The Dunkin' Brands LBO: How Bain, Carlyle and Thomas H. Lee Pulled a $500M Dividend Out of a Franchise Royalty Stream Seven Months Before the IPO

Most of the peak-cycle LBOs on this site have the same arc. Sponsors write a big equity check into a company with real operating risk — hotel rooms to fill, stores to staff, pipelines to run, distress to work through — leverage it to the gills, and then spend the next decade either riding out a cycle (Hilton, Dollar General, Freescale) or watching the equity burn (Toys R Us, TXU, Harrah's, Chrysler, Envision). The structural question in every one of those deals is the same: whose balance sheet absorbs the operating risk when demand goes sideways?

The Dunkin' Brands buyout is the deal that answered that question differently. In March 2006, Bain Capital Partners, The Carlyle Group and Thomas H. Lee Partners paid Pernod Ricard $2.425 billion in cash for Dunkin' Brands — Dunkin' Donuts, Baskin-Robbins and Togo's. Nearly every one of the ~9,000 Dunkin' Donuts locations and ~5,600 Baskin-Robbins locations worldwide was franchised. The sponsors did not buy stores. They bought the royalty stream the stores paid upstream.

That one structural feature — a 100% franchised model where the operating risk sits on thousands of independent franchisee balance sheets rather than on the parent — is what made the rest of the deal work. It is what let JPMorgan and Lehman Brothers package the whole buyout into a $1.5 billion securitization of the royalty receivables themselves, rather than a conventional LBO bank-plus-bond stack. It is what let the sponsors pull $500 million of cash out in a pre-IPO dividend recap seven months before the July 2011 IPO. And it is what let Carlyle fully exit by August 2012 — 6.5 years after close — into one of the stronger realizations of any 2005-2007 vintage mega-LBO.

What the sponsors actually bought

Pernod Ricard did not want Dunkin' Brands. The French spirits group had inherited it eight months earlier through its 2005 takeover of Allied Domecq — a British spirits-and-quick-service company that had owned Dunkin' Donuts and Baskin-Robbins for decades and bolted on Togo's in 1997. Pernod's strategic case was to be a focused global spirits company. A doughnut franchise with three sub-brands and nearly 15,000 locations across 50+ countries was off-strategy from day one.

The deal was announced December 12, 2005 and closed March 1, 2006. The buyers were a three-equal consortium: Bain Capital Partners, The Carlyle Group and Thomas H. Lee Partners. Each firm took three board seats. Pernod Ricard applied approximately $1.7 billion in net proceeds after tax to paying down debt from the Allied Domecq transaction.

The buyout capital structure was unusual, and the unusual parts are where most of the lessons live.

Three features of this stack deserve a closer read.

First, the equity split was unusually symmetric. Three sponsors, each writing $345.5 million. No lead. No co-lead with a minority partner. No sub-$100M fourth sponsor rounding out the stack. Each firm took three of the nine board seats. The governance model was explicitly a three-way equal partnership — closer in shape to the seven-sponsor club deal that took SunGard private than to the two-sponsor-plus-management governance at Harrah's or the solo-sponsor model at Burger King. Three equal votes work on a franchised cash-flow business, where most strategic decisions are about capital allocation and franchise system policy; they would have worked less well on a turnaround where someone needed to break ties on operating calls every week.

Second, the headline leverage ratio looked peak-cycle, but the underlying asset was not. Breakingviews in December 2005 flagged the deal as being underwritten at approximately 8.5x EBITDA — well above the mid-cycle 6-7x range and in line with the leverage ratios that broke Toys R Us and TXU. But those deals carried 8x leverage on EBITDA that was cyclical, operationally levered and sensitive to end-market demand. Dunkin's EBITDA was a franchise royalty stream, overwhelmingly fee-based and contractually bound to franchisee gross sales. That is a fundamentally different debt-capacity question. To see what the math looks like when you move the same leverage ratio across businesses with different cash-conversion characteristics, our free debt capacity calculator is a quick way to pressure-test what DSCR and interest coverage actually hold up at 8.5x.

Third, the permanent financing was a whole-business securitization — not a traditional LBO debt stack. JPMorgan and Lehman Brothers initially bridged the deal with an $850 million interim credit facility priced at LIBOR plus 2.50%. They then refinanced into approximately $1.5 billion of asset-backed notes collateralized directly by the franchise royalty receivables. The notes were secured by the right to receive royalty payments from Dunkin' Donuts, Baskin-Robbins and Togo's franchisees — not by the enterprise cash flow of the parent, and not by the parent's unsecured credit. Economically, this is closer to the music-royalty securitizations that David Bowie pioneered in 1997 than to a typical leveraged buyout bond. Investors looked through the holding company to the royalty stream itself.

