Use AI for Stock Research & Investment Analysis

1. 1

   Set Up Your Investor Profile and Research Framework

   Before asking AI about any stock, establish your investing identity -- goals, risk tolerance, time horizon, and analytical framework. This context ensures every AI response is calibrated to YOUR situation, not textbook advice.

   ChatGPT Claude

   I want to use you as my investment research assistant. Before we analyze any stocks, I need you to understand my investing context so your analysis is relevant to MY situation.
   
   **IMPORTANT DISCLAIMER: I understand you are an AI providing research assistance and information organization — NOT financial advice. All investment decisions are my own responsibility. I will verify all data you provide with primary sources before making any decisions.**
   
   **My investor profile:**
   - Experience level: [choose: Complete beginner (I've never bought a stock) / Novice (I have a 401k and maybe a few stocks) / Intermediate (I've been investing 2-5 years and understand basic concepts) / Experienced (5+ years, I understand valuation and read financial statements)]
   - Investment goal: [e.g., "Long-term wealth building for retirement in 25 years" or "Generate passive income from dividends" or "Growth — I want to find the next big winners" or "I have $10,000 to start investing and don't know where to begin"]
   - Risk tolerance: [choose: Conservative (I lose sleep if my portfolio drops 10%) / Moderate (I can handle 20% drawdowns if the long-term thesis is intact) / Aggressive (I'm young, I can handle volatility for higher expected returns)]
   - Time horizon: [choose: Short-term (under 1 year) / Medium-term (1-5 years) / Long-term (5-20 years) / Very long-term (20+ years, retirement)]
   - Current portfolio: [e.g., "$50k in index funds, $10k in individual stocks (Apple, Tesla, Microsoft), $5k cash to deploy" or "Nothing yet — starting from zero" or "I'd rather not share specifics"]
   - Sectors I'm interested in: [e.g., "AI and technology, clean energy, healthcare" or "I don't know enough to have preferences yet"]
   - Sectors I want to avoid: [e.g., "Tobacco, weapons, gambling" or "No restrictions"]
   - How much time I'll spend on research: [e.g., "2-3 hours per week" or "I want a low-maintenance approach"]
   
   **Based on my profile, help me with:**
   
   1. **My Research Framework**: What type of analysis best matches my profile? (value investing, growth investing, dividend investing, index investing, etc.) Explain each in 2 sentences and recommend the best fit for me with rationale.
   
   2. **My Key Metrics**: Based on my chosen framework, what are the 8-10 financial metrics I should always check before buying a stock? For each metric:
      - What it measures (one sentence, no jargon)
      - What "good" looks like (a benchmark range)
      - What a red flag looks like
      - Where to find this data for free
   
   3. **My Research Checklist**: Create a simple checklist I can run through for any stock before buying. Keep it under 15 items. This becomes my standard operating procedure.
   
   4. **Tools Setup**: Recommend 3-5 free tools/websites for stock research (screening, financial data, news, filings) that complement AI analysis. Tell me specifically what to use each one for.
   
   5. **What AI Can and Cannot Do**: Be honest about where AI analysis is useful (pattern recognition, data summarization, scenario analysis) and where it's dangerous (predicting future prices, timing the market, emotional decision-making). What should I NEVER rely on AI for?

   Copy

   Tip: The most important thing AI can do for you as an investor is NOT pick stocks — it's help you build a consistent research process that you follow every time, removing emotional decision-making. Warren Buffett's edge isn't genius stock picks; it's a disciplined framework he applies consistently for 60 years. Your AI research assistant should help you build YOUR framework, then keep you honest about following it.

2. 2

   Screen Stocks Using AI-Powered Fundamental Analysis

   Stock screening filters thousands of stocks into a manageable watchlist based on specific criteria. AI helps you define smart filters, interpret results, and identify which stocks deserve deeper research -- saving hours of manual work.

   ChatGPT Claude Perplexity

   I want to screen for stocks that match my investment criteria. Help me build and run a smart screening process.
   
