Session Ready
Exercise

Volatility and correlation of FX returns

In this exercise, you will discover evidence of volatility and serial dependence in daily and weekly exchange-rate log-returns. The dataset fx contains daily log-returns for the "EUR_USD", "GBP_USD", "JPY_USD" and "CHF_USD" exchange rates while fx_w contains the corresponding weekly log-returns. Both are in your workspace.

Note that foreign exchange trading takes place every day of the week although the lower volume of trading at weekends leads to an unusual weekly correlation cycle which will be evident in one of the pictures you create.

Instructions
100 XP
  • Plot the multivariate time series fx and fx_w with the option type = "h".
  • Use acf() to make acf plots for both fx and the absolute values of fx.
  • Use apply() to conduct the Ljung-Box test, with a lag of 10, on the components of fx and then their absolute values.
  • Use acf() to make acf plots for both fx_w and the absolute values of fx_w.
  • Use apply() to conduct the Ljung-Box test, with a lag of 10, on the components of fx_w and then their absolute values.