Anyone on /biz/ algorithm trade? How does one get started with this? Any resources or courses to recommend?
>>1320704
Are you the best at math in you country? yeah didn't think so... you might want to look into some good quality kneepads.
>>1320767
Do, but I have an Asian that lives next door that very well could be. Assuming, as unlikely as it is, that the Asian is both the sharpest mathematician and cooperative, how would I go about this?
>>1320704
I gave it a try a while ago. A halfarsed try but a try none the less. I tried making an algorithm that traded the companies in the s&p500 with trades lasting for 3-10 days. I got mixed results during testing and i had other things to do so i stopped working on it.
Now for the advice part:
You can get free historic days ohlc data from yahoo. This was also the reason i tried making an algorithm that makes trades over multiple days.
I suggest you try to make an algorithm for day trading because there's less chance of news fucking your shit up.
Lastly managing risk is arguably more important than finding good trades
>>1321211
Thanks. Did you program it yourself? I know intermediate Java and C++, I get how to go about writing the algorithm, but I'm a bit lost how to code actual controls of live-monitoring of stocks and collection of data to process and how to code the actual purchasing and selling of stocks--say, if I was using Robinhood in an emulated environment.
>>1320704
I am an algo trader.
Here are the facts:
> You need a minimum of $30,000.
> Interactive Brokers is literally the only choice.
> It takes 3-6 years to get good at it if you're already a good programmer.
> You'll need to build your own software from scratch.
> Most of the hard work is developing strategies. This is where most people fail.
My advice:
> Read Robert Pardo's book.
> Don't make a "bot," build an Engine.
I'm looking at you, >>1321211
> Read everything you can find about trading. Build a library.
Most important:
> This is a solo mission.
You'll see what I mean if you stick with it.
>>1321338
Thanks anon
Can you recommend any other resources?
>>1321338
What about the programming and strictly computer science side of it all?
>>1321407
Go take a look at the Interactive Brokers API.
It's all very straight forward.
>>1321445
Thanks, I appreciate it anon
>>1321338
Do you work at a firm or do you run it yourself at home with your own money.
I'm a student doing quant work in chicago but my firm is mainly modeling long term trends, so I'd love to move to a more active role (algo trading, quant research)
>>1321338
You need a minimum of $30,000.
no
> Interactive Brokers is literally the only choice.
no
> It takes 3-6 years to get good at it if you're already a good programmer.
nobodys fucking generating algos from the scratch nowadays
> You'll need to build your own software from scratch.
HAHAHA
> Most of the hard work is developing strategies. This is where most people fail.
Orly?
>>1321559
Could you elaborate on what you mean? Is there framework software out there? My google skills be weak and frail.
What do you think of Quantopian?
>>1321559
>https://www.interactivebrokers.com/en/index.php?f=4745&p=daytrade
He actually does need $30k if he goes with Interactive Brokers. But I agree with you in that there are other (cheaper) options out there.
>>1320704
consult /g/