Model Review and QA Workflow
Executive Summary
Key Takeaways
- ✓ Model review and QA workflow is the internal build, self-check, peer-review, and sign-off lifecycle a modelling team runs on its own models, distinct from FMAE's own audit-engine methodology covered on Audit Methodologies for Financial Models.
- ✓ The workflow applies independent of any named modelling standard and independent of whether the model is later submitted for external audit at all.
- ✓ Self-check and peer review test different things — the model's own builder cannot reliably catch every error, since familiarity with a model's intended logic makes it easy to read what was intended rather than what was actually built.
- ✓ Sign-off should be a documented, named event, not an informal assumption that a model is "done" once no further changes are being made to it.
- ✓ This workflow is a construction-and-quality-control process, not a substitute for independent structural audit where one is required for a material decision.
Institutional Definition¶
Model review and QA workflow is the internal lifecycle a modelling team runs on a financial model before it is relied on: build, self-check, peer review, and documented sign-off. This page is deliberately distinct from Audit Methodologies for Financial Models, which describes FMAE's own deterministic rule-based engine and the broader external audit methodologies (manual, automated, deterministic) used to independently examine a model that already exists. This guide addresses a different, earlier stage: the process a modelling team runs on its own model, internally, during and immediately after construction — a general practice independent of any specific standard, tool, or whether the model is ever submitted for external audit at all.
Construction, review, and audit are related but separate activities in this Knowledge Centre's terminology. This page covers review; Financial Model Auditing covers audit. Confusing the two — treating an internal peer review as equivalent to an independent external audit — is one of the more consequential misconceptions addressed below.
Why It Matters¶
A financial model accumulates risk continuously during construction: a formula copied across a row with one cell left unadjusted, an assumption hardcoded during a late change and never moved back to the input sheet, a circular reference introduced to solve an immediate problem and never revisited. None of these require external scrutiny to catch — a disciplined internal workflow catches most of them before the model ever reaches a reviewer outside the team.
A modelling team without a defined review workflow relies on informal habit: the builder reads back through the model once, feels satisfied, and calls it complete. This is not a workflow — it is the absence of one, and it systematically misses the class of error a builder is least likely to catch in their own work, because they know what the model is supposed to do and unconsciously read that intent into cells that do not actually implement it.
The Workflow¶
1. Build¶
The construction phase, informed by the layout and structural decisions covered in Workbook Design and Model Architecture and Spreadsheet Engineering. Build-phase discipline — consistent formulas, no hardcodes in calculation cells, documented assumptions as they are entered — reduces the volume of issues the later stages need to catch, since fewer errors are cheaper to prevent than to find afterward.
2. Self-Check¶
The model's own builder reviews their own work systematically rather than casually: tracing key outputs back to their inputs, checking formula consistency across each row, and confirming that scenario or sensitivity toggles actually propagate through every downstream calculation. Self-check is useful and necessary, but limited by a structural blind spot: the builder's familiarity with the model's intended logic makes it easy to read what was meant rather than what was actually written, particularly in a formula that is close to correct but subtly wrong.
3. Peer Review¶
An independent second reviewer — a colleague with no prior involvement in the model's construction — examines the model with fresh eyes, working from the model's stated purpose and outputs rather than from familiarity with how it was built. Peer review catches a different class of error than self-check precisely because the reviewer has no existing assumption about what a given formula is "supposed" to do; they can only assess what it actually does. A Model Reviewer Checklist structures this stage so it is systematic rather than an unguided read-through.
4. Sign-Off¶
A documented, named event: a specific individual confirms, in writing, that the model has passed the team's internal review process and is ready for its intended use. Sign-off should record who reviewed the model, what was checked, what (if anything) was found and corrected, and the date. An undocumented, informal sense that a model is "done" once active changes stop is not a sign-off — it leaves no record a later reader can rely on to confirm the model was actually reviewed rather than simply finished.
Internal QA and External Audit¶
Internal review and external audit are complementary, not substitutable. Internal QA is appropriate, proportionate quality control before any model is relied on for any purpose, however minor. Where a model supports a material external decision — a financing close, an investment committee vote, a regulatory submission — independent external verification is typically still required in addition to internal QA, not instead of it, because an internal team, however disciplined, shares the same blind spots that produced the model in the first place. See Financial Model Auditing and Audit Methodologies for Financial Models for the independent-verification layer this workflow sits beneath, not on top of.
