Here is a system I tried with ninjascript , it worked nicely, I am improving. Basically I created an RSI indicator by subtracting a fast RSI from slow RSI. Fast RSI has half the period of slow RSI. This way you have a relative RSI that is sensitive to momentum changes. Then I used standard error band on this new RSI. Long when RSI crosses above the upper band, short when RSI crosses below the lower band. These crosses indicate momentum initiations. I use TTM squeeze to filter for better signals. I use NQ 133 tick data. 100s of trades a day, very high win rate.
Use stochastic RSI indicator instead how's your RSI that you customized to your liking. Instead This stochastic RSI is the proper version of what you were trying to do you want to set it 80 and 20 overbought oversold and when the lines cross either at the bottom or top. It's one of the best indicators you can use
Hey, I mean the algorithmic trading, you are killin it. I can see we need to use indicators available to build a strategy, but I wanted to understand whether we can build a algorithm based on smc concept and our own entry types in that. Thank you.
Do you think this performance is really appliable in real life? As you mentioned, only selected stocks perform well on this setup, but when we work with the present, it looks difficult to guess what stock is going to work, even when a stock gives us a good pattern in the past, it doesnt garantee that the future will follow the same pattern
The best is to find let's 5 different patterns, and run them in parallel on let's say 10 different assets. We will always have loosing versus winning assets, but the candles patterns will make the difference in a way that the total is always in the positive returns side. This however requires more work to test lots of patterns and on lots of assets to make the selection. It's like any investment strategy you try to diversify all your parameters.
Thanks for the great video! Keep going on please! I have a question. If the agregated return is 186% in 7 years, then someone with 100€ would have 286€ after the 7 years of trading?
Had some success, but the key is understanding the market. Trading offers more potential than holding, especially with Trump’s policies changing things up. Charts alone won’t cut it, Honestly,James Whisman completely transformed my approach to trading. his insights aren’t just helpful but the personalized guidance really helped me see things clearly and grab opportunities
Hi, daily timeframe, the script as is is just for backtesting, but you can translate it to MQL language to run it in mt4 or pinescript is you want to run it in tradingview.
Thank you for a video it will be second strstegy for my tradnig bot. However I write mine with TypeScript. Python is too hard for me to maintain - simple thing like importing in python is overkilled.
Hey just a question . What do you think about option trading? Have you ever tried that? Also for metatrader 5 do you know any good broker for us residents?
Hi, I haven't traded options myself but I have friends who did and it's a specific mindset, you need to read related books and check tutorials for this.
It's a good question, I am sure it's reliable because adding a fee percentage is not complicated but most importantly you need to read their documentation to know what and how exactly are they treating the data in the background, sometimes it's good enough just to know.
26% profit after 7 years with 7% max drowdown is pretty horrible. And you didn't calculate fees for the 211 trades. I mean in a savings account you make 4% p.a. right now risk free.
Hi, fees are accounted for in the backtest (commission parameter in the code), usually when using the daily timeframe fees effect is minimal because the price range is relatively large so the trade covers the fees and spread comfortably. I agree that the profits are limited but this is one pattern among others, this type of trading needs to identify let's say 5 patterns and run simultaneously on different assets to increase the signal frequency. I hope this clarifies a bit the potential of the indicator.
No worries, I don't think it would work on smaller timeframes, the candles patterns are usually stronger/more certain on daily. We could try it though.
Not specifically, but I do have ML videos using models for trading check this one for example : ruclips.net/video/tNFnACpzxVw/видео.htmlsi=VmY75U2TC9eQ0nc0
Hi, this is taken into account in the commission parameter, the average spread is included in the backtest, and usually it's not an issue when backtesting on daily timeframe because the price range is relatively large, it does however make a difference if we are running on minutes timeframe for sure.
Hi, great video! I’m really intrigued by the 'Simple Candles Strategy' that achieved a 90% win rate with a 7-year automated backtest as presented by Larry Williams. If we were to implement this strategy on the stock market with random stocks, do you think it would produce the same 90% win rate, or would the results vary significantly based on the asset or market conditions? Would love to hear your thoughts on this!
Hi, thank you. In my opinion it might fail on different assets, I haven't seen any patter/strategy that works on all the assets, this is usually part of the robustness testing where we test the strategy on different assets to see if it still works. For this one (presented in this video) I don't think all the assets are good candidates, we could backtest it and run it on the optimal selection.
