Prompt 1 — Full DCF Valve (From Historicals to Fair Value)

Prompt text (production-ready):

Act as a professional equity analyst. Using the provided historical financials and market data, build a 10-year discounted cash flow (DCF) model. Provide (1) projected unlevered free cash flows for each year, (2) a WACC calculation with components, (3) terminal value using both Gordon Growth and Exit Multiple methods, (4) enterprise value (EV), equity value and per-share fair value. Output MUST be valid JSON with keys: assumptions, projections (yearly), wacc, terminal_values, valuation, sensitivity_table. Include calculation formulae and any rounding instructions. Do not include extraneous commentary.

Context variables (supply these)

• company_name: “Acme Co.”
• fiscal_year_end: “2025-12-31”
• historical_revenue: [2021: 2,100,000; 2022: 2,450,000; 2023: 2,900,000; 2024: 3,300,000] (in USD)
• historical_operating_income: [ … ]
• capex_schedule: {“2025”: 120000, “2026”: 150000, …}
• depreciation_schedule, working_capital_rules (days receivable, payable), tax_rate (e.g., 21%)
• risk_free_rate (example 4.10% — replace with real-time), market_premium (6.0%), beta (1.05), pre-tax_cost_of_debt (6.5%), target_debt/total_capital (25%)
• terminal_GG_growth (example 2.5%) and exit_multiple (example 10.5x NTM EBITDA)
• shares_outstanding (excl. dilutive instruments) and net_debt (cash minus debt)

Expected output format (JSON schema)

{
  "assumptions": {
    "discount_rate_description": "WACC (inputs...)",
    "tax_rate": 0.21,
    "growth_assumptions": {"year1":0.08,...}
  },
  "projections": [
    {"year":2026,"revenue":3300000,"unlevered_fcf":265000,...},
    ...
  ],
  "wacc": {"cost_of_equity":0.085,"cost_of_debt_after_tax":0.0525,"wacc":0.075},
  "terminal_values": {"gordon_growth": 123456789, "exit_multiple": 130000000},
  "valuation": {"enterprise_value": 250000000,"equity_value": 200000000,"per_share": 24.50},
  "sensitivity_table": {"g_growth": {"1.5%":..., "2.5%":...}, "wacc": {"7.0%":...}}
}

Usage notes

• Set temperature to 0.0 to avoid variances in numeric formatting.
• Validate that the model’s implied terminal value does not exceed logical industry caps (e.g., terminal EV/EBITDA > 30x requires review).
• For large models, request outputs in pages (split decades into multi-call workflow).

Example input snippet

{
  "company_name":"Acme Co.",
  "fiscal_year_end":"2025-12-31",
  "historical_revenue":{"2021":2100000,"2022":2450000,"2023":2900000,"2024":3300000},
  "tax_rate":0.21,
  "shares_outstanding":8200000,
  "net_debt":50000000,
  "risk_free_rate":0.041,
  "market_premium":0.06,
  "beta":1.05
}

Example output (abridged)

{
  "assumptions":{"tax_rate":0.21,"terminal_growth":0.025,"wacc":0.075},
  "projections":[{"year":2026,"unlevered_fcf":265000},{"year":2027,"unlevered_fcf":285500}...],
  "wacc":{"cost_of_equity":0.085,"cost_of_debt_after_tax":0.0525,"wacc":0.075},
  "terminal_values":{"gordon_growth":125000000,"exit_multiple":138000000},
  "valuation":{"enterprise_value":260000000,"equity_value":210000000,"per_share":25.61}
}

Prompt 2 — WACC Calculator with Sensitivity Grid

Prompt text:

Compute a detailed WACC for the target using the inputs below. Show the cost of equity using the CAPM formula (cost_of_equity = risk_free_rate + beta * market_premium), cost of debt (pre- and post-tax), and weightings. Output a sensitivity grid (matrix) showing WACC across beta (0.8 to 1.6 step 0.1) and market_premium (4% to 8% step 0.5%). Provide a CSV-compatible table with headers: beta, market_premium, wacc.

Context variables

• risk_free_rate (example 4.10%)
• base_beta (1.05)
• market_premium_base (6.0%)
• pre_tax_cost_of_debt (6.5%)
• tax_rate (21%)
• debt_weight (25%), equity_weight (75%)

Expected output

{
  "base_wacc":0.075,
  "components":{"cost_of_equity":0.085,"cost_of_debt_pre_tax":0.065,"cost_of_debt_after_tax":0.05135},
  "sensitivity_grid": [
    {"beta":0.8,"market_premium":0.04,"wacc":0.0623},
    {"beta":0.9,"market_premium":0.045,"wacc":0.0644}, ...
  ]
}

Usage tips

• Supply current market_premium and risk-free rate from a trusted data source on the run date.
• For leverage changes, re-run with adjusted debt weights and recalculate effective tax shield contribution.

