Key takeaways
A strong record buys time, not proof. Lampert's reputation raised the bar for anyone questioning him as Sears declined.
Cutting investment can imitate skill. Buybacks and thin store spending lifted early numbers before the starvation became visible.
Related-party flows are readable in real time. Lampert stood on both sides of Sears's biggest deals as lender, landlord, and buyer.
The dull signals predict better than the story. Where cash goes and who benefits are far harder to spin than a turnaround thesis.
How a Buffett comparison bought a decade of patience
In November 2004, a business magazine put a hedge fund manager on its cover and asked whether he was the next Warren Buffett. His name was Eddie Lampert. Within a year he had combined Kmart and Sears into one company and set out to prove a great investor could turn around two aging retailers.
Early on, plenty of people liked the idea. Over the years that followed, even as the stores emptied and the stock slid, the story kept enough believers and enough financing to keep Sears alive. Companies fail all the time. Far fewer keep drawing patience and fresh money years after the business underneath has come apart.
A strong record made the slide easy to excuse
Lampert did not show up as an outsider with a pitch. He showed up with a record. Earlier bets on AutoZone and AutoNation had made him a billionaire and earned him comparisons to the best value investors alive.
That reputation raised the bar for anyone who wanted to argue he was wrong. When results started slipping, the credential bought him the benefit of the doubt, and the benefit of the doubt bought time. Sears got years of it.
Allocators run into a milder version of this constantly. A strong track record gets you in the room, yet it is where underwriting starts, not where it ends. Sears shows what happens when the starting point gets treated as the whole case.
Cutting investment can look like running things well
For the first stretch, Lampert's Sears threw off cash and the numbers looked healthy. Instead of pouring money into stores, the company handed it back to shareholders. Between 2005 and its 2018 bankruptcy, Sears repurchased roughly $6 billion of its own stock.
Spending on the stores themselves stayed light. In the first quarter of 2005, for instance, the whole company spent about $66 million on capex while running a business that booked tens of billions in annual sales. For a while, that flatters the picture. Fewer shares lift per-share earnings. Less spending lifts free cash flow. A company that stops feeding itself may look lean before it looks starved.
Economist William Lazonick, who studies buybacks, later summed up the cost. Had that money gone into stores and the website, he told CNN, Sears could have been in “a very different position”. When reading a set of early numbers, outperformance that comes from not spending is hard to tell apart from outperformance that comes from skill.
Inside Sears, the plan turned staff against each other
Underinvestment was only half of what hollowed out the operation. The other half was a management experiment.
Starting around 2008, Lampert broke Sears into more than 30 separate units, each with its own budget, its own boss, and its own profit target. A believer in market competition, he expected the internal contest to sharpen everyone. Reporting drawn from interviews with dozens of former executives later on documented the opposite. The divisions turned against each other. One unit gave better placement to rival products than to Sears's own Kenmore line. Teams bid against each other for space in the weekly sales flyer.
A company built to run like an open market ran like a turf war, and shoppers walked into stores that showed it. Two things were failing at once: the stores were starved of cash, and the organization was turning on itself. From the outside, the investment story still looked intact.
A theory that survives every bad quarter has stopped being tested
Lampert had a persuasive answer for every weak result, which was the trouble. The story went that Sears was secretly worth far more than it looked, and that the market was too slow to see it. Lampert pointed to the pieces the numbers supposedly missed:
the real estate under the stores
brands like Kenmore and Craftsman
and, later, a loyalty program meant to pull shoppers back
Framed that way, almost nothing could disprove him. A weak quarter meant the market still hadn't caught on. Falling foot traffic meant the real value sat in the land rather than the sales floor. Every disappointment fed the thesis instead of testing it.
A story built to survive any result cannot be argued down. It can only be checked against something outside the story, and at Sears one such thing was sitting in a public filing.
Three roles, one person
In July 2015, Sears sold 235 stores and interests in 31 more to a brand new real estate company called Seritage for about $2.7 billion, then rented the space back. Lampert chaired Seritage and was a major shareholder in it. Value moved out of a retailer that was losing money and into a property company he controlled, with him on both sides of the table.
The same pattern ran through the whole saga. Lampert and his fund were:
The lender. They became Sears's largest creditor, owed at least $1.3 billion, and were still arranging a fresh $300 million loan as Sears entered bankruptcy.
The landlord. Through Seritage, they collected rent on the stores Sears kept operating.
The buyer. From the Seritage real estate to a controlling stake in Lands' End to the leftover company itself in 2019, the fund kept picking up the pieces worth having while controlling the seller.
Operating results were genuinely murky in the moment. The direction of those transfers was not. When one person sits on both sides of a deal, you can often tell which way value is flowing before you know how the business turns out.
When Seritage was spun off, Warren Buffett personally bought about 8% of it. The real Buffett put his money in the property and passed on the stores. The bankruptcy estate later alleged Lampert had underpriced that real estate by at least $649 million, and in a broader suit that insiders drained value across roughly $2 billion of transactions. ESL rejected the claims as “baseless allegations”, and the case settled in 2022 for $175 million with no admission of wrongdoing. The point is that the structure was on file in 2015, long before any judge weighed in.
The same moves built an empire for someone else
None of Lampert's tools were exotic. Spin-offs, real estate carve-outs, sum-of-the-parts thinking, patient bets against the crowd: this is the standard kit of a certain kind of investor. John Malone used a near-identical toolkit for decades and built a media fortune with it.
So the tools themselves were not the problem. What set Malone's version apart was whether the core business was being fed or drained while all the clever structuring happened around it.
Where the recession explanation runs out
Former Sears chief executive Alan Lacy argues that the 2008 crash and the rise of Amazon were the real killers, calling the recession the “nail in the coffin” and largely separate from how Lampert ran the place. He has a point about difficulty. Retail did turn brutal in those years.
Difficulty, though, does not explain the choices. A hard economy did not decide to send billions to buybacks rather than into stores. It did not route rent and asset sales toward companies the chief executive controlled. The environment set how hard the game would be. Lampert's decisions set where the money went.
The condition that makes this repeatable
Sears was not a one-off failure of judgment. It was a predictable outcome of a structure that recurs across private markets, where the thing being valued cannot be checked until long after capital is committed. As one Collateral Partners piece notes, financial products are goods you cannot fully assess before you buy them.
That delay is the opening. When the payoff sits years out and the core claim rests on a number no one can mark today, a credible reputation can carry a valuation well past the point performance would otherwise support it. Lampert had the reputation and the unmarkable number at once. Most managers who look like him have only one of the two, which is why the full version is rare and worth recognizing when it appears.
Bottom line
The uncomfortable thing about Sears is how little detective work the ending required. The two facts that mattered, that the stores were being starved of capital and that one man sat on both sides of the largest deals, were sitting in public filings for years. They were also, frankly, dull. A rights offering and a sale-leaseback do not make for a good argument at an investment committee. The turnaround thesis did.
Diligence attention tends to pool around the parts of a manager's story that are most debatable, because that often seems more interesting. The parts that are hardest to spin, related-party flows, the direction of cash, who is buying what from whom, generate no debate and therefore attract less scrutiny than their predictive value deserves.
A proven reputation widens that asymmetry rather than closing it. The stronger the pedigree, the more willing an allocator is to treat the story as the thing to underwrite and the plumbing as a formality. Sears suggests the inversion is the safer approach. When a manager is credentialed enough that the narrative feels settled, the unglamorous columns deserve a second read, because they are the only ones the reputation cannot color.
