Your funded account is a commercial contract with a private company, not a brokerage account holding segregated customer money. Everything traders find surprising about prop firms follows from that one fact. The rules can change because the agreement you accepted says they can, and profit you have earned but not withdrawn is a claim against a company rather than property held for your benefit. That is not an accusation. It stays true when the firm on the other side is well capitalized, well run and entirely honest.
A prop firm evaluation or funded program is a private service contract with a company, so both the firm's rule changes and its solvency sit outside the trader's control and outside the fund-segregation regime that covers money held at a registered futures commission merchant. Archived terms and an independent trade log strengthen a trader's position against a firm that is still operating and still paying, and do nothing against one that has stopped. The only defense that works against insolvency is not holding a large unwithdrawn balance when it arrives.
What kind of arrangement is a funded account?
It is a service agreement. You pay a private company for access to a program with performance conditions attached, and the company promises a share of measured profit if you meet them. The company writes the conditions, publishes them, and in nearly every case reserves the right to change them. That is ordinary commercial contracting, disclosed at purchase.
The confusion comes from the interface. A funded program has a login, a balance, an equity curve and a margin figure, so traders assume the balance carries the same legal character as a balance at a broker. It does not. The reason is which body of rules applies, not anyone's integrity. For the structural background on the account itself, our piece on whether funded accounts are real money or simulated covers that ground. This article covers the risk that follows from it.
Where does the money actually sit?
At a registered futures commission merchant, customer money is walled off from the firm's own money by rule. The CFTC states it plainly: all customer funds for trading on designated contract markets must be kept apart, or segregated, from the FCM's own funds. Those accounts must be titled for the benefit of the FCM's customers, and acknowledgements must preclude a bank or clearinghouse from recognizing a right of offset against the account for the FCM's own debts.
The scope argument is worth stating precisely. Under 17 CFR 1.3, futures customer funds means all money, securities and property received by a futures commission merchant or by a derivatives clearing organization from, for, or on behalf of futures customers to margin, guarantee or secure contracts for future delivery. An evaluation fee is revenue received by a company for a service. It is not money received by an FCM to margin a position, so the segregation rules are not the rules that describe it. That is a statement about which regulation applies, not about anyone's conduct.
Follow two payments and the difference becomes physical rather than legal. Payment one: a trader wires margin to a registered FCM. It lands in a bank account titled for the benefit of that FCM's customers. An acknowledgement letter sits between that bank and the FCM's creditors, blocking any right of offset against the balance, and the money moves onward to the clearing organization as margin. Payment two: a trader pays an evaluation fee by card. It lands in the program company's general operating account, where it mixes with subscription revenue and flows out again as payroll, advertising spend and payouts to other traders. Same trader, two payments, two completely different destinations. There is no wall anywhere on the second path, because no rule puts one there.
Both things can be true at once, and usually are. Where a firm routes live orders through a registered FCM, segregation attaches to that underlying brokerage relationship and the money held there. The program's own terms, meaning targets, consistency rules, payout schedule and profit split, stay private contract between you and the program company. Two relationships, two rule sets, one dashboard login. The shared login is why traders merge them mentally.
Do not overstate the protected side either. The same CFTC page continues: segregated customer funds have a bankruptcy preference if the FCM fails, and to the extent those funds are not sufficient to pay customer claims, the remainder participates pro rata in distributions to unsecured creditors. Segregation is a preference and an insulation from the firm's own creditors. It is not a guarantee of full recovery.
The four things that actually change
Four different events get lumped together as "the firm changed the rules," and they differ in kind, not degree. Only two touch money you already consider yours.
| Event | What it alters | Realistic recourse |
|---|---|---|
| Rules change mid-evaluation | The target you are currently working toward | Amendment clause governs; possible fee dispute inside the billing window |
| Terms change at renewal or reset | The deal you re-enter, going forward | Strong: do not re-enter. Read before paying again |
| Payout policy changes | Money already earned but not yet withdrawn | Weak. Contractual argument only, and only against a solvent firm |
| Firm ceases operations | Everything, including accrued balance | Effectively none. This is an insolvency question, not a rules question |
Consistency rules deserve separate treatment. They are the change most likely to land mid-evaluation and the one traders least expect. A consistency rule caps the share of total profit that any single day or single trade may represent. Adding one does not only constrain future trading. It can invalidate progress already recorded.
Worked example, with illustrative figures you should replace with your firm's actual numbers. A trader runs a $50,000 evaluation with a $3,000 profit target and reaches exactly $3,000, with the best single day contributing $1,500. The firm then caps any single day at 30 percent of total profit. Check the existing book: $1,500 / $3,000 = 50 percent, over the cap. To bring that day inside 30 percent, total profit must reach at least $1,500 / 0.30 = $5,000. Verify: $1,500 / $5,000 = 30 percent exactly. Additional profit required is $5,000 - $3,000 = $2,000, which is 66.7 percent more than the original target, and it has to be earned without any new day breaching the cap either. The trader hit the stated target and is now two thirds short of a target that did not exist on the day the fee was paid.
