How an LBO Model Works

An LBO model works out the most a private equity firm can pay for a company, funded mostly with borrowed money, and still earn its target return when it sells the company about five years later. Every part of the spreadsheet feeds that answer: how much the firm can borrow, how much cash the business produces to pay the debt down, and what the equity is worth at the end.

I'm Stephen Turban, founder of WSG and a Harvard and McKinsey alum. I went through McKinsey's on-cycle recruiting myself, and I now guide students through PE and VC recruiting.

Once you understand how the model fits together, practice saying it out loud with the 6-step LBO answer interviewers want to hear here. LBO questions come up most often in private equity recruiting, so see how undergrads actually land those internships here.

How Much of an LBO Model You Need for IB and PE Interviews

Banking interviews mostly ask you to explain how an LBO works, while private equity interviews add paper LBOs and, often, a timed test where you build the model yourself. In investment banking, "walk me through an LBO" is a standard question (our walk me through an LBO guide covers the verbal answer), but Mergers & Inquisitions' interview guide ranks LBO questions as less important than accounting and valuation, and it calls Excel tests "quite unlikely for entry-level roles in the U.S." In private equity, M&I's LBO modeling test guide describes three formats: simple paper LBOs, timed tests that are "often between 1 and 3 hours," and open-ended take-home case studies. It says the timed tests are most common in on-cycle recruiting at large US firms.

LBO Returns Come From EBITDA Growth, Debt Paydown, and the Exit Multiple

Every LBO makes its money from some mix of EBITDA growth, debt paydown, and selling at a higher multiple than the buyer paid, and because the lenders are owed a fixed amount, debt makes each of those gains count for more on the equity. (EBITDA is earnings before interest, taxes, depreciation, and amortization, a rough measure of the cash a business produces before financing and accounting choices; our EBITDA guide explains why buyers price companies off it.)

Any gain or loss in the company's value lands on the equity, so borrowing makes an LBO's returns bigger in both directions. Say a private equity firm (the "sponsor") buys a company for $1,000 million and sells it five years later for $1,300 million. If it paid entirely with its own money, it made 30 percent. If it put in $400 million and borrowed the other $600 million, it repays the $600 million out of the sale and keeps $700 million, so the same company at the same price turned $400 million into $700 million, a 75 percent gain. If the company sells for less, the lenders still get their $600 million back first, so the loss lands on the equity too.

That example leaves out interest. In a real deal the company pays interest on the $600 million out of its cash flow, so borrowing raises the equity's return only while the business earns more than the debt costs. When the business's return falls below the cost of debt, the leverage effect goes into reverse and debt lowers the return on equity.

Here's each driver on its own, starting from that same deal: $100 million of EBITDA bought at 10.0x, with $600 million of debt and $400 million of equity, sold after five years.

• EBITDA growth: this is the case above. EBITDA rises to $130 million at a flat 10.0x, so the company sells for $1,300 million and the $400 million of equity becomes $700 million, 1.75x.

• Debt paydown: EBITDA stays at $100 million, so the company is still worth $1,000 million, but assume five years of cash flow cut the debt from $600 million to $450 million. The equity is worth $550 million, about 1.4x.

• Multiple expansion: EBITDA and debt stay where they were, but a buyer pays 12.0x, so the company sells for $1,200 million and the equity is worth $600 million, 1.5x.

The first two drivers come from running the business: growing it, running it for cash, and paying down the loans. The third depends on what buyers will pay in five years, which the sponsor can't control. That's why a model that needs a higher exit multiple to work is weaker than one that works with the multiple held flat, and why Wall Street Prep calls setting the exit multiple equal to the purchase multiple "the standard convention."

The Six Parts of an LBO Model

An LBO model runs in the order the deal happens: set the price, line up the money, forecast the cash, pay down the debt, measure what the equity is worth at the end, and then test how much that answer depends on the assumptions.

