Hi everyone,
I have a time series data with companies and stock returns and I would like to check if returns from a strategy is significantly different from zero. I was going to do the t-test but the t-test assumes indecent data, while my data is autocorrelated.
Is there a command to run a t-test when you have autocorrelated data?
My data looks like this:
don't mind the log, I want to look at the WML1 which contains returns for every month. I also have a variable WML11 which only contains one value every month. I used this one for the mean and the t-test.
the command I used before was test WML11==0
I hope you can help me
I have a time series data with companies and stock returns and I would like to check if returns from a strategy is significantly different from zero. I was going to do the t-test but the t-test assumes indecent data, while my data is autocorrelated.
Is there a command to run a t-test when you have autocorrelated data?
My data looks like this:
don't mind the log, I want to look at the WML1 which contains returns for every month. I also have a variable WML11 which only contains one value every month. I used this one for the mean and the t-test.
Code:
* Example generated by -dataex-. For more info, type help dataex clear input str6 Code float(mofd WML1 logWML1) "9215PR" 731 -.2165781 -.2440839 "9526R7" 731 -.2165781 -.2440839 "692612" 731 -.2165781 -.2440839 "9216VU" 731 -.2165781 -.2440839 "29509W" 731 -.2165781 -.2440839 "912326" 731 -.2165781 -.2440839 "255383" 731 -.2165781 -.2440839 "74085P" 731 -.2165781 -.2440839 "9459TV" 731 -.2165781 -.2440839 "923388" 731 -.2165781 -.2440839 "9406X3" 731 -.2165781 -.2440839 "533153" 731 -.2165781 -.2440839 "8694XQ" 731 -.2165781 -.2440839 "9445LM" 731 -.2165781 -.2440839 "9535FE" 731 -.2165781 -.2440839 "697661" 731 -.2165781 -.2440839 "698992" 731 -.2165781 -.2440839 "8725EZ" 731 -.2165781 -.2440839 end format %tm mofd
I hope you can help me
