Underwriting · 11 min read · Updated November 2026
Management Case vs Base Case vs Downside Case: How Private Equity Underwrites a Recession
Management case vs base case vs downside case in an LBO: who builds each, how much to haircut, how to build a recession case and test it against covenants.
The management case is the seller’s plan, the base case is the buyer’s own view, and the downside case is the bad year the deal has to survive. Private equity firms price the deal and size the debt on the base case, then run the downside to check that the company stays inside its covenants and the equity is still worth something.
On a simplified teaching schedule of the house case on this site (assumptions below), the same $150 million buyout returns 3.79x in the management case, 3.06x in the base case and 2.09x in a year-2 recession, with leverage peaking at 2.37x against a 3.0x covenant. Add one slow year of collections to that recession and the covenant breaches while the return barely moves: the downside’s first job is liquidity, not the return.
Management Case vs Base Case vs Downside Case
Putting the cases on a single switch lets you see which operating assumptions change the investment decision. They differ in who builds them, what they assume and what they are used for.
| Case | Who builds it | What it assumes | What it is used for |
|---|---|---|---|
| Management case | The seller and its management, with the banker | The business plan in the marketing materials | The starting point, and the upside the sponsor is not paying for |
| Base case (sponsor or investment case) | The buyer’s deal team | What the team believes will happen, after diligence and haircuts | The price, the target return and the debt package |
| Bank case | The lenders, from the sponsor’s base case | Often the base case, sometimes trimmed further | Sizing the debt and setting covenant levels |
| Upside case | The buyer’s deal team | Management’s plan or the value creation plan delivered | What the deal could return if additional improvements materialize |
| Downside case | The buyer’s deal team | A recession or a combined set of bad assumptions | Testing covenants, liquidity and the equity’s survival |
Terms vary by firm. Roberto Ippolito, writing from the lender’s side, calls the case the debt is sized on the base case, also called the bank case, and sets covenants on it (Private Capital Investing, ch. 11). What matters is that the case that sets the price is the one the team believes, not the one the seller hopes for.
Why the Management Case Is the Starting Point, Not the Case
Management plans are written to sell a company. Tamara Sakovska, a private equity investor whose career runs through Warburg Pincus and Permira, writes that in her own transactions she met only one CEO her team thought too conservative (The Private Equity Toolkit, 2022, ch. 7). She lists five problems that recur in the rest:
- Excessive optimism. Hockey-stick projections, usually driven by fast market growth and share gains.
- An unrealistic timeline. Sales cycles that are too short, launches that never fail and more initiatives than one holding period can carry.
- An idealized view of the company’s capabilities. Plans that assume skills or talent the business does not have yet, such as an engineering company that turns into a sales organization on schedule.
- Missing costs and investments. Restructuring that saves money but costs nothing to execute, acquisitions with no deal costs, no cost inflation, too little capex. Her rule for any plan that leans on a new ERP system: double the money and the time.
- Weak capital allocation and cash management. Detailed revenue lines with working capital taken as a flat percentage of sales and capex left on autopilot.
In a study published in 2016, Gompers, Kaplan and Mukharlyamov surveyed 79 buyout investors. Of the 44 who gave a numerical discount to management forecasts, both the average and median were about 20%, measured against EBITDA. Another 11 said they varied the discount by deal. Apply the adjustment where the evidence points: for example, shift a delayed launch by one year rather than cutting every year’s earnings by the same percentage. Authors’ working paper, Table 6
You have one hour. Build an integrated three-statement LBO from a blank Excel workbook.
Prepare a base case and a second case with a general economic recession in the second year.
How to Build a Downside Case
A downside case is not one sensitivity. Sakovska separates the two: sensitivities show which assumptions move the outcome most, and the downside combines four or five of them into one punitive scenario, an outcome she puts at roughly a 10% to 15% probability. Her bar for the base case is the mirror image: one the team believes it can hit 60% to 70% of the time (pp. 154-157). These are one investor’s yardsticks, not an industry standard; what matters is that a team states the odds it is underwriting to.
