Limit-only scalping: being the one who gets paid the spread

[Fairman]

Limit-only scalping: being the one who gets paid the spread

Most retail scalpers cross the spread: buying at the ask, selling at the bid, paying the cost every time. A different approach posts limit orders and waits to be filled, earning the spread instead of paying it.

In theory, the maths is attractive. In practice, it runs into adverse selection. Limit orders fill most easily when price is moving through them, which tends to be when the trade is going wrong. The orders that do not fill are the ones that would have been winners.

Retail platforms also lack the queue priority and fast cancellation that professional market makers rely on. A retail limit strategy competes against machines with far better tools.

Still, some traders find a niche, often in calm conditions on pairs with a wide spread, using limit entries at levels where price has shown a tendency to bounce.

Measuring the real fill rate and the performance of filled versus unfilled orders is essential. A strategy that appears to work only in the filled sample may be hiding all the misses.

The honest measurement is a record of every intended limit order, whether it filled, and the result of the trade if it had filled at market instead. The comparison shows how much the strategy's profits depend on the orders that never happened.

Has a limit-based approach ever looked better in the fills than in the full set of intentions?