Abstract
This paper aims at modeling and forecasting volatility in both oil and USD exchange rate markets using high frequency data. We test whether extreme co-move-ments (co-jumps) between these markets, as well as intraday unexpected news, help to improve volatility forecasting or not. Accordingly, we propose different extensions of Corsi (2009)’s model by including co-jumps and news. Our analysis provides two interesting findings. First, we find that both markets exhibit significant co-jumps driven by unexpected macroeconomic news. Second, we show that our model outperforms Corsi (2009)’s model and provides more accurate forecasts. In particular, while co-jumps constitute a key variable in forecasting oil price volatility, the unexpected news is relevant to forecasts of USD exchange rate volatility.
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have