Born to Be Claude: 5 Best Prompts For Analyzing a Stock
The 5 best prompts to make AI your stock research partner
Hi, Investor 👋🏻
If you’ve been following me for a while, you know I’ve spent over 12 years analyzing companies:
Thousands of 10-Ks
Dozens of earnings calls
And more late nights than I’d like to admit
It was slow. Draining.
If you’ve ever tried digging into a company yourself, I’m sure some of this sounds familiar.
Now things are changing…
LLMs are evolving. They’re no longer toys.
Instead of getting lost in filings, we can use them to:
Organize information
Summarize what matters
Surface insights faster
I’ve tested GPT, Gemini and even Grok.
But the one that impressed me the most was Claude Opus.
And that’s why I’m writing this article.
Today I’ll show you how it can actually help in your day-to-day as an investor.
What’s Claude?
Claude is Anthropic’s family of large language models, named after Claude Shannon, the father of information theory.
If Shannon laid the theoretical foundations for how we measure and transmit information, Anthropic is trying to apply that same rigor to how AI reasons and communicates.
The version I use most - Claude Opus - is the top tier of Anthropic’s lineup (above Sonnet and Haiku).
Think of it as the “analyst upgrade package”: slower than the lighter models, but smarter, more careful and more reliable on complex reasoning tasks.
Here’s what makes Claude unique:
Long Context: with a context window of 200K+ tokens, Claude can handle entire annual reports, proxy statements or multiple earnings transcripts in a single thread.
No need to chop documents into tiny fragments (yes, even full 10-Ks!!!)
Instruction Fidelity: where some models drift into creative tangents, Claude tends to stick to the brief.
If you ask for a table with valuation comps, you’ll likely get it - structured, neat, and labeled.
Agentic Capabilities: Claude has built-in features like:
Computer Use: lets Claude interact directly with apps and your desktop environment - automating repetitive research tasks.
Tool Calling: enables Claude to pull in external data sources (like EDGAR or Google), boosting accuracy and context.
Artifacts: packages outputs - tables, memos, charts - into clean, shareable documents you can use with your team.
Training Philosophy: Anthropic is big on “Constitutional AI,” which makes Claude feel cautious, but also thoughtful.
It’s less likely to hallucinate aggressively when it’s unsure, often flagging uncertainty instead of faking precision (!!!)
In short, Claude is less of a “creative assistant” and more of a disciplined research partner.
For equity analysts, that distinction matters.
Found this content valuable? Share it with your network! Help others discover these insights by sharing the newsletter. Your support makes all the difference!
How Does It Stand Out?
The latest benchmarks show one thing:
Claude Opus 4.1 ranks at the very top - right next to GPT-5 and Gemini 2.5.
In the Arena Overview, it ranks first overall with 1,447 points, alongside GPT-5 and Gemini 2.5.
Among all models, it stands out as the most balanced - performing strongly across hard prompts, math-heavy reasoning, text generation, and more.
Within the MMLU Pro leaderboard, a benchmark for advanced professional reasoning, Claude (both thinking and nonthinking) scores close to 88%, ahead of GPT-5 at 87%.
Finance-focused benchmarks point in the same direction. On the Finance Agent test - 537 analyst-style questions - Claude performs especially well in the kind of work that actually matters: parsing 10-Ks, pulling out margins and reconciling share counts.
Its ability to handle long documents, stick to instructions, and reason cautiously translates into real productivity for analysts.
There are, of course, trade-offs.
The cost per input/output is higher ($15/$75 vs. GPT-5’s $1.25/$10).
But the latency is much better - usually 10-18 seconds compared to 30+ for GPT-5.
That balance is why many professionals, myself included, see Claude as the most “analyst-like” model today.
It may not have GPT’s flashy multimodal demos, but when it comes to actual research, it’s the one I trust more.
Enjoying the content? Don’t miss out on more exclusive insights and analyses. Subscribe now and stay updated.