@@CodeTradingCafe Thanks for your input! I agree that not all assets are suitable for every strategy, and robustness testing is key. While I can’t run the backtest directly, you could test it on NVIDIA stock?
@CodeTradingCafe thank you, I am already into trading, but I want to automate and remove the human errors that comes from discretionary trading. It's killing me. Machines don't think, don't argue, don't hesitate etc.
If I look at your short setup entry graph, it feels like you are describing it wrong. It is not the same as the reference video's. The Buy setup entry is also wrong.
Hi, actually in the referred video if you check their description text there is a link where you can download a pdf where they provide more details, this is one of the variations they described I will publish the others as well soon. Numerically there is no difference between our code and the strategy described in their docs.
It's very interesting (especially considering its simplicity), but personally I'd never put any real money behind something like this. At just 0.06% average profit it will probably still result in a net loss (considering trading fees), and such a small percentage is well within the margin of error. Also with just 211 trades in back testing, the statistical significance doesn't seem super convincing to me (even with that high win-rate). In my personal experience, getting a high win-rate like this is relatively easy: Just be happy with many very small profits and don't care about occasional big losses that undo all those small profits (which those 10% of losing trades have a high chance to be). Not a criticism of this video mind you, which was I think very well presented (and actually addresses most of my concerns). It's certainly an interesting exercise, but don't put real money in an overly simplistic strategy like this!
@@CodeTradingCafe Ah, yes I should've been clear that my concern about fees was actually included in the test and addressed in the video. (Note to self: start writing after watching the whole video). Still, that average profit is so incredibly low that I think it's worth pointing out that it could easily be destroyed by market spread alone (taking down that high win-rate in the process), right?
@@CodeTradingCafe Thanks for your reply. I'm not sure I understand you correctly though. I mean spread fluctuates continuously, regardless of what timeframe the candlestick graph is in, right? I suppose what mainly matters there is what asset you trade in (i.e. trading random low-cap crypto meme-coin probably has higher risk of high spread than, say, trading some large-cap stock). Probably market conditions too. My personal algo-trading efforts (FWIW) are in crypto, where spread >0.06% is not uncommon.
Depends on how busy I am, some months I don't trade at all, sometimes I have 12 hours working days and trading is time consuming so I try to avoid it, last year I had to skip over 6 months of trading. It depends on the season I guess, but recently I have been less and less active (in trading I mean) other projects are taking the priority (less stress). But if you are looking for a number it's around 8-14% profit per year on average with some exceptions (reaching up to 24% per year).
Hello author, can you turn this indicator into multiple timeframe putting the logic exactly same? I mean it will plot 3 pivot lines in smae indicator based on different timeframe such as 3 minutes 5 minutes and 15 minutes. This one below 👇 "The next pivot by kiosefftrading"
Here is a system I tried with ninjascript , it worked nicely, I am improving. Basically I created an RSI indicator by subtracting a fast RSI from slow RSI. Fast RSI has half the period of slow RSI. This way you have a relative RSI that is sensitive to momentum changes. Then I used standard error band on this new RSI. Long when RSI crosses above the upper band, short when RSI crosses below the lower band. These crosses indicate momentum initiations. I use TTM squeeze to filter for better signals. I use NQ 133 tick data. 100s of trades a day, very high win rate.
Using fixed stop along with dynamic stop and target based on ATR trailing stop indicator
@@KEMALIST-1911 Sonuc nasil paylasa bilirmisin?. Ugrasmaya degerse mt5 kodlayip denemeyi dusunuyorum. Bu tarz stratejilerin genellikle sorunu spread oluyor.
Use stochastic RSI indicator instead how's your RSI that you customized to your liking. Instead This stochastic RSI is the proper version of what you were trying to do you want to set it 80 and 20 overbought oversold and when the lines cross either at the bottom or top. It's one of the best indicators you can use
Source code, please. @KEMALIST-1911
üstad 15 dakika ve 30 dakika 1 saatlik zaman diliminde rsi periodunu kac ayarlıyorsun..
Thanks for all the great work your doing for the community
Thank you for your support! Glad this is useful.