Purchase Price

The purchase price is usually set as a multiple of the last twelve months' (LTM) EBITDA, and the result is the enterprise value, the price of the whole business, debt and equity together. At $100 million of LTM EBITDA and 10.0x, enterprise value is $1,000 million.

The current owners receive less than that when the company has debt, because the buyer either takes the debt over or pays it off. Private deals are often priced "cash-free, debt-free": the price is set on enterprise value, the seller pays off its debt at closing out of the proceeds, and the seller keeps the excess cash. For a public company you work from the other direction: the offer price per share times the diluted share count gives the equity value, and adding net debt (debt minus cash) gives enterprise value.

Fees come in two kinds, and the model treats them differently. Transaction fees, paid to advisors, lawyers, and accountants, are expensed at closing, so they come out of the sponsor's equity on day one and never come back. Financing fees, paid to the banks that arrange the debt, are spread over the life of the loans, so they show up each year as a non-cash expense.

Sources and Uses

The sources and uses table lists what the deal costs and where the money comes from, and the sponsor's equity is the plug that makes the two columns equal. Uses are the purchase of the equity, repayment of the company's existing debt, the two kinds of fees, and the minimum cash the company needs on its balance sheet on day one. Sources are the new debt, listed from most senior to least, and the sponsor's equity.

Debt is sized first, as a multiple of EBITDA (each 1.0x is called a "turn"), because the lenders decide how much they'll lend. The equity is whatever's left, total uses minus the debt, so it has to be a formula. If you type it in and then add a fee to uses, the table stops balancing.

The debt usually comes in layers, called tranches, with different terms:

• Revolver: a line of credit the company can draw on when cash runs short. It's usually undrawn at closing and listed at zero so the reader knows it's there.

• Term loan B: a floating-rate loan priced at a benchmark rate (now usually SOFR, the Secured Overnight Financing Rate) plus a spread, with required repayments of usually about 1 percent a year and a maturity of about six to eight years. Leveraged loans usually require the borrower to prepay them out of excess cash flow, and teaching models, including the one below, often send all spare cash there, so this is the tranche that gets paid down.

• Senior notes: high-yield bonds with a fixed coupon, repaid in one payment at maturity and often protected from early repayment for the first two or three years, so they sit at their original balance through most models.

Operating Model and Cash Available for Debt Repayment

The operating model turns revenue into cash available for debt repayment, the cash the business has left to pay its lenders after taxes, capital spending, and working capital. Each line runs off one assumption: revenue grows at a rate, EBITDA is revenue times a margin, and depreciation and amortization (D&A) is a percentage of revenue. Subtracting D&A gives EBIT (operating income). Interest comes in from the debt schedule, which is where the two parts connect.

Interest comes off before taxes because it's tax-deductible: paying it reduces "the taxable income and the amount of taxes due," so a company with debt pays less tax than the same company without it. Subtract interest and the financing fee amortization from EBIT to get pre-tax income, then tax it to get net income.

Net income isn't cash, so three adjustments turn it into cash:

• Add back D&A and the financing fee amortization: both reduced net income, but neither was a cash payment this year.

• Subtract capital expenditures (capex): the cash spent on equipment and buildings doesn't appear on the income statement, so it comes out here.

• Subtract the increase in net working capital: as a company grows, more cash gets tied up in receivables and inventory than suppliers' payables give back, so growth uses cash even in a profitable year.

What's left is the line the whole debt schedule runs on. Some models call it levered free cash flow before debt repayment.

Debt Schedule

The debt schedule tracks each tranche's balance year by year and sends every spare dollar to prepay the term loan, which is called the cash sweep. Each tranche gets its own block with a beginning balance, repayments, and an ending balance, and interest is calculated off the balance.

Any revolver borrowing is repaid first, because it's meant to be short-term. The term loan then takes its required repayment, and whatever cash is left above the minimum cash balance sweeps to prepay it. The sweep formula is a MIN of the cash left and the loan balance left, so the loan can't go below zero once cash flow outgrows the remaining balance. The notes just collect their coupon. If a year's cash doesn't cover the required payments, the revolver draws the shortfall, with a MAX(0, shortfall) formula that reads zero in a normal year.

