What Grant Thornton Must Believe: An LBO of the $5 Billion CBIZ Take-Private

The wire hit on the morning of July 28th: Grant Thornton Advisors, backed by New Mountain Capital, is taking CBIZ (NYSE: CBZ) private for $55.00 a share, all cash, $5.0 billion of enterprise value. My first reaction wasn’t about the premium — a chunky 54% over the 30-day VWAP, 17.8% over the last undisturbed close — it was the itch every deal like this gives me. Before I’d opened a single spreadsheet, I wanted to know whether $55 actually works.

Not “is it a fair price.” Fair is a fairness-opinion word, and Goldman Sachs is already being paid to answer that one for CBIZ’s board. I wanted to know whether the number holds up as an investment — whether the sponsor writing the equity check can plausibly get paid for it.

Why this has to be an LBO, not a merger model

The instinct with any big-ticket M&A announcement is to reach for an accretion/dilution model — model the combined entity, check whether the deal adds to or subtracts from the acquirer’s earnings per share. That instinct breaks here immediately. Grant Thornton Advisors isn’t a public company. There’s no ticker, no market EPS, nothing to accrete or dilute into. The buyer is a private partnership, itself owned by a private equity firm (New Mountain has controlled Grant Thornton since May 2024), funding this through a parent vehicle called Viking ParentCo with New Mountain Partners VII equity and third-party debt.

Take away the public acquirer and what’s left is a much older, much more honest question: does the sponsor’s return clear its hurdle? That’s not a merger model. That’s a leveraged buyout — new debt raised against the target’s cash flows, an equity check sized to make the financing balance, and a bet that the business grows and delevers enough over a hold period to hand back a multiple of that equity. It’s the correct lens for this specific deal, not a stylistic preference.

And because the price is already public, this isn’t the usual forward-looking LBO where you solve for what a sponsor could pay. I’m holding the entry fixed at the announced $55.00 and asking the genuinely open question: given that price, what return is New Mountain actually underwriting, and what has to happen in the business for it to work? Call it an as-announced reconstruction. The entry is ex-post — it already happened. Everything downstream of it is ex-ante, mine to build and stress.

The entry bridge, and a share-count wrinkle that almost got me

CBIZ’s FY2025 numbers are the foundation: $2,758.0 million of revenue, $446.9 million of Adjusted EBITDA (a 16.2% margin), $234.0 million of GAAP operating income, $115.4 million of net income, and a balance sheet carrying $1,472.4 million of gross debt against just $18.3 million of cash — net debt of $1,454.1 million.

Bridge that against the announced $5.0 billion enterprise value and the equity value falls straight out: $5,000.0M minus $1,454.1M of net debt is $3,545.9M of equity going to CBIZ shareholders. Divide by the $55.00 offer price and you get roughly 64.5 million fully diluted shares.

Here’s where I tripped, briefly. CBIZ’s 10-K cover page reports a basic share count closer to 50 million — the standard “shares outstanding as of the most recent practicable date” line every 10-K opens with. Run the deal math off that number and the implied equity value doesn’t reconcile at all; you’d back into a price that isn’t $55. The gap is real and it has a specific cause: CBIZ still owes stock to Marcum’s former partners, delivered monthly over a 36-month schedule tied to the 2024 acquisition of Marcum’s non-attest business. That earn-out stock is dilutive and forward-looking in a way the cover-page basic count doesn’t capture, and it’s the fully diluted figure — not the headline cover number — that actually reconciles to the deal. It’s a small thing, but it’s the kind of small thing that quietly wrecks a model if you grab the first share count Google hands you instead of the one the transaction is actually priced against.

Exhibit A — Entry: EV → equity → shares

Entry multiples land at 1.8x revenue and 11.2x Adjusted EBITDA — full, but not outlandish for a professional-services roll-up with a clean growth story.

Funding it: reconciling to the $5.2 billion committed

The 8-K discloses $5.2 billion of committed financing — New Mountain Partners VII equity plus third-party debt — without breaking out the split. So I built one from first principles instead of taking it on faith.

At 5.5x FY2025 Adjusted EBITDA, a new term loan comes to $2,458.0 million. Layer on financing fees (roughly 2.5% of the new debt) and advisory and other transaction costs (roughly 1.5%, on a base I’ve been conservative about), and total fees land around $155 million. Add that to the $5.0 billion enterprise value being paid for, and total uses come to $5,155.0 million. New Mountain’s sponsor equity is the plug that makes sources equal uses: $2,697.0 million.

Total financing in my build comes to $5,155 million against the $5.2 billion actually committed — within 0.9%. I want to be honest about what that gap means and doesn’t mean: it’s not proof my leverage or fee assumptions are correct, since I built the capital structure bottom-up without seeing the commitment letters. But a bottom-up build landing within a percentage point of a disclosed number I never touched is a decent sanity check, not a coincidence I’d bet the model’s credibility on.

The operating plan New Mountain is actually buying

Debt sizing tells you what was paid for. The operating model tells you what has to happen next for that to have been a good idea.

I modeled 5% annual organic revenue growth — no bolt-on acquisitions, deliberately, since add-on M&A is exactly the kind of upside a sponsor might want to keep separate from the base case rather than baked into it. That takes revenue from $2,758.0 million to roughly $3,520 million by FY2030. Alongside it, Adjusted EBITDA margin expands from 16.2% to 17.4%, pushing EBITDA to about $612 million.

