AI Prompts for Investment Research: Earnings Analysis, Market Trends, and Due Diligence

Investment research has always been about information processing — reading filings, analyzing earnings, tracking industry trends, and building the mosaic of data that supports an investment thesis. The volume of available information has grown enormously. The analyst's time has not.

AI is useful here not because it has opinions about markets — it does not — but because it processes and structures information faster than manual reading. It can summarize a 200-page 10-K, organize an earnings call transcript by theme, or structure a due diligence checklist from a set of criteria. The analytical judgment remains entirely with the researcher.

This guide covers three investment research workflows:

• Earnings analysis — quarterly results frameworks, transcript analysis, and multi-quarter pattern identification
• Industry and market trends — sector overviews, competitive landscape mapping, and thematic research briefs
• Due diligence — comprehensive checklists, 10-K deep dives, management assessment, and thesis construction

Each section provides prompt templates with [PLACEHOLDERS] for your specific data. The prompts are designed to structure your research, not to generate conclusions. You provide the source material and judgment; the AI provides the organizational framework.

Every prompt works with ChatGPT, Claude, Gemini, or any general-purpose LLM. Build custom research prompts with the AI Prompt Generator.

Earnings Analysis Prompts

Earnings season generates massive amounts of structured data — financial results, management commentary, guidance updates, analyst Q&A. These prompts help organize that information into usable analysis.

What AI can do with earnings data:

• Summarize earnings call transcripts by theme
• Compare reported results to consensus estimates you provide
• Structure financial data into consistent analysis frameworks
• Identify changes in management language between periods
• Generate initial questions and areas for deeper research

What AI cannot do:

• Access real-time earnings data, analyst estimates, or stock prices
• Determine whether a stock is overvalued or undervalued
• Predict future earnings or price movements
• Provide proprietary consensus estimates or market sentiment data
• Replace the judgment of a trained investment analyst

The prompts below follow a consistent pattern: you provide the raw data (financial results, transcripts, estimates), specify the analytical framework, and the AI structures the output. This keeps you in control of what data enters the analysis and ensures the AI is organizing information rather than generating it.

1. Earnings Report Analysis Framework

You are an equity research analyst reviewing quarterly results for [COMPANY] ([TICKER]).

QUARTER: [Q_ FY____]

I will provide: key financial results, guidance updates, and consensus expectations.

[PASTE FINANCIAL RESULTS]
[PASTE GUIDANCE]
[PASTE CONSENSUS — e.g., "Street expected revenue of $X and EPS of $Y"]

Analyze using this framework:

1. HEADLINE (3 sentences) — beat/miss on revenue and earnings, most important data point, guidance direction
2. REVENUE — total vs. expectation, segment breakdown, revenue quality (recurring vs. one-time, organic vs. acquired)
3. PROFITABILITY — gross margin trajectory, operating margin vs. prior periods, one-time items to flag
4. BALANCE SHEET & CASH FLOW — cash/debt/working capital changes, capital allocation, free cash flow
5. MANAGEMENT THEMES — what they emphasized, what they avoided, tone vs. prior quarter
6. QUESTIONS FOR FURTHER RESEARCH — 3-5 specific follow-ups

No buy/sell/hold recommendations. Facts and structured observations only.

Why this prompt works: It separates analysis into discrete sections that mirror how analysts actually process earnings — headline first, then drilling into revenue, profitability, balance sheet, and management commentary. The "questions for further research" section ensures the output opens new lines of inquiry rather than closing the analysis prematurely.

2. Earnings Call Transcript Analysis

You are a research analyst extracting insights from an earnings call.

COMPANY: [COMPANY] ([TICKER]) | QUARTER: [Q_ FY____]

[PASTE TRANSCRIPT]

Provide:

1. PREPARED REMARKS — CEO key messages (3-4 bullets), CFO financial commentary (3-4 bullets)
2. STRATEGIC THEMES — 3-5 themes with direct quotes; note whether each is new, recurring, or shifted
3. Q&A ANALYSIS — most-asked topics, which got direct vs. evasive answers (evasion is informative)
4. FORWARD-LOOKING STATEMENTS — all specific targets or commitments, any guidance changes
5. LANGUAGE SIGNALS — confidence shifts ("we expect" vs. "we hope"), topics dropped from prior quarter, new topics introduced

Include direct quotes with references where the transcript supports them.

