SkillsSkill 2 of 5
Know where the return comes from
Split every return into growth, debt paydown and multiple before you trust it.
Worked example
The model says 2.5x. Where does the return come from?
A model can produce the answer in seconds. The interview tests whether you know which assumption is doing the work.
The question
Quillfield, a fictional claims software and document-review business, earned $40m of EBITDA in FY25. The seller’s price is 10.0x, or $400m. An AI tool built the model below. The fund underwrites to 2.5x and a 20% gross IRR over five years.
| Entry | Exit, model | Exit, flat multiple | |
|---|---|---|---|
| EBITDA | 40 | 50 | 50 |
| Multiple (x) | 10 | 12 | 10 |
| Enterprise value | 400 | 600 | 500 |
| Net debt | 200 | 100 | 100 |
| Equity value | 200 | 500 | 400 |
Invented data. Debt is 5.0x entry EBITDA; $100m of free cash flow repays debt over five years. No fees, dividends or management equity. MOIC is exit equity ÷ $200m; IRR = MOIC^(1/5) − 1, rounded to 0.1%: 2.50x is 20.1%, 2.00x is 14.9%.
“The model shows 2.50x and a 20.1% IRR at a $400m price. Walk me through where that return comes from, and would you pay $400m?”
Your turn
Say your answer out loud, or jot notes below. Both work. Or skip to the worked answer.
Not saved by this site. Your browser may restore text when you navigate back.
A strong answer, step by step
Open each step in turn. This is one strong answer, not the only one.
Step 1: Do the paper LBO first
What would you do here? Open the move you’d make.
Quote the model: 2.50x and 20.1% clears the fund’s 2.5x and 20% hurdle.
Tempting, but it misses something. It repeats the output without testing it. The interviewer wants to see you get from $200m to $500m yourself, and to notice that the answer depends on a 12.0x exit.
Rebuild it: $200m of equity becomes $500m, 2.50x and 20.1%. At a flat 10.0x it’s $400m, 2.00x and 14.9%.
The strong move. Doing the arithmetic exposes the input that matters: the same deal at an unchanged multiple misses both hurdles.
Audit the model’s IRR formula and cell links before discussing the return.
Tempting, but it misses something. Auditing is useful later, but the formula is right: 2.50x over five years is 20.1%. The risk is the 12.0x assumption, not the spreadsheet.
See how a strong candidate says it
“$200m of equity becomes $500m: 2.50x, about 20.1% over five years. At a flat 10.0x exit it’s $400m, 2.00x and 14.9%, below our hurdle.”
Why it works: It shows you can reproduce the model’s answer without the model.
Step 2: Split the $300m gain by source
What would you do here? Open the move you’d make.
Credit the return to EBITDA growing from $40m to $50m: the business got better.
Tempting, but it misses something. Growth at the entry multiple is only 10 × 10 = $100m of the $300m gain. The other $200m is debt paydown and the move from 10.0x to 12.0x.
Credit the leverage: $200m of debt on a $400m price is what makes the equity return work.
Tempting, but it misses something. Debt paydown is $100m, a third of the gain. Stopping there hides that another $100m comes only from multiple expansion.
Split the $300m three ways: growth at 10.0x is $100m, debt paydown is $100m, and 2 turns of expansion on $50m is $100m.
The strong move. Attribution shows that a third of the gain, and the whole gap between 2.00x and 2.50x, rests on the exit multiple.
See how a strong candidate says it
“EBITDA growth at the entry multiple is 10 times 10, $100m. Debt paydown is $100m. Multiple expansion, 2 turns on $50m, is the other $100m.”
Why it works: Attribution shows that a third of the gain, and all of the hurdle, rests on the exit multiple.
Step 3: Challenge the assumption doing the work
What would you do here? Open the move you’d make.
Ask why a buyer would pay 12.0x for what we buy at 10.0x, when over 40% of revenue is review that AI could shrink.
The strong move. It tests the assumption doing the work against the business: 102 of 240 of FY25 revenue, 42.5%, is document review, which argues for a flat or lower multiple.
Accept 12.0x because comparable software deals traded there.
Tempting, but it misses something. Comps price pure software. Quillfield has 42.5% of revenue in review; underwriting a software multiple at exit means paying today for a re-rating nobody has earned.
Use the midpoint, an 11.0x exit, as a compromise base case.
Tempting, but it misses something. Splitting the difference isn’t a view. At 11.0x exit equity is $450m, 2.25x, still below 2.5x, and it skips the question of which multiple the business deserves.
See how a strong candidate says it
“Why would a buyer pay 12.0x for a business we buy at 10.0x? Over 40% of its revenue is document review that AI could shrink, so I’d assume a flat or lower multiple.”
Why it works: It ties the mechanics back to the investment judgment.
Step 4: Turn it into a price
What would you do here? Open the move you’d make.
Pay $400m and negotiate cheaper debt to lift the return.
Tempting, but it misses something. Financing moves the return at the margin. The gap here is 2.00x against 2.5x at a flat multiple, and it comes from the price, not the debt.
Solve for the price: at a flat multiple, 2.5x needs an 8.0x entry, $320m, so $120m of equity becomes $300m.
The strong move. It ends in a number the team can act on: 2.5 × (40 × 8.0 − 200) = 50 × 8.0 − 100 = $300m.
Call the deal “interesting at the right price” and leave the number to the partners.
Tempting, but it misses something. Without a number the analysis stops short. The table already holds everything you need to solve for 8.0x, or $320m.
See how a strong candidate says it
“At a flat multiple, 2.5x needs an 8.0x entry: $320m, so $120m of equity becomes $300m. I wouldn’t pay $400m unless diligence supports multiple expansion.”
Why it works: A scoped answer ends in a number the deal team can act on.
The answer, in short
- Decision
- Don’t pay $400m on this model. At a flat exit multiple, the hurdle return implies about 8.0x, or $320m.
- Because
- The model clears 2.5x only because it assumes the multiple rises from 10.0x to 12.0x; without that it returns 2.00x.
- What would change it
- Evidence that the software business will be a larger, faster-growing share of EBITDA at exit, which could support a higher multiple.
Common traps
- Quoting the model’s IRR without checking it.
- Treating multiple expansion as a base case.
- Forgetting that leverage adds risk as well as return.
- Giving a range of views but no price.
Check your answer
- Could I reproduce 2.50x and 20.1% without the model?
- Did I split the gain into growth, paydown and multiple?
- Did I test the exit multiple against the business?
- Did I end with a price and what would change it?
Want feedback on how you’d answer? Practice “Know where the return comes from” with an AI interviewer on Socratify.
Last reviewed Nov 9, 2026