STATA if you know is a STAT program, and we use that in our ECON STAT class. Just wondering if it would help me/be beneficial for me in my future (in my pursuit) to give that extra 10% on it. If you never heard of STATA you could find more info here at www.stata.com
For investors, I might see an inkling of a utility, but for trader no. Stata 7.0 Intercool is more geared towards time series analysis and mulitvariate regression for STATIONARY variable series. (Qualifer: You could convert nonstationary into stationary) If you want to do "regression" for intraday ... Excel functions are more efficient IMO. Hope that helps. P.S. what school do you go to? Stata is usually reserved for Graduate Seminars.
used stata as well while in school. more relevant as a tool for academic research than trading per se. good program though, couldn't hurt to be comfortable using it.
Who said academic research is not good for a trader? STATA's a good package. I think you'll benefit if you know how to apply econometrics in developing and testing strategies. I personally stick with SAS and TSP with some R and Splus on top. Didn't have the money to buy STATA at first and after I got used to other packages, I can do pretty all it does elsewhere. The interface is much better though, or so I heard. Good Luck.
Excel regression utility is extremely rudimentary. If he wants to do anything more fancy (e.g. fixed or radom effects etc), stata might be a good pick.
Its rudimentary for a reason. Someone with only an undergraduate degree would only scratch the surface of Stata. Stata is a powerful powerful resource that is manufactured with rigorous academics in mind. With Excel, you have your basic: covariance-variance, correlation, simple regression. If you want anything else, use VBA.
I agree, that's why I was gonna say (but didn't for lack of time) that by learning how to use stata he might also learn a whole lot of additional techniques. Some of them are a bit sophisticated, but not all that much. PS I use VBA a bit too, but mainly to prepare the data for other packages and for simulations.
I'm also a bit biased b/c to me if you are putting your (or others') money on the line, using smth rudimentary to do the analysis doesn't seem appropriate. Excel won't tell much about how sensitive your results are to influential observations etc etc. In short, you won't know how reliable the findings are. And I like them to be to based trading decisions on them
I think he has to get a grad degree or at the very least a very dedicated ambition to learn intermediate time series analysis (aka books, ex. Hamilton's classic) I cant imagine an undergrad running time series decomposition or markov process, or basic bootstrapping.... you now what i mean? To us, it seems so easy, but the theory and concept must be learned. Implementation is pretty useless unless the underlying factors are understood. There's really no short cut. well, that just my two cents
In response to your latest comment, very very few traders think along the same line - even though they should. most traders just need the basic, i.e. simple rsquare, coefficient, alpha, correlation ... But if you're into statistical arbitrage, statistical models, and what have you.... then yes I agree. But these areas are "reserved" - for a lack of a better term - for Quant Shop.