Skip to main content
Strategy — scalping latency

Scalping latency — why consistent matters more than fast for M1 and M5 traders.

Raw latency numbers don't tell the scalping story. The number that matters is jitter — the wobble around your average. A session that is always 45ms beats one that swings between 10ms and 200ms.

Scalpers care about fills. Fills depend on the route arriving at the broker's matching engine before the price reprices. That is a jitter problem, not a pure speed problem.

See the plansLatency and jitter guide
Jitter
The wobble that makes fills inconsistent even when the average looks fine.
M1/M5
Timeframes where route noise shows up directly in fill quality.
Consistent
A steady path beats a fast-but-variable one for active scalpers.

What scalping actually demands of your network

Every scalp trade needs to reach the broker's matching engine before the price moves. On tight pairs like EURUSD during a session overlap, prices reprice multiple times per second. The window for a clean fill is narrow. A route that arrives at the consistent 40ms mark gets the same fill quality each time. A route that arrives at 40ms on one order and 190ms on the next introduces unpredictable slippage that compounds over a full session of trades.

Tight spreads make the problem worse. When the spread is 0.1 pips, slippage of 0.3 pips from a late order arrival is a significant cost. Scale that across 20 or 30 trades in a session and the route noise has a measurable effect on the P&L, separate from the trading decisions themselves.

Why residential connections create problems for scalpers

Residential ISPs route your traffic through shared transit networks. The path changes based on what your carrier is doing at that moment — congestion, peering shifts, maintenance. During session overlaps, when you most need a clean route, your ISP's shared infrastructure is also under higher load. The jitter that results is not visible on a speed test but shows up directly in your fill quality.

WiFi adds another layer of local variance before your packets even leave your building. A 2.4GHz connection in a building with other networks can add 5-30ms of jitter on top of whatever your ISP produces. For scalpers, this is the layer of free noise that is simplest to eliminate — plug in an Ethernet cable.

The honest hierarchy of improvements

A VPS at your broker's data centre is the best raw latency solution. If you run automated EAs on EURUSD at Equinix LD4, a VPS in London gives you 1-3ms round trips. The trade-off is remote desktop — your charts and interface live on a remote machine. For EA-based scalpers, that is an acceptable trade. For manual scalpers, it adds friction that often negates the latency improvement.

A proxy keeps your own setup — your charts, your keyboard shortcuts, your monitor layout — and routes only the trading traffic through a clean, dedicated path. The round-trip will not be 1ms, but it will be consistent. Consistent is what helps scalping fills. TradersProxy nodes are co-located at the same Equinix data centres as the major retail brokers.

How to measure your route for scalping
  • Ping your broker's MT server hostname for 60 seconds during your session window.
  • Watch the min-max spread, not just the average.
  • Run the same test at three times: quiet, London open, London/NY overlap.
  • If the spread widens during busy windows, ISP routing is your variable.
The scalper's routing question

A steady 45ms route is more useful than a route that swings between 10ms and 200ms.

Scalpers sometimes optimise for the best-case number when the worst-case number is what costs them fills. The question is not "what is my fastest ping?" — it is "what does my route look like at the moment I need it most?" That moment is during a session overlap or news event, not in a quiet afternoon test.

A proxy matches the answer to that question. The route is dedicated, not shared with residential traffic. The path to the broker's data centre is direct. The jitter that residential ISPs add at peak load is removed from the equation.

  • Test during peak hours, not quiet periods.
  • Ethernet eliminates local WiFi jitter for free.
  • The proxy is the middle option — better than a raw residential route, no remote desktop required.
FAQ

Under 30ms average with low variance is a reasonable target for most retail brokers. But the consistency matters more than the number — a steady 45ms route is more useful for fills than a route that swings between 10ms and 200ms.

It matters for M1 and M5 scalpers during session overlaps and news events. For traders working M15 or above with room for the market to move, route noise is rarely the primary issue.

A VPS at your broker's data centre is the best raw latency solution for EA-based scalpers. A proxy is better for manual scalpers who want to keep their own machine and chart setup while still routing trading traffic through a clean path.