How to Use ChatGPT to Analyze a Stock (Step-by-Step)
Most people who try to analyze a stock with ChatGPT do it wrong.
They type “analyze Tesla for me” and wait. What comes back is a paragraph they could have found on Wikipedia. Revenue is growing. Competition is intense. Valuation is debated. Thanks, I guess.
That’s not a ChatGPT problem. That’s a prompting problem.
Done right, ChatGPT can turn a two-hour research session into twenty minutes. It can help you make sense of financial statements you’d normally skim past, catch things buried in earnings calls that analysts flagged twice, and pull together a coherent view of a business from a pile of messy information.
But it needs two things from you: the actual data, and the right questions.
Here’s exactly how to do it.
Before You Start: Two Things You Need
The materials. Grab the company’s most recent annual report (10-K for US stocks), the latest earnings call transcript, and a few key numbers — P/E ratio, revenue growth, gross margin. All free. Macrotrends, the company’s investor relations page, or just a Google search gets you there.
The mindset. ChatGPT is a research partner, not a fortune teller. It will help you think more clearly. It will not tell you whether to buy. That part stays with you.
Step 1: Set the Scene
Before you paste anything, give ChatGPT context. One message. This small step makes every response that follows noticeably better.
“I’m researching [Company Name] ([ticker]) as a potential long-term investment. I’ll be sharing financial data and company documents with you. Your job is to help me understand the business, identify risks, and think critically — not to give me a buy or sell recommendation.”
Now it knows what it’s doing. You’re not asking it to guess. You’re bringing it in as a thinking partner.
Step 2: Understand What the Business Actually Does
Even if you think you know what a company does, run this step anyway. You’ll almost always learn something.
“Here is the business description from [Company]’s most recent annual report: [paste it]. In plain language, explain: how this company makes money, who its customers are, what keeps them coming back, and what would have to be true for this business to still exist in 10 years.”
That last question is the one that matters. “What would have to be true for this to exist in 10 years” cuts through corporate language and forces a real answer. If the explanation sounds shaky — dependent on trends that might not hold, or on a single customer type — you know before you’ve looked at a single number.
Step 3: Dig Into the Numbers
This is where most investors get lost or skip ahead too fast. Don’t skip it, and don’t drown in it either.
Paste in five years of the basics — revenue, gross profit, operating income, net income, free cash flow, and total debt. Then ask:
“Here are [Company]’s financial results for the last five years: [paste data]. Analyze the trends. Is the business growing? Is profitability improving or getting worse? Is free cash flow tracking net income, or is there a gap? Flag anything that looks unusual.”
The gap between free cash flow and net income is the one most people miss. When net income looks great but free cash flow doesn’t keep up, it often means the company is recognizing revenue it hasn’t actually collected yet. ChatGPT will catch this if you ask.
Then go one level deeper:
“Gross margin has moved from [X]% to [Y]% over this period. What might explain that shift? Pricing power, competitive pressure, cost inflation? And how does it match what management said in their last earnings call?”
Step 4: Check If the Balance Sheet Can Take a Hit
A company can look great on the income statement and still blow up if the balance sheet is fragile. Five minutes here can save a lot of pain.
“Here is [Company]’s balance sheet summary: [paste assets, liabilities, debt, cash, current assets and liabilities]. How much debt is the company carrying? Could it survive a 20-30% revenue decline without being forced to raise capital or default? What does the current ratio tell us?”
This step matters most for companies that look exciting. High-growth stocks often carry significant debt. The question isn’t whether they’re growing — it’s whether they survive long enough for that growth to matter.
Step 5: Read the Earnings Call Like an Analyst
Earnings transcripts are where management gives itself away. Press releases are polished. The Q&A isn’t.
“Here is the Q&A section from [Company]’s most recent earnings call: [paste it]. Summarize what management seems most confident about and what they seem uncomfortable discussing. Which analyst questions did they answer directly, and which did they dodge? Were there topics that came up in multiple questions?”
If three different analysts asked about the same concern, that’s not a coincidence. And when a CEO answers a slightly different question than the one they were asked, they’re usually avoiding something. Worth noting.
