AI-Backbone
There are only a handful of genuinely high-value books about trading, and the ones written by Al Brooks belong on everybody's shelf. The problem: each of them runs to 400 pages or more.
I started reading a couple of them a few years ago, but I never managed to build anything solid from them. They contain so many ideas that I simply got overwhelmed. And even once you have picked an idea, building a trading strategy around it used to take months before it worked as intended and was ready for optimisation and incubation.
Until now.
Today we have AI. This article shows how I used it to read a 478-page Al Brooks book on price action trends — and believe me, that book offers far more than a single idea about when to enter or exit a trade.
Spoiler: you can download the full strategy source code for free at the end of this article.
The Idea
I want to build a lot of strategies, let them run, and trade the best of them with real money. Quickly, cheaply and reliably.
So I built an AI harness that can build, improve, run and monitor many strategies on one platform.
How I Did It
Step 1
Build a knowledge base the AI can actually use
I fed some of the very best trading books into a vectorised semantic search database. The AI can then query it by meaning rather than by exact wording, so it finds the passages that are relevant to a question even when they are phrased in completely different words.
The prompt that starts the whole thing off.
Step 2
Ask for the ideas, not for the summary
I asked the AI to find trend-following ideas in those books. It came back with several genuinely interesting approaches. Have a look:
The ideas the AI pulled out of the book — each one traceable back to the passage it came from.
Step 3
Build it, optimise it, run it
From there the harness does the heavy lifting. Here is the strategy that came out of it.
The equity curve. The strategy finds something to trade on nearly every day.
The statistics. Note the run of consecutive losses — and that the strategy remains profitable.
Key insight: the strategy stays profitable even after fourteen consecutive losses. That is possible because it keeps its losses small relative to its wins. You can read it straight off the histogram.
The trade list. Many positions are managed as a package and closed together, either at the stop loss or at the take profit.
The trades on the chart — several positions stacked into the same move.
This last one is the part I find remarkable. Brooks supplied the entry; the AI fused it with an exposure management scheme that opens several positions so that larger gains are provoked, while still closing the whole package within a sensible stop loss. Putting this kind of risk management, entries and exits together is exactly the work that used to take me months.
I want everyone to become part of the new AI for trading opportunity. That's why I give the source code of this strategy for free.
Download the Strategy
The full source code, the parameters and the prompts behind this strategy are yours to take. Read the code, run the backtest yourself, and change whatever you disagree with.
Download the prompt and the MT5 source code here.
Start Building Your Trading Operation
Join the waitlist to get early access and build strategies like this one.
If you want to build a strategy like this one, join the waitlist and get ahead of the other traders.