Exchange Rate Models: Fundamentals, Horizons, and Research Limits
Summary
The document surveys a broad question about stochastic processes for currency exchange rates and how well they perform. Its main perspective is that exchange-rate research often emphasizes economic fundamentals, such as balance of payments, and that useful model choice depends on the investor's or trader's horizon. It distinguishes longer-horizon modeling from higher-frequency currency strategies, while noting that the public literature on recent research and high-frequency FX models may be limited.
The responses recommend older papers and an overview as starting points, but provide no specific process, calibration method, performance metric, or empirical results. The answer therefore functions as guidance on scoping a literature review rather than as a tested forecasting approach. The cited material is described as dated, and the response explicitly leaves room for other researchers to identify stronger high-frequency work. Readers should treat its assessment of the literature as time-bound and seek sources suited to their intended horizon and use case.
Key ideas
- Exchange-rate model choice depends in part on the trading or investment horizon.
- Currency research often incorporates macroeconomic fundamentals such as balance of payments.
- The response describes publicly available research on high-frequency FX models as limited.
- The suggested readings are starting points, and the document reports no model comparison or performance evidence.
Tags
Full text
# Statistical models for exchange rates? # Statistical models for exchange rates? What stochastics processes (if any) are used to model currencies exchange rates and how good are such models ? References on subj. are also highly welcome. ## Answer by Matt Wolf (score 7, accepted) https://quant.stackexchange.com/a/8080 I work extensively with currency models and have to admit there is not much in the public domain regarding recent published research that may satisfy your needs. Some of the below mentioned models incorporate stochastic components but please keep in mind that most research on currencies focuses on fundamentals (such as balance of payments) and depending on your investment and trading horizon they may suit your needs. There is hardly anything published that deals with high frequency currency models that I would deem noteworthy but I am happy to stand corrected by others if they can dig up such research. Having said that may I point you to some dated papers that may generate ideas and on which you may expand: - Olson is a highly respected data supplier as well as research house (they recently also offer investment services). They provide access to some of their older research: Olson - Another dated paper that walks through some of the more basic methodologies: Link The following papers provide somewhat of an introduction to modeling fx strategies in higher frequency space though I did not derive much added value out of those. See for yourself whether you like any of those: - washington.edu - bis - ucdavis.edu - essex.ac ## Answer by Bob Jansen (score 4) https://quant.stackexchange.com/a/8081 I'm not exactly sure what you are looking for but if you're interested in models for longer periods this overview paper might be a good starting point for you. If you can tell use more about what you want to do with the model it might be helpful to give more specific answers.
Shown in full with attribution under the source's licence. Licence: CC BY-SA 4.0 (Stack Exchange)
This summary was written by Stratmill's research agent from the original; it is not a copy of the source.