How to Analyze a Stock with Claude:
At the end of the day, benchmarks and rankings only get you so far.
What really matters is how these models perform in practice - when you drop in a ticker and ask them to help build a case.
And if you’ve ever tried that with a single, simple question, you know it doesn’t work.
The chatbot drifts off, hallucinates or gives you something completely different from what you wanted.
That’s because in this game, inputs matter far more than outputs.
Think of it like this:
Input + Energy - Wastes = Output
A vague input leads to a noisy, shallow response (it uses the same energy, but the waste and by-products are disproportionately large %)
A disciplined input leads to a structured, useful analysis.
The Priming Prompt:
That’s why the very first step is always contextualization.
Before diving into ratios, comps or risk maps, you need to “prime” the model - essentially, you’re educating it, or even creating the persona it will adopt for the task ahead.
And that brings us to the first (and most important) prompt in the playbook: the priming prompt.
You are a senior equity research analyst at a top-tier investment fund with over 15 years of experience. Your role is to deliver professional, disciplined, and actionable equity research analysis.
Always integrate fundamental and macroeconomic perspectives into your work, with a focus on clarity and depth. You are data-driven, rational, and guided by insights—every conclusion must be supported by evidence or clearly stated assumptions.
Was I clear enough? Can I proceed with the next instructions?Once trained with the right context, the AI is ready to support us in the equity research process.
Prompt 1: Business Model Overview
We couldn’t start anywhere else.
After all, what’s the point of investing in stocks if not for the sheer pleasure of studying business models?
I’ll admit… that’s my favorite part.
This is where great companies stand out:
Strong management. Competitive Advantages. Strategy. Long-term perspective.
And that’s exactly what we’ll explore in our first prompt:
Explain how [Insert company name here] makes money by breaking down its main products, services, and geographies, highlighting revenue contribution by segment where possible. Identify the key revenue drivers (pricing power, volume growth, renewals, unit economics) and describe the cost structure (fixed vs. variable, operating leverage, COGS vs. SG&A).
Outline who the customers are, how concentrated or diversified the base is, and the distribution model (direct, retail, e-commerce, partnerships). Clarify the pricing logic—whether one-off, subscription, transaction, or usage-based.
Then analyze the company’s competitive positioning: highlight the top 2–3 moats such as brand, scale, network effects, IP, or switching costs. Explain how these advantages sustain profitability and whether they create customer lock-in or premium pricing power. Finally, compare with closest peers, showing how their models differ and what makes this company distinct.
Be didactic. You are explaining this to an 18-year-old with no experience in the stock market.Prompt 2: Sector Overview
A company never operates in a vacuum.
Even the strongest business model can be limited - or supercharged - by the sector it belongs to.
Understanding the industry context helps you see whether growth is driven by company-specific execution or simply by a rising tide lifting all boats.
That brings us to prompt #2:
Explain the overall sector environment in which the company operates. Describe market size, growth rates, and key demand drivers.
Highlight competitive dynamics: number of players, market concentration, barriers to entry, and degree of pricing power. Assess regulatory or technological trends shaping the industry.
Identify major risks that affect the entire sector (commodity cycles, FX, consumer behavior, interest rates, new entrants). Then, provide a forward-looking perspective: how the sector is expected to evolve in the next 3–5 years, and what role the company could play in this landscape.
Conclude with 2–3 key indicators or macro variables to monitor going forward.Prompt 3: Financials and Valuation
At the end of the day, valuation is where great companies turn into great investments.
A strong business model and durable moats are critical, but if you pay the wrong price, returns will disappoint.
That’s why analyzing financials and valuation side by side is essential - numbers tell the real story of growth, profitability and capital efficiency.
Review the company’s financial performance over the last 3–5 years. Highlight revenue growth, margin trends (gross, operating, net), free cash flow generation, and capital structure (net debt/EBITDA, leverage ratios). Assess returns on invested capital (ROIC vs. WACC) to judge whether the company is creating or destroying value.
