What Is Copy Trading and how does it work?

Open any copy trading platform and the first thing you see is a list sorted by return. Rank 1 at the top. Rank 2 below it. A percentage next to each name, a small risk badge, a button that says Copy.

That list is the product. Not the traders on it — the list itself. And the sorting rule that builds it is the single most important thing to understand about what is copy trading, because ranking people by realised return does something specific and predictable to who ends up at the top. It promotes whoever took the most risk and has not yet been punished for it. Not the best trader. The luckiest survivor of the riskiest strategy.

That is not a cynical reading. It is a measured one. There is a controlled experiment on exactly this mechanism, published in a peer-reviewed management journal, and its numbers are worse than most sceptics would guess. Handing a decision to a sorted list you cannot audit sits closer to the mathematical signature that decides whether forex trading is gambling than the marketing admits.

This article by PropLynq Prop Firm walks through what copy trading is, how the money actually moves, and what the evidence says about the gap between the return on the leaderboard and the return in your account. It matters most to anyone weighing how much money to start trading with, because this is routinely pitched as the shortcut that makes a small account viable. If you already trade seriously — including on a prop trading evaluation — there is a section on why this is a rules problem before it is a returns problem.

Copy trading is a service that automatically replicates another trader’s positions in your account, sized in proportion to the funds you allocate. When the trader you follow opens, adjusts, or closes a position, the same action executes for you without further input. You keep market risk, pay platform and provider costs, and delegate every decision.

What Copy Trading Is and How the Money Actually Moves

Copy trading links a portion of your account to a chosen trader’s account and reproduces their actions proportionally. Allocate $5,000 to a leader who risks 1% of their own equity on a position, and 1% of your $5,000 goes at risk on the same instrument, at close to the same moment.

The word “proportionally” is where the first misunderstanding lives. Proportional sizing preserves percentage risk exactly — it does not reduce it. If the leader draws down 18% of their equity, you draw down 18% of your allocation. Allocating less money does not make the strategy safer; it makes the same strategy smaller in dollars. Percentage exposure is imported at full strength.

The instruction path matters too. Your platform receives a signal, translates it into an order, and routes it — and none of that is free or instant. The leader’s fill and your fill are different events at different prices. On a fast entry that difference shows up as slippage, and it works against the follower more often than the leader because the leader’s order hits the book first. You also cannot use the tools a discretionary trader would reach for. You cannot set your own pending order inside a position you did not open, move a stop because the structure changed, or sit out a trade you dislike. The relationship is binary. Take every position, or exit and close them all at market.

The Leaderboard Is the Product, Not the Trader

Ranking by return is a selection filter, and it selects for variance rather than skill. Two traders can post the same twelve-month figure — one by grinding a small consistent edge, one by holding oversized positions that happened not to blow up. The ranking cannot tell them apart, because realised return is the only input.

Consider what that does over a population. If a hundred traders run a genuinely reckless strategy that ends badly four times in five, roughly twenty of them will still be sitting on strong returns at any given moment. Those twenty are the ones the sort order puts on screen. The eighty are not on the list at all, because the list only ranks people who are still ranked. The visible top of the leaderboard is a survivorship artefact before it is a performance signal.

Platform data makes the concentration concrete. In a survey of four large platforms, the share of ranked users who were actually being copied ran from 1.13% to 8.71% — on eToro, 2,417 leaders out of 193,701 ranked users, or 1.25%. And the distribution within that small group is extreme: the top 5% of leaders accounted for between 61.1% and 92.8% of all such relationships.

A few dozen accounts absorb almost all of the delegated capital, chosen by a sort order that cannot distinguish a durable edge from a fortunate run. This is the same evidence problem as judging prop firm income from payout screenshots, or judging a firm by its landing page rather than its rules — the failures are simply not in the sample. Working out how to choose a prop firm runs into the identical trap.

What a Controlled Experiment Found About Who Gets Copied

In 2020, Apesteguia, Oechssler and Weidenholzer published a laboratory study of imitation in financial markets in Management Science. They gave 176 participants a market with four assets ranging from a fixed-payout safe option to a high-volatility asset with a 4% per-period crash probability — an asset that ended in a crash roughly 80% of the time when held to term. Participants first completed a standard risk-elicitation task, so the researchers knew each person’s actual risk tolerance before any trading began.

One group then saw a ranking list of earlier participants, sorted highest earnings first, with a Copy button next to each name. The design is a stripped-down copy trading platform with the social noise removed.

Thirty-five percent of that group used it. Of them, 88% chose a participant holding the riskiest available asset, and 71% took the single highest earner on the list. Twenty-nine percent never scrolled past the first five names, which meant they never saw what the same asset did to the people it destroyed. Just seeing the ranking was enough on its own: participants who declined still shifted toward riskier assets than the control group, who saw no list at all. The ranking did not inform the decision. It reframed it, in much the way an unmissable move on a screen reframes a decision into FOMO.