Prompt 3 — Terminal Value Method Comparison

Prompt text:

Given projected year-N EBITDA and the company's peer group multiples, compute terminal value using both (a) Gordon Growth Model (specify growth rate and WACC) and (b) Exit Multiple approach. For the multiple method, normalize peer EV/EBITDA by median and 25th/75th percentiles. Return a comparison table with implied per-share values for three scenarios: conservative (25th percentile), base (median), aggressive (75th percentile). Output structured JSON and include the peer multiple list used.

Context variables

• last_projected_ebitda (example 2029 EBITDA = 120,000,000)
• peer_ev_ebitda_list (e.g., [8.5, 9.8, 11.2, 14.0, 10.3])
• wacc, terminal_growth
• shares_outstanding, net_debt

Expected output

{
  "peer_stats": {"25th":9.3,"median":10.3,"75th":11.6},
  "terminal_values": {"conservative":950000000,"base":1236000000,"aggressive":1392000000},
  "per_share": {"conservative":12.34,"base":16.04,"aggressive":18.08}
}

Usage notes

• Flag cases where the exit multiple implies unrealistic growth relative to GDP/industry (~>20-year CAGR).
• If peer sample <5 companies, request the model to expand the peer list or return a confidence flag.

Prompt 4 — FCF Bridge & Three-Statement Reconciliation

Prompt text:

Given income statement, balance sheet, and cash flow statement items (annual or quarterly), reconcile net income to unlevered free cash flow for each period. Output a step-by-step bridge table with each line item (net income, add: depreciation, less: capex, change in working capital, add/subtract taxes saved, etc.). Output must be a CSV-friendly table and JSON array with keys: period, net_income, d_and_a, capex, working_cap_delta, unlevered_fcf.

Context variables

• periods: [“2023-Q4″,”2024-Q1”,”2024-Q2″…]
• net_income, depreciation_and_amortization, capex, changes_in_working_capital, interest_expense (for tax adjustments)

Expected output

[
  {"period":"2024","net_income":34000000,"d&a":6500000,"capex":12000000,"wc_change":-1500000,"unlevered_fcf":37000000},
  ...
]

Usage notes

• Mark any missing or negative working capital line items clearly and return an error field if required data is missing.

Prompt 5 — Scenario & Sensitivity Batch Generator

Prompt text:

Create three full DCF scenarios (Base, Bear, Bull) given central assumptions. For each scenario, generate: projected FCF series (10-year), WACC, terminal value (Gordon growth), EV, equity value, per-share value, and a short narrative (2–3 sentences) explaining drivers. Output as JSON with scenario keys and include a sensitivity matrix for WACC vs. terminal growth for each scenario.

Context variables

• base_assumptions (growth rates, margins), downside_adjustments (-200bps margin), upside_adjustments (+150bps)
• scenario_probabilities (optional: base 50%, bear 30%, bull 20%)

Expected output

{
  "Base":{"fcf":[...],"wacc":0.075,"terminal":120000000,"equity_value":200000000,"per_share":24.4,"narrative":"Base case expects margin expansion..."},
  "Bear":{...},
  "Bull":{...}
}

Usage notes

• Use scenario probabilities to compute probability-weighted equity value when requested.
• Ensure narratives reference the concrete drivers (revenue growth, margin, capex) and avoid generic adjectives.

Prompt 6 — Earnings Call Summary & KPI Extraction

Prompt text:

Given a transcript of an earnings call (or bullet points), extract and summarize the top 8 actionable items in this output structure: {items: [{title, type (guidance/product/ops/finance), quantitative_impact_estimate, confidence (low/med/high), citation (timestamp or paragraph id)}]}. Also extract any updated guidance and numeric ranges. Output JSON only.

Context variables

• transcript_text (UTF-8, up to the model token limit)
• recent_guidance (previous guidance values for comparison)

Expected output

{
  "items":[
    {"title":"Revenue guide raised","type":"guidance","quantitative_impact_estimate":"+3.5% FY rev","confidence":"high","citation":"Q&A 12:34"},
    ...
  ],
  "updated_guidance":{"FY2026_revenue":[320000000,330000000]}
}

Usage notes

• When transcript length exceeds token window, chunk and aggregate with topic-based indexing—ask ChatGPT-5.5 to produce a master summary from stitched outputs.
• For quoted quantitative items, include exact timestamps to support data lineage.

