I trade stocks and forex alongside a day job. Been doing it a few years. For the past several months I've been using ChatGPT and Claude as a structured part of my daily trading routine — not to get signals, not to predict prices, but as a thinking partner with specific, repeatable workflows.
There's a way to do this that actually works, and a way that wastes your time. Here's the difference.
Why most traders get nothing useful from AI
Ask ChatGPT "should I buy AAPL?" and you'll get something like:
"AAPL has strong fundamentals, but as with any investment, you should consider your personal risk tolerance and consult a qualified financial advisor."
Useless. But it's not the AI's fault — it's the prompt's fault.
Trading decisions require context the AI doesn't have unless you give it: your entry and stop prices, your account size, your timeframe, the current chart structure, and your emotional state. Feed it those inputs with a structured request and a defined output format, and you get a completely different quality of response.
The pattern I use for every prompt I build:
1. Context block — everything the AI needs to know
2. Specific request — exactly what I want analysed
3. Output format — exactly how I want it structured
4. Constraint — "be direct, no hedging, no flattery"
Here are the prompts that made it into my permanent rotation.
The Pre-Market Plan
Every morning before open. Fill in the variables, run it, print the output:
Build my pre-market plan for today, [DATE].
Watchlist: [TICKERS]
Market context: [overnight moves, macro events, VIX level]
My strategy: [your approach in plain English]
Account status: [e.g. up 2R this week / at daily loss limit / fresh]
Structure the plan as:
1. MARKET BIAS — overall tone and what it means for my tickers
2. PER-TICKER LEVELS — key S/R zones, what I need to see to trade
3. SCENARIOS — bull case and bear case for today's session
4. TODAY'S RULES — specific rules given current conditions
5. PASS/TRADE DECISION TREE — criteria for trading vs sitting on hands
Concise. Actionable. Printable.
Takes four minutes. The output is a written session plan I trade from. Before this habit I was opening my platform cold and reacting to whatever moved. Now I have a plan before the open, and the number of impulsive trades I take has dropped measurably.
The Post-Trade Journal Entry
This one changed my journaling consistency more than anything else:
I just closed a trade. Raw details:
Ticker: [TICKER] | Direction: [LONG/SHORT] | Timeframe: [TF]
Entry: [PRICE] | Exit: [PRICE] | Stop: [PRICE] | TP: [PRICE]
P&L: [$X] | R-Multiple: [e.g. +1.8R or -1R]
Emotional state during the trade: [honest — e.g. "anxious, wanted to exit early at +1R"]
Generate a structured journal entry:
1. TRADE SUMMARY (2–3 factual sentences)
2. SETUP QUALITY SCORE (1–10 with explanation)
3. EXECUTION QUALITY SCORE (did I follow my plan?)
4. PSYCHOLOGICAL AUDIT (what emotions influenced my decisions?)
5. KEY LESSON (one specific, actionable takeaway)
6. REPLICATE / AVOID
Write as a direct trading coach. No flattery.
That last line — "No flattery" — is doing a lot of heavy lifting. Without it, AI finds the silver lining in every bad trade. With it, you get the honest post-mortem you actually need.
I've journaled every trade for four months straight because this prompt removed all the friction. Consistency has compounded.
Position Size Calculator
Before every trade, without exception:
Calculate my exact position size and give me a full risk summary.
Account balance: $[BALANCE]
Risk per trade: [RISK %]% of account
Entry: $[ENTRY] | Stop: $[STOP] | Target: $[TARGET]
Instrument type: [STOCKS / FOREX / CRYPTO / FUTURES]
Calculate and return:
1. Dollar amount at risk: $X
2. Correct position size (shares/units/contracts): X
3. R:R ratio: X:1
4. Break-even win rate required at this R:R: X%
5. Portfolio heat (% of total account in this position): X%
Flag any concerns. Is this sizing appropriate for the setup?
The math is simple enough to do mentally. The point is being forced to read the numbers back as a structured output before you enter. How often do you think you know your R:R without actually calculating it? More often than you'd admit.
The Stop Loss Placement Prompt
When you're unsure exactly where to put your stop:
Help me place my stop loss for this trade.
Ticker: [TICKER] | Direction: [LONG/SHORT] | Timeframe: [TF]
Entry: [PRICE] | Setup type: [e.g. breakout retest, support bounce]
Nearest support below entry: [LEVEL]
Nearest resistance above entry: [LEVEL]
My target: [PRICE] | Risk per trade: [%]%
Give me:
1. Logical stop loss price with reasoning
2. Why this is the structural invalidation level (not arbitrary)
3. The R:R this placement gives me
4. Whether R:R is worth taking or I should skip this trade
5. The most common stop placement mistake for this setup type
The key thing this prompt does: forces you to think about the stop as a structural invalidation level, not a random number below entry. That shift in framing alone is worth running it.
The Revenge Trading Circuit Breaker
The one I use most but talk about least:
I just took a [R-MULTIPLE] loss on [TICKER].
I'm feeling [HONEST EMOTIONAL STATE — e.g. "frustrated, want to get it back"].
I have the urge to re-enter immediately or find another trade.
1. IMMEDIATE INTERVENTION — what do I do in the next 10 minutes?
2. PHYSIOLOGICAL RESET — a specific 3-step process to calm my system
3. RATIONAL REALITY CHECK — reframe this loss in context of my account and stats
4. DECISION GATE — 5 questions I must answer YES to before trading again today
5. HONEST RECOMMENDATION — should I close the platform for today?
Be direct. My capital is at stake.
Impulsive trades after losses are one of the most common ways retail traders turn a manageable bad day into a blown week. Having something that interrupts that loop — even just 60 seconds of reading a structured response — has stopped enough bad decisions that it's earned a permanent spot in my workflow.
The Monthly Bias Audit
Run this once a month. Describe your own behaviour honestly:
Act as a professional trading psychologist.
I will describe my recent trading behaviour. Identify my top 3
cognitive biases with specific evidence from my descriptions.
My recent behaviour:
[DESCRIBE IN DETAIL — be specific and honest. Examples:
"I move my stop when price gets close to it."
"I exit winners before target when I'm up 1R."
"I hold losing trades far longer than my rules allow."]
For each bias:
1. Name it precisely
2. Show the specific evidence from my description
3. The psychological root cause
4. One countermeasure I can implement this week
5. A self-check question to ask before each trade
Then create a BIAS WATCHLIST — 3 items I can print and keep at my desk.
The output is uncomfortable to read. That's the point. You already know your patterns. This makes you write them down, name them, and get a specific plan to address each one.
What this won't do
Worth being clear:
- AI cannot read your charts — only what you describe to it
- It won't create an edge where none exists
- It can be wrong, especially on calculations — always verify
- It's a thinking tool, not an authority
The value is in consistency, structure, and honest self-assessment — not signals.
The Pack
I built 15 of these prompts — refined over months of actual daily use — into a PDF. Fully selectable text, copy-paste ready, each one with a real example output so you know what a useful response looks like.
Covers: price action analysis, trade journaling, risk management, psychology, and strategy validation.
→ https://anusiempreciouso.gumroad.com/l/TraderMind_AI_Prompt — $17 one-time
Or use the prompts above and build your own. The pattern is the same throughout: context + specific request + output format + no-hedging constraint.
Not financial advice. These are process and thinking tools — not signals or trading recommendations.
Top comments (0)