Across five decades and four generational cohorts, the companies that the loudest segments of the public have most wanted to see fail have a curious habit of outperforming. This isn't a moral judgment. It's an empirical pattern with a well-understood mechanism. Here's what it means for a 2026 portfolio.
In 2009, finance professors Harrison Hong and Marcin Kacperczyk published "The Price of Sin: The Effects of Social Norms on Markets" in the Journal of Financial Economics. They examined the long-run returns of so-called "sin stocks" — alcohol, tobacco, gambling — and found that these stocks had outperformed the broad market by roughly 2-3 percentage points annualized over multiple decades. Not by trivial amounts. Persistently.
The mechanism they documented: a meaningful slice of institutional capital is contractually forbidden from owning these companies (university endowments, public pension funds, certain mutual-fund mandates, faith-based investors). Suppressed demand depresses prices. Lower prices mean higher expected forward returns for investors who will own them. Combine that with inelastic addiction-driven revenue and high regulatory moats blocking new entrants, and you get a structural premium.
Warren Buffett has been explicit about the pattern for fifty years. His See's Candies, Coca-Cola, and tobacco-adjacent positions weren't morally chosen — they were chosen because customers couldn't or wouldn't switch. The "moral" reading came later, as a side effect of how the moats worked.
The pattern hasn't gone away. The targets have rotated.
This is not a comprehensive sociology. It's a sketch of the dominant moral targets each cohort identified during their formative consumer years, drawn from longitudinal trust surveys (Pew Research, Gallup, the Edelman Trust Barometer), academic survey literature, and the public record. Each generation's list reflected their lived concerns. Each list, looking back, also reflected which companies were hardest to dislodge.
Note the pattern: the moral target rotates with each cohort's formative experience, but the trade is the same. The companies people most wanted to see fail tended to be the ones hardest to compete against. The two facts are connected — that's the point of the next section.
The correlation between moral controversy and outperformance isn't mystical. It runs through five well-documented mechanisms.
A growing slice of institutional capital is contractually prohibited from owning the targeted names. With less buying pressure, prices stay lower than fundamentals justify. Investors who do own the stock collect the difference as forward return. Hong & Kacperczyk's mechanism, generalized.
The tobacco master settlement made surviving cigarette companies more profitable, not less. Compliance burdens that follow major regulatory hits make new market entry prohibitive. The companies that survive the moral hit emerge with the field cleared.
Customers of controversial products often can't or won't switch. Addictive products, network-effect products, government-mandated products, irreplaceable platforms. Buffett's whole career rests on this observation.
For founder-led companies (Tesla / xAI under Musk, OpenAI under Altman), the controversy itself drives capital, talent, and customer mindshare. The hate is functional, not incidental — opposition keeps the names in the news as effectively as fans do.
The newest mechanism, magnified in 2025-26. When the state becomes the buyer of last resort (Palantir, defense, hyperscaler cloud, AI labs taking Pentagon contracts), the customer cannot leave. National security framing converts ordinary corporate revenue into structurally protected revenue.
Companies that externalize costs onto society (environmental damage, mental health impact, surveillance costs to civil liberties) operate with lower internal costs than competitors who internalize them. The market rewards externalization until regulation forces internalization — usually decades later, after compounded outperformance.
None of these requires the company in question to actually be morally bad. They only require the company to be perceived as morally bad by enough of the public to trigger the screening, regulatory, and selection effects above.
Which is the perspective shift this page exists to make: moral controversy is more reliably mispriced as a risk than it is correctly priced as one.
For each of the names that draw the strongest current public moral concern, here is which mechanism above is doing the work — viewed neutrally, mechanism-first.
Mechanism: founder-personality network effects (#4) plus government-aligned positioning (#5, especially after the 2025 administration change). The controversy fuels engagement; the engagement fuels capital and talent inflow. Customer base self-selects through the controversy — buyers are people for whom Musk's posture is a feature.
