Measuring and Hedging
Foreign Exchange Risk
Foreign exchange risk is an important component of international portfolio risk. In this chapter we study the empirical properties of currency risk and develop risk model architecture that includes it.
Section 7.1 describes an approximation method for decomposing the total return on a foreign investment into currency-unrelated return (called local return) and currency-only return. This decomposition can be employed in building a portfolio risk model with three components: a local risk component, a currency risk component, and a component measuring the covariances between local and currency returns. Section 7.2 discusses currency hedging models from both short-horizon and long-horizon investment perspectives. Section 7.3 reviews empirical research on the covariances between currency returns and pervasive factor returns in stock and bond markets. Section 7.4 discusses the relationships between macroeconomic variables and currency returns.
Economic researchers use the term “peso problem” to refer to any situation in which the true relationship between economic variables is difficult to measure accurately because one or more of the variables is subject to large, infrequent jumps. The term has its origins in economic research on currency markets (in particular, the Mexican peso–U.S. dollar foreign exchange market (see Lewis 1995)). Currency markets, particularly those under pegged exchange rates, are the classic case of an economic environment subject to“peso problems,” where reliable inference requires historical samples of unfeasible length. Short-term currency risk management is reasonably well understood, but for long-horizon investors there are many open questions about optimal risk-management policies.
Suppose that our home currency is the U.S. dollar, so we measure risk and return in dollar units. In this case the U.S. dollar is called our numeraire