Everything is a casino

Open your phone and look at the bottom row of apps. There is a fair chance one of them lets you put $20 on tonight's game, another lets you buy a weekly call option that expires worthless by Friday, a third sells you a coin named after a frog, and a fourth quotes you live odds on whether the Fed cuts in December. Four different icons, four different regulators, four different stories about what kind of person you are when you tap them — an investor, a fan, a degenerate, a forecaster. Same thumb. Same little jolt when the number moves. Same screen, glowing in the same dark room at the same 1 a.m.

I want to argue something that sounds like a metaphor but is closer to an accounting fact: those four felts are becoming one table. The walls that used to separate investing from gambling from entertainment have come down — in the product design, in the regulation, and most of all in the head of the person tapping. We are living through the casino-ification of ordinary financial life, and the most consequential effect isn't moral. It's that capital, which spent a century holding its nose at anything that smelled like gambling, has noticed the smell is now everywhere — and started to follow it.

This is the first of three pieces on online gambling as a business. Before we get to the unit economics — and they are extraordinary unit economics — and before we follow the money around the world to see where it actually pools, we have to start with the strange cultural moment that is making a once-untouchable industry suddenly investable. Start, in other words, with the convergence.

One screen, every table

Take the inventory in order of how respectable each one is still allowed to pretend to be.

Sports betting is the loud one. Seven years after the Supreme Court struck down the federal ban in Murphy v. NCAA,¹ legal sports wagering runs in 38 states and Washington, D.C. Americans bet a record $149.9 billion on sports in 2024 — handle that simply did not legally exist in 2017 — and the books kept $13.8 billion of it.² FanDuel and DraftKings turned the second screen of every game into a live wager. A parlay is a lottery ticket you build yourself, and it is marketed as fandom.

Retail trading is the respectable one, and it has quietly become the wildest. The signature instrument of the era is the zero-day-to-expiry option — a contract that lives and dies inside a single trading session, a pure bet on the next few hours of the S&P. Robinhood built a generation's idea of "investing" around the same dopamine architecture as a sportsbook: confetti, streaks, a number that moves while you watch. And the line is no longer even rhetorical. In March 2025 Robinhood launched a prediction-markets hub — event contracts, routed through the CFTC-regulated exchange Kalshi — and by August it was offering markets on pro and college football inside the same app you buy Apple stock in.³ The brokerage and the book are now one login.

Crypto dropped the pretense entirely. The defining asset of 2025 was the memecoin — a token with no cash flow, no project, sometimes no pretense beyond the joke in its name. In January a coin branded $TRUMP launched and briefly carried a paper market value in the tens of billions before most of it evaporated;⁴ it was, transparently, a slot machine with a ticker. Coinbase, a regulated public company, is the on-ramp. (The strange path that got us here — from a 2013 dog joke to a president's coin — is its own story, which I told in a joke with a ticker.)

And prediction markets — the category I spend most of my time writing about — sit in the most interesting seat of all. Kalshi, the first federally-regulated event exchange, raised at a $2 billion valuation in June 2025;⁵ Polymarket runs the same game offshore. To a regulator, an event contract is a financial derivative. To the kid clicking it, it is a bet on the news. Both descriptions are correct, and that is exactly the point: the same object is a security, a wager, and a piece of content depending only on who is looking at it.

Lay them on a line from one honest extreme to the other — from owning a slice of a productive company to feeding a machine engineered to take your money — and you do not get neat categories. You get a smear. The middle, where most of the new products live, is where the words stop meaning anything.

Financial nihilism

Why now? The products are slicker than they used to be, but slick alone doesn't explain a culture-wide shift. The deeper engine is an attitude, and it already has a name. The investor and podcaster Demetri Kofinas popularized the term "financial nihilism" around 2021: the worldview of a generation that has concluded the patient, sensible path — save, compound, buy a house, retire — quietly stopped working, and that if the game is rigged anyway, you may as well play the version with the best story and the fastest payout.⁶

It is easy to sneer at this, and wrong to. Look at it from the inside. If a house costs eleven times your salary instead of three, if the index fund returns a number that feels theoretical against rent that is extremely real, then a lottery ticket is not obviously the irrational choice. It is a small, voluntary tax on hope, and hope is one of the few things still on sale at a price you can afford. A $50 long-shot parlay that could pay $5,000 is, for someone who will never accumulate $5,000 by saving $50 at a time, a coherent — if losing — strategy. Wall Street built index funds for people who believe tomorrow will reward patience. Financial nihilism is the product line for people who don't.

