Collaborative approach...

tere

tere

Hi,

Have been developing datamining/backtesting/trading related algorithms for a while now...

Lately started to look into collaborative approaches and found it hard to find people with decent trading related insight while outsourcing some tedious aspects.
Also specialization would take time and effort.
At the end of the day there might not be as much motivation present when working for others gain and you are left only with project/component based or montly pay.

So i am hoping someone on similar journey happens to read and would consider collaboration to achieve benefit for both parties.

My background:
I am located in balitics. GMT+3 zone.
My english is not perfect..

Writing mainly C++ ,and most aspects are developed from ground up, in order to have maximal flexibility.

The project started as a openGL market player...tool to improve my manual trading via speed up market. Soon i realized i might not handle trading lifestyle and developing automation seemed easyer at first.


So far i have various tools worked out , in order to automate various tasks involving strategy testing.
Historical data access, data conversion, data compression, data filtering, historical simulation , graphing and market connections etc.
Not much more that i would need, but everything could be improved further.
Most of my time has been in rewriting existing code.

My coding style for performance critical components is rather interconnected and not following OOP guidelines.
Most parts of codebase would be harder to root out compared to more modular designs.


If anyone is interested in working out common grounds for future goals and collaboration. Let me know.

Perhaps developing data filtering algo using ideas and experinece of multiple people?
Or shared endpoint for market access to win on trading costs?

Il consider whatever is your missing piece or idea.
 
Last edited:
i highly doubt you can find a partner who is competent With mutual trust. Trading is a lonely journey. I have been coding my strategy for 10 years，I did not seek someone’ s help. First if he is not expert I have to teach him, which wastes time； if he is good I do not trust to disclose my strategy and my reasoning.
 
  • Like
Reactions: d08
trend2009 said:
i highly doubt you can find a partner who is competent With mutual trust. Trading is a lonely journey. I have been coding my strategy for 10 years，I did not seek someone’ s help. First if he is not expert I have to teach him, which wastes time； if he is good I do not trust to disclose my strategy and my reasoning.
More...
Dont worry about trust, its not like someone is going to throw their strategic sources.

But trading source components surrounding testing environment for strategies would be an option.

My idea is more about finding common goals for developing surrounding tools .
Not so much on sharing strategies.

Id assume 90%+ code is testing tools anyway and core components of strategy might be much smaller proportion.

In 10 years you probably already have decently worked out many of the the tools you need, and could have less to gain.
Your advice and experience could be valuable.
Id be interested in your failed strategies for example...

One thing to consider is that various people see things differently, and dif perspectives can be used for generating new ideas.
 
Last edited:
tere said:
Hi,

Have been developing datamining/backtesting/trading related algorithms for a while now...

Lately started to look into collaborative approaches and found it hard to find people with decent trading related insight while outsourcing some tedious aspects.
Also specialization would take time and effort.
At the end of the day there might not be as much motivation present when working for others gain and you are left only with project/component based or montly pay.

So i am hoping someone on similar journey happens to read and would consider collaboration to achieve benefit for both parties.

My background:
I am located in balitics. GMT+3 zone.
My english is not perfect..

Writing mainly C++ ,and most aspects are developed from ground up, in order to have maximal flexibility.

The project started as a openGL market player...tool to improve my manual trading via speed up market. Soon i realized i might not handle trading lifestyle and developing automation seemed easyer at first.


So far i have various tools worked out , in order to automate various tasks involving strategy testing.
Historical data access, data conversion, data compression, data filtering, historical simulation , graphing and market connections etc.
Not much more that i would need, but everything could be improved further.
Most of my time has been in rewriting existing code.

My coding style for performance critical components is rather interconnected and not following OOP guidelines.
Most parts of codebase would be harder to root out compared to more modular designs.


If anyone is interested in working out common grounds for future goals and collaboration. Let me know.

Perhaps developing data filtering algo using ideas and experinece of multiple people?
Or shared endpoint for market access to win on trading costs?

Il consider whatever is your missing piece or idea.
More...
Post your questions in the "Tools Of The Trade" and you will get answers to any technical questions you need help on.

I received all the helps I needed in software/hardware/coding by posting my questions there. The number of expert techies on ET is amazing.
 
You must log in or register to reply here.
Back
Top