Why whole-business securitization matters for a franchise LBO

The financing choice sounds technical. It is not. It is the single most important feature of the deal from a sponsor return perspective.

A conventional LBO debt stack — senior secured term loan plus revolver plus unsecured high-yield notes — prices off the enterprise credit of the company: its EBITDA, its coverage ratios, its ability to carry debt through a cycle. The lender is underwriting the business as a going concern, with claims on the enterprise. Pricing reflects that.

A whole-business securitization prices off the collateral itself. The lender is not primarily underwriting Dunkin' Brands as an operating company. The lender is underwriting the cash flow from franchise royalty receivables — contractually specified at roughly 4-6% of franchisee gross sales, paid by thousands of independent franchisee operators, each of them a separate credit — and lending against that cash flow as a dedicated collateral pool. Default-remote structuring, bankruptcy-remote SPVs, and triggered reserve mechanisms mean the collateral is isolated from the operating company's credit even if the operating company itself gets into trouble. Pricing reflects that.

The pricing gap between the two approaches is substantial. For a franchise business dominated by recurring royalty revenue — Dunkin' Brands reported in its 2011 S-1 that approximately 62% of FY2010 revenue came from royalty income and franchise fees, with another 16% from franchisee rental income — the whole-business securitization route is both cheaper and more flexible. The sponsor gets more leverage at a lower blended cost of debt, with covenants that depend on the collateral pool rather than on operating-company financial ratios that could come under pressure in a recession.

This mattered acutely in 2008-2009. When the financial crisis arrived, the Harrah's, TXU, Chrysler, Clear Channel and Freescale capital structures — all conventional bank-plus-bond stacks — faced covenant pressure, forced debt exchanges, or outright default, because the parent-level EBITDA was deteriorating. The Dunkin' stack did not. Franchisee comp-sales declined, but royalty receipts held up far better than enterprise EBITDA at operating-heavy peers, because each franchisee was absorbing the margin compression at the store level while paying the same percentage royalty on whatever sales they generated.

Our companion piece on sources and uses in an LBO walks through the full capital structure mechanics; what matters here is that Dunkin's financing route fit the collateral. That is rarer than it should be in mega-LBO history.

The hold: franchise expansion, international scale, and operating leverage

Between March 2006 and the July 2011 IPO, Dunkin' did what franchise businesses are supposed to do under PE ownership: grow units with other people's capital, discipline the cost structure, and let the royalty stream compound.

The 2011 S-1 gives a partial picture of the expansion. Dunkin' Donuts global points of distribution grew from 8,835 at the end of FY2008 to 9,760 at the end of FY2010 — roughly 10% unit growth over two years, almost entirely franchisee-financed. Baskin-Robbins' international footprint, particularly in Asia, expanded materially. The Togo's sandwich chain, a strategic misfit for a coffee-and-doughnut franchisor, was divested in 2007 to focus the portfolio.

Carlyle's own marketing case study — promotional and self-reported, but directionally useful — describes LTM EBITDA growing approximately 71% during the firm's ownership and EBITDA margins expanding approximately 866 basis points. A good share of that margin expansion came from the mechanical operating leverage of a franchise business: if your revenue is a 5% royalty on franchisee sales and your incremental cost of supporting another unit is close to zero, then each new franchised unit drops almost its entire royalty contribution into operating income.

The combination — moderate unit growth, strong unit economics, and a securitized capital structure that was not under covenant pressure through 2008-2009 — is what let the sponsors do what they did next.

The December 2010 dividend recap

On December 3, 2010, Dunkin' Brands paid a $500.0 million cash dividend on its outstanding Class L common stock. The S-1/A disclosed the transaction to IPO investors in 2011. Fortune at the time characterized the recap as debt-funded — new leverage layered onto the company's balance sheet to fund the distribution, with the sponsors as the recipients.

The math on this is worth staring at. The sponsors' combined equity check was approximately $1.036 billion. A $500 million dividend recap returned roughly 48% of that in cash — seven months before the IPO even priced. On a weighted-average basis, the sponsors had recovered close to half of their original equity in cash by December 2010, before any public market exit, and before any value crystallized at all. In MOIC vs IRR terms, this is the mechanic — pull cash early to compound the return on the residual stake — that separates good PE sponsors from average ones.

