anthropic just released the Fable 5 model, so I put it to the test for algo trading by having it research, backtest, optimize, and run Monte Carlo simulations on a trend-following strategy entirely on its own. Using the Jesse framework and an MCP server, the model builds a simple Bollinger Bands breakout strategy and validates it against overfitting to hit a Sharpe ratio above 1.2. I also break down whether Fable 5 is actually worth the cost compared to cheaper models like Opus and Sonnet for trading research.

👉 Jesse (MCP):
https://jesse.trade/?utm_source=fable5

👉 Get this strategy’s code:
https://jesse.trade/strategies/ethtrendbb?utm_source=fable5

👉 Join our FREE Discord community:
https://jesse.trade/discord

👉 Follow our Telegram:
https://jesse.trade/telegram

👉 Apex signup URL for 25% fee discounts:
https://jesse.trade/apex

👉 Explore my other strategies:
https://jesse.trade/strategies?utm_source=fable5

## Chapters
00: 00 Introduction
00: 49 Setting Up Fable 5 & Jesse
02: 55 The Strategy Research Prompt
04: 32 Reviewing the Results
07: 06 Optimization & Overfitting
07: 43 Monte Carlo Simulations
11: 13 Inside the Strategy Code
16: 30 Examining the Trades
18: 21 Should You Trade This?
19: 10 Is Fable 5 Worth It?