The file S03_64.xlsx lists monthly data since 1950 on the well-known Dow Jones Industrial Average (DJIA), as well as the less well-known Dow Jones Transportation Average (DJTA) and Dow Jones Utilities Average (DJUA). Each of these is an index based on 20 to 30 leading companies (which change over time).
a. Create monthly differences in three new columns. The Jan-50 values will be blank because there are no Dec-49 values. Then, for example, the Feb-50 difference is the Feb-50 value minus the Jan-50 value.
b. Create a table of correlations of the three difference columns. Does it appear that the three Dow indexes tend to move together through time?
c. It is possible (and has been claimed) that one of the indexes is a “leading indicator” of another. For example, a change in the DJUA in September might predict a similar change in the DJIA in the following December. To check for such behavior, create “lags” of the difference variables. To do so, select Lag from the StatTools Data Utilities dropdown list, select one of the difference variables, and enter the number of lags you want. For this problem, try four lags. Then press OK and accept the StatTools warnings. Do this for each of the three difference variables. You should end up with 12 lag variables. Explain in words what these lag variables contain. For example, what is the Dec-50 lag3 of the DJIA difference?
d. Create a table of correlations of the three differences and the 12 lags. Use conditional formatting to color green all correlations greater than 0.5 (or any other cutoff you choose). Does it appear that any index is indeed a leading indicator of any other? Explain.