LBO (Leveraged Buyout)

Key takeaways

Buy a company mostly with borrowed money. Use the company's cash to pay it down. Sell the company years later. The leverage amplifies returns.
Three return drivers: debt paydown, EBITDA growth, multiple expansion. The last is the most fragile — count on it least.
The non-negotiable mechanics: Sources = Uses (must balance exactly), interest computed on beginning debt balance (avoids circularity), debt balances capped at zero with =MAX(0, ...), returns measured by IRR (typically XIRR) and MOIC.
Cash sweep follows tranche priority — senior debt before sub before mezz — and excess FCF after mandatory amortisation flows to whichever tranche is next in the waterfall.
Sensitise entry multiple × exit multiple on both IRR and MOIC. That's the single biggest swing factor in returns and the variable sponsors negotiate hardest.

The model in four sections

A working LBO is built as four linked sheets, in dependency order. Don't build end-to-end then check; verify at each section break.

Sources & Uses — how the deal is financed. Must balance exactly.
Operating Model — five-year projection of revenue, EBITDA, capex, working capital, taxes, and the free cash flow available for debt service.
Debt Schedule — tranches, interest, mandatory amortisation, optional cash sweeps, ending balances, leverage covenants.
Returns Analysis — exit equity value, IRR, MOIC, sensitivity tables.

Sources & Uses

The first sanity check: every dollar paying for the deal has to be sourced. Sources and uses must balance to the cent — if they don't, the model is wrong.

Uses (where the money goes)

Purchase price = LTM EBITDA × Entry Multiple
Transaction fees — typically 2–4% of EV (advisor, legal, accounting, due diligence)
Financing fees — typically 1–2% (loan origination, OID, commitment fees)
Refinancing existing debt — assume the buyer takes out the existing capital structure unless told otherwise

Sources (where the money comes from)

Senior / Term Loan — typically 4–5× EBITDA, 7-year maturity, 1% annual mandatory amortisation, SOFR + spread
Subordinated debt / High-yield notes — typically 1–2× EBITDA, bullet maturity, fixed rate
Mezzanine — typically 0.5–1× EBITDA, often with PIK toggle, equity warrants
Revolver — committed line, drawn only if needed
Sponsor equity — the plug. Computed as residual: Sponsor Equity = Total Uses − All Other Sources
Management rollover — sometimes part of the equity

The arithmetic check at the bottom of the sheet: Sources − Uses = 0. If it's not zero, you have a missing line item.

Debt schedule mechanics

Interest on the beginning balance — not the average

The temptation is to compute interest on the average of beginning and ending debt balances. Don't. Ending balance depends on free cash flow, which depends on net income, which depends on interest. You've created a circular reference — Excel will warn and your formulas may resolve to nonsense.

Interest_t = Beginning Balance_t × Rate
(NOT Average × Rate, NOT Ending × Rate — both create circularity)

Cash sweep waterfall

After mandatory amortisation, any excess free cash flow "sweeps" into prepaying debt — but it must respect tranche priority:

Pay mandatory amortisation on senior
Pay scheduled interest on all tranches
If excess FCF remains, prepay senior to zero
Then prepay subordinated to zero
Then prepay mezzanine to zero

The mechanical issue: if mandatory amortisation plus the cash sweep exceeds the beginning balance, the formula naively goes negative. Wrap each balance in a floor:

Ending Balance = MAX(0, Beginning − Mandatory − Sweep)

Leverage covenants

Most senior debt carries financial maintenance covenants. Track each year:

Total Debt / EBITDA — typical covenant: stays under 6.0× initially, stepping down to 4.5× by year 4
Senior Debt / EBITDA — typical covenant: stays under 4.0×
EBITDA / Interest (coverage) — typical covenant: stays above 2.0×

Build a covenant compliance row that flags red if any year violates. A breach in the model means the structure is too aggressive.

Returns analysis

The bridge from exit to equity returns

Exit EV = Exit Year EBITDA × Exit Multiple
Exit Equity = Exit EV − Net Debt_exit
MOIC = Exit Equity / Entry Sponsor Equity
IRR = XIRR(cash flows, dates)

Cash-flow signs for IRR

For Excel XIRR:

Entry sponsor equity = negative at the deal date
Any interim distributions (dividend recaps, leveraged dividends) = positive at their dates
Exit proceeds = positive at the exit date

Reversed signs give meaningless results. XIRR is preferred over IRR because it handles real dates, not just annual periods — material when entry and exit aren't on calendar year-ends.

