Copier documentation covers linking, sizing and latency. It skips the first decision: which account originates the trade. That choice sets who holds the off switch on your signal, whether your measured edge survives real liquidity, and how much of your own capital sits outside the funded programs. Three options are realistic. All three fail, in different directions. Picking a master is picking the failure you can least afford.
The master account must be an account the trader personally owns, and for a trader running several funded accounts at once, a personal live account or a dedicated demo account is structurally safer than designating one funded account as master. A funded master puts the prop firm's risk engine inside the signal source, so a daily loss limit or trailing drawdown breach can flatten and freeze the whole fleet at once. A demo master removes that single point of failure but replaces it with simulated fills that overstate the strategy's edge in one consistent direction.
What does the master account actually decide?
Four things: fill realism, who has authority to disable the signal, how finely exits can be expressed, and where your capital sits. No option scores well on all four.
| Master type | Fill realism | Who can disable it | Size and granularity | Primary failure mode |
|---|---|---|---|---|
| Funded account | Real fills, real queue, real slippage | The firm's risk engine, without warning | Capped by the firm's plan and scaling rules | Signal stops mid position |
| Demo account | Modelled, and the model varies by platform | Nobody, unless the broker retires the account | Free, you choose it deliberately | Signal is systematically optimistic |
| Personal live account | Real fills, real queue, real slippage | You and your broker only | Free, but funded by your own capital | Costs real money to operate |
Column three has no comfortable answer: whoever can disable the master disables every follower behind it, and that authority outranks any feature in the copier above it. Read this next to the practical limits on running a copier across multiple prop firm accounts.
Option A: a funded account as master
A funded master gives the most honest signal on the most fragile foundation. Every fill is a real fill against real liquidity, so prices in the master's trade log come from the same distribution as the followers' prices, and the decision-maker faces genuine consequences. Then the structural problem arrives: the firm's risk engine lives inside your signal source.
Daily loss limit reached. Trailing drawdown breached. Position over the plan cap. Consistency flagged at payout review. Account frozen pending review. Any one of those can force-flatten or lock the master while open positions sit across the fleet. Risk engines fire during volatility, exactly when open exposure is largest and re-entry matters most. The fleet does not simply stop receiving new signals. It receives a flatten instruction nobody wanted, then finds the origin disabled so nothing can be re-entered.
Picture the chain. A box labelled firm risk engine holds the rules: daily loss limit, trailing drawdown, account review. One arrow runs from that box into the funded master, and it is the only arrow in the picture you do not control. From the master an arrow runs into the copier, and from the copier arrows fan out to follower 1 through follower N. Break the master node and the whole tree goes dark in one step.
The freeze is worse than the silence. Topstep's TopstepX help centre states that when copy trading is enabled you will not be able to create, close or adjust orders in a follower account, and that the restriction covers all components including the DOM, the order component and the chart. Master dead, followers holding inherited positions, manual intervention blocked until the link is broken. Behaviour differs by firm and plan, so check your rulebook, but the shape repeats wherever the copier runs inside the firm's own platform.
The second cost is granularity, and almost nobody mentions it. Futures contracts are integers. A master holding one contract has exactly two expressible states, in and flat. It cannot signal a partial exit, so no follower can scale out, however large it is and however sophisticated the copier is. The same Topstep page adds a matching constraint from the other side: follower accounts must have a greater than or equal margin and max position size compared to the lead account in order to follow.
The master's contract count is the resolution of the entire fleet's exit management.
A funded master held to a small position cap by an early scaling tier does not just cap your size. It caps your strategy's exit vocabulary.
Option B: a demo account as master
A demo master is the only option no third party can switch off. No drawdown, no daily loss limit, no contract cap, no consistency rule, no forced liquidation, no payout processing window interrupting the link. You also set the master's equity and contract count deliberately instead of inheriting whatever the firm permits, so you choose the sizing ratio rather than discovering it.
The cost is a fill engine instead of a market. The master's entries and exits come from a model of how orders would have filled, and that model's quality is the whole question. Careful simulators track bid and ask volume, traded volume and queue position before granting a fill. Careless ones fill on touch, full size, every time. Assume nothing about yours: run the strategy on the simulator and on a small live account in parallel for a month, then compare fill prices trade by trade.
Why does demo optimism compound instead of averaging out?
Because a naive fill model errs in the same direction every time. A simulator that fills when price touches your level fills you on 100% of the occasions price trades there. Live, you are behind a queue at that price, and on the fast prints that matter most you are not filled at all. The simulator grants entries live traders never obtained, and full size where live gets partial. Random noise averages toward zero over 400 trades. A touch-fill assumption accumulates over 400 trades.