This isn't a fringe read anymore. As I write, even the business press has caught up: this week CNBC ran a piece on young adults embracing exactly this posture — piling into crypto, memecoins, and options not in spite of the risk but because the slow path feels foreclosed.⁶ When the safe move stops feeling safe, the risky move stops feeling risky. That is the psychological precondition for everything that follows.

What the house actually sells

Strip the skins off all of these products and most of them share one mechanical organ: the edge. A casino game is built so that, on average, every dollar wagered returns slightly less than a dollar. That gap is the entire business. A slot machine advertised at a 96% return to player keeps four cents of every dollar that flows through it:

The genius — and the menace — is in that little \(N\). One spin loses you four cents on average; you would never notice. But a slot resolves in three seconds, and a phone never closes, so \(N\) climbs into the thousands in an evening, and the law of large numbers does the rest. The house doesn't need to win any single hand. It needs you to keep playing, and it has engineered the most compelling reasons-to-keep-playing humanity has ever built. The product is not the game. The product is volume.

A sportsbook hides the same edge inside the odds. When both sides of a game are priced at the standard −110, each side's quoted odds imply a 52.4% chance — and the two implied probabilities add up to more than 100%. That overhang, the vig, is the hold:

Now hold that against the one product in the inventory that is built differently — and this distinction is the whole reason I keep insisting prediction markets are not just another casino game. In a real market, the two sides of a binary contract — YES and NO — are priced by traders against each other, and in a healthy book the prices sum to about 100%, not 105%. There is no structural overhang skimmed by the house, only a thin fee. The venue is a counterparty to nothing; it takes a toll for running the road, not a cut of who wins:

That gap — between a price the house sets to guarantee its take, and a price the crowd discovers with no built-in rake — is the line between gambling and a market. It is a real line. But I'd be kidding you if I pretended the average person tapping the average icon feels it. To the user, the dopamine is identical. The convergence is happening in the part of the brain that doesn't read the prospectus.

Why this is the backdrop

Here is why a research desk that mostly writes about markets and AI is spending three pieces on gambling. The convergence isn't a curiosity. It is the macro condition that makes the next two essays possible.

Because the stigma is dissolving, the capital can move. For decades, serious investors treated gambling as untouchable — not because the returns were bad (they were spectacular) but because their own backers wouldn't allow it. When sports betting becomes a halftime ad read by a beloved ex-quarterback, when "investing" and "betting" share an app icon, when a regulated brokerage lists the odds on a football game, the thing that kept money out — that it was disreputable — quietly expires. That is the subject of the next piece: a business with software margins that venture capital ignored on grounds that no longer hold.

And because the behavior is now universal, the market is enormous and global — and most of it is happening somewhere its own government can't see or tax. That is the subject of the third: where the money actually pools, from the offshore two-thirds of the American market to the parts of Asia rewriting their gambling laws in real time.

55%

of US adults gambled in the past year (2024) — up from 49% a year earlier⁷

$149.9B

wagered on US sports in 2024 — a record; legally zero before 2018²

38+DC

states with legal sports betting — from a single state seven years ago²

What's hard — and what I won't pretend

I find this industry genuinely fascinating as a business. I also think it does real harm, and I'm not going to launder one feeling into the other.

The harm is not hypothetical. The wave of research following legalization is sobering and consistent: in states that legalized sports betting, households — especially financially stretched ones — show measurably lower savings and investment, higher credit-card debt, and rising bankruptcy filings.⁸ Jonathan Cohen's Losing Big and a year of front-page reporting have made the human version of that statistic impossible to ignore.⁸ A business whose margin is, mechanically, other people's losses cannot be discussed honestly as though it were a SaaS company that happens to sell fun. The volume that makes the economics beautiful is, on the other side of the ledger, someone's rent.

The honest intellectual move is to hold two things at once. The mechanism — a priced, voluntary, mostly-legal transaction — is morally neutral in the way a credit-card rail is neutral. The externality — what happens when you point that mechanism at human compulsion and remove every point of friction — is alarming, and pretending otherwise is how an industry gets its reckoning. I can think the casino is one of the best business models ever designed and think we should be far more careful about pointing it at everyone, all the time, through a device that never sleeps. Both are true. The pieces that follow keep both in view.

One more guardrail, because it matters to everything I write: blurry is not the same as identical. The fact that a prediction market feels like a bet to the person tapping it does not make it one, any more than the fact that an index fund feels like gambling to a nervous first-timer makes it one. Some of these products, underneath the identical dopamine, actually produce something — a price, a forecast, a hedge. Most produce only churn. Telling them apart, when they've been deliberately designed to look the same, is going to be one of the defining literacy problems of the next decade. The convergence is real. Refusing to think clearly inside it is optional.

So: everything is becoming a casino. The interesting questions are the ones that follow from taking that literally — as a market, as a business, as a map.