M&A Tax Strategy Builder

Before we dive into prompts, here’s the flow in plain English. You’ll give ChatGPT your key inputs once. Prompt 1 collects them all and grabs fresh IRS and state context. After that, the system uses those inputs to build the model, the negotiation playbook, and your decision memo. Your outcome: an 8–10 slide executive deck plus a 2-page appendix, built step by step. Have these ready: User inputs: One-page purchase summary: headline price, payment mix, expected close date Entity type and ownership: C/S corp/LLC taxed as corp/partnership, owner percents, owner states Allocation outline or asset schedule: inventory, fixed assets, identifiable intangibles, goodwill, liabilities assumed Tax attributes: NOLs and limitations, QSBS facts, audits or open positions State footprint: formation state, operating states, nexus/apportionment notes Existing LOI terms affecting structure: elections, escrows, reps, indemnity caps [if any] Last two years financials or tax depreciation schedules [if available] Integration intent: buyer’s hold period and step-up value horizon Auto web pulls (the model will fetch): IRC §§338(h)(10), 336(e), 1202 operative text and Form 8023 instructions Recent IRS/Treasury guidance or notable Tax Court decisions on these elections and QSBS Current federal corporate and individual rate tables and named state rates Authoritative PPA guidance and sample allocation templates or safe-harbor references 💡 Pro tip Answer in natural language or voice notes. Don’t polish... substance beats style. How the System Works Short rails. Real outputs. No fluff. Deal-phase diligence: capture facts and rank unknowns by tax-dollar impact. Valuation & structuring matrix: model Asset vs Stock vs 338/336 with sensitivities. Negotiation & integration runbook: translate math to talk-tracks, reps/escrows, and a doc checklist. (If we need a fourth: fold in QSBS and state quirks so they don’t bite you later.) Your Task: Use These Prompts Paste each prompt in order. Use GPT-4 class with browsing on for P1 and P2. The rest runs offline from your collected inputs. Prompt 1 — Collect & Search Goal: Build a one-page fact pack and grab the latest rules and rates. Why: Model garbage in: garbage out. We capture the minimum set once, plus authoritative web context. > Action: Copy, paste, fill in the bold fields, then run. DESCRIPTION Create a concise one-page deal fact pack and pull authoritative tax context. Prioritize unknowns by potential tax-dollar impact. TASK Assemble inputs for Asset vs Stock vs 338/336 modeling and compile current IRS/state guidance and rates. ROLE You are DealIntel-GPT: a senior M&A tax advisor with web search and summarization skills. INPUTS * PURCHASE SUMMARY: [Price, currency, expected close date] * PAYMENT MIX: [Cash at close %, Escrow %, Seller note %, Earnout %, timing terms] * ENTITY & OWNERSHIP: [Entity type, tax status, % owners, owner states/residency] * DEAL FORM INTENT: [Asset sale vs Stock sale preference if any, election appetite] * ASSET SCHEDULE OR OUTLINE: [Inventory, fixed assets (book/tax lives if known), intangibles, goodwill, liabilities assumed] * TAX ATTRIBUTES: [Federal/state NOLs and limits, QSBS facts, audits/open positions] * STATE FOOTPRINT: [Formation state, operating states, nexus/apportionment notes] * FINANCIALS/DEPR SCHEDULES: [Links or note “N/A”] * INTEGRATION INTENT: [Buyer hold period, step-up use horizon] Auto Web Pulls: Fetch and summarize only what materially improves output; prefer last 12–24 months. 1. IRC §§338(h)(10), 336(e), 1202 and Form 8023 instructions (official sources) 2. Recent IRS/Treasury guidance and notable Tax Court decisions affecting these elections/QSBS 3. Current federal corporate and individual rates for the current tax year 4. State corporate/individual rates for named states and key quirks (apportionment, franchise, addbacks) 5. Authoritative PPA guidance and sample templates/safe-harbor references WORKFLOW (no steps skipped) 1. Normalize inputs into a one-page Deal Snapshot. Flag any missing or ambiguous items. 2. Rank unknowns by potential tax-dollar impact at close and over five years. Show why each matters. 3. Browse official/primary sources for the listed web pulls. Capture only operative text and rule-of-thumb implications. 4. Extract current rate tables: federal + named states. Note any material recent changes. 5. Produce a short Web Context Digest with inline citations (Source, Year). 6. Output a clean fact pack + prioritized unknowns + digest. OUTPUT FORMAT Title: Deal Fact Pack A) Deal Snapshot * Price and Timing: ... * Payment Mix: ... * Entity & Ownership: ... * Intent and Constraints: ... * Asset Outline: ... * Tax Attributes: ... * State Footprint: ... * Integration Intent: ... B) Prioritized Unknowns (Ranked by Estimated $ Impact) 1. Unknown: ... Why it matters: ... Likely range: ... Validation doc(s): ... 2. ... (Include 6–12 items if applicable.) C) Web Context Digest (short, authoritative) * §338(h)(10) operative points: ... (IRS, 2024) * §336(e) operative points: ... (IRS, 2024) * §1202 QSBS notes: ... (IRS/Treasury, 2023) * Form 8023 highlights: filing, timing, parties ... (IRS, 2024) * Federal rates snapshot: ... (IRS, 2025) * State rates/quirks: [State]: ... (State DOR, 2024) * PPA guidance/safe-harbor: ... (Rev. Proc./ASC ref, Year) D) Assumptions (Replace when docs arrive) RULES * No chain-of-thought. Be concise and factual. * Prefer official sources; include short inline citations (Source, Year). * U.S. deals only. Mask PII. No legal advice. * Keep Web Context Digest ≤ 250 words total. FALLBACK If browsing is unavailable: output a Web Pull Plan with source URLs as [PLACEHOLDER], and proceed with the fact pack using clearly labeled [ASSUMPTION] notes. Instruct the user to rerun with browsing to replace placeholders. ❕ Quality gate If you can’t estimate the $ impact for an unknown, give a low/base/high bracket and mark it as [Range]. This is deal triage for grown-ups: capture the facts, fetch the rules, price the tax, and script the asks. It belongs in your LOI-to-definitives window where ambiguity kills leverage and speed wins concessions. Make structure a choice, not a debate. Then go get your terms.