Two mechanisms produce the gap, both documented at venue and platform level. CME Group's order type documentation explains that a market order with protection is not guaranteed at the touch: for a buy order, protection points are added to the current best offer to calculate a protection price limit, and if the entire order cannot be filled within the protected range immediately, the unfilled quantity remains in the order book as a limit order at the limit of the protected range. MQL5's order properties documentation defines the fill policies a live order runs under: Fill or Kill executes in the specified volume only and will not execute at all if that amount is unavailable in the market, while Immediate or Cancel executes the volume available and cancels the remainder. A demo master that always receives full size has never exercised either branch.
Follow one order intent down two routes. Route A goes demo master, simulator fill engine, filled at the touch, full size, instantly, position marked green. Route B goes live follower, exchange match engine, price and time priority queue, and then splits again: partially filled at a worse average price, or left resting at the protection price limit while price walks away. The two routes reconverge at a divergence bar comparing master fill price against follower fill price. The gap never points in the follower's favour.
Price that gap in dollars, with an illustrative assumption you should replace with your own measured fills. A scalping strategy measures an average edge of 2 ticks per round turn on ES. At $12.50 per tick that is 2 x $12.50 = $25.00 per contract per round turn before commissions. Assume live followers give up 1 tick on entry and 1 tick on exit, also 2 ticks, also $25.00. Net live edge is $25.00 - $25.00 = $0.00 before commissions, so negative after them. At 20 round turns a day across 20 sessions, 400 round turns produce $10,000 of demo edge per contract and $10,000 of slippage against it. On a ten contract follower, 400 x 10 x $25.00 = $100,000 of projected profit meets the same $100,000 of slippage. The master's equity curve climbs the whole time. It is not lying. It is measuring a market that does not exist.
Demo master optimism never shows up in the master's equity curve, because the master is not lying about its own fills. It appears only when you compare master fill price against follower fill price trade by trade, so build that reconciliation before you scale a demo originated strategy.
Option C: a personal live account as master
A personal live account at a retail broker is the only option that gets real fills without handing anyone else the off switch. Master and follower slippage are drawn from the same distribution, so the reconciliation problem above disappears. No prop firm rule engine has authority over the origin. No payout processing pause interrupts the link. And you size it for the granularity the strategy needs rather than the size a scaling plan permits, so a master trading three or five contracts makes staged exits expressible across every follower.
The costs are real. Capital in the master is capital not spent on evaluation fees or held as a buffer. The master takes losses no prop firm absorbs. And if the copier sizes proportionally off master equity, a small personal master paired with large funded followers produces a large multiplier that amplifies every mistake as faithfully as every good trade.
One constraint gets ignored until it bites: the master should trade the same instrument family and ideally the same venue as the followers. A master on a retail CFD feed originating signals for futures followers inserts a symbol mapping and tick size translation layer between origin and destination, and every translation layer is a place where a rounding decision quietly changes follower size.
Master size is also a behavioural dial. One live master trading 1 ES contract feeding five funded followers at 10 contracts each carries fleet exposure of 1 + (5 x 10) = 51 contracts, of which the master holds 1 / 51 = 1.96%. The decision-maker absorbs under two percent of the risk while making 100% of the decisions. On a 10 point adverse move the master loses 1 x 10 x $50 = $500 while the followers lose 50 x 10 x $50 = $25,000 combined. Size the master at 5 contracts and its share becomes 5 / 55 = 9.09%, a $2,500 loss on the same move.
How does master equity set every follower's size?
Under proportional sizing, master equity is a live input to every order the fleet sends. MetaTrader's signal documentation is explicit: the volume is multiplied by the ratio of subscriber's and provider's balances considering deposit currency and allowable deposit load specified in the platform settings. The same page adds a second stage after that: subscriber and provider leverages are taken into account as well, which can shrink the copied volume again. It documents no fixed volume mode. Choosing a master is therefore choosing a denominator, which is why the position sizing method your copier uses has to be settled in the same sitting as the master itself.
Take a master with 5,000 USD equity and a follower with 50,000 USD. The ratio is 50,000 / 5,000 = 10.0, so a 1 contract master trade becomes 1 x 10.0 = 10 contracts on the follower. Now the error case. The trader intends 1 contract and sends 3. The follower receives 3 x 10.0 = 30. On ES at $50 per point, a 4 point adverse move costs the master 1 x 4 x $50 = $200 on the intended trade and 3 x 4 x $50 = $600 on the mistake, a difference of $400. The follower loses 10 x 4 x $50 = $2,000 intended and 30 x 4 x $50 = $6,000 on the mistake, a difference of $4,000. The extra damage is exactly ten times larger on the follower, because the ratio applies to errors with the same fidelity it applies to edge.
