Partial fills, and why a scalper should care even in retail

[Fairman]

Partial fills, and why a scalper should care even in retail

Retail platforms advertise full fills, and most of the time that is what shows up. But on larger sizes, or during thin liquidity, some brokers split a market order into several chunks at slightly different prices.

The platform then shows a single averaged entry. That average hides the fact that the last chunk may have filled several tenths of a pip worse than the first.

For a one-lot trade on EURUSD this is invisible. For a size that is large relative to the book on a minor or exotic pair, it becomes a real part of the cost. The deeper problem is the exit. Closing the same size produces the same splitting, so the cost doubles quietly.

A scalper whose target is two pips can lose a quarter of the target before the market even moves.

Two habits help. First, check the deal history, not just the position window, since individual fills are listed there with separate prices. Second, test by sizing up gradually on demo or a small live account and watching where the average entry starts drifting from the quoted price.

The size where fills begin to split is a useful number to know.

A good way to find the threshold is a ladder test on a quiet afternoon: place trades of increasing size, say 0.5, 1, 2, 5 and 10 lots on a pair, and note how each deal appears in the history. The first size at which multiple deals show up marks the practical depth available at the broker. It differs by pair and by time of day, so repeating the test at a busy hour and a quiet hour gives a fuller picture. Results from the test belong in the cost model alongside spread and commission.

At what lot size did fills start looking different on the account being traded, and was it a gradual change or a sudden step?