Is Day Trading Gambling? The House Edge, Computed

For a trader without a measured edge, yes: after costs, a day trade is a losing bet with a house edge (the average loss per trade as a share of the amount risked) of 8.25% on an E-mini S&P 500 contract with a 4-point stop and an illustrative $16.50 of round-turn costs, against 5.26% at American roulette. The qualification is skill: roulette's odds are fixed for every player, while fewer than 1% of day traders in a 15-year study of Taiwan's stock market earned returns after fees that held up from year to year.

Key takeaway

Without a proven edge, every day trade loses its costs on average, which makes it a wager whose house edge depends on the stop: at an illustrative $16.50 of costs per E-mini round turn, any stop tighter than about 6.27 points gives worse odds than American roulette's 5.26%. What roulette lacks is persistent skill, which Barber, Lee, Liu and Odean (2014) found in fewer than 1% of Taiwan's day traders. Your own results after costs, over a sample fixed in advance, decide which one you are doing.

What is the difference between trading and gambling?

The difference lies in expected value and in whether skill persists, not in the act of risking money. The American Psychiatric Association describes gambling as taking a risk on an unclear outcome by "risking something of value (usually money) with the intent of trying to win an item of higher value." Every speculative trade fits that sentence.

Reviewing the evidence, Arthur, Williams and Delfabbro (2016) concluded: "Investment is distinct from gambling, but speculation and gambling have conceptual overlap and a strong empirical relationship." They rate the expected return of gambling as usually negative and of speculation as mixed and highly variable. Day trading is speculation, so its sign depends on the trader.

The house sets roulette's odds with 38 pockets (37 on a European wheel); other traders and your costs set a trade's. Roulette is negative-sum by design and futures zero-sum before costs, yet both charge a cost per play: a fixed 5.26% or 2.70% at the wheel, costs divided by risk for a trade (2.06% to 16.5% below).

What sets them apart is skill, which persisted for fewer than 1% of Taiwan's day traders, and the regulator: roulette answers to a state board such as the Nevada Gaming Control Board, while price-based futures such as the E-mini S&P 500 answer to the CFTC, under a Commodity Exchange Act that lets the CFTC bar event contracts involving "gaming". Event contracts are where that line is contested in court as of October 2026: in April 2026 the CFTC sued Arizona, Connecticut and Illinois over their actions against CFTC-registered exchanges that list event contracts.

Both columns charge every player a cost per play, so a trade without skill sits in the same family as a spin. The row that can change the verdict is skill that persists, and in a 15-year Taiwan study it did for fewer than 1% of day traders.

What is the house edge of a day trade?

For a trader without skill, the house edge of a day trade is its round-turn cost divided by the amount risked: 8.25% for one E-mini S&P 500 (ES) contract with a 4-point stop and $16.50 of costs. The model treats price as a fair game over the trade's life, so before costs any bracket expects to break even, and after costs it expects to lose the friction (commissions, exchange fees and the spread you cross).

Step by step, with ES worth $50 per point and $12.50 per 0.25-point tick under CME Group's contract specification:

Risk 4 points: R, the amount risked in R-multiple terms, is 4 x $50 = $200. Set the target at 4 points too.
Add costs: $4.00 of commissions and exchange fees plus one $12.50 tick of spread, the $16.50 round turn from our scalping comparison. Topstep's help center lists $3.80 for an ES round turn in its Live Funded Account as of October 2026.
A winner nets $200 - $16.50 = $183.50; a loser costs $200 + $16.50 = $216.50.
At a 50% win rate, the average trade is ($183.50 - $216.50) / 2 = -$16.50, or 8.25% of the $200 at risk.
Breaking even takes $216.50 / $400 = 54.13% winners, more than the 52.38% a sports bettor needs at -110 (staking $110 to win $100).

