If anyone is using time series analysis for stocks, how many instruments are you trading? And, how do you decide how much capital to allocate to each instrument?
All liquid stocks. Portfolio / 2 at any given time, no leverage. My algo generates ~7 signals every day, so most of the time they don't overlap.
Don't listen to me, I only made a 50% return last year with no losses at all, but if you are dumb then go ahead and use your stoneage tools
How about this time series analysis. lol I know this isn't the same thing but I couldn't resist. RTY high today was made 3:13am EDT and low at 2:19pm EDT So you say? 11 hours and 6 minutes. Again so? 666 minutes. I track such minutia, for other purposes, is how I know. Jumped out at me from Excel page during market close so didn't do me a bit of good today.
For choosing potential assets, I pick ETFs as in this post where each ETF is correlated to any of the other ETFs below a threshold. A similar idea could be used for individual stocks. For position sizing, I use the method from this book:
Sounds good. Where can retail players get this model applied to indices or stocks like this website? https://www.macroaxis.com/forecast/spy Any insights/comments on this website simple model? What is the difference between the 2 models explained in layman terms?
Everything is shit if you don't know how to use it. So shit or not depends on YOU, not on themath behind the indicator. For me stochastics don't work...
This article gives a brief comparison between deterministic and stochastic models. https://www.kddanalytics.com/practical-time-series-forecasting-deterministic-stochastic-trend/ But how does a trading program keep switching between deterministic to stochastic model after price shocks when there are so many?