Interest is the one tricky part. It's more accurate to charge a year's interest on the average of the beginning and ending balances, because the company pays debt down through the year. But the ending balance depends on the sweep, the sweep depends on net income, and net income depends on interest, so the calculation loops back on itself. Excel calls that a circular reference. There are two ways to handle it:

• Turn on iterative calculation, with a switch: in Excel, open File, then Options, then Formulas, and select "Enable iterative calculation," then add a cell that flips interest between the average balance and the beginning balance. If an error gets into the loop and spreads, flip the switch to break the loop, let the errors clear, and flip it back.

• Charge interest on the beginning balance: there's no loop at all, and interest comes out slightly higher than the average-balance version in years when debt is falling. In a short timed test, this is the simpler choice unless the prompt asks for average balances.

Returns

The returns block measures what the sponsor's equity is worth at exit, as a multiple of what it put in (MOIC) and as an annual rate (IRR). Exit enterprise value is the exit year's EBITDA times the exit multiple, and exit equity value is that minus the debt still outstanding plus the cash on hand.

MOIC, the multiple on invested capital, is exit equity divided by the equity the sponsor put in, so 2.5x means $2.50 back for every dollar. It ignores how long the money was tied up. IRR, the internal rate of return, is the annual compound rate that turns the money in into the money out. With one payment in at the start and one out at the end, IRR is MOIC raised to the power of one over the number of years, minus one. When there are cash flows along the way or the exit doesn't land on an anniversary, use Excel's XIRR with dates, because its IRR function assumes the cash flows come at regular intervals.

This table converts MOIC to IRR for the hold periods you'll see most:

MOIC	3 years	4 years	5 years	6 years	7 years
1.5x	14.5%	10.7%	8.4%	7.0%	6.0%
2.0x	26.0%	18.9%	14.9%	12.2%	10.4%
2.5x	35.7%	25.7%	20.1%	16.5%	14.0%
3.0x	44.2%	31.6%	24.6%	20.1%	17.0%
4.0x	58.7%	41.4%	32.0%	26.0%	21.9%

The five-year column is the one to know without thinking: 2.0x is about 15 percent, 2.5x about 20, and 3.0x about 25. Wall Street Prep's private equity interview questions put the usual target at about 20 to 25 percent, though each firm sets its own hurdle, so a five-year deal needs to roughly 2.5x to 3x the equity. That gives you a quick way to tell whether a model's answer is plausible.

Sensitivity Tables

A sensitivity table reruns the model across a grid of two inputs, and the most telling one sets the entry multiple against the exit multiple, because it shows how much of the return depends on price. In Excel it's Data, What-If Analysis, Data Table, with one input across the top, one down the side, and the top-left cell linked to the IRR. You can pair any two inputs the same way, such as revenue growth against margin.

A Five-Year LBO Model, Worked Line by Line

Ridgeway, a made-up company bought at 10.0x EBITDA with 6.0x of debt, turns the sponsor's $450 million into about $1,164 million over five years, a 2.59x MOIC and a 20.9 percent IRR, with all of the gain coming from EBITDA growth and debt paydown. The model below is built with exactly the parts above. The tables round to one decimal, so a line can differ by 0.1 from the sum of the lines above it; the Year 1 walk-throughs use exact figures.