The same arithmetic at four stop widths, ranked against edges from game rules and Nevada's published slot hold:

Bet	Risked or staked	House edge	Break-even win rate
ES day trade, 2-point stop	$100	16.5%	58.25%
ES day trade, 4-point stop	$200	8.25%	54.13%
Nevada slot machines, 12 months to August 2026	$1 played	7.14%	n/a
American roulette (38 pockets)	$1	5.26%	n/a
Sports bet at -110	$110	4.55%	52.38%
ES day trade, 8-point stop	$400	4.13%	52.06%
European roulette (37 pockets)	$1	2.70%	n/a
ES day trade, 16-point stop	$800	2.06%	51.03%

Computed by Phoenix Technologies; method: ES edge = $16.50 / (stop x $50), break-even = (R + $16.50) / 2R; roulette and -110 from game rules (2/38, 1/37, 5/110); slots from the Nevada Gaming Control Board's Monthly Revenue Report (win divided by amount played). ES costs are illustrative.

Each doubling of the stop halves the edge, because the $16.50 stays fixed while the risk doubles, so on one baseline the ES bars step down from 16.5% to 2.06% past every casino bar. With a 4-point stop, a trader without skill pays more per dollar risked than a Nevada slot player pays per dollar played.

Each doubling of the stop halves the edge, because the $16.50 stays fixed while the risk doubles. With a 4-point stop, a trader without skill pays more per dollar risked than a Nevada slot player pays per dollar played.

The claim that trading has no house edge is half right: no casino banks your bet. The $4.00 goes to the broker, exchange and clearing house and the $12.50 to the trader whose quote you hit, but to you the cost is the same, charged on every trade, win or lose.

Why does a tighter stop make the odds worse?

A tighter stop raises the house edge because costs are fixed per trade while the amount at risk shrinks: at $16.50 per round turn, any ES stop tighter than about 6.27 points (25 ticks) gives a zero-skill trade a bigger edge against it than American roulette. The cost is 1.32 ticks whatever the stop.

Each crossover on the chart below is $16.50 divided by a game's edge, converted at $50 a point. Below 4.62 points a trade without skill is worse than Nevada's 7.14% slot hold, below 6.27 points worse than American roulette, below 7.26 points worse than the -110 bet, and below 12.21 points worse than European roulette's 2.70%.

The fixed cost is 1.32 ticks whatever the stop, so the edge falls as the stop widens. Every stop in the red zone gives a trade without skill worse odds than a double-zero roulette wheel.

A stop loss does not separate trading from gambling: it caps each loss but leaves the sign of the average unchanged. A bigger profit target does not fix the odds either, because with no skill a wider target is reached less often in exact proportion, so the gross expectation stays at zero and the $16.50 still comes off every trade.

A stop caps the loss on one trade; it does not change the sign of the average.

How fast does a zero-skill day trader fall behind?

Faster than a roulette player: a zero-skill ES trader with a 4-point stop first has a 90% chance of being net negative after 231 trades, while a player betting red on an American wheel reaches 90% after 593 spins. At 3 trades a day, 231 trades is 77 trading days.

After 10 plays, 62.30% of zero-skill ES traders are behind against 44.32% of red bettors, whose tied records count as not behind. After 100 the shares are 81.59% and 66.57%, after 500 they are 96.67% and 87.15%, and after 1,000 they are 99.57% and 94.89%. Because results come in whole wins and losses, each share wobbles near 90%: ES stays at or above 90% from 254 trades, roulette from 629 spins.

The ES trader falls behind faster because the edge is larger relative to the swings, 0.0825 standard deviations per trade against 0.0527. Tied roulette records count as not behind, which is why the wheel starts lower at 10 spins.

Method: exact binomial, each ES trade a 50% chance of +$183.50 or -$216.50, each spin an 18-in-38 chance. With a symmetric bracket the standard deviation per trade equals R, so the 8.25% edge is 0.0825 standard deviations per trade against 0.0527 for the wheel, which is why the trader falls behind faster. With a 2-point stop, 90% are behind after 55 to 66 trades.

The per-trade edge also understates what you lose from the money you bring. Nevada's roulette tables reported an 18.44% win percent over the 12 months to August 2026, against a double-zero wheel's 5.26% edge per bet; gaming mathematician Michael Shackleford explains on Wizard of Odds that hold "depends on how long the player will sit at the table and the same money circulates back and forth."

