hello all

wordd

wordd

Member
Been lightly trading stocks/options for 10 years or so, passively making money on indexes and actively losing it whenever I do things myself. Many times because I'm either asleep or in a meeting whenever the action happens... and the rest of the time I'm just wrong.

@jack — I have to commend you for all the work you do. Of all things I've learned in my life, algo/trading is by far the most secretive and difficult. Most of the information out there is bs and the rest is wrong. I understand and agree with why but holy hell it's hard to start from the ground level.

Your posts + others on this forum have been the most help in laying some of the groundwork.. so, thanks.

edit: am about 6 months into my algotrading journey.. building services and whatnot and have done a few test trades. I've got it to work but haven't made any money yet and am on rewrite #2. Probably another 12 months of work ahead of me at least.
 
Thanks for the kind words! :)

I will say this: It's going to be really easy to keep burning time on code that is nice to have, but not absolutely required.

As much as it might feel like "so much more needs to be done" try to remind yourself: "Not every job worth doing, is worth doing well."

A common trap when coding your first algo or system is to avoid going live though adding more features and constantly being in a 'almost done' state.

Just focus on what's immediately practical.. hammer it out.. dirty.. don't worry about making a fancy interface (you're the primary user and you know what it does..) just get it working.

You have no idea how much momentum you'll pick up once you have something, even if it's basic, that trades for you.

My very first algo platform ran without a GUI, entirely in the command line.. that means I had a separate CMD window open for every strategy I was running, plus one extra CMD window open for the quote and execution service (they all talked to each other via IPC..)

It was ugly as hell... but it worked.

Once I had a dozen+ strategies / windows running at the same time, and managing windows became a bother, only THEN I work on an interface to manage the whole operation within a single app...

The takeaway is this: """great artists ship""" ...get something running in production.. even if it's ugly. Do this, and experience wise you'll be ahead of the overwhelming majority of novice algo traders (most of which will spend years building a backtesting engines or playing with libraries, but never get a system running in production.)
 
Last edited:
You must log in or register to reply here.
Back
Top