Copy trading, and how it compares to running a bot

What actually happens when you copy a trader

A provider trades their own live account. Copier software watches it and opens a proportional position on every subscriber account, normally within a second or two. You are not receiving advice or licensing a strategy — you are duplicating executions, whatever they turn out to be.

Sizing is proportional, not identical. If the provider risks 2% of their equity and your account is a tenth of theirs, you get a position a tenth the size risking the same 2%. This is the detail most people get backwards: copy trading does not reduce risk, it reproduces it exactly. A provider who blows up takes their subscribers down at the same percentage.

The delivery mechanism varies. MQL5 Signals runs it inside MetaTrader itself, a broker copier such as Exness Social Trading runs it internally, and a Telegram channel is not copy trading at all — it is a signal service you execute by hand.

Costs, including the one that is not on the price list

The visible cost is a performance fee, typically ten to thirty percent of profit. What matters more than the rate is whether it is charged against a high water mark: without one, a provider can lose 20% and then bill you on the recovery of their own loss.

The invisible cost is execution. Your fill is never the provider's fill. Between their execution and yours sits network latency, your broker's spread rather than theirs, and possible requoting. On gold that gap is real money, and it works against the subscriber structurally — the provider's entry is the better one by definition.

On a small account the fee is genuinely cheaper. As the account grows, a percentage of everything stops being the economical option surprisingly quickly.

How to assess a provider without being sold to

Ignore the headline return. Open the monthly table instead: twelve unremarkable positive months describe a system, while one enormous month and eleven flat ones describe an event that will not repeat.

• ·Account age and total trades first — a two-month record has told you nothing except that two months went well.
• ·Maximum drawdown, which is what you will actually have to sit through.
• ·Deposit load — a smooth curve at high load means the punishing move simply has not arrived yet.
• ·Whether the record is verified and whether it is live or demo.
• ·Whether losing periods are shown, or the record conveniently begins after them.

Be especially careful with leaderboards. They sort by recent return, and the quickest route to the top of a return-sorted list is enormous risk that has not failed yet — see social trading.

Which one suits you

Copy trading suits someone who wants no involvement at all and accepts that they cannot inspect or adjust what is happening. It is genuinely simpler, and simplicity has value.

A bot suits someone who wants to know why a position was opened, control their own risk settings, and keep costs fixed as the account grows. It requires more of you — a running terminal or a VPS, and a willingness to read documentation.

The dishonest answer is that one is better. The honest one is that they fail differently: a bot fails when its rules stop matching the market, and copy trading fails when a person you cannot see takes a risk you would not have taken.