Pitch decks make every LBO look the same: a sources & uses table, a pro-forma EBITDA build, a returns waterfall. The framework is identical whether the target rents hotel rooms or sells toys. So when Blackstone and KKR went out hunting consumer assets in the mid-2000s, the model template they used was effectively the same one.
The outcomes were not. Blackstone's Hilton deal returned $14.0B in proceeds at a 2.5x MOIC. KKR + Bain + Vornado's Toys R Us deal returned zero — full equity wipeout for LPs, 33,000 layoffs at liquidation, and a multi-year national debate about whether PE should be allowed to own retail at all.
The standard explanation is leverage. But Hilton was financed at 8.4x EBITDA — meaningfully more than Toys at 5.7x. If raw leverage killed the company, it should have killed the hotel deal first. It didn't. The difference is structural, and it's the most important screen any junior associate can run before underwriting a consumer LBO.
What a "retail LBO" actually is
Retail LBOs aren't a single category. They span hospitality (Hilton, Four Seasons), specialty retail (Toys R Us, Petco, Neiman Marcus), grocery, and quick-service. What unites them is that the underlying business sells to end consumers, so cash flow tracks discretionary spending more directly than B2B revenue does.
That isn't, on its own, a problem. Consumer demand is volatile but not toxic — plenty of consumer PE deals work. What kills these deals is the interaction between two things:
- The variability of EBITDA — how much it can drop in a bad year.
- The asset base supporting the debt — what creditors can attach to, and whether the sponsor can monetize it, if EBITDA breaks.
Hilton sat on one side of both dimensions. Toys R Us sat on the other. The capital structures looked similar on the surface; the businesses underneath did not.
How the Hilton math actually worked
Blackstone closed Hilton on October 24, 2007 — weeks before the financial crisis broke. The headline metrics:
| Entry | Value |
|---|---|
| Enterprise value | $26.0B |
| LTM EBITDA | $1.2B |
| Entry multiple | 21.7x |
| Leverage | 8.4x EBITDA |
| Sponsor equity | $5.6B |
| Cash at close | $0.3B |
That is an extreme entry multiple at the peak of a cycle. The deal looked dead inside 18 months. Hotel demand collapsed in 2008-09, and Blackstone had to restructure roughly $4B of debt in 2010 to avoid default.
But Hilton had something Toys R Us did not: an asset base that allowed restructuring without liquidation. Hotels have real-estate basis, brand value, and a franchise model that produces fee streams even when room nights fall. Blackstone separated the business along those lines — spinning out Park Hotels & Resorts (the real-estate vehicle) and Hilton Grand Vacations (the timeshare business) before the final IPO sell-down — turning a single distressed bet into three discrete pools that could be optimized separately.
| Exit | Value |
|---|---|
| IPO date | December 12, 2013 |
| EV at IPO | $32.0B |
| Total proceeds | $14.0B |
| MOIC | 2.5x |
| IRR | 18.7% |
| Hold | 6.2 years |
A 2.5x MOIC is solid but not heroic. The deal became famous because the dollar return was historic — the largest single-deal profit in PE history at the time. If you want a deeper read on why MOIC and IRR diverge on long holds like this one, our primer on MOIC vs IRR walks through the same Hilton numbers with the math behind them.
The point for the retail thesis isn't whether 18.7% IRR is a great return. It is whether the deal could have survived 2009 at all. The answer was yes — barely — because there was something to restructure against.
How the Toys R Us math actually broke
KKR + Bain Capital + Vornado closed Toys R Us on July 21, 2005, two years before Hilton. The deal looked far more conservative on paper:
| Entry | Value |
|---|---|
| Enterprise value | $6.6B |
| LTM EBITDA | $0.55B |
| Entry multiple | 12.0x |
| Leverage | 5.7x EBITDA |
| Sponsor equity | $1.3B |
| Cash at close | $0.0B |
Compare line by line to Hilton: lower multiple, lower leverage, lower absolute dollar size. By the metrics that show up first on a pitch deck, this was the safer deal.
The problem was what the leverage cost against what the business could pay. The $5.3B of debt produced roughly $400M per year of interest expense against $550M of LTM EBITDA. That left about $150M of coverage cushion in a flat year. In a bad year — and there were several once Amazon launched its toy warehouse two years before the deal closed — EBITDA fell toward the interest line and the cushion vanished:
| Base year | Down year | |
|---|---|---|
| EBITDA | $550M | $400M |
| Cash interest | $400M | $400M |
| Coverage cushion | $150M | $0M |
There was no free cash flow to redirect into an e-commerce transformation. The 5.7x leverage that read conservative in 2005 became a stranglehold by 2010. Toys filed for Chapter 11 in September 2017 and liquidated in June 2018. LP equity went to zero. The sponsors had separately collected roughly $470M in management fees and dividend recaps over the 12-year hold — a dynamic the dividend recaps explainer decomposes more fully — but the fund-level MOIC was 0.0x.
