Real Exchange Rates and Foreign Portfolio Investment

Chapter 1 uses real interest rates to show that the long-term real exchange rate and a risk premium are more volatile than the nominal exchange rate for four developing countries. Chapter 2 finds country-level effects of foreign portfolio investment in 18 developing countries and proposes a channel for real exchange rate volatility driven by foreign portfolio investment. Chapter 3 shows that there are plant-level productivity gains in Chile from small levels and small increases in foreign ownership typical of portfolio investment.Nominal exchange rate changes can be decomposed into inflation differentials and real components, each having different causes and consequences. Although we have explanations for each component, the relative importance of the long-run real exchange rate has previously not been quantified. Chapter 1 uses inflation-linked bond data for Brazil, Mexico, South Africa, and Turkey to quantify the contribution of those components to exchange rate changes against the United States. We find that the long-term expected real exchange rate plus its risk premium is even more volatile than the nominal exchange rate.Volatility and unforecastability of real exchange rates is a fundamental puzzle of international economics, and these features are even more pronounced in developing countries. In chapter 2, we present a new fact, that inward portfolio investment in equities predicts the extensive margin of imports in 18 developing countries. With this we build a purely real model in whichforeign investors finance new intermediate inputs and increase productivity in the tradeable goods sector. We express the Balassa-Samuelson determined real exchange rate as a function of foreign investment. We use the established fact that investors in developed countries exhibit positive feedback or operate under portfolio constraints and show how therefore the real exchange ratereacts positively to equity prices. We test and confirm the prediction of the model that equity portfolio investment predicts exchange rates at quarterly frequencies. We also confirm that the extensive margin of imports comoves with the real exchange rate at annual frequencies, and do not reject that equity portfolio investment acts through this channel only.In chapter 3, we measure the effect of foreign ownership on the productivity of Chilean manufacturers between 1998 and 2001. Total factor productivity is measured at the plant level using multiple alternative methods. Both the level of foreign ownership and increases in foreign ownership are significant predictors of productivity. These effects remain when restricting the sample to plants with low foreign ownership and small changes in foreign ownership, showing the importance of foreign portfolio investment on productivity. A non-parametric fit suggests that low levels of foreign ownership are as significant as larger stakes. When controlling for endogeneityin the foreign investment decision by instrumenting with total developed countries’ portfolio outflows the positive effects on productivity remain. Finally we show that foreign ownership predicts that firms source more of their intermediate inputs from abroad