   **What I'm looking for:**
   - Investment style: [from your profile in Step 1, e.g., "Growth stocks" or "Dividend income" or "Undervalued companies"]
   - Market cap preference: [choose: Large-cap (>$10B, safer) / Mid-cap ($2-10B, balanced) / Small-cap (<$2B, higher risk/reward) / No preference]
   - Sector focus: [e.g., "Technology and Healthcare" or "Broad — across all sectors"]
   - Geographic focus: [e.g., "US stocks only" or "US + International"]
   - Specific criteria: [e.g., "Revenue growing >15% annually, profitable, P/E under 30" or "Dividend yield >3%, dividend growing for 10+ years, payout ratio under 60%" or "I don't know what criteria to use — suggest some based on my profile"]
   
   **Help me screen in 3 phases:**
   
   1. **Define My Screen** (criteria selection):
      Based on my investment style, suggest 6-8 screening criteria with specific thresholds. For each criterion, explain:
      - Why this matters for MY strategy
      - The threshold I should set (and why)
      - Which criteria are must-haves vs nice-to-haves
      
      Example format:
      | Criterion | Threshold | Why It Matters | Must-Have? |
      |-----------|-----------|---------------|------------|
      | Revenue growth | >15% YoY | Shows demand is growing | Yes |
   
   2. **Run the Screen** (generate candidates):
      Based on these criteria, suggest 8-12 stocks that likely meet most of them as of your last data update. For each stock:
      - Company name, ticker, and what they do (one sentence)
      - Which of my criteria it passes and which it fails
      - A one-sentence bull case and one-sentence bear case
      - How well-known or followed this stock is (heavily covered vs under the radar)
      
      **IMPORTANT: Your training data has a cutoff date. All stock data you provide may be outdated. Tell me your data cutoff and remind me to verify ALL numbers on [Yahoo Finance / Finviz / SEC filings] before making any decisions.**
   
   3. **Narrow to a Short List** (prioritize):
      From the 8-12 candidates, help me pick 3-5 for deep research. Rank them by:
      - Fit with my investment criteria
      - Quality of the business (competitive moat, management, market position)
      - Current valuation attractiveness
      - Conviction level (how confident you are in the thesis, and why)
   
   For my top 3, provide the specific questions I should research next (we'll do that in Steps 3-4).

   Copy

   Tip: Never buy a stock just because AI suggested it. Use AI screening as the first step, then verify every data point on primary sources: Yahoo Finance for real-time prices and key ratios, SEC EDGAR for actual financial filings, and the company's own investor relations page for earnings transcripts. AI is excellent at organizing information and suggesting what to look at — but the numbers it cites may be outdated or hallucinated. Always verify.

3. 3

   Analyze Earnings Calls and Financial Reports

   Earnings calls and quarterly reports are goldmines, but they're dense and time-consuming. AI summarizes these documents, highlights key takeaways, and spots the signals that matter -- turning a 90-minute earnings call into a 5-minute briefing.

   ChatGPT Claude Perplexity

   I want to analyze a company's recent performance. Help me break down their financial report and/or earnings call.
   
   **The company:** [ticker and name, e.g., "NVDA — Nvidia"]
   
   **What I have access to (paste or describe):**
   [Choose what applies:]
   - I'll paste the earnings call transcript below: [paste text, or say "I'll paste it in the next message"]
   - I'll paste key financial data: [paste revenue, EPS, margins, etc. from Yahoo Finance or the 10-Q]
   - I just want you to analyze based on your training data (I understand this may be outdated)
   - I have the 10-K/10-Q filing — I'll paste relevant sections
   
   **Analyze using this framework:**
   
   1. **The Numbers** (quantitative):
      - Revenue: What was it? Growth rate? Beat or miss analyst expectations? Any one-time items inflating/deflating it?
      - Earnings (EPS): Reported vs expected? What's driving earnings growth or decline?
      - Margins: Gross margin, operating margin, net margin — are they expanding or compressing? Why?
      - Cash flow: Is the company generating real cash, or are earnings just accounting tricks?
      - Guidance: What did management project for next quarter/year? Is guidance above or below analyst consensus?
      - The ONE number that matters most: Which single metric best captures this company's health right now?
   
   2. **The Story** (qualitative — from earnings call):
      - Management tone: Were they confident, cautious, evasive, or defensive? Quote specific language that reveals tone.
      - Key initiatives: What is management focused on? What did they spend the most time discussing?
      - Risks acknowledged: What concerns did they raise? What questions from analysts seemed to make them uncomfortable?
      - Buzzword vs substance: Did they say anything concrete about AI/growth/innovation, or was it just buzzwords?
      - What they DIDN'T say: What obvious topics were conspicuously absent from the discussion?
   
   3. **The Context** (competitive landscape):
      - How does this performance compare to their main competitors this quarter?
      - Are they gaining or losing market share?
      - Any industry-wide trends that explain the results (not company-specific)?
   
