How to Measure the Bid-Ask Spread

Direct answer

To measure the bid-ask spread, subtract the best bid from the best ask, then divide by the midpoint and multiply by 100:

spread %  = (best ask − best bid) / midpoint × 100
midpoint  = (best ask + best bid) / 2

With a best bid of 99.50 and a best ask of 100.50, the absolute spread is 1.00, the midpoint is 100.00, and the spread is 1%. Always compare spreads as percentages — absolute spreads are meaningless across assets at different price levels.

Step 1 — Read the top of the book

You need exactly two numbers:

• Best bid — the highest price anyone is currently offering to pay.
• Best ask — the lowest price anyone is currently offering to accept.

These sit at the top of each side of the order book, closest to each other.

Step 2 — Calculate the absolute spread

absolute spread = best ask − best bid

Example: 100.50 − 99.50 = 1.00

This number is only useful within a single market. A spread of 1.00 is enormous on an asset priced at 2 and negligible on one priced at 60,000.

Step 3 — Calculate the midpoint

midpoint = (best ask + best bid) / 2

Example: (100.50 + 99.50) / 2 = 100.00

The midpoint is the conventional reference price. It is not a price you can trade at — it is the theoretical centre between the two prices you can trade at.

Step 4 — Express as a percentage

spread % = absolute spread / midpoint × 100

Example: 1.00 / 100.00 × 100 = 1%

This is the comparable form. Now you can compare this market against any other, regardless of price level.

Some sources divide by the best bid or the best ask instead of the midpoint. The difference is small at tight spreads and material at wide ones. Use the midpoint, and state your convention when publishing a figure.

Step 5 — Measure the effective spread at your size

The step that separates a useful measurement from a decorative one.

The quoted spread applies to the quantity available at the top of the book, which is often small. Your effective spread is what you would actually pay to enter and exit at your size.

To calculate it:

1. Walk the ask side, accumulating quantity until you reach your order size. Record the average price. Call it buy_avg.
2. Walk the bid side the same way. Call it sell_avg.
3. Then:

effective spread % = (buy_avg − sell_avg) / midpoint × 100

Worked example

Quoted spread looks excellent: 0.999 / 1.001, so 0.2%.

But the book holds only 500 tokens on each side at those prices, and you want 5,000.

Buying 5,000: (500×1.001 + 1,500×1.010 + 3,000×1.030) / 5,000 = 1.0210 Selling 5,000: (500×0.999 + 1,500×0.990 + 3,000×0.970) / 5,000 = 0.9790

effective spread = (1.0210 − 0.9790) / 1.00 × 100 = 4.2%

The quoted spread was 0.2%. Your real cost is 4.2% — twenty-one times worse. The quoted figure described a market that did not exist at your size.

Step 6 — Sample over time

A single spread reading is a snapshot. Spreads widen during volatility, around news, and in thin hours.

For any decision that matters, sample repeatedly across a period and record the range and typical value, not one observation. A pair with a 0.3% spread most of the time and 5% during every move is a different proposition from one steady at 1%.

What is a normal spread?

There is no universal answer, but rough expectations for reference:

Treat these as orientation, not benchmarks. The right comparison is against the same pair over time, and against your own effective spread at size.

Common mistakes

• Comparing absolute spreads across assets at different prices.
• Quoting the top-of-book spread as though it applies at any size.
• Taking one reading. Spreads move constantly.
• Reading a tight spread as a liquid market. A 0.02% spread on 5 tokens is decoration. See market depth guide.
• Forgetting fees. The spread is the cost of crossing the market; the exchange fee is charged on top.

Related

• Market Depth Guide — the other half of liquidity.
• Understanding Slippage — what effective spread becomes at size.
• What Is Liquidity — why both measures are needed.