Then evaluate valuation metrics relative to peers. Present key multiples such as P/E, EV/EBITDA, EV/FCF, and P/B, noting where the company trades at a premium or discount and why. Compare valuation against financial quality (growth, margins, returns) to see if the premium or discount is justified.
Conclude with a fair value discussion: does the current market price reflect fundamentals, and under what conditions would the stock look undervalued or overvalued?Prompt 4: Risks and Perspectives
Every investment case looks good in a PowerPoint (back in my days as an analyst, we had already abandoned ppt).
What separates serious analysis from wishful thinking is the ability to stress-test the thesis.
By mapping risks and looking ahead, you build conviction not only in the upside but also in how much pain you’re willing to accept if things go wrong.
Identify the top risks facing the company across categories such as demand, pricing power, input costs, regulatory changes, technological disruption, and competitive pressure. For each risk, explain:
1. The probability of occurrence (high, medium, low).
2. The potential financial impact (on revenue, margins, EPS, or FCF).
3. The early-warning indicators that an analyst should monitor.
Then, provide a forward-looking perspective: how the company is positioned over the next 3–5 years, considering macro trends, sector dynamics, and internal strategy. Highlight the main opportunities that could accelerate growth and the threats that could undermine it.
Conclude with a short summary: Which risks are “make or break”? Which opportunities are realistic? What would invalidate the investment thesis?Prompt 5: Earnings Calls
Earnings calls are where management shows its hand.
Beyond the numbers, you hear tone, strategy and how executives handle tough questions.
By reviewing the last few calls together, you can spot patterns: where guidance is shifting, what risks management downplays and which themes keep coming back.
And that brings us to our last one - prompt #5:
Summarize the last four company's earnings calls. For each call, extract:
Management’s key messages
Strategy updates, guidance changes, main themes.
Highlights and surprises (positive or negative).
Analyst questions that revealed important information.
Then, compare across the four calls:
Identify recurring themes and management priorities.
Highlight shifts in tone or changes in guidance.
Point out gaps—issues analysts keep asking about but management avoids.
Finally, conclude with:
The 3 biggest positive developments since the first of the four calls.
The 3 most concerning signals that could impact the thesis.
5 sharp questions an analyst should ask on the next call.Found this content valuable? Share it with your network! Help others discover these insights by sharing the newsletter. Your support makes all the difference!
DLocal ($DLO) Investment Thesis:
Now let’s put it to the test with a real case.
It’s a stock I’ve been following for quite some time - one of the rising stars in Latin America, though still not as widely known as it deserves to be.
✍🏻 Jimmy’s Notes: I had to ask the model to limit itself to 200 words (each prompt) - there was simply too much information for a single post.
I suggest you try it yourself with your own AI setup and investment thesis as well.
1. Business Model:
Think of DLocal ($DLO) like a specialized translator for money transactions. Just as you need a translator to communicate in a foreign country, global companies need DLocal to "translate" their payments when doing business in emerging markets like Brazil, Mexico, or India.
What DLocal Actually Does:
Imagine you're Netflix and want to expand to Brazil. Brazilian customers prefer paying with local methods like PIX (Brazil's instant payment system) or Boleto (payment slips), not just credit cards.
DLocal acts as the bridge—it takes Netflix's global payment system and connects it to all these local payment methods, currencies, and banking systems.
Revenue Model:
DLocal operates a transaction-based model earning money on every payment processed. In 2024, they processed $26 billion in Total Payment Volume (TPV) and generated $746 million in revenue—keeping approximately 2.9% of each transaction as their "take rate."
Example: When someone in Brazil pays $100 for Netflix, DLocal keeps $2.90 and passes $97.10 to Netflix.
