Lessons from Latin American Capital Flows
Capital controls, as used in Chile and Colombia, have simply changed the composition of capital inflows, and had little effect on overall flows.
The recent history of capital flows to Latin America is a tumultuous one: first there was the flood of dollars from newly rich Middle Eastern oil exporters, recycled as loans from U.S. and European banks, that washed over the region starting in the mid-1970s. Then came the Mexican debt default of 1982, which set off an economically devastating outflow of capital from the region that lasted for most of the 1980s. In the early 1990s the capital flows resumed, with portfolio and direct investment this time playing a bigger role than bank loans. The Mexican peso crisis of 1994 slowed this inflow, but unlike the debt crisis of the 1980s it hasn't stopped it.
These roller-coaster flows of capital have posed huge challenges for Latin American governments, and the effectiveness of those governments' responses provide lessons for other emerging-market nations, contends NBER Research Associate Sebastian Edwards. In Capital Inflows Into Latin America: A Stop-Go Story? (NBER Working Paper No. 6441), Edwards reviews the history of the region's capital flow experiences and concludes that there are signs that both Latin American governments and outside investors are getting better at keeping capital mobility from wreaking havoc.
"During the last few years the Latin American countries have been a laboratory of sorts, where almost every possible approach towards capital mobility has been tried," Edwards writes. The danger of one approach, letting capital flow as it will, is that capital flows into a country exert upward pressure on the country's currency. If economic growth keeps up with that pressure, this isn't a problem. If it does not, then the country's currency becomes overvalued and its economy loses competitiveness. That means that at some point, capital flows must reverse and the exchange rate must fall, which seldom happens smoothly.
One response to this volatility has been the use of capital controls, such as minimum stay requirements for foreign direct investment and reserve requirements on loans. But Edwards argues that such capital controls, as used in Chile and Colombia, have simply changed the composition of capital inflows, and had little effect on overall flows.
Another strategy is sterilized intervention, by which a country's central bank buys foreign exchange (often issuing securities to pay for it) in an attempt to drive down the price of its own currency. This tactic can be really expensive, since interest earnings on international reserves are low, while a Latin American central bank generally has to pay high interest rates to get investors to buy its securities. Also, the high interest rates on the securities issued by the central bank can bring even higher capital inflows, thus defeating the entire purpose of the intervention.
The government of Chile, however, has had some success with a nominal exchange rate band. That is, the government commits to keeping the Chilean peso exchange rate from rising too high or falling too low against a basket of three foreign currencies. The band is not fixed; its rate of "crawl" is determined by inflation rates in Chile and abroad. And while the band has not entirely prevented real appreciation of the Chilean peso as capital has poured into the country in recent years, it has maintained it at a controlled level.
Probably the clearest way in which developing countries can prevent changes in capital flows from leading to financial panics is by strengthening their banking systems. The foreign investors putting money into Latin America appear to have become more sophisticated over time, coming to understand that there are significant differences among regions and countries. But the weak financial situation and inadequate government supervision of commercial banks in many Latin American countries mean that those countries are still extremely vulnerable to financial crises when capital inflows reverse.
-- Justin Fox