A free quick-check: plug 5.7x leverage and $550M EBITDA into the debt capacity calculator at the implied ~7.5% blended rate ($400M ÷ $5.3B). Coverage lands around 1.4x — fine in a steady state, instantly dangerous on any EBITDA decline. The sponsors had run that math. They did not stress it against Amazon.
The structural difference: what supports the debt
Strip away the multiples and look at what was under each capital structure.
Hilton's assets had standalone value. Branded hotel real estate in major cities is hard to replicate and retains value through a cycle. The franchise model produces fees that don't move 1-for-1 with room nights. When 2008 hit, Blackstone could restructure debt against real, valuable, separable assets — and later carve them into vehicles (Park, Grand Vacations) sized to different buyer pools. Lenders had something to attach to that wasn't shrinking.
Toys R Us's assets were trapped. The company did own significant store-level real estate — Vornado was brought into the consortium for exactly that reason. But bank covenants written at deal close prevented sales during the eventual distress, so the real-estate spinoff thesis (Propco) never delivered the value it was supposed to. Meanwhile the operating business itself was being structurally degraded by Amazon and Walmart category pricing. By 2017 enterprise value was a function of distressed lease assignments and inventory liquidations.
The two deals exposed the same line in a credit memo to wildly different outcomes:
| Hilton (2007) | Toys R Us (2005) | |
|---|---|---|
| Entry multiple | 21.7x | 12.0x |
| Leverage | 8.4x | 5.7x |
| Cash interest vs LTM EBITDA | manageable | $400M on $550M |
| Asset realization optionality | Yes (RE + brand spins) | Blocked (covenants) |
| Demand secular trend | Cyclical (travel recovers) | Declining (e-commerce) |
| Restructuring path | Amend & extend (2010) | Chapter 11 (2017) |
| MOIC outcome | 2.5x | 0.0x |
The headline ratios are misleading. The real underwriting question is not "what's my coverage ratio at close" but "what does my asset base look like if I have to negotiate with creditors in year 3."
What this means for an LBO underwriter
If you are modeling a consumer or retail deal, three checks that tighten the screen:
-
Run the leverage decision off downside EBITDA, not LTM. What's the coverage ratio if revenue drops 15% and gross margin compresses 200 bps? If that scenario breaches a covenant, the deal is already over — you just don't know it yet. The LBO returns calculator lets you stress this in under a minute by flexing exit-year EBITDA and watching the equity bridge collapse in real time.
-
Underwrite the asset base separately from the cash flow. Real estate, brand IP, customer data, and franchise fee streams are restructuring optionality. Operating leases without saleable property are not. Hilton survived 2009 because there was something to amend and extend against. Toys had property too — but covenants made it unsellable when it mattered, which is functionally the same as not having it.
-
Treat secular trend as a bright line. Hilton bet on a cyclical recovery. Toys bet against a secular decline. Cyclical bets recover; secular bets compound the wrong way. Underwriting an LBO into a secular decline is structurally different from underwriting through a cycle, and most LBO templates do not differentiate.
The sources & uses table looks the same on both deals. The capital structure does too. What's under the capital structure is the actual deal. Our sources & uses walkthrough covers the mechanical version of the table; this post is about the layer beneath it.
What this looks like in a real model
The All-in-One PE Model builds the EV-to-equity bridge, the debt schedule, and the returns waterfall as one connected sheet — and the sensitivity tables on the LBO tab let you flex EBITDA-to-cash conversion and exit-year EBITDA separately, which is exactly the variable that separated Hilton from Toys. See the sheet preview before buying.
Retail LBOs aren't structurally riskier than industrial or B2B deals. They're structurally different. The sponsors who understand that difference end up running Hilton playbooks. The ones who don't end up running Toys R Us postmortems.
Sources:
- Hilton S-1 Filing (Dec 2013), SEC EDGAR — entry capital structure, IPO valuation
- Bloomberg, "How Blackstone Made One of the Greatest Deals in PE History," May 2018, Bloomberg — total proceeds and hold-period detail
- Blackstone Q4 2018 Earnings Release, IR.Blackstone.com — final exit returns
- Toys R Us Chapter 11 Petition (S.D. NY 2017), Prime Clerk docket — original capital structure detail
- House Financial Services Committee Hearing (Feb 2019), "The Aftermath of the Toys R Us Bankruptcy," House Financial Services — interest expense and EBITDA figures
- Bloomberg, "How Vulture Capitalists Ate Toys R Us," March 2018, Bloomberg — sponsor fee history and bankruptcy context