Prompt 7 — Consensus vs. Reported Earnings Reconciliation

Prompt text:

Compare reported quarterly results with a provided consensus dataset. Produce a table with: line_item, consensus_value, reported_value, variance_absolute, variance_pct, and commentary explaining main drivers. Highlight non-GAAP adjustments (EPS, EBITDA adjustments) and quantify their impact on headline metrics. Output JSON and a CSV table string.

Context variables

• consensus (dictionary of line items: revenue, EPS, EBITDA)
• reported_results (same keys)

Expected output

{
  "comparison":[
    {"line_item":"revenue","consensus":320000000,"reported":327500000,"var_abs":7500000,"var_pct":2.34,"commentary":"higher ASP in North America"}
  ],
  "csv":"line_item,consensus,reported,var_abs,var_pct,commentary\n..."
}

Usage notes

• Require the model to flag any variance >2 standard deviations from the trailing 8-quarter mean as an ‘outlier’.

Prompt 8 — Non-GAAP to GAAP Reconciliation and Normalization

Prompt text:

Given reported non-GAAP metrics and disclosed adjustments, produce a reconciliation table to GAAP (EPS, EBITDA). Additionally, normalize by removing one-off items specified in the 'adjustments' list and show adjusted margins. Output JSON with fields: gaap_metric, non_gaap_metric, adjustments (list), normalized_metric.

Context variables

• non_gaap_items (list of named adjustments with amounts and descriptions)
• reported_gaap_metrics

Expected output

{
  "reconciliation":[
    {"metric":"net_income","gaap":35000000,"non_gaap":37000000,"adjustments":[{"name":"restructuring","amount":2000000}]}
  ],
  "normalized":{"adjusted_eps":1.24,"adjusted_ebitda_margin":0.185}
}

Usage notes

• Ask the model to flag adjustments exceeding 1% of revenue or those recurring across >3 periods as potentially structural.

Prompt 9 — Quarter-over-Quarter KPI Trend Analysis

Prompt text:

Analyze provided quarterly KPIs for the last 12 quarters. For each KPI (revenue, gross margin, AR days, churn, ARPU, etc.), compute QoQ and YoY percent changes, 4-quarter rolling average, and a Z-score relative to the 12-quarter mean. Output a JSON array with KPI analysis and a short one-sentence signal (improving/declining/stable) and recommended action.

Context variables

• kpi_time_series: {“revenue”:[…], “gross_margin”:[…], “churn”: […]}
• reporting_calendar to map quarters

Expected output

[
  {"kpi":"revenue","qoq_change":"1.8%","yoy_change":"6.5%","rolling_avg":"3.1%","z_score":1.4,"signal":"improving","action":"Monitor product pricing trends"}
]

Usage notes

• Use rolling averages to reduce seasonality noise and require the model to state the sample size used for statistics (n=12 for quarterly).

Prompt 10 — Earnings Note Draft (Analyst Distribution)

Prompt text:

Draft a 300–450 word earnings note suitable for client distribution. Use the provided numeric results and reconciliations. The structure must include: key facts (bulleted), financial highlights (table-like bullets), investment implications (3 bullets: catalyst, risk, valuation change), and a suggested analyst stance (Buy/Hold/Sell + price target and rationale). At the end, include a 2-line methodological note describing model version and data sources. Output JSON: {headline, key_facts, financial_highlights, implications, stance, methodology}.

Context variables

• headline_metrics (EPS, revenue, guidance change, percent variances)
• analyst_prior_rating_and_target

Expected output

{
  "headline":"Acme Co. beats Q2 estimates; raises guidance",
  "key_facts":["Revenue $327.5m (+2.3% q/q)","EPS $0.48 vs $0.42 est"],
  "financial_highlights":[{"line":"Revenue","value":"327.5"}],
  "implications":[{"catalyst":"New product ramping Q4","risk":"Supply chain constraints affecting margin"}],
  "stance":{"rating":"Buy","new_target":28.00},
  "methodology":"DCF v3.1; data: company release, consensus via Refinitiv; model run 2026-07-12"
}

Usage notes

• Specify desired word count and tone (concise neutral vs. sales-oriented) in the input for consistent output.

Prompt 11 — Factor Attribution for Returns

Prompt text:

Given monthly returns for a portfolio and benchmark over the last 60 months and factor exposures (e.g., market beta, size, value, momentum), conduct a performance attribution decomposition. Provide results in JSON with fields: total_return, benchmark_return, active_return, allocation_effect, selection_effect, interaction_effect, factor_contributions (list of factor:contribution). Include a short 3-line interpretation of top drivers.

