Cleveland Fed Paper: Crypto Buying Tracks Beliefs, Not Demographics
A working paper from the Federal Reserve Bank of Cleveland argues that Americans buy cryptocurrency because of what they believe about its future returns — not primarily because of their age, income or risk tolerance. Two outlets, Cointelegraph and Crypto Briefing, published independent accounts of the paper on August 23, 2026. Read together, they confirm the study’s core claim while diverging sharply on the granular numbers behind it.
The paper, titled Do You Even Crypto, Bro? Cryptocurrencies in Household Finance, was authored by a team that both outlets say includes economists Michael Weber and Bernardo Candia; Cointelegraph additionally names co-authors Olivier Coibion and Yuriy Gorodnichenko. Crypto Briefing reports the paper was released on July 14, 2026 — a detail Cointelegraph’s account does not mention.
What both outlets confirm
Cointelegraph and Crypto Briefing agree on the paper’s central mechanism: expectations about crypto’s future returns explain more of the variation in who owns it than a broad set of demographic characteristics. Both describe this as unusual compared with traditional assets like stocks, bonds and gold, where demographic and financial characteristics normally carry more explanatory weight.
Both outlets also describe an embedded randomized experiment in which researchers gave some survey participants information about Bitcoin’s historical returns while a control group received none. In both accounts, the group that received the information increased its desired allocation to crypto, and that intention translated into actual purchases in later survey waves.
On demographics, both outlets describe crypto holders as skewing younger and male relative to non-holders. Both also report that crypto holders perceive digital assets as less risky than non-holders believe them to be.
On the wealth-effect finding, both outlets report that rising Bitcoin prices are associated with increased spending on durable goods — cars, appliances, furniture — among existing holders, but that this effect does not carry over into ordinary day-to-day spending. Cointelegraph reports the researchers likened the pattern to
“gambling income”
rather than a lasting increase in wealth.
Where the numbers diverge
The two outlets part ways sharply once the story moves from qualitative findings to specific figures. Crypto Briefing’s account contains almost no numerical detail beyond a single figure — a 35% viewability reference embedded in its own site metadata, unrelated to the study’s findings — leaving most of the paper’s quantitative results reported by Cointelegraph alone.
According to Cointelegraph, the paper drew on repeated surveys of as many as 25,000 US households per wave. Cointelegraph reports that in the researchers’ 2021 survey, 87% of non-owners said they did not know what return to expect from crypto over the following year, compared with 54% among owners. Among those willing to forecast a number, Cointelegraph reports crypto owners expected an average 22% return over the following year, versus 7% among non-owners.
Cointelegraph also reports that each additional percentage point of expected return raised the odds of owning crypto by 0.8 percentage points, according to the paper’s findings, and that people under 40 were 13 percentage points more likely to own crypto than people over 60, with men about 4 percentage points more likely than women to hold it.
On the information experiment specifically, Cointelegraph reports that showing participants Bitcoin’s return over the prior year lifted their desired crypto allocation by about 2 percentage points — a roughly 47% jump from the control group’s 4.3% desired allocation — and pushed actual subsequent purchases up by about 2.5 percentage points. Cointelegraph quotes the authors describing the result as “providing information about recent Bitcoin returns induces some households to start buying cryptocurrency.”
On the durable-goods effect, Cointelegraph puts a number on it that Crypto Briefing does not: a doubling in Bitcoin’s price made a household whose entire financial portfolio sat in crypto 1.4 percentage points more likely to buy a durable good, which Cointelegraph describes as roughly a 7% increase relative to the unconditional probability of such a purchase.
Crypto Briefing, for its part, supplies context Cointelegraph’s account lacks: it describes crypto holders as more libertarian in political orientation than non-holders, and it places the paper as building on a 2023 National Bureau of Economic Research study on household expectations around digital assets. Neither of those framings appears in Cointelegraph’s reporting.
Why it matters
If the paper’s findings hold up, they offer a Fed-branded explanation for why crypto rallies can feed on themselves. Cointelegraph quotes the authors writing that “positive returns attract new participants, which raises the price further,” and separately quotes them concluding that the absence of shared information and beliefs among crypto investors means price volatility will likely remain a defining feature of the asset class. That reading suggests retail demand for crypto may depend less on fundamentals and more on which headlines about past performance reach potential buyers.
Cointelegraph notes the information-treatment effect was concentrated among people who said they avoided crypto for lack of information; those who already believed crypto was a bad investment did not respond to being shown Bitcoin’s returns, according to Cointelegraph’s account.
What remains unresolved
Neither outlet’s account addresses whether the working paper has undergone or will undergo peer review or formal journal publication. Neither specifies the exact survey vehicle or the full time span of the underlying data — Cointelegraph refers only to survey waves without dating their range, and Crypto Briefing does not detail the sampling method beyond calling it a large-scale household survey. It also remains unclear whether the effects measured in the 2025 experiment persist beyond that specific survey wave, since both outlets describe only the immediate post-treatment results.
Sources
Every fact above is attributed to one of these reports. Where they disagree, the article says so.
Read More
