How this model actually works
An LBO isn't a valuation method — it's a returns model. You're not asking "what is this company worth," you're asking "if I buy it at this price, finance most of the purchase with debt, and sell it in N years, what return does the equity check produce." The mechanics: purchase price = entry EBITDA × entry multiple, split into a debt tranche and an equity tranche (the sponsor's actual cash outlay). Every year, EBITDA less D&A, interest, and tax gives net income; add D&A back and subtract CapEx and working-capital investment to get free cash flow, which sweeps against the debt balance. At exit, enterprise value (exit EBITDA × exit multiple) minus whatever debt is still outstanding is the equity's exit value — divide by the entry equity check for the MOIC (multiple on invested capital), and annualise that over the hold period for the IRR.
The sensitivity table is the actual point of the exercise. Three things drive an LBO's return: how much you pay (entry multiple), how much debt you use (leverage amplifies both gains and losses on the equity check), and what you sell for (exit multiple) — everything else is detail. A deal that only works if the exit multiple expands beyond the entry multiple is a bet on the market re-rating the business, not on operational improvement, which is exactly the distinction a real diligence process is trying to pin down.
Where this model is honest about its limits
This is a single-tranche, single-cash-sweep model — real LBOs have senior/subordinated debt tranches with different rates and covenants, PIK interest, management rollover equity, dividend recaps, and add-on acquisitions mid-hold, none of which are modelled here. It also assumes a clean entry and exit with no interim distributions. For a real deal, this is the skeleton every more detailed model builds on top of, not the finished analysis.