A $100 Account or a Prop Firm Challenge? 400,000 Simulated Years and a Gold Scalper on 23.0 Million Real Ticks

00The short answer

• With an edge, a cheap challenge paid far more than a $100 account. In a simulated year, the trader with a good edge made +$208 on average on a $100 account and +$1,562 after the fee on one $35 challenge; the trader with a small edge made +$47 against +$337.
• Without an edge, most years end behind on both, in different ways. The break-even trader ended the year ahead 46.7% of the time on the $100 account and 25.1% of the time on one challenge, whose average was positive (+$63) only because a few large payouts outweigh many lost fees. The trader with a slightly negative edge lost $38 on average on the $100 account, $20 on one challenge, and $91 buying a new challenge after every failure.
• The $100 account's real handicap is the smallest trade. On gold, 0.01 lot with a $10 stop risks $10, a tenth of the account. With that minimum, 77.0% of the break-even traders and 94.8% of the negative-edge traders fell below the smallest trade within the year, and even the good-edge trader lost the account 14.6% of the time.
• The gold scalper never passed the $5,000 challenge. At 0.01 lot, out at +$5 or after 40 seconds, 15 trades a day, not one of 80,000 simulated runs (40,000 periods on the 148 trading days of real ticks, with and without a commission) passed both phases, at any accuracy we tried: $5 scalps on 0.01 lot are too small for a $400 target. The fee was lost every time.
• On the $100 account, the scalper had to be right about 60% of the time just to break even after the spread (63.5% with a $0.07 commission). Trading at random, the account fell below the margin of one 0.01 lot in every simulation; right 65% of the time, it ended ahead in 99.8%.
• No stop means no limit. The worst 40-second scalp in the data lost $81.71 on 0.01 lot, 82% of a $100 account: a buy on Friday 4 September at 08:29:41 New York time, 19 seconds before 08:30, when US economic figures are released.

01Who we are, and why you can check these numbers

RIP Backtest is an independent site run by one retail futures trader. We test the trading strategies, indicators and claims that people share, under one fixed protocol: the rules are written down and dated before any data is run, costs are charged on every trade, and every version we try is reported, including the ones that fail. So far that is 404 strategy variants in 59 tests, and 2 of them have passed every gate.

We are not a prop firm or a broker, and we have nothing to sell you. There are no affiliate links on this site, no firm pays us, and this article names no firm. Most pages that compare small accounts with prop firm challenges are written by someone who earns from one of the two. Every number here comes from a simulation whose scenarios we fixed in writing before running it; the specification, the code, the checks and the complete results are linked at the end. Read more about the site and our testing protocol.

02The two ways to start with a little money

These are generic rules, not any firm's. Fees, payout splits, payout schedules and loss limits differ by firm and change often; the fee is illustrative ($5,000 two-step challenges were offered for roughly $29 to $36 in October 2026), and section 05 shows how the results change with it. The funded account has the same loss limits as the challenge, and it ends at the first breach. We compare three paths: A, the $100 account; B, one challenge; and B-R, buying a new challenge after every failure or breach, as long as the year lasts (at most 12 attempts).

03How we simulated it

We wrote down every scenario before running anything (sa_spec.md, frozen 16 October 2026). The study has two parts.

Part 1: traders by edge

• Four traders, as in our prop firm challenge guide: every trade wins 1.5 times the amount risked or loses it, after costs; the traders win 38% (a slightly negative result, −0.05 R a trade), 40% (break-even), 42% (a small edge, +0.05 R) and 46% (a good edge, +0.15 R) of their trades. Three trades a day, 250 trading days.
• The $100 account risks 1% of the balance a trade, but never less than the smallest position allows: $1 (A1: 0.01 lot of EUR/USD with a 10-pip stop) or $10 (A2: 0.01 lot of gold with a $10 stop). Below that minimum, the account stops.
• The challenge risks $25 a trade (0.5% of $5,000).
• 400,000 simulated years, 100,000 per trader. Each year's trades are drawn once and go through every path, so the paths are compared on the same luck.

