The useful answer is short: the Coinbase Bitcoin series, labeled CBBTCUSD, is described as U.S. dollars, not seasonally adjusted, at a daily, 7-day frequency, sourced from Coinbase. That description is the starting point for any careful reading of the series. It tells you the unit, the cadence, and the publisher, and it does not tell you anything about price levels, methodology changes, or revision history.
This guide explains how to translate those labels into checkable steps. It separates what the documentation states from what you would still need to confirm elsewhere, so you do not mistake a label for a full methodology.
What the labels actually say
The series carries the unit line "Dollars, Not Seasonally Adjusted | Daily, 7-Day." Three pieces of metadata sit in that one line, and each has a practical consequence for how you read a value.
- Units: U.S. dollars. A value in this series is expressed in dollars, not in satoshis, not in index points, and not in a foreign currency.
- Adjustment: not seasonally adjusted. No statistical adjustment for recurring calendar patterns is applied. If you see a pattern that repeats weekly or seasonally, the series has not been smoothed to remove it.
- Frequency: daily, 7-day. The series is intended to carry one observation per day, including weekends.
The source line adds: "Source: Coinbase Release: Coinbase Cryptocurrencies Units: U.S." This confirms Coinbase as the originating publisher and places the series within a broader Coinbase cryptocurrencies release.
Units are a claim, not a conversion
"Dollars" tells you the denomination of the number. It does not tell you which market, which timestamp, or which side of the order book the number was drawn from. Nor does it tell you whether the value is a single trade, a consolidated print, or some other reference.
A practical check when you already hold a value: treat the unit label as the denominator, and treat everything else as unstated. If you are comparing the series to another dollar-denominated Bitcoin reference, the comparison is only meaningful once both sides share the same timestamp convention and the same frequency definition. The label alone does not guarantee that.
Why 'not seasonally adjusted' matters here
Seasonal adjustment removes recurring calendar effects so that month-to-month or week-to-week changes reflect underlying movement rather than routine patterns. For a series described as not seasonally adjusted, no such removal has occurred. Any regular rhythm you observe, whether weekly or tied to particular calendar days, is present in the raw numbers.
This is a statement about processing, not about volatility. It means you should not expect the series to have been smoothed in that specific way, and you should not read a repeating pattern as evidence of some separate adjustment.
Frequency and the 7-day cadence
The "daily, 7-day" description implies one observation per day across the full week. A checklist for anyone building a dataset from this description:
- Confirm that each day in your window carries exactly one observation, including the weekend days.
- Check whether any days are absent, and if so, whether they are documented as gaps rather than treated as zero.
- Keep the daily cadence intact when comparing with other daily series; do not resample one series to weekly and compare it with another daily series without marking the change.
- Record the observation date and the frequency label together, so a later reader can tell them apart.
None of those steps verifies the values themselves. They verify that your handling of the series matches what the documentation claims about cadence.
A worked example using only labels
Suppose a reader wants to compare this series with a separate daily dollar-denominated Bitcoin reference. The comparison plan needs four recorded fields per observation: the series identifier (CBBTCUSD), the unit (U.S. dollars), the adjustment status (not seasonally adjusted), and the frequency (daily, 7-day).
That gives you a minimum viable record. It does not give you a price, and it does not justify any statement about what the number should be. A comparison built on matching labels is a consistency check, not a valuation exercise.
What the source does not say
The documentation shown does not establish several things that a reader might reasonably want:
- It does not describe how each daily value is calculated from Coinbase market data, or which trades or quotes feed the value.
- It does not state the time of day at which an observation is captured.
- It does not state whether historical values have been revised, or how revisions are handled.
- It does not give any price level, return, or trend, and this article does not either.
- It does not grant permission to reproduce the data. The copyright notes state that reproduction of Coinbase data in any form is prohibited except with the prior written permission of Coinbase, and that the copyright information in the series notes should be reviewed before sharing.
Use and sharing constraints
The copyright statements are part of the record, not fine print to skip. The notes indicate that reproduction of Coinbase data in any form is prohibited except with the prior written permission of Coinbase, and that the copyright information in the series notes should be reviewed before sharing. Readers who publish or redistribute derived material should treat that as a permission question to resolve, not a formatting question.
Updating the series: automatic or static
The documentation offers a choice between automatic updates and a static time frame. The consequence is simple to state and easy to forget. If you pull an automatically updated series, the number changes without notice and your record of the value depends on the moment of capture. If you pull a static time frame, the record of the value is stable for that window, but you must record which window you captured.
For reproducible work, record the capture mode alongside the series identifier, unit, adjustment status, and frequency. Those five fields, all derived from what the source states, are the honest minimum for a record that a later reader could follow.
The article's scope is deliberately narrow. It covers what the labels say and the checks they imply. It does not fill in the methodology the source leaves unstated.