A very small master paired with much larger followers maximises the multiplier on every mistake while minimising the contract granularity available for exits. If you are choosing master equity freely, choose it large enough that its own sizing errors are not magnified into the fleet.
The reverse configuration fails differently. Master equity 50,000 USD, follower equity 5,000 USD, ratio 5,000 / 50,000 = 0.1. A 1 contract master trade becomes 0.1 contracts. Futures are integers, so it floors to 0 and the follower takes nothing. A 3 contract signal becomes 0.3, which also floors to 0. The master must trade 10 contracts before the follower receives 1. An oversized master produces a follower that silently ignores small signals, then jumps from 0 to 1 contract in a single step. On fractional lot venues such as forex and CFD the symptom differs but the root cause is the same: the ratio basis, not the copier.
Granularity arithmetic runs the same way. A 1 contract master at a 10.0 ratio gives a follower holding 10 contracts exactly two states, so ten contracts behave as one indivisible block. Give the master 3 contracts and closing 1 of 3 is 1 / 3 = 33.3333%, which on a 30 contract follower closes 0.333333 x 30 = 10 contracts and leaves 20. Change one variable: a follower holding 25 contracts receives 0.333333 x 25 = 8.3333 contracts, which floors to 8 and leaves 17 open where strict proportionality wanted 25 - 8.3333 = 16.6667. That follower carries 0.3333 contracts more risk than the master's profile implies, on that leg, and again on the next one.
Follower side settings move the denominator too. In MetaTrader terms, a provider balance of 15,000 USD and a subscriber balance of 40,000 USD with deposit load set to 50% gives usable funds of 40,000 x 0.50 = 20,000 USD and a ratio of 20,000 / 15,000 = 1.3333, or 133.33%. MetaQuotes' own worked example runs the other way: 4,800 USD of usable subscriber funds against a 10,000 USD provider copies 1.00 lot as 4,800 / 10,000 = 0.48 lots. The computed figure is then rounded to the instrument's volume step, so 1.33333 lands on 1.33 at a 0.01 step and 1.3 at a 0.1 step. Verify the volume step and the rounding behaviour in your own platform build before relying on either.
Whose account is allowed to originate the signal?
Yours, and only yours, whichever of the three types you pick. Firms that permit copy trading at all generally permit it only across accounts owned and traded by the same individual. Copying a third party's signals, subscribing to a signal service, or letting anyone else originate trades on your funded account is commonly prohibited and classified as account management or group trading, whatever software carries the order. Rulebooks in this space frequently require that the account be traded only by the individual named on it and bar direction from any other person, service, bot or mirroring tool, but the exact wording, the scope and the penalty are firm-specific and they change. Read the rulebook for the plan you actually hold rather than a summary of it.
This is a detection question as much as a policy one, because firms reconcile fill timestamps and order sequences across accounts. The mechanics behind how prop firms detect copy trading tell you more than the wording of any single clause.
Nothing here is legal advice, and none of it substitutes for a lawyer who has read your actual agreement. In US futures, the framework governing advising or directing other people's trading is commodity trading advisor registration under the Commodity Exchange Act, administered by the CFTC with NFA membership, and the available exemptions turn on facts a qualified professional has to weigh. Trading only accounts you personally own is structurally a different activity from directing accounts belonging to other people, but where a specific setup lands is a question for counsel. Bring three. Am I directing accounts belonging to other people, or only my own? Is my advice tailored to their circumstances, and how many people receive it? Does my prop firm agreement permit any of this, and what happens to a payout if it does not? Ask before you run signals into someone else's account, not after.
So which master should you actually pick?
If you run several funded accounts at once, designating one of them as master is the weakest of the three choices. It hands a third party the ability to disable your entire fleet's signal source at the moment of maximum open risk, and no copier can prevent that, including this one. A firm's risk engine will act on the account it governs. The only mitigation is architectural: keep the firm's risk engine out of your signal origin. A personal live account, or a demo account used strictly as a signal origin, does exactly that.
