Under 17 CFR 4.41(b), printed in the 2023 Code of Federal Regulations, any simulated or hypothetical commodity trading record shown to the public carries a disclaimer placed in immediate proximity to it, stating that the results do not represent actual trading, that they were designed with the benefit of hindsight, and that market factors such as lack of liquidity are not fully reflected. That is the regulator's summary of what a paper record measures: the arithmetic of a rule set, never a fill. The condition that decides what one is worth is whether every cost and delay it omitted can be named and priced.

The limitation is already written into the rules

The wording exists because simulated records are systematically flattering. CFTC Regulation 4.41 requires the disclaimer to be prominently disclosed and placed next to the numbers, not in a footnote. The National Futures Association goes further on presentation: its Interpretive Notice 9025, issued under Compliance Rule 2-29, requires the disclaimer to be displayed as prominently as the hypothetical results themselves and printed in a type size at least as large, to precede the hypothetical data when the actual results shown are under 12 months old, and it bars hypothetical results altogether for a program with three months or more of actual trading results, with narrow exceptions.

The securities side reached the same place by a different route. The SEC's Investment Adviser Marketing rule, Rule 206(4)-1, was adopted on 22 December 2020 and took effect on 4 May 2021. The final adopting release, Release No. IA-5653, sets the compliance date eighteen months after the effective date, 4 November 2022, and governs hypothetical performance, including model and back-tested figures, in adviser advertisements. A simulator on a phone is not an advertisement, and the arithmetic of what it leaves out is identical.

Three fills a simulator cannot promise

A simulated fill is a price the platform chose: the quote at the signal, the next bar's open, or the midpoint of the bid and ask. None of those is a transaction. FINRA's page on order types states that a market order generally executes at or near the current bid or ask during regular hours of 9:30 a.m. to 4 p.m. Eastern, and that the main risk is not getting the price that was quoted. The same page states a limit order executes only at the limit price or better, so the order can fail to execute at all, and that a stop order converts to a market order once triggered, inheriting that execution uncertainty. A simulator that records every signal as a fill has assumed away the second and third of those.

Queue position, latency between signal and submission, and orders that execute in pieces are the other three omissions, each one checkable against the platform's own documentation.

The costs that sit outside the result

Zero commission is not zero cost. FINRA's page on fees and commissions states plainly that free trading does not mean free investing, and lists the charges that remain: markups and spreads when a firm sells from inventory, sales loads, surrender charges, asset-based advisory fees and annual operating expenses. That page also prices the ongoing kind — a fee of only 1 percent per year for 20 years on a $100,000 investment earning 4 percent costs $28,000.

Scroll sideways to see the full table →

CostCharged onAppears in a simulator?
SpreadEvery entry and exitOnly if fills use the bid and ask, not the midpoint
SlippageFast or thin marketsOnly with an explicit slippage model
Exchange and regulatory feesPer executionRarely
Platform or data subscriptionPer monthNever

Example: $480 simulated, $135 after friction

An example with stated assumptions, and a mechanism rather than a strategy. A three-month simulation on a $10,000 balance logs 40 round trips of 50 shares each, filling both sides at the bid-ask midpoint with no commission, and reports a gain of $480, or 4.8 percent. Four frictions get added back at assumed rates, no taxes modeled.

  • Slippage at $0.05 per share beyond the quote, two sides per round trip: $0.05 x 50 x 2 x 40 = $200.
  • Spread at a constant $0.04, one full spread per round trip relative to midpoint fills: $0.04 x 50 x 40 = $80.
  • Exchange and regulatory fees at $0.50 per round trip: $0.50 x 40 = $20.
  • Platform subscription at $15 a month for three months: $45.

Total: $345. The adjusted result is $480 - $345 = $135, or 1.35 percent of the $10,000 balance. Friction removed 71.9 percent of the simulated gain, and the rates above are assumptions, not measurements. This is the same subtraction that separates a gross result from a net one.

What the platform does not enforce

A simulated account cannot be lost, and brokers say so in their marketing: TradeStation's simulated trading page advertises unlimited paper dollars and balance resets. Margin is the other absentee. Real accounts run against FINRA's margin rules, and those rules are changing: Regulatory Notice 26-10, published 20 April 2026, replaces the $25,000 pattern day trader minimum equity requirement and the day-trade count with an intraday margin deficit standard under new Rule 4210(d)(2), effective 4 June 2026 with an 18-month phase-in ending 20 October 2027. A simulator applying neither the old threshold nor the new standard has removed a constraint that decides whether a real position stays open, which is why the transition itself is worth reading before comparing the two records.

What the number is evidence of

Inside its stated assumptions, a simulation shows how a fixed rule set behaves over a price sequence: how often it trades, how large its gains and losses run, how long positions stay open. Win count answers none of that. The result is evidence about the rules; the four frictions above are the price of turning it into a claim about money.

Source references

Primary sources used for this guide. Follow the links for the original scope, definitions and conditions.