For an operating business, a useful downside moves the same lines a recession would:
| Lever | What to change | Why it matters |
|---|---|---|
| Revenue | A drop in year 1 or 2, then a slower recovery than the base case | Sets the size of the EBITDA hole |
| Gross margin | Lower pricing power and less absorption of fixed costs | Margins fall faster than revenue in most recessions |
| Working capital | Slower collections, inventory that does not turn | Decides whether the revolver gets drawn |
| Capex | Maintenance capex stays; growth capex is cut | Shows how much cash management can protect |
| Exit | Same multiple, or one turn lower | Separates an operating miss from a valuation miss |
Then report the four numbers a credit committee reads first: peak leverage, the lowest interest cover, the first covenant breach and the largest revolver draw. Only after those does the equity return matter.
The Downside’s First Job: Which Covenant Breaks, and When
A covenant tells you how much room the loan documents allow. An uncured breach can give lenders remedies, including acceleration. A waiver or amendment may carry a fee, higher interest, tighter terms or a requirement to contribute equity. The agreement sets the test date, cure rights and remedies.
A maintenance covenant is tested on specified reporting dates. A springing test can depend on revolver usage or remaining availability. An incurrence covenant applies when the borrower takes a specified action, such as issuing debt or paying a dividend. Read the actual financing documents to identify which tests apply. Example of a lender’s covenant-lite disclosure
The house case under five cases
Industrial Inc. is bought for $150 million (6.0x $25 million of EBITDA) on $200 million of revenue, with $80 million of debt and $80 million of equity. Each case changes the operating inputs below and runs through the same teaching schedule.
| Case | Revenue | EBITDA margin | Working capital and capex |
|---|---|---|---|
| Management | Up 8% a year, to $293.9M in year 5 | 13.9%, rising to 15.5% | Inside the 66.5% cash conversion |
| Base | Up 5% a year, to $255.3M | 13.8%, rising to 15.0% | Inside the 66.5% cash conversion |
| Downside (year-2 recession) | $210.0M, then down 10% to $189.0M in year 2; up 3%, 4% and 4% after | 12.3% in year 2, back to 14.0% by year 5 | Inside the 66.5% cash conversion |
| Downside with slow collections | As the downside | As the downside | Receivables 30 days slower in year 2: $15.5M more cash tied up, released in year 3 |
| Flat | $200.0M every year | 12.5% | Inside the 66.5% cash conversion |
The schedule converts 66.5% of each year’s EBITDA into cash before financing. This shortcut combines operating tax, capex and normal working capital into a single assumption. It does not forecast each one independently. The fourth case adds what that shortcut hides: a year of slow collections, the same 30-day slip the full house case models as its own one-year shock.
| Management | Base | Downside | Downside, slow collections | Flat | |
|---|---|---|---|---|---|
| Year-5 EBITDA | $45.5M | $38.3M | $29.5M | $29.5M | $25.0M |
| MOIC | 3.79x | 3.06x | 2.09x | 2.08x | 1.68x |
| IRR | 30.6% | 25.1% | 15.9% | 15.7% | 10.9% |
| Highest forecast year-end debt ÷ EBITDA | 2.18x | 2.28x | 2.37x | 3.04x | 2.75x |
| Lowest EBITDA ÷ interest | 4.3x | 4.1x | 3.9x | 3.9x | 3.5x |
| Tightest covenant headroom | 33% | 30% | 21% | Breach (Y2) | 15% |
| Largest new revolver draw after closing | None | None | None | $5.3M | None |
- The base case clears this example’s targets. It delivers 3.06x and 25.1% against the assumed 3.0x and 25% hurdles.
- The recession alone never breaches. Leverage peaks at 2.37x in year 2 against a 3.0x covenant, 21% headroom at the tightest point, and the sponsor still gets 2.1x its money back.
- Slow collections do what the recession did not. Thirty more days of receivables in the recession year tie up $15.5 million that debt has to fund: the sweep stops, the revolver draws $5.3 million, leverage reaches 3.04x and the covenant breaches. At year-end the revolver has funded $5.3 million and the term loan has repaid less principal. Cash comes back in year 3, so the five-year return moves little. The modeled IRR assumes the breach is resolved without a fee, repricing or an equity cure.
