Descriptions:
You’ve seen the ads — someone posts a chart screenshot to ChatGPT and asks whether to go long or short. So I actually tested it. I fed 2 years of BTC/USDT 5-minute candle data to Claude, had it make a trading decision at each step based only on recent price history, and tracked every trade. No cherry-picking, no future data leakage. Just raw results.
726 trades. Here’s what happened.
What I tested:
— AI model (Claude) reading raw candle data and calling LONG / SHORT / FLAT
— Random coin flip on the same schedule (the real benchmark)
— Buy & hold over the same period
Support my work and get access to the biggest quant finance codebase (no brainer deal):
https://www.youtube.com/channel/UC87aeHqMrlR6ED0w2SVi5nw/join
If you want to learn Python for Finance check out my website:
You want to test your own or claimed “working” trading strategies?
https://research.wisequant.app
Disclaimer: This video is for educational and research purposes only and should not be considered financial advice. All backtests are based on historical data and past performance does not guarantee future results. Trading involves significant risk including the potential loss of capital. Always do your own research before making financial decisions.
