How I Use ChatGPT to Research Stocks in 15 Minutes

Most retail investors spend too long on research that doesn’t move the needle, or too little time and buy stocks they don’t understand. ChatGPT (and Claude) don’t replace fundamental analysis — but they dramatically compress the time to develop a first-pass thesis on any stock.

Here’s the exact 15-minute research workflow, with the specific prompts that work best. This isn’t about getting AI to make your investment decisions. It’s about using AI to get to an informed starting point 10x faster than reading every document yourself.

Step 1: Company Overview (2 minutes)

Start with a clean business model summary. Use this prompt in ChatGPT-4o (with web search enabled) or Perplexity:

Prompt: “Give me a business model overview of [Company Name] ([TICKER]). Include: (1) How they make money — revenue streams with approximate % of total, (2) Who their customers are, (3) Their main competitive advantages, (4) Key metrics the market watches, (5) What the company was worth at last close and approximate revenue multiple. Be specific and concise.”

What you’re looking for: Can you explain how this company makes money in two sentences? If the AI can’t generate a clear answer, the business model may be genuinely confusing — which is itself a red flag for long-term investors.

Step 2: Competitive Moat Assessment (3 minutes)

Prompt: “Analyze [TICKER]’s competitive moat. For each of the following moat types, rate it as Strong / Moderate / Weak / None and give one sentence of evidence: (1) Cost advantage, (2) Switching costs, (3) Network effects, (4) Intangible assets (patents, brand, licenses), (5) Efficient scale. Then give an overall moat rating and the biggest threat to it.”

This gives you a structured moat analysis in 60 seconds that would take 30 minutes of reading to approximate. The AI will sometimes hallucinate specific details — verify key claims, but the framework output is reliable.

Step 3: Financial Health Check (3 minutes)

For this step, use a combination of ChatGPT with web access and a direct data source (Macrotrends or Wisesheets for actual numbers).

Prompt: “Analyze [TICKER]’s financial health. Based on publicly available data, provide: (1) Revenue growth rate (TTM and 3-year CAGR), (2) Gross margin and trend, (3) Free cash flow (positive/negative, trend), (4) Debt-to-equity ratio, (5) Cash and cash equivalents, (6) Any concerning items from recent balance sheet or cash flow statement. Flag any red flags.”

Always verify the numbers independently. AI can miscalculate or use stale data. Use the output as a checklist of what to verify, not as ground truth.

Key Financial Red Flags to Watch For

• Rapidly growing accounts receivable relative to revenue (revenue recognition concerns)
• Declining gross margins while revenue grows (pricing pressure)
• Rapidly increasing share count (aggressive dilution)
• Cash balance declining faster than EBITDA losses (burn acceleration)
• Revenue growth slowing while operating expenses grow (scaling problems)

Step 4: Risk Factor Extraction (3 minutes)

Every 10-K contains a risk factors section. Reading it cover-to-cover takes an hour. AI can compress this:

Prompt: “What are the top 5 investment risks for [TICKER]? Include both company-specific risks and industry/macro risks. For each risk, rate the probability (Low/Medium/High) and potential impact (Low/Medium/High). Flag any that have materialized recently or show signs of materializing.”

Follow up with: “Has [TICKER] faced any recent developments (last 6 months) related to these risks?” (Use web-enabled AI for this.)

Step 5: Valuation Sanity Check (4 minutes)

You don’t need a full DCF model for a first-pass. You need to know if the valuation makes sense given the growth rate and business quality.

Prompt: “Do a quick valuation sanity check for [TICKER]. Provide: (1) Current P/E, P/S, EV/EBITDA ratios vs. sector average, (2) PEG ratio if applicable, (3) If analyst consensus price targets exist, what’s the range and implied upside/downside, (4) What growth rate does the current valuation imply (reverse DCF), (5) Is this stock cheap, fairly valued, or expensive relative to its growth rate? Be direct with your assessment.”

The goal isn’t precision — it’s calibration. Is this a 10x revenue multiple requiring hypergrowth justification, or a 2x revenue stock that’s unloved? That context shapes what questions matter most.

Step 6: Investment Thesis in One Paragraph (1 minute)

Prompt: “Based on everything above, write a one-paragraph bull thesis for [TICKER] and a one-paragraph bear thesis. Then state: If the bull case plays out perfectly over 3 years, what is the stock price? If the bear case plays out, what is the downside scenario?”

This forces the AI to synthesize and gives you two bookend scenarios for position sizing decisions.

Putting It Together: The 15-Minute Review Sheet

After running through these steps, you have:

• Business model clarity
• Moat assessment
• Financial health overview
• Top 5 risks rated by probability/impact
• Valuation context
• Bull/bear thesis with price scenarios

That’s enough to decide: (1) Is this worth deeper research?, (2) If I already hold it, should I be more/less concerned than I was?, (3) What are the 1-2 questions I need to answer to get conviction?

What AI Stock Research Can’t Do

AI models have knowledge cutoffs and can hallucinate specific data points. Always verify key financial figures independently. AI can’t tell you how insiders are actually executing their strategy — read earnings call transcripts for management tone and candor. And AI can’t give you the intuition that comes from following a company for years.

Use AI as a research accelerator, not a research replacement. The goal is to get you to the right questions faster, so your limited time goes to high-value analysis rather than reading every filing cover-to-cover.

For our own space stock research process — including how we track RKLB, ASTS, PL, and others — see the Orbital Returns Space Watchlist and our full blog archive. Sign up for the free weekly newsletter for regular stock analysis delivered to your inbox.