Model Audit
Find the errors in a financial model before someone senior does.
Seat
IB · PE
Level
Analyst / Associate
Runtime
~30 min
Output
Structured error/risk report
What this helps you accomplish
A systematic review of an Excel model for hardcodes, broken links, circularity, sign errors, inconsistent assumptions and balance-sheet breaks — ordered so the highest-impact errors surface first.
Delegate · Verify · Decide
The core L3VLUP principle: AI output is never automatically correct. Know what to hand off, what to check, and what only you can own.
Delegate
AI is good enough to do this.
- • Scanning for hardcodes, broken links and inconsistent formulas across rows
- • Locating circularity and volatile functions
- • Diffing formula structure between periods/columns
- • Drafting the issue log with cell references
Verify
AI accelerates you here, but a professional checks the work.
- • Each flagged "error" is actually wrong — some hardcodes are deliberate inputs
- • The balance check formula itself isn’t a plug
- • Suggested fixes don’t break dependent calculations
Decide
Judgment the human owns. This is the skill.
- • Which issues are material to the decision vs cosmetic
- • Whether the model’s core architecture is trustworthy or needs a rebuild
- • What to say to the model’s owner, and how
Inputs required
- The model file (.xlsx)
- What the model is for (LBO, merger, operating, DCF)
- Which outputs matter most (returns, EPS, valuation range)
The workflow
- 1
Check structural integrity
Balance sheet balances in every period. Cash flow ties to balance-sheet cash. Circularity is intentional and switched.
- 2
Hunt hardcodes
Plugs inside formula chains are the classic silent killer. Every hardcode is either an input (move it) or an error (fix it).
- 3
Trace the drivers
Do key outputs respond correctly to key inputs? Flex growth, margin and leverage and watch the directionality.
- 4
Audit the assumptions
Internally consistent? Revenue growth vs capacity, margins vs history, exit multiple vs entry logic.
- 5
Stress the edges
Zero growth, downside case, covenant limits — models that break at the edges have hidden fragility in the base case.
- 6
Report by impact
Errors ranked by effect on the decision output, each with location, issue and suggested fix.
Run this skill
Any listed company, anywhere. Ticker or name.
Quality checklist
The output isn’t done until every box ticks.
- Issues ranked by output impact, not discovery order
- Every issue has a cell reference and a proposed fix
- Structural checks (B/S, CF tie) documented as pass/fail
Practise the fundamentals first
Free, no sign-up — in the Labs.