Build a GARCH(1,1) model with Gaussian innovation for the resent series. Use the Ljung-Box statistics up to lag 10 to test if the estimated innovations $\hat{e_t}$ are serially correlated. Plot the ACF of $\hat{e_t}$ and the rejection boundaries of a 5% level test of zero autocorrelation at each lag. (data available at http://web.stanford.edu/~xing/statfinbook/_BookData/Chap06/intel_d_logret.txt)