Pretty cool, you showed a comment from my friend, Vassilis, I know him for over 10 years )
Great Guy! I wonder where he is now, he's usually one of the first to leave comments in here :)
Hey, I mean the algorithmic trading, you are killin it. I can see we need to use indicators available to build a strategy, but I wanted to understand whether we can build a algorithm based on smc concept and our own entry types in that. Thank you.
Short answer definitely yes, we can code anything in python!
Yes you can ,i’m an Mql5/4 developer and i can build you your strategy if you are interested
This is the most underrated channel ever...
Thank you :)
Love this strat! The only bad thing about it is that the 250 pip SL does not work. It stops trading with 50-86 % fall from top.
I agree it's not a complete strategy yet, there is room for improvement.
@@CodeTradingCafe But its the best strat I have come across if I know a bullmarket is coming :)
I am a trader from chain, this is very very very useful and high success rate. excellent❤❤❤
Glad to hear that! Thank you for your comment.
In the last 6 months backtesting gives 58% winrate on exactly 50 trades, with a -18% balance loss overall.
It also depends on the assets, but this is just one candles pattern it can be added to other indicators in parallel.
This is the best video out there!! Thank you so much!! Really appreciate it!!!
Glad you enjoyed it!
Do you think this performance is really appliable in real life? As you mentioned, only selected stocks perform well on this setup, but when we work with the present, it looks difficult to guess what stock is going to work, even when a stock gives us a good pattern in the past, it doesnt garantee that the future will follow the same pattern
The best is to find let's 5 different patterns, and run them in parallel on let's say 10 different assets. We will always have loosing versus winning assets, but the candles patterns will make the difference in a way that the total is always in the positive returns side. This however requires more work to test lots of patterns and on lots of assets to make the selection. It's like any investment strategy you try to diversify all your parameters.
Great work Dr Ziad, very informative channel!
So nice of you :)
Thanks for the great video! Keep going on please! I have a question. If the agregated return is 186% in 7 years, then someone with 100€ would have 286€ after the 7 years of trading?
Yes that is correct, I know it's not impressive but this is how these algos work.
doesn't matter, if you know how to trade, people will practically be begging you to manage their money. its really easy to pass a prop firm challenge
and i dont say that to discourage you ! if you take the time and effort and really lock in theres nothing holding you back!!!!!!!
Had some success, but the key is understanding the market. Trading offers more potential than holding, especially with Trump’s policies changing things up. Charts alone won’t cut it, Honestly,James Whisman completely transformed my approach to trading. his insights aren’t just helpful but the personalized guidance really helped me see things clearly and grab opportunities
by the way, how i can execute exactly using this code for Metatrader 5? btw thx for that helpful information to the community!
Thank you! this code only runs on python it needs to be translated to MQL before running on MT5.
@@CodeTradingCafe thx mate
you can also use the mt5 library for python, but you can only operate it on a windows platform
@@mmustap3use AI to rewrite it to mql
In which TMF works this stratgy? And how to implement the script? And where must be implementet on mt4 or tradingview?
tradingview
indicator("turendu", overlay = true)
test = open[1]>close[1] and high[0]>high[1] and close[0]>low[1] and close[0]
Hi, daily timeframe, the script as is is just for backtesting, but you can translate it to MQL language to run it in mt4 or pinescript is you want to run it in tradingview.
@@marianosanchez7488 i am trying too finde those candelsticks but i cant find it ...
Thank you for a video it will be second strstegy for my tradnig bot. However I write mine with TypeScript. Python is too hard for me to maintain - simple thing like importing in python is overkilled.
Thank you for your support, good luck!
Hey just a question . What do you think about option trading? Have you ever tried that? Also for metatrader 5 do you know any good broker for us residents?
Hi, I haven't traded options myself but I have friends who did and it's a specific mindset, you need to read related books and check tutorials for this.
@@CodeTradingCafecould you ask them for resources that are worth to read?
I have question, is the backtest library reliable. Has anyone compared to the real time, assuming fixed spread and slippage, commission etc?
It's a good question, I am sure it's reliable because adding a fee percentage is not complicated but most importantly you need to read their documentation to know what and how exactly are they treating the data in the background, sometimes it's good enough just to know.
@CodeTradingCafe thanks🙏
Awesome. Any pine script for the same python code that will help
Not yet, you might want to try chatGPT for translating python into pine it might work.
Thanks very much for this informative video, can you please point me to the code link ?😊
Hi, yes the link is in the description of the video. Thank you for your support!