Geographic Focus & Customers
Revenue comes primarily from Latin America (Brazil, Mexico, Argentina, Colombia), with expansion into Asia and Africa. Their customer base includes global enterprises like streaming services (Netflix, Spotify), e-commerce platforms (Amazon), gaming companies, and travel/hospitality firms—companies with significant transaction volumes that make the take rate model profitable.
Cost Structure
Variable costs include payment processing fees, fraud prevention, and currency conversion. Fixed costs cover technology infrastructure, regulatory compliance (licensing in 48+ countries), and sales/marketing. With 40% gross margins ($295M gross profit on $746M revenue), DLocal benefits from operating leverage as volumes grow.
Competitive Moats
Regulatory Network Effects: Licenses in 48+ countries—extremely expensive and time-consuming to replicate, requiring local regulatory approval, banking partnerships, and compliance infrastructure.
Local Payment Expertise: Deep understanding of regional preferences (PIX in Brazil, UPI in India, OXXO in Mexico) and unique fraud patterns.
High Switching Costs: Integration requires months of technical work, staff training, and risks payment failures during transition.
Competitive Differentiation
Unlike global players Stripe/Adyen (focused on developed markets), DLocal specializes in emerging market complexity. Key competitors include Rapyd (broader geographic focus), EBANX (Latin America specialist), and Airwallex (business payments).
DLocal's unique value proposition combines deep emerging market specialization, local payment method expertise, regulatory relationships in complex jurisdictions, and currency/fraud management tailored specifically to emerging market risks—capabilities that take years to build and are difficult for generalist competitors to replicate.
Enjoying the content? Don’t miss out on more exclusive insights and analyses. Subscribe now and stay updated.
2. Sector Overview:
Market Dynamics
The global payment processing market is projected to reach $914.91 billion by 2034, growing at 20.30% CAGR. Emerging markets in Latin America, Africa, and Southeast Asia show the highest growth potential driven by smartphone adoption and financial inclusion efforts.
Competitive Landscape
The sector is moderately fragmented with global giants (Stripe, Adyen) focusing on developed markets while specialists like DLocal dominate emerging market niches. High regulatory barriers create meaningful moats—Brazil's Open Finance processes 1.5 billion weekly API calls, showing rapid digitization.
Key Risks
FX volatility, regulatory changes, fraud/security breaches, and potential disruption from CBDCs or big tech entrants pose sector-wide threats.
3-5 Year Outlook
Latin America's fintech market alone is projected to reach $102.16 billion by 2030 (6.44% CAGR). Expect continued emerging market digitization, consolidation among regional players, and expansion into value-added services beyond pure transaction processing.
Key Monitoring Indicators
Digital payment penetration rates in emerging markets
Cross-border e-commerce growth from developed to emerging markets
Regulatory environment changes affecting payment licensing and open banking implementation
3. Financials and Valuation:
3-Year Financial Performance
DLocal achieved $746M revenue in 2024 with 14.7% growth, showing deceleration from earlier explosive growth. Gross margins compressed to 39% in Q1 2025 from 41% in Q4 2024, reflecting competitive pressures and scale investments.
Margin Trends: Operating leverage has been inconsistent. Q2 2024 operating income dropped 37% YoY despite revenue growth, with operating expenses surging 72%, indicating aggressive investment phase.
Cash Position: Strong balance sheet with $425M cash as of December 2024, providing financial flexibility.
Valuation Metrics
DLocal trades at 17-21x P/E ratio, which appears attractive versus peer average of 40.1x. Average analyst price target of $12.18 suggests 15% upside with "Buy" consensus.
Fair Value Assessment
The discount to payment processor peers reflects growth deceleration and margin pressure concerns. However, the valuation appears reasonable given DLocal's emerging market exposure and regulatory moats.
Undervalued if: Emerging market digitization accelerates, margins stabilize above 35%
Overvalued if: Competition intensifies, regulatory advantages erode, or EM growth disappoints
Current pricing appears fair, reflecting balanced risk-reward in a transitioning business model.