Common Mistakes¶
| Common Mistake | Why It Matters |
|---|---|
| Treating self-check as sufficient without peer review | The builder's own familiarity with the model's intended logic is precisely what makes certain errors invisible to them; a second, independent reviewer is required to catch that class of error. |
| Leaving sign-off informal or undocumented | Without a named individual and a recorded date, a later reader cannot confirm the model was actually reviewed, only that changes eventually stopped. |
| Running the workflow only once, at the very end of a long build | Errors introduced early in a long build compound; periodic internal review during construction, not only at the end, catches issues while they are still cheap to fix. |
| Assuming internal QA removes the need for external audit on a material decision | Internal review and external audit test for different things and share different blind spots; a model supporting a material external decision generally needs both. |
Best Practices¶
| Best Practice | Why It Matters |
|---|---|
| Assign self-check and peer review to different people, always | This is the structural reason the two stages exist separately — collapsing them back into one person's judgment removes the benefit of an independent perspective. |
| Use a structured checklist for peer review rather than an unguided read-through | A checklist ensures the same categories of risk are checked on every model, regardless of which reviewer is assigned — see the Model Reviewer Checklist linked below. |
| Record sign-off as a specific, dated, named event | This is what makes the review auditable later — a later reader should be able to confirm who reviewed the model, when, and what was found. |
| Run the workflow at defined checkpoints during a long build, not only at the end | Catching an error early in a multi-week build is materially cheaper than catching the same error after every downstream schedule has been built on top of it. |
Continue Reading¶
Prerequisites¶
- Financial Modelling Best Practices — the parent pillar
Related Technical Guides¶
- Audit Methodologies for Financial Models — FMAE's own audit-engine methodology; a distinct domain from this page's internal review workflow
- Workbook Design and Model Architecture
- Spreadsheet Engineering
Related Checklists¶
Related Pillars¶
Related Products¶
- Financial Model Audit Engine (FMAE) — independent structural auditing, the external-verification layer this internal workflow sits beneath
How OXXON tests thisRun a free structural check with FMAE
Frequently Asked Questions
What is model review and QA workflow?
The internal process a modelling team runs on a financial model before it is relied on externally — building the model, the builder's own self-check, an independent peer review by a second team member, and a documented sign-off confirming the model is ready for its intended use.
How is this different from FMAE's audit methodology?
FMAE's audit methodology (see Audit Methodologies for Financial Models) is a deterministic rule-based engine's own process for independently examining a model that already exists. This guide covers the internal workflow a modelling team runs on its own model during and after construction, independent of whether the model is ever submitted to FMAE or any other external auditor.
Is peer review the same as self-check?
No. Self-check is the model's own builder reviewing their own work, which is useful but limited, since familiarity with the model's intended logic makes it easy to read what was intended rather than what was actually built. Peer review is an independent second reviewer examining the model with no prior involvement in its construction, which catches a different class of error.
Does a model need external audit if it has already been through internal QA?
Internal QA and external audit address related but distinct risks. Internal QA is appropriate quality control before a model is relied on for any purpose. Where a model supports a material external decision — financing, an investment committee vote, a regulatory submission — independent external verification is typically still required, since internal reviewers share the team's blind spots regardless of how rigorously the internal process was followed.
What does "sign-off" mean in this workflow?
A documented, named event in which a specific individual confirms the model has passed the team's internal review process and is ready for its intended use, as distinct from an informal assumption that a model is finished once active changes stop.
Does this workflow depend on following a specific modelling standard?
No. It applies whether or not the model was built to a named standard such as the FAST Standard or the ICAEW Financial Modelling Code — see Financial Modelling Best Practices for the standards landscape this workflow is independent of.
Related Articles
Audit Methodologies for Financial Models
Financial model audit methodologies fall into three primary categories: manual line-by-line review, automated structural analysis, and deterministic rule-based checking. Each methodology differs in scope, speed, consistency, and the types of errors it is designed to detect. The appropriate methodology depends on transaction complexity, time constraints, and institutional risk appetite.
Financial Modelling Best Practices — Standards Compared
Financial modelling best practice is not a single document but a landscape of named institutional standards, each publishing its own conventions for how a model should be structured, formatted, and documented. This page defines that landscape — what a named modelling standard actually is, how the FAST Standard and the ICAEW Financial Modelling Code differ in approach and scope, and how a practitioner chooses between them or applies more than one. It sits beside, not instead of, the Knowledge Centre's structural-foundation page on what makes an Excel financial model reliable — this page is about who has codified that discipline into a named standard, and how those standards compare to one another.
Financial Modelling Best-Practice Checklist
This checklist sets out the construction-time disciplines a financial modelling team should apply while a model is being built, synthesising the common ground across the FAST Standard, the ICAEW Financial Modelling Code, and general spreadsheet engineering practice. It is not a certification checklist and does not test whether a model's calculations are correct — it is a builder's self-check aid, distinct from the Financial Model Audit Checklist, which is used for independent, post-hoc structural verification rather than during construction.
Model Reviewer Checklist
This checklist structures the independent peer-review stage of a modelling team's internal quality-assurance workflow. It is role-neutral and standard-agnostic — written for whoever is acting as the second, independent reviewer of a model, rather than for a specific external audience such as a lender or investment committee. It complements, rather than duplicates, the audience-specific review checklists elsewhere in this section and the Financial Modelling Best-Practice Checklist used at the construction stage that precedes peer review.
Real Estate Financial Modelling
Real estate financial modelling spans two structurally distinct disciplines: development appraisals, built forward from land and construction cost through phased sales or leasing velocity to a gross development value, and income-producing asset models, built from stabilised net operating income to an exit value using direct capitalization or a discounted cash flow. This page is the hub for the Knowledge Centre's real estate modelling content: the two model families, how gross development value and residual land value are built, waterfall and promote mechanics, and how each major property type — residential, office, retail, industrial and logistics — specializes the base structure to its own revenue drivers.