Financial flows in the exchange rate market may explain future exchange rate movements at longer horizons(Evans & Lyons 2009). Such persistence seems to call for a more fundamental explanation than varyingexpectations alone. To what extent do these flows reflect subtle changes in the fundamentals, yet to berevealed and announced through public statistics, caused by the aggregation of thousands of individualinvestment or purchase-timing decisions? The central bank of Norway is currently the only institutio n to publish extensive data on such financial flows. Could proper filtering and modeling of these data, through an extension of the Evans-Lyons portfolio shifts model or by a hidden Markov model for the real time state of the economy extending the informa tional space, predict movements of additional economic aggregates, either by themselves or combined with other financial flows data like investment portfolios and measures of international capital movements: data that is not reflected in the exchange rate but may reflect underlying
fundamentals?
