Many venues, one headline number
A coin can trade on dozens or even hundreds of exchanges and pools simultaneously, each with its own recent trades and resting orders. Yet when you open a price website or app, you usually see one figure. That figure is an index or aggregate price: a calculated reference value built from many sources, rather than a quote from any one market.
The reason for combining sources is practical. Each venue is a partial view, and some are thin or temporarily out of line. A blend is usually steadier and more representative than any single feed, although it is still a model with choices built in.
The basic steps
Providers differ in details and often do not publish every one of them, but most follow a similar outline.
- Choose sources. Decide which venues and trading pairs to include. Some providers screen for reliability, regulatory standing, data quality or trading activity.
- Collect prices. Pull recent trades or best bid and ask quotes from each source, often many times per minute or faster.
- Convert to a common currency. Venues quote against different currencies, including US dollars, dollar-pegged tokens, and others. The provider converts each into one reference currency, which relies on exchange rates that themselves have uncertainty. This is one consequence of the way crypto prices are quoted against dollar tokens.
- Clean the data. Remove stale prices, obvious errors and outliers.
- Weight and combine. Compute a single number from what remains.
- Publish and update. Show the result and refresh it regularly.
Common ways to combine
Simple average. Add up the prices and divide by the count. This is easy but gives a tiny venue the same influence as a huge one.
Volume-weighted average. Give each venue influence in proportion to its trading volume. The idea is that busier markets say more about the going price. The weakness is that the weights depend on reported volume, and volume figures are not always reliable, as discussed in the article on trading volume.
Median. Take the middle value of the prices. This is robust against a single venue showing an extreme number.
Liquidity-based weights. Some methods weight by order book depth or by other measures of how much can really be traded near the price, rather than by reported volume.
A small illustration with invented numbers. Three venues report these prices and volumes:
| Venue | Price | Volume |
|---|---|---|
| A | 100.0 | 50 |
| B | 100.2 | 30 |
| C | 99.8 | 20 |
- Simple average: (100.0 + 100.2 + 99.8) / 3 = 100.00
- Volume-weighted average: (100.0 * 50 + 100.2 * 30 + 99.8 * 20) / 100 = (5,000 + 3,006 + 1,996) / 100 = 10,002 / 100 = 100.02
- Median: 100.0
All three land close together here because the inputs agree. They would diverge if one venue were far off, and that is where method choice matters.
Dealing with bad data
Aggregators have to guard against several problems.
- Outliers. A venue showing a price far from the rest is more likely to be thin or malfunctioning than correct. Typical defences include discarding values beyond some distance from the median.
- Stale feeds. If a venue stops updating, its last price should drop out of the calculation after a while.
- Thin markets. Small venues can be moved by tiny trades, so they are often given less weight or excluded.
- Manipulation. Someone could try to move the price on a minor venue to distort an index that uses it. Weighting, filtering and wide source lists make this harder but do not eliminate the possibility. See the article on price manipulation for more on this theme.
- Faulty volume. If reported volume is inflated, volume-based weights can be skewed.
Why two sites disagree
Because of these choices, two reputable websites can show slightly different prices for the same coin at the same moment. They might include different venues, weight them differently, update at different speeds, or use different currency conversions. Small differences are normal. A large, persistent gap is worth investigating, for example by checking which pair and which venues each site uses.
The same logic explains why a chart from a single exchange can differ from an index. The exchange chart shows one marketplace, a point explained in the article on how crypto exchanges set prices.
Reference prices in other products
Index prices are used beyond display. Derivatives platforms often use an index to calculate funding payments and to decide when positions are liquidated, because a robust reference is harder to push around than a single venue’s last price. Funds and other products may use benchmark rates for valuation. In each case the method matters, which is why reputable providers usually document how their index is built.
What an index is not
An aggregate price is a reference, not something you can necessarily trade at. Your own fill depends on the venue you use, its order book, fees and your order size, which is the topic of slippage. It also does not predict future movement.
Cryptocurrency prices are volatile and you can lose some or all of the money you invest. This article is general education and not financial, tax or legal advice.