A U.S. Federal Reserve (Fed) study found that the key variable separating participation in crypto investing is perceptions of expected returns and risk, rather than income or age.

Cointelegraph reported on Aug. 23 that the Federal Reserve Bank of Cleveland analysed that crypto holders and non-holders fundamentally perceive the future returns of digital assets differently.

In repeated surveys of as many as 25,000 households, researchers found that expectations of future returns had a bigger effect on whether someone held crypto than age, income or gender. The pattern became clearer when perceptions of risk were also considered. This differs from stocks, bonds and gold, which are relatively more influenced by demographic characteristics.

The study also confirmed that investors’ views of crypto’s future value vary widely. It explained that past price gains can spur new buying, and increased buying can then push prices up again, creating a feedback loop.

A lack of understanding of crypto was also evident. In a 2021 survey, 87 percent of respondents who did not hold crypto said it was difficult to predict returns over the next year. Among holders, 54 percent gave the same answer.

Among those who provided an outlook, holders’ average expected return was 22 percent, well above non-holders’ 7 percent. Holders also tended to see crypto as less risky than non-holders did.

The impact of expected returns on the likelihood of holding crypto was also confirmed in numbers. When an individual’s expected crypto return rose by 1 percentage point, the probability of holding increased by 0.8 percentage point. The researchers said this relationship was relatively weaker for traditional financial assets.

The influence of demographic characteristics did not disappear entirely. Participation was higher among younger people, men and higher-income, higher-asset households. Even after controlling for other conditions, those under 40 were 13 percentage points more likely to hold crypto than those 60 and older. Men were also about 4 percentage points more likely to hold it than women.

A randomised experiment also confirmed that investment information changes actual behaviour. In 2025, the researchers randomly provided participating households with information related to bitcoin, stocks, GameStop and inflation.

Participants who saw bitcoin’s return over the prior 12 months raised their desired share to allocate to crypto by about 2 percentage points. That was about a 47 percent increase compared with the control group’s 4.3 percent. The likelihood of actually buying crypto later also rose by about 2.5 percentage points. By contrast, respondents who had already judged crypto to be a negative investment target showed little reaction even after seeing the same information.

The study also found some impact of rising crypto prices on household consumption. If the bitcoin price doubled, households holding their entire financial portfolio in crypto were 1.4 percentage points more likely to buy durables such as cars and furniture. That was about a 7 percent increase compared with the average purchase probability.

Crypto gains did not lead to increased everyday spending. The researchers analysed that investors tend to view crypto gains as closer to 'gambling income' like lottery winnings than as sustained wealth accumulation.

The study shows that volatility in the crypto market is difficult to explain by fundamentals alone. The researchers judged that because investors lack shared information and beliefs, price volatility is likely to remain a prominent feature of crypto. They also expected that future demand for retail investment would be influenced not only by the bitcoin price itself but also by how investors interpret past price movements as information.