That margin ramp isn’t a rounding assumption — it’s New Mountain’s actual thesis, in numbers. Offshoring back-office and delivery functions, layering in AI-assisted workflow tools, and finishing the integration of Marcum’s practice into CBIZ’s platform are the standard playbook a professional-services sponsor runs, and 120 basis points of margin over five years is a believable, not heroic, read on that playbook working. Capex stays light at 0.6% of revenue, appropriate for an asset-light services business, and working capital drags about 8% of incremental revenue — the ordinary cost of a growing services book carrying more receivables.

Where the equity actually gets made: deleveraging

A buyout doesn’t generate its return primarily from the operating plan above — it generates it from paying down debt with the cash that plan throws off, so more of a growing enterprise value accrues to a shrinking equity check.

The term loan carries an all-in rate around 8.5% (SOFR near 4% plus a 450 basis-point spread), with 1% scheduled amortization and a 100% cash sweep absorbing everything else. I charged interest on the opening balance each year rather than an average balance, which avoids a circular reference between interest expense and the cash available to pay it down — a modeling choice, not a real-world one, but the kind that matters if you want the sheet to actually calculate instead of fighting Excel’s iterative-calc settings.

The result is a genuinely steep delevering path: net debt-to-EBITDA starts at 5.5x at entry and finishes the five-year hold around 1.9x, with roughly $1,305 million of debt retired and exit net debt down to about $1,135 million. That’s the unglamorous machinery every LBO return actually runs on.

The number that made me sit up: 16.2%

Five-year hold, exit at the same 11.2x multiple CBIZ was bought at — a deliberately “multiple-neutral” exit, so the model isn’t quietly assuming a friendlier buyer shows up later. Exit enterprise value comes to about $6,860 million off $612 million of exit EBITDA. Subtract exit net debt of $1,135 million and exit equity value is roughly $5,725 million.

Against New Mountain’s $2,697 million entry check, that’s a 2.12x MOIC and a 16.2% IRR.

I expected this deal, at the price actually announced, to clear a normal buyout hurdle without much drama. It didn’t — not comfortably. Most sponsors underwrite new platform deals to something north of 20% IRR, and 16.2% sits meaningfully below that, even though 2.12x is a perfectly respectable multiple of money on its own. That gap between “solid” and “what we actually underwrite to” is the most interesting number in the whole model, and it’s the point where the exercise stops being a valuation check and turns into a question about what New Mountain believes.

Exhibit B — Returns bridge

So what does New Mountain actually need to believe?

This is the part I built the model to get to. If the base case falls short of a normal hurdle, the useful question isn’t “is the deal bad” — it’s “what has to be true for it to be good.”

Running the same structure and solving backward for the exit multiple each target IRR requires, the picture sharpens fast. Clearing 15% only needs 10.7x — actually a touch of multiple contraction from the 11.2x entry, which the base case already beats. Clearing 20%, the more standard sponsor hurdle, needs about 12.8x — roughly 1.6 turns of multiple expansion over the entry price. Clearing 25% needs 15.3x, over four turns of expansion, which starts to look like a very different bet than “buy a stable professional-services roll-up.”

Exhibit C — What must be true: required exit multiple by target IRR

So the model’s real output isn’t a price target — CBIZ already has one, and it’s $55.00. It’s a reverse-engineered thesis. To underwrite anything close to a normal buyout return, New Mountain is either betting on roughly 1.6 turns of multiple expansion by exit, an operating plan that beats CBIZ’s own guided range, meaningful add-on M&A I deliberately didn’t model, or — most likely, knowing how these deals get built — some blend of all three. None of that makes the deal irrational. It makes it a bet with real, specific, nameable conditions attached, instead of a number that just sounds big enough to justify itself.

Where I could be wrong

A model like this is only as honest as its caveats section, so here’s what I’d push back on if I were reviewing my own work. The 5.5x entry leverage and the 8.5% all-in debt cost are assumptions, not facts — I won’t know the real commitment letter terms until the merger proxy is filed. The margin ramp from 16.2% to 17.4% is New Mountain’s thesis as I’ve modeled it, not a disclosed target; it could be conservative or aggressive, and I have no way to check which yet. I deliberately kept D&A and the tax line simple, ahead of the purchase-accounting step-up that a real post-close balance sheet will apply — that line is likely to move once real numbers exist. And the required-exit-multiple sensitivity in Exhibit C holds exit net debt fixed on the base-case deleveraging path, so it isn’t a full two-dimensional grid against leverage as well as multiple — a reasonable simplification, but one worth naming rather than hiding.

None of this makes the 16.2% number fake. It makes it a well-reasoned first pass, built entirely from what’s public as of July 29th, that I plan to test against better information the moment it exists.

What I’m checking next

CBIZ’s go-shop window runs through August 27th, and the real test of everything above lands when the merger proxy — the PREM14A, then the DEFM14A — actually files. That document will carry Goldman Sachs’s fairness opinion, the real financing commitment letters, and very likely management’s own projections instead of my reverse-engineered growth and margin assumptions. Once it’s out, I want to lay my number next to the Street’s and see how much of the gap between “solid” and “what sponsors actually underwrite to” survives contact with real numbers.

Sources: CBIZ, Inc. Form 10-K for fiscal year 2025 (SEC EDGAR); CBIZ Q4 / full-year 2025 results release; the merger 8-K filed July 28, 2026; and the DEFA14A filed July 29, 2026. Filed documents can be located directly through the SEC’s EDGAR full-text search under CBIZ, Inc.

This is an educational analysis built entirely from public filings and disclosures. It is not investment advice, and it is not affiliated with CBIZ, Grant Thornton Advisors, New Mountain Capital, or Goldman Sachs. Every operating, financing, and exit assumption in the model is my own and independent of any party to the transaction. All figures above are subject to revision once the merger proxy (PREM14A/DEFM14A) and underlying financing commitment letters are filed.