You sir are a brilliant teacher, you explain us the strategies in lucid and easy layman language that even a novice trader can understand the nuances of trading. Keep on doing the brilliant work. Cheers.
13.11.2021. Martyn: I have tested the ideas in your video and I can tell you they work. I have taken a mean reversion strategy and I have introduced a filter based on SMA and the Calmar ratio has improved by a 30%. Awesome. Thanks a lot. 14.11.2021. I have to add that the efectiveness of the filter seems to depend somewhat on the asset.
How do you deal with the regime classification delay? How long after a cross is it no longer considered a ranging regime? As far as I can tell with the crosses shown, you would still be classified as a range regime while the explosive move was happening, and even using relatively short delay periods, by the time it's classified as a momentum regime, the value in the explosive move is over. If the delay period is dropped low enough that it can classify momentum regimes in a fast enough manner to be useful, the cross filter will constantly mis-classify ranging regimes as momentum regimes.
Will you cover how to detect that the regime is broken? The issue with mean reversion is the penny in front of the steam roller problem. Do you use stops on the positions and maybe build in slowly while looking for slowing momentum on a shorter timeframe. I can think of other things but was interested in your take.
When price goes up you should think sell sell sell. When price goes down you should think buy buy buy. What actually occurs is FOMO amongst private savers and eveyone wants to get in whereas traders are looking to get out. When you see a price increase in your local retailchain for goods you dont buy more but instead look for alternatives. But somehow that is not the case when it comes to stocks. Everyone wants to get in on the action and that drives up the price. When something goes up to fast it usally comes down as fast the reason being that the market dont like these movements though every price point should be traded. So sell after 2-3 days and buy when it comes down again.
Martyn, I believe that one of the best robustness tests is to test whether the strategy is profitable on another market ( with less restricted conditions), but in our context of mean reversion, isn't that over optimistic ? Mean reversion strategies are by their nature fitted to the market condition, so trying to test them ALSO on another market wouldn't that limit our passed strategies? that works with me with breakouts, but I'm having a hard time with mean reversion.. I would appreciate to hear your insight !
You sir are a brilliant teacher, you explain us the strategies in lucid and easy layman language that even a novice trader can understand the nuances of trading.
Keep on doing the brilliant work.
Cheers.
13.11.2021. Martyn: I have tested the ideas in your video and I can tell you they work. I have taken a mean reversion strategy and I have introduced a filter based on SMA and the Calmar ratio has improved by a 30%. Awesome. Thanks a lot. 14.11.2021. I have to add that the efectiveness of the filter seems to depend somewhat on the asset.
You know it's a good video when your mind sets to coding ideas before it's even finished. 😉
Algo Trading University ! Thank you Darwinex and Martin for the knowledge you provide. Always grateful
Amazing content as always. Always happy to see your notifications.
Very well explained, Knowledge to word count ratio is through the roof with this one.
How do you deal with the regime classification delay? How long after a cross is it no longer considered a ranging regime? As far as I can tell with the crosses shown, you would still be classified as a range regime while the explosive move was happening, and even using relatively short delay periods, by the time it's classified as a momentum regime, the value in the explosive move is over.
If the delay period is dropped low enough that it can classify momentum regimes in a fast enough manner to be useful, the cross filter will constantly mis-classify ranging regimes as momentum regimes.
Thanks Martyn! really looking forward to this series
Really great tutorial with good pieces to follow. Nice to be with you again. Thanks! 🤓
Hi Whats the time scale of the chart?
Gracias, muy bueno
Will you cover how to detect that the regime is broken? The issue with mean reversion is the penny in front of the steam roller problem. Do you use stops on the positions and maybe build in slowly while looking for slowing momentum on a shorter timeframe. I can think of other things but was interested in your take.
Layer your mean reversion strategies on top of low correlated strategies Aka trend following strategies , that’s one of my solutions
Thank you Martyn , Nice informative video
Using Aroon as a market filter shows promise.
Love this!
Kuafman Efficiency instead of MA?
waiting for next one
When price goes up you should think sell sell sell. When price goes down you should think buy buy buy. What actually occurs is FOMO amongst private savers and eveyone wants to get in whereas traders are looking to get out. When you see a price increase in your local retailchain for goods you dont buy more but instead look for alternatives. But somehow that is not the case when it comes to stocks. Everyone wants to get in on the action and that drives up the price. When something goes up to fast it usally comes down as fast the reason being that the market dont like these movements though every price point should be traded. So sell after 2-3 days and buy when it comes down again.
Martyn,
I believe that one of the best robustness tests is to test whether the strategy is profitable on another market ( with less restricted conditions), but in our context of mean reversion, isn't that over optimistic ? Mean reversion strategies are by their nature fitted to the market condition, so trying to test them ALSO on another market wouldn't that limit our passed strategies?
that works with me with breakouts, but I'm having a hard time with mean reversion.. I would appreciate to hear your insight !
this the one
I can tell by the EURUSD price on the left hat you filmed this on the 10th of November 2021. That's all.