Obtain closing data and compute daily returns for NFLX and for GLD for the first calendar quarter 2018.

(a) Bivariate boxplot; bagplot.

Produce a bagplot of the returns of NFLX and GLD for that period.

(b) Bivariate kernel density estimators.

Choose four different smoothing matrices for a bivariate kernel density estimator. For the four different matrices, choose three that are diagonal (with equal values, then the first one larger, then the second one larger) and choose one proportional to the inverse of the sample variance-covariance matrix.

Fit a bivariate kernel density using four different values for the smoothing matrix. (This is the \(\mathrm{H}\) argument in the \(\mathrm{R}\) function kde.) Make contour plots of the fitted densities.

(c) Bivariate densities.

Now, compute some relevant summary statistics, and in light of those statistics, as well of the density plots, describe the bivariate distribution of the NFLX and GLD returns.