The close of a business combination is a moment of triumph—and, for most finance teams, the beginning of a uniquely stressful compliance exercise. Purchase Price Allocation (PPA) under ASC 805 or IFRS 3 requires you to take the total consideration paid and assign fair value to every identifiable tangible and intangible asset, as well as assumed liabilities. The gap between the purchase price and the sum of those fair values becomes goodwill. On paper, this sounds like a straightforward accounting entry. In practice, it is a minefield of valuation methodologies, interim useful life determinations, and audit scrutiny that can stretch for months.
The friction is not the math. The friction is the judgment. Your team must analyze dozens of intangibles—customer relationships, backlog, trade names, technology, non-compete agreements—and justify each fair value estimate against the Income Approach, Market Approach, or Cost Approach. Every assumption about discount rates, revenue attrition, and royalty rates is a potential audit finding. Meanwhile, the clock is ticking toward the 45-day measurement period, and the external valuation firm you hired is billing you by the hour. The CFO is asking for a high-level summary, the auditors are asking for source documents, and your spreadsheet model has twelve tabs of linked assumptions that no one else can follow.
This is where a large language model (LLM) changes the workflow. An AI tool does not replace the valuation expert, but it does eliminate the repetitive, error-prone scaffolding around the PPA: drafting the initial asset list from the purchase agreement, generating a first-pass intangible asset identification memo, stress-testing your assumptions for reasonableness, and preparing audit-ready documentation that explains the “why” behind each number. Instead of starting from a blank page, you start from a structured draft that reflects current valuation practice. Instead of manually reconciling the balance sheet, you ask the AI to flag inconsistencies. The result is a PPA process that is faster, more defensible, and far less prone to oversight.
To get that result, however, you cannot just type “do a PPA for me” into a chatbot. The output quality depends entirely on the quality of your instruction. You need to treat the AI like a new senior associate on your team—one who has read every valuation textbook but knows nothing about your specific deal. Your prompt must contain context, constraints, and a clear definition of success. Below is a structured prompt template that works for the initial intangible asset identification phase of a PPA.
First, read these files completely before responding:
[purchase_agreement_draft.md] — the full APA including purchase price, indemnification clauses, and schedule of acquired contracts
[target_company_financials.md] — last three years of audited financials, including revenue by product line and customer concentration data
[valuation_policy_manual.md] — my firm’s internal policy on intangible asset categories, valuation approaches, and documentation standards
Here is a reference for what I want to achieve:
[Upload reference file: a completed PPA intangible asset memo from a prior deal (anonymized) as markdown]
Here’s what makes this reference work:
– It lists each intangible asset separately with a clear fair value conclusion and a one-paragraph rationale
– It separates contractual intangibles (backlog, customer contracts) from non-contractual intangibles (trade name, technology)
– It explicitly notes which assets are being subsumed into goodwill and why
– It uses a table format for the asset listing with columns: Asset Class, Fair Value, Valuation Approach, Useful Life, Key Assumption
Here’s what I need for my version / SUCCESS BRIEF:
Type of output + length: A memo of 1,500–2,000 words, with an executive summary table and detailed asset-by-asset analysis
Recipient’s reaction: The audit partner should be able to trace every asset on the list back to a specific clause in the APA or a line item in the financials
Does NOT sound like: Generic boilerplate from a textbook; no vague language like “appropriate fair value” without a supporting calculation
Success means: I can hand this memo to the external valuation firm as the starting point for their detailed work, and they will not need to re-read the APA to understand the asset population
My context file contains my standards, constraints, audience. Read it fully before starting.
DO NOT start executing yet. Ask clarifying questions first.
Give me your execution plan (5 steps max) before you begin.
This first prompt forces the AI to read the deal documents before it writes a single word. The clarifying questions step is critical—if the purchase price includes contingent consideration or an earn-out, the AI should ask whether you want that treated as a separate liability or as part of the consideration transferred. If the target company has a heavily R&D-centric cost structure, the AI should ask whether in-process R&D will be measured separately. These questions surface gaps in the source data that you might otherwise miss until the audit is well underway.
