1. Download monthly data for the REIT return index in file reit_data.xls. The data can also be found on the FTSE /NAREIT website.

2. Compute the simple return (R_t), unfiltered dividend yield, DY , and price appreciation.

3. Compute the smoothed DY over the previous 12 months, denoted by DY^s.

4. Compute the log return (rt), log DY, and log DY s

5. Compute annualized summary statistics for the simple return (R_t), log return (r_t), DY, DY s, log DY, and log DY^s- mean, standard deviation, autocorrelation, skewness, kurtosis.

6. Run the following regressions for simple returns

(a) R_t on R_t-1
(b) R_t on R_t-1 and R_t-2
(c) R_t on R_t-1 and R_t-2..., R_t-12
(d) How do you test for persistence in returns in AR(1) model?
(e) How do you test for persistence of returns in AR(p) model?

7. Run the same regressions as in 6.a-6.e but for log returns. Are the results di?erent?

8. Are REIT returns predictable by their own lags?

9. Run the following returns-dividend yield regressions:

(a) R_t on DY_t-1
(b) R_t on DY^s_t-1
(c) r_t on log DY_t-1
(d) r_t on log DY^s_t-1
-

10. Are returns predictable by the dividend yield? If you had to rely on one specification, which one would you choose?

11. Should you adjust your results in question 9? How? Display the adjusted results. How large is the bias?

12. Redo the regressions in questions 6, 7, and 9 for the 1972-1993 and 1994-2014 subsamples.

Attachment:- reit_data.xls