The mattress Roosevelt slept on

Posted on: 22 April 2026

The Simmons Bedding Company was founded in 1870 in Kenosha, Wisconsin, by Zalmon G. Simmons, who originally manufactured telegraph insulators and cheese boxes. He drifted into mattress-making almost by accident, receiving as settlement of a debt the patent for a woven wire bedspring. On the back of that patent he built a company that for over a century was synonymous with quality in American bedding. Eleanor Roosevelt praised the virtues of the Beautyrest. Presidents slept on Simmons mattresses aboard Air Force One, and the Lincoln Bedroom at the White House was fitted out with them. Cole Porter referenced the brand in "Anything Goes" on Broadway. Henry Ford appeared in its advertisements, as did H.G. Wells. This was an industrial asset, certainly, but it was also a piece of American material culture, the sort of brand that passes from one generation to the next the way one passes down well-built household appliances.

In 2009 the Simmons Bedding Company went bankrupt. In 2023 it went bankrupt again, trading under the name Serta Simmons Bedding after the merger with the competitor that took place in 2012 under a new owner. In the preceding two decades it had been sold seven times in just over twenty years, each time to a different private equity firm, each time at a higher price, each time with more debt loaded onto the balance sheet of the operating company. In 1991, when the music started, Simmons carried 164 million dollars of debt. By 2009, at the first bankruptcy, it was carrying 1.3 billion. The New York Times headlined the article which has since become the case study taught in every American business school "Buyout Firms Profited as a Company's Debt Soared". The various private equity owners had between them extracted roughly 750 million dollars in profits, while the bondholders of the final round lost 575 million and the employees watched their pension plans collapse, because the previous owner, as part of an employee stock ownership plan, had stopped contributing to the pension fund on the assumption that the share value would be sufficient to cover it. Ça va sans dire, it was not.

Robert Hellyer, Simmons's former president, who had worked for several of the private equity owners before being asked to leave the company in 2005, put it disarmingly to the Times. "From my experience, none of the private equity firms were building a brand for the future. Plus, the mind-set was, since the money was practically free, why not leverage the company to the maximum?" Hellyer was speaking from inside the room, not from academic distance. He was one of those who had seen the mechanism from within, and the mechanism is what deserves to be described with surgical precision.

A leveraged buyout, or LBO, works as follows. The private equity fund decides to acquire the target company putting up a relatively modest equity contribution, typically between ten and twenty per cent of the purchase price. The remainder, eighty per cent or more, is borrowed. The critical point, the one most non-financiers fail to absorb the first time they hear it, is that the debt is not loaded onto the fund's balance sheet, it is loaded onto the balance sheet of the acquired company. The operating company, the one that manufactures mattresses or shoes or runs restaurants or publishes magazines, finds itself the day after closing with a mass of debt it did not previously carry, and from that moment onward has to service that debt out of its own cash flows. The interest is tax-deductible, which makes the operation even more efficient for the fund. Technically elegant, structurally brutal.

When the fund then resells the company, which it typically does after five to seven years, the new acquirer, often another PE firm, repeats the operation, layering fresh debt to finance the higher price of the next transaction. The original company never sees a penny of this capital gain, which goes entirely to the exiting fund. But it inherits the additional layer of debt to service. After three or four changes of ownership, the operating company has become a machine whose primary purpose is to pay interest to bondholders who have never set foot in it, rather than to reinvest in its product, its research, its workforce, its community. Simmons was bought and sold seven times. At some point the model stopped working because it could not continue working, and the company collapsed.

What the Simmons case demonstrates with almost pedagogic clarity is that the collapse had nothing to do with the quality of the underlying asset. Simmons was operationally sound, the Beautyrest was a serious product, the brand still carried significant market value, distribution functioned. What had become unsustainable was the financial structure superimposed onto the asset, which had turned a solid manufacturing business into a pure rent-extraction vehicle.

