Hello, This is my first post on this forum. I've searched and this looks like the best forum for people interested in ATS to share their thoughts. My partner and I have been playing with ATS for a few years now. We have programmed and back tested about 5 different systems based on some simple principles. The results look great. Weâve back tested back a few decades and have significant out of sample data. We've also written in re-optimizing code which is also very clean and the system adapts to changing markets. I'm trying to do some testing myself on the overall portfolio and unclear how to determine the volatility for a portfolio of ATS's. I have the standard deviation of each system's profits; I am trying to determine the standard deviation of the whole portfolio. Is this as trivial as determining the covariance between each system and then using basic statistics? I'm an actuary and work extensively with variances (though nothing related to trading) and feel like I might be over-thinking the whole thing. I look forward to any input! Thanks
Isn't it just the traditional formula for the calculation of portfolio variance? I mean this: You should be able to find it here: http://en.wikipedia.org/wiki/Modern_portfolio_theory
you know, one would think. But i have a feeling it can not be that trivial Now im going to get this 5X5 covariance matrix. Fun!