Accused of latency arbitrage? What records actually help

[Amanhasnoname]

The term gets thrown around loosely. The textbook version means exploiting a stale price feed at the broker, buying where the quote lags the real market. A scalper trading off a normal chart on a normal connection is doing something different, but a risk desk looking at a stream of tiny, fast winners may not see the distinction.

If a warning email lands, paperwork matters more than arguments. Records that carry weight:

- Entry and exit timestamps from the platform, with the server timezone noted
- The setup rules in writing, dated before the trades in question
- Chart evidence that the signal formed before entry, not after a quote comparison
- The losing trades too, since a history of only fast winners looks odd

One more thing worth doing is splitting results by holding time. If every winner closes inside three seconds while losers run for minutes, that distribution resembles feed exploitation even when the logic is plain price action. Better to spot that pattern in a personal review than to have a broker point it out first.

Keep the journal consistent and never tidy up old entries after the fact. Edited records are worse than messy ones.

Has anyone had to answer a broker query about trading patterns, and which records ended up mattering?