Liquidity Filters in Options Backtests

Abstract

Liquidity filters are not there to make a backtest look conservative. They define whether the strategy could plausibly trade the contract. For options, that means spreads, quote age, size, open interest, volume, DTE, and price guards.

Developers should treat liquidity as part of the strategy contract, not a cleanup step after results are known.

Filters That Matter

The basic filters are quote freshness, max spread percentage, minimum bid or ask, open interest, volume, and contract price. More advanced filters can include delta range, expected move, intrinsic value, and minimum number of valid quote observations.

The key is consistency. If the filter is part of research, it should also exist in paper trading.

Liquidity Can Change The Winner

A strategy family may produce a strong stock-signal winner that fails once option liquidity is enforced. Another branch may have lower raw return but better quote quality and fewer rejects.

That second branch is often the better paper candidate. The market does not pay for untradable theoretical edge.

Avoid Post-Hoc Filter Mining

Liquidity filters can be overfit too. If a threshold is chosen only because it rescues the winner, it should be treated with suspicion. Prefer round, explainable thresholds and then test robustness around them.

Takeaway

Liquidity filters are part of the evidence standard for options backtesting. They should be explicit, stable, logged, and carried forward into paper validation.