Worked example — the math, end to end

Target: $100M of LTM EBITDA. Entry multiple of 10×. Hold for 5 years. Exit at 11× on year-5 EBITDA of $150M (50% growth). Capital structure on entry: 5× senior debt, 1× subordinated, the rest sponsor equity.

Entry sources & uses	$M
Purchase price (10× × $100M)	1,000
Transaction & financing fees (~3%)	30
Total uses	1,030
Senior debt (5× × $100M)	500
Subordinated debt (1× × $100M)	100
Sponsor equity (plug)	430
Total sources	1,030

Through the hold, assume the company generates $300M of cumulative FCF after interest and capex, all used for debt paydown. Net debt at exit drops from $600M entry to $300M.

Exit math	$M
Year-5 EBITDA	150
× Exit multiple	11.0×
= Exit EV	1,650
− Net debt at exit	(300)
= Exit equity	1,350
÷ Entry sponsor equity	430
MOIC	3.14×
IRR (5-year hold)	~25.7%

Decomposing the return drivers:

Multiple expansion (10× → 11×): adds ~$150M of equity value (~10% of returns)
EBITDA growth ($100M → $150M at 10× constant): adds ~$500M of EV, after net debt (~50% of returns)
Debt paydown ($600M → $300M): adds $300M directly to equity (~40% of returns)

Multiple expansion is the smallest driver and the one you control least. EBITDA growth requires a real operational thesis. Debt paydown is the most reliable — it's just arithmetic if the cash flows materialise.

Sensitivity — minimum two 5×5 tables

Table	Axes	Output
1	Entry multiple × Exit multiple	IRR
2	Entry multiple × Exit multiple	MOIC
3 (optional)	Revenue growth × EBITDA margin	IRR
4 (optional)	Hold period × Exit multiple	IRR

Color-code the IRR table: green > 20%, yellow 15–20%, red < 15%. Sponsors typically underwrite 20%+ on the base case to leave room for the bear; below 15% means the deal probably doesn't pencil.

The most common LBO errors

Circular interest — using average or ending debt balance for the interest calc. Always use beginning.
Sources don't balance Uses — every model needs a plug (usually sponsor equity); compute it as residual, then verify the difference is zero.
Negative debt balances — if mandatory + sweep exceeds the balance, the cell goes negative. Cap at zero with MAX.
Reversed cash-flow signs in IRR — entry must be negative, exit must be positive. Reversed signs return meaningless IRR.
Forgotten transaction fees — uses should include the 2–4% transaction fees and 1–2% financing fees. Forgetting them inflates returns by hundreds of basis points.
Multiple expansion baked in as base case — assuming the exit multiple equals the entry multiple is the conservative anchor. Underwriting expansion (entry 10×, exit 12×) at base is sponsor's wish, not investor's reality.
No covenant compliance check — building the model without flagging covenant breaches produces a structure that wouldn't actually clear lender approval.
Hardcoded interest expense — should be = Beginning Balance × Rate per tranche, then summed. Not a single hardcoded number.

How Smalt AI builds it

Tell your AI coworker the target, the entry multiple, and the leverage assumption — "LBO ABC at $42/share, 6.5× total leverage, 5-year hold." What you get:

Sources & Uses with sponsor equity computed as residual; balance check at the bottom is enforced (Sources − Uses = 0 to the cent).
Operating Model — 5-year revenue / EBITDA / capex / working-capital projection, with margins reasonable vs historical, scenario blocks for bear / base / bull.
Debt Schedule with senior, subordinated, and revolver tranches; interest computed on beginning balance per tranche; cash sweep waterfall respects priority; =MAX(0, ...) floors prevent negative balances; leverage and coverage covenants tracked per year with red flags on breaches.
Returns Analysis — full equity bridge, IRR via XIRR (real dates), MOIC, decomposition of return drivers (debt paydown vs EBITDA growth vs multiple expansion).
Two minimum 5×5 sensitivity tables: entry × exit multiple → IRR and MOIC, both with full recalculations and color-coded by return band.
Quality checks auto-run before delivery: Sources = Uses, debt schedule ties to balance sheet, no negative balances, IRR signs correct, covenants flagged.