Closer to home, Debenhams offers the British reader a parallel almost uncomfortably exact. The department store chain, founded in 1778 and for two centuries a fixture of British high street life, was taken private in 2003 by a consortium of CVC Capital Partners, Texas Pacific Group and Merrill Lynch Global Private Equity. In the three years of their ownership the consortium extracted over one billion pounds in dividends while selling off the freeholds of most Debenhams stores and leasing them back, thereby transforming fixed property assets into long-term rental liabilities. When the consortium refloated Debenhams in 2006, the company returned to the stock market with a debt structure and a property portfolio that would prove fatal. Twelve years of struggle followed, two administrations, the final liquidation in 2021, and thousands of redundancies across the country. The consortium members had long since pocketed their gains. The employees, the pensioners and the high streets left without their anchor tenant bore the cost.

Here, however, lies a dimension of the phenomenon that mainstream economic literature systematically overlooks, and it is the point that deserves particular attention. When a fund arrives with an offer for a founder, or for the heirs of a founder, or for an historic owning family, the offer is almost always generous. Multiples are above comparables, cash at signing is substantial, the retention package for existing management is attractive, and it is often accompanied by a multi-year earn-out which, should the company continue to perform, can roughly double the total value as the seller perceives it. The founder reads the term sheet, looks at the number, runs the calculations and almost inevitably signs. In front of him, blindingly, sits the economic realisation of a lifetime of work. That he should be dazzled is entirely normal. That he should not be dazzled is almost impossible.

The problem is that what he is signing, beyond the equity transfer, is a series of clauses whose significance only becomes apparent later, when nothing can be done about them. The earn-out is structured around short-term operational metrics, typically EBITDA for the first two or three years, which means that the founder, if he wants to collect the second tranche of the price, must chase the very quarterly numbers he had spent decades as an owner avoiding. The post-acquisition board is controlled by the fund's partners, who hold the majority of votes on strategic decisions. The founder, even if he formally retains the CEO title or becomes honorary chairman, no longer decides the things that matter. He signs the lock-up, which keeps him tied to the company for three to five years, during which he cannot leave without forfeiting the residual earn-out. He signs the non-compete, which prevents him from rebuilding a similar business for two or three years beyond the term of his engagement. He signs the claw-back provisions, which allow the fund to reach back into the purchase price if issues emerge that were not flagged in due diligence.

Within a couple of years the founder finds himself in a position no term sheet has described honestly. He is still inside the company, formally holding a senior role, but the decisions that matter are not his. He watches the new owners cut R&D budgets, consolidate longstanding suppliers, renegotiate distributor contracts downward, sack veteran sales executives and replace them with younger, cheaper, more compliant figures, reduce after-sales service, multiply retail locations in places that erode the positioning. To the fund's accounts, all this is operational optimisation, virtuous and documentable. To the founder, it is the methodical dismantling of what he built. But the founder is contractually bound, and if he rebels openly he loses the earn-out, breaches the non-disparagement clause, and probably finds himself in litigation with people for whom that kind of litigation is the day job. So he stays, he signs off communications he would never have written, he approves decisions he does not endorse, and he watches in silence.

The psychological dimension of this dynamic is almost never told, for an obvious reason: founders who emerge damaged have no incentive to speak publicly about it. To admit that one sold badly, or that one ceded too much control, or that one overestimated one's capacity to influence post-acquisition outcomes, is to compromise the residual reputation and, frequently, the ability to secure new senior advisory or board positions. So founders who have watched their brand be consumed by the new owner tend to retreat into silence, write guarded memoirs, give interviews that dwell on the glorious past and glide over the recent decline. The result is that the available literature on this phenomenon is dramatically understated relative to its actual prevalence.

The Simmons case also shows something else, which is perhaps the hardest thing to accept for anyone viewing the phenomenon through a moralising lens. The individual PE partners involved were not evil, not incompetent, not criminal. They were doing precisely what their limited partners, who are pension funds, university endowments and family office endowments, pay them to do. A typical closed-end PE fund has a ten-year life, with a four-to-five-year investment period and a five-year exit window. Limited partners expect a net IRR, the internal rate of return, typically above fifteen per cent per annum. To deliver that sort of return the fund has to extract value from its portfolio companies and return it as cash within the life of the vehicle, or the next fund-raise becomes impossible. It is not a moral choice, it is a structural constraint imposed by the legal form of the vehicle itself. Asking a closed-end PE fund to behave like Pictet or like a third-generation family office is like asking a watchmaker to work at the speed of an assembly line. Optimised for different things.