Step 6: Think About the Competition
No company lives in a vacuum. Who else wants what this company has?
“Based on what you now know about [Company]’s business model and financials, help me think through the competition. Who are the main rivals? Is this company’s position getting stronger or weaker? And — this is the important one — what could a well-funded competitor do to take real market share in the next three years?”
That last question is uncomfortable on purpose. The best businesses are the ones where it’s genuinely hard to answer. If you can easily imagine a competitor replicating what this company does, that’s worth sitting with.
Step 7: Find the Risks That Aren’t in the Filing
Annual reports list every risk in exhaustive legal language. They’re designed to protect the company, not to actually inform you. The real risks are often the ones that aren’t listed.
“Here is the risk factors section from [Company]’s annual report: [paste it]. Tell me which risks you think are most material and most likely — not just the longest list. Then tell me: what risks are conspicuously absent? What could go wrong that management hasn’t mentioned?”
That second question is where ChatGPT earns its keep. It will cross-reference the industry context, the competitive dynamics, and the macro environment — and flag things that aren’t in the filing. Not always. But often enough to be worth asking every time.
Step 8: Look at the Valuation — Last, Not First
Most people start here. That’s backwards.
Knowing whether a stock is cheap or expensive only makes sense once you understand what you’re paying for. Valuation without context is just a number.
“The stock is currently trading at a P/E of [X], EV/EBITDA of [Y], and price-to-free-cash-flow of [Z]. Revenue is growing at [X]% annually. How do these multiples compare to the company’s historical range? What growth rate does the current price imply? At what price would this start to look genuinely attractive?”
Push for a specific answer on that last question. Having a price target in mind before you buy — one based on analysis, not hope — is one of the most useful habits you can build.
Step 9: Write the Investment Thesis
You’ve done the work. Now get it out of your head and onto paper.
“Based on our full analysis of [Company], write a concise investment thesis covering: what the business does and why it matters, the bull case in three sentences, the bear case in three sentences, the single biggest risk that could make the bull case wrong, and what you’d need to see over the next 12 months to stay convinced. Under 300 words.”
Save this somewhere you’ll find it.
Because here’s what happens. The stock drops 15% on a mediocre earnings quarter. Your instinct says sell. But you have this document — written by you, based on research you actually did — asking whether the investment case has changed or whether the market is just being emotional.
Most of the time, it’s the latter. The thesis you wrote is what keeps you from confusing noise with a real reason to act.
Common Mistakes Worth Avoiding
Asking for a recommendation. “Should I buy this?” produces nothing useful. Ask instead: “Make the strongest possible case against buying this stock right now.”
Skipping your own data. ChatGPT’s training has a cutoff. Quarterly numbers change. If you’re not pasting in current data, you’re working with information that might be a year out of date.
Accepting the first answer. The first response is usually decent. The third or fourth — after follow-ups and pushback — is where the real insight lives. It’s a conversation, not a search query.
Only looking for confirmation. If you’re excited about a stock, you’ll want ChatGPT to agree. Resist it. Ask for the bear case. It exists whether you ask for it or not.
Quick Reference: The 9-Step Workflow
Step 1 — Orient. Set ChatGPT up as your research partner.
Step 2 — Business model. How it makes money and what would have to be true for it to exist in 10 years.
Step 3 — Financials. Five years of key numbers. Free cash flow vs net income.
Step 4 — Balance sheet. Debt levels. Stress test against a 20-30% revenue drop.
Step 5 — Earnings call. Deflections, repeated concerns, uncomfortable answers.
Step 6 — Competition. What could a well-funded rival do in three years?
Step 7 — Hidden risks. What’s absent from the risk factors section?
Step 8 — Valuation. Last. What does the current price imply about growth?
Step 9 — Write the thesis. Bull case, bear case, biggest risk. Under 300 words. Save it.
First time through: about 45 minutes. For a quick first-pass, steps 1, 2, 5, and 9 take fifteen minutes and tell you whether it’s worth going further. That alone is worth the effort.
Nothing in this article is financial advice. Always do your own research before making any investment decision.