Final Report Summary - FOREX & FUNDAMENTALS (Exchange rates and fundamentals)
This project built upon the seminal work of Engel and West (2005, Journal of Political Economy 113, 485-517) and in particular on the relationship between FX rates (FXRs) and fundamentals (FMs). A new line of attack was taken on the question of co-movement between FXRs and FMs as well as among FXRs. A novel multivariate filtering approach was implemented, whilst linear and non-linear non-parametric tests were applied to detect the dynamic causal relationships. The methodology incorporated time-varying parameters, structural breaks, regime switches and other advanced econometric features and was extensively tested and validated on simulated and empirical data. First, the results of the extensive simulations provided some counterbalance to the suitability - especially in the short run - of rational expectations present-value models of currency rates. It might be that FXRs and FMs are linked in a way that is broadly consistent with asset pricing models of the FXRs, but no evidence was found of a prevailing direction in the examined causalities. Second, an empirical exercise was conducted with data covering the Great Moderation period, the introduction of the euro, the rise and fall of the 'dot-com' bubble, the financial crisis of 2007 - 2010 and the Eurozone sovereign debt crisis. The country-specific FMs were the money supply, industrial production, consumer price index and the three-month rate. The project explored different forecasting horizons and frequencies among the 6 most widely traded FXRs (FX majors), their country-specific FMs and their United-States (US) differentials. The empirical result was not uniformly strong and overall the evidence was inconclusive that there exists a global direction in the examined causalities. Additionally, linear and non-linear links differed significantly and multivariate second-moment filtering purged most of the non-linear interdependencies. However, linkages in some cases emerged and most importantly persisted even after proper filtering with various models. This indicated that if non-linear effects were accounted for, neither FX market lead or lagged the other consistently and returns may exhibit statistically significant higher-order moments and asymmetries. In accordance with the FXRs literature, there was not much evidence that the FXRs are explained only by the observable FMs. Finally, as an extension to the initial work plan, causal interrelations of the FXRs return and volatility series in different scales were also explored with the use of advanced spectral techniques. In contrast to simple disaggregation, a multiscale spectral analysis was used. This project attempted to probe into the micro-foundations of cross-scale heterogeneity in the variability pattern, on the basis of long- and short-term trader behaviour with different time horizons and information flow across time scales. It was shown that markets 'cool off' after a shock at higher frequencies in a much shorter period than after a significant structural change. Overall, FMs may be important determinants of FXRs, however there might be some other unobservable variables driving the currency rates that current asset-pricing models have not yet captured. In general, these results, apart from offering a much better understanding of the dynamic linear and non-linear heterogeneous relationships underlying the major currency markets and their relationships with fundamental macro-variables, may have important implications for market efficiency. They may be useful in future research to quantify the process of financial integration or may influence the greater predictability of these markets. Overall, the knowledge of the nature of interdependency between the currency markets and the degree of their integration will expand the information set available to practitioners and policymakers for decision making.