Mean reversion on seconds charts and the spread filter that decides everything

[Amanhasnoname]

Fading small spikes on very short timeframes can work. Price overshoots, liquidity providers adjust, and the move often retraces within seconds.

The catch is cost. If the retrace averages 1.5 pips and the spread plus slippage is 1 pip, the edge is mostly gone. On a quiet day with a 0.2 pip spread, the same trade has room. During a wide-spread period, it has none.

That's why a spread filter often matters more than the entry signal. A simple rule such as only taking trades when the spread is below a fixed fraction of the expected move turns a marginal idea into a tradeable one.

Another issue is trend days. Mean reversion works until a larger move starts, and then every fade becomes a loss. A volatility filter or a limit on consecutive losses is a sensible safeguard.

Strategy designers sometimes spend months on the entry and a few minutes on the filters, which is backwards.

Which filters have made the biggest difference to your short-term reversion setups?