What the mechanism cannot handle, and here we return to the starting point, is the communitarian and relational capital of certain assets. A company like Simmons was not merely factory and brand, it was a network of independent retailers fostered over three generations, a long-standing relationship with hotel chains that specified Beautyrest in their procurement manuals, a reputation among craftsmen in the bedding trade that distinguished it from cheaper alternatives. All of this is capital that appears on no balance sheet but that generates the cash flows which make the asset attractive to PE in the first place. The paradox is that the PE buys the company precisely because of that capital, and then, in order to service the debt, begins systematically to consume it. The independent retailers are consolidated into centralised contracts with thinner margins, and many simply stop pushing the brand. The hotels renegotiate their contracts on terms that shift volume but not value. The craftsmen notice the gradual decline in quality and recommend alternatives. The end consumer, eventually, finds on the shop floor a product bearing the Simmons name but no longer carrying what that name used to stand for. By that point the reputational capital has been withdrawn entirely, and what remains is an empty brand sitting on a debt-laden balance sheet.

Toys R Us under Bain Capital, KKR and Vornado followed exactly this pattern until the 2017 bankruptcy, leaving generations of American and British families, who had made the shop an institution of childhood, suddenly without a reference point. Red Lobster under Golden Gate Capital, with the parallel ownership of Thai Union which supplied shrimp to the restaurant chain, imposed the "endless shrimp" promotion at a structurally negative margin to generate the volumes that drove sales at the related party, until bankruptcy in 2024. Payless ShoeSource under Blum Capital and Golden Gate, same pattern, bankrupt in 2017 and again in 2019. Every time the mechanism is identical, every time the community of long-standing customers finds itself holding a product no longer resembling what it remembered, every time the fund extracts value and moves on. This is not serial bad luck, it is the systematic execution of a fully documented playbook that works precisely as it should work given its structural premises.

The counterfactual exists here too, for those who want to see it. Tata bought Jaguar Land Rover after the Ford disaster and, despite real difficulties, has invested in research and product in a way no closed-end PE fund could structurally have afforded to do. Berkshire Hathaway holds its portfolio companies for decades without ever selling, and the managers of Berkshire subsidiaries know that their horizon is not three years but potentially thirty. Long-standing family holdings, whether Italian, German or Swiss, when they still exist, operate on the same kind of horizon and indeed tend not to sell to PE if they can avoid it. The difference is not moral, it is structural, different vehicle, different incentives, different outcomes.

What the founder ought to have in mind when the generous term sheet arrives, therefore, is not whether the figure on paper is right. It is whether the buyer is the right type of buyer for the right type of asset. A commodity steel mill may be a reasonable LBO target, flows are predictable, debt is serviceable, brand does not matter. A 150-year-old company with a loyal community, a recognisable cultural identity, a relational web built over generations, is a wrong LBO target by definition, because the acquiring vehicle cannot structurally realise the very thing that makes the company valuable. Selling that sort of asset to a closed-end PE fund is not a financial transaction, it is a transfer of custodianship that cannot go well, for the simple reason that the successor custodian is optimised to do the opposite of custodianship.

This the founder should know before he signs, not after. And if he really must sell, he should at least read carefully who he has across the table from him, and who will be sitting on his board for the next five years. Because the millions in the bank, when they arrive, are real. But what one sees in the mirror in the morning, after three years of being a silent witness to the dismantling of one's own life's work, is quite another matter. And for some things, the bank balance does not compensate enough.

Caveat emptor, the old lawyers used to say. Buyer beware. It was a principle of Roman law before it was a principle of corporate finance. And it described with some exactness a thing the M&A manuals rarely say: that there exist economically generous offers whose real cost reveals itself only after the signature, when signing was the last decision one could still make autonomously.

Reference: Julie Creswell, "Buyout Firms Profited as a Company's Debt Soared", The New York Times, 5 October 2009. Full version also accessible via cnbc.com.