How I prepare a trade with AI: the 5 step method (with prompts)

Why a method and not a "signal"

The problem of almost every trader I know, me included for years, is not a lack of ideas. It is the lack of a repeatable process. You enter on a feeling, move the stop, exit out of fear. AI does not fix this alone, but it is a perfect tool to follow the same steps every time.

Here is the method I use. It takes a few minutes.

Step 1: write the thesis in one sentence

If you cannot explain the trade in one sentence, it is not a trade, it is a hunch. The sentence must include what you buy or sell, why, and in which scenario you are wrong.

Prompt to copy:

Here is my trade idea. Help me rewrite it as one clear sentence that includes: instrument, direction, main reason and what must happen for me to be wrong. Do not tell me what to do, just ask me questions if something is missing.
Idea: [write your idea here]

Step 2: ask AI to play devil's advocate

This is the step that saved me from mistakes most often. The goal is not to hear yes, it is to find the weak spots.

Prompt to copy:

This is my trade thesis: [thesis].
Play devil's advocate. Give me the 3 strongest reasons I could be wrong, the scheduled events that could invalidate it and the information I should check before entering. If you quote data or numbers, tell me where they come from so I can check them at the source.

Important: every figure AI suggests must be checked on a reliable source. AI can get numbers and dates wrong.

Step 3: calculate risk before entering

No intuition needed here, just arithmetic. Two numbers decide everything: how much you are willing to lose on this trade and where you place the stop.

The formula:

• Risk in money = capital × risk percentage per trade
• Risk per share = entry price - stop price
• Number of shares = risk in money ÷ risk per share

Purely illustrative example: capital of 10,000, risk of 1% per trade, so 100. Entry at 50, stop at 47: risk per share is 3. 100 divided by 3 is 33.3: the position is 33 shares.

Prompt to copy:

Calculate my position size. Capital: [X]. Maximum risk per trade: [Y]%. Entry: [price]. Stop: [price]. Target: [price].
Show me the steps, the number of shares or contracts rounded down, the maximum loss and the risk to reward ratio. Do not suggest changing my numbers.

Always check the result: it is simple maths and worth redoing yourself.

Step 4: the checklist before you click

A short checklist prevents the silly mistakes, the ones that cost the most.

Prompt to copy:

Before entering I want to run my checklist. Ask me these questions one at a time and wait for my answer:
1. Are there earnings, macro data or major events before the trade's horizon?
2. Is the instrument liquid enough to enter and exit easily?
3. Is this trade correlated with positions I already hold?
4. Is the stop at a technically meaningful level or did I place it only to risk less?
5. Am I entering because of the plan or because of an emotion (hurry, fear of missing out, wanting to win back losses)?
At the end, tell me only which answers are weak.

Step 5: write the journal (before and after)

The journal is what turns experience into method. Write before entering and complete it after the exit.

Template to copy:

Date and time:
Instrument and direction:
Thesis in one sentence:
Entry / Stop / Target:
Size and risk in money:
Emotional state before entering (1-5):
Outcome:
What I did well:
What I would change:

After a few weeks you can give all your entries to AI and ask which patterns repeat in losing trades. That is where you learn the most.

Mistakes to avoid

• Asking AI "should I enter?": the decision stays yours.
• Using a price or figure without checking it at the source.
• Moving the stop after entering.
• Skipping the journal when the trade goes well: wins need to be understood too.

This article is for educational and informational purposes only and is not financial advice. Trading involves a high risk of losing capital. The numerical example is purely illustrative.