Peak-Multiple LBOs: How Blackstone Paid 21.7x EBITDA for Hilton and Still Won

There's a rule senior partners drill into associates the first month on the desk: don't pay peak multiples. The arithmetic is unforgiving — buy a business at 12× EBITDA, sell it at 9×, and the multiple contraction alone wipes out most of your operating improvement and a chunk of your debt paydown on top of that. The exit math is a one-way ratchet, and the way you avoid getting caught in it is to never close a deal where the multiple is materially above the long-run average for the sector.

Blackstone broke that rule on Hilton in October 2007. They closed at a $26.0B enterprise value on $1.2B of LTM EBITDA — a 21.7× entry multiple, at the peak of the late-2000s hospitality cycle, six weeks before the global hotel demand chart fell off the table.

It became the largest dollar return in PE history.

The lesson isn't "go ahead and pay peak." It's that peak-multiple LBOs occasionally clear because of a specific set of structural advantages, and the discipline lives in knowing which ones you have and which ones you don't.

What "peak multiple" actually means

Practitioners use the phrase loosely. The precise version is: an entry EV/EBITDA multiple that sits at the high end of the trailing 10-year range for comparable transactions in the sector, in an environment where the public comps are also at the top of their decade.

Three things distinguish a peak-multiple deal from an aggressive deal:

1. Sector multiple is at a cyclical high. The whole industry is being priced for a continuation of the current expansion. There is no "the multiple compressed because the market hates this sector"; the multiple is high because the market loves it.
2. The buyer's underwriting assumes the current EBITDA is a fair representation of run-rate earnings. No haircut for cycle.
3. There is no margin in the exit assumption. The base-case exit multiple is at or near the entry multiple. There is no "we'll get rerated higher" thesis available.

Hilton checked all three boxes in 2007. Hospitality multiples were at the top of the decade, the $1.2B EBITDA was a peak-cycle number, and the exit thesis depended on operating improvement and debt paydown — not multiple expansion.

The arithmetic that scares people away

To see why paying peak is conventionally fatal, hold every other variable constant and walk through what a 3-turn multiple contraction does to the equity.

Take a $1B EBITDA business bought at 18× ($18B EV) with 7× leverage ($7B equity, $11B debt). Hold five years. Assume EBITDA grows 5% per year (so it hits $1.28B at exit), and the company pays down 30% of the debt with cumulative FCF.

Every turn of exit-multiple contraction costs roughly 2 percentage points of IRR. Sponsors who underwrite to a 20% IRR at entry multiple = exit multiple have effectively zero margin against the most common bad outcome in a peak-cycle buy: rerating down to the long-run mean. That's the discipline argument, and it's right most of the time. If you don't have a clear answer to "what happens at a 4-turn compression," you don't have a deal — you have a hope.

Multiple expansion in reverse is also the cleanest part of the LBO returns decomposition. Operating improvement is a forecast. Debt paydown depends on cash flow. Multiple expansion (or contraction) is a single number, applied at exit, that swings equity around more than anything else in the model.

How Hilton broke the rule

Blackstone closed Hilton at 21.7× — higher than the 18× in the table above, with 8.4× leverage, into the worst hospitality downturn since the early 1990s. The exit multiple at the December 2013 IPO came in at roughly 24× LTM EBITDA on the $32.0B IPO EV. The multiple didn't compress at all. It expanded slightly.

That outcome wasn't luck. It was the product of four structural advantages that converted a peak-multiple buy from a hope into a survivable position.

1. Real estate-collateralized debt. Hilton's balance sheet was loaded with mortgageable hotel assets. When Blackstone restructured roughly $4B of debt in 2010, the lender syndicate had something tangible to anchor the workout against. A peak-multiple LBO of a pure cash-flow business in 2009 didn't get that conversation — the lender's only recourse was the equity holder's willingness to keep going.

2. Sponsor balance sheet capable of injecting fresh equity. Blackstone could write a follow-on equity check at the 2010 trough. Most sponsors can't. The LPA prohibits material follow-ons, and even funds that can structurally support them are reluctant to mark the position to the new lower basis. Peak-multiple deals where the sponsor is locked out of defending the position at the bottom don't survive a multiple contraction; they wipe out.

3. Cyclical, not structural, EBITDA decline. Hospitality EBITDA in 2009 was depressed by the cycle, not by a permanent shift in the business. Once travel recovered, RevPAR — the industry's primary revenue metric — re-accelerated, and EBITDA grew through it. Compare that to a peak-multiple LBO of a retail business in 2007: the EBITDA decline that started in 2008 was partly cyclical, partly the structural shift to e-commerce. You can't wait out a structural decline. You can wait out a cyclical one if the cap structure survives the wait.

