# Trade copier latency: how long a copied trade takes

> Trade copier latency in Copilink's data: followers fill at a median of 103 ms after the leader, and on average about 120 ms in North America and 200 ms in Europe.

Last verified: 2026-10-27
Canonical page: https://copilink.com/articles/trade-copier-latency
Product: Copilink Standalone and Copilink for NinjaTrader 8
Markdown of: https://copilink.com/articles/trade-copier-latency


## Quick facts

- Canonical HTML: https://copilink.com/articles/trade-copier-latency
- Last verified: 2026-10-27
- Publisher: Copilink
- Focus: practical, first-party guidance for multi-account futures traders

Trade copier latency, or copy delay, is the time between the leader's order filling and the follower's copy filling. Across about 2,000 copied follower trades from Copilink Standalone users who sync their data, half of the copies filled within 103 ms of the leader and 9 in 10 within 391 ms. Location sets most of the difference between traders: copies average about 120 ms in North America and about 200 ms in Europe, and from what we see in support, traders in Asia can see 200 to 300 ms.

[Copilink Standalone](https://copilink.com/standalone) runs on your own computer and sends follower orders straight to your broker, with no Copilink server in between. Its own processing is about 5 ms of that time; the rest is distance and the broker filling the order.

I copy my own trades from Europe, and the delay isn't noticeable. It has never affected me.

## Trade copier latency at a glance

| Measure | Copilink's data |
| --- | --- |
| Half of copies filled within | 103 ms |
| 9 in 10 copies filled within | 391 ms |
| Average, stalled trader-days left out | 142 ms |
| North America, average | about 120 ms |
| Europe, average | about 200 ms |
| Follower entry vs leader entry, median | same price (0 points) |
| 9 in 10 follower entries | no more than half a point worse |
| Copilink's own processing | about 5 ms |

All figures are from the leader's fill to the follower's fill, measured with the broker's own fill timestamps. How we measured it is at the end of the page.

## What trade copier latency measures

A copied trade makes four steps between the two fills:

1. The leader's fill notice travels from the broker to the copier.
2. The copier works out each follower's order. This is the only step the copier itself controls.
3. The follower order travels from the copier to the broker.
4. The broker accepts the order and fills it.

Copy delay covers all four. Processing time covers only step 2. That's the figure Copilink has published until now: about 5 ms in Copilink Standalone and about 1.6 ms in Copilink for NinjaTrader 8 ([how we measure processing time](https://copilink.com/methodology)), which is what our [fastest trade copier](https://copilink.com/fastest-trade-copier) page quotes. Both are real, but they answer a different question: how much the copier adds, not how long the follower waits.

When you compare speed claims, check where each timer starts and stops. A figure that doesn't say can't be compared with one that does.

## How long copied trades take

Half of all copies filled within 103 ms of the leader, and 9 in 10 within 391 ms. With stalled trader-days left out, the average was 142 ms.

The slow tail comes from a handful of trader-days when copies took seconds instead of milliseconds. We treat those as connection or platform stalls, since the copier's own step takes about 5 ms. Counted in, they push a plain average to about 3 seconds, which says nothing about normal copying. So the headline figures are the median and the 9-in-10 figure, and the averages leave out trader-days that averaged over 500 ms, about 8% of follower trades.

## Copy delay by region

| Region | Average over all copies | Typical trader | Sample |
| --- | --- | --- | --- |
| North America | 117 ms | 127 ms | largest |
| Latin America & Caribbean | 175 ms | 129 ms | smaller |
| Europe | 198 ms | 171 ms | smaller |
| Asia, Australia | not shown | not shown | too few to show |

"Average over all copies" weights every copied trade equally, so very active traders count more. "Typical trader" is the middle trader in each region, which keeps one busy account from setting the number. Region is where each trader visits copilink.com from, so a European trader whose copier runs on a server in Chicago counts as Europe. That may be part of why the typical European trader sits closer to North America than the all-copies average.

The gap between North America and Europe, about 45 to 80 ms depending on the measure, is the extra distance across the Atlantic. Every copy pays one round trip between the copier and the broker's servers in the Chicago area: about 20 to 50 ms from most of the US and about 90 to 130 ms from most of Europe ([WonderNetwork pings to Chicago](https://wondernetwork.com/pings/Chicago)). Our [NinjaTrader VPS guide](https://copilink.com/articles/ninjatrader-vps) has the round-trip times city by city and what moving your copier changes.

From what I see in support, traders close to the servers they trade on usually get Copilink's order to the broker in about 20 ms. Traders in Asia can see copy delays of 200 to 300 ms. It depends heavily on location.

