Trade copier latency: what the numbers really mean
Where the time goes between a signal and a fill, why latency is not slippage, and how to measure your own copier on a demo account.
Trade copier latency is the time between a trade appearing on the master side, a Telegram post or a position on another account, and the matching order being sent to your broker. It is a sum of several short delays, and a number quoted for one of them says little about the fill price you get.
CopySignals is a cloud Telegram signal copier for MT4, MT5 and eTrader: it reads the rooms you already follow through your own Telegram account and places the trades on your own broker account. We do not publish a latency figure, and this page explains why you should be careful with anyone's. It covers where the time goes, which part you can influence and how to measure your own setup on a demo account.
What does trade copier latency actually measure?
It measures one step in a chain, and each vendor picks the step that looks best. A copier can only start counting when it knows about the trade, and it can only stop counting when your broker has accepted an order. Everything before and after those two points is outside the copier.
For a Telegram copier the chain looks like this:
- The person running the room writes the post and sends it.
- Telegram delivers it to your account.
- The copier reads the text and works out instrument, side, entry, stop and targets.
- The copier checks your rules: filters, maximum spread, open position cap, lot size.
- The order travels to your broker and the broker executes it at a price.
Steps 3 and 4 are the copier's own work. Step 1 is the human, step 2 is Telegram and step 5 is your broker. A figure for "copy time" normally covers a slice of 3 to 5. A trader who waits for a fill sees the whole chain, and the largest part of it is often the person typing the post.
Why is a latency number not the same as slippage?
Latency is time. Slippage is the difference between the price you expected and the price you got. They are related only when the market is moving, and in a quiet market a longer delay costs nothing while in a fast one a short delay can cost a lot.
Take a gold buy posted at 2,400.00 with an entry zone in the text. If the quote moves two dollars in the seconds before your order reaches the broker, you pay two dollars more than the post suggests, whether the copier took a fraction of a second or three. The same trade in a flat afternoon fills at the posted price either way. This is why the useful question is not "how fast is the copier" but "what price did my account get compared with the post, across a week of the rooms I follow".
The spread adds to this. You buy at the ask and sell at the bid, and around news releases brokers widen it. A copier that fires into a stretched spread fills badly however quick it was, which is the reason for a maximum spread filter.
Where does a cloud copier spend its time, and where does a local one?
A local copier runs inside a terminal on your own computer or a rented VPS, and the same machine talks to Telegram or to the other terminal and to your broker. A cloud copier runs the terminal on the vendor's servers. Neither layout is faster by definition. Distance to the broker's server, load on the machine and the quality of the connection decide it, and a laptop on home wifi will behave differently from a server in a data centre.
What changes is who carries the risk of the machine. With a local setup, a closed lid, a Windows update or a dropped connection turns into a missed or late trade, and you find out afterwards. On CopySignals the MetaTrader terminals run on our servers, one per connected account, so there is nothing at home to keep on. We wrote up the trade-offs in comparing cloud and local copiers, and the no-VPS side is on the Telegram copier without a VPS page.
eTrader works differently from the two MetaTrader platforms. There the account is driven through eTrader's copytrading API rather than a terminal, so the path to the broker is another one, and you should measure it separately if you use both.
Is a faster copier always better?
No. Speed helps when the first price is the good price. It hurts when the first seconds are the worst ones, which is common for room signals posted into a news spike or posted with a typo the author fixes a moment later. CopySignals has a delay before placing for this reason: you choose how long the copier waits after reading a post. The reasoning is in delay before placing copied trades.
A wait of a few seconds is a trade you make on purpose: a little worse on the quiet fills and a little better on the jumpy ones. Whether it pays for your rooms is a question for your own logs, not for us, and trading carries risk whichever way you set it.
Account to account copying has its own version of the question. If a master account opens a position and the copier mirrors it to a second account at its own size, the mirrored fill comes after the master's, and a wide spread on the second broker adds to the gap. Two brokers quote slightly different prices, and the gap between them is a cost of the setup that has nothing to do with the software. See copying trades between brokers for that case.
How do you measure copier latency yourself?
Measure it on a demo account with the rooms you really follow, for a week, and compare three things per trade. This takes about as long as setting up the copier.
- The time of the post in the Telegram room.
- The time of the event on the copier's log and the time the order opened on the account. On CopySignals the event log shows what each message was read as, including the posts it skipped and why.
- The posted entry against the price your account actually got, in pips or points.
Put the results in a sheet. Sort by the price difference, not by the time. If the large differences cluster around news times, the cause is the market and a delay or a spread filter will help. If they appear in quiet hours, look at the broker: its execution on that symbol, its server and its spread. If a trade is missing, the log tells you whether the post was skipped or never read.
The guide to testing signals on a demo account lists what else to check before going live. Remember that demo servers do not always fill like live ones, so repeat the exercise at a small size on the real account.
What should you ask a vendor who quotes a latency figure?
Ask four things, and expect a straight answer to each:
- Which two moments does the figure sit between?
- Was it measured on a live broker account or in a test that skips the broker?
- Is it an average, a median or a best case, and over how many trades?
- Does it include the time to read and parse a free-text post, or only to forward an order that is already structured?
The last point matters for Telegram. A copier that has to read free text, with the entry zone, several take profits and a stop in whatever layout the room uses, does more work than one that copies a position from another terminal. The two figures are not comparable, which is why a table of latencies across both kinds of product misleads. If a number comes with no answer to these questions, treat it as a slogan.
What to do next
Pick the three rooms you copy most, run them on a demo account for a week and read the price difference per trade. Set a maximum spread first, add a short delay if the news-time trades are the bad ones, and keep the rest of the settings as they are so you can tell which change did what. If you want to see the product, the MT4 and MT5 trade copier page covers the platform side, and every plan includes all of it. Plans start at $14 a month, with a 30 day money back guarantee, on the pricing page. CopySignals supports MT4, MT5 and eTrader only.