Transaction assumptions

Assumption	Value
LTM revenue and EBITDA	$500.0 million and $100.0 million (20.0% margin)
Entry multiple and enterprise value	10.0x, so $1,000.0 million
Existing debt	$150.0 million, repaid at closing (cash-free, debt-free)
Transaction fees	$20.0 million, expensed at closing
Financing fees	$10.0 million, amortized over the 5-year hold
Minimum cash, funded at closing	$20.0 million
Term loan B	$450.0 million (4.5x), interest held flat at 8.0%, 1% annual repayment, 100% cash sweep
Senior notes	$150.0 million (1.5x), 9.0% fixed, repaid at maturity
Revolver	undrawn
Revenue growth	6.0% a year
EBITDA margin	up 0.5 points a year, from 20.5% in Year 1 to 22.5% in Year 5
D&A and capex	each 3.0% of revenue
Net working capital	10.0% of revenue
Tax rate	25.0%
Exit	end of Year 5 at 10.0x EBITDA

Sponsor equity is the plug, so it covers whatever the term loan and notes don't and the check row reads 0.0.

Total debt is $600.0 million, 6.0x EBITDA, and the sponsor's $450.0 million is about 43 percent of the total. That's more debt than recent US buyouts, which a PitchBook LCD analysis published in August 2025 put just under 5x EBITDA; the round 6.0x keeps the arithmetic easy to follow.

The equity check is also bigger than the $850.0 million paid to the sellers minus the $600.0 million borrowed, which would be $250.0 million. The other $200.0 million pays off the $150.0 million of old debt and covers $50.0 million of fees and starting cash, because the sponsor funds everything the lenders don't.

Term loan interest falls from $36.0 million to $22.3 million as the loan is paid down, which helps net income more than double over the five years.

In Year 1, revenue is $500.0 million times 1.06, or $530.0 million, and EBITDA at a 20.5 percent margin is $108.65 million. D&A of $15.9 million leaves EBIT of $92.75 million. Interest is 8 percent of the $450.0 million term loan, $36.0 million, plus 9 percent of the $150.0 million of notes, $13.5 million, and the financing fees amortize at $2.0 million a year. That leaves pre-tax income of $41.25 million, taxes of $10.31 million, and net income of $30.94 million.

Every dollar above the $20.0 million minimum goes to the term loan, so ending cash holds at $20.0 million every year.

Year 1's $30.94 million of net income, plus $15.9 million of D&A and $2.0 million of fee amortization, minus $15.9 million of capex and a $3.0 million working capital build (10 percent of the $30.0 million revenue increase), leaves $29.94 million of cash available for debt repayment. The required $4.5 million comes out first, and the other $25.44 million sweeps to the term loan, so it ends the year at $420.06 million. The debt schedule tracks that for each tranche, every year.

Only the term loan moves: it takes the 1 percent repayment and the whole sweep, while the notes stay at $150.0 million and the revolver is never drawn.

Cash available rises every year for two reasons, which is why the paydown speeds up. In Year 2, EBIT grows by about $8.4 million, which adds about $6.3 million after tax, and interest falls because it's charged on a lower balance, which adds about $1.8 million; a slightly bigger working capital build takes back about $0.2 million. D&A and capex cancel out here because both are 3 percent of revenue, as they would for a business that replaces equipment about as fast as it wears out.

Over five years the term loan comes down by $238.8 million, total debt falls from 6.0x EBITDA to 2.4x, and EBITDA covers interest 4.2 times by Year 5 instead of 2.2.

Each year's step down is bigger than the last, from $29.9 million in Year 1 to $67.7 million in Year 5, because cash available rises as EBITDA grows and interest falls.

At exit, the returns block turns the Year 5 numbers into the sponsor's equity.

Excel's IRR function on the sponsor's cash flows gives the same 20.9 percent as the MOIC shortcut, because nothing comes back to the sponsor until the exit.

MOIC is $1,164.3 million divided by $450.0 million, or 2.59x, and 2.59 raised to the one-fifth power, minus one, is 20.9 percent. That lines up with the conversion table.

The sponsor's equity grew by $714.3 million, and you can trace every dollar. EBITDA rose by $50.55 million, which is worth $505.5 million at 10.0x, and debt fell by $238.8 million, for $744.3 million in total. The $30.0 million difference is the transaction and financing fees, which the sponsor paid at closing and never got back.