3. Earnings Pattern Analysis

You are a quantitative analyst studying earnings patterns for [COMPANY] ([TICKER]).

[PASTE DATA — for each of [NUMBER] quarters: reported EPS, consensus, revenue reported, revenue estimate, guidance]

Analyze:
1. BEAT/MISS HISTORY — table per quarter, average surprise %, conservative guidance pattern
2. GUIDANCE ACCURACY — compare forward guidance to actuals, average error, over/under tendency
3. MARGIN TRENDS — gross and operating margin trajectory, quarters outside typical range
4. SEASONAL PATTERNS — consistent seasonality, strongest/weakest quarters, guidance timing patterns
5. KEY OBSERVATIONS — 3-4 points on earnings quality and predictability

Present data in tables. Verify all percentages and averages.

How to chain these earnings prompts: Start with Prompt 1 for the financial results, then run Prompt 2 on the transcript to understand the qualitative context. If the company has been in your coverage for multiple quarters, run Prompt 3 to identify patterns. The combination of quantitative results, qualitative commentary, and historical patterns gives you a much more complete picture than any single analysis pass.

Industry and Market Trend Prompts

Understanding the broader industry and market context is essential for evaluating any individual investment. These prompts help structure industry research from publicly available information.

A critical note on AI and market data: AI models do not have access to real-time market prices, current analyst estimates, or live economic data. Any market figures, growth rates, or statistical claims in AI output should be treated as approximate and requiring verification against current sources — Bloomberg, FactSet, S&P Capital IQ, or the relevant industry research provider.

The prompts below work best when you have already identified the industry or theme of interest and need help organizing the research. They are structuring tools for information you will verify, not substitutes for primary research.

4. Industry Landscape Overview

You are an industry analyst preparing a sector overview for [INDUSTRY/SECTOR].

PURPOSE: [screening opportunities / evaluating competitive positioning / due diligence context]

Structure:
1. MARKET STRUCTURE — approximate size and growth (note source year), concentration, key segments
2. VALUE CHAIN — primary steps, where value is captured, dependencies
3. COMPETITIVE DYNAMICS — Porter's Five Forces, each rated HIGH/MEDIUM/LOW with one-sentence justification
4. SECULAR TRENDS — 3-5 major trends, what drives each, who benefits, who's at risk
5. KEY METRICS — what analysts track and healthy benchmarks

[Any market figures require verification against current industry reports. AI may present approximate or outdated data as precise.]

Label uncertain figures as "approximate — verify against current sources."

5. Competitive Landscape Mapping

You are an analyst mapping the competitive landscape for [COMPANY] in [INDUSTRY].

COMPETITORS: [LIST 4-6]

For each company:
1. OVERVIEW — 2-3 sentence business model
2. SCALE — revenue, employees, geography (note data may be outdated)
3. COMPETITIVE ADVANTAGE — primary moat (cost, network effects, switching costs, brand, tech, regulatory)
4. VULNERABILITY — single biggest competitive risk
5. RECENT MOVES — significant actions in past 12 months

Then provide:
A. POSITIONING — where each sits on two relevant dimensions (e.g., premium vs. value, breadth vs. specialization)
B. SHARE DYNAMICS — who's gaining, losing, or stable
C. WHITE SPACE — 1-2 underserved segments

[Verify all company-specific claims against current filings and press releases.]

Tip: Run this prompt once with publicly known competitors, then ask a follow-up prompt: "Based on this landscape, what types of companies would represent the biggest competitive threat if they entered this market?" This second pass often surfaces risks that a static competitor list misses.

6. Thematic Research Brief

You are a thematic analyst preparing a brief on [INVESTMENT THEME].

AUDIENCE: [portfolio managers / investment committee / client-facing]

Structure:
1. DEFINITION — what it is, what's driving it, time horizon
2. MARKET SIZING — TAM framework, estimation method, key assumptions
   [Do not fabricate numbers. If uncertain, say "source from [PROVIDERS]."]
3. VALUE CHAIN — beneficiaries categorized as pure plays, diversified, picks-and-shovels, tangential
4. INVESTMENT CONSIDERATIONS — catalysts, risks, where consensus might be wrong, what disconfirms the thesis
5. MONITORING — 5-7 indicators to track, data sources, what changes would shift the view

Balanced between bull and bear. No buy/sell recommendations or price targets.

