Essays on exchange rates: Central banks' interventions, effects on gold mining activity, and anticipating market risk

The essays in this work explore exchange rate related topics of different areas of study. We start exploring the notion of central banks' interventions in the exchange market. Particularly, we are interested in testing the well known Fear of Floating hypothesis. Then, we change gears to study the effects of a volatile exchange rate on the extraction of a natural resource. Gold mining activity around the world is our subject of interest. Last, we contribute to the risk analysis literature by adding a new parameter (market trend) into the volatility-weighted historical simulation approach to better anticipate market risk for different groups of assets including exchange rates.;The first essay uses a new approach to identify the exchange policies of central banks around the world. We focus directly on the lagged relationship between changes in international reserves and changes in the exchange rate. The evidence does not support the fear of floating hypothesis. Central banks have accumulated reserves after exchange rate depreciations as well as after appreciations. Rather than constantly intervening in the exchange market, our evidence indicates that central banks appear to be accumulating reserves as insurance against future financial crises.;The second essay poses and tests a dynamic optimal control production/location model of a representative global gold mining company. The firm must decide where, when, and by how much to expand gold production in each country given each country's in-ground gold reserves. First, we assume that both gold prices and exchange rates are known in advance with certainty. The optimal extraction path in this case balances revenues with production costs and the "user cost" of depleting the known reserve, in common currency terms, in each country. The more realistic model formalizes the real-world uncertainty about gold prices and exchange rate risk. In this model, exchange rates follow a stochastic differential process with a direct dependence on time and space. The results indicate that prices of gold and country exchange rates have been significant determinants of the location and intensity of gold mining activity around the world.;The third essay aims at verifying empirically the applicability of adding a new parameter (market trend) into the volatility-weighted historical simulation approach to anticipate market risk (VaR and CVaR) for different groups of assets (exchange rates, financials, metals, and energies). Volatilities are forecasted using EWMA and GARCH models. The results show that adding the market trend parameter can improve VaR estimates, especially for extreme VaRs (99% and 1%). We also observe that overall using a well calibrated EWMA model to forecast volatility provides better VaR estimates for the volatility-weighted historical simulation approach than using the GARCH model. In terms of CVaR estimation the results are ambiguous. There is no prevalence of one model over the other. Also, the contribution of the trend parameter to these models in estimating CVaRs is not evident.