Now the counterpoint, because it decides the choice for many readers. A demo master does not delete risk. It relocates risk from mechanical to behavioural. The person clicking the button on an account that cannot lose money behaves differently: larger size, later exits, more tolerance for a losing position, more revenge entries after a bad print. A funded master at least imposes on the decision-maker the same consequences the followers will face. If you know from experience that you trade differently on simulated money, a demo master is the wrong answer for you specifically, and no structural argument overrides that.
The honest limit of this article: it assumes a fleet. A trader running one funded account has no fleet, no single point of failure and no ratio problem, and for that trader a copier is usually the wrong tool. It adds an account to manage, a mapping layer to maintain and a new failure surface, in exchange for duplicating a position that could have been sized correctly to begin with. Add the copier when account count makes manual execution the binding constraint, not before.
Once you do have a fleet, the decision reduces to one question with three honest answers. If a signal that stops mid position is your worst outcome, do not use a funded master. If a signal that flatters its own fills is your worst outcome, do not use a demo master. If capital tied up outside the funded programs is your worst outcome, do not use a personal live master. Pick the failure you can absorb, size the master large enough to express your exits, and reconcile master fills against follower fills every month so you learn early which failure you actually chose.
Frequently asked questions
Should my master account be one of my funded accounts?
Only if you can afford the signal to stop without warning, because a funded master puts the prop firm's risk engine inside your signal source. A daily loss limit, trailing drawdown breach, position cap violation or account review can force-flatten or lock the master while every follower holds an open position. If you run several funded accounts at once, a personal live account or a dedicated demo account removes that single point of failure.
Can I use a demo account as the master for copy trading?
Yes, and it is the only master type nobody else can switch off, but its fills are modelled rather than matched. If the simulator fills on price touch, the master records entries live followers never obtained, so follower results sit below master results by a consistent amount per trade. Run the strategy on your simulator and on a small live account in parallel before trusting the simulator as a signal origin.
Is copy trading between my own prop firm accounts allowed?
Firms that permit copy trading at all generally permit it only across accounts owned and traded by the same individual, so the master must be your own account. Rules differ by firm, by product line and by plan, and they change without much notice, so read the current rulebook for the plan you hold rather than a summary of it. Some firms prohibit copying entirely on certain account types.
Can I copy someone else's signals into my funded account?
Usually no, because most futures prop firms classify third party signals as account management or group trading regardless of the software used. Firm rulebooks commonly require that the account be traded only by the individual listed on it and bar direction from any other person, service, bot or mirroring tool, though the exact wording and the penalty vary by firm. Breaching such a clause is typically grounds for denying a payout or closing the account, so check your own agreement.
How big should my master account be relative to the followers?
Large enough that its own sizing errors are not amplified into the fleet, and large enough in contract count to express the exits your strategy needs. A 5,000 USD master with a 50,000 USD follower produces a ratio of 10.0, so a mis-clicked extra 2 contracts becomes 20 extra contracts on that follower. A small master paired with much larger followers is the most fragile common configuration.
Why did my follower account skip a trade the master took?
Most often the proportional ratio produced a fractional contract that floored to zero. A 50,000 USD master with a 5,000 USD follower gives a ratio of 0.1, so a 1 contract master trade computes to 0.1 contracts and futures cannot be fractional. The master would need 10 contracts before that follower receives 1, which is a sizing basis problem rather than a copier fault.
Why can't my followers scale out of a position?
Because the master cannot express a partial exit at its current contract count. A master holding 1 contract has exactly two states, in and flat, so every follower inherits those two states no matter how many contracts it holds. Give the master 3 contracts and a 1 of 3 close maps to 33.3333% across the fleet, which is how staged exits become available.
Can a master on a forex or CFD account feed futures followers?
It can, but it inserts a symbol mapping and tick size translation layer between origin and destination, and every translation layer is a place where rounding quietly changes follower size. Keeping the master in the same instrument family and ideally the same venue as the followers removes an entire class of silent sizing errors. If you must cross venues, reconcile follower size against intended size after every session for the first few weeks.
What happens to followers if the master gets flattened by the firm?
Followers receive the flatten and then have no source to re-enter from, which is the worst case during volatility. On some firm platforms the followers are also blocked from creating, closing or adjusting orders while copying is enabled, so the positions cannot be managed by hand until the link is broken. No copier can prevent a firm's risk engine from acting on the account it governs, which is why the master should sit outside that authority.
Do I need a copier if I only run one funded account?
No, and adding one usually makes the setup worse. With a single account there is no fleet, no single point of failure and no sizing ratio to manage, so a copier adds a second account, a mapping layer and a new failure surface in exchange for nothing. Add a copier when the number of accounts makes manual execution the binding constraint.