About two-thirds of the value came from EBITDA growth, about a third from debt paydown, and none from the exit multiple.

The first data table varies the entry and exit multiples and keeps everything else fixed: debt stays at 6.0x EBITDA and the operating case doesn't change, so a higher entry price means a bigger equity check for the same company.

Down the 10.0x exit column, each turn taken off the entry price adds more IRR than the turn before it.

Read across the 10.0x row and the exit multiple moves IRR from 13.9 percent to 26.6 percent. Read down the 10.0x column and paying 9.0x instead of 10.0x lifts IRR from 20.9 to 27.2 percent. That one turn adds about six points of IRR, more than raising revenue growth from 6 to 8 percent (about three points) or doubling the margin improvement (about four).

The same model gives you the maximum price. Set the IRR to the sponsor's target and solve for the entry multiple, either by reading down the 10.0x exit column of the table, which brackets the answer, or exactly with Excel's Goal Seek (Data, What-If Analysis, Goal Seek). With debt held at $600.0 million and a 10.0x exit, Ridgeway clears a 20 percent IRR up to an entry multiple of just under 10.2x, or roughly $1,018 million, and a 25 percent IRR up to about 9.3x, or roughly $932 million.

In Ridgeway's case, holding longer moves the two return measures in opposite directions. Selling after three years gives 1.81x and 21.9 percent, while selling after seven, if growth and the margin gains continue for two more years, gives 3.55x and 19.9 percent. The longer hold earns more dollars at a lower annual rate, because by then much of the debt is paid down and the equity grows more slowly each year.

A second data table sets the amount of debt against the exit year, so its 6.0x row is the hold-period case above. Entry and exit stay at 10.0x, the debt keeps the same 75/25 split between the term loan and the notes, and the financing fees scale with it. Each extra turn of debt adds about 1.5 to 3.5 points of IRR, and a little more with each turn, because the same price is paid with a smaller equity check. The grid keeps the same interest rates at every level, so it leaves out any rise in borrowing costs as the debt goes up.

More debt raises IRR at every hold period, and a longer hold lowers IRR at every level of debt.

How to Do a Paper LBO

A paper LBO is the same model worked with round numbers and a pen, usually in five to ten minutes, so rounding is fine as long as each step is clear. Wall Street Prep's paper LBO guide gives that time range and says rounding to whole numbers is reasonable. Here's a typical prompt: a company has $100 million of EBITDA growing $10 million a year. You buy it at 8.0x with 5.0x of debt at 8 percent interest. D&A and capex are both $20 million a year, taxes are 25 percent, working capital doesn't change, and you sell at 8.0x after five years. What's your return?

1. Price and equity. $100 million times 8.0x is $800 million. Debt is 5.0x, or $500 million, so equity is $300 million. Say all three numbers so the interviewer can follow.

2. Year 1 cash flow. EBITDA of $110 million minus $20 million of D&A is $90 million of EBIT. Interest is 8 percent of $500 million, $40 million, so pre-tax income is $50 million and net income after 25 percent tax is $37.5 million. D&A and capex cancel, so free cash flow is about $38 million, and all of it pays down debt.

3. Years 2 to 5. Repeat, charging interest on each year's beginning debt and rounding to the nearest million. Say the rounding rule once before you start.

Year	EBITDA	EBIT	Interest	Pre-tax income	Free cash flow	Debt, end of year
1	110	90	40	50	38	462
2	120	100	37	63	47	415
3	130	110	33	77	58	357
4	140	120	29	91	68	289
5	150	130	23	107	80	209

4. Exit. Year 5 EBITDA of $150 million at 8.0x is $1,200 million. Minus $209 million of remaining debt leaves $991 million of equity.

5. Returns. $991 million on $300 million is about 3.3x. 3.0x over five years is about 25 percent, so 3.3x is a little more, about 27 percent. Without rounding, the answer is 3.30x and 27.0 percent.

