Bucketed

A bucket shop, the Supreme Court wrote, is “an establishment, nominally for the transaction of a stock exchange business, or business of similar character, but really for the registration of bets, or wagers, usually for small amounts, on the rise or fall of the prices of stocks, grain, oil, etc., there being no transfer or delivery of the stock or commodities nominally dealt in.”

That is Gatewood v. North Carolina, decided 24 December 1906. Read the last clause again — there being no transfer or delivery of the stock or commodities nominally dealt in — and then read the terms of any funded-account programme sold to retail traders this year. The clause is in both. In 1906 it was the thing the state had to prove in court. In 2026 the firm writes it into the contract itself, and sells a few hundred dollars’ worth of it to a customer who has read it and does not mind.

The resemblance is closer than the industry likes. It is also less damning than the comparison is usually made to imply, and on one point it runs firmly the other way. But the part worth the most attention is neither: it is how the first version actually ended. It was not ended by a regulator. The thing that ended it has a precise modern equivalent, and that equivalent has already been at work for two years.

The same trade, a hundred and twenty years apart

The bucket shop’s customer walked in, put money on the counter, and “bought” a hundred shares of Union Pacific at the price coming off the ticker. Nothing was bought. The shop recorded a wager against the tape’s next prints and took the other side of it. Leverage ran as far as 100 to 1 — a dollar deposited controlled a hundred dollars of stock. The shop’s revenue was its customers’ losses plus commission on both ends of a transaction that never happened. It carried no exchange risk, no clearing obligation and no counterparty except the people standing in front of it.

The funded-account firm’s customer pays a fee — fifty dollars for a small account, past a thousand for a six-figure one — for the right to trade a simulated account to a profit target under a set of house risk rules. Pass, and he is “funded”. At most firms the funded account is also simulated: payouts come out of the fee pool and the firm’s own book, and the firm decides for itself what fraction of customer flow, if any, it hedges into a live market. The ticket does not reach an exchange. Not during the evaluation, and usually not after it.

Both businesses need exactly two inputs. A price feed good enough to settle against, and a house rule that clears the customer out often enough to keep the fee cycle turning.

The rule that clears the customer out

The bucket shops did not make margin calls. That detail is worth sitting with, because it was not an oversight — it was the entire mechanism. A legitimate broker telephones you for more money because he wants the position kept open; he earns on the position. The bucket shop earned on the closing-out. So a move through your margin produced no telephone call. It produced a forfeiture: the whole deposit, at once, to the house. At 100 to 1, one per cent against you was the ticket. Lefèvre’s account of the 1890s, published in 1923 and generally read as Livermore’s own, describes the other half of the arrangement too — what happened to a customer who was consistently right, which was that shop after shop refused his orders and then barred him from the premises.

The modern equivalent is the daily loss limit and the maximum drawdown. FTMO’s two-step challenge is representative: a ten per cent profit target in the first phase, a five per cent daily loss limit measured from each day’s opening balance, and a ten per cent overall loss limit. The daily limit counts open positions, not just closed trades. Touch it intraday, on an unrealised excursion, and the account is finished — not called, finished. There is no version of that rule under which a trader who is right at the end survives being wrong in the middle, which is a fair description of nearly every trader who is right at the end.

The outcomes follow from the rule. FTMO has cited pass rates around nine to ten per cent for the two-step evaluation; Topstep’s own site discloses that 16.8 per cent of Trading Combines begun in 2025 were completed. A third-party study circulated this year, covering some 300,000 accounts across ten firms, put the pass rate near fourteen per cent and found that about 45 per cent of funded traders ever received a payout — roughly seven per cent of everyone who bought a challenge. All of these numbers are self-reported or unaudited, and the firms that publish are the firms with numbers worth publishing.

Note what those figures do not establish. They are not evidence of a rigged simulator. They are approximately what you would expect from a five per cent daily liquidation rule applied to leveraged retail traders, which is precisely the point: a house that writes the stop does not need to cheat.

The difference, and it runs the other way

The sympathy in this comparison usually flows to the bucket-shop customer, as the victim of a cruder age. That has it backwards.

The bucket shop existed because the real market was shut. In 1900 the round lot on the New York Stock Exchange was a hundred shares, commissions were written for customers the house already knew, and there was no product on the exchange sized for a man with two hundred dollars in his pocket. The bucket shop was the only door open to him, and it was a fraudulent one. What it sold — dishonestly, but genuinely in demand — was access. That demand is why the shops stood on every corner and why the exchanges could not shame them out of business.

The door is open now. A Micro E-mini S&P contract is five dollars an index point: a tenth of the E-mini and a fiftieth of the old full-size contract. The tick is $1.25. CME maintenance margin runs near $1,500 overnight and retail futures brokers quote day-trade margin in the tens of dollars. Micros exist across the board — gold, crude, the currency pairs, the ten-year note. Every fill is a real fill, cleared, against a real counterparty, in a market whose open interest the CFTC publishes every Friday.

So the 2026 customer pays four hundred dollars for a simulation of a market he could have entered for less, and entered for real. That is not the bucket shop’s trade. The bucket shop sold access to a market that was closed. The funded-account firm sells something else, and it is worth naming exactly: it sells the proposition that what stands between this customer and a living is capital rather than skill. That is a far more comfortable thing to be told, and people pay more for comfort than they ever paid for access.

