Linear regression model, volatility measure, stock return predictability

[Filip Pierzgalski]

Hi, Statalist!


I am writing a thesis on the stock market predictability, using a linear regression model:

I have a monthly data set for the stock market. I am using 17 different macroeconomic and technical variables to predict the stock market. One of them that has been advised by my supervisor was the ''volatility'' measure. I have been trying to construct such a predicting variable, but it seems that there is a lot of different methods. Three of them that I came across were rolling standard deviation:

''ARCH model'' and ''realized volatility estimated from daily data for each month'' ( i only have monthly data).

I have been trying to produce one of these methods in Stata but I don't know which approach is best in my situation since a lot of literature that I found on the topic is using daily data. Do you have any advice on what to do?

. dataex t return, count(30)

* Example generated by -dataex-. To install: ssc install dataex
clear
input float t double return
 1  .01242389
 2  .02019115
 3  .02964644
 4  .03405454
 5  .02848361
 6 -.01542547
 7 -.00863808
 8  .01915629
 9  .01136751
10 -.05886838
11  .03742423
12   .0257288
13  .02951672
14  .00101633
15  .04072984
16  .03142682
17 -.00213872
18 -.00795212
19 -.00034236
20 -.00761387
21 -.01335902
22  .01832136
23 -.00212152
24  .03346823
25  .01082923
26  .01575765
27 -.01307993
28 -.00878622
29  .01496919
30 -.01535528
end

Listed 30 out of 1764 observations


Thank you in advance for all the help, and if I am unclear in any way, please let me know.

[Phil Bromiley]

You didn't get a quick answer. You'll increase your chances of a useful answer by following the FAQ on asking questions - provide Stata code in code delimiters, readable Stata output, and sample data using dataex. Assume we are not from your area.

While some come from finance, your question is really about finance not Stata. You are supposed to be the subject expert. Ask your adviser.