Premium
Difference‐Equations Solution of Exchange Rate Dynamics
Author(s) -
Akiba Hiroya
Publication year - 2002
Publication title -
review of international economics
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.513
H-Index - 58
eISSN - 1467-9396
pISSN - 0965-7576
DOI - 10.1111/1467-9396.00329
Subject(s) - economics , saddle point , stability (learning theory) , exchange rate , asset (computer security) , rational expectations , econometrics , replicate , mathematics , forcing (mathematics) , contrast (vision) , infinity , mathematical economics , statistical physics , computer science , mathematical analysis , physics , statistics , geometry , computer security , machine learning , artificial intelligence , macroeconomics
The paper focuses on an appropriate and consistent theoretical as well as empirical model of the rational‐expectations version of the asset‐market approach to exchange rate determination. Using a “sticky‐price” monetary model, and formulating the model by a difference‐equations system, explicit solutions are obtained as functions of forcing variables extending to past dates only. This “backward‐looking” characteristic of the solution is in stark contrast to conventional “forward‐looking” models, and alleviates empirical investigation because of the need for past data only. This discrete dynamic model is superior to the corresponding continuous model, because its solutions neither exhibit the empirically unfounded “overshooting” behavior, nor have saddle‐point (in)stability. Rather, the exchange rate is shown to follow an oscillatory path with asymptotic stability, and this seems to replicate the actual movements closely approximated by a random‐walk process.