The out-of-sample success of term structure models as exchange rate predictors: a step beyond

Central Bank Intervention and Heterogeneous Exchange Rate Expectations: Evidence from the Daily DEM/US-Dollar Exchange Rate

“…As stated in Clarida et al (2001), the Markov regimeswitching model suggested by Engel and Hamilton (1990) and developed further by, among others, Engel (1994), Vigfusson (1997), and Dewachter (2001) is a natural candidate to characterize exchange rate behavior. In our model, the conditional mean µ t and the conditional variance h t of exchange rate changes e t are allowed to follow two different regimes-a chartist and a fundamentalist regime-indicated by an unobservable state variable S t .…”

Markov-switching regimes and the monetary model of exchange rate determination: Evidence from the Central and Eastern European markets

“…Frydman and Goldberg (2001) have shown that such regime changes occurred in the case of the dollar-deutschemark exchange rate over the recent float. Mahavan and Wagner (1999), Marsh (2000), Taylor and Peel (2000), Taylor et al (2001), Clarida et al (2003), Frömmel et al (2005a,b) and De Grauwe and Vansteenkiste (2007) were among the first studies to analyze the monetary model in a Markov-switching model for a set of main bilateral exchange rates and they provided support in favour of a fundamental model. Bacchetta and van Wincoop (2009) argued that large and frequent variations in the relationship between the exchange rate and macroeconomic fundamentals become evident when structural parameters in the economy are unknown and subject to changes.…”

Can Markov switching models predict excess foreign exchange returns?

“…2 The second strand uses lower frequency Markov switching models to forecast exchange rates (Engel and Hamilton, 1990;Engel, 1994); this has enjoyed limited success. 3 We view our work as complementary to that of Dewachter (2001) and Clarida et al (2003), who use elements of both threads in very different ways. Dewachter (2001) constructs a Markov switching model that is able to produce simulated weekly data on which trend-following trading rules are successful.…”