Permitted-instrument and permitted-copying changes are the sneakiest category, because they turn a compliant setup into a breach with no action by the trader at all. A rule that bans copying between accounts, drops an instrument from the permitted list, or forbids holding through a session boundary can make a running configuration non-compliant overnight.
Re-read the restricted-activity section on every renewal and after any notice of a terms update. A rule change can breach a configuration you set up months ago and have not touched since.
What recourse actually exists?
Start with the amendment clause, the single most underread paragraph in any program agreement. Most terms reserve the right to modify with notice by posting, often with continued use constituting acceptance. Find it before you pay and note three things: whether notice is individual, such as an email, or constructive, such as a posted update you are expected to check; whether changes apply to accounts already open or only to new purchases; and whether any grandfathering period is stated. A clause that applies changes to open accounts on posting is not hidden or improper. It is disclosed, and accepting it should be a decision you make knowingly rather than a discovery in month three.
Payment-method recourse is real, narrow and time-boxed. Under Regulation Z, a written billing error notice must be received by the creditor no later than 60 days after the creditor transmitted the first periodic statement that reflects the alleged billing error, and the listed errors include an extension of credit for services not delivered to the consumer as agreed. The creditor must acknowledge in writing within 30 days and resolve within two complete billing cycles and in no event later than 90 days, and you need not pay the disputed amount or face adverse credit reporting on it while the dispute is open.
Whether a mid-evaluation rule change fits that description is contested, not settled. The issuer decides, and the amendment clause you accepted at checkout is the firm's answer to any claim that the service was not delivered as agreed. Treat the dispute channel as a possibility worth preserving, not an entitlement.
Put a clock on it. A trader pays a $500 evaluation fee on day 0, and the statement first reflecting the charge is transmitted on day 10. The Regulation Z deadline is day 10 + 60 = day 70. The firm changes the profit target on day 45, leaving 70 - 45 = 25 days to file. Card network dispute windows are commonly cited at around 120 days from the transaction, which would leave 120 - 45 = 75 days, but network rules are not law, they vary by dispute type and they change, so confirm the current window with your issuer. Either way the window is short, and it recovers a $500 fee, not a profit balance.
Regulation Z's billing error procedure applies to open-end consumer credit under US law, which is what a consumer credit card account is. Debit cards, bank transfers and cryptocurrency payments carry materially weaker protection or none, and readers outside the US fall under different regimes entirely. Paying an evaluation fee by credit card preserves a recourse channel that a crypto payment does not. That is worth something even if you never use it.
Documentation improves your position against a solvent company and does nothing against an insolvent one.
Every defensive measure below wins arguments with firms that are still operating and still paying. None of them recovers money from a company that has stopped. And a contract dispute of any real size is a question for a lawyer in the relevant jurisdiction, not for a blog.
The defensive practice that works
Archive the terms in force on the day the payment clears. The useful artifact is not a screenshot of the marketing page. It is a timestamped capture of the actual terms document and the rule set as displayed in your dashboard, saved alongside the payment receipt and the payment confirmation email, which carries a third-party timestamp. A capture with an independent timestamp is materially stronger than an image you could have taken on any day.
What it protects against is a version divergence you cannot see from inside the dashboard. On day 0 you pay and archive rulebook v1. On day 30 the firm publishes v2, and the dashboard now serves v2 to everyone, including on the pages describing your already-open account. On day 45 a dispute arises: you hold v1, the live page shows v2, and the amendment clause is the single edge that decides which version governs an account opened under the first one. Without the day 0 archive there is no v1 to point at, and the disagreement is about memory.
Keep your own trade log. The firm's dashboard is a derived record, not a primary one. Every execution originates at the platform or bridge that placed it, so the fill exists upstream of whatever the dashboard renders. Export platform statements on a fixed schedule, weekly is enough, and your P&L history survives the dashboard becoming unavailable. Traders running a copier get this almost free, because the copier's execution log is a second, independently produced record of what was sent and what filled. That is a reason to retain platform-level logs rather than truncate them, and it is directly useful when a payout is questioned, a scenario covered in our piece on why prop firm payouts get denied.
Verify claims rather than assuming registration. The CFTC advises traders to verify that an entity is properly registered before working with it and points to the NFA BASIC database for registration, disciplinary or regulatory history and financial information. Be careful with the inference: many program companies carry out no activity requiring registration, so absence from BASIC is not evidence of anything. The defensible version is narrow. When a firm claims a registration, an NFA membership or an affiliated FCM, BASIC is where you confirm that specific claim against the record. Note the asymmetry too. NFA Compliance Rule 2-29 governs communications with the public and promotional material, and NFA guidance addresses performance results that are hypothetical or simulated rather than actual, but those rules bind NFA Members, which is exactly why they do not automatically reach a company that is not one.
