I have built something pretty cool over at qntlab.app

Please have fun with it and lets me know what you would like improved or if there are bugs.

Please note the backtesting results and visuals are DEMO but the code generation is real. You must export & run the strategy in a quantiac environment to test against real data.

I am not affiliated with Quantiacs Q24 contest nor do I represent them in any way. I am just having fun with code and would like to share with this community.

If anyone would like to work on this with me please reach out to @Svyable on x or reddit etc.

I'm currently running some long-term optimizations on my strategies (using genetic algorithms, so it takes forever to converge). I find myself staring at the progress bar way too often.

What do you guys do to kill time during these periods? I've started getting into some casual gaming to keep my mind off the charts.
Currently revisiting some classics and looking forward to Hytale Game (really hyped for the sandbox creativity). Also found myself clicking away on Cookie Clicker 2 when I need a mindless break, or checking random browser games on Funboxie.

Do you have any other recommendations for "low commitment" activities that don't distract too much from the coding mindset? Or should I just be reading more papers?

Cheers!

I am new to building strategies using ML/DL on Quantiacs and am very impressed with the OS performance of your ML strategies. I hope you can give me your contact (mail, limkedin,...) so I can learn from your experience in building an ML/DL strategy.

Sincerely thank.

Please note that according to the rules at https://quantiacs.com/contest/19

A Trading System will be deemed to be a “unique“ Trading System if it was not submitted by the same user to a previous Contest and it was not published by the Sponsor itself and it was not submitted by another user to a previous Contest or to the current Contest. The Sponsor will run on submissions a correlation filter and will have to right to disqualify submissions which are not deemed to be unique.

So re-submitting the same system will result into a system which is not eligible for a prize.

Spyder should be run under conda environment

an alternative way is to clone the library from https://github.com/quantiacs/toolbox
and develop strategies inside qnt. But I recommend using the approach from the documentation.

You can simply create a new environment:

conda create -n qntdev quantiacs-source::qnt conda-forge::ta-lib conda-forge::pandas==1.2.5 conda-forge::dash==1.18 python==3.7

The basic reason is that xarray works really well with array with an arbitrary number of dimensions. For financial data we have normally 3 dimensions:

and xarray works really well with cross-sectional indicators ranging over several assets.

However, for speeding up the evaluation, we are using a simplified model based on the gaps, it fits well with historical data.

Clearly slippage will be a function of the allocated USD capital to a strategy.

Other strategies have been funded, and agreements are in place between quantiacs, investors and quants. We cannot disclose more details now, sorry.

5M USD is a reasonable capacity a strategy could handle, yes.