For 682 pt a: Could you elaborate on what you meant by "bootstrap in order to determine the standard errors of the median"? Do you mean for us to perform random sampling with replacement on the returns data and then apply the 'Studentized Bootstrap'?

Great question! As you know from your research into what the bootstrap is, it provides a way to get at the sampling varibality of various statistics. The statistic in question here is the median. The general idea is as you say to perform random sampling with replacement and calculate the statistic B=5000 times to get a distribution for the statistic. The standard deviation of this distribution is the standard error of the statistic. Programming the bootstrap is relatively simple, see for example http://www.statmethods.net/advstats/bootstrapping.html (Links to an external site.)Links to an external site..