Bid-ask spread explained: bid, ask, mid price and spread in bps

The bid-ask spread is the difference between the highest price a buyer is willing to pay (the bid) and the lowest price a seller will accept (the ask). Open any quote from a market data API and you get both numbers. On a Monday morning in late September 2026, EURUSD quoted 1.136965 to sell and 1.137005 to buy: a spread of 0.00004, which forex traders call 0.4 pips.

That gap tells you how easy it is to trade at this moment, what a round trip costs before any fee, and whether the quote in front of you is healthy. This guide covers how to express the spread so markets can be compared, why it changes through the day, what NBBO and slippage mean, and how to track it in a few lines of Python.

Bid vs ask: two prices, one market

If you want to buy right now you pay the ask (also called the offer). If you want to sell right now you receive the bid. The mid price halfway between them is what most charts draw, but nobody actually trades at the mid. Every quote endpoint returns both sides and the size available at each:

{
  "symbol": "EURUSD",
  "bid": 1.136965,1
  "ask": 1.137005,2
  "bid_size": 85,3
  "ask_size": 61,
  "timestamp": 17905911140024
}

A quote is not a trade. A quote is an offer that is still standing; a trade is a deal that already happened. The last trade price can sit anywhere inside, or briefly outside, the current bid and ask. The OHLC explainer shows how trades become the bars on a chart, and why a bar never tells you what the spread was.

The arithmetic: pips, cents and basis points

Forex traders usually quote spreads in pips. For most pairs a pip is the fourth decimal (0.0001); for yen pairs it is the second (0.01), because the yen trades at a much larger number per dollar. Our pip explainer has the full rule and a calculator. Stocks quote spreads in cents, and crypto in whatever the quote currency is.

None of those units compare across markets. Divide the spread by the mid and express it in basis points instead: one basis point is 0.01%, and a 2 bps spread costs the same fraction of a trade whether the instrument is priced at 1.13 or at 82,000. Here are five live quotes taken a few seconds apart:

The last row is the interesting one. The same stock that trades a cent or two wide in the middle of a regular session was quoting 19 cents wide before the open: seventeen times wider than gold and sixty times wider than EURUSD in relative terms. Nothing was wrong with the data; there were simply very few people quoting it at 06:25 in New York. The after-hours and pre-market guide goes deeper on why.

How wide is normal?

There is no universal "good" spread. It depends on how many people trade the instrument and how much risk a market maker takes by standing in the middle. The most traded currency pairs and the largest stocks sit at a basis point or less; thinly traded shares, exotic currencies and small coins can be fifty times wider.

Why spreads widen at night and at the rollover

The same pair can be tight at 14:00 UTC and wide at 22:00 UTC. Liquidity follows people: when London and New York are both open, many banks and funds quote currencies and competition squeezes the spread. When only one region is awake, fewer quotes compete and the gap opens up. The forex market hours guide has a live session clock.

The widest moment of the forex day is usually the rollover at 17:00 New York time, when dealers settle positions and roll them to the next value date. For a few minutes liquidity thins out and spreads can jump to several times their daytime width. If your alerting fires on price moves, it will fire at the rollover for reasons that have nothing to do with the market.

Stocks follow the same logic on a different clock. Spreads are widest in the first minutes after the open, while the order book is still filling, and in pre-market and after-hours sessions. Crypto trades all week, so it has no open, but weekend nights are visibly thinner than weekday afternoons.

NBBO: the best bid and offer across US venues

A US stock trades on many venues at once, and each has its own best bid and ask. The national best bid and offer (NBBO) is the highest bid and the lowest ask across all of them. It is the reference brokers are measured against, and the tightest spread anyone can see for that stock at that instant.

TickerLayer quotes are derived from aggregated sources rather than being the official NBBO, so we publish how close they come: the US stocks data quality benchmark compares our quotes with a consolidated US market reference, state for state, including the share of samples where our midpoint sits inside the NBBO.

Slippage: when the spread is not the whole cost

The spread is the cost of trading a small amount. The size fields tell you how small. The pre-market US:KO quote above showed 200 shares at the 88.30 ask. Buy 1,000 shares and the first 200 fill at the ask; the rest fill at worse prices deeper in the book. That difference between the price you expected and the average price you got is slippage.

Quote data does not show the levels behind the best price, but it tells you when to worry: when your order is several times the displayed size, or when the spread is already wide, expect to pay more than the ask.

Track the spread yourself in Python

You need a key and one endpoint. GET /{asset}/quote/{symbol} returns the current bid and ask for any asset class, with the same shape for forex, stocks, crypto, indices, ETFs and commodities. This script prints the spread in basis points for a few instruments:

import os
import requests

API = "https://api.tickerlayer.com"
HEADERS = {"x-api-key": os.environ["TICKERLAYER_API_KEY"]}


def spread_bps(asset: str, symbol: str) -> float:
    r = requests.get(f"{API}/{asset}/quote/{symbol}", headers=HEADERS, timeout=5)
    r.raise_for_status()
    q = r.json()
    bid, ask = float(q["bid"]), float(q["ask"])
    mid = (bid + ask) / 2
    return (ask - bid) / mid * 10_000


for asset, symbol in [
    ("forex", "EURUSD"),
    ("forex", "USDJPY"),
    ("commodities", "XAUUSD"),
    ("crypto", "BTCUSD"),
    ("stocks", "US:KO"),
]:
    print(f"{symbol:<8} {spread_bps(asset, symbol):7.2f} bps")

EURUSD      0.35 bps
USDJPY      0.45 bps
XAUUSD      1.25 bps
BTCUSD      0.00 bps
US:KO      21.54 bps

Polling is fine for a spot check. To watch a spread change tick by tick, subscribe to the quote channel over WebSocket instead. Stream frames carry the same bid and ask, sent as strings on the forex, crypto and stocks channels, so convert them before doing arithmetic. The WebSocket market data guide walks through connecting, subscribing and reconnecting, and the exchange rate API guide covers forex quotes in depth.

What the spread tells your code

• The cost of tradingCrossing the spread is a real cost. A strategy that makes 1 bp per trade on a 2 bps spread loses money before fees.
• A liquidity gaugeA spread three times its usual width means the book is thin. Size down, or wait.
• A quality checkZero, negative or absurd spreads point at a stale or broken quote. Drop it instead of plotting it.
• A calmer alertAlert on the mid, but only when the spread is normal. Most false alarms happen when it is not.

The quality check deserves code of its own. A crossed quote (bid above ask) or a locked one (bid equal to ask) is not a price anyone can trade. It is usually a moment where the two sides were updated at slightly different times, and a reason to wait for the next update. The bad ticks guide extends this into a full stale-price and spike filter.

def usable(quote: dict, max_bps: float = 50.0) -> bool:
    bid, ask = float(quote["bid"]), float(quote["ask"])
    if bid <= 0 or ask <= 0 or ask <= bid:
        return False  # empty, crossed or locked
    mid = (bid + ask) / 2
    return (ask - bid) / mid * 10_000 <= max_bps