Using AI For Stock Trading

Good luck buddy
Are you using this in crypto or shares ?
my goal is to get 1-3% of my capital each day it that possible ?

This is strictly US Stock market.

The number 1 stock trader in the world "George Soros" (imo)

Has an average return of 20% a year over his 40+ years of trading.

That's 0.05% a day.

Your goal is unrealistic.
 
This is just gambling. Unless you have the software the quant funds have, you're not gonna make any money doing this. You will lose your ass.
It's not even gambling. Gambling risks real money. This is delusion. Systems that make money on paper are a dime a dozen. When you have skin in the game, it's a whole new ball game. But you're right about the funds. On the other side of his trades - if he ever makes them for real - are Market Makers and fund managers with a trillion times the human and AI brain power that he has.
 
This is just gambling. Unless you have the software the quant funds have, you're not gonna make any money doing this. You will lose your ass.

Doing good so far.

It's not even gambling. Gambling risks real money. This is delusion. Systems that make money on paper are a dime a dozen. When you have skin in the game, it's a whole new ball game. But you're right about the funds. On the other side of his trades - if he ever makes them for real - are Market Makers and fund managers with a trillion times the human and AI brain power that he has.
Don't need to prove anything to you, but I will anyways.

Account.png
 
Exactly 2 Months in and so far are my results:

P&L: +14.91%
Trading Record: 16 - 12
Highest Winner: +$409.XX
Highest Loser: -$407.XX
Trading Days: 60 days
 
I think it’s okay just to implement some repetitive operations, but to allow it to automatically analyze the entire stock market requires a very powerful algorithm and is prone to errors.
 
As stated I'm sure it's just a lucky first few weeks.

Also I only use roughly 3% per trade of my total portfolio.

So if you do the math let's assume I have $10k balance:
- 3% of portfolio, which is $300
- 15.77% of that $300, which is $47.31 average per trade.
- 5 Trades so 5*$47.31, which is $236.55
- $236.55 / $10,000, which is 0.023655
- Convert to percentage looking at about 2.36% actual gain

Hardly going to make me rich, not to mention capital gains tax here in the states.
interesting way to size positions.. is there any reason you're doing it that way? what is your rationale? simplicity?

i size dynamically according to total capital and risk/stop. usually 1% of account risk. say my entry price is $5 and my stop is $4.50. this gives me -10% drawdown until i am stopped out. i have $10k capital. how many shares to buy where if i experience that -10% drawdown i only risk 1% of the account? 200 shares, which would be deploying 10% of my capital to the trade. i also have a hard limit on that % allocation so i don't size too big (can happen if stop is very tight, eg: if below i had my stop at $4.95 it would have put the full $10k in with only -1% drawdown max). example:

1722718810024.png

i would be curious to hear what kind of data the LLM is being fed, but you don't seem to know that. i would be careful to not blindly trust a system you have zero visibility into. i have not yet tried to use LLM for trade signals.. currently only using it to summarize/analyze text (PR, news, earnings calls, etc). when i am finished my market breadth project i will attempt to feed those data reports into LLM to tell me what parts of the market are doing well, which are doing shit, and which are flat (i believe taking trade signals in parts of the market that are doing well will improve probability of win).

anyways good luck with your journey, i am getting back to mine: trying to figure out how to get 1,000 robinhood accounts to scale my "free money" method. having tested since march i know it works, but trying to scale it is getting rather complex! :eek:

1722719106508.png

have a great weekend bhw!
 
Good luck buddy
Are you using this in crypto or shares ?
my goal is to get 1-3% of my capital each day it that possible ?
1% a day is NOT happening. We can prove this with numbers.
Grab a calculator and start with $1,000 then compound a 1% gain every day.
After a few short years the calculator runs out of numbers, and you are richer than Elon Musk.

This is just gambling. Unless you have the software the quant funds have, you're not gonna make any money doing this. You will lose your ass.
You don't need the software quant funds have; you can also just follow what they do and trade using sentiment.
Plus, bots are available to the general public, last year I made 42% using ones from AlgoDatabase.

Most likely you'll need to self-fund though as the drawdowns are too high for so-called "prop firms".
I also make my own which is super simple with TradingView.
Then have the alert system send a webhook and it'll auto trade; or you can send the alerts to Discord/Telegram if you wanted.
 
interesting way to size positions.. is there any reason you're doing it that way? what is your rationale? simplicity?

i size dynamically according to total capital and risk/stop. usually 1% of account risk. say my entry price is $5 and my stop is $4.50. this gives me -10% drawdown until i am stopped out. i have $10k capital. how many shares to buy where if i experience that -10% drawdown i only risk 1% of the account? 200 shares, which would be deploying 10% of my capital to the trade. i also have a hard limit on that % allocation so i don't size too big (can happen if stop is very tight, eg: if below i had my stop at $4.95 it would have put the full $10k in with only -1% drawdown max). example:

View attachment 364701

i would be curious to hear what kind of data the LLM is being fed, but you don't seem to know that. i would be careful to not blindly trust a system you have zero visibility into. i have not yet tried to use LLM for trade signals.. currently only using it to summarize/analyze text (PR, news, earnings calls, etc). when i am finished my market breadth project i will attempt to feed those data reports into LLM to tell me what parts of the market are doing well, which are doing shit, and which are flat (i believe taking trade signals in parts of the market that are doing well will improve probability of win).

anyways good luck with your journey, i am getting back to mine: trying to figure out how to get 1,000 robinhood accounts to scale my "free money" method. having tested since march i know it works, but trying to scale it is getting rather complex! :eek:

View attachment 364704

have a great weekend bhw!

It isn't always 3% it is variable sizing based on the confidents of the trade. Ranges from 10% to 1.5% sizing.
I don't know exactly how the AI works that is correct. I have been told it is more technical based.

I am pretty much blindly following the advice it sends, however I am pretty confident with the guys doing the strategizing of it.

Also another update here:

P&L: -0.84%
Trading Record: 18 - 25
Highest Winner: +$625.XX
Highest Loser: -$407.XX
Trading Days: 109 days
 
I have no clue what's going on in the stock market world right now.

P&L: +3.53%
Trading Record: 20 - 26
Highest Winner: +$625.XX
Highest Loser: -$407.XX
Trading Days: 117 days
 
I have no clue what's going on in the stock market world right now.

P&L: +3.53%
Trading Record: 20 - 26
Highest Winner: +$625.XX
Highest Loser: -$407.XX
Trading Days: 117 days
To me, it's setting up for a crash... Watch NVDA as they have a pretty big pull and high volatility.
 
P&L: +14.76%
Trading Record: 30 - 26
Highest Winner: +$625.XX
Highest Loser: -$407.XX
Trading Days: 129 days

Had a flawless last 2 weeks!
 
P&L: +21.76%
Trading Record: 34 - 28
Highest Winner: +$625.XX
Highest Loser: -$407.XX
Trading Days: 137 days

Most of the wins are from yesterday. And still holding some unrealized gains from yesterday's fed run.
 
P&L: +2.76%
Trading Record: 35 - 32
Highest Winner: +$625.XX
Highest Loser: -$832.XX
Trading Days: 144 days

Took a massive L this week losing nearly 20% of my P&L.
 
Back
Top