Tools · Model Health Check

Would your modelsurvive a review?

Twenty yes/no questions built on the principles of the ICAEW Financial Modelling Code. Five minutes, a score out of 100, a breakdown by category, and a report that tells you what to fix first.

BE HONEST. THE SCORE IS ONLY USEFUL IF THE ANSWERS ARE.

Model health score

00 / 20 ANSWERED

Answer the twenty questions

DESIGNINPUTSCALCSCHECKSCONTROL
  • DESIGN0/4
  • INPUTS0/4
  • CALCS0/4
  • CHECKS0/4
  • CONTROL0/4

01 · PURPOSE & DESIGN

Whether the model is built for a defined decision and laid out so that a reviewer can follow it.

0 / 4
  1. 01The purpose of the model, the decisions it supports and its intended users are written down inside the model.

    A model with no stated purpose cannot be judged fit for it. · ICAEW · Purpose

  2. 02Inputs, calculations and outputs are kept on separate sheets or clearly separated, labelled areas.

    Mixing them is the most common source of hidden errors and silent overrides. · ICAEW · Design

  3. 03Every sheet follows the same layout and the same timeline: the same period sits in the same column everywhere.

    A shifted column on one sheet breaks every cross-sheet link without warning. · ICAEW · Design

  4. 04The model is no more complex than the decision requires, and nothing in it is hidden: no hidden sheets, rows, columns or grouped-away logic.

    Complexity that the user cannot see is complexity the user cannot challenge. · ICAEW · Design

02 · INPUTS

Whether assumptions are visible, sourced and owned.

0 / 4
  1. 05No numbers are hard-coded inside formulas; every assumption sits in its own labelled input cell.

    A number buried in a formula cannot be found, flexed or defended. · ICAEW · Inputs

  2. 06Inputs are visually distinct through consistent formatting, and only input cells are intended to be changed by the user.

    Formatting is how a reviewer knows what they are allowed to touch. · ICAEW · Inputs

  3. 07Every input carries a unit, a source, an owner and the date it was last updated.

    An assumption with no provenance cannot be defended in front of a board. · ICAEW · Inputs

  4. 08Alternative cases are run from a single scenario selector, with the base case preserved and never overwritten.

    Without this, scenarios become copies of the file and drift apart within a week. · ICAEW · Inputs

03 · CALCULATIONS

Whether the arithmetic is transparent, consistent and self-contained.

0 / 4
  1. 09Formulas are short enough to read in the formula bar, and longer logic is broken into labelled intermediate steps.

    A formula nobody can read is a formula nobody can check. · ICAEW · Calculations

  2. 10Each calculation row uses one formula, applied consistently across every period.

    Consistency across the row is what makes a model auditable in seconds rather than days. · ICAEW · Calculations

  3. 11Balances are built as opening plus movements equals closing, in the same way for cash, debt, working capital and fixed assets.

    Consistent roll-forwards are why the balance sheet balances by construction, not by luck. · ICAEW · Calculations

  4. 12The model contains no circular references, no links to other workbooks and no macros required to calculate.

    Circularity, external links and macros all hide logic from the reviewer and break when files move. · ICAEW · Calculations

04 · CHECKS & TESTING

Whether the model can tell you when it is wrong, and whether anyone has tried to break it.

0 / 4
  1. 13A dedicated checks area exists, and a single overall error flag is visible from every sheet.

    If the model can go wrong silently, it will. · ICAEW · Checks

  2. 14The balance sheet balances and cash reconciles in every period, and a check would fire if either stopped.

    The balance check is the model's heartbeat. · ICAEW · Checks

  3. 15Sensitivity and stress tests have been run, and the outputs respond sensibly to extreme inputs.

    A model that only works on the base case is a calculator, not a model. · ICAEW · Testing

  4. 16Someone other than the builder has reviewed the model, and their findings have been addressed.

    Builders cannot see their own blind spots. Every serious model has a second pair of eyes. · ICAEW · Review

05 · DOCUMENTATION & CONTROL

Whether someone else could pick it up, trust it and run it.

0 / 4
  1. 17A cover or guide sheet explains purpose, structure, key assumptions, limitations and how to use the model.

    Documentation is what turns a personal spreadsheet into a business asset. · ICAEW · Documentation

  2. 18Version control is in place: consistent file naming, a change log inside the model and one clearly current version.

    "Which version is this?" is the most expensive question in a board meeting. · ICAEW · Version control

  3. 19Changes are logged, re-tested and the checks re-run before a new version is released to users.

    A change that skips the checks is a new error waiting to be found. · ICAEW · Change control

  4. 20A competent person new to the model could understand and run it within an hour.

    Transparency is the whole point of the Code. This is the test of all the others. · ICAEW · Transparency

YOUR REPORT

Answer all twenty questions and your report appears here: the score, what to fix first, and what we would recommend.

Email me the report

We will email you the score, the category breakdown, every fix in priority order and our recommendation, with a PDF of the full report attached. Handy for the file, or for whoever built the model.

The standard behind the score

Built to be questioned, not admired.

Every question maps to a principle in the ICAEW Financial Modelling Code, the standard Naked Finance builds to. The score is not a judgement on the person who built the model; it is a measure of how much the business can trust the numbers it is making decisions on, and how quickly someone new could pick it up.

The Code

ICAEW Financial Modelling Code

The Institute's principles for models that decisions depend on: a defined purpose, a clear structure, visible and sourced inputs, transparent calculations, checks that catch errors, documentation, version control and independent review.

  • Purpose & Design. Whether the model is built for a defined decision and laid out so that a reviewer can follow it.
  • Inputs. Whether assumptions are visible, sourced and owned.
  • Calculations. Whether the arithmetic is transparent, consistent and self-contained.
  • Checks & Testing. Whether the model can tell you when it is wrong, and whether anyone has tried to break it.
  • Documentation & Control. Whether someone else could pick it up, trust it and run it.
Read the ICAEW Financial Modelling Code

Opens the Code on icaew.com (PDF). The ICAEW Financial Modelling Code is published by the Institute of Chartered Accountants in England and Wales. Naked Finance is not affiliated with or endorsed by ICAEW.

Nothing you answer is stored unless you ask us to email the report. The score is a self-assessment, not a review or certification of your model. · All tools · Training and resources