The management case is upside the sponsor is not paying for. If the same deal carried 5.0x of debt, the recession alone would breach the covenant in years 3 and 4; the debt capacity guide works that comparison through.
Horizontal bars of sponsor IRR: management case 30.6% (3.79x), base case 25.1% (3.06x, highlighted), downside case 15.9% (2.09x), downside with slow collections 15.7% (2.08x) and flat case 10.9% (1.68x). A reference line marks a 25% target.
A Real Case: Two Debt Structures, Two Answers
Agilent announced the December 2005 sale of its semiconductor products group to KKR and Silver Lake for $2.66 billion, subject to closing adjustments. A teaching case on the deal models a $2,578.8 million enterprise value, or 9.6x its $267.7 million of LTM EBITDA. It compares two ways to raise $1,725 million of debt. The returns below are the teaching case’s scenarios. Agilent closing announcement · INSEAD case
| Bank loan structure | Hybrid structure | |
|---|---|---|
| Debt | $1,200M of bank loans and $525M of subordinated notes | $725M of bank loans and $1,000M of senior and subordinated notes |
| Maintenance covenants | Secured debt 4.0x, total debt 6.0x and interest cover 2.0x | Secured debt 4.0x only |
| Base case gross IRR (2009 to 2011 exit) | 12.9% to 13.1% | 11.9% to 12.1% |
| Downside total debt ÷ EBITDA, peak | 9.9x | 10.3x |
| Downside gross IRR (2009 to 2011 exit) | −23.9% to 0.0% | −27.6% to −1.7% |
In the downside, revenue falls 10% and then 5%, and EBITDA drops from $238.0 million to $129.8 million in two years. Read the exhibits together and the base case prefers the bank loans, which are cheaper and return about a point more. The bank structure has the higher modeled equity return in the downside too, but its maintenance covenants breach. The hybrid has more room under its modeled tests. Under the bank structure, total debt reaches 9.5x in the first recession year against a 6.0x maintenance test, and interest cover falls to 1.3x against 2.0x; both stay in breach for four years. Under the hybrid structure, bank debt peaks at 2.2x EBITDA against its 4.0x secured debt test, and the bonds carry only incurrence covenants, so the modeled secured-debt test still passes. Other obligations and liquidity would still need to be checked.
The lesson travels to any size of deal. A structure chosen on the base case alone optimizes the outcome you expect; the downside tells you what the structure does in the outcome you fear.
How to Show Cases in a Modeling Test
Several timed tests ask for more than one case, and reviewers read them the same way:
- Put the case switch at the top of the assumptions, with every case on one tab so a reviewer can flip it and watch the outputs move.
- Name what changed in each case in one line: “Downside: revenue −10% in year 2, gross margin −200 bps, DSO +10 days.”
- Show the outputs side by side: IRR and MOIC, peak leverage, lowest coverage, first breach, revolver peak.
- Say which case the price rests on. If it is the management case, say so and expect the question.
- Close with the call: “The deal survives a year-2 recession with 21% covenant headroom and returns 2.1x; the base case clears 25%. A 30-day slip in collections that year would breach the covenant, so receivables and revolver availability are the first diligence items.”
Underwriting case traps
| Trap | What goes wrong | How to catch it | Fix |
|---|---|---|---|
| Pricing on the management case | The returns page leads with the seller's plan and the base case appears only as a sensitivity. | Ask which case sets the bid and the debt. | Price and size the debt on the base case; show management as upside. |
| A base case doing the upside case's work | The base case assumes every initiative lands on time. | List each initiative in the base case and its odds. | Identify the assumptions the deal team can support with evidence, then isolate the initiatives it treats as upside. |
| A downside that only trims growth | The downside grows a little slower and never tests the debt. | Check whether revenue, margin and working capital all move. | Combine several bad assumptions into one recession case. |
| No covenant line in the model | Leverage and coverage are shown, but nothing to compare them with. | Look for a covenant row with headroom by period. | Add the covenant schedule and report headroom and the first breach. |
| Same exit multiple in every case | The downside exits at the entry multiple after a recession. | Ask whether a buyer would pay the same multiple for that business. | Show the downside at the entry multiple and one turn lower. |
Before you send a model with cases
- One switch that runs every case through the same schedule.