Mechanism: government as customer (#5), with a regulatory/clearance moat that blocks new entrants (#2). Every administration since 2010 has expanded Palantir's contracts. The civil-liberties-concern community provides an ideological floor on competitive entry — few well-funded competitors have wanted to chase the surveillance-state customer base, leaving Palantir nearly uncontested in its category.
Mechanism: structural lock-in via passive index dominance (a variant of #5 — the financial system itself is the captured customer). The "BlackRock controversy" of recent years coexists with the simple structural fact that index funds are how most retirement money invests, and the largest index manager has nowhere to go but bigger.
Mechanism: government as customer (#5), plus negative externality cost advantage (#6, in the form of war-cost externalization). Performance is essentially uncorrelated with administration. The "Cold War 2.0" framing of 2024-26 has hardened the moat further.
Mechanism: late-stage migration into #5. As frontier labs face the recoupment math we cover in our trillion-dollar-question analysis, the laboratories that take the controversial defense contracts become the ones with reliable revenue. The Anthropic-to-OpenAI Pentagon contract swap of March 2026 is the data point that crystallized this — the lab that drew red lines lost contracts within hours to the lab that didn't.
Funds that screen out these names consistently underperformed unscreened benchmarks 2020-2025 by amounts well-documented in financial press. The capital that exited the controversial names had to go somewhere. It went into less-controversial alternatives that, on average, lacked the same moats.
The controversy premium isn't permanent. It ends in one of three ways. Recognizing which way matters more than recognizing that it ends.
The tobacco master settlement of 1998 didn't kill the survivors — it made them more profitable for years. But certain industries did get broken: lead paint, asbestos, leaded gasoline. The distinguishing feature: regulation that imposes legal liability extending backward through time, not just forward. A cosmetic compliance burden hardens the moat. A retroactive liability regime breaks it.
Cigarettes were "evil" in 1995 and culturally unremarkable by 2020. The moral panic faded as the generations who lived through the controversy aged out. When the moral panic ends, the screening exemption ends, and the structural premium collapses. The companies don't lose their customers — they lose their forward-return advantage as ESG dollars start flowing in again.
The rarest case. Patagonia was supposed to disrupt fast fashion. It didn't. Beyond Meat was supposed to disrupt factory farming. It hasn't, at scale. The reason ethical competitors usually fail: they internalize costs that incumbents externalize, which makes them structurally more expensive, which makes them niche.
For each of the 2026 names above, the question to track is which of the three inflections is closest. For Palantir, it's almost certainly inflection 2 — generational replacement is the longest-dated risk. For Big Tech, it's a closer call between inflection 1 (antitrust hardening) and inflection 2. For AI labs, it's inflection 3 — whether a genuinely-aligned, well-funded competitor emerges. None of these are imminent.
Five takeaways, useful regardless of which side of any controversy you sit on:
28 industries scored across 5 horizons (1yr, 2yr, 3yr, 5yr, 10yr) using 8 analytical dimensions and 167 cross-industry effects — moat structure, demand elasticity, regulatory positioning, and capital intensity all included. Free portfolio scan available.
See the Matrix Scores Free Portfolio ScanIf you're a market-first investor who finds moral discourse around stocks tiresome — this page is the analytical version of what you already suspect. The "evil company outperforms" pattern is real, well-documented, and structurally caused. Stop apologizing for understanding it.
If you have ethical concerns about specific companies in your portfolio — this page doesn't tell you to abandon those concerns. It tells you that abandoning them is, on average, the higher-return choice, so that you can decide consciously rather than have the choice made for you by an ESG-screened fund. You can still pay the premium for moral comfort. Just price it accurately.
If you're a skeptic of the whole enterprise — fair. The framework here is sympathetic to your skepticism. Markets don't reward moral superiority. They reward durability of demand and structural protection. That observation has held across five decades. The 2026 configuration just makes it harder to ignore.