Once the intangible asset identification memo is complete, your next bottleneck is usually the valuation methodology itself. Most PPAs get challenged on the reasonableness of the assumptions used in the multi-period excess earnings method (MEEM) for customer relationships, or in the relief-from-royalty method for technology and trade names. The discount rate, the attrition rate, and the forecast period are the three levers that move the fair value the most. If you get these wrong, the goodwill balance will be misstated, and the auditors will push back. The second prompt below focuses on validating and stress-testing those assumptions using market data and industry benchmarks.
Stress-Testing Your Valuation Assumptions
The output of the first prompt is a comprehensive asset list. The second prompt is about pressure-testing the logic behind the numbers. This is where you, as the finance professional, add the most value—the AI can generate a sensitivity analysis, but you must interpret it in the context of the market, the target’s growth trajectory, and the risk profile implied by the purchase price.
First, read these files completely before responding:
[intangible_asset_memo_draft.md] — the output from the previous identification phase, including the draft fair value estimates and useful life assumptions
[discount_rate_calculation.md] — my WACC build-up model, including risk-free rate, beta, size premium, and company-specific risk adjustments
[comparable_transactions.md] — recent M&A transactions in the same industry with disclosed PPA metrics (customer attrition rates, royalty rates, useful lives)
[valuation_policy_manual.md] — my firm’s internal policy on acceptable valuation approaches and documentation requirements
Here is a reference for what I want to achieve:
[Upload reference file: a published valuation guide excerpt (e.g., from a Big 4 firm or AICPA) showing typical attrition rates and royalty ranges for the target’s industry]
Here’s what makes this reference work:
– It provides a range of observable market data points, not just a single point estimate
– It explains the qualitative factors that justify moving to the high or low end of the range
– It explicitly links the discount rate adjustment to company-specific risk factors (customer concentration, technology obsolescence risk)
– It provides a sensitivity table showing the impact of a +/- 1% change in attrition rate and a +/- 0.5% change in discount rate on the final fair value
Here’s what I need for my version / SUCCESS BRIEF:
Type of output + length: A 2,000-word validation memo with three sections: (1) Assumption Benchmarking table, (2) Sensitivity Analysis with tornado chart description, (3) Auditor Challenge Q&A — the ten most likely questions and my recommended responses
Recipient’s reaction: The audit manager should feel that every assumption is supported by either internal data or external market evidence, and that I have already considered the “what if” scenarios they would raise
Does NOT sound like: A defensive justification of my original numbers; no “we believe this is reasonable” without showing the underlying reasoning and comparable data
Success means: I can send this memo to the auditors pre-emptively, and they either accept the assumptions without further inquiry or ask questions that I have already answered in the memo
My context file contains my standards, constraints, audience. Read it fully before starting.
DO NOT start executing yet. Ask clarifying questions first.
Give me your execution plan (5 steps max) before you begin.
When you run this second prompt, expect the AI to ask for clarification on whether your WACC build-up already includes a size premium that overlaps with the company-specific risk adjustment, or whether the attrition rate you used reflects historical customer churn or a forward-looking projection. These are the exact questions an auditor will ask. The value of the AI here is not that it knows the “right” answer—it does not—but that it forces you to articulate and document the reasoning behind your choices in a concise, traceable way.
A practical tip for both prompts: always upload the actual source documents as markdown or PDF files rather than pasting text into the chat. The AI reads the full context much more reliably when it has the original files to reference. If you cannot upload the purchase agreement due to confidentiality, create a redacted version that removes the parties’ names but keeps the economic terms and asset definitions intact. The more complete the input, the less hallucination you will see in the output.
What to try next after you have mastered these two prompts: ask the AI to draft the goodwill impairment analysis memo for the following year-end, using the PPA output as the starting point. The same asset list and assumptions will feed directly into the CGU (cash-generating unit) determination and the recoverable amount assessment. If you build the PPA documentation properly now, you will save yourself an entire quarter of work at the next reporting date.
Published on 8 September 2026 on growwithgpt.com