Why does the same coin have a different price on every exchange?
In US equities, a broker owes you the national best bid and offer, because the venues are wired into a consolidated tape and bound by rules that make one price the reference. Crypto has no consolidated tape, no NBBO, and no regulation requiring one. Each venue is an island with its own book, and the prices differ — permanently, structurally, and for reasons worth enumerating.
Four reasons the same asset quotes differently
1. Separate books. Nothing links them. A trade on venue A does not touch venue B's order book. Prices converge only to the extent that someone actively trades both.
2. Different quote currencies. A "BTC price" may be quoted in USD, in USDT, in USDC, in EUR or in KRW. A BTC/USDT quote is a bitcoin price multiplied by a tether price. If USDT itself is trading at 0.997, then a 0.3% "bitcoin discrepancy" between a USD venue and a USDT venue is not a bitcoin discrepancy at all — it is a stablecoin discrepancy wearing bitcoin's clothes. Check the quote currency before you explain the gap.
3. Trapped capital. To sell on venue B you must already hold the asset on venue B. Moving it takes blockchain confirmations. Moving fiat takes banking hours. Capital cannot flow instantly to where the price is wrong, so the price stays wrong for as long as the flow takes.
4. Jurisdiction. Where capital controls, banking restrictions or licensing regimes fence a market off, its price can detach persistently. The Korean market's premium over global venues — the so-called kimchi premium — has at times run to several per cent for extended periods, precisely because the arbitrage required moving money across a border that restricted exactly that.
The size of a typical gap
Between two large, well-connected venues in normal conditions, majors trade within a few basis points of each other. Illustratively:
- Venue A: $60,000.00
- Venue B: $60,050.00
- Gap: 50 / 60,000 = 0.083% = 8.3 basis points
That looks like free money written on the screen. Take one taker fee of 5 bps on each side and the round trip costs 10 bps — more than the entire gap, before you have moved a single coin anywhere. The next lesson does that arithmetic properly, because "why doesn't arbitrage fix this?" has a precise answer.
What it means for your data
There is no such thing as "the price of bitcoin" the way there is a closing price for a listed share. There is a price on each venue at each instant, and every single figure you have ever seen quoted as the price is a construction — someone's rule for combining venues. Two providers using different rules produce different numbers for the same instant, and neither is wrong.
Practical consequence: when two datasets disagree, the first suspects are venue composition, quote currency and timestamp — the same where/when discipline the Foundations course installed, applied to a market that has hundreds of wheres.
In the data
The quote currency is part of the symbol, not a display setting. Here is bitcoin against the dollar and against the euro over the last year, two separate series rather than two views of one:
Each has its own previous close and its own daily change. Most crypto pairs in this data are quoted in dollars: 1,860 of 1,868, against five in pounds, two in bitcoin and one in euros (29 September 2026). Compare a euro-quoted series against a dollar-quoted one and part of what you are measuring is the euro.
Try it now
- Set the two charts above side by side and find a month where their shapes differ. That difference is the currency, not the coin.
- The table below holds both bitcoin quotes and the euro, taken at the same moment. Convert the euro price into dollars (BTC-EUR close × EURUSD close), then express the gap to the dollar quote in basis points: (converted ÷ BTC-USD close − 1) × 10,000.
- The currency leg on its own over the same year is below. The euro explains the difference between the two bitcoin charts' shapes; what it cannot explain is the residual you computed in step 2. Say how large that residual is next to the euro's own daily moves on this chart, and whether it looks like a difference in the asset's price or in the moment each quote was taken.
Build it yourself
Build a list that carries instruments from different venues, and see the problem on your own screen. Your tickers, in one call