26% profit after 7 years with 7% max drowdown is pretty horrible. And you didn't calculate fees for the 211 trades. I mean in a savings account you make 4% p.a. right now risk free.
Hi, fees are accounted for in the backtest (commission parameter in the code), usually when using the daily timeframe fees effect is minimal because the price range is relatively large so the trade covers the fees and spread comfortably. I agree that the profits are limited but this is one pattern among others, this type of trading needs to identify let's say 5 patterns and run simultaneously on different assets to increase the signal frequency. I hope this clarifies a bit the potential of the indicator.
@@CodeTradingCafe Sorry mate, I overlooked the commission parameter in the code. I'm curios how the strategy performs on smaller timeframes.
No worries, I don't think it would work on smaller timeframes, the candles patterns are usually stronger/more certain on daily. We could try it though.
Do you have 500k in your savings account?
Did you forget about leverage?
Hello sir what platform do you use to trade so that I can use it for experimental purposes?
I trade with Oanda and Binance ( these are for algorithmic trading). Others offer options of trading as well though (IB...)
Can you do a vide using machine learning to optimise oversold overbought indicators for an asset?
Not specifically, but I do have ML videos using models for trading check this one for example : ruclips.net/video/tNFnACpzxVw/видео.htmlsi=VmY75U2TC9eQ0nc0
But in real market, you have to work on bid and ask, can you please incorporate it ?
Hi, this is taken into account in the commission parameter, the average spread is included in the backtest, and usually it's not an issue when backtesting on daily timeframe because the price range is relatively large, it does however make a difference if we are running on minutes timeframe for sure.
The best channel ''🙌🙌😎😎
Thank you !
could someone explain how to add this to the trading view? It's quite interesting, Thank you Code trading...
Hi, sorry from my side I don't code in pinescript but you can easily try to use chatGPT to translate this code to pinescript indicator.
Great video
Glad you enjoyed it! Thank you.
It's missing some information...
Hi, great video! I’m really intrigued by the 'Simple Candles Strategy' that achieved a 90% win rate with a 7-year automated backtest as presented by Larry Williams. If we were to implement this strategy on the stock market with random stocks, do you think it would produce the same 90% win rate, or would the results vary significantly based on the asset or market conditions? Would love to hear your thoughts on this!
Hi, thank you. In my opinion it might fail on different assets, I haven't seen any patter/strategy that works on all the assets, this is usually part of the robustness testing where we test the strategy on different assets to see if it still works. For this one (presented in this video) I don't think all the assets are good candidates, we could backtest it and run it on the optimal selection.
@@CodeTradingCafe Thanks for your input! I agree that not all assets are suitable for every strategy, and robustness testing is key. While I can’t run the backtest directly, you could test it on NVIDIA stock?
How we can get data folder for back testing ?
There is a link in the description of the video.
Thank you sir..Always awaiting for your new video....
Always welcome! Thank you for your support.
Is it Possible to implement ICT concepts in trading with algo way
Definitely Yes!
@ can you please make one video on this
might take some time to prepare though.
@@CodeTradingCafe ok take your time. Eagerly waiting
Keep it up!
That's the plan! Thank you :)
Awesome Dr Ziad, we just found your channel and it's very inspiring!!
Busted :)
What do I need to learn to be able to write these programs?
First you need to learn Python language, then some data analysis skills, and finally dive into the trading part.
@CodeTradingCafe thank you, I am already into trading, but I want to automate and remove the human errors that comes from discretionary trading. It's killing me. Machines don't think, don't argue, don't hesitate etc.
That's good then, you can go for python as a good start.
Impressive!
The winrate yes, the returns maybe not that much but still positive.
Did backtest included tradeing fee?
Yes fees are included, with the daily timeframe fees are not an issue.
How to setup can you pls guide
This is just a backtest, you can rewrite for live trading or live signals.
But can you guide me ?
I don't want to rewrite, when your strategy is win rate is above 80 percent
How to get in contact with you? What is the other pair?
You can't ... I will find you :) ... joking I can't commit to additional projects at the moment.
Hi,
Can we make indicator on pine script for simple candles strategy ?
Yeah use gpt it will make it for you
@hardkoz6243 thank you
Indicator is just a tool, better to use SMC
ChatGPT is a good idea to translate into pinescript, let us know how it goes for you :)
감사합니다
Thank you for your nice comment!
