Broker tick feed or third-party tick data: which one belongs in a scalping test?

[Shadow Trader]

Both options have a case, and they answer different questions.

A broker feed, where the platform lets you export or record it, reflects the prices that broker actually quoted. For a scalper who trades with that broker, this matters a lot. Spread behavior around rollover, the way quotes thin out before news, the occasional stale quote: all of it is embedded. The downside is coverage. Brokers often keep limited history, sometimes only what the platform cached, and the earlier years can be reconstructed from lower-resolution data rather than true ticks. Depth of history is usually the weak spot.

Third-party data tends to go back much further and is often assembled from liquidity provider or interbank sources. It is clean, consistent and easy to share, which makes results reproducible. The catch is that it is not what any single retail broker showed. Spreads in the file may be a generic model, and the quote path during fast moves can differ from what a given broker delivered, since each broker filters and marks up its own stream.

A practical split looks like this:

- Use third-party data for broad strategy research, where the question is whether the idea has any structure at all across many years and regimes.
- Use the broker's own recorded feed for final validation, where the question is whether the idea survives the specific costs and quote behavior of the account.

One more thing worth doing: compare both sources over the same week. Count how often the highs and lows differ by more than the strategy's stop distance. For a 2 pip scalp, differences that look tiny on a chart can flip dozens of outcomes.

Whichever source is used, it helps to know how it was built. Ask the provider whether quotes are raw or filtered, and whether spreads are real or modeled.

Which source do you lean on for final validation, and has the two-feed comparison ever changed a conclusion for you?