forex-scalping.com

Price feed aggregation and why two quotes at the same moment differ

Retail brokers do not make prices. They aggregate quotes from several liquidity providers, apply a markup, and publish a single feed. Different brokers use different providers, different aggregation rules, and different markup logic, so the same second of the same pair can show visibly different bid and ask values.

Differences of 0.1 to 0.3 pips are normal in quiet markets. During fast moves, one feed may lag by 200 ms and show a stale price that the other already left behind.

For scalpers, this implies two things. First, a strategy tested on one broker's data may behave differently on another's, particularly for tick-level logic. Second, the feed can be a source of edge or drag. A slow feed means signals arrive late, and a feed with filtering may smooth away the very ticks a system reacts to.

Compare feeds by logging ticks side by side for an hour during a volatile window. The lag and the spread differences will be obvious in the data.

Has anyone noticed a feed that consistently led or lagged another one?

I ran two terminals side by side for about three weeks, one at my main broker and one at a smaller ECN, and logged ticks on EURUSD during the London open. The ECN feed led more often than not, by roughly 100 to 150 milliseconds on fast moves, and the main broker's quotes looked smoother, almost as if it was filtering a few ticks out.

The practical effect showed up in my entry timing. A breakout signal that fired on the ECN feed would sometimes fire on the other broker one or two ticks later at a worse price, and that cost me around half a pip on average. Not huge, but for a scalper aiming for six or seven pips it adds up quickly.

What I did was stop optimizing on one feed and then trading on another. I now export tick history from the broker I actually trade with and backtest only on that. When the two feeds disagree by more than a pip in quiet conditions I treat it as a warning sign and skip the signal.