poison

In a toy aisle, Pikachu outlasts any actor who ever voiced it. On a studio greenlight spreadsheet, a single lead credit can mean the difference between a franchise extension and a write-down. Pokémon earns $91 billion lifetime across 107 tracked franchises in Artometrics' TidyTuesday media-franchise file — and 61 percent of that total comes from merchandise, not from any one person's opening weekend. A six-film slate led or co-led by Jared Leto in the 2020s tells the opposite structural story: worldwide gross against reported budget ranging from 0.65x to 2.04x, with half the titles at or below breakeven and no licensing floor underneath.

This report is an editorial synthesis. It reads two published Artometrics datasets — franchise revenue and horror-tagged box office — alongside a small, separately compiled case study of one actor's recent theatrical credits. The argument is about revenue architecture, not scandal; personal-conduct questions belong in The Protection Threshold, not here.

Film economists have long treated theatrical returns as high-variance bets — Arthur De Vany and W. David Walls documented the "wild uncertainty" of box-office outcomes in The Journal of Economic Perspectives (2004). Transmedia scholars describe why character-based IP can absorb format shifts without collapsing on a single performance. Those literatures frame the charts below; they do not prove anything about any individual named in the case study.

RESEARCH QUESTION

Does diversifying a franchise's revenue away from any single star's box-office draw — toward merchandise and licensing, the way the top-earning franchises do — measurably reduce exposure to concentrated "star risk"? And does a small, star-fronted film slate show what the absence of that cushion looks like in practice?

The numbers that set the scale for this report:

MERCHANDISE IS THE INSULATION IP FRANCHISES HAVE AND STAR VEHICLES DON'T

Artometrics' franchise-revenue report established that merchandise is not one revenue stream among several — it is the dominant one, at 61 percent of everything tracked across 107 franchises earning at least $4 billion lifetime. Pokémon's $91 billion total, the highest of any tracked property, was built almost entirely on cards, games, and retail rather than any theatrical run. Henry Jenkins's transmedia framework (Convergence Culture, NYU Press, 2006) helps explain why: revenue attached to a character or mascot can move across games, apparel, and film without routing through a single performer's weekend gross. Contrast that with a franchise or vehicle built around one performer's draw. When merchandise, licensing, and games are absent or minor, the entire revenue case rests on box office and, downstream, on home video and streaming licensing tied to that same theatrical result. There is no secondary economy running in parallel the way there is for Pokémon or, per the same report, for Star Wars — the one franchise in that dataset the report specifically flagged as fully diversified across box office, merchandise, home video, and games. A star-fronted film without that diversification is, in revenue-structure terms, closer to a single bet than to a franchise.

WHAT CONCENTRATED STAR RISK LOOKS LIKE: A SIX-FILM CASE STUDY

The case-study dataset is small and deliberately narrow: six of actor Jared Leto's 2020s film credits, each expressed as worldwide gross divided by reported production budget, using trade-reported budget bands from Box Office Mojo. House of Gucci (2021, ensemble) returned 2.04x on a $75 million budget. Morbius (2022, lead) returned 2.02x on $83 million. Those two clear breakeven with room. The other four sit closer to the line or under it: The Little Things (2021, lead, released into a pandemic-narrowed theatrical window) landed at 1.03x; Haunted Mansion (2023, ensemble) at 0.79x; Masters of the Universe (2026, supporting) at 0.66x; and Tron: Ares (2025, lead) at 0.65x, on the largest reported budget in the slate at $220 million. Read individually, each figure is a data point about one film. Read as a slate, the pattern is a distribution centered near or below 1.0x rather than comfortably above it — half the titles below breakeven, and only two of six clearing 2.0x. That is the shape of concentrated star risk: revenue tied tightly to one theatrical outcome per title, with no licensing or merchandise base underneath it the way Pokémon's $91 billion or Star Wars' diversified mix has. When a title in this kind of slate underperforms, the studio absorbs the loss directly — there is no secondary revenue stream to blunt it. This is a small, non-random sample, and the report treats it as illustration, not proof; see Limitations below.

WHY GENRE AND CAST METADATA CAN MISLEAD SINGLE-TITLE CONCLUSIONS

Artometrics' margins report already demonstrated a caution that applies directly here: community-cleaned genre and metadata tags in box-office tables are unreliable enough to mislead a reader trying to draw conclusions about any single title or person. That report's dataset, nominally tagged "horror," was topped by Star Wars Episode I: The Phantom Menace at $474,544,677 domestic — a science-fiction adventure film, not a horror film, sitting atop a genre-labeled table because of a tagging error in the community-cleaned merge. The median across all 3,401 films in that file was $25,533,818 domestic, spanning 1936–2019. The lesson generalizes beyond genre tags: any table that mixes lead, supporting, and ensemble billing, or that blends pandemic-window releases with ordinary theatrical runs, risks producing headline numbers that look like a verdict on one person when they are really an artifact of how the underlying rows were labeled and merged. The six-film case study above mixes exactly those roles — lead, supporting, and ensemble — and one pandemic-era release. That mix is disclosed, not hidden, because the margins report's central caution is that undisclosed metadata quietly drives conclusions readers think are about content, when they are really about cleaning decisions.

LIMITATIONS

Six films is a small sample. It cannot generalize to "all star vehicles" or to any actor's career as a whole, and it should not be read as a ranking or grade of any individual film. The gross-to-budget ratios use trade-reported budget bands, which studios do not always disclose precisely or consistently; treat them as order-of-magnitude figures, not audited production accounting. No causal link between reputational risk and box-office softness is established, tested, or claimed in this report — the framework below is an editorial hypothesis about revenue structure, not a measured finding, and is labeled as such. The franchise and horror datasets describe different populations (107 large franchises; 3,401 individual films) and different time windows; combining them illustrates a structural argument, not a single unified statistical model.

CONCLUSION

Editorial framework, not measured finding: the evidence assembled here is consistent with the idea that merchandise-driven IP carries structurally less exposure to any one person's performance than a star-fronted vehicle does. The six-film case study illustrates concentrated theatrical exposure when that cushion is absent — not proof about any particular result. De Vany and Walls (2004) show why theatrical outcomes alone behave like lottery tickets; this report asks whether merchandise share is the hedge studios buy when they want something sturdier than a lottery ticket. A future report with a larger sample could test whether merchandise share predicts how much a single underperforming title costs the studio that owns it.

DATA, METHODS & SOURCES