Part 2: a gold scalper on real ticks

• The data: the gold tick history (every bid and ask) of a retail broker's MetaTrader 5 server, 23.0 million ticks. The server keeps no ticks before 24 March 2026, so Part 2 covers 24 March to 16 October 2026: 148 trading days, not a full year. In London hours, the spread averaged $0.22 per 0.01 lot (median $0.19), highest in April ($0.32) and lowest in October ($0.17); the longest stretch without a tick was 25 seconds, and no day was dropped (data check).
• The trades: 15 a day at random seconds between 08:00 and 17:00 London time, one in each 36-minute part of the day, never overlapping. 0.01 lot is one ounce: $1 for every $1 the price moves. Buys fill at the ask and sells at the bid, so the real spread of that second is paid. A trade closes at +$5 when the price gets there, otherwise after 40 seconds. No stop. A variant adds a $0.07 commission a trade ($7 a lot).
• The traders: one picks the direction at random. Three others know where the price will be 40 seconds later on 10%, 20% or 30% of their trades and flip a coin on the rest, so they are right about 55%, 60% and 65% of the time.
• The $100 account stops when the balance falls below the margin of one 0.01 lot at 1:500 leverage (the gold price divided by 500: $7.92 to $9.78 in this period). The challenge is the same $5,000 one, with the same 0.01-lot trades.
• 40,000 simulated periods, 10,000 per trader, each with new random entry times and directions on the same real days.

Checks before any result: the $100 account's average matches the textbook formula; the challenge loop is the one of our prop firm challenge guide, unchanged, and reproduces its published pass rates exactly; 24 hand-made cases (payouts, resets, breaches, the end of the year, the fills, the +$5 and 40-second exits, the London clock); and a second implementation, written separately, agreed with the engine on every number of every cell and on all 4,350,847 possible entries in the tick data (self-test result).

04Part 1: on average, one challenge beat the $100 account for every trader

The result of the year is the money made minus the money put in: for the $100 account, the final balance minus the $100; for the challenge, the payouts minus the fees. Average result, and the share of simulated years that ended ahead:

Share of simulated years that ended ahead, 100,000 years per trader. “$100, gold minimum” is the $100 account that cannot risk less than $10 a trade (section 06).

Read across the rows and, for every trader with an edge, the challenge paid several times more than the $100 account. Read the “ahead” columns and you see what the averages hide: the challenge's average is carried by fewer, larger wins. For the break-even trader, 74.9% of single challenges never paid anything and the median result was the lost fee; on the $100 account, the same trader's median year ended at $97.49, with 90% of years between $44.90 and $164.99. For the trader with a slightly negative edge, one challenge lost less on average than the $100 account; buying again lost the most.

The $100 account in detail

The challenge in detail

Passing is not the end of it: 12.3% of even the good-edge traders broke a rule in the funded account during the year (it then closes), as did 21.9% of the break-even ones.

05Why the challenge comes out ahead on paper

• Size. The challenge risks $25 a trade; the $100 account at 1% risks $1. For a $35 fee, the challenge trades 25 times bigger, and an edge earns in proportion to size.
• The loss is capped at the fee. When a funded account breaks a rule it is closed, and the trader owes nothing; payouts already made are kept. The year can never be worse than the fees paid, and it can be much better.
• The fee at which one attempt breaks even is the average payout per attempt: $15 for the negative-edge trader, $98 for the break-even one, $372 and $1,597 for the small and good edges. Every dollar of fee above it is a dollar lost on average, per attempt; that is how our results scale with the fee.

And why most buyers still lose: the fee is certain and the payout is not. A trader whose results are slightly negative after costs, where many traders end up once costs are counted, did not come out ahead on a single challenge in 94.1% of years. Buying a new challenge after every failure raised the chance of a payout from 5.9% to 13.3%, but cost 3.5 attempts and $122 in fees on average, for $31 paid out: the average loss grew from $20 to $91, worse than the $38 lost on the $100 account.

06The $100 account's handicap: the smallest trade

On a $100 account, 1% is $1. That pays for a 10-pip stop on 0.01 lot of EUR/USD, but not for gold: 0.01 lot of gold with a $10 stop risks $10, a tenth of the account. So we ran the same traders with a $10 minimum (A2): the account risks $10 a trade until 1% of the balance is more than that, which takes a balance above $1,000.

