How to Build an LBO Model: The Build Order That Survives a Modelling Test

Step 1 — Transaction assumptions and sources and uses

Start with entry: LTM EBITDA, the entry multiple, and therefore entry enterprise value. Then the funding: how much debt by tranche, and sponsor equity as the plug that makes sources equal uses.

Uses are the purchase of enterprise value, the refinancing of any existing debt, and transaction fees — both advisory fees and financing fees, which are treated differently in the accounting. Sources are each debt tranche and the sponsor equity.

A worked case: $50m of EBITDA at an 8.0x entry multiple is $400m of enterprise value. With $240m of debt (4.8x leverage) and $15m of fees, uses total $415m and sponsor equity is the $175m balance.

Build the check here: sources less uses must equal zero, with conditional formatting that makes a mismatch impossible to miss.

Step 2 — The opening balance sheet

Purchase accounting adjustments: write off the existing goodwill, calculate new goodwill as purchase price less the fair value of net identifiable assets, write up identifiable intangibles where the case specifies it, recognise the deferred tax liability that the write-up creates, capitalise financing fees as an asset to be amortised, and expense advisory fees against equity.

In a timed test, the examiner usually simplifies this — often to "assume no asset write-ups". Read the instructions before building it. Time spent on purchase accounting the case has told you to skip is time not spent on the returns calculation.

Step 3 — The operating build, kept proportionate

Project revenue, EBITDA and the cash flow items needed for debt service: D&A, capex, and the change in working capital. Nothing more elaborate than the case supports.

This is where most candidates over-invest. The returns are driven by three things — EBITDA growth, deleveraging and the exit multiple — and a highly granular revenue build changes only the first, usually marginally. A simple growth-rate build that lets you finish the model beats a segment-level build that leaves the debt schedule empty.

Stop at EBITDA and free cash flow before debt service. Interest needs the debt schedule.

Step 4 — The debt schedule and cash sweep

This is the engine of the model. For each tranche in seniority order: opening balance, mandatory amortisation, optional prepayment from the cash sweep, closing balance, and interest calculated on the balance.

The waterfall of cash: start from cash available for debt service, pay cash interest, pay mandatory amortisation, then sweep whatever remains against the most senior prepayable tranche, respecting the minimum cash balance the business needs to operate. If cash would go below the minimum, the revolver draws.

Deleveraging is where a meaningful share of the equity return comes from, so this schedule is not a supporting detail — it is a primary driver of the answer. It is also where the circularity between interest and cash flow lives, so decide up front whether you are enabling iterative calculation or using opening balances for interest.

• Tranches in strict seniority: revolver, term loan A, term loan B, then subordinated or mezzanine.
• Mandatory amortisation before any sweep — it is contractual.
• A minimum cash balance the sweep is not allowed to breach.
• Cash interest and PIK interest handled separately; PIK accretes to the balance rather than consuming cash.

Step 5 — Exit and returns

Exit enterprise value is exit-year EBITDA times the exit multiple. Subtract net debt at exit to get exit equity value. MoIC is exit equity divided by the sponsor’s entry equity; IRR is the annualised equivalent over the hold period.

Continuing the earlier case: EBITDA grows from $50m to roughly $73m over five years and cumulative free cash flow of about $110m takes debt from $240m to $130m. Exiting at the same 8.0x gives $588m of enterprise value, less $130m of debt, for $458m of equity. Against $175m invested that is 2.6x, roughly a 21% IRR.

Hold the exit multiple flat unless the case gives you a reason not to. Assuming multiple expansion to make the returns work is the single fastest way to lose credibility with a private equity reviewer, because it is the one return driver the sponsor does not control.

Step 6 — The value creation bridge

This is the output that separates a model from an answer, and the one most candidates omit. Decompose the equity return into its three sources: how much came from EBITDA growth, how much from debt paydown, and how much from multiple change.

The bridge is what an investment committee actually discusses. A deal whose return comes 70% from deleveraging is a fundamentally different proposition from one that depends on growing EBITDA by half, and the two carry different risks. Presenting the decomposition shows you understand what the model is telling you rather than merely that you can operate it.

It is also the fastest sanity check on your own work. If the bridge does not reconcile to the total equity gain, something upstream is wrong.

Sensitivities, and what to do when time runs out

Finish with a sensitivity table on IRR across entry multiple and exit multiple, and a second across leverage and EBITDA growth if time allows. Those four variables cover most of what a reviewer will ask.

If the clock is running out, the priority order is unambiguous: a working returns calculation with rough operating assumptions beats a sophisticated operating model with no returns. Get to an IRR, then improve. And state a recommendation — most tests ask what you would do, and a model that reaches no conclusion scores poorly regardless of how clean the mechanics are.