   4. **The Verdict** (for MY portfolio):
      - Based on my investor profile (from Step 1), is this company still a good fit for me?
      - Key risks I should monitor going forward
      - What would change your thesis? (What would make this a buy, hold, or sell?)
      - One specific metric to watch next quarter that will tell me if the thesis is playing out
   
   **DATA ACCURACY WARNING: Remind me which specific data points you're confident about vs. which I should verify. If you're working from training data rather than the actual document, flag that clearly.**

   Copy

   Tip: For the most accurate analysis, paste the ACTUAL earnings call transcript or 10-K text into the AI rather than asking it to work from memory. You can find free earnings transcripts at seekingalpha.com, fool.com, or the company's investor relations page. Perplexity is particularly useful here because it can search the web for recent data. Pro move: ask the AI to compare how management described the same business segment last quarter vs. this quarter — tone shifts often signal inflection points before the numbers do.

4. 4

   Build a Portfolio Risk Assessment

   Most individual investors have no idea how risky their portfolio actually is -- they own stocks they've heard of and hope for the best. This step helps you analyze your portfolio's risk profile, identify hidden concentrations, and stress-test against realistic scenarios.

   ChatGPT Claude

   I want to assess the risk profile of my current (or planned) portfolio. Help me understand what I'm actually exposed to and whether my portfolio matches my stated risk tolerance.
   
   **My portfolio:**
   [List your holdings — be as specific or general as you're comfortable with:]
   - [Ticker] — [approximate % of portfolio] — [why I bought it]
   - [Ticker] — [approximate % of portfolio] — [why I bought it]
   - [Continue for all holdings...]
   - Cash: [% of portfolio]
   
   Example:
   - AAPL — 25% — I've held it for years, it's my biggest winner
   - NVDA — 20% — AI hype, bought in 2024
   - VOO (S&P 500 ETF) — 30% — Core holding
   - MSFT — 15% — Diversification from Apple
   - Cash — 10% — Waiting for a dip
   
   **My stated risk tolerance (from Step 1):** [Conservative / Moderate / Aggressive]
   **My time horizon:** [Short / Medium / Long / Very long]
   
   **Analyze my portfolio risk across these dimensions:**
   
   1. **Concentration Risk**:
      - How diversified am I really? (number of holdings is not the same as diversification)
      - Do I have hidden concentration? (e.g., 4 tech stocks = not actually diversified)
      - What % of my portfolio is in my top 3 holdings?
      - Am I overweight in any single sector, geography, or market cap?
      - Comparison: how does my sector allocation compare to the S&P 500?
   
   2. **Correlation Risk**:
      - Which of my holdings tend to move together? (If A drops, does B also drop?)
      - In a tech crash (like 2022), what % of my portfolio would likely decline simultaneously?
      - Do I have any true diversifiers — holdings that go UP when the rest goes DOWN?
   
   3. **Scenario Stress Tests**: What would happen to my portfolio in each scenario?
      - Scenario A: Broad market correction (-20% like early 2022)
      - Scenario B: Interest rates spike (how rate-sensitive are my holdings?)
      - Scenario C: AI hype bubble deflates (if applicable)
      - Scenario D: Recession (revenue decline across the economy)
      - Scenario E: My worst-case scenario — what's the maximum realistic drawdown?
      For each scenario, estimate the approximate portfolio impact and which holdings are most vulnerable.
   
   4. **Risk-Return Assessment**:
      - Does my portfolio risk level match my stated risk tolerance? If I said "moderate" but my portfolio is concentrated tech, flag the mismatch.
      - Am I being compensated for the risk I'm taking? (Higher risk should mean higher expected return)
      - What's the simplest change I could make to better align risk with my goals?
   
   5. **Action Items**:
      - Top 3 specific adjustments to improve my portfolio's risk profile
      - What to add for better diversification (asset classes, sectors, geographies I'm missing)
      - What to reduce if I'm overconcentrated
      - A rebalancing schedule recommendation (how often should I review?)
   
   **DISCLAIMER: This is a framework for thinking about risk, not a recommendation to buy or sell anything. Portfolio construction is personal and should account for tax implications, which I should discuss with a qualified financial advisor.**

   Copy

   Tip: The most dangerous portfolio risk is the one you don't see. Most individual investors think they're diversified because they own 10 stocks — but if 8 of them are US tech companies, they're actually making one concentrated bet on US tech. The correlation analysis in this step is the most valuable part: it reveals whether your 'diversified' portfolio would actually all crash at the same time. If it would, you need real diversifiers: bonds, international stocks, commodities, or REITs.

