Just had one doubt - since you are using greater than -2 so if 2 of the systems are giving +1 then one enters. Will not that result in a trend indicator which gives strength of direction like ADX which can go beyond 20 in a strong down trend and RSI or 5 period moving average again giving positives which could also trigger during market downturns. Just curious why you did not use a regime filter like 200 SMA that you usually employ to ensure that one does not go long in bear markets. Anyways learnt a lot from this very interesting video Ali. Thanks for posting it.
it depends on what type of systems you are using. so if I want only long mean reversion, then I build multiple mean reversion long systems using different methods and combine them to get one signal to go long on mean reversion (pull back)
Great Techique, Ali. I think the next step is optimise the weight of each votaing. I mean, maybe the SMA is more import of the RSI, the RSI is more important of the ADX, etc.
you are correct, that is great idea, you can over weight the signal that is important to you, say one signal has high win %, so you can overweight that to get higher win %, of course as always you need to test 🙂
Just an idea, don't know if you have already seen or test something like that: some hedge fund add a constraint that come control system theory for limiting overtrading and lost, when a sequence is too good or too bad. It can be summarized like that: if you have a sequence of bad trade, skip the next, similar for winning trades. We understand that winning trades can be longer than loser trades since the biais for the markets go up. This principle is used in many area like training swedish pilots which miss or hit their targets. Never seen any results in trading, just read taht many hedge finds use it and I have been alway curious how good or not is it.. Best regards and continue your good working!
turtle trades used that feature in their system 1, so if they lose two trades in a row they skip the next trades until they get a winning trade. it works for short term system, provided that you have another entry on the long term system so as not to miss it. it can be programmed, but I haven't tested it. will add it to post ideas.
This is similar to a strategy that Art Collins talks about and he refers to it "multiple biases". My back testing using some very basic rules not much different to yours has proved it to be a very effective strategy in trending markets.
great content as usual! by the way, lines 25 and 31 can be omitted because edge is bounded in -4..4. I wonder how did you choose which boolean expression will add 1 and which will subtract 1?
Great video. Pardon a possibly silly question, but how do you know hedge funds use this technique? And for a hedge fund is it all built into an algo, or is there ever actually voting, like each person gets a vote?
Great content! Had the same idea to have such voting eng. why did you choose to go long at -2? It means that 2 votes are for the short side.. am I right?
I just ran an optimization and found about 40% of profitable runs get in at a negative value. I think it is because of the mean reverting nature of the instrument. the values for the other sets were positive, but they were cut in editing 🙃
Hi Ali, could you kind share the mt4 ex4 file for the strategy pls it benefits us as well, its like treat for us for watching you videos and liking your videos
![BLACK BAG - Official Trailer [HD] - Only in Theaters March 14](http://i.ytimg.com/vi/Du0Xp8WX_7I/mqdefault.jpg)
watch more secrets ruclips.net/video/6jzQck5lbDA/видео.html
how can i watch this
Just had one doubt - since you are using greater than -2 so if 2 of the systems are giving +1 then one enters. Will not that result in a trend indicator which gives strength of direction like ADX which can go beyond 20 in a strong down trend and RSI or 5 period moving average again giving positives which could also trigger during market downturns. Just curious why you did not use a regime filter like 200 SMA that you usually employ to ensure that one does not go long in bear markets. Anyways learnt a lot from this very interesting video Ali. Thanks for posting it.
it depends on what type of systems you are using.
so if I want only long mean reversion, then I build multiple mean reversion long systems using different methods and combine them to get one signal to go long on mean reversion (pull back)
Great Techique, Ali. I think the next step is optimise the weight of each votaing.
I mean, maybe the SMA is more import of the RSI, the RSI is more important of the ADX, etc.
you are correct, that is great idea, you can over weight the signal that is important to you, say one signal has high win %, so you can overweight that to get higher win %, of course as always you need to test 🙂
Just an idea, don't know if you have already seen or test something like that: some hedge fund add a constraint that come control system theory for limiting overtrading and lost, when a sequence is too good or too bad. It can be summarized like that: if you have a sequence of bad trade, skip the next, similar for winning trades. We understand that winning trades can be longer than loser trades since the biais for the markets go up.
This principle is used in many area like training swedish pilots which miss or hit their targets. Never seen any results in trading, just read taht many hedge finds use it and I have been alway curious how good or not is it..
Best regards and continue your good working!
turtle trades used that feature in their system 1, so if they lose two trades in a row they skip the next trades until they get a winning trade.
it works for short term system, provided that you have another entry on the long term system so as not to miss it.
it can be programmed, but I haven't tested it. will add it to post ideas.
@@StatOasis I remembered it, it's a long time ago that I've read it
👏👏👏 It would be great to have a video about this strategy but for sqx
sure, future video
This is similar to a strategy that Art Collins talks about and he refers to it "multiple biases". My back testing using some very basic rules not much different to yours has proved it to be a very effective strategy in trending markets.
You are correct, art collins walk the same line in his book
great content as usual! by the way, lines 25 and 31 can be omitted because edge is bounded in -4..4. I wonder how did you choose which boolean expression will add 1 and which will subtract 1?
Will add one when expression make money long and will deduct one when it makes money short
Great video. Pardon a possibly silly question, but how do you know hedge funds use this technique? And for a hedge fund is it all built into an algo, or is there ever actually voting, like each person gets a vote?
I have good relations with a handful of hedge fund managers.
It is all built in the algo, there is no human interventions
@@StatOasis I appreciate your videos and how you always answer comments.
Marvelous! Thank you Ali Casey!
Thank you for watching
Great content!
Had the same idea to have such voting eng.
why did you choose to go long at -2?
It means that 2 votes are for the short side.. am I right?
I just ran an optimization and found about 40% of profitable runs get in at a negative value. I think it is because of the mean reverting nature of the instrument.
the values for the other sets were positive, but they were cut in editing 🙃
Great video! Thanks Ali
Thank you for watching
great video
Thank you for watching
Hi Ali, could you kind share the mt4 ex4 file for the strategy pls it benefits us as well, its like treat for us for watching you videos and liking your videos
If strategy is built with SQX then it can be exported to MT4, but if i build using Easy Language (as in this video) then i cant export to MT4
AMAZING !!!!
Thank you
this is fuzzy logic in Strategy quant x
Yes, but fuzzy logic doesn’t vet each condition separately