Broker bridge latency - Glitch Executor

Bridge latency is the time between Glitch Executor emitting an order intent and the broker acknowledging it. Round-trip (RTT) varies by execution path; the right path depends on the strategy class.

Latency profile by path

Path	Typical RTT	Region variance	What’s in the path
cTrader OpenAPI (OAuth)	20–40 ms	±15 ms regional	Glitch Executor → cTrader Open API → broker
TradeLocker REST	40–80 ms	±20 ms	Glitch Executor → TradeLocker auth + REST → broker
DXtrade REST	50–90 ms	±25 ms	Same shape as TradeLocker
MT4/MT5 via MetaApi	80–150 ms	±50 ms (regional)	Glitch Executor → MetaApi → MetaApi MT terminal → broker
Broker-direct EA (VPS)	5–15 ms	minimal	EA on same-region VPS → broker directly

Why MetaApi is slower

MetaApi runs the MT4/MT5 terminal in their cloud. Your order has to travel from Glitch Executor → MetaApi → MetaApi’s MT terminal → the broker. Every hop adds ~20–40 ms. The trade-off is operational simplicity: no VPS, no MT terminal to maintain, full API access to the MT account.

When latency matters

Strategy class	Latency tolerance	Acceptable path
Tick scalping (sub-1-second hold)	< 20 ms	Broker-direct EA only — banned by firm policy anyway
Standard scalping (5–60 sec)	< 50 ms	cTrader OpenAPI, broker-direct EA
Intraday momentum	< 200 ms	Any of the bridges
Swing trading	< 1000 ms	All paths fine
Multi-day algo	< 5 sec	All paths fine

For scalping strategies, 100 ms of latency on a fast-moving instrument can move the price 0.5–2 pips against you — eating 10–40% of average winner size. For swing strategies, 100 ms is invisible against multi-pip moves over hours.

Regional considerations

MetaApi’s regional endpoints (us-east-1, eu-west-2, ap-southeast-1) determine baseline latency. Deploy Glitch Executor orchestration in the same region as both MetaApi and the broker for the lowest end-to-end RTT. cTrader OpenAPI servers are clustered in EU; us-east clients see ~40–50 ms baseline; ap clients see 100+ ms. TradeLocker is hosted regionally with the broker so latency follows the broker’s region, not TradeLocker’s globally.

How to measure

The Glitch Executor dashboard exposes per-account bridge-latency metrics (95p RTT over the last hour). For a deep-dive:

Take 100 dummy entry-then-cancel orders during a normal market session.
Record the timestamp of intent emission and broker acknowledgement.
Compute p50 / p95 / p99 RTT.
Compare against the table above; investigate if 30%+ higher than typical.

What latency is NOT solving

A high-latency bridge slows execution but doesn’t change firm rules. A strategy that’s profitable at 100 ms latency stays profitable at 200 ms — it just gives back more on slippage. Don’t blame latency for unprofitable strategies; blame edge. See related: MetaApi internals, EA / cBot policy, brokers overview.

​Latency profile by path

​Why MetaApi is slower

​When latency matters

​Regional considerations

​How to measure

​What latency is NOT solving