- A base case you would back most of the time, with the haircut to management explained.
- A downside that moves revenue, margin and working capital together.
- A covenant schedule, with headroom by period and the first breach flagged.
- Peak revolver draw and lowest liquidity in the downside.
- Returns for every case side by side, with the case that sets the price named.
Frequently Asked Questions
What is the difference between the management case and the base case? The management case is the seller’s business plan, usually built to support a sale. The base case is the buyer’s own view of the same business, built to be achieved most of the time. Sponsors price the deal and size the debt on the base case and treat anything above it as upside.
How much do private equity firms haircut management projections? It depends on the plan and the evidence. In the Gompers, Kaplan and Mukharlyamov survey published in 2016, 44 respondents quantified their discount to management forecasts. The mean and median were about 20%, measured against EBITDA. Another 11 said the discount varied by deal. Rebuild the forecast from specific assumptions before comparing the result with that historical sample.
What is a downside case in an LBO? A punitive scenario that combines several bad assumptions at once, such as a recession year with lower revenue, thinner margins and more working capital. Sakovska describes one with roughly a 10% to 15% probability. Its job is to show whether the company breaches its covenants or runs out of liquidity, and what the equity is worth if it does not.
What is the difference between a sponsor case and a bank case? The sponsor case is the private equity firm’s investment case. The bank case is the plan the lenders size the debt on, often the sponsor’s base case or a more conservative version of it. Covenants are set off that case with headroom, so the downside test shows how much room the business has.
How do you stress test an LBO? Change one assumption at a time to see which ones move returns most, then combine the worst plausible changes into one downside case. Report peak leverage, the lowest interest cover, the first covenant breach and the revolver draw, then the equity return if the business survives.
What probability should a base case have? Tamara Sakovska calls a base case solid if the deal team believes it can be achieved with a probability of 60% to 70%, and treats everything in the management plan beyond it as potential upside. That is her working yardstick, not an industry standard; firms calibrate their own.
Why can a working capital shock matter more than an EBITDA drop? Slow collections can force the company to borrow. Here, delaying collections by 30 days on $189 million of annual sales ties up $15.5 million in year two. Closing debt rises to 3.04x EBITDA against a 3.0x limit. Collections recover in year three, so the five-year IRR drops only from 15.9% to 15.7%, before any cost of resolving the breach.
Sources
Books and frameworks (the case probabilities are one author’s yardsticks):
- Tamara Sakovska, The Private Equity Toolkit (Wiley, 2022), ch. 7, “Analyzing the Business Plan”: the five recurring problems in business plans; the 20% haircut reported by 44 funds, p. 149; base and downside case probabilities, pp. 154-157.
- John Gilligan and Mike Wright, Private Equity Demystified, 1st ed. (ICAEW Corporate Finance Faculty, 2008): covenants, defaults and renegotiation, pp. 54-55; headroom of 20% to 50%, p. 67.
- Roberto Ippolito, Private Capital Investing (Wiley, 2020), ch. 11: the base or bank case and setting covenants on it.
Research:
- Paul Gompers, Steven Kaplan and Vladimir Mukharlyamov, “What do private equity firms say they do?” Journal of Financial Economics 121(3), 2016: a survey of 79 private equity investors, the source Sakovska cites for the haircut.
A historical transaction:
- Claudia Zeisberger, Michael Prahl and Bowen White, Private Equity in Action (Wiley, 2017), Case 12, “Chips on the Side (B): The Buyout of Avago Technologies”: Exhibits 12.1, 12.7, 12.8 and 12.9, and the discussion of maintenance and incurrence covenants. Agilent announced the sale to KKR and Silver Lake in August 2005 at $2.66 billion, subject to closing adjustments (Agilent).
Our calculations:
- The house case: PE Case Study: $150M Distribution LBO. The five cases, their inputs and results are our own illustration on the teaching schedule described above, not the case’s full model.
Revision History
- : Checked the original forecast survey, reconciled the five operating cases and separated Agilent's closing price from the Avago teaching case.
- : First published, with operating cases on the house distribution case and the Avago financing exhibits.