Coinbase premium data: U.S. bid versus global price

What Coinbase Premium measures.

Coinbase Premium is the dollar gap between BTC/USD on Coinbase Pro (US institutional venue) and BTC/USDT on Binance (offshore reference venue), normalized for any USDT-versus-USD basis. A positive premium means US buyers are willing to pay above the offshore market — typically because they cannot easily settle on offshore venues. A negative premium means the offshore market is more aggressive, often during Asian session selling or US risk-off sentiment.

The metric was popularized by CryptoQuant in 2020 and remains the cleanest single read on US institutional spot demand. It is now widely cited by The Block, CoinDesk and Glassnode analysts in their market commentary. The premium can stay persistently positive or negative for weeks at a time, and those regimes carry information about which side of the market is leading.

Why this is an institutional flow signal.

US institutions face regulatory barriers to offshore venues. After the 2023 CFTC and SEC actions against Binance and its US affiliate, institutional capital became effectively locked into compliant US venues — Coinbase, Kraken, and the spot ETF complex. When that locked capital wants to buy aggressively, Coinbase order books absorb the demand and the spot price rises above Binance's offshore price. That spread is the premium. The mechanism makes it a cleaner read on US institutional positioning than aggregated volume.

ETF authorized participants are a particularly important flow source. When BlackRock IBIT or Fidelity FBTC see large net inflows, the APs need to acquire spot BTC on regulated venues — primarily Coinbase, which serves as custodian for most of the spot ETF complex. SoSoValue and Farside Investors track these flows daily, and historical correlation between ETF inflows and Coinbase Premium runs above 0.6 since the January 2024 ETF launch.

How to read the number.

1. Sustained positive premium (+$50 to +$200).

This regime indicates US institutional accumulation. It correlated strongly with the November 2023 - March 2024 rally that preceded and accompanied ETF launches. When premium holds above +$50 for more than three sessions, treat it as a structural risk-on signal worth cross-checking with ETF flow data and stablecoin issuance.

2. Neutral band (within ±$30).

The modal state. Markets are pricing US and offshore demand at parity. No strong signal in either direction; the dashboard adds little to the trading thesis in this state.

3. Sustained negative premium (-$30 to -$50, extreme to -$80).

US capital is exiting or sitting on the sidelines while offshore venues lead. The August 2024 yen-carry unwind saw premium briefly touch -$120, and the November 2022 FTX collapse drove premium to -$90 for nearly two weeks. Sustained negative premium during a price rally is a quiet warning — the rally lacks US institutional confirmation.

4. Sudden sharp moves (intraday doubling or sign flip).

These usually mark news events. The March 2024 spot ETF approval saw premium swing from +$30 to +$180 within four hours. Sharp moves should be cross-checked with the wire — Bloomberg Crypto and CoinDesk usually have the news within 15 minutes. Trading off the premium move before identifying the catalyst is reading shadows, not light.

5. Magnitude versus percentage.

Premium is quoted in dollars, not percentage. A $100 premium at BTC $100,000 is 0.1% — a meaningful but not extreme move. A $100 premium at BTC $30,000 (as occurred in late 2022) was 0.33%, three times more meaningful. Always convert to percentage when comparing across price regimes.

6. Weekly rhythm versus real signal.

Premium has a weekly pattern: Asian session (Coinbase off-hours) often shows mild negative premium, US session often shows mild positive premium. The signal is in deviations from that baseline, not in the raw number at any given hour. Three-session moving average filters out the rhythm.

Four practical use patterns.

Use 1 — ETF inflow confirmation.

SoSoValue and Farside Investors publish ETF net flow data with a one-day lag. Coinbase Premium reacts in real time. If ETF flow turns positive and premium expands the same session, the signal is high-confidence. If premium does not respond, suspect the flow data and wait for confirmation.

Use 2 — Top warning.

Premium often peaks before price peaks. The November 2021 top was preceded by premium collapse from +$150 to -$30 in the two weeks before BTC's actual high. A pattern of price making new highs while premium declines (premium divergence) is the cleanest top warning this dashboard can produce.

Use 3 — Bottom verification.

Sharp negative premium often marks panic selling. When premium hits -$80 or below and starts compressing back toward zero, US institutional buying is returning. This sequence preceded each of the 2022-2023 bear-market local bottoms verified after the fact.

Use 4 — Cross-indicator confirmation.

Premium pairs naturally with funding rates and Coinglass open interest. Positive premium plus low funding plus rising OI is a clean accumulation profile. Negative premium plus high funding plus falling OI is a clean distribution profile.

Historical case studies.

