A leveraged buyout model answers one question: if a buyer funds an acquisition mostly with debt, can the business carry that debt and still return enough to the equity? Private equity funds, independent sponsors and search fund buyers all build some version of it, whether the deal is a large carve-out or a small company bought with an SBA loan and a seller note.
I'm Sergei Mochtchenkov, a CFA Charterholder. I build and review acquisition models for buyers, sponsors and lenders. This guide walks through how an LBO model is put together, the order to build it in, and the mistakes that most often make the returns look better than they are.
What an LBO model does
An LBO model is a three-statement model with a transaction layered on top. It starts with the business's normalized earnings, adds the purchase price and financing structure, projects cash flow for a holding period (often five years), and ends with an assumed exit. The outputs are the equity return measures, usually internal rate of return (IRR) and multiple on invested capital (MOIC), and the credit measures a lender cares about, such as leverage and debt service coverage.
The model works in both directions. A sponsor uses it to find the maximum price that still clears a target return. A lender uses the same mechanics to test whether projected cash flow repays the debt under a weaker case.
Step 1: Start from normalized EBITDA
Every LBO model rests on the earnings figure the price and the debt are sized from. In private deals, that is usually adjusted EBITDA, not reported EBITDA. If the add-backs behind it are weak, every output downstream is overstated, because the same number drives the purchase price, the debt capacity and the exit value.
Before building the transaction, test each adjustment against evidence. My guide to EBITDA add-backs covers which adjustments usually hold up and which do not. For owner-operated businesses, also decide whether you are working from SDE or EBITDA, since the difference between SDE and EBITDA changes both the multiple and the cash available for debt service.
Step 2: Build sources and uses
The sources and uses table is the deal on one page. Uses are what the money pays for: the purchase price (enterprise value), transaction fees, financing fees, and any cash funded to the balance sheet for working capital. Sources are where the money comes from: senior debt, subordinated or mezzanine debt, a seller note, rollover equity from the seller, and new sponsor equity. The two sides must balance.
Here is a simplified illustration. The figures are hypothetical, chosen only to show the mechanics.
- Adjusted EBITDA: 5.0 million
- Purchase price at 7.0x EBITDA: 35.0 million
- Transaction and financing fees: 1.0 million
- Total uses: 36.0 million
- Senior term loan (3.0x EBITDA): 15.0 million
- Seller note (1.0x EBITDA): 5.0 million
- Sponsor equity: 16.0 million
- Total sources: 36.0 million
Keep the purchase price as enterprise value and handle cash, debt and debt-like items in the closing mechanics. The bridge from EBITDA multiples to equity value explains why the headline price and the cheque to the seller are rarely the same number.
Step 3: Project operating cash flow
Next, project the business for the holding period: revenue, gross margin, operating expenses, EBITDA, capital expenditure and changes in working capital. Tie the base year to the normalized figures, drive revenue from volume and price assumptions you can explain, and model working capital from days metrics rather than as a plug.
Working capital is often the most underestimated line in a growing business. A company that grows revenue usually ties up more cash in receivables and inventory, which reduces the cash available to repay debt. If the deal includes a working capital peg, make sure the closing level in the model matches the working capital peg in the purchase agreement.
Step 4: Build the debt schedule
The debt schedule is where an LBO model differs most from a normal forecast. Each tranche needs its own block:
- Opening balance from the sources and uses table.
- Interest, at a fixed rate or a floating base rate plus a spread. Calculate interest on the opening balance, or on the average balance with a circularity switch, and be consistent.
- Mandatory amortization, the scheduled principal payments.
- Optional prepayment or cash sweep, where excess cash flow repays debt in order of seniority.
- Closing balance, which becomes the next period's opening balance.
A cash sweep needs a clear waterfall. Start with cash flow from operations, deduct capital expenditure, mandatory debt service and any minimum cash balance, and only then apply the remainder to optional prepayments. Seller notes are often subordinated and may have standstill terms, so check the documents rather than assuming they amortize like a bank loan.
Step 5: Model taxes and interest deductibility
Interest is deductible, but not without limit. Under Section 163(j), business interest expense is generally limited to business interest income plus 30% of adjusted taxable income. The IRS explains that for tax years beginning after December 31, 2024, depreciation, amortization and depletion are again added back when calculating adjusted taxable income, which generally increases the deductible amount (IRS Q&A on the business interest limitation). Small businesses that meet the gross receipts test can be exempt.
