This presentation is part of: C10-1 Econometric and Statistical Studies

Cointegrating-VAR Monetary Models for the GBP/USD Exchange Rate

Viet Hoang Nguyen, M.A, Leeds University Business School (LUBS) - Economics Division & Centre for Advanced Studies in Finance (CASIF), The University of Leeds, Maurice Keyworth Building, Leeds, LS2 9JT, United Kingdom

The macro-modeling of exchange rates has been one of the most challenging research areas in economics over the last thirty years with the prevailing failure of macro-fundamentals to forecast exchange rates. Under the cointegrating VAR framework with exogenous I(1) variables initiated by Pesaran et al. (2000), Pesaran and Shin (2002), we examine popular monetary models of GBP/USD bilateral exchange rate, most notably the sticky-price model which incorporates current account balances (Hooper-Morton). The GBP/USD bilateral exchange rate will be analysed from the UK’s perspective within the system of core macroeconomic variables of the UK and US’s economies in the period from 1980Q1 to 2006Q4. In particular, we explicitly allow for the presence of structural break in the model, considering the Black-Wednesday event as a change in monetary policy regime of the UK. This econometric framework allows us to: firstly analyse the theory-predicted long-term relationship between exchange rate and macro-fundamentals, secondly examine the impulse response functions with respect to shocks of interest such as oil price and domestic monetary policy shocks, and finally provide a broad picture of exchange rate forecasting based on several in- and out-of-sample forecasting exercises using parametric simulation. Our empirical findings show that: firstly, there exists a theory-predicted long-run relationship between exchange rate and economic fundamentals, but both exchange rate and its long-run cointegrating relation with economic fundamentals are very sensitive to shocks, which could be an explanation for the difficulty of forecasting exchange rate; secondly, macroeconomic fundamentals can explain around 50% of quarterly exchange rate variation; thirdly, through the impulse response analysis, most of variables response to the shocks of interest in a theory-consistent fashion; fourthly, among the examined forecasting exercises for exchange rate, the in-sample directional change forecasts bring about promising results.