Skip to Main Content
 

Global Search Box

 
 
 
 

Files

ETD Abstract Container

Abstract Header

Predicting short term exchange rates with Bayesian autoregressive state space models: an investigation of the Metropolis Hastings algorithm forecasting efficiency

Gendre, Victor Hugues

Abstract Details

2015, Master of Science, Ohio State University, Statistics.
In this thesis, we propose four Bayesian AR(1) models, namely an AR(1) with constant variance and constant slope coefficient, an AR(1) with stochastic variance and constant slope coefficient, an AR(1) with constant variance and time varying slope coefficient and finally an AR(1) with stochastic variance and time varying slope coefficient. These models were used to examine the daily predictability of the EUR/USD and USD/JPY exchange rate returns. These autoregressive models were fitted using the classic Metropolis Hastings (MH) within Gibbs algorithm as well as the more novel approach developed by Golightly and Wilkinson (2008). We found that while all the models outperform a driftless random walk in terms of MSPE, they fail to capture the volatility of exchange rate returns. We also observed the superiority of the Golightly and Wilkinson technique compared to the MH within Gibbs algorithm regarding the speed of the inference procedure.
Radu Herbei, Ph.D (Advisor)
Laura Kubatko, Ph.D (Committee Member)
161 p.

Recommended Citations

Citations

  • Gendre, V. H. (2015). Predicting short term exchange rates with Bayesian autoregressive state space models: an investigation of the Metropolis Hastings algorithm forecasting efficiency [Master's thesis, Ohio State University]. OhioLINK Electronic Theses and Dissertations Center. http://rave.ohiolink.edu/etdc/view?acc_num=osu1437399395

    APA Style (7th edition)

  • Gendre, Victor. Predicting short term exchange rates with Bayesian autoregressive state space models: an investigation of the Metropolis Hastings algorithm forecasting efficiency. 2015. Ohio State University, Master's thesis. OhioLINK Electronic Theses and Dissertations Center, http://rave.ohiolink.edu/etdc/view?acc_num=osu1437399395.

    MLA Style (8th edition)

  • Gendre, Victor. "Predicting short term exchange rates with Bayesian autoregressive state space models: an investigation of the Metropolis Hastings algorithm forecasting efficiency." Master's thesis, Ohio State University, 2015. http://rave.ohiolink.edu/etdc/view?acc_num=osu1437399395

    Chicago Manual of Style (17th edition)