When to use thematic briefs: These work best at the beginning of a research process — when you are trying to understand a space before diving into individual companies. Run the thematic brief first, identify the most interesting part of the value chain, then use the competitive landscape prompt to map the players in that specific segment.

Due Diligence Prompts

Due diligence is the most structured phase of investment research. Whether evaluating a potential acquisition target, a private investment, or a new portfolio position, AI can help organize the research process and ensure comprehensive coverage. The prompts below generate checklists and frameworks — the actual due diligence work of verifying claims, analyzing data, and making judgments remains entirely with the analyst.

7. Due Diligence Checklist

You are a due diligence analyst creating a research checklist for [COMPANY] as a potential [INVESTMENT TYPE].

Context: Industry: [INDUSTRY] | Stage: [public/private/early] | Size: [AMOUNT] | Horizon: [HOLDING PERIOD]

Checklist:

1. BUSINESS MODEL — revenue model, customer concentration, unit economics, competitive position, management team
2. FINANCIAL — 3-5yr performance, revenue quality, margins, working capital, capital structure, cash flow
3. MARKET — TAM, market share, competitive dynamics, regulatory environment
4. RISKS — key person, technology, regulatory, concentration, legal/litigation, ESG
5. VALUATION — comparable companies, precedent transactions, DCF assumptions, key sensitivities
6. LEGAL/STRUCTURAL — governance, IP, material contracts, litigation, insurance

Each item: question to answer, suggested data source, [STATUS] and [ANALYST] placeholders.

Customization tip: The checklist above is a general framework. For specific investment types, add domain-specific sections: technology due diligence for software companies, regulatory pathway analysis for biotech, reserve analysis for oil and gas, or same-store sales analysis for retail. The more specific the checklist, the more useful the output.

8. 10-K Section Analysis

You are a research analyst reading the [SECTION — e.g., Item 1, Item 1A, Item 7] of [COMPANY]'s ([TICKER]) 10-K for FY[YEAR].

[PASTE THE SECTION]

Provide:
1. KEY FACTS — 5-7 most important data points, one sentence each
2. NOTABLE LANGUAGE — unusually specific, hedged, new vs. boilerplate, or quantitative where peers are vague
3. RISKS — disclosed or implied, categorized as well-known/priced-in, company-specific, or potentially under-appreciated
4. FOLLOW-UP — 3-5 questions requiring other sources
5. YEAR-OVER-YEAR CHANGES (if prior year provided) — additions, deletions, wording shifts, new risk factors

Cite specific paragraphs from the text provided.

How to use this iteratively: Process the 10-K section by section rather than pasting the entire document. Start with Item 1A (Risk Factors) to understand what the company is worried about, then move to Item 7 (MD&A) for the financial narrative, then Item 1 (Business) for the operational overview. Each pass builds on the context from prior sections.

9. Management Team Assessment

You are a research analyst evaluating the management team of [COMPANY] for investment due diligence.

EXECUTIVES: [LIST KEY NAMES, TITLES, TENURE, AND KNOWN BACKGROUND]

Assess:
1. COMPOSITION — do key roles have relevant experience? Average tenure? Recent C-suite departures?
2. TRACK RECORD — 2-3 sentence professional history per executive; have they scaled this type of business before?
3. ALIGNMENT — insider ownership level (I'll provide proxy data), compensation structure, recent buying/selling
4. QUALITY SIGNALS — communication quality on calls, forecast accuracy, capital allocation track record, transparency about mistakes
5. RED FLAGS — turnover, related-party transactions, excessive comp, restatement history, promotional language without substance

[Verify executive backgrounds against SEC filings and public records. AI may generate inaccurate biographical details.]

Present evidence, not ratings. Let the analyst draw conclusions.

Why management assessment matters: Management quality is one of the most important and hardest-to-quantify factors in investment research. This prompt does not attempt to score management — instead, it organizes the evidence so you can form your own view. Pay particular attention to the gap between what management says and what the data shows. The most useful signal is often the inconsistency between confident forward guidance and deteriorating operational metrics.

10. Investment Thesis Construction

You are a senior analyst structuring a thesis for [COMPANY] ([TICKER]).

