I Built A Stock Trading Bot And Gave It $30,000 To Trade

Introduction

Why does no one even like me, dude? Not even my dog... Oh, hey! How you doing? I know what you're thinking – this is not possible by any means. Guess what? You're wrong. I'm going to explain how we can make an AI trading bot that will make stock trades for me and hopefully, a whole lot of money. We're going to code it, but first, let me explain how this AI is actually going to work.

The Basics of the AI Trading Bot

We'll be using a programming language called Python. Don't worry too much about the technicalities. The main method we'll employ is called machine learning. Machine learning is basically like a human brain inside a computer. We're going to use it to make predictions on the trades we're doing.

Paper Trading: Our Safe Start

Let's address the cat in the room. Life isn't going so great right now – I only have $1.42 to my name. But, all is good because we can use a paper trading account. Paper trading is free and uses virtual money, which isn't real, but at least we can track the process. Whether you're $500 in the negative or positive, you can use my link below to sign up for Robin Hood and get a chance to win $1,000 in stocks.

Coding the AI Trading Bot

A project like this would take a lot of time if I started it from scratch. Instead, we'll be using open source code, which is basically code that people put on the internet for free. We'll tweak it to fit our needs. Without further ado, let's get coding!

Skimming Through the Code

The bot is complete after about 14 hours of work and a day to spare. We used open source code and modified it extensively to fit our needs. Now, let's quickly skim through the code.

Understanding the Model

If you don't know code, don't worry. I'll keep it beginner-friendly. For those who do, we're using a logistic regression model and libraries like NumPy and pandas. This model works by giving it data so it can understand patterns. We load stock prices from various dates and let the model make the best trading decisions.

Preparing the Data

We have many different stocks, including Apple from 2013 to 2018, Disney, IBM, JP Morgan, Microsoft, and Nike. The model uses this historical data to predict future trades, but it doesn't account for news or hype, just historical patterns.

Placing Trades

Now, let's place some trades based on the bot's suggestions. We'll use a paper trading account with $30,000, making three trades at $10,000 each.

Trade Recap

First Trade: Apple

We bought $10,000 worth of Apple shares on Monday at $137 per share and sold them on Wednesday at $141 per share, making a profit of $291.

Second Trade: Nike

We bought $10,000 worth of Nike shares on Tuesday at $128 per share but sold them at $126 per share, resulting in a loss of $158.

Third Trade: Caterpillar (CAT)

We bought $10,000 worth of CAT shares on Tuesday at $252 per share and sold them on Wednesday at $258 per share, making a profit of $258.

Results and Conclusion

In three days, we made three trades. Our trades were from Apple, to Nike, to CAT. We spent a total of $30,000 and sold for a combined total of $30,391, resulting in a profit of $391. This was all with fake money in a paper trading account.

Final Thoughts

If you want to see me put real money on this project, make sure to hit the like button. If this video gets 800 likes, I'll do it. The accuracy of the model is around 42%, so it should be a fun experiment.

Leave a like, subscribe, and share with your friends. I hope you enjoyed the video. Coding and modifying the project took a while, but it was worth it. Who knows? In the next video, I could actually make some money or lose a lot.

Peace out!