Context variables

• portfolio_returns (monthly array)
• benchmark_returns (monthly array)
• factor_returns (matrix: factors x months)
• exposures (current exposure vector)

Expected output

{
  "total_return":0.182,
  "benchmark_return":0.145,
  "active_return":0.037,
  "factor_contributions":[{"market":0.025},{"size":0.004},{"value":-0.002}]
}

Usage notes

• Specify the attribution methodology (Brinson-Fachler or factor model), and ask the model to include the formula selection in the response.
• Cross-check numeric contributions sum to active return ± rounding tolerance (e.g., 1e-4).

Prompt 12 — Portfolio VaR & CVaR Estimation (Historical & Parametric)

Prompt text:

Estimate 1-day and 10-day Value at Risk (VaR) and Conditional VaR (CVaR) for the provided portfolio using (1) historical simulation (last 500 trading days) and (2) parametric (variance-covariance) approach assuming normal and t-distributed returns (df=5). Output JSON with VaR and CVaR values at 95% and 99% confidence levels, plus assumptions and a small code snippet (Python) showing the calculation used.

Context variables

• positions (list: ticker, weight, price series)
• return_series (daily returns aligned to tickers)
• holding_period_days (1,10)

Expected output

{
  "historical_VaR":{"1d":{"95%":0.024,"99%":0.042},"10d":{...}},
  "parametric_VaR_normal":{...},
  "parametric_VaR_tdist":{...},
  "python_snippet":"import numpy as np\n..."
}

Usage notes

• Ask the model to check for return stationarity and to signal when returns have structural breaks (e.g., regime change around 2020 pandemic) that invalidate simple historical simulation.

Prompt 13 — Stress Test Scenario Generator

Prompt text:

Generate five stress-test scenarios tailored to the client's portfolio and sector exposures. For each scenario supply: scenario_name, macro_shock_description (GDP, rates, FX, commodity moves), expected P&L impact (% portfolio NAV), rationale, and recommended hedging actions. Output JSON. Base scenarios on historically-observed extremes and plausible near-term risks (next 12 months as of July 2026).

Context variables

• sector_weights, geographic_exposure, leverage (if any)
• current_holding_period and liquidity constraints

Expected output

{
  "scenarios":[
    {"name":"Rapid Rate Spike","macro_shock":"10yr +150bps","pnl_impact":"-4.2%","hedge":"Increase duration hedges via swaps"}
  ]
}

Usage notes

• Require the model to provide an actionability score for each recommendation (1–5) and an estimated cost to implement hedges.

Prompt 14 — Portfolio Optimization Suggestion (Mean-Variance & Constraints)

Prompt text:

Given expected returns, covariance matrix, and constraints (min/max weight per asset, sector caps, turnover limits), propose an optimal portfolio allocation under mean-variance optimization for target risk levels (low/medium/high). Provide allocations for each risk level, expected return, expected volatility, and a list of active constraints that bind. Output JSON and include a convex optimization formulation summary (objective, constraints).

Context variables

• expected_returns (vector), covariance_matrix, current_weights, constraints (dict)
• risk_targets: {“low”: vol 6%,”medium”:10%,”high”:15%}

Expected output

{
  "allocations":{"low":{"AAPL":0.12,"MSFT":0.10,...},"medium":{...},"high":{...}},
  "metrics":{"low":{"exp_return":0.085,"vol":0.06} ...},
  "binding_constraints":["sector_cap_tech_30%"]
}

Usage notes

• For production, pass suggested allocations to an optimizer (cvxpy) and validate feasibility; use the model’s formulation as a human-readable plan rather than a numeric solver unless cross-validated.

Prompt 15 — Backtest Summary & Trade Blame Report

Prompt text:

Summarize a backtest of the strategy over the last N years. Produce: total return, annualized return, annualized volatility, max drawdown, Sharpe ratio (assume 1.5% risk-free baseline unless provided), and a 'trade blame' section listing top 10 trades/positions that most contributed to underperformance (by contribution to tracking error). Output JSON and a short action plan to address recurring failure modes.

Context variables

• backtest_trades (list: date, ticker, action, size, pnl)
• benchmark_returns

Expected output

{
  "backtest_metrics":{"total_return":0.42,"ann_return":0.073,"vol":0.12,"max_dd":0.23,"sharpe":0.47},
  "trade_blame":[{"trade_id":...,"contribution":-0.024,...}],
  "action_plan":["Reduce exposure to X during Y conditions"]
}

Usage notes

• Ask the model to compute contribution to active return and to rank trades by absolute impact for clarity in governance discussions.

