The Coinbase Premium is one of the few widely followed onchain-adjacent metrics that answers a genuinely specific question: is the buying pressure in Bitcoin coming from the United States, or from somewhere else? It is simple to calculate, easy to misread, and useful in a narrower set of circumstances than its popularity suggests.
For the basics of how the metric is built, see our Coinbase Premium explainer. This guide is about using it well and avoiding the usual misreadings.
How it is calculated
The Coinbase Premium is the difference between Bitcoin’s price on Coinbase Pro, quoted in US dollars, and its price on Binance, quoted in the dollar-pegged stablecoin Tether. It is usually expressed as a percentage or as a raw dollar gap, and it oscillates around zero.
The logic behind it rests on who uses each venue. Coinbase is the primary on-ramp for US retail and, historically, for a meaningful share of US institutional flow, because it operates under US regulation and connects directly to dollar banking rails. Binance serves a global, largely non-US user base transacting mostly in stablecoins. When Bitcoin trades persistently more expensive on Coinbase than on Binance, it implies that dollar buyers are lifting offers faster than arbitrageurs can level the two books, which is a reasonable proxy for US demand exceeding global demand.
What a positive and negative premium actually mean
A sustained positive premium is the interesting case. Arbitrage between two liquid venues is fast and competitive, so a gap that persists for hours or days rather than seconds means real, ongoing dollar-denominated buying is absorbing supply. That pattern has historically shown up around periods of institutional accumulation and around ETF creation activity, where authorised participants source spot Bitcoin to meet fund inflows.
A sustained negative premium implies the opposite: dollar sellers are more aggressive than global buyers, or offshore demand is leading. Neither reading is a signal on its own. Both are descriptions of where the marginal buyer currently lives.
The three ways people misread it
First, intraday noise. The premium flickers constantly and most of that movement is meaningless. Only sustained multi-day deviations carry information.
Second, the Tether assumption. Because the Binance leg is priced in USDT, the premium partly reflects whatever discount or premium USDT itself is carrying against the dollar. In calm markets this is negligible. In periods of stablecoin stress it is not, and the metric quietly stops measuring what you think it measures.
Third, treating it as a directional signal. The premium tells you the composition of demand, not its magnitude or its persistence. US buyers can be enthusiastic into a falling market, and often are.
Where it fits alongside the rest
The premium is a flow metric, and flow metrics are most useful when they corroborate something structural. A positive premium at the same time as exchange reserves are draining is a coherent story: dollar buyers are taking supply and that supply is not coming back. A positive premium while reserves build is a contradiction worth taking seriously, because it usually means the buying is being met by equally large distribution somewhere else in the market.
Pairing it with realized cap is similarly useful. Realized cap confirms whether capital is genuinely entering the network at these levels; the premium suggests where that capital is coming from. Neither answers the other’s question, and together they describe more than either does alone.
A reasonable way to use it
Check it weekly, not hourly. Look for multi-day persistence rather than spikes. Treat it as one input into a picture rather than a trigger, and be sceptical of any interpretation that requires the premium to predict rather than describe. If you are assembling a regular set of checks, the free onchain tools guide covers where to find it alongside the more structural metrics.
Common questions
Is the Coinbase Premium an onchain metric? Not strictly. It is derived from exchange order books rather than blockchain data, which is why it responds faster than genuine onchain metrics and carries less structural information.
Does a positive premium mean ETFs are buying? It is consistent with that but does not prove it. ETF creation activity is one plausible source of persistent dollar buying; ordinary US retail and corporate treasury purchases are others.
How large a premium is meaningful? There is no fixed threshold. What matters is the size relative to its own recent range and how long it persists. A small premium holding for a week says more than a large one lasting an hour.
Leave a Reply