Looking for NinjaTrader users who are willing to share / swap strategy ideas

[burtoninlondon]

Hello,

I have been developing trading algorithms for about the last 2 years. I've seen success with some, others have been flat out losers and the remainder have floated in and out of profitability depending on the overall market context. I currently have 12 profitable operational algos.



I'm at the stage where I'm trying to diversify my portfolio so that I can spread my risk by increasing the number of uncorrelated algos. To achieve this I'm trying to find an online community where I can connect with likeminded people who would be willing to share/swap profitable algo strategy ideas. I am willing to share my strategy code with users in return for access to theirs, assuming they can demonstrate profitable performance.

If this is something you'd be interested in then please let me know, alternatively if you know of any online communities where this post would be better suited then let me know.

Thanks

Neil

p.s to NT staff - I hope the content of this post is appropriate, if not then please let me know.

[NinjaTrader_ChelseaB]

Hello Neil,

Asking the community for trading advice or logic advice is very appropriate for the forums.

This thread will remain open for any community members that would like to share their opinion.

[Fbraun378]

I think Algo's B,G,K,L are the most profitable long term. I haven't been doing algos for as long as you but I am currently working on developing a couple. Did you backtest using High resolution or standard mode (in Historical data)? I would definitely jump at the chance to coordinate trade ideas. LMK

[burtoninlondon]

Asking the community for trading advice or logic advice is very appropriate for the forums.

This thread will remain open for any community members that would like to share their opinion.

Thanks NinjaTrader_ChelseaB

[burtoninlondon]

Did you backtest using High resolution or standard mode (in Historical data)?

Fbraun378, I use standard mode. To be honest I haven't ever used High resolution mode but I've just reran some backtests using the High resolution mode and the results are largely the same.

What trading methodologies have you seen most success with? I have tried using indicators to generate signals but have not found a combination which works consistenly so my better strategies tend to be simpler and use price action as the signal. Having said that I do use a handful of volume related indicators (volume, vwap, volume moving average, etc.) and Laguerre RSI for momentum.

[Samario]

Hi, burtoninlondon

Thanks for opening this thread.

I've just started to develop strategies using Strategy Builder, not coding yet. I am now working on a strategy using Parabolic SAR indicator. I ran it for one month and it has been flat, I know that I need to test it longer, but I think by making a few changes I will be able to make a profit.

In these few months that I have been developing strategies what I learned is that the fewer indicators you use the better results. The simplest is the best.

Hope to get a better profit and share ideas and strategies with you and the community.

Thank you.
Jose

[Leeroy_Jenkins]

Results that have 20-30 trades belong in the trash bin. You can't draw any conclusions based on that small sample size. That's just one of the few mistakes you've made.

[Samario]

I know, Leeroy_Jenkins

I didn't have the time and resources to run it for a more long time.

I am working on having a server on the cloud to keep the strategies running so I can get to better conclusions.

Thanks.

[WattMan]

Hello,

I'm at the stage where I'm trying to diversify my portfolio so that I can spread my risk by increasing the number of uncorrelated algos. To achieve this I'm trying to find an online community where I can connect with likeminded people who would be willing to share/swap profitable algo strategy ideas. I am willing to share my strategy code with users in return for access to theirs, assuming they can demonstrate profitable performance.

Thanks for starting this thread; I am game for participating and sharing ideas - implementation IMO will differ for everyone per their risk appetite, asset preference etc. I am curious though to learn/understand what makes most sense and why when evaluating algo performance. Personally I started with the obvious Net profit, then preferred Profit factor and have recently started considering Win rate as well in the mix. Evaluating other performance metrics like MAE, MFE, and Ulcer at this stage. How do others approach and any context to your approach would be helpful to understand.

Background: My algo for ES delivered me a very nominal net profit (approx. $5K across all of 2022), but more importantly validated my idea and approach. Not a profitable trader by any measure. Now am looking to sharpen the same through this thread. Looking forward to hear more from others.

[sundae]

I have been increasingly enthused by developing strategies for less than a year now. When I trade manually, I am not profitable and let my emotions get the best of me. I have some good and some bad days. Thats why I started looking into semi or fully automated as a filter for my downfalls. I do have some strats that show potential and some that I just can't get to work no matter what I do. I look forward to sharing some ideas with you.

