Skip to Main content Skip to Navigation
Preprints, Working Papers, ...

Fundamentals and exchange rate forecastability with machine learning methods

Abstract : Using methods from machine learning -- sequential ridge regression and the exponentially weighted average strategy both with discount factors -- that do not estimate a model but directly output forecasts we show that fundamentals from simple exchange rate models (PPP, UIRP and monetary models) consistently allow to improve exchange rate forecasts for major currencies over the floating period era 1973--2014 at a 1 month forecast and allow to beat the no-change forecast. ``Classic'' fundamentals hence contain useful information about exchange rates even for short forecasting horizons. Such conclusions cannot be obtained when using rolling or recursive OLS regressions as in the literature.
Document type :
Preprints, Working Papers, ...
Complete list of metadata

https://halshs.archives-ouvertes.fr/halshs-01003914
Contributor : Gilles Stoltz <>
Submitted on : Thursday, April 23, 2015 - 3:23:52 PM
Last modification on : Friday, November 6, 2020 - 3:10:07 PM
Long-term archiving on: : Monday, September 14, 2015 - 12:56:44 PM

File

ExchangeRates_2015_HAL.pdf
Files produced by the author(s)

Identifiers

  • HAL Id : halshs-01003914, version 3

Citation

Christophe Amat, Tomasz Michalski, Gilles Stoltz. Fundamentals and exchange rate forecastability with machine learning methods. 2015. ⟨halshs-01003914v3⟩

Share

Metrics

Record views

100

Files downloads

76