The result is in the tables above. With a $1 minimum, few accounts were stopped: 4.6% for the negative-edge trader, 0.3% or fewer for the others. With a $10 minimum, 94.8% of the negative-edge accounts and 77.0% of the break-even ones fell below the smallest trade within the year, and so did 49.9% of the small-edge accounts. The traders with an edge gained from the bigger bets on average (small edge: $322 against $147; good edge: $1,143 against $308), but the median small-edge account ended the year at $50, and even the good-edge trader lost the whole account 14.6% of the time.

This is the honest case for a challenge on small money: a $5,000 account can risk 0.5% on gold, a $100 account cannot. It is also the honest case against trading gold with $100: the market forces a size that turns a small edge into a coin flip on the whole account.

07Part 2: what a 40-second gold scalp costs

Over all 8,701,694 possible trades in the data (every second with a tick that can start a trade in London hours, long and short):

• The spread is the cost: $0.22 a trade on average at the entry second. A trade in a random direction lost $0.23 on average: about the spread, because over 40 seconds the price is as likely to go one way as the other.
• $5 is far in 40 seconds: the target was reached in 1.1% of all possible trades. Almost every trade closes at 40 seconds, so its result is the 40-second move minus the spread.
• Without a stop, the rare trade is the dangerous one: one possible trade in 119 lost $5 or more, and the worst lost $81.71 in 40 seconds, on Friday 4 September just before 08:30 New York time. A trader who scalps through the release times of US data meets these seconds sooner or later.

08The $100 account: you have to be right about 60% of the time

148 trading days, 15 trades a day, 0.01 lot, starting with $100:

“Stopped” means the balance fell below the margin of one 0.01 lot, so no further trade was possible. The break-even point follows from two averages over all the possible trades: a random side lost $0.23 a trade, the side the price was about to move to made $0.87. Between the two, the scalper breaks even when right 60.3% of the time with the spread only, and 63.5% with a $0.07 commission. Being right 55% of the time, which would be a remarkable skill over 40 seconds, still lost the account in 99.3% of the periods.

Three small things decide this: the spread is paid on every one of the 2,220 trades of the period, the +$5 target is almost never reached, and the $100 account has only about $90 of room before it cannot open a trade at all.

09The $5,000 challenge at 0.01 lot: out of reach

Phase 1 asks for $400. The best scalper in the study, right 65% of the time, made $0.10 a trade on average, $1.55 a day: about 258 trading days to reach $400, far more than the 148 days of data. The same scalper reached phase 2 by the end of the data in 1.0% of the periods. The others lost money, slowly: the random scalper broke a loss limit in 53.7% of the periods (98.6% with the commission). At 0.01 lot, the $5,000 challenge was a $35 fee for a target this scalper could not reach in seven months.

To make the target reachable, the scalper would have to trade bigger. Five times the size multiplies every trade by five, the losses too, including the $81.71 one, which at 0.05 lot is more than the $250 daily limit. We did not simulate other sizes; the specification fixed 0.01 lot on both accounts.

A rule that matters more than the numbers: some firms do not pay profits from trades held under one or two minutes, or forbid “tick scalping” and trading around news releases. Under such a rule this style is not paid at all, whatever its results. Read the firm's rules on holding times before you buy.

10What the simulation leaves out, on both sides

• The challenge's fine print. Holding-time and news rules, consistency rules, payout reviews in which profits can be refused for a rule breach, accounts that are simulated rather than live, rule changes, and the firm's own ability to pay. Our challenge pays every payout it owes.
• The $100 account's fine print. Real fills can be worse than the quoted price, most of all around news; brokers close positions when equity falls below a share of the margin; and some brokers have minimum deposits or charge fees for small accounts. Our $100 account fills at the quote and simply stops below the margin.
• What the $100 account has that the numbers do not count: no rules, no review, no firm between you and your money, and every dollar of profit paid at once. What the challenge has: other people's money, which is worth a lot to a trader with a real edge and a size problem.

11So which one should you choose?

