Attention Programmers

ElectricSavant

ElectricSavant

Well? Have you done it? Did you crack the code? There are many engineers fascinated with this quest.

Do you just watch re-runs of Star Wars all day? What...yeAH?

CAN YOU TURN YOUR AUTOMATED TRADER ON AND IN TEN YEARS PICK UP YOUR PROFIT?

ES
 
ElectricSavant said:
CAN YOU TURN YOUR AUTOMATED TRADER ON AND IN TEN YEARS PICK UP YOUR PROFIT?
More...

You can definitely turn on your automated system. But unless you doing some sort of almost risk free trading you have cannot predict the future results. You can only hope your edge continues to remain profitable in the future like it did in the past (or less profitable than it did in the past but still good enough).

I suppose your typical engineer ATS newbie is looking for an ATS that can make say 100% a year system with a stdev of 10% (sharpe of 10.0), or double or triple those numbers when using leverage. Unless they can do HFT they are not going to find anything close to that.
 
Last edited:
d08 said:
I'm at the point now where I can leave things running and often have, however I prefer to keep an eye on it. I absolutely hated organizing my life around market hours, it gets even worse when trading multiple timezones - manually it would be impossible.
More...

Are you using a custom built platform?
 
traider said:
Are you using a custom built platform?
More...

You can call it that but it has no GUI (which I don't need), for many that's not a platform. I'm using Python and IB API. I'd never want to be tied to a commercial closed-sourced application because when the support stops or they're sold, you're usually screwed - has happened before.
 
ElectricSavant said:
Well? Have you done it? Did you crack the code? There are many engineers fascinated with this quest.

Do you just watch re-runs of Star Wars all day? What...yeAH?

CAN YOU TURN YOUR AUTOMATED TRADER ON AND IN TEN YEARS PICK UP YOUR PROFIT?

ES
More...
There is a concept/technique in software development called CICD (Continuous Integration Continuous Deployment). An example of a CIDD pipeline would be a developer checks code into a development source control branch, another developer reviews it and when that code is merged into the master branch, a chain reaction is initiated where code is deployed to a test server and automated tests are run. If the tests pass, the new code is deployed to production. In practice, it might be partial automation where a human has to push a button to deploy to production.

Back to trading...I thought it would be cool if you could add a new indicator or piece of data and could categorize it and the platform would "know" how to test it. The platform would also regularly test existing systems and weed out ones that were no longer performing and add new strategies that it discovered and were performing. I intend on having a separate platform for live trading so a new strategy would have to be deployed to that environment, backtested and then the results compared against the platform that discovered the system. I suspect some very large hedge funds have something like this set up. In practice, for someone without a yearly million dollar plus tech budget, the best thing to do is to find the "low hanging fruit" that can easily be automated.
 
fan27...gets it (very fast)...I do not even have to post anymore from my seed post.

Awesome.

ES

P.S. If i had enough money how much would software like that cost? Every year there is a new harvest and the low hanging fruit changes. I need automation that optimizes after searching for the best set up and then calibrates and trades accordingly. It can change like a chameleon and no problem I can press a button or two. fan27 nailed it. can anybody have this? I am trading Futures.
 
Last edited:
Traders...

Would you pay a million dollars for software like fan27 explained? (CICD for Trading)

ES


Back to trading...I thought it would be cool if you could add a new indicator or piece of data and could categorize it and the platform would "know" how to test it. The platform would also regularly test existing systems and weed out ones that were no longer performing and add new strategies that it discovered and were performing. I intend on having a separate platform for live trading so a new strategy would have to be deployed to that environment, backtested and then the results compared against the platform that discovered the system. I suspect some very large hedge funds have something like this set up. In practice, for someone without a yearly million dollar plus tech budget, the best thing to do is to find the "low hanging fruit" that can easily be automated.
 
Let's just pause a minute and think " how much would I pay for a gold plated profitable EA running on my machine " ?
OK let's not do this in actual $$$s as everyone has different resources. So let's do it in percentages of your trading fund. Would 70% seem reasonable ? Leaving the other 30% for actual dealing.
 
You must log in or register to reply here.
Back
Top