Copy Trading vs Mirror Trading vs Social Trading vs Managed Accounts!

These four terms get used interchangeably in marketing copy and mean four different things in practice. The difference is who or what you are delegating to, and how much discretion moves with the money.

Copy trading ties your allocation to a specific person. You inherit their judgement, their emotional state, and their bad weeks along with their good ones. Mirror trading follows a defined strategy or algorithm rather than an individual, so the logic is fixed and disclosed in advance and there is no human deciding to break the rules on a Tuesday.

Social trading is the low-commitment version — you see what others are doing and decide for yourself whether to act, which keeps you as the decision-maker and turns the platform into a discussion board with charts. Managed structures, including PAMM and MAM arrangements, pool capital under one manager who allocates across sub-accounts, and they usually sit inside a formal contractual and regulatory wrapper.

The practical ranking on control is straightforward. The social version leaves you the most, the mirrored version fixes the logic but removes your input, copy trading removes your input and keeps the human variability, and the pooled structures remove both but at least name the manager. Someone running an algorithm through their own terminal after installing a custom indicator on MT5 still owns every parameter. The follower owns nothing but the exit button. That distinction matters most for anyone still working out how to start trading, because this route quietly skips the stage where you learn what a parameter is.

Why the Most Risk-Averse Traders End Up With the Riskiest Exposure

The experiment’s most uncomfortable finding is about who presses the button. Across every regression the researchers ran, the one consistent predictor was risk aversion — and it pointed the wrong way. The more risk-averse a participant was, the more likely they were to follow someone else. Each step down the risk-tolerance scale raised that probability by roughly 26% to 33%.

Now combine that with where the money went. Based on their own elicited risk preferences, exactly one participant out of 48 should have ended up holding the riskiest asset. Once the button was available, 31 of 48 either chose it directly or followed someone who had — thirty-one times the number the group’s own stated risk tolerance predicted.

Read that sequence carefully, because it inverts the entire sales pitch. Copy trading is marketed to cautious people who do not want to take chances alone. It is disproportionately adopted by them. And it routes them into the highest-variance exposure on the platform. The mechanism does not require anyone to be dishonest — the ranking is accurate, the returns are real, the button works exactly as advertised.

It is the sort order doing the damage, in the same structural way that martingale sizing looks safest right up until the moment it is not. Delegating the decision does not remove the emotional problem either. It relocates it, and the relocated version arrives without any of the context that would let you sit through it.

The Return You See Is Not the Return You Get

A leader’s advertised figure and a copier’s realised figure diverge for reasons that compound quietly. Field research on eToro data found that such positions finished positive more often than ordinary ones — but that the return on the successful ones was smaller, and the losses ran larger. More winners, smaller winners, bigger losers.

That combination is a specific arithmetic trap, and it is worth doing on paper. Take a strategy that wins 60% of the time, makes 0.80R on winners and loses 1.30R on losers. The win rate looks excellent. The expectancy is −0.04R per trade, which is roughly −4R over a hundred trades. To break even on that payoff profile you would need a 61.9% win rate — so a strategy sitting at 60% is losing money while displaying a headline number most traders would envy. A win rate published without an average win-to-loss ratio is not a performance statistic. It is decoration.

Three further leaks sit between the leader’s number and yours. Entry price differs because your order arrives after theirs. Position sizing rounds — small allocations get rounded lot sizes that shift your true risk away from the advertised percentage, which is the same arithmetic anyone learns from a forex lot size calculator. And spread and swap conditions on your side of the platform are yours, not the leader’s, so a strategy that is marginally profitable at their costs can be flatly negative at yours.

If you have never checked pip value by pair, those differences are invisible until they show up in the balance. Copy trading does not transmit an outcome. It transmits an instruction, and the outcome is computed in your account under your conditions.

How Copy Trading Platforms and Signal Providers Get Paid

Follow the compensation and the incentives stop being mysterious. These platforms typically do not charge subscription fees — they earn on the bid-ask spread of transactions, which means platform revenue rises with trading volume regardless of whether copiers make money.

Leaders are paid on similar terms. ZuluTrade has offered signal providers a commission of 0.5 pip on volume executed through copiers. eToro’s popular investor programme has paid fixed amounts plus up to 2% of the equity copying that investor, along with spread rebates on their own trading. Put numbers on it. Half a pip on a standard lot is $5, which against an eight-pip average net edge is a 6.25% haircut on every trade.

The 2% figure is larger and structurally different: on $50,000 of copied equity that is $1,000 a year paid for attracting capital, not for making it grow — a fee that arrives whether followers profit or not. Neither scheme pays for risk-adjusted return or for drawdown control. Both pay for volume and for assets attracted, and a leader optimising honestly against those incentives will trade more and post the kind of returns that climb a ranking. That is also why a leader’s requote and execution environment can differ from yours without anyone breaking a rule.

