Image not available

513x513

Ai0c4pp.jpg

๐Ÿงต algorithmic trading

Anonymous No. 16294309

suggest me some interesting concepts to implement in trading algorithms
Hard mode: no common bs, no cringe random walks, no coin flips, amaze me

Anonymous No. 16294378

if RSI(6) drops below 5, buy.
if the RSI(6) pump is above 90, market sell.

Anonymous No. 16294534

>>16294378
nigger

Anonymous No. 16294650

Biden is a piece of shit

Anonymous No. 16294714

>>16294309
no point in questioning these autistic twelve year olds OP, they don't even know how to tie their own shoes

Anonymous No. 16294718

>>16294309
stock = 5$, buy
wait X amount of time until stock = 6$, sell
keep 1 dollar
buy another stock for 5$
wait X amount of time until stock = 6$, sell
keep 1 dollar
repeat

Anonymous No. 16294760

>>16294718
be born
grow up to read this message
kys
repeat

Anonymous No. 16294787

moving average are key

Anonymous No. 16294848

>>16294309
I have a marvelous algorithm but the margins of this textbox are too small to contain it.

Image not available

1496x1182

1720927562600955.png

Anonymous No. 16294851

>>16294309
do a flip faggot

Anonymous No. 16294860

>>16294309
Use this
https://global-mind.org/gcpdot/

Anonymous No. 16294905

>>16294378
This is literally the golden formula that made me 7 figures. Just do this anon, everything else is cringe

Anonymous No. 16294934

>>16294905
kys nigger

Anonymous No. 16295015

>>16294309
WHERE THE FUCK IS BIDEN?

Anonymous No. 16295363

this fucking board is completely useless, kek

Anonymous No. 16295375

>>16295363
>Start a retarded thread
>Get retarded answers
Impossible to predict, just like the stock market.

Anonymous No. 16295412

>>16295375
>Entirely legit thread
>Retarded answers from obnoxious losers

Anonymous No. 16296018

>>16294309
Look into time series analysis. I believe deep learning still doesn't perform that well and is very expensive. Tree based methods like XGBoost are basically SOTA.

I'd opt for short to medium term trading, like 1 to 5 days. Create tabular data such as a stock open price, then the target will be the price 5 days out. Add in as many useful features as possible incorporating technical indicators, fundamentals, sentiment analysis, etc. Avoid data leakage when training, especially in time series forecasting where you train on future time points and predict on past time points.

That said, you'll most likely fail to make money and the big boys have more brains and resources than you.

In my experience the best strategy has been to buy and hold a few things you're solid about. I bought bitcoin 10 years ago, and generally do dollar cost averaging mostly in tech stocks.

Anonymous No. 16296394

>>16294309
Follow this putorial:
https://youtu.be/fhwoIQTKCyE
Warning: being filthy rich is not for everyone.

Anonymous No. 16296461

>>16294309
Have you tried buying low and selling high?

Anonymous No. 16296465

>>16296461
It can be arranged, but this process costs a lot of money and you will want more money than you do now, after, which thus means you need to ask your boss.

Anonymous No. 16296977

>>16294309
Martingale every time downside is realized

Anonymous No. 16297093

>>16296977
Isn't profitable on crypto, I've been simulating that.

Anonymous No. 16297326

>>16296977
This only "works" if you have deep pockets and a very liquid market. Still, it's statistically possible to get 10 heads in a row. If you started with an initial bet of 100, you'd need about 200k on hand to recoup the loss. Assuming even chances, this should only occur like 0.1% of the time, but you never want to risk getting wiped out. For building long-term wealth preventing losses is far more important than securing gains.

Image not available

720x886

1721988993326189.jpg

Anonymous No. 16297586

You model the stock price as brownian motion and then do a MANOVA on indexe funds with their stock prices as cofactors. Why funds? Because they have nearly guarantees yields which means they are robust tp noise. Then you do monte carlo simulation of the strongest interacting cofactors to determine which stocks are likely part of a loop (like if company A goes down B goes up and then C goes down which causes A to go up) and eliminate those. With the remains you literally just run a GLM on the remaining stocks and lasso away the less significant and bet on the remains. So if the price of index I is down then that means stocks X and Y and so on are the culprits, so buy those at low and then trade when the index goes up again.

Anonymous No. 16297605

>>16296977
Yes... sure...
>>16297326
This just doesn't work

Image not available

1026x960

9.jpg

Anonymous No. 16297869

>>16297586

Anonymous No. 16298097

>>16297869
>levy processes
one day man, one day
im getting to the bottom

Anonymous No. 16298136

>>16298097
what's that?