Reference Case Number: FOI2025/00604
Freedom of Information: Right to know request:
Please provide data from waves 1-5 of your Crypto Assets Consumer Research, specifically on the mean and median value of crypto holdings in the UK for each year from 2020 to 2024.
Some of this data is available sporadically in FCA press releases, but I would like a continuous series of data from 2020–24 tracking both mean and median values. All on the mean and median value of crypto holdings in the UK for each year from 2020 to 2024. Some of this data is available sporadically in FCA press releases, but I would like a continuous series of data from 2020–24 tracking both mean and median values.
FCA response:
As you will be aware the mean represents the average value of holdings. Mean averages are useful when comparing different groups of consumers and when tracking changes over time. However, mean averages can be skewed due to the presence of outliers. As such, relying on mean averages alone can result in misleading results, which is why we often rely on median values. The median value is generally seen as a better indicator of the typical individual, as due to how it’s calculated, outliers tend not to skew results.
It is also worth noting that the FCA has moved away from reporting mean, or median averages from banded numeric data in its flagship Financial Lives survey (FLS). Almost all monetary value questions in the Financial Lives survey, such as amounts held in savings, ask respondents to select from pre-defined ranges rather than provide an exact figure. Our principal goal in reporting these questions in the FLS is to show the proportion of respondents selecting each range. This is standard practice in surveys and is the approach used in recent waves of our regular cryptoasset survey.
We have chosen not to report means in the FLS because calculating a mean requires selecting a single point within each range, which is not straightforward. Responses within any given range (e.g. £30,000-£50,000) may not be clustered around the midpoint, and for open-ended top bands (e.g. £50,000 or more), any assigned value would need to be based on assumptions rather than actual data. Since different choices of the value points to use in calculations can lead to different mean estimates, we have chosen to report medians instead. For FLS, we report the median band. This is the range in which the middle respondent falls and provides an objective summary of the data, without requiring an assumption to be made about the distribution of responses within each range.
Turning now to the specific information requested. We can confirm we hold survey data on the amount of crypto owned by individuals covering the following years:
• 2020 (wave 2)
• 2021 (wave 3)
• 2022 (wave 4)
• 2024 (wave 5)
We did not produce a survey for 2023, and while Wave 1 was produced in 2019, it did not ask people the value of their cryptoassets.
The below table sets out the mean and median value of crypto holdings in the UK for each year we hold this information for:
| Median | Mean | Non-response rate | |
|---|---|---|---|
| 2020 | £260 | £1,575 | 32% |
| 2021* | £300 | £1,775 | 42% |
| 2022** | £175.5 | £1,595 | 13% |
| 2023 | Not held | ||
| 2024** | £375.5 | £1,845 | 13% |
*Mean value for this year calculated based on midpoint value of indicated range of ownership
**Median and Mean values for this year calculated based on midpoint value of indicated range of ownership