Thanks Rob - next question(s). On one of the slides you listed the following questions where all but #6 had a red line through them. What period of time do trends last for? When should we enter trends? When should we exit trends? Should we have a stop loss rule? What is it? How do we identify markets that are, or aren’t “trend friendly”? How do we identify how strong the trend is? What size should our positions be? 1) What does the red line mean? Not covered in your presentation. Not an important question. Not enough time to discuss in the time allotted, etc? 2) #6 was not red lined. It didn't seem that you answered that question though maybe you did and it wasn't obvious enough for me. Was that question answered?
Crikey - I don't have time to walk you through in this much detail. Are you watching the video (here or here) or just looking at the slides? GAT
What period of time do trends last for? - as long as they last When should we enter trends? - we gradually enter positions based on the forecast strength When should we exit trends? - and vice versa Should we have a stop loss rule? What is it? - no, there are no binary trades. How do we identify markets that are, or aren’t “trend friendly”? we don't, we assume that all markets have the same risk-adjusted return (probably not true, but a reasonable assumption). Focus then on diversifying, which will roughly 3x the performance. How do we identify how strong the trend is? Forecast strength What size should our positions be? Whatever is necessary to meet our volatility target, that we specify beforehand. Forecast = (ewma_fast-ewma_slow) / ewma_vol (closely related to the Sharpe ratio) - this is the moving average crossover.
Hi Rob, Was just looking at the slides. I didn't have questions about all the points just the meaning of the red line strike through and #6 whether, as it wasn't red lined, that it was described in the rest of the slides as I wasn't sure. I would think that #5 is an interesting and relevant question. My thought on #5 is that, in general, if you put a N% swing chart over several markets that the one with the fewest swings is the most trend-friendly. Though what N is is an open question. I would also think that the higher the r-squared the better/stronger the trend. Your thoughts?
I'd strongly suggest watching the video (the New York one is easier as the slides are visible), and watching it all the way through. Then if you still have questions I'll be happy to answer them. GAT
I just watched the video. Very informative. Tip: Watch at 1.25speed to save sometime. Also, probably can be edited to take out some Q/A which were obvious or not contributing too much value to the video.
Glad you enjoyed it. Not my video (recording of a live presentation at QuantCon, so owned my quantopian) so I can't do that. Also some subjectivity about what is obvious... GAT
BTW, you got me into using swigibpy a few years ago when I was looking around for some framework for IB Python for my automated trading system. Thank you for that!