Found this content valuable? Share it with your network! Help others discover these insights by sharing the newsletter. Your support makes all the difference!
4. Risks and Perspectives:
Regulatory Risk (High Probability, High Impact): Brazil regulatory changes already impacted a top merchant, reducing credit card volumes.
Monitor: License renewals, local compliance changes, cross-border payment regulations.
Client Concentration Risk (Medium Probability, High Impact): Reliance on few major clients threatens revenue stability.
Monitor: Customer diversification metrics, top-10 client revenue mix.
FX Volatility (High Probability, Medium Impact): Currency fluctuations affect cross-border transactions; potential FX spread tightening could impact performance.
Monitor: Emerging market currency stability, hedging ratios.
Margin Pressure (Medium Probability, Medium Impact): Q1 2025 gross margins dipped to 39% from 41%.
Monitor: Take rates, competitive pricing dynamics.
3-5 Year Positioning
Opportunities: 70% of unbanked population lives in emerging economies with exploding digital payments. Q1 2025 showed record TPV of $8B (+53% YoY).
Threats: Rising protectionism, digital currency adoption, intensifying competition from larger payment processors.
Make-or-Break Risks: Brazilian regulatory stability, top merchant retention
Realistic Opportunities: Geographic expansion, value-added services
Thesis Invalidation: Loss of regulatory licenses, major client departures, sustained margin compression below 35%
5. Earnings Calls:
Q4 2024: Missed EPS expectations ($0.10 vs $0.15), but achieved 45% TPV growth. Management warned of continued take rate compression but expects it to be "more muted" going forward, not a "rapid race to the bottom".
Q3 2024: Record TPV despite Brazilian regulatory challenges and currency devaluation. Management "cautiously optimistic" about Q4.
Q2 2024: Strong TPV ($6B, +38% YoY) but revenue growth slowed (6% YoY) with 100% net revenue retention.
Recurring Patterns
Consistent: Record TPV growth, emerging market expansion, regulatory navigation
Shifting Tone: Increasing caution about margin compression, more emphasis on diversification
Key Developments & Concerns
Positive: Sustained volume growth, geographic expansion, strong balance sheet
Concerning: Accelerating take rate compression, EPS misses, regulatory headwinds in key markets
Critical Questions for Next Call:
What's the floor for take rate compression?
Brazil regulatory impact timeline?
Client concentration reduction progress?
Margin stabilization timeline?
Geographic diversification strategy specifics?
Enjoying the content? Don’t miss out on more exclusive insights and analyses. Subscribe now and stay updated.
Conclusion:
After years of reading filings the hard way, I see Claude for what it is: a support tool.
It won’t replace judgment or do the job for me, but it makes the work lighter.
Parsing a 200-page 10-K, organizing comps or summarizing calls - it does the heavy lifting so I can focus on decisions.
It’s not perfect and it never will be. But as a research partner, it earns its place.
If you enjoyed this article, please share it with a friend who might find it useful.
Thanks for sticking with me until the end. See you in the next one.
Cheers
Jimmy
Disclaimer
As a reader of Jimmy’s Journal, you agree with our disclaimer. You can read the full disclaimer here.
Fantastic breakdown, Jimmy. Your explanation of ROIC and ROIIC really highlights why high-quality businesses keep compounding over long periods. I love how you connect profitability, capital efficiency, and competitive advantage into one clear framework. The reminder that growth only creates value when ROIC exceeds the cost of capital is spot-on, simple, but so often overlooked. Thanks for continuing to make complex investing concepts incredibly accessible.
Hi Jimmy, I'm wondering if you've tried using the same prompts for each LLMs (ChatGPT vs Gemini vs Fiscal AI vs Claude) to compare the output produced side by side? Also, I'm curious to know how you balance the use of Claude as a supporting tool vs manually reading 10-Ks?