Trading recycles money through frequency. At 3 trades a day for 20 trading days, $16.50 of costs adds up to $990 a month: 4.95R at a $200 risk, or 49.5% of an illustrative $2,000 drawdown buffer on a $50K account. At 20 trades a day it is $6,600, 330% of the buffer.

Is day trading luck or skill?

For most day traders the results look like luck minus costs, but a small group shows skill that persists. Studying day traders in Taiwan from 1992 to 2006, Barber, Lee, Liu and Odean (2014) concluded: "Less than 1% of the day trader population is able to predictably and reliably earn positive abnormal returns net of fees."

Ranked on one year's returns, Taiwan's top 500 day traders earned 61.3 basis points (hundredths of a percent) a day before fees and 37.9 after in the next year; the bottom-ranked earned -11.5 and -28.9. As a group they still lost: their 2017 working paper with Zhang found that "the aggregate performance of day traders net of fees is negative in each of the 15 years that we study."

These were stock traders, not futures traders as one addiction-information page says: the 2017 paper defines day traders as those "who buy and sell the same stock within a day." In futures, a Brazilian study found that 97% of those who began day trading mini-Ibovespa futures in 2013 to 2015 and persisted at least 300 days lost money; the other loss-rate studies have their own article. The often-repeated line that 80% of day traders fail within a year fits the 2014 Taiwan paper only loosely: about 20% of its day traders earned profits net of fees in a typical year, so about 80% did not, a measure of one year's results rather than of quitting or ruin.

Skill also has to be shown, and costs slow the proof. A trader who truly wins 55% of 1:1 trades needs 396 trades to show it before costs (the method is in how many trades it takes to prove an edge). After $16.50 of costs that trader loses $6.50 a trade with a 2-point stop and needs 12,931 trades with a 4-point stop, about 17 years at 3 a day, but 629 with a 16-point stop.

Do people day trade for the thrill?

Some trading is done for the thrill: when a big lottery jackpot offers the same thrill elsewhere, individual investors trade less. Gao and Lin (2015) tested this in the Review of Financial Studies: "We hypothesize that individual investors treat trading as a fun and exciting gambling activity, implying substitution between this activity and alternative gambling opportunities."

On days when Taiwan's lottery jackpot exceeded NT$500 million, trading volume fell 5.2% to 9.1% in stocks individual investors prefer and 6.8% to 8.6% in lottery-like stocks. Buying and selling fell by statistically indistinguishable amounts, which fits people stepping away from the activity rather than changing their view of the market.

Among 1,000 German brokerage clients, Dorn and Sengmueller (2009) found that investors who report enjoying investing or gambling turn over their portfolios at twice the rate of their peers, and Dorn, Dorn and Sengmueller (2015) found the jackpot effect again in US, Californian and German data. We found no such study of futures traders.

Is day trading addictive like gambling?

Frequent trading goes with problem gambling in several studies, but we found no public study that measures how common gambling problems are among US futures day traders. A South Australian study, reported in the 2016 review above, found problem gambling in 7.6% of day traders against 1.7% of people who did not day trade, about 4.5 times the rate: one region, and an association rather than proof of cause.

Among 795 US adults who gamble and hold stocks, Mosenhauer, Newall and Walasek (2021) found higher problem-gambling scores in those who traded more often, after allowing for financial literacy, overconfidence and demographics: "This study provides support for the hypothesis that behavioral addiction to gambling-like activities is associated with frequent stock market trading." Because the sample was drawn from gamblers, it cannot say what share of all traders is affected.

Losses rarely stop experienced day traders, even though unprofitable traders quit more often than profitable ones overall. Among Taiwanese day traders with 50 or more days of experience, the previously unprofitable had a 95.3% probability of trading again within 12 months, against 96.4% for the previously profitable, while more than 75% of all day traders quit within two years.

The APA's criteria for gambling disorder require at least four symptoms in the past year, and one is chasing: "After losing money gambling, often returning to get even." This is research context, not a diagnosis; revenge trading is chasing in trading form, and trading psychology for funded accounts covers the habits that interrupt it.