Three things should be said plainly against the easy version of this argument. The product is legal. The fee is a disclosed and capped loss — the customer cannot lose more than he paid, which the 1900 customer could not say, and which is a real difference rather than a technical one. And firms do pay out; money moves, in both directions. The objection here is not that the industry is uniformly fraudulent. It is that an honest description of the product is “an examination fee for a job that does not exist”, and that almost nobody sells it that way.

What actually closed the shops

Here is the part that gets left out of the comparison, and it is the useful part.

The shops were not legislated away first. The direct route was tried and it failed: the New York Stock Exchange cut the telegraphic tickers in 1889 to starve them of quotations and abandoned the attempt within days, because the same feed was what its legitimate customers were paying for. You cannot switch off the price for the bucket shop without switching it off for the broker.

What worked was doing it selectively, and that took the Supreme Court. On 8 May 1905, in Board of Trade of the City of Chicago v. Christie Grain & Stock Co., Holmes held that the Board’s continuously collected quotations were the Board’s property. “It stands like a trade secret.” “The plaintiff has the right to keep the work which it has done, or paid for doing, to itself.” And it did not forfeit that right by distributing the quotations under contract to people in confidential relation with it. Which meant the Board could sell the feed to brokers, deny it to bucket shops, and have a court behind the distinction.

The statutes came afterwards and buried what was already dying — state anti-bucket-shop acts through the 1900s and 1910s, then New York’s Martin Act in 1921, passed against the bucket shops and the boiler rooms and far better known now for everything it has been used for since.

The sequence, then: the input was cut, the business became unworkable, and the law arrived at the funeral. Not the other way round.

The regulator’s turn, and how it went

Worth being exact about this, because it does not end the way an argument of this shape usually wants it to.

On 29 August 2023 the CFTC sued Traders Global Group, trading as My Forex Funds, alleging it had taken more than $310m in fees from over 135,000 customers while presenting itself as a prop firm that funded successful traders — when in the Commission’s telling it was a simulated retail platform that profited from their losses and interfered with execution to make sure the winners failed. A judge froze the assets the same day and appointed a receiver. It was the industry’s landmark case.

It collapsed. On 13 May 2025 Judge Edward Kiel dismissed the complaint with prejudice and sanctioned the Commission, following a Special Master’s finding that the agency had made false representations to the court and acted “willfully and in bad faith”, including holding back for more than six months an Ontario Securities Commission email confirming that a C$31.5m transfer the CFTC had painted as suspicious was a tax payment. The court ordered the CFTC to pay the defendants over $3m in fees and costs.

Read that as narrowly as it deserves: the case was thrown out for the agency’s conduct, not on a finding that the underlying business was sound. But the distinction is cold comfort to anyone waiting for a verdict, because there is none — and after a defeat of that shape, the next enforcement action will be a while coming. The CFTC, the FCA, ESMA and ASIC are all reported to be examining the funded-account model; in July the chairman of CySEC said retail prop trading is not an ESMA priority. That is the state of the law. Nothing decided, and nobody in a hurry.

Where the chokepoint went

Meanwhile, between 2024 and this year, somewhere between eighty and a hundred firms went dark.

Not one of them was closed by a court. The trigger was MetaQuotes. On 2 February 2024 it terminated True Forex Funds’ MT4 and MT5 licences without notice, and worked down the grey-label chain through that year: brokers who had been sub-licensing the platform onward to prop firms were made to stop, and Blackbull was pushed into cutting off Funding Pips. Firms that had built an entire product on a platform they did not license directly found out they had not owned a business, only a dependency. True Forex Funds shut in May 2024 owing about $1.2m to roughly three hundred traders. The Funded Trader and SurgeTrader went the same year. The survivors migrated to cTrader and Match-Trader, which is what you would expect — a chokepoint with an alternative is a toll rather than a chokepoint.

But the shape of the event is 1905’s exactly. The party that owned the input decided who could have it, and an industry reorganised itself around that decision inside a year, while the regulators were still consulting.

The lesson is not that a software vendor makes a good regulator. It is that in this business the entity holding the price feed has always been able to do in one quarter what a commission cannot do in three years — and that the feed is now held by platform vendors and liquidity providers, whose interests align with the public’s occasionally, by coincidence, and never by obligation.

Not in the record

One last thing, which is this desk’s own concern rather than the law’s.

Every Friday the CFTC publishes the Commitments of Traders report. For each futures market it sets out who is long and who is short, split into commercial hedgers, managed money, other reportables, and the non-reportable positions — the small traders, everyone below the reporting threshold. It is the only public register of who actually holds what. This page is built on it.

The funded-account population does not appear in it. Not among the non-reportables, not inside a broker’s aggregate, not anywhere. Those tickets never become contracts, so they never become open interest, so they leave no trace in the record at all. Exactly the same was true of the man at the counter in 1906: his money moved the shop’s book, and never once touched the tape he was reading it from.

That is the cleanest way to put the whole comparison, and it requires accusing nobody of anything. You can establish whether someone is in a market by looking for them in the register of who holds it. If they are not there, they are not in the market. They are in a commercial arrangement with somebody who is watching one.