For a highly leveraged deal, model the limitation explicitly rather than assuming all interest is deductible. Disallowed interest can generally be carried forward, so the effect is often timing, but it still reduces cash in the early years when debt is highest.
Step 6: Set the exit and calculate returns
The exit assumption converts projected EBITDA back into enterprise value, usually with an exit multiple. Subtract the remaining net debt to get equity value at exit, then compare it with the equity invested.
Continuing the illustration: suppose EBITDA grows 5% a year to about 6.4 million in year five, the business exits at the same 7.0x multiple for an enterprise value of about 44.7 million, and cumulative free cash flow has repaid 12.0 million of the 20.0 million of debt.
- Equity value at exit: 44.7 − 8.0 = 36.7 million
- MOIC: 36.7 ÷ 16.0 = about 2.3x
- IRR over five years: about 18%
It helps to split the 20.7 million equity gain into its sources: about 9.7 million from EBITDA growth at a constant multiple, 12.0 million from debt paydown, less the 1.0 million of fees paid at entry. If most of the return depends on a higher exit multiple than the entry multiple, the deal relies on market conditions the buyer does not control.
Use the XIRR function with actual dates if there are interim dividends, recapitalizations or staged equity contributions. A simple compound growth formula only works for a single entry and a single exit.
Step 7: Stress-test the case
A single base case tells you very little. Build at least a downside case with slower growth, lower margins and a lower exit multiple, then run sensitivity tables on the drivers that matter most: entry multiple, exit multiple, EBITDA growth and interest rate. In the illustration, cutting the exit multiple from 7.0x to 6.0x lowers equity at exit to about 30.3 million, a MOIC of about 1.9x and an IRR of about 14%. My guide to scenario and sensitivity analysis covers how to set these tables up so they stay connected to the model.
Lenders look at the same model differently. They focus on leverage (total debt to EBITDA), interest coverage, fixed charge or debt service coverage, and whether the business can de-lever over a reasonable period. In December 2025 the OCC and FDIC withdrew the 2013 interagency leveraged lending guidance and told banks to apply general safe-and-sound lending principles, including analysis of a borrower's "past and current performance compared with projections" and the "capacity to de-lever over a reasonable period" (FDIC interagency statement). Each lender now sets its own leverage appetite, so ask early rather than assuming a market limit.
For SBA-financed acquisitions, the lender also applies a minimum debt service coverage ratio under SBA SOP 50 10 8.1, so the LBO model should report DSC by year alongside IRR.
Common LBO model mistakes
- Unsupported EBITDA. The model is only as good as the earnings it starts from. A quality of earnings review before the model is final protects both price and debt sizing.
- Ignoring working capital and capex. EBITDA is not cash. Growth that consumes working capital, or deferred maintenance that must be caught up, reduces debt paydown.
- Exit multiple above entry multiple without a reason. Multiple expansion is possible, but it should not carry the base case.
- Fees left out. Transaction and financing fees are real uses of cash and lower returns.
- Hard-coded plugs. Balance sheets that balance only because of a plug hide errors. Every line should flow from the statements and the debt schedule.
- No downside case. If the deal only works in the base case, it does not work.
If you want an independent check, a financial model audit reviews formulas, links and assumptions before the model goes to investors or a credit committee. If you're working on a deal now, you can book a call to talk through your model.
FAQ
What is a good IRR in an LBO model?
There is no single answer. The required return depends on the fund or investor, the risk of the business and current financing costs. The useful question is how much of the return comes from operating improvement and debt paydown rather than an assumed higher exit multiple.
What is the difference between IRR and MOIC?
MOIC measures how many times the equity is returned, regardless of time. IRR measures the annualized return and accounts for timing, so the same MOIC produces a lower IRR over a longer hold.
How much debt can an LBO use?
It depends on the lender, the stability of the cash flow and the debt service coverage the business can support. Model debt from cash flow capacity, not from a target leverage multiple alone.
Do small business acquisitions use LBO models?
Yes. A search fund or individual buying a business with an SBA loan and a seller note is doing a leveraged buyout. The model is simpler, but the same mechanics of sources and uses, debt service and downside testing apply.