Evaluation of Exchange Rate Point and Density Forecasts: an application to Brazil

In this paper, we construct multi-step-ahead point and density forecasts of the exchange rate, from statistical or economic-driven approaches, using financial or macroeconomic data and using parametric or nonparametric distributions. We employ a set of statistical tools, from different strands of the literature, to identify which models work in practice, in terms of forecast accuracy across different data frequencies and forecasting horizons. We propose a novel full-density/local analysis approach to collect the many test results, and deploy a simple risk based decision rule to rank models. An empirical exercise with Brazilian daily and monthly data reveals that macro fundamentals matter when modeling the risk of exchange rate appreciation, whereas models using survey information or financial data are the best way to account for the depreciation risk. These findings have relevance for econometricians, risk managers or policymakers interested in evaluating the accuracy of competing exchange rate models.