What Copy Trading Looks Like Under Financial Regulation

Regulators do not classify this as a learning tool. They classify it as fund management. The FCA has stated that mirror and copy trading produce “people-based portfolios” and supports ESMA’s position that automatic execution of trade signals falls within MiFID.

Where the service provider has investment discretion — the trades execute without further intervention from you — the activity regularly qualifies as portfolio management, and ESMA issued a dedicated supervisory briefing on firms offering these services in March 2023. In 2025, ESMA confirmed the same reasoning applies to copy trading in crypto-assets under MiCA.

Two things follow. First, the moment execution happens without your input, you have hired a discretionary manager, and the correct comparison is not “cheaper than learning to trade” but “how does this manager compare to other managers.” Second, the regulatory perimeter is where the consumer protections live — suitability assessments, disclosure standards, complaint routes.

A service operating outside it offers none of them, which is the same due-diligence question that decides whether a prop firm is a scam or a legitimate business. Check the licence and what it actually authorises before allocating anything.

Why Copy Trading Breaks Inside a Prop Firm Evaluation

If you trade a funded account, copying is a rules problem before it is a returns problem — and the rule is close to universal. Firms draw a line at ownership. Internal duplication — across accounts you personally own — is permitted at many futures firms and restricted at most forex firms. The external kind, meaning third-party signals, someone else managing your account, or pass-your-challenge services, is prohibited essentially everywhere, and the penalty is usually termination with profits voided and fees unrefunded. Firms detect it by comparing entry timestamps, lot sizes and sequencing across unrelated accounts. The logic is not arbitrary: an evaluation exists to measure your edge, and an imported edge measures nothing.

Even where a firm allowed it, the mechanics would work against you. Published drawdown figures usually describe closed-trade equity. Evaluation limits are measured on live equity, floating positions included. A leader who holds losers through a drawdown and closes green records a winning day at zero drawdown — and would have run your account through an intraday breach on the way. Take a $100,000 PropLynq Two-Step account, where the structure is a 5% daily loss limit and a 10% maximum drawdown.

That is $5,000 of daily room. A leader floating 6.2% against you before recovering puts $6,200 of unrealised loss on your equity curve — $1,200 past the limit, with the account closed before the recovery arrives. Their public record for that day reads +0.4%. Yours reads breached. This is the same distinction that makes static vs trailing drawdown worth understanding before you pick a challenge, and it is why how to pass a prop firm challenge is a question about risk mechanics rather than strategy selection. An evaluation is a test of one specific person’s process under specific constraints, which is exactly what an imported signal cannot supply.

How to Assess a Signal Provider If You Still Want to Copy

None of this makes the practice indefensible. Delegating to a disclosed, regulated manager whose method you understand is an ordinary financial decision. Handing money to an anonymous account because it currently sits at the top of a sorted list is not. Five questions separate the two.

Ask for the maximum floating drawdown, not the closed-equity figure, because the difference is exactly where accounts die. Ask for track record length measured in market regimes rather than months — twelve months that contained no volatility shock tells you nothing about behaviour in one. Ask for average win versus average loss alongside the win rate, since neither number means anything alone. Ask how the provider is paid, and whether any part of that compensation depends on your outcome rather than your volume. Ask what happens on the worst day in the record, and whether that day would have breached your own limits.

If a platform will not surface those five, the ranking is the only information on offer, and you now know what the ranking selects for. Keeping a trading journal of your own fills against the leader’s, rather than the platform’s summary, is the only way to find out whether the divergence is small or ruinous. Set an exit rule in advance too, because the decision to uncopy after a bad run is made under exactly the conditions that produce revenge trading.

Conclusion

The most common justification here is educational — watch a professional work and absorb the method. The experimental evidence points the other way. Twenty-nine percent never looked past the top five names, and those who followed someone were disproportionately the least risk-tolerant people in the room. Nobody in that group was studying anything. They were outsourcing a decision that made them uncomfortable.

Learning requires a hypothesis you formed, a decision you made, and feedback you can attribute to your own reasoning. A trade you did not choose gives you an outcome with no attributable cause — you cannot tell whether the leader was right, lucky, or wrong-and-bailed-out, because you never saw their reasoning and the platform does not publish it. Marking up your own chart and being wrong about a trendline teaches more than a hundred flawless imported entries, because the error is yours and therefore traceable. The same is true for a mismarked order block — the mistake is the lesson.

So the honest summary is this. Copy trading is a delegation product wearing an education product’s marketing. As delegation it can be reasonable, if the manager is disclosed, regulated, paid on outcomes, and transparent about floating drawdown. As education it does not work, and the ranking that draws people in selects for precisely the variance they were trying to avoid. Decide which one you are buying before you allocate.

If you would rather build a record that belongs to you, with defined loss limits and rules published in advance, you can get a funded account and trade your own process under evaluation conditions.