OPEC News and Exchange Rate Forecasting Using Dynamic Bayesian Learning

We consider whether a newspaper article count index related to the organization of petroleum exporting countries (OPEC), which rises in response important OPEC meetings and events connected with production levels, contains predictive power for foreign exchange rates G10 countries. The applied Bayesian inference methodology synthesizes wide array established approaches modelling rate dynamics, whereby various vector-autoregressive models are considered. Monthly data from 1996:01 2020:08 (given an in-sample 1986:02 1995:12), shows that incorporating news-related into proposed leads statistical gains out-of-sample forecasts.