it applies to all instruments, but you need a portfolio to get into trend following, because you don't know which market will have the biggest trend at any moment in time
there is no best market for trend following, that's why you need to trade +50 markets in a portfolio so hopefully one or two will have a big trend to pay for all.
@@athurgirard1790 I wonder if he doesn’t show the whole development process. I always use in-sample vs out of sample testing to ensure something is robust and can help reduce overfitting. I think using oscillators, moving averages, or other signal processing techniques are more prone to overfitting since they have outputs that are optimizable such as the lookback period. I think IF one is to use such indicators as an entrance signal, they have to start with a fixed lookback length and not optimize it. Then add pattern filters from a pattern library to find when the crossover is best and worse. ALL backtesting, optimization, walk-forward testing, machine learning, etc is some form of curve fitting. The trick is to find ways to reduce the number of optimizable inputs that have numeric values such as period length, overbought/oversold levels, etc to keep it simple.
Good inputs for the development of trendfollowing strategies here. THX.
Thank you for being a loyal viewer of the channel 🙏
Always very interesting analyses.Thank you Ali
My pleasure!
Keep them coming. Loving seeing more fx. I am finding it easy to generate strats on the indicies. FX always seems to require more work.
More to come!
Thank you Ali, always interested, and very well researched and presented 😀
Glad you enjoyed it
thanks for this, i wonder if this is something that only works with forex, i don't really trade forex
it applies to all instruments, but you need a portfolio to get into trend following, because you don't know which market will have the biggest trend at any moment in time
Hello! I love you videos
What would be the best markets for this strategy?
there is no best market for trend following, that's why you need to trade +50 markets in a portfolio so hopefully one or two will have a big trend to pay for all.
Too technical for me
that's ok, there are million ways to make money in the market.
I like your videos for ideas but i trully think you're approach is curve fitting way too much which makes it hard to trust and take seriously
all strategies on the channel are not fully baked, you can easily take any of them and apply your robustness workflow before committing to it live.
Most of these trading channels give you just a little bit to keep coming back, but nothing really profitable. Good for ideas though
@@athurgirard1790 I wonder if he doesn’t show the whole development process. I always use in-sample vs out of sample testing to ensure something is robust and can help reduce overfitting. I think using oscillators, moving averages, or other signal processing techniques are more prone to overfitting since they have outputs that are optimizable such as the lookback period. I think IF one is to use such indicators as an entrance signal, they have to start with a fixed lookback length and not optimize it. Then add pattern filters from a pattern library to find when the crossover is best and worse. ALL backtesting, optimization, walk-forward testing, machine learning, etc is some form of curve fitting. The trick is to find ways to reduce the number of optimizable inputs that have numeric values such as period length, overbought/oversold levels, etc to keep it simple.