How can you tell if your trading is gambling?

Ask five questions: a "no" to any of them means your trading currently works like a bet, whatever analysis sits behind it.

Do you know your average result per trade after costs, over a sample size you fixed in advance, recorded in R in a trading journal?
Is your risk per trade set before you enter?
Do you stop for the day at a preset loss?
After a loss, can you stop instead of trading on to get even?
Would you still trade this way if it were boring?

Every exit lands in the same box, so one weak answer is enough. The first question is the slowest to answer honestly, because a small edge after costs needs thousands of trades to show.

Then apply the numbers:

If you cannot answer the first question with a number, assume the house edge in the table above applies to every trade you take.
If your ES stop is tighter than about 6.27 points at $16.50 of costs, a trade without skill gives up more than American roulette's 5.26%: widen the stop and size down, or cut costs, before adding trades.
If you take 20 trades a day at $16.50 each, your costs are about $6,600 a month per contract; without a measured edge, that is also your expected loss.
If questions 3 to 5 worry you, the National Council on Problem Gambling runs the National Problem Gambling Helpline (1-800-MY-RESET) by call, text or online chat, and its help page offers a self-assessment.

Go deeper

Frequently asked questions

Is scalping more like gambling than swing trading?

For a trader without an edge, yes, because scalping pairs the tightest stops with the most trades. At $16.50 of costs, a 2-point ES stop gives up 16.5% of the risk on every trade, and 90% of zero-skill traders are behind after 55 to 66 trades. A scalper with a measured edge is not gambling, but has the least room for error.

Is the stock market a casino?

Not for long-term investors. Arthur, Williams and Delfabbro (2016) class investment as distinct from gambling, and Dorn, Dorn and Sengmueller (2015) found that trading tied to long-term savings, such as trading in retirement accounts, does not respond to lottery jackpots. Short-term speculation in individual stocks is where the overlap with gambling shows up.

Who gambles most in the stock market?

Studying US individual investors, Kumar (2009) found that their propensity to gamble shows up in their portfolios and that "state lotteries and lottery-type stocks attract very similar socioeconomic clienteles." Demand for lottery-type stocks also rises in downturns, as lottery demand does. Because those stocks underperform, he found the resulting underperformance greater among low-income investors who excessively overweight them.

Is buying a prop firm challenge gambling?

Without a measured edge it works like one: the fee buys a chance of passing under the firm's rules, and a zero-skill trader still pays costs on every trade inside it. With an edge measured after costs, the fee is a priced attempt you can judge on your own numbers. No prop firm publishes what share of challenge buyers trade for entertainment rather than income.

Does using limit orders remove the house edge?

Only part of it. A limit order can save the $12.50 tick of spread that a market order crosses, but it takes adverse selection instead, tending to fill when price is about to move against it. The $4.00 of commissions and exchange fees stays on every round turn, so a zero-skill trader's edge shrinks without turning positive.

What house edge do real day traders actually pay?

No public dataset measures it. Every edge figure here is computed from an illustrative $16.50 of costs and chosen stop widths. Your own figure is your average cost per round turn divided by your average risk per trade, and your fills and stop distances already record both.

Can you make a living from day trading?

Very few people do on the available evidence. Studying Brazilian index-futures day traders, Chague, De-Losso and Giovannetti wrote: "We show that it is virtually impossible for individuals to compete with HFTs and day trade for a living, contrary to what course providers claim." HFTs are high-frequency trading firms, and 97% of the persistent traders in their sample lost money.

Does copying your trades to more accounts change the odds?

No. A trade copier repeats each trade on every connected account, so it multiplies the average result per trade, house edge included, and each account pays its own costs. Copying belongs after an edge is measured, never as a way to find one.

Is options trading gambling?

Short-term options speculation shows the same lottery link as stock trading. Gao and Lin (2015) found that large Taiwan lottery jackpots go with less trading in options with high sensitivity to volatility, the same substitution they saw in stocks. Whether a given options trade is a bet still depends on its expected value after costs.