What Is a VPS in Trading? Why Uptime Decides Results
A VPS, or virtual private server, is a slice of a physical machine in a datacentre that you rent and use like a remote computer. In trading it has one job: run your MetaTrader terminal continuously, on a connection that does not drop, in a building close to your broker's trade server. For an automated strategy it is not an optimisation — it is the environment the strategy actually lives in.
How it works
You connect to the VPS over remote desktop, install the terminal there, log in and leave it running. Your own laptop can then be closed, updated or thrown away without the strategy noticing. The two things you are buying are uptime and physical proximity to the trade server, and the second one is measurable as round-trip latency.
Ping = terminal → trade server → terminal, round trip
There is no official threshold for a good number, and the one that matters is relative: the same figure that is unremarkable for a strategy holding positions for days is disqualifying for one working the spread. What is not arguable is the direction — as the round trip grows, fills arrive further from the price the strategy asked for.
What happens when the terminal goes down mid-position is the part people get wrong. In MetaTrader, a stop loss or take profit attached to an open position is held by the broker's server, so it still triggers with the terminal off. Everything else is client-side and stops dead: new entries, trailing stops, partial closes, break-even moves, virtual stops the Expert Advisor keeps in memory, grid and recovery logic. A martingale system that loses its terminal halfway through a sequence does not pause — it holds an unmanaged basket until the terminal comes back, and the position it would have opened to balance the book never exists.
Latency and packet loss matter for a separate reason: they widen the gap between the price your EA decided on and the price it received. That gap is slippage, and no amount of strategy tuning trades around it.
Why it matters
An automated strategy's live result is the strategy plus its execution environment. Two accounts running identical code on different machines produce different curves, and the difference belongs to ping, uptime and fill quality rather than to logic. This is one of the reasons backtest results overstate live performance: a backtest never disconnects, never misses a bar, and always fills at the modelled price.
It also matters for anyone reading someone else's record. A track record produced on a 40 ms VPS is not reproducible on a home laptop with 300 ms ping and nightly Windows updates, even with the same EA and the same broker.
What the data shows
Automation is the norm on our platform, not the exception. Across the public accounts on ShowMyTrades with trading history (August 2026), the median autotrading share is 99%, and 53.9% of accounts are more than 90% automated. The distribution is barbell-shaped: another 42.2% are under 10% automated, with very little in between. Traders are either fully hands-off or fully manual.
More than half of the accounts publishing results here therefore depend on a terminal that has to stay up, connected, and permitted to trade — across 703 distinct broker servers and 15,436,464 synchronised trades.
Where you see it on ShowMyTrades
The Terminal & Open Charts module on the account page reports the live state of the machine the strategy runs on: Ping to the trade server, Retransmission (the share of network packets that had to be sent again), Terminal build, Memory of the host machine, and whether Auto trading and DLL imports are allowed. A connection badge reads Connected, No trade server or Offline, next to how long ago the terminal last reported, and the header counts how many of your terminals are online.
Below it, the open charts table lists each chart with its symbol, timeframe, Expert Advisor and indicators — so you can see what is actually loaded, not what is claimed. In the account header, a badge shows the automation split read from the trades themselves, for example 99% autotrading.
This is also what a verified track record looks like when it is honest about its own conditions: the numbers and the environment that produced them, on the same page.
Common misunderstandings
- A VPS does not make a strategy profitable. It removes missed executions and downtime. If the edge is not there, a faster machine loses money more reliably.
- Ping to a website is not ping to the trade server. The number that matters is the round trip to your broker's server, which is what the terminal reports — a VPS in the wrong city can be slower than your desk.
- Low latency does not eliminate slippage. Spread, liquidity and the broker's execution model still apply. A VPS removes your share of the delay, not theirs.
- "My PC never crashes" is not the claim being made. Windows updates, sleep settings, ISP outages and power cuts all disconnect a terminal, and each one happens at a random moment relative to your open positions.
To read the rest of an account page with the same scepticism: how to read a trading account dashboard.
Related terms
Verified Track Record
A verified track record is performance data pulled straight from the broker through a read-only link, continuously, with nothing the publisher can edit.
Backtesting
Backtesting simulates a trading strategy on historical price data. What it can prove, what it cannot, and why live broker-synced results almost always differ.
Slippage
Slippage is the difference between the price you expected and the price you got. Why it is worst when it hurts most, and how latency makes it measurable.