![Trade copier latency by component: in North America about 117 ms splits into ~5 ms of Copilink processing, ~20 to 50 ms of travel and an estimated ~80 ms for the broker to confirm and fill; in Europe about 198 ms has the same processing and broker share but ~90 to 130 ms of travel across the Atlantic and back.](https://copilink.com/media/articles/trade-copier-latency-where-time-goes.webp)

## What sets copy delay, and what doesn't

**Where the copier runs.** Distance is the biggest part you can change. Moving a European copier next to the broker removes the Atlantic round trip, not the broker's time, so it gets you close to North American numbers rather than to zero.

**The broker.** Once distance is taken out, most of the remaining 100 ms or so is the broker accepting and filling the order. No copier or server changes that.

**The order type.** It decides when the copy delay happens:

- A market order is copied when the leader's order fills, so the follower pays the whole copy delay.
- Limit, stop and stop-limit orders are copied as working orders. The follower's copy sits at the broker at the same price as the leader's, moves when you move the leader's and is cancelled when you cancel it. When price reaches the level, each account fills on its own, and a follower can fill before the leader. When a follower's copy fills first, there's no delay to measure, so those copies are left out. Working-order copies that fill after the leader are counted, so the figures above mix both order types.

The catch with working orders is a fast move that runs straight through the price. The leader's order fills and a follower's copy, placed a moment later, may not. Copilink Standalone waits a set time for the copy to fill, 10 seconds by default, and then cancels it, so the follower sits that trade out instead of chasing at a worse price. The [Entry fill wait setting](https://copilink.com/standalone/user-guide/how-copying-works#entry-fill-wait) controls how long.

![Trade copier latency by order type: a market order is copied only after the leader fills, so the follower fill comes one copy delay later; a limit or stop order is copied as a working order at the broker, so leader and follower fill as price reaches the level.](https://copilink.com/media/articles/trade-copier-latency-market-vs-working.webp)

**The number of followers.** Copilink Standalone copies each follower on its own, so a slow broker response on one account doesn't hold up orders or stops for the others.

**What barely matters.** The copier's processing, at about 5 ms. The power of your PC: copiers use very little of it. Cloud sync, which isn't in the copy path at all; it uploads finished trades for the journal after the fact.

## Does copy delay cost you money?

Very little, in our data. The median follower entry filled at exactly the leader's price, and 9 in 10 follower entries filled no more than half a point worse. On average, followers came out very slightly ahead of the leader. On MNQ, half a point is $1 per contract; on NQ it's $10. These figures include limit and stop copies, which rest at the leader's price; a market-order copy pays the delay in full, as above.

That matches my own trading. My trades are short MNQ and NQ scalps from Europe, and the delay on the followers costs me a few dollars of slippage. I'm fine with that.

When a trader tells us copies are slow or slipping, we check the trades one by one, and it usually turns out to be something else:

- **A follower account refusing orders.** One trader reported a lot of slippage. On the followers that copied, fills matched the leader to within about a tick. Two large exits at the open filled a few ticks apart, because a follower can only exit after the leader's exit fills, and with 21 contracts a few ticks add up. The real problem was a broker restriction on some follower accounts, which refused the orders. Copilink now switches a follower off by itself when its broker rejects or cancels three of its copied orders, so a refused account stops piling up rejected orders while the others keep copying.
- **A limit entry that price ran through.** Another trader thought Copilink ran in the cloud and was too slow for scalping. It doesn't run in the cloud, and their copies were reaching the broker in about 25 ms. What had happened was a fast drop: NQ fell 10 ticks in under 2 seconds, and the followers' limit entries never filled. That's how the Entry fill wait above came about.

One trial user compared Copilink with another copier on the same accounts. Both had some delay, they said, but with Copilink both followers got the same entry, while the other copier filled each account at a different price.

If your followers do drift from the leader, our [trade copier troubleshooting guide](https://copilink.com/articles/trade-copier-troubleshooting) goes through the causes.

## Copying through news and fast opens

Around a big release such as CPI, NFP or FOMC, spreads widen and price can jump several ticks while an order travels. Copy delay costs the most there, on market orders and on exits.

I never hold a trade through those releases. That's asking to be wrecked.

Some prop firms also make you flat around news, and a copied position counts on every account:

- **Lucid Trading:** LucidDaily funded accounts must be flat from 1 minute before to 1 minute after US high-impact news. Breaking it is a hard breach ([Lucid's LucidDaily funded rules](https://support.lucidtrading.com/en/articles/15997244-luciddaily-funded-account)). See the [Lucid Trading rules](https://copilink.com/prop-firms/lucid-trading).
- **My Funded Futures:** sim funded accounts on the Rapid and Pro plans must be flat, with no working orders, from 2 minutes before to 2 minutes after Tier 1 releases ([MFFU's news trading policy](https://help.myfundedfutures.com/en/articles/8230009-news-trading-policy)). See the [My Funded Futures rules](https://copilink.com/prop-firms/my-funded-futures).
- **Tradeify:** no news rule; trading news is at your own risk ([Tradeify's news rules](https://help.tradeify.co/en/articles/10495874-rules-news-trading)).