4. Long enough hold to outlast the cycle. Blackstone held Hilton on a 6.2-year weighted-average basis — well past the typical 4-5 year PE hold. A 4-year hold to a 2011 exit would have produced a sub-1× MOIC, because EBITDA still hadn't recovered. The extra two-plus years of patience let the cycle catch up to the underwriting. Hold-period flexibility is a structural advantage; funds late in their investment period don't have it.

These conditions are the framework. A peak-multiple buy without all four is the version that doesn't end well.

The Hilton scorecard

The full deal:

The headline number is the $14B back on $5.6B in — roughly $8B of fund profit, the largest single-position dollar return in PE history at the time. But the more useful number for an associate trying to learn from this is the 18.7% IRR. Most mid-market funds underwrite to 20%+. Hilton, the "greatest deal of all time," cleared the bar by a hair on an annualized basis. The headline came from the absolute dollars, not from the rate.

That's the trade you make when you pay peak: even if everything works, the IRR will look mediocre, because the entry valuation took the easy alpha off the table. If you also need MOIC and IRR to print at the same time, peak multiples don't get you there.

The underwriting test you should actually run

The practical framework for screening a deal where you suspect you're at or near the peak:

1. Underwrite the base case at entry multiple = exit multiple, plus a downside case at 4 turns of compression. If the 4-turn downside still produces a 1.5× MOIC over a 6-year hold, you have a deal. If it produces a sub-1× MOIC, you have a hope. The compression case is the one your IC should grill, not the base case.

2. Confirm the EBITDA decline is cyclical, not structural. If you can't articulate the recovery thesis in two sentences — "demand falls 20% in a recession, but the secular driver of X means it recovers by year 4" — the decline probably isn't cyclical. Structural EBITDA decline never gets bailed out by hold-period extension.

3. Stress-test the debt capacity at a 25-30% EBITDA decline. Most peak-cycle LBOs don't fail at the entry; they fail when EBITDA falls 25% and the coverage covenant trips. The free debt capacity calculator lets you set leverage and run a downside EBITDA scenario in 30 seconds — at 8× leverage on cyclically peaked EBITDA, the binding constraint is almost always the interest coverage ratio, not the leverage covenant.

4. Verify you have a hold-period extension option. Funds in years 4-5 of their investment period don't have flexibility to hold a deal eight years. If your fund vintage doesn't permit a long hold, a peak-multiple buy that needs cycle recovery to clear is a structural mismatch even if the deal underwrites.

5. Score your sponsor's ability to inject follow-on equity at the trough. If your LPA caps follow-ons at 10% of original investment and the trough requires 30% additional capital to restructure, the math doesn't work no matter what the operating thesis says.

If you can answer all five, you're at the high end of the discipline curve and a peak-multiple buy is defensible. If you can't, you're underwriting the deal Blackstone ran, without the structural advantages Blackstone had.

What this looks like in a real model

The arithmetic of multiple compression vs. operating improvement vs. debt paydown is exactly what an LBO sensitivity table is built to surface. Our All-in-One PE Model runs dual 5×5 sensitivity tables across the entry-multiple/exit-multiple grid, with an embedded downside-EBITDA stress on the same axes — so you can see, at a peak entry, what hold period and operating trajectory you need to clear a hurdle rate under a 4-turn compression. The LBO sheet also handles multi-stage exits and debt restructuring toggles, which is what you need to model a Hilton-style deal honestly. You can see the LBO sheet preview before buying.

For a faster screen — entry multiple, leverage, hold length, exit assumptions — the free LBO returns calculator surfaces the same dynamic in 30 seconds. Try 21× entry / 21× exit at 8× leverage and a 6-year hold, then drop the exit multiple to 17× and watch the IRR move. The Hilton trade lives in the gap between those two lines, and so does the discipline.

The Hilton deal isn't an argument for paying peak. It's an argument that peak-multiple buys can work when four specific structural conditions are present — and that most sponsors most of the time don't have all four. Pay peak only when you can name them.

---

Sources:

1. Hilton S-1 Filing (Dec 2013), SEC EDGAR — entry transaction detail and capital structure
2. Bloomberg, "How Blackstone Made One of the Greatest Deals in PE History," May 2018, Bloomberg.com — total proceeds and exit sequence
3. Blackstone Q4 2018 Earnings Release, IR.Blackstone.com — final realization and IRR