility in the financial markets associated with free capital movements and advocate imposing capital controls to limit capital mobility. With capital movements under check, intermediate exchange arrangements such as conventional pegs, crawling pegs or bands, and target zones remain a viable and attractive option for many countries (Wyplosz, 2001; Williamson, 2000). These two strands of research both rely on the proposition that exchange rate stability and monetary policy autonomy are not jointly achievable under free capital mobility, the so-called impossible trinity, but they look at it from different angles. The “bi-polar” view emphasizes the trade-off between exchange rate stability and monetary autonomy given free capital mobility and questions the viability of intermediate exchange rate regimes without the support of capital controls. An implication of this view is a possible hump-shaped relationship between the flexibility of exchange rate regimes and the intensity of capital controls. While fixed and flexible regimes can live with high capital mobility, intermediate regimes are expected to be associated with higher intensity of capital controls. The views favoring capital controls challenge the desirability of unrestricted international capital movements in the first place. They argue that restrictions on capital mobility may be required to achieve second-best solutions in the face of capital market distortions, and that the choice of the exchange rate regime should take existing restrictions into account. The empirical literature has so far neglected the interdependence between the choice of exchange rate regimes and restrictions on capital mobility as well as the possibility of a non-monotonic relationship between the two. In this paper we attempt to fill this blank in an analysis of the determination of capital controls in transition economies during the 1990s.1 In order to integrate the analysis of the two topics, we construct a simultaneous equations model to describe the joint determination of capital controls and the choices of exchange rate regimes. After establishing the recursive structure of that model empirically, we estimate a single equation model for capital controls that explicitly allows for a non-monotonic effect of the rigidity of exchange rate regimes on the intensity of capital controls. The rest of the paper is organized as follows. Section 2 provides a brief review of the theoretical underpinnings of capital controls as well as the existing empirical evidence. Section 3 discusses the measurement of capital controls and explains their determinants. The simultaneous equations model for both capital controls and exchange rate regime choices is introduced in Section 4, while the single equation model for capital controls is presented in Section 5. Conclusions are summarized in Section 6.

We have examined the role of exchange rate regime choices in the determination of capital controls in transition economies. We develop a simultaneous equations model to account for the interactions between decisions on exchange rate regimes and on capital controls. While the exchange rate regime choices are discrete-valued variables, we develop a continuous index to measure the intensity of capital controls. The discrete-continuous simultaneous equations model is estimated using the two-stage estimation procedure suggested by Heckman (1978). After finding that exchange rate regime choices are not affected by decisions on capital controls, we develop a single equation model to analyze the non-linear influence from exchange rate regime choices on the intensity of capital controls. The results of the simultaneous equations model show a strong influence from exchange rate regime choices on capital controls, while the feed-back effects from capital controls on exchange rate regime choices are absent. The weak response of exchange rate regime choices to capital account liberalization suggests that governments tend to utilize capital controls to help manage the exchange rate regimes, rather than adjusting the latter passively to accommodate the changing degree of capital mobility. This weak response also implies that the exchange rate regime choices can be used as an exogenous explanatory variable in the single equation model for capital controls. The results of the single equation model provide evidence for a non-monotonic relationship between capital controls intensity and exchange rate regime choices. The overall evidences suggest that intermediate regimes are typically associated with the most intensive capital controls, and hard pegs are associated with the most liberal capital accounts. Both models show that strong central bank independence and current account liberalization are associated with substantially lower intensity of capital controls. In contrast, advances in the development of financial institutions, current account surpluses, and heavy burden of external debt are associated with tighter capital controls in transition economies. Turning to country groups, the EU accession candidates, especially those advanced in this process, maintain much more open capital accounts than the non-accession countries, most of them being member states of the CIS. There is also evidence that the crises-ridden late 1990s witnessed a slight tightening of capital controls in many transition countries.