The recap is also structurally different from the KKR and Bain dividend recaps at Toys R Us. At Toys R Us, successive dividend recaps extracted capital out of a declining retail business whose debt service was eating the cash generation; the recaps functionally transferred value from LPs and creditors to sponsors and management fees, and the equity was ultimately wiped. At Dunkin', the recap came out of a franchise royalty stream that was intact, cash-generative, and about to be re-rated by public markets. One is value extraction out of a stressed balance sheet; the other is a reasonable mid-hold distribution funded by a company that could carry additional leverage without impairing its going-concern value.

The difference is not the mechanic. It is the quality of the underlying cash flow.

The July 2011 IPO

Dunkin' Brands filed its S-1 on May 4, 2011 with the Securities and Exchange Commission. The initial filing contained a placeholder maximum offering amount of $400 million. By early July, the company had filed an estimated price range of $16.00 to $18.00 per share. On July 27, 2011, after the roadshow, the deal priced at $19.00 per share — above the top of the filed range.

The shares opened at $24.97, traded as high as approximately $29 intraday, and closed their first session at $27.85 — a 46.6% gain versus the $19.00 pricing. Bloomberg reported the open and close in real time; Renaissance Capital subsequently listed the first-day return at +46.6%. By any 2011 public-offering standard, this was a strong debut.

Two features of the IPO are worth noting from a sponsor-exit perspective.

First, the IPO proceeds did not flow to the sponsors. The 22.25 million shares sold in the offering were newly issued shares sold by the company. The ~$422.8 million in gross proceeds — approximately $348 million in projected net proceeds at the earlier S-1/A midpoint, before the pricing moved higher — went to Dunkin' Brands itself, which used it (together with additional term loan borrowings) to retire the 9 5/8% senior notes due 2018. In other words, the IPO was primarily a recapitalization transaction that strengthened Dunkin's balance sheet. The sponsors' return would come from subsequent secondary offerings against their retained ownership, not from this first transaction.

Second, after the IPO, each of the three sponsors still held approximately 26.1% of the equity — three roughly-equal controlling blocks adding up to approximately 78% of the pro forma public float. That retained position is what would eventually produce the sponsor return.

The sponsor sell-down: four offerings over 13 months

The pattern from here is a standard post-IPO PE sell-down, executed at compressed speed and rising prices.

The November 2011 secondary priced at $25.62, approximately 35% above the IPO price three months earlier. The April 2012 secondary priced at $29.50, approximately 55% above the IPO price. The August 2012 transaction was structurally different — the company repurchased 15 million of the roughly 21 million offered shares from the selling stockholders, with Citigroup as sole underwriter — but it realized another approximately $652.6 million in gross proceeds to the sellers.

Carlyle's own case study states that the firm fully exited the investment in August 2012 — at the third secondary. From the firm's perspective, the hold was approximately 6.5 years from the March 2006 close.

Bain Capital and Thomas H. Lee Partners retained smaller residual stakes that were sold down over 2013 and into 2014, though the SEC filings and trade press do not cleanly document each firm's final exit date. The aggregate sponsor realization through August 2012 — the December 2010 recap dividend plus the three secondaries — totaled approximately $2.58 billion against an original combined equity check of approximately $1.04 billion. Before counting any 2013-2014 Bain or THL sell-downs, that is roughly 2.5x cash multiple on the original equity, over a 6.5-year hold.

In MOIC and IRR terms, 2.5x over 6.5 years maps to roughly a 15% gross IRR on undiscounted cash flows — modest-looking in isolation, but the actual distribution-weighted IRR is higher because the $500M dividend arrived in Year 4.8 and the bulk of the secondaries cleared in Years 5.7 to 6.4. Proceeds concentrated in the back half of the hold at compounding prices produce a meaningfully better cash-on-cash IRR than a flat distribution across the hold. The LBO returns calculator linked below walks through the exact mechanic.

Why the deal worked

Three things, in decreasing order of importance.

First, the franchise model. Nearly 100% franchised, with the operating risk pushed down to independent franchisee balance sheets, is the single most important feature of the deal. It is what let the sponsors carry 8.5x leverage on a securitized royalty pool through a recession without covenant trouble. It is what let margins expand through the hold on the operating leverage of a fee-based model with near-zero incremental cost per new unit. And it is what distinguishes the Dunkin' return profile from the operating-heavy Toys R Us, J. Crew and Neiman Marcus outcomes over the same cycle. Franchise model is not a tactic. It is a structural fact about the business that makes peak-cycle leverage survivable.

Second, the whole-business securitization. The sponsors financed the deal against the royalty collateral pool itself, not against the operating company's enterprise credit. This gave them more leverage at lower cost, with covenants that did not tighten when the underlying operating company came under stress. The 2008-2009 performance test that broke so many 2005-2008 vintage capital structures did not break this one, because the capital structure was not underwritten to parent-level EBITDA. For any PE sponsor evaluating a royalty-dominant target — restaurants, music catalogs, pharmaceutical royalties, consumer franchise networks — the whole-business securitization route is worth understanding before falling back on a conventional LBO debt stack.