What leaving a balance actually costs
An accrued but unwithdrawn balance is an unsecured claim against a private company. If that company becomes unable to pay, the balance ranks with its other unsecured obligations and carries none of the bankruptcy preference the CFTC describes for segregated futures customer funds. Compounding inside a program grows the size of that claim month by month while the probability of ever needing to collect on it stays exactly whatever it was.
Price it. Assume a trader clears $2,000 per month of withdrawable profit for six months. All figures here are illustrative; substitute your own.
| Path | Balance held at month end | Peak exposure | Total exposure (dollar-months) |
|---|---|---|---|
| A: leave it in to compound | 2,000 / 4,000 / 6,000 / 8,000 / 10,000 / 12,000 | $12,000 | $42,000 |
| B: withdraw every month | 2,000 each month, reset to 0 | $2,000 | $12,000 |
Path A sums to 2,000 + 4,000 + 6,000 + 8,000 + 10,000 + 12,000 = $42,000 dollar-months. Path B is 6 x $2,000 = $12,000. Peak exposure ratio: $12,000 / $2,000 = 6.0x. Total exposure ratio: $42,000 / $12,000 = 3.5x. Identical trading, identical profit, six times the peak unsecured claim and three and a half times the time-weighted exposure.
Now apply a failure. Assume, purely as an illustration, that a liquidation returns 10 cents on the dollar to unsecured claimants. Path A: $12,000 x 0.10 = $1,200 recovered and $10,800 lost. Path B, failing at the worst moment for that path with a full month accrued: $2,000 x 0.10 = $200 recovered and $1,800 lost. The difference is $10,800 - $1,800 = $9,000, on identical earnings. The 10 percent figure is a chosen illustration and not data. Real recoveries from an insolvent private company vary enormously and are frequently zero. The ratio between the paths is the point, not the recovery rate.
Frequent withdrawals cost money, so price that too. Same $2,000 monthly accrual, an illustrative $30 payout fee (verify your firm's actual fee before relying on this), and profit treated as accruing evenly across each month. Monthly withdrawals over six months: 6 x $30 = $180 in fees, with the balance sawtoothing between $0 and $2,000, so average exposure is $1,000. Quarterly withdrawals: 2 x $30 = $60 in fees, with the balance running $0 to $6,000 across each leg, so average exposure is $3,000. Going monthly costs $180 - $60 = $120 more over six months, or $20 per month, and it removes $3,000 - $1,000 = $2,000 of average exposure. That is $20 / $2,000 = 1.0 percent per month, or 12 percent per year, of the exposure removed. If you would not pay a 12 percent annual premium to insure that balance, withdraw less often. If you would, withdraw monthly. That is a decision, not a guess.
The honest counterweight: where a firm's scaling plan keys off account balance or a sustained-profit threshold, frequent withdrawals genuinely slow scaling and can reset progress toward a larger account. Frequent payouts are correct when your exposure is large relative to your ability to absorb losing it, and wrong when the scaling curve you are climbing is worth more than the exposure you are carrying. Check how your firm's scaling plan treats withdrawals before deciding.
Does spreading across firms help?
It cuts the worst case and leaves the average untouched, which is not what most traders think it does. Let B be the total balance you hold across program companies and p the annual probability that any one of them becomes unable to pay, assumed equal and independent. Concentrated at one firm, expected loss is p x B and the worst single-event loss is B, or 100 percent. Split across three firms at B/3 each, expected loss is 3 x p x (B/3) = p x B, identical, while the worst single-event loss falls to B/3, or 33.3 percent.
The cost side is concrete. Three firms taking one payout each per month at an illustrative $30 fee is 3 x $30 = $90 per month against $30 for one firm, an extra $60 per month or $720 per year, before counting three evaluation fees, three rulebooks, three daily loss limits and three sets of restricted-activity clauses to track. Every extra rulebook is an extra way to breach a rule you forgot applied on that particular account. That operational tax is why spreading is not free. Our guide to trading with multiple prop firms covers how to run that spread without drowning in it.
Where a copier is not the answer
A trade copier does nothing about counterparty risk. It replicates orders. If a firm tightens a consistency rule or stops paying, the copier is irrelevant to both outcomes, and no amount of execution quality changes a contractual exposure. Anyone selling copying as a diversification answer to firm risk is selling the wrong tool for the problem.
Worse, copying between accounts or between firms is among the activities most often restricted by program terms, so wiring accounts together can create a violation where none would otherwise exist, and a terms update can ban it retroactively for a setup that has run compliantly for months. Copying across accounts is a scaling and execution tool, and its correct precondition is that every destination account's terms permit it in writing. Check the permitted-copy-trading clause per firm, per renewal, before you connect anything.
