Most financial models fail twice. First, in the spreadsheet — bad assumptions, broken links, circular references that nobody notices until the data room. Second, in the room — when the analyst who built it cannot explain in writing what the model actually says, why the numbers move the way they do, and what a decision-maker is supposed to do with the output.
The model memo is what bridges those two failures. It is not a pitch deck. It is not an executive summary. It is the document that makes your model defensible — in writing, before anyone opens the file.
What a Model Memo Is — and What It Is Not
A model memo is a structured written document that translates the analytical logic of a financial model into a format decision-makers can evaluate without navigating the spreadsheet. It accompanies the model — it does not replace it.
This distinction matters because most analysts confuse model documentation with three other things:
A pitch deck is a persuasion tool. It selects favorable data and packages it for an audience that has not yet committed to the transaction. A model memo assumes the reader is already inside the process — lender, CFO, investment committee — and needs to evaluate assumptions, not be sold to.
An executive summary is a two-paragraph overview. A model memo is a working document, often five to fifteen pages, that a lender’s analyst or an IC associate will read line by line against the model.
A model audit is a backward-looking error check. A model memo is a forward-looking analytical argument — here is what we assume, here is why, here is how sensitive the output is to those assumptions being wrong.
What this is the decisive point: the memo must be written so that someone who has never seen your model can reconstruct the analytical logic — and stress-test it — from the document alone.
Why Most Model Memos Fail Before They Are Read
The most common failure is not incorrect numbers. It is structural mismatch — the memo is organized differently from the model, uses different terminology, and references outputs the reader cannot trace back to a line item.
We see this consistently in corporate finance engagements: a model that has been built with precision — three-statement integration, scenario architecture, WACC built from first principles — arrives with a memo that reads like a sales document. Broad claims about market opportunity. A revenue forecast presented without a single assumption statement. A terminal value that appears in the output tab but is never explained in the written document.
A lender’s credit analyst will stop reading at that point. An investment committee associate will flag it before the meeting starts.
The fix is not better writing. It is better structure. [INTERNAL LINK → How to Build a Financial Model Review Process Your CFO Won’t Tear Apart]
The Five-Layer Structure of a Professional Model Memo
A bank-grade model memo follows a consistent architecture. Each layer answers a specific question the reader will ask — in the order they will ask it.
Layer 1 — Transaction or Situation Overview (½ to 1 page)
What is being modeled, and why does it matter? This section identifies the company or asset, the transaction structure or analytical purpose, and the time horizon of the model. It should be written in three to five sentences — not paragraphs.
One sentence that belongs here: “This model projects the three-statement performance of [entity] under three scenarios over a five-year period, with a terminal value derived using the Gordon Growth Method at a 2.5% perpetuity growth rate.” One sentence that does not belong here: anything about market opportunity, competitive positioning, or management quality.
Layer 2 — Model Architecture Note (½ page)
How is the model built? This is not a tour of the tabs. It is a statement of structural choices: whether the model is integrated or modular, how assumptions flow into outputs, how circularity (if any) is handled, and what the model cannot do — its stated limitations.
Stating what a model cannot do is not weakness. It is the single clearest signal of analytical discipline. A model memo that claims the model covers every scenario is read by experienced reviewers as a model memo that covers none of them rigorously.
Layer 3 — Key Assumptions Register (2 to 4 pages)
This is the core of the memo — and the section most analysts underwrite. The assumptions register lists every material input that drives the model output, states the source or rationale for each input, and flags which assumptions are most sensitive to being wrong.
A professional assumptions register is not a table of numbers. It is a written argument. For each material assumption, the memo should answer: where did this number come from, what is the range of defensible alternatives, and what happens to the output if this assumption moves by 10 percent?
If the answer to any of those three questions is “we don’t know” — that is not a problem with the memo. It is a problem with the model. The memo makes it visible. [INTERNAL LINK → Building Sensitivity Tables the Right Way]
Layer 4 — Output Summary and Scenario Comparison (1 to 2 pages)
What does the model say? This section presents the key outputs — IRR, equity value, DSCR, EBITDA margin trajectory, free cash flow yield — across the defined scenarios. It does not interpret them. It states them.
The interpretation comes in one paragraph at the end of this section: which scenario is the base case, why, and what the base case implies for the decision at hand. One paragraph. Not five.
What most analysts get wrong here: they present the upside scenario first because it looks better. A professional model memo presents the base case first, the downside second, and the upside third. That ordering signals that the analyst has thought about risk before optimism.
Layer 5 — Limitations, Open Items, and Next Steps (½ page)
Every model memo ends with what is unresolved. What data was unavailable. What assumptions were proxied from comparable transactions because company-specific data did not exist. What changes to the deal structure would require the model to be rebuilt rather than updated.
This section is where model memos earn long-term credibility — and where most analysts leave value on the table by simply omitting it.
Writing the Assumptions Register: The Section That Separates Practitioner Memos from Analyst Memos
The assumptions register deserves its own treatment because it is where the analytical quality of the model memo is evaluated.
Most assumptions registers are formatted as tables: assumption name, value, source. That format is necessary but not sufficient. A table tells a reviewer what you assumed. A professional register tells them why — and what happens if you are wrong.
Here is the difference in practice:
Table format (insufficient): Revenue growth rate: 8.0% — Management guidance.
Register format (defensible): Revenue growth rate: 8.0% per annum, based on management guidance for years one through three and tapered to 4.5% in years four and five to reflect sector mean reversion. The 8.0% figure is consistent with the trailing three-year CAGR of 7.3% but above the peer median of 6.1%. A 200bps reduction in the growth assumption reduces base-case equity value by approximately 12%.
That paragraph takes forty-five seconds to write once the model is built. It takes a lender’s analyst four minutes to verify. It is the difference between a model memo that passes first review and one that generates a list of thirty follow-up questions before the IC meeting.
When to Write the Model Memo — and Who Should Write It
The model memo should be written by the person who built the model. Not by a junior analyst cleaning up the work of a senior. Not by a communications team translating technical content for a general audience.
The reason is simple: the memo requires the author to defend every assumption in writing. That defense process almost always surfaces errors, inconsistencies, and unstated assumptions that were invisible in the spreadsheet. Writing the memo is, functionally, the final quality check on the model.
In terms of timing: the memo draft should be completed before the model is shared with any external party. A model sent to a lender, an investment committee, or a counterparty without a memo is a model that will generate questions the analyst is not prepared to answer in writing.
If you are working on a transaction or planning exercise where the model needs to survive external scrutiny — and the written documentation is not keeping pace with the analytical work — that is the inflection point where structured support adds the most value.
FAQ — Financial Model Memo
What is the difference between a model memo and an investment memo?
An investment memo makes a case for a transaction. A model memo documents the analytical logic behind the financial model — assumptions, architecture, outputs, and limitations — for a technically literate reviewer.
How long should a financial model memo be?
Five to fifteen pages, depending on model complexity. A three-statement business plan model typically requires eight pages. An LBO with multiple debt tranches and scenario architecture may require twelve to fifteen.
Who reads the model memo?
Lenders’ credit analysts, investment committee associates, CFOs, and external auditors. The memo is written for a reader who will open the model after reading it — not instead of it.
What is the most common mistake in model memos?
Presenting outputs without stating assumptions. A model memo that shows an IRR of 22% without explaining what growth rate, exit multiple, and leverage assumption produces that result is analytically incomplete and will not survive lender or IC review.
For organizations where the gap between model quality and documentation quality is creating friction in the deal or approval process — we work with finance teams to build both the analytical framework and the written output that makes it defensible.