Picture a storefront in a busy American city sometime after the Civil War. A clerk chalks the latest stock prices on a big board as they come in over the wire. Another takes a few dollars from a factory worker who is sure a railroad stock is about to rise. If it goes up, he collects. If it goes down, the house keeps his money.
He never owns a single share. No stock is ever bought or sold. He is simply betting on a number.
That was a bucket shop. And if it sounds familiar, it should. Open almost any prediction market app today and you'll find the same basic idea, rebuilt for smartphones: bet on what a number, an election or a game will do, without owning anything at all.
How Bucket Shops Worked
Bucket shops spread after the stock ticker made real-time prices widely available. By the turn of the century, several hundred were operating across the country.
The pitch was access. The official exchanges were effectively closed to ordinary people, with large minimum trade sizes, high margin requirements and brokers who didn't want small accounts. Bucket shops welcomed anyone with a few dollars. Some claimed thousands of customers.
But the customer wasn't trading with the market. They were trading against the shop. The prices came from the exchange ticker, yet the bets never touched the exchange. When customers lost, the shop won. And shops had every incentive to make sure customers lost, sometimes by closing positions at convenient moments or simply disappearing when the house was on the wrong side.
Translation: it looked like investing, it used the vocabulary of investing, and it was gambling.
How America Shut Them Down
The exchanges hated bucket shops, partly on principle and partly because the shops were siphoning off business using the exchanges' own price data.
The decisive blow came in 1905, when the Supreme Court ruled in Board of Trade v. Christie Grain & Stock Co. that exchanges owned their price quotations and could control who received them. Exchanges cut the shops off from the data they depended on, and states passed laws against them. Within about a decade, the bucket shops were essentially gone.
Something else happened too. Legitimate brokers, having watched how much ordinary people wanted access, started courting small investors themselves. The demand bucket shops had revealed didn't disappear. It got channeled into regulated markets.
Enter the Prediction Market
Fast forward to today. Platforms like Kalshi and Polymarket let users buy "event contracts" that pay out depending on whether something happens: an election result, an interest rate decision, an award show, a sports game.
Their argument is that these are financial instruments, regulated as derivatives by the Commodity Futures Trading Commission, not bets regulated by state gambling commissions. They describe the result as a market that aggregates information and produces useful forecasts.
There's something to that. Markets can be good at pricing probabilities, and some research suggests prediction markets forecast certain events reasonably well.
But look at where the money actually goes. According to the Associated Press, sports made up about 90% of Kalshi's trading volume as of early 2026, and roughly half of Polymarket's. That is not a market discovering the future of monetary policy. That is a sportsbook with a finance-flavored interface.
The Fight Over What These Things Are
This is exactly the question that ended the bucket shop era: is it investing or is it gambling?
Several states have answered "gambling." They've sued or ordered platforms to stop offering sports contracts, arguing they're unlicensed sports betting that skips state rules, taxes and age limits. Many state gambling laws require players to be 21, while prediction platforms have often allowed users as young as 18.
The federal government has largely taken the platforms' side, with the CFTC arguing it has exclusive authority over these contracts. But in August 2026, a federal appeals court sided with Nevada, finding that Kalshi was unlikely to succeed in arguing that federal law overrides the state's gambling rules for sports contracts. Other courts have reached different conclusions, and many observers expect the fight to end up at the Supreme Court.
In other words, we're rerunning a 120-year-old argument, with apps instead of chalkboards.
The Lessons We Seem Determined to Relearn
Calling it a market doesn't make it investing. Bucket shops used ticker prices and trading language. Prediction markets use contracts, order books and "probabilities." The question that matters is whether participants are allocating capital or placing bets.
Access cuts both ways. Bucket shops genuinely served people shut out of Wall Street. Prediction markets genuinely make it easy to put money on your views. But easy access to a product designed so most participants lose isn't democratization. It's distribution.
Follow who wins. In a bucket shop, the house profited from customer losses. On today's platforms, revenue often comes from fees and from the flow of trading activity, which grows when people trade often, emotionally and on impulse. That's a familiar pattern from the AI gold rush too: the platform gets rich whether or not the users do.
Regulation arrives late and messy. The bucket shops weren't stopped by a single elegant rule. It took court fights, state laws and pressure from established players. Expect the same here.
Trust is the real product. Markets only work when people believe the game is fair. Insider information, manipulation and murky rules can erode that quickly. As I wrote in What AI-Powered Due Diligence Still Cannot Detect, no amount of technology substitutes for honest participants.
So Where Is This Headed?
Let's be clear: prediction markets aren't identical to bucket shops. They operate openly, under at least some federal oversight, and they can produce genuinely useful forecasts on non-sports questions.
But the core tension is the same one America faced a century ago. When a product lets ordinary people bet on outcomes, wraps it in the language of finance and grows fastest on the most gambling-like uses, regulators eventually have to decide what it really is.
Last time, the answer reshaped American investing. This time, the answer will decide whether prediction markets become a serious financial tool, or just the bucket shop with better branding.