Prompt 16 — Industry Landscape Snapshot (Top-Down)

Prompt text:

Produce a concise industry landscape snapshot for [industry_name] as of [date]. Include: market size (current and 5-year CAGR), top 5 competitors by revenue, key secular trends (3 bullets), regulatory risks (2 bullets), and an estimated TAM/SAM/SOM matrix with short justification. Output JSON and a 5-bullet executive summary for slides.

Context variables

• industry_name (e.g., “Electric Vehicle Charging”)
• date (e.g., “2026-07-01”)
• market_data_source_note (where you pulled numbers if known)

Expected output

{
  "market_size":{"2026":43.2e9,"cagr_2021_2026":0.24},
  "top_competitors":[{"name":"ChargeCo","revenue":2.4e9},...],
  "trends":["fleet electrification","grid constraints"],
  "tam_sam_som":{"tam":500e9,"sam":120e9,"som":15e9}
}

Usage notes

• When using real numbers, validate with primary sources (industry reports, government stats) and provide citations in the methodology field.

Prompt 17 — Peer Benchmarking Table & Valuation Multiples

Prompt text:

Construct a peer benchmarking table for the provided peer tickers with the following columns: ticker, market_cap, revenue_TTM, ebitda_TTM, ev, ev_ebitda, pe, net_debt_to_ebitda. Sort by market_cap descending. Output JSON array and a CSV string. Highlight peers with EV/EBITDA > 75th percentile and those with negative EBITDA.

Context variables

• peer_tickers (list)
• pricing_date (e.g., “2026-07-10”)
• data_provider (for traceability)

Expected output

[
  {"ticker":"ABC","market_cap":25000000000,"revenue_TTM":12000000000,"ev_ebitda":11.6,...}
]

Usage notes

• Request the model to normalize TTM vs. LTM definitions and to note any fiscal calendar mismatches (fiscal-year vs calendar-year).

Prompt 18 — Revenue Deck: Product & Channel Decomposition

Prompt text:

Breakdown total revenue into product lines and sales channels using the provided segment disclosures (past 4 quarters). For each segment, compute percent of revenue, 4-quarter CAGR, gross margin estimate, and comment on growth drivers. Provide an output table and JSON with segment line items and a short marketing/operational risk factor for each.

Context variables

• segment_disclosures (quarterly revenue per segment)
• segment_gross_margin_estimates (if available)

Expected output

{
  "segments":[{"name":"Hardware","pct_of_revenue":0.45,"4q_cagr":0.08,"gross_margin_est":0.18,"risk":"Component shortages"}]
}

Usage notes

• Flag segments with YoY/4q_cagr divergence greater than 20% for follow-up due diligence.

Prompt 19 — Catalyst Calendar & Event Impact Scoring

Prompt text:

Create a 12-month catalyst calendar for the company and sector with event_name, date_estimate, probability_of_occurrence (0-100%), expected_impact_on_price (%), and required_monitoring_actions. Rank events by potential impact and provide suggested positioning strategy for each event (hedge, increase, reduce, monitor).

Context variables

• corporate_events (e.g., product_launch Q4 2026, regulatory_decision Q3 2026)
• sector_events (e.g., rate decisions, industry conferences)

Expected output

{
  "calendar":[{"event":"Q3 earnings","date":"2026-10-22","prob":100,"impact":2.5,"action":"Monitor, avoid large changes within 3 days prior"}]
}

Usage notes

• Include a brief rationale (one sentence) for each probability and impact estimate referencing comparable events when possible.

Prompt 20 — Investment Memo Draft (Long/Short Thesis)

Prompt text:

Draft a 1,200–1,800 word investment memo with the following sections: Executive Summary, Investment Thesis (3 bullets), Valuation Summary (DCF and multiples), Key Catalysts (3), Key Risks (4, ranked), Sensitivity Analysis (table), and Recommended Positioning with target price and time horizon. Use provided model outputs and peer data. Output must be in JSON with 'sections' keyed to each subheading and a CSV-friendly sensitivity table.

Context variables

• model_outputs (valuation JSON from DCF prompts)
• peer_benchmark_table
• analyst_prior_view

Expected output

{
  "sections":{"executive_summary":"...", "investment_thesis":[...],"valuation_summary":{...},"recommendation":{"rating":"Buy","target":28,"horizon":"12 months"}}
}

Usage notes

• Set temperature 0.2–0.3 to allow concise narrative variety while keeping facts consistent. Insist the model include exact numeric sources for any factual assertions.