[WattMan]

I have been increasingly enthused by developing strategies for less than a year now. When I trade manually, I am not profitable and let my emotions get the best of me. I have some good and some bad days. Thats why I started looking into semi or fully automated as a filter for my downfalls. I do have some strats that show potential and some that I just can't get to work no matter what I do. I look forward to sharing some ideas with you.

Been there and for those exact reasons I am here as seriously my day-job/day-life will permit :-)

[jeronymite]

burtoninlondon Good backtest results are great to have. But there are backtests and there are backtests! And more importantly, good backtests do not in any way "guarantee" profitable trading in the live market. There are so many real-world market conditions that affect how trades play out, and it is very difficult to simulate many of them in backtest. Simple things like network latency, broker processing times, market volume and liquidity, real-time events that affect markets, unexpected real-time spread and margin changes, slippage, etc, etc.

Keep on developing and backtesting, definitely, and use the strategies with the best backtest results; but only real-world, live market trading will prove the profitability (or otherwise) of any strategy. If backtests really did correlate with actual profitability, there'd be many more happy traders than there are! {"Backtest billionaires" can be some of the least profitable real-world traders!)

Thanks.

[burtoninlondon]

burtoninlondon Good backtest results are great to have. But there are backtests and there are backtests! And more importantly, good backtests do not in any way "guarantee" profitable trading in the live market. There are so many real-world market conditions that affect how trades play out, and it is very difficult to simulate many of them in backtest. Simple things like network latency, broker processing times, market volume and liquidity, real-time events that affect markets, unexpected real-time spread and margin changes, slippage, etc, etc.

Keep on developing and backtesting, definitely, and use the strategies with the best backtest results; but only real-world, live market trading will prove the profitability (or otherwise) of any strategy. If backtests really did correlate with actual profitability, there'd be many more happy traders than there are! {"Backtest billionaires" can be some of the least profitable real-world traders!)

Thanks.

jeronymite, I couldn't agree more and I always test my strategies out in the sim and then if all ok on a live account with 1 share/contract to understand performance and assess against the backtested results. Thats what I was trying to demonstrate in my green columns (forward test) in the table I shared in my initial post where these trades have been made on a live account..

[burtoninlondon]

Results that have 20-30 trades belong in the trash bin. You can't draw any conclusions based on that small sample size. That's just one of the few mistakes you've made.

Hi Leeroy_Jenkins, thanks for your feedback and I agree with your QTY trades comment. For what it's worth the QTY of trades continues to increase as I continue my forward test.

[rtwave]

this thread could be the beginning of something interesting. i had not been able to reply before but i will share some information.



- over the time i have been a member of these fora i have seen practically no cooperation - collaboration between members. there are several fora on reddit where there is interesting content but unfortunately, like of most of the internet that website is a cesspool of aggression, garbage and imperialistic propaganda. nt is a great platform to develop and trade strategies and it would be great to have a sizable collective of proficient developers programming principally on nt and collaborating as much as possible.



- there's no need to risk real funds to learn whether a strategy is profitable or not. it is enough to leave 30% of the historical data out of sample when developing - optimizing and then evaluate the strategy over the entire expanse of data.


- regarding performance metrics, the most widely used are not really good. i would actually like to develop my own performance metric where strategies whose cumulative profits grew very close to a positive diagonal would have the highest scores and large losses would greatly decrease scores. it would be perfectly possible to program that into nt and i have thought of asking for help on fora like stckovrflw but i have never made it a priority. something like this:










when one has strategies like this above the issues and limitations with nt's optimization engine do become glaring.


- the key for results like those in the images above is to understand that the trajectories that the prices for financial garbage are made to follow can be treated like fractals. and strategies that assume every trade as if it was one more fractal can deliver results like these.


- furthermore, it is evident to me that fractals are best traded with manual - discretionary means. there are a number of individuals on twatter who focus on pivots, diagonals, continuation, reversals and such, and seemingly those are the best methodologies that exist.



very well, this is it for now for me, we'll see if participation ever increases in these nt fora. regards.