Rules differ by plan and change often. Check yours in the [prop firm rules directory](https://copilink.com/prop-firms) before a release day.

## How to cut copy delay

- **Move the copier closer to the broker, if distance is your problem.** The closest to zero delay you will get is a copier in Chicago, and even then you can't get to zero: the broker's part stays. A Chicago VPS does this; most traders don't need one.
- **If you trade through NinjaTrader or TradingView, keep the copier on the same machine as the platform.** Every relay between machines adds its own trip.
- **Use a wired connection** if you copy from home. Wi-Fi adds a little delay and drops more often.
- **Use working orders where your strategy allows,** so the copies already sit at the broker when price arrives.
- **Check after any change:** place one micro contract trade and compare each follower's fill with the leader's.

## How we measured it

- **Data:** copied trades from Copilink Standalone users who sync their trades to the cloud, on prop firm evaluation and sim-funded accounts at Tradovate and Rithmic.
- **Timing:** the broker's own fill timestamps on the leader and the follower, for copies in the same direction on the same instrument.
- **Left out entirely:** copies where the follower filled before the leader, which happens with working orders that price reached first.
- **Left out of the averages only:** trader-days that averaged over 500 ms, about 8% of follower trades, which we treat as connection or platform stalls. The median and the 9-in-10 figure use every remaining sample.
- **Region:** where each trader visits copilink.com from, which can differ from where the copier runs. Regions with too few traders aren't shown.
- **Slippage:** the follower's entry price against the leader's, in points, counted as worse when it's against the follower.

Copilink's processing benchmarks, which time only the copier's own step, are in the [Tradovate copy latency benchmark](https://copilink.com/research/tradovate-copy-latency-benchmark). Your own copy delay depends on your broker, connection, location and market conditions.

## Copy delay on NinjaTrader 8

The figures above come from Copilink Standalone. [Copilink for NinjaTrader 8](https://copilink.com/trade-copier) runs inside NinjaTrader and sends follower orders through NinjaTrader's own connections, so the same four steps apply, but we haven't measured its fill-to-fill delay across users.

Its Market setting decides how much delay the follower pays. With Market off, the setting we recommend for most users, a follower's working order is placed as soon as the leader's working order is placed. With Market on, followers get market orders only after the leader fills, so they pay the full copy delay and can see more slippage.

One NinjaTrader detail I've noticed: Rithmic P&L updates arrive more slowly in NinjaTrader than Tradovate's do. That comes from how NinjaTrader is built inside, not from the copier.

## FAQ

**Is there any delay when copying trades?**
Yes. A market-order copy is sent only after the leader's order fills; limit and stop copies already rest at the broker. In Copilink's data, half of copies filled within 103 ms of the leader and 9 in 10 within 391 ms.

**How fast does NinjaTrader 8 fill orders?**
NinjaTrader sends the order; the broker fills it, so the time depends on your connection, the broker and the market. We haven't measured NinjaTrader's own fill time. Copilink for NinjaTrader 8 adds about 1.6 ms of its own processing, and in Copilink Standalone data a follower filled a median 103 ms after the leader.

**Is a local or a cloud trade copier faster?**
It depends on where each one runs relative to your broker. A local copier sends orders from the machine your platform runs on; a cloud copier relays them through its own server, so that server's location sits in the path too. A local copier on a computer near the broker has the shortest path. See [how Copilink compares with other copiers](https://copilink.com/compare-trade-copiers).

**Does a VPS reduce copy delay?**
Yes, if you're far from the broker. A Chicago VPS removes the round trip across the ocean, about 90 to 130 ms from most of Europe, but not the broker's time.

**Why did my follower fill at a different price than the leader?**
Price moved during the copy delay, the follower's limit copy filled at a different moment, or a large order filled in pieces. In our data the median follower entry matched the leader's price exactly.

**What is latency arbitrage?**
Profiting from a speed edge over other traders, usually with high-frequency strategies. It has nothing to do with copy trading, where speed matters for one reason: keeping followers' fills close to the leader's.

## Why traders choose Copilink

- Choose a direct Tradovate and Rithmic desktop workflow on Windows or macOS, or stay entirely inside NinjaTrader 8 with the dedicated add-on.
- Manage exact or ratio-based follower sizing, per-account safeguards, and prop-firm-style risk rules from one focused workflow.
- Keep copying local on the trader's machine while cloud sync powers web analytics and journaling.
- Start with a no-card trial and clear plan options for both Standalone and NinjaTrader 8 workflows.

## Sources

- https://copilink.com/articles/trade-copier-latency