PURPOSE: [initiating coverage / investment committee / position paper]

I will provide: [FINANCIAL DATA] | [INDUSTRY CONTEXT] | [MY PRELIMINARY VIEWS]

Structure:

1. THESIS STATEMENT — one paragraph, why the opportunity exists
2. INVESTMENT MERITS (3-5) — argument, evidence, and data that would strengthen it
3. KEY RISKS (3-5) — description, likelihood (qualitative), mitigants, and disqualifying threshold
4. VARIANT PERCEPTION — consensus view, where this thesis disagrees, what isn't priced in
5. CATALYSTS — near-term (0-6mo), medium-term (6-18mo), long-term (18mo+)
6. DISCONFIRMATION — what developments invalidate the thesis, monitoring data points
7. INFORMATION GAPS — unanswered questions, additional research needed

Rigorous and balanced. Present bull and bear fairly. No price targets or buy/sell/hold.

The thesis construction prompt is a synthesis tool, not a starting point. Run it after you have already done the fundamental work — analyzed the earnings, read the filings, mapped the competitive landscape, and formed preliminary views. The AI structures your thinking; it does not generate the thinking for you. The most valuable output from this prompt is often the "information gaps" section — it highlights what you have not yet verified, which is exactly what due diligence should surface.

Working With AI in Investment Research

AI works best in investment research when it is embedded into your existing process as a productivity layer, not as a replacement for any step.

The goal is not to produce AI-generated research. The goal is to use AI to do the structured, repetitive parts of research faster so you can spend more time on the judgment-intensive parts that actually differentiate your analysis. Three specific limitations to keep in mind:

Temporal Blindness

AI models have a training data cutoff. They do not know about last quarter's earnings, yesterday's FDA approval, or this morning's market movement. Always provide current data yourself rather than asking AI what happened recently. The prompts above are designed with this in mind — they ask you to paste data rather than asking AI to retrieve it.

False Precision

AI will generate specific-sounding numbers — market sizes, growth rates, competitive shares — that may be approximations, outdated, or entirely fabricated. Treat any AI-generated statistic as "requires source verification" by default. This is especially dangerous in investment research, where precise numbers matter.

Confidence Without Basis

AI models generate text that sounds confident even when the underlying analysis is weak. Maintain the same skepticism you would apply to a junior analyst's first draft — the structure may be sound, but every substantive claim needs a source.

This is particularly dangerous in investment research because confident-sounding analysis can influence decision-making even when the underlying data is wrong. Build a habit of asking "what is the source for this claim?" for every factual statement in AI output.

The Research Assembly Line

A practical workflow for integrating these prompts:

• Data gathering (human) — Pull the 10-K, earnings transcript, and financial data from primary sources
• Initial structuring (AI-assisted) — Use the prompts above to organize raw data into analysis frameworks
• Deep analysis (human) — Read AI output critically, challenge assumptions, verify calculations
• Synthesis (AI-assisted) — Use the thesis construction prompt to organize your conclusions
• Review and distribution (human) — All output goes through compliance review before any external use

Compliance Considerations

Before integrating AI into your research workflow, confirm your firm's policies on:

• Which AI tools are approved for use (consumer vs. enterprise tier)
• What types of data can be input (public filings vs. proprietary models vs. client data)
• Disclosure requirements for AI-assisted research in client-facing materials
• Record-keeping obligations for AI inputs and outputs used in investment decisions

These policies vary significantly across firms and regulatory jurisdictions. When in doubt, consult your compliance department before using AI with any investment-related data.

For more templates, explore our AI prompts for finance and business prompts collection. Build custom research prompts with the AI Prompt Generator.

FAQ

Can AI predict stock prices or market movements?

No. AI language models are not predictive financial models. They do not have access to real-time market data, order flow, or proprietary signals. These prompts help structure analysis of data you provide — they do not generate forecasts, price targets, or trading signals. Any forward-looking output from AI is speculative narrative, not quantitative prediction.

Is AI-generated investment research suitable for client distribution?

Not without significant human review and compliance oversight. AI output may contain errors, unsupported claims, or outdated information. Any research distributed to clients or used in investment decision-making must be reviewed by a qualified analyst, verified against primary sources, and cleared through your firm's compliance process. AI is a drafting tool, not an authoring tool for regulated communications.

How should I handle material non-public information when using AI?

Never input material non-public information (MNPI) into any AI tool, including enterprise-tier products. MNPI includes earnings previews, M&A discussions, insider knowledge, and any information that has not been publicly disclosed. AI tools process data on remote servers and may log inputs. Inputting MNPI into AI tools could constitute a violation of insider trading regulations and your firm's information barriers.