Detecting Bubbles in the USD-JPY Exchange Rate by Sequential Monte Carlo Methods

Date

2021-02-12

Journal Title

Journal ISSN

Volume Title

Publisher

Abstract

This paper uses recently developed Bayesian techniques in the analysis of stochastic bubbles in the USD-JPY exchange rate. After the fundamental value of the price series is removed, the exchange rate is subject to two regimes. The first regime follows a mean-reverting process around a longterm moving average. The second regime is an autoregressive process with an explosive root. The SMC2 particle filter jointly estimates the hidden state and model parameters in real time. This method can readily deal with changes in market behavior and provides a measure of parameter uncertainty. Significant evidence of bubbles in the USD-JPY exchange rate were found. Furthermore, two trading strategies are devised and tested. Both strategies produce higher Sharpe ratios than a directional-trading benchmark.

Description

Keywords

Particle filter, MCMC, SMC, Monte Carlo, Exchange Rate, Trading, Finance, FOREX, Bubble,

Citation