1. Measure your edge first, on a demo account. It is free. In our numbers, without an edge both paths end the year behind most of the time, and with a slightly negative edge, buying challenge after challenge lost the most.
2. With a measured edge, a fixed small risk and a stop, a cheap challenge whose rules fit your style paid several times more than a $100 account. Read the payout and holding-time rules before you buy, and price the attempt by its odds, not by its best case: you can estimate them for your own rules with our prop firm challenge simulator.
3. If you scalp gold at 0.01 lot with small targets, a $5,000 challenge is not reachable at that size, and on your own $100 account you need to be right about 60% of the time after the spread to break even, 63% with a commission. Check that over hundreds of trades on a demo account before you trust it.
4. Use a stop. The worst 40-second gold scalp in seven months of ticks lost $81.71 on 0.01 lot.
5. Do not buy challenge after challenge to get lucky. With a slightly negative edge, buying again turned an average loss of $20 into $91.
6. On $100, trade where 1% is possible: 0.01 lot of EUR/USD with a 10-pip stop is about $1; 0.01 lot of gold with a $10 stop is a tenth of the account.

12What these simulations cannot tell you

• Real trades are not fixed bets. Part 1's trades win 1.5 R or lose 1 R; real results vary in size, cluster in streaks and depend on the market.
• Seven months are not a year. Part 2 covers the 148 trading days the broker's server keeps; another period, another broker's spread or another time of day would give other numbers.
• Skill is modelled crudely. Our skilled scalpers know where the price will be in 40 seconds on a fixed share of trades. Real skill, if it exists, is uneven and changes with the market.
• Behaviour is left out. The simulated trader never revenge-trades, never doubles the size and never stops after a bad morning.
• Your edge. The traders are assumptions. The study shows what an edge of a given size does on each path; it cannot tell you what yours is.

This is educational content, not financial advice, and not a recommendation to open an account or to buy or avoid any challenge.

FAQCommon questions

Is a $100 account better than a prop firm challenge?

It depends on how you trade. For traders with a stop and a fixed risk, one $35 challenge on a $5,000 account did better on average than a $100 account for all four edges we modelled, because it trades 25 times bigger and the loss is capped at the fee; but without an edge the usual result was the lost fee (the break-even trader got no payout from 74.9% of challenges). For a gold scalper at 0.01 lot it was the other way round: the challenge was never passed, and only the $100 account could make money, for a scalper right more than about 60% of the time.

Can you trade gold with a $100 account?

You can open trades, but not at a sensible risk: 0.01 lot of gold with a $10 stop risks a tenth of $100. With that minimum, 77.0% of break-even traders and 94.8% of slightly negative ones fell below the smallest trade within a year in our simulation.

Can you pass a prop firm challenge by scalping gold?

Not at 0.01 lot on a $5,000 challenge in our test: none of 80,000 simulated runs of 40-second scalps on real gold ticks passed both phases in 148 trading days. Even a scalper right 65% of the time averaged about $0.10 a trade, too little for the $400 target. Some firms also refuse profits from trades held under a minute or two.

What accuracy do you need to scalp gold?

For 40-second scalps with a $5 target and no stop, about 60.3% of trades on the right side just to break even after the spread, and 63.5% with a $0.07 commission, on 148 days of real gold ticks. In a random direction, the average trade lost $0.23, about the spread.

How much does a $5,000 prop firm challenge cost?

Two-step $5,000 challenges were offered for roughly $29 to $36 in October 2026; prices, rules and payout terms vary by firm and change often. In our simulation one attempt breaks even on average only if the fee is below the average payout per attempt: $15 for a slightly negative trader and $98 for a break-even one.

Should you buy another challenge after failing one?

Not without an edge. For a trader slightly negative after costs, buying a new $35 challenge after every failure meant 3.5 attempts and $122 in fees on average in a year, for $31 paid out.

Should a beginner start with a small live account or a prop firm?

Neither first: measure your edge on a demo account, which is free. With a measured edge after costs and a stop, a cheap challenge paid several times more than a $100 account in our simulation; without one, both lost money in most years.

13Check our arithmetic

Code provenance: our own code. The challenge rules are generic types described in this article, not any firm's rules.

14Corrections

21 Oct 2026, navigation: added a link to the prop firm challenge simulator in section 11. No result changed.

+More tests and guides

• How to pass a prop firm challenge
• Prop firm challenge pass rate
• Why do forex traders lose money
• Ttm squeeze strategy
• Free prop firm challenge calculator
• What all the backtests found so far
• The backtesting protocol
• How to backtest a trading strategy without fooling yourself
• Backtesting glossary