Comparison with Kimchi Premium.

Kimchi Premium measures the Korean exchange (Upbit, Bithumb) versus offshore reference price. It serves the same function as Coinbase Premium but for the Korean retail market, which is structurally siloed by Korea's capital controls. The two premia frequently disagree: Coinbase Premium tracks US institutional flow, Kimchi Premium tracks Korean retail euphoria. Reading them together gives a global flow picture — both positive is broad demand, both negative is broad capitulation, divergence reveals regional rotation.

Companion tools and further reading.

• Funding rate guide — derivatives positioning that often diverges from spot demand.
• Whale tracking — large wallet behavior that often precedes premium expansion.
• Stablecoin market cap — the dollar liquidity that funds premium-driving demand.
• Long-short ratio — positioning sentiment cross-check.

Data methodology — how this dashboard is computed.

The dashboard pulls BTC/USD price from Coinbase's public ticker endpoint every 30 seconds and BTC/USDT price from Binance's public ticker endpoint at the same interval. USDT-to-USD basis (rarely material but occasionally meaningful) is fetched from CoinGecko's USDT ticker. The premium is computed as (Coinbase BTC/USD) minus (Binance BTC/USDT × USDT/USD). The 90-day historical series is rebuilt daily from each venue's market chart endpoint.

Known limitations.

One: weekend and US-holiday data is less reliable because Coinbase order book depth thins out. Two: the premium can be transiently distorted by exchange-specific events (a Binance withdrawal pause, a Coinbase outage). Three: the published CryptoQuant Coinbase Premium Index uses a slightly different basket and may print 5%-10% differently in absolute terms — direction and trend agreement is high, absolute level agreement is partial. Four: during ETF outflow days, premium can stay negative while market structure is otherwise healthy; treat it as one input, not the whole picture.

Frequently asked questions.

How often does the data refresh?

Live ticker every 30 seconds; historical series rebuilds daily at 00:00 UTC.

Why not just use CryptoQuant's official index?

CryptoQuant's index requires a paid subscription for real-time access. This dashboard provides a free public-data approximation that tracks the official index direction with high correlation. For audit-grade trading decisions, the CryptoQuant index is the authoritative source.

Does positive premium always mean buy?

No. Premium tells you US institutions are buying aggressively, not whether you should follow. The signal works best as confirmation of an existing thesis, not a standalone trigger.

Can you show premium for other coins?

The dashboard currently tracks only BTC. ETH premium tracks similarly during normal regimes but diverges meaningfully during ETH-specific events (Shapella upgrade, ETF approval timeline, Layer 2 narrative shifts).

Can historical data be downloaded?

Not directly from this dashboard. Both Coinbase and Binance public APIs serve their respective price histories; reconstructing the premium series locally is straightforward.

The dashboard and CryptoQuant disagree — who is right?

Neither is wrong; they use different bases. CryptoQuant typically averages multiple offshore venues; this dashboard uses Binance only. Use whichever methodology is documented and stable for your purpose.

Premium near zero for weeks — is the indicator broken?

No. Extended neutral bands occur when US and offshore demand are balanced. This is information itself — typically corresponding to consolidation regimes.

Does the chart distort on mobile?

The 90-day chart uses responsive SVG with viewBox preservation; absolute values display correctly on mobile but the visual density of recent data may compress. Tablet or desktop is recommended for reading the chart at higher resolution.

Related reading.

For the historical context of premium signals across cycles, the CoinDesk "On the Margin" series and The Block's quarterly institutional flow report are both worth bookmarking. Glassnode's "Week On-Chain" newsletter often references Coinbase Premium when discussing US-led versus Asia-led market phases.

How to combine premium with on-chain reads.

Coinbase Premium is a venue-level signal; on-chain metrics are a global supply-side signal. The two layers tell different stories and combining them yields a cleaner read than either alone. When premium is positive and Glassnode's "Realized Cap HODL Waves" shows long-term holder supply expanding, the implication is institutional accumulation meeting holder conviction — a high-confidence accumulation regime. When premium is positive but HODL waves show long-term holders distributing, the implication is institutional buying being absorbed by smart-money selling — a much less convincing setup.

Coinglass open-interest data adds a third layer. OI rising alongside positive premium is healthy bid expansion; OI rising while premium falls is leveraged speculation getting ahead of the spot market — a fragile configuration that has historically resolved with sharp deleveraging events. The August 5, 2024 unwind was an extreme example: OI had built up through July while premium was already turning negative by late July. The setup was visible in real time on this dashboard cross-checked with the Binance and Bybit OI feeds.