What a copier does contribute is a record. It produces an independent execution log of what was sent and what filled, upstream of any firm's dashboard, which is the single artifact most worth having when a payout is questioned or a platform goes dark. That is a documentation benefit, not a protection benefit, and the distinction is the whole article. Documentation wins arguments with solvent firms. Against insolvency, the only thing that works is not having a large balance sitting there when it arrives.
Frequently asked questions
Is money in a prop firm funded account protected like a brokerage account?
No. Money at a registered futures commission merchant must be segregated from the firm's own funds and held in accounts titled for the benefit of customers, and the CFTC states that segregated funds carry a bankruptcy preference if the FCM fails. A profit balance accrued inside an evaluation or funded program is a claim against a private company under a service agreement, so the segregation rules are not the rules describing it. Where a firm routes live orders through a registered FCM, those protections can attach to that underlying brokerage relationship while the program's own terms stay private contract.
Can a prop firm change its rules in the middle of an evaluation?
In most cases yes, because the agreement you accepted contains an amendment clause reserving that right. Read that clause before paying and note whether notice is individual or by posting, whether changes apply to already-open accounts or only new purchases, and whether any grandfathering is stated. A clause that applies changes to open accounts is disclosed and standard, so price it knowingly rather than discover it later.
What happens to my balance if a prop firm shuts down?
An unwithdrawn balance becomes an unsecured claim ranked with the company's other unsecured obligations, and realistic recovery is often little or nothing. It does not carry the bankruptcy preference that segregated futures customer funds carry at a registered FCM. Documentation, archived terms and trade logs improve your position against a solvent firm and change nothing against an insolvent one.
Can I chargeback a prop firm evaluation fee?
There is a real but narrow window, and it recovers a fee rather than a profit balance. Under Regulation Z, a written billing error notice must reach the creditor no later than 60 days after the creditor transmitted the first periodic statement reflecting the alleged error, and the listed errors include services not delivered to the consumer as agreed. Whether a rule change fits that description is contested, the issuer decides, and the amendment clause you accepted is the firm's counterargument. Card network dispute windows are commonly cited at around 120 days from the transaction, but network rules are not law and change, so confirm the current window with your issuer.
How often should I withdraw profits from a funded account?
Often enough that your unwithdrawn balance stays small relative to what you can afford to lose entirely. On an illustrative $2,000 monthly accrual with a $30 payout fee, switching from quarterly to monthly withdrawals costs $20 per month and removes $2,000 of average exposure, which prices the reduction at 12 percent per year. The counterweight is that some scaling plans key off balance growth or a sustained-profit threshold, so frequent withdrawals can genuinely slow progress toward a larger account.
Does trading with multiple prop firms reduce risk?
It reduces the worst single-event loss and leaves expected loss unchanged. Splitting a balance equally across three independent counterparties with the same failure probability cuts the worst case from 100 percent of the balance to about 33 percent, but expected loss stays p x B either way. The cost is multiplied rulebooks, evaluation fees, payout fees and daily loss limits to track, and every extra rulebook is an extra way to breach a rule you forgot applied.
How do I verify whether a prop firm is registered?
Use the NFA BASIC database to check any registration, membership or affiliated FCM the firm specifically claims, which is where the CFTC directs traders for registration, disciplinary or regulatory history and financial information. Be careful with the inference in the other direction: many program companies conduct no activity requiring registration, so absence from BASIC is not evidence of a problem. Verifying a claim is a different exercise from expecting every counterparty to be registered.
What should I document when I pay for an evaluation?
Capture the actual terms document and the dashboard rule set as displayed on the day the payment clears, saved with the payment receipt and the confirmation email, since that email carries a third-party timestamp. A marketing page screenshot is close to worthless; the terms page and rules panel are what an amendment dispute turns on. A capture with an independent timestamp is materially stronger than an image you could have created any day.
Why do consistency rules cause so many mid-evaluation failures?
Because adding one does not just constrain future trading, it can invalidate results already recorded. A trader at exactly $3,000 profit whose best day contributed $1,500 is at 50 percent concentration, so a new 30 percent cap forces total profit up to $1,500 divided by 0.30, which is $5,000, an extra $2,000 or 66.7 percent above the original target. That additional profit also has to be earned without any new day breaching the cap.
Does a trade copier protect me from prop firm risk?
No. A copier replicates orders and has no effect on whether a firm changes a rule or pays a withdrawal. Copying between accounts or firms is also among the activities most often restricted by program terms, so wiring accounts together can create a violation, and a terms update can ban it retroactively for a setup that was compliant last month. Confirm the permitted-copy-trading clause per firm in writing before connecting anything.