If I look at your short setup entry graph, it feels like you are describing it wrong. It is not the same as the reference video's. The Buy setup entry is also wrong.
Spot on they got that completely wrong wonder if they back tested the wrong strategy
Hi, actually in the referred video if you check their description text there is a link where you can download a pdf where they provide more details, this is one of the variations they described I will publish the others as well soon. Numerically there is no difference between our code and the strategy described in their docs.
what this candlestick pattern called as ??
I don't really know just took it from the book as is.
@CodeTradingCafe ok ty...
Which time frame?
Daily the safest
What is the strategy name?
I don't recap it has a specific name, just a candles pattern.
Can it be tested with crypto
Ofcourse it cant, its forbidden.
@@holdem235ahahahah 🤣
@@holdem235 mmmh ok
It can be tested on any chart regardless the asset... it doesn't mean it will always work though.
@@CodeTradingCafe Worth a try. Thanks
Name of the candles
So many names I wouldn't remember :)
What timeframe?
Daily my favorite.
@@CodeTradingCafe thanks
It's very interesting (especially considering its simplicity), but personally I'd never put any real money behind something like this.
At just 0.06% average profit it will probably still result in a net loss (considering trading fees), and such a small percentage is well within the margin of error.
Also with just 211 trades in back testing, the statistical significance doesn't seem super convincing to me (even with that high win-rate).
In my personal experience, getting a high win-rate like this is relatively easy: Just be happy with many very small profits and don't care about occasional big losses that undo all those small profits (which those 10% of losing trades have a high chance to be).
Not a criticism of this video mind you, which was I think very well presented (and actually addresses most of my concerns). It's certainly an interesting exercise, but don't put real money in an overly simplistic strategy like this!
Hi thank you I agree on the stats part and definitely this is not a complete strategy, just to be precise the fees were included in the backtest.
@@CodeTradingCafe Ah, yes I should've been clear that my concern about fees was actually included in the test and addressed in the video. (Note to self: start writing after watching the whole video).
Still, that average profit is so incredibly low that I think it's worth pointing out that it could easily be destroyed by market spread alone (taking down that high win-rate in the process), right?
Not on the daily timeframe, usually the price range covers spread and fees very comfortably, but if you use lower timeframes the risk is higher.
@@CodeTradingCafe Thanks for your reply. I'm not sure I understand you correctly though. I mean spread fluctuates continuously, regardless of what timeframe the candlestick graph is in, right? I suppose what mainly matters there is what asset you trade in (i.e. trading random low-cap crypto meme-coin probably has higher risk of high spread than, say, trading some large-cap stock). Probably market conditions too.
My personal algo-trading efforts (FWIW) are in crypto, where spread >0.06% is not uncommon.
excelente ❤
Thank you !
So basically 30 trades a year over the 7 years?
Yes, the idea is to run let's say 5 patterns in parallel and do we can get 5 times that.
Yes, the idea is to run let's say 5 patterns in parallel and do we can get 5 times that.
Hello sir please create the same statargy for Mt5
Hi, I am not using MQL, but you can easily use chatGPT yo translate this code for MQL5
Hi, not using MT for a while
Thank you!
The second indicator looks to me like a trailing stop loss
No trailing, we just fixed the SL in relation to the entry position.
Ea name?
No name so far :)
Curious to know what your live portfolio is like if you would like to share!
Depends on how busy I am, some months I don't trade at all, sometimes I have 12 hours working days and trading is time consuming so I try to avoid it, last year I had to skip over 6 months of trading. It depends on the season I guess, but recently I have been less and less active (in trading I mean) other projects are taking the priority (less stress).
But if you are looking for a number it's around 8-14% profit per year on average with some exceptions (reaching up to 24% per year).
Hello author, can you turn this indicator into multiple timeframe putting the logic exactly same? I mean it will plot 3 pivot lines in smae indicator based on different timeframe such as 3 minutes 5 minutes and 15 minutes. This one below 👇
"The next pivot by kiosefftrading"
I will have to look into the details.
@@CodeTradingCafe Can we discuss on messenger or WhatsApp?
Wow ❤
Thank you :)
intresting
Simple is always interesting :)
can't you give us the .py files too please?
Hi, I didn't use any .py file it's just the Jupyter notebook that was presented in the video.
Another useless high win rate strategy.
But the numbers are there.