First venture into ARMA models on real data
Here I apply a few concepts that I learnt in my courses to see if they actually make sense in the real world. It's hard to digest the fact that even doing an "a+b" in the real world is not that easy - acquiring data, ensuring that it is exactly what you want, cleaning it, checking if it makes sense and clipping out those that don't and still be left with enough meaningful numbers involve a series of massive challenges that only a practitioner would comprehend.
I attempt to implement ARMA time series model on large cap stock returns in an attempt to predict their future returns. It's a first attempt but nevertheless gives an excellent idea about the issues I described above and if the time series concepts work by a brute force methodology.