1. Do the differenced series appear stationary according to the augmented Dickey–Fuller tests?  2. Do you see evidence of autocorrelations in the differenced series? If so, describe these...

1. Do the differenced series appear stationary according to the augmented Dickey–Fuller tests?

 2. Do you see evidence of autocorrelations in the differenced series? If so, describe these correlations.

For the remainder of this lab, we will focus on the analysis of the 91-day T-bill rate. Since the time series are quarterly, it is good to see whether the mean depends on the quarter. One way to check for such effects is to compare boxplots of the four quarters. The following code does this. Note the use of the cycle() function to obtain the quarterly period of each observation; this information is embedded in the data and cycle() simply extracts it.