Financial model review
Check Excel #DIV/0!, #REF! and #N/A errors
Review saved #DIV/0!, #REF! and #N/A errors in Excel financial models. Find the cause, inspect the formula and understand the limits of saved results.

A saved Excel error is evidence to investigate at its exact sheet and cell. It does not prove that today's calculation still fails: the file may contain an old saved result, or the error may be deliberate. Start with the formula and its inputs before changing the model.
Division by zero: inspect the denominator
In an invented model, Metrics!D9 contains:
=B9/C9
If C9 is zero or blank, Excel can return #DIV/0!. Review what that denominator represents and whether zero or missing data is expected for the period. Microsoft's division-error guide explains this cause.
Do not choose a replacement result merely to remove the warning. A missing activity volume might require a visible unavailable result; an intentionally inactive period might have an agreed display convention. Those are model decisions, not conclusions a static checker can make.
Reference errors: trace the original target
Suppose Forecast!F14 contains:
=SUM(B14,#REF!)
A #REF! token can indicate an invalid reference after a change such as deleting referenced cells. Review the original layout and intended target before substituting a nearby cell. Microsoft's reference-error guide describes common causes.
A token and a saved result are different observations. For example, =IFERROR(#REF!,0) contains a reference-error token even when Excel returns zero. A report observing the token does not claim that this branch currently produces an error.
#N/A: distinguish a missing match from a deliberate marker
A saved #N/A may mean a lookup did not find its expected value. Examine the lookup key and source range before deciding that zero is the right substitute. Microsoft's #N/A guide explains this investigation.
Excel also has an NA() function that intentionally returns #N/A, as listed in Microsoft's function reference. An intentional missing-data marker still deserves explanation, but it is not automatically a financial-model defect.
Read the report's evidence and limitations
FinModelCheck distinguishes an explicitly stored cell error from a saved formula-error cache. Text that merely spells an error code is not treated as a typed Excel error. Missing formula caches are unavailable evidence, not errors or successful calculations. SpreadsheetML can store a formula and its saved value separately; see Microsoft's format documentation.
The checker does not recalculate Excel. Download the free detailed PDF to locate each reported Finding by sheet and cell, then compare its saved formula or error with the workbook in Excel. Save a separate corrected copy and compare your changes. Keep unexplained missing data visible until its meaning is resolved.
Explore the synthetic example report or open the checker. Related guides cover external dependencies and embedded assumptions.
Short answers
Should I wrap every formula in IFERROR?
No. IFERROR substitutes a chosen value for several error types, as Microsoft's function documentation explains. Use it only after choosing an appropriate result; broad suppression can conceal a cause you still need to resolve.
Is an absent formula cache a zero or an error?
Neither. It means the saved result is unavailable to this inspection. It does not establish what Excel would calculate now.
Does a high score certify the model?
No. The score is provisional and limited to supported saved evidence. Formula logic, financial assumptions and unassessed Excel features still need review.


