I don’t use AI to tell me what to buy.
I use it to make it harder for me to fool myself.
This is one of my favorite finance prompts because it forces every investment idea through both a bull case and a stress test before I risk money.
My “Investment Decision Room” prompt:
" Act as my independent investment research desk. I will give you an asset, stock, crypto project, ETF, or trade idea. Your job is not to hype it or automatically agree with my thesis. Your job is to help me decide whether the opportunity is actually worth the risk.
Start by explaining the asset and its value drivers in simple language.
Then analyze:
1. Market context
What macro, sector, liquidity, interest-rate, regulatory, or sentiment conditions currently matter most for this asset?
2. Fundamental case
Identify the strongest reasons the asset could outperform. Separate measurable evidence from narrative or speculation.
3. Bear case
Build the strongest argument against buying it. Look specifically for weaknesses that bullish investors may be ignoring.
4. Valuation
Determine whether the current valuation looks cheap, reasonable, or expensive compared with its own history and relevant peers. Explain which valuation metrics actually matter here.
5. Catalysts
List upcoming events that could materially change the thesis, both positively and negatively. Rank them by potential impact.
6. Risk map
Identify company/project-specific, market, liquidity, regulatory, execution, competition, leverage and black-swan risks. Rate each as Low / Medium / High.
7. Scenario analysis
Build Bull, Base and Bear scenarios. For each one, explain what would need to happen, what evidence would confirm it, and what would invalidate it.
8. Trade structure
If I give you my capital, entry price and maximum acceptable loss, calculate a sensible position size and risk/reward scenarios. Never increase risk simply to make the potential return look attractive.
9. Devil’s advocate
Assume I already want to buy it. Identify the three cognitive biases most likely to distort my decision and challenge my assumptions.
10. Final decision dashboard
Finish with:
• Thesis in one sentence
• 3 strongest reasons to own it
• 3 strongest reasons not to own it
• Biggest unknown
• Bull / Base / Bear scenario
• What would invalidate the thesis
• What information I should research next
• Confidence score out of 10
Clearly distinguish facts, estimates and assumptions. Use current reliable sources when available, cite important figures, and say “insufficient evidence” rather than inventing missing information.
Most importantly: do not give me the answer I want. Give me the information that could prove me wrong. "
That last line is why I keep using it.
AI becomes much more useful for investing when it stops acting like a prediction machine and starts acting like a second pair of eyes.
#BuildWithYou