5. 5

   Create a Watchlist with AI-Monitored Signals

   A watchlist isn't just stocks you like -- it's a tracking system with clear criteria for when to buy, when to add, and when to walk away. AI helps you define those triggers upfront so you decide on logic, not emotion.

   ChatGPT Claude Perplexity

   Help me build a structured watchlist for stocks I'm interested in but haven't bought yet (or want to add to). I want clear entry criteria so I'm not just checking prices and guessing.
   
   **Stocks on my watchlist:**
   [List 3-8 stocks you're watching:]
   1. [Ticker] — [one-sentence reason I'm interested]
   2. [Ticker] — [one-sentence reason I'm interested]
   3. [Ticker] — [one-sentence reason I'm interested]
   [Continue...]
   
   **For each stock on my watchlist, create a structured research card:**
   
   1. **The Thesis** (in 3 sentences):
      - What does this company do and why is it well-positioned?
      - What's the specific catalyst or trend that could drive the stock higher?
      - What makes NOW a potentially interesting time to look at it?
   
   2. **Entry Criteria** (specific and measurable — when would I buy?):
      - Valuation trigger: At what P/E, P/S, or other valuation metric does this become attractive?
      - Technical trigger: What price level or pattern would signal a good entry? (support level, breakout, pullback to moving average, etc.)
      - Fundamental trigger: What business milestone would confirm the thesis? (e.g., "Revenue growth accelerates above 20%" or "They win a major contract" or "New product launches successfully")
      - Set a specific price alert: "I'd start researching a buy if [TICKER] reaches $[price]" — what price?
   
   3. **Kill Criteria** (when would I remove this from my watchlist?):
      - What would break the thesis? Be specific.
      - What's the maximum I'd pay? (Above this, I'm not interested.)
      - Time limit: If the thesis hasn't played out in [X months], reassess.
   
   4. **Position Sizing** (how much would I invest?):
      - Based on my portfolio size and risk tolerance, what's the maximum allocation?
      - Would I buy all at once or scale in? (suggest a scaling plan)
      - Where's my stop-loss? (At what point do I admit I'm wrong and sell?)
   
   5. **Monitoring Schedule**:
      - What should I check weekly? (price, key metric, news?)
      - What events should I watch for? (earnings dates, product launches, regulatory decisions)
      - What would prompt me to do a full re-analysis? (e.g., stock drops/jumps 15%+, management change, competitor news)
   
   **Finally, create a one-page watchlist summary table:**
   | Ticker | Thesis (5 words) | Entry Price | Kill Price | Next Catalyst | Review Date |
   
   This table should fit on one screen so I can glance at it daily.
   
   **REMINDER: All prices and valuations should be verified with current market data. My AI training data has a cutoff and may not reflect current prices. Always check real-time data before acting.**

   Copy

   Tip: The most valuable part of this exercise is defining your KILL criteria BEFORE you have money on the line. Investors lose the most money on stocks they should have walked away from but didn't because they fell in love with the story. Writing down 'I'll remove this from my watchlist if X happens' while you're calm and rational is a gift to your future emotional self. Print this watchlist and check it monthly — the discipline of a systematic process is worth more than any single stock pick.

Recommended Tools for This Scenario

ChatGPT

Freemium

The AI assistant that started the generative AI revolution

• GPT-4o multimodal model with text, vision, and audio
• DALL-E 3 image generation
• Code Interpreter for data analysis and visualization

Get Started →

Claude

Freemium

Anthropic's AI assistant built for thoughtful analysis and safe, nuanced conversations

• 200K token context window for massive document processing
• Artifacts — interactive side-panel for code, docs, and visualizations
• Projects with persistent context and custom instructions

Get Started →

Perplexity

Freemium

AI-powered search engine that answers questions with cited sources

• Real-time web search with inline source citations
• Pro Search multi-step deep research automation
• Multiple model options (Sonar, GPT-4o, Claude)

Get Started →

Google Gemini

Freemium

Google's multimodal AI assistant with deep ecosystem integration

• Gemini 2.0 multimodal model (text, image, audio, video)
• 1 million token context window
• Deep Google Workspace integration (Gmail, Docs, Sheets, Slides)

Get Started →