time series data HELP!!! please..

[wipaberry]

Hey guys, I am working on a research about how the economic indicators ( inflation rate, CPI, interest rate) are influencing the stock price index of my
country. The data are secondary time-series data dated back for 6 years. I am very new to econometric and I really need help.
I basically know that we have to conduct a unit root test to test for the stationary of the data using Dicky-Fuller test
however, what happen next? I mean if the data is not stationary what do I do?
Also after the data is being adjusted to be stationary, can I use OLS to run the data? After running the OLS, then I shall test for autocorrelation?
In the old paper that I've found, they just take the data and run and OLS. After the results are made, they just test for Durbin-Watson
and if there are autocorrelation, they just do an AR(1) or AR(2) and get the results.
Please help me here, I will be really appreciated :)

[paolodio]

I can only answer to your first question : take log returns of your data to make it stationary, that is, take the logs of the first differences.