> > > > >

Capital Controls in Emerging Market Economies - A Review of the Chilean and Malaysian Experiences

Matthias Müller

Although financial globalisation is generally seen as the key economic trend of recent times, restrictions on international capital movements remain the rule rather than the exception in international finance. In 1999, well over half of the countries in the world were still deemed by the IMF to have significant controls on capital account transactions. Yet increased capital mobility has created a highly unstable international financial system. Large reversals of short-term capital flows were at the heart of every major financial crisis during the 1990s -- Mexico in 1994, East Asia in 1997-98, Russia in 1998, and Brazil in 1999.

With Article VI, Section 3 of the IMF Articles of Agreement permitting the use of capital controls, the idea of restricting capital mobility as a means of reducing macroeconomic instability is not new in modern policy discussions. In the 1970s, James Tobin argued that a tax on foreign exchange transactions would reduce destabilising speculation in international financial markets. While the theoretical rationale for sprinkling "sand in the wheels of international finance" is largely uncontested, the prospect of asset and geographical substitution render Tobin-taxes technically and politically unfeasible. Discussions in the wider context of the debate about the new financial architecture have tended to shy away from grand and universal schemes, focussing instead on the merits of more modest proposals aimed at unilaterally restricting capital mobility in emerging market economies. In particular, two type of controls on short-term cross-border capital movements have been considered: controls on capital inflows, similar to those implemented in Chile between 1991 and 1998, and controls on capital outflows of the type Malaysia imposed in mid-1998.

Capital mobility and financial crises

In economic theory, the fundamental case for capital mobility is undisputed. The efficiency gains deriving from capital account liberalisation include static resource allocation benefits through the global intermediation of resources from high-saving to low-saving countries and the international specialisation in the production of financial services, static financial gains through international portfolio diversification, dynamic (or X-efficiency) gains through the introduction of competition in the financial sector, and gains from inter-temporal trade, i.e. the trading of goods today for goods in the future. However, against the benefit of improved efficiency must be set equally clear the greater susceptibility to sharp reversals in the direction of short-term capital inflows, leading to currency and liquidity crises.

Currency Crises

Most economists think about currency crises using one of two standard models of speculative attack on fixed or heavily managed exchange rates. First-generation models explain currency crises as the result of a fundamental inconsistency between domestic policies, i.e. the government's need for seignorage to cover its budget deficit, and the attempt to maintain a fixed exchange rate. This inconsistency can be temporarily papered over if the central bank has sufficient foreign exchange reserves, but the eventual collapse of the exchange rate peg is inevitable. If a devaluation is in prospect, all holders of domestic currency assets will try to switch into foreign currency assets before foreign exchange reserves were exhausted in the natural course of events, and in doing so will advance exhaustion. The result is that if reserves fall to some critical level, there is an abrupt speculative attack that quickly drives reserves to zero and forces the abandonment of the fixed exchange rate.

Second-generation models view currency crises as shifts between multiple equilibria in response to self-fulfilling speculative attacks. The speculative attack, driven by expectations of devaluation, is itself the main proximate reason for devaluation. In the absence of an attack, monetary policies remain unchanged and the exchange rate peg is maintained forever -- the "good" equilibrium. An attack occurs because market participants rationally anticipate that, if (and only if) attacked, monetary policy will move in a more expansionary direction causing the exchange rate to depreciate, thus shifting the economy to a "bad" equilibrium. If most agents believe that the others will act likewise, it will be in the interest of everyone to participate in a run on the currency to avoid the loss they will suffer if they keep a depreciating currency. A related, but separate explanation for the onset of a currency attack is herding when information in foreign exchange markets is incomplete or asymmetrical. A wave of selling, whatever its initial cause, could be magnified through sheer imitation and turn into a currency crisis.

Liquidity Crises

By contrast to the well-established first- and second- generation models of currency crises, models of liquidity crises are still in their infancy. This line of work relies on a version of the Dybvig-Diamond model of bank runs. Banks provide liquidity transformation services by pooling short-term deposits and relending the deposited funds to long-term investors. If most depositors maintain confidence in the liquidity of the bank, they will maintain their deposits in the bank, and the bank will be able to meet the short-term random demands for high-powered money that arise at any point from a small proportion of bank depositors. However, if each depositor suspects that the other depositors want to withdraw their funds in the near future, the individual depositor will expect that the bank will soon be facing a liquidity crisis. It then makes sense for the individual depositor to race to the bank to withdraw her deposits ahead of the rest, thereby salvaging the value of the deposits. The result is a run on the bank in which each depositor rationally races to withdraw ahead of the rest. As a consequence, even a solvent and healthy bank can be forced to liquidate its assets prematurely. By analogy, a country may be subject to self-fulfilling shifts in creditor confidence if its financial system is internationally illiquid, i.e. if its potential short-term obligations in foreign currency exceed the amount for foreign currency it can have access to on short notice. Whereas in a domestic context the central bank can protect against bank panics by acting as a lender of last resort, the same is not true in an international setting where the short-term debt is denominated in foreign currency.

Fortsetzung: Capital Controls in Emerging Market Economies - A Review of the Chilean and Malaysian Experiences