A signal arrives in a Telegram channel at 09:31: "XAUUSD BUY 2412.50 SL 2405.00 TP1 2418 TP2 2426". Four seconds later there is a position on the account. Between those two moments a piece of software read that line as plain text, decided it was an instruction, extracted five numbers, mapped a symbol name, calculated a volume and sent an order to the trading server. Nothing about that chain is intelligent. It is pattern matching, and it is fragile in exactly the places you would expect.
What the parser is actually looking for
Most Telegram-to-MetaTrader copiers work in three stages. First they connect to the message source, either through a bot added to the channel or through a user session that reads the channel the way a phone client would. Second, they run the message body through a parser. Third, they hand a structured order to a bridge that talks to the terminal.
The parser looks for a small set of tokens. A direction word (buy, sell, long, short, and sometimes an emoji). A symbol, matched against a list of aliases, because a provider writing GOLD means the same thing your broker calls XAUUSD.m or XAUUSD.pro. A price for entry, or the absence of one, which the parser reads as "market". A stop loss, usually behind the letters SL or the words stop loss. One or more take profit levels. Everything else in the message, the commentary, the risk warning, the channel branding, is noise that has to be ignored without accidentally swallowing a number.
That last part is where the failures live. A message that says "TP1 2418 TP2 2426, aiming for 100 pips today" gives the parser a third number to think about. A message that says "risk 1% max" hands it a stray digit. Good parsers anchor on labels rather than on position in the line, and they refuse to act when a required field is missing rather than filling it with a default.
The four failure modes worth knowing
The first is symbol mapping. Your provider trades on a broker whose gold symbol is XAUUSD. Yours has a suffix. If the alias table is not maintained, the copier either errors out or, on some platforms, matches a different instrument entirely. Always check the mapping list before you go live rather than after.
The second is the edit. Telegram lets a sender edit a message after it is posted. A provider who types 2405 and corrects it to 2450 thirty seconds later has produced two different signals from one message. Some copiers reprocess edits, some ignore them, and the ones that reprocess can open a duplicate position. Find out which behaviour yours has.
The third is the follow-up. "Move SL to entry" and "close half" are instructions about an existing trade, and they only work if the copier remembered which order it opened from which message. Copiers that track the reply chain handle this well. Copiers that treat every message independently will silently drop every management instruction the provider sends, which means you get the entries but none of the exits. That is a materially different strategy from the one you subscribed to.
The fourth is timing. A copier that polls for messages once a minute is not a copier, it is a delay. Webhook-driven and session-driven pipelines deliver in seconds. On gold at a news release, seconds are the whole trade. Read our note on slippage and why fills differ from quotes before assuming the difference is the copier's fault.
A copier does not reduce risk. It removes the delay between someone else's opinion and your money being at risk, which is the opposite. Every position it opens is yours, and leveraged trading carries a high risk of loss.
Volume is the setting people get wrong
The signal says buy. It does not say how much. Volume comes from your configuration, and the three common models behave very differently.
Fixed lot is what most people pick because it is simple. It ignores your account size, so a provider whose stops are 400 points wide on gold and 15 pips wide on EURUSD will produce wildly different losses from the same lot size. Risk-percent sizing calculates volume backwards from the stop distance, which is the only model that keeps the loss per trade constant, and it fails closed when the message has no stop loss. Proportional sizing copies the provider's lot scaled by the ratio of the two account balances, which only makes sense when the provider actually publishes real lot sizes.
If you use risk-percent, decide in advance what the copier should do with a signal that has no stop. Skipping it is the defensible answer. Our guide to position size calculators covers the arithmetic, and the risk rules themselves are the part no software can decide for you.
Bot source or account source
There are two ways to get the messages. A bot added to the channel as an administrator receives posts through the official bot API. It is clean, it is permitted, and it requires the channel owner to cooperate. A user session reads the channel as a logged-in account, which works for any channel you can personally join, and which puts your Telegram account inside an automation. Providers who do not want their signals redistributed often ban this, and the platform's own rules on automation apply.
If you run a channel yourself and want subscribers to receive a consistent, machine-readable format, publishing through a bot is the better engineering decision. It also lets you control the format, which is the single biggest determinant of whether copiers downstream work at all. Running a signal channel and running a copier are related problems with the same root cause.
Testing before you trust it
Point the copier at a demo account and let it run for at least two weeks against the live channel. You are not testing profitability. You are testing whether every message the provider sent in that window produced the trade you would have taken by hand. Count the misses. Count the duplicates. Count the management instructions that were dropped. If the answer to any of those is not zero, either the format or the parser needs work.
Keep the log. A copier that does not write a line for every message it saw, including the ones it rejected and why, is impossible to debug. When something goes wrong at 09:31 on a Friday, the rejection log is the only evidence you have. Evaluating the signal service itself is a separate exercise, and one worth doing first.
"Ninety percent of copier complaints are not bugs in the copier. They are a provider who changed how he types his messages on a Tuesday."
— Alex Onta, Executive Director, SINGUARD
Key Takeaways
- A copier is a text parser, so the provider's message format determines whether it works at all.
- Check symbol aliases, message edits, follow-up instructions and delivery latency before going live.
- Volume comes from your settings, not the signal, and risk-percent sizing should skip any message with no stop loss.
- Run two weeks on demo and count misses, duplicates and dropped management messages rather than profit.
Frequently Asked Questions
Can a Telegram copier handle messages with no entry price?
Most treat a missing entry as a market order. That is fine when the signal is meant to be taken immediately, and dangerous when it was a pending setup posted hours in advance, so check which assumption your copier makes.
What happens if the provider edits a signal after posting?
It depends on the copier. Some reprocess the edited text and can open a second position, some ignore edits entirely and leave you on the original numbers. Ask the vendor directly, because both behaviours cause different problems.
Do I need a bot in the channel or can it read as my account?
A bot added by the channel owner uses the official API and is the cleaner route. Reading through your own logged-in session works on any channel you can join but puts your personal account inside an automation, and many providers prohibit it.
About the Author
Alex Onta is an Executive Director at SINGUARD. He built eTrader, the terminal, the mobile apps, eTrader Broker, Copytrading, Business and Community, along with the worldwide clustered-server infrastructure it all runs on, with his brother Roman Onta helping on the design, and he leads that division today. Together with Roman he builds the Prop Firm CRM, the Broker CRM, Scalegram and CopySignals, and the two of them carry worldwide compliance, payment processing and international business structuring side by side. He lives and works in Dubai for most of the year. Meet the executive duo leading Singuard's five divisions.