Bid-ask spread

How it's measured

Spread = ask − bid

To compare instruments at different prices, express it as a percentage of the midpoint:

Spread % = (ask − bid) ÷ ((ask + bid) ÷ 2) × 100

Some sources divide by the ask or the last price instead. For tight spreads the difference is negligible; for wide ones it isn't.

Why it matters

A round trip that buys at the ask and sells at the bid loses the full spread before price moves at all, which makes the spread the most basic source of slippage. For short-term trades with small targets, a wide spread can consume much of the expected move. Spreads widen when liquidity drops: before the open, after the close, around news and after halts. A wide spread also makes chart prices less reliable, because the last trade may have printed at either edge of the quote.

Example

For example, a hypothetical stock is quoted $4.96 bid and $5.04 ask. The spread is $0.08 on a midpoint of $5.00, or 1.6%. Buy at $5.04 and, if the quote doesn't move, you can only sell at $4.96. A trade aiming for a 3% move has given up more than half of it to the spread.

Common mistakes

• Watching the last price instead of the bid and ask.
• Assuming the spread at 08:00 will still be there at 09:30, or the other way round.
• Ignoring the size quoted: a tight spread with 100 shares on each side may not hold for your order.

How PatternX uses it

On chart-pattern signals, PatternX's universal filter rejects any setup whose spread is wider than 0.25%, and spread is one of the 13 factors in the PatternX Score (weight .05). In Momentum Lab, the maximum spread defaults to 1% of mid; a stock without a two-sided quote fails the check, and crossed quotes are rejected by the data-quality gate. The stock in the example above would be skipped at the default. Momentum Lab backtests assume a 0.2% spread. For the premarket side, see premarket liquidity and spreads.