My Experience Building a Python Stock Trading Bot
I built an automated cryptocurrency trading bot for Upbit using pyupbit.
At first, I thought running the bot would make me some money...
- Week 1: +8% return (got excited and bragged to friends)
- Week 2: -12% (panic sold)
- Week 3: Volatility was too high, so I just turned the bot off
Conclusion: You need to thoroughly backtest your quant strategies, and bots are poison in highly volatile markets.
Now I just buy ETFs for long-term investing lol
5 answers
Agreed. Not backtesting is basically a recipe for disaster.
I've had a similar experience lol. I was building a bot with pyupbit and was happy when it hit +15% in the first month, but then it suddenly dropped -20% and I shut it down right away. Now I just stick to dollar ETFs.
But isn't that old news? I heard the volatility on Upbit has decreased recently, making it harder to run bots.
It's cute that you're boasting about an 8% return lol. Anyway, good work. The strategy definitely needs to be verified, but it's a good lesson for beginners.
This is a critical opinion, but isn't the issue more about the strategy being too simple rather than the bot itself? When running it in a volatile market, you should add more conditions. Still, thanks for sharing your real-world experience.