Bankrupt in 1996. Out of A-list characters. So it borrowed $525 million against the leftovers and bet the company on a guy in a metal suit.

There’s a version of the Marvel story that gets told at conferences, and it goes: great IP always wins in the end.

That version is wrong in an important way. In 1996, Marvel owned essentially the same characters it owns today — Iron Man, Thor, Captain America, Black Widow, the Hulk, the entire Avengers roster. The IP was already there. It was worth close to nothing, because nobody had figured out how to convert it into cash flow.

The turnaround wasn’t about discovering the assets. It was about changing the business model wrapped around them — and being willing to risk the assets to do it.

The Setup: A Bubble Made of Comic Books

The early 1990s comic market was a speculative mania. Collectors bought multiple copies of #1 issues in polybags, convinced they were buying appreciating assets. Publishers, Marvel loudest among them, printed straight into the frenzy: variant covers, foil covers, endless #1s, event crossovers.

Marvel’s owner, Ron Perelman, financed aggressively against that revenue and pushed the company into adjacent businesses — trading cards, toys, stickers. Then in 1994 came the move that turned a downturn into a catastrophe: Marvel bought Heroes World, a distributor, and moved to distribute its own comics directly.

It alienated the existing distribution network overnight, and it turned out Marvel wasn’t good at logistics. Retailers got books late or not at all. Competitors signed exclusive deals elsewhere. And underneath it all, the bubble popped exactly as bubbles do. Speculators discovered a print run of two million “rare” comics is not rare. Shops closed by the thousand.

Revenue collapsed into a debt load of roughly $610 million.

The Collapse

Marvel filed for Chapter 11 on December 27, 1996.

What followed was less a restructuring than a knife fight: Perelman and Carl Icahn battled for control of the carcass while the company laid off more than a third of its staff. The eventual resolution, in 1998, handed control to Toy Biz — meaning Isaac Perlmutter and Avi Arad, a toy executive and a toy designer, now ran the House of Ideas.

To generate cash, Marvel did the obvious thing: it licensed its most valuable characters to studios that could actually make films. Spider-Man went to Sony. X-Men and the Fantastic Four went to Fox.

It worked, sort of. X-Men (2000) and Spider-Man (2002) were enormous hits. Marvel collected a licensing fee — a thin slice of a very large pie it had baked. Marvel watched other people get rich off Marvel characters and booked what amounted to a rental income.

By the mid-2000s, that left the company in the strangest possible position: proof that its IP could print money at the box office, and a contractual structure that guaranteed it would capture almost none of that money.

The Turn: Stop Renting. Start Producing.

The architect of what came next was David Maisel, who joined Marvel with a simple, borderline reckless proposition: Marvel should finance and produce its own films.

The obstacles were stacked. Marvel had no studio infrastructure. It had already sold off its best-known characters. Its remaining roster was, in 2005 Hollywood’s assessment, a bin of second-stringers — Iron Man was best known to non-readers, if at all, as a hero with an alcohol problem. And film production is the single most capital-intensive way to lose money ever devised.

Maisel’s pitch to a skeptical board was essentially a floor argument: if this fails, we’re roughly where we already are. If it works, we own the whole upside instead of a licensing fee.

In 2005, Marvel closed a $525 million non-recourse credit facility with Merrill Lynch, structured over seven years to fund up to ten films with budgets between $45 million and $180 million. Paramount agreed to distribute and fund marketing without contributing production capital, leaving Marvel with creative control.

The collateral was the point. Marvel pledged the film rights to ten core characters — Iron Man, Thor, Captain America, Black Widow, the Hulk and others. If the films failed, a bank would own the Avengers.

They then made the least obvious choice available: the first film out of the gate would be Iron Man, starring an actor whose recent history had made him close to uninsurable.

The Payoff

Iron Man opened May 2, 2008, and grossed roughly $585 million worldwide.

More importantly, it worked as an architecture. The post-credits scene establishing a shared universe converted every subsequent film into a marketing vehicle for every other film — a compounding asset structure that no rival studio had, because no rival studio owned a coherent stable of interlocking characters.

In 2009, Disney acquired Marvel Entertainment for about $4.24 billion. The Marvel Cinematic Universe built from those “leftover” heroes has since grossed over $30 billion at the global box office, the highest-grossing film franchise in history.

Sony and Fox had bought the characters everyone wanted. Marvel got rich on the ones nobody did.

What Operators Can Steal From This

Owning the asset is not the same as owning the business model. Marvel held the IP the entire time it was worthless. Value showed up when it changed where in the value chain it sat — from licensor to producer.

Licensing is comfortable, and comfort is the trap. Fees are low-risk, low-capital, and low-ceiling. If your best-performing product line is someone else’s business built on your inputs, you have a strategy problem, not a revenue problem.

“We’re already at the bottom” is legitimate risk-adjusted reasoning. Maisel’s argument only works because the downside case was close to the status quo. Run that test honestly before you dismiss a bet as too aggressive — and just as honestly before you tell yourself you’re at the bottom when you aren’t.

Build for the second product, not the first. Iron Man‘s real innovation was that it made the next nine films cheaper to sell. Ask what your current launch makes possible, not just what it earns.

Vertical integration cuts both ways. Buying Heroes World nearly killed Marvel. Producing its own films saved it. The difference wasn’t integration — it was whether the company had any competence in the thing it was absorbing, and whether it had the balance sheet to survive learning.