Third, the governance fit. Three equal sponsors with three equal board blocks worked for Dunkin' because the strategic agenda was predominantly capital allocation and franchise-system policy, not operating turnaround. The sponsors did not need to move quickly on store-level operating decisions. They needed to agree on capital structure, IPO timing, and dividend distributions. On those questions, three equal votes were workable. On the day-to-day operating calls that a turnaround would have required, three equal votes would have been a coordination disaster. The governance form has to match the operating challenge, not the other way around.

The counterfactual is informative. If the same consortium had taken a 2006 specialty retail LBO — say, picked up the equivalent of Michaels or J. Crew at the same leverage — and run the same three-equal governance over a six-year hold, the operating demands of running a specialty retail business through 2008-2009 would almost certainly have exceeded what three equal sponsor votes could coordinate. The Dunkin' structure fit the Dunkin' business. It would not have fit everything.

What to take from this

Two things.

First, structural features of a business matter more than entry multiples when it comes to what capital structure survives a cycle. A 100% franchised, royalty-dominant business can carry 8.5x leverage through 2008-2009 because the operating risk sits on thousands of franchisee balance sheets, not the parent. A company-operated retail business at the same 8.5x leverage cannot. The industry-adjusted leverage question — not the absolute leverage ratio — is what separates survivors from write-offs in the 2005-2007 vintage.

Second, financing-structure innovation (the whole-business securitization in particular) is one of the few PE technical advantages that is genuinely scarce and durable. Most of what PE does operationally — refranchising, zero-based budgeting, add-on M&A, CEO replacement — can be done by strategics and public companies with roughly similar competence. Financing-structure innovation requires a team that understands collateralized debt markets, rating agency criteria and bankruptcy-remote SPV structuring in combination with corporate M&A. That skill set is rare. The Dunkin' deal is one of the places where it clearly earned its fees.

Model this yourself

If you want to build a fully-populated LBO model — with purchase price waterfall, whole-business securitization versus conventional bank-plus-bond stack comparison, franchise-royalty revenue build, mid-hold dividend recap mechanics, and IPO-plus-secondary-offering exit sequencing — that lets you replicate the Dunkin' structure and stress-test how much of the return comes from the franchise model versus the financing versus the dividend recap timing, our the All-in-One PE Model is designed for exactly that. The 26 sheets include a dedicated LBO sheet with dual 5x5 MOIC and IRR sensitivity, a sources-and-uses waterfall with recap mechanics, and franchise-royalty revenue build worksheets. See the sheet preview before buying.

---

Sources:

1. Dunkin' Brands Group, Inc. — Form S-1 (May 4, 2011), SEC EDGAR
2. Dunkin' Brands Group, Inc. — Form S-1/A (July 2011), SEC EDGAR
3. Dunkin' Brands Group, Inc. — Form 424B4 Final IPO Prospectus (July 2011), SEC EDGAR
4. Dunkin' Brands Group, Inc. — Form 10-Q Q2 FY2011, SEC EDGAR
5. Dunkin' Brands Group, Inc. — Form 10-Q Q2 FY2012, SEC EDGAR
6. Dunkin' Brands Group, Inc. — Form 424B4 April 2012 Secondary Prospectus, SEC EDGAR
7. Dunkin' Brands Group, Inc. — Announces Pricing of Initial Public Offering (press release, July 27 2011), PR Newswire
8. "France: US$2.425bn sale of Dunkin' Brands completed," just-food, just-food.com
9. "Dunkin' Donuts Brand Acquired for $2.43 Billion," Chief Marketer, chiefmarketer.com
10. "Dunkin' Jumps 47% in First Day After $422.8 Million IPO," Bloomberg, July 27, 2011, Bloomberg
11. "LBOs: JPM and Lehman Brothers Dunkin' into royalty pool to fund buyout," Euromoney, euromoney.com
12. "Throwing iced coffee on the Dunkin' IPO," Fortune, July 26, 2011, fortune.com
13. "Simpson Thacher Represents Underwriters in Follow-On Offering by Dunkin' Brands Group, Inc." (April 9, 2012), stblaw.com
14. "Simpson Thacher Represents Underwriter in Follow-On Offering by Dunkin' Brands Group, Inc." (August 20, 2012), stblaw.com
15. The Carlyle Group — Dunkin' Brands Case Study (PDF), carlyle.com