Then add the sentence that shows you understand the deal: EBITDA grew 50 percent and the company paid down almost $300 million of debt, and the multiple didn't change, so the return doesn't depend on the market paying more at exit.

How to Build an LBO Model in a Timed Modeling Test

In a timed test, a finished, simple model counts for more than an ambitious one that runs out of time, so lay out every part first and fill them in order. M&I's test guide puts it plainly: "if you can finish the model and get most of it correct, you should be in good shape." Put the parts on one tab, top to bottom in the order above, with a header on each, so the grader can trace any formula without clicking between sheets. Filling them in order means most formulas point at cells that already exist. Interest is the exception: build the operating model down to EBIT, then the debt schedule, then link interest back. Keeping every part simple, with beginning-balance interest, one sweep, and nothing the prompt doesn't ask for, is what gets you to the returns block in time.

A few habits make a model easy to check:

• Color code: the common convention is blue for numbers you typed in and black for formulas (plus green for links to other sheets, if you use more than one), so a grader can find every assumption at a glance.

• One assumption per row: each projection formula points at an input cell (prior year times one plus the growth cell), so changing the input updates the whole model.

• One sign convention: either costs are negative everywhere and you add the column, or they're positive and you subtract them. Mixing the two is how an expense gets added instead of subtracted.

• Units in every header: "$ in millions" on each block, so figures in thousands never mix with figures in millions.

• State gaps: if the prompt leaves out an input, put your assumption in a labeled input cell and keep going.

Leave out anything the prompt doesn't ask for, such as PIK interest, net operating loss schedules, management rollover and option pools, dividend recapitalizations, purchase price allocation, a full balance sheet, or quarterly periods. Each one takes a real slice of the time without changing the core answer.

Before you hand the model in, run these checks, each of which catches a mechanical error:

• Sources minus uses equals zero: if it doesn't, the equity was typed in instead of calculated as the plug.

• No debt balance goes below zero: the sweep needs a MIN against the balance left.

• Cash stays at the minimum while the term loan falls: if cash piles up instead, the sweep isn't connected.

• Taxes are never negative: in a loss year, a MAX against zero keeps the model from paying the company a refund.

• Working capital uses cash as revenue grows: if growth adds cash, the sign is flipped.

• The answer is believable: IRR near the sponsor's target and debt falling as a multiple of EBITDA every year.

The best practice is to rebuild the Ridgeway example from a blank sheet. Type the assumptions as inputs, build each part in order, and check that Year 5 total debt comes to $361.2 million and IRR to 20.9 percent. Then add a switch cell and flip the interest to average balances. Compare your formulas with the Year 1 block below. Its switch is set to 0, so rows 82 to 87 are the block with beginning-balance interest, and rows 89 to 92 work out Year 1 with the switch on, which is what your block should show once you flip it. The cells with dollar signs, such as $F$24, are fixed references to the assumptions at the top of the sheet.

Average-balance interest cuts Year 1 term loan interest from $36.0 million to $34.8 million, raises cash available to $30.9 million, and leaves the loan at $419.1 million.

Follow-Up Questions That Test Whether You Understand the Model

What makes a good LBO candidate? Steady, predictable cash flow, so the company can carry a lot of debt; low capex and working capital needs, so that cash goes to lenders rather than back into the business; room to grow EBITDA or improve margins; a strong management team; and a purchase multiple that isn't stretched. Assets that can be pledged as collateral help with borrowing too.

Why is an LBO valuation treated as a floor? A DCF asks what the business is worth from its cash flows. An LBO asks what a financial sponsor could pay and still hit its required return with a realistic amount of debt. A sponsor should be willing to buy at any price up to that level, even if no strategic buyer bids, as Macabacus explains, so a seller has little reason to accept less. That's why Breaking Into Wall Street says the LBO valuation "normally sets the 'floor' for a company's value in a deal." Our DCF walkthrough covers the other side of that comparison.

For the questions that tend to come after these, see our 25 private equity interview questions.