According to a working paper published by the Cleveland Federal Reserve, the past increase in bitcoin would not only affect the views of United States households on encrypted assets, but could also directly facilitate subsequent purchases. The study found that, after seeing a return on Bitcoin over the past 12 months, the desired distribution of encrypted assets had increased significantly, followed by an increase in the real purchasing probability.
Random experiments indicate that the configuration will rise
The study was based on multiple rounds of surveys, each with a sample of approximately 15 to 25 million people, interviewed from Nielsen Homescan Panel. The research team joined the random information experiment in the second quarter of 2025 to present to the different interviewees information on bitcoin, standard 500, GameStop or inflation projections.
Of these, the Bitcoin Group saw a return of 14.3 per cent over the past year, or a price trend. The results show that, following access to such information, an average increase of about 2 percentage points was expected in the distribution of encrypted assets. This represents an increase of approximately 47 per cent compared to the average target configuration of 4.3 per cent for the control group.
The study also found that part of the adjustment came from a decline in the allocation of cash, current accounts and savings accounts. At the same time, there has been an increase in the target allocation of equities, indicating that such information not only increases interest in encrypted assets, but also leads to a broader risk asset preference.
The expected gap between the holder and the non-holder is significant
The paper showed that there was a clear gap between the judgement of the holders of encrypted assets and that of non-possessors regarding future returns. In the third quarter of 2021, respondents who had held encrypted assets expected an average return of 22 per cent for the coming year, compared with 7 per cent for those who did not.
By 2025, the expectations of both groups had fallen, but the gap remained. The average expected return for holders was 13.8 per cent and for non-holders 4.7 per cent. According to the study, the statistical correlation between the expected return and the possession of encrypted assets is stronger than personal characteristics such as age, income, gender or wealth.
According to the paper estimates, the probability of holding an encrypted asset increases by an average of 0.8 percentage points for each 1 percentage point increase expected by the interviewee. However, the author also points out that this result reflects mainly relevance and that it cannot be assumed that optimistic expectations explain the full silo formation process.
There's an increase in the probability of subsequent actual purchases.
In the follow-up survey, the probability of actually buying encrypted assets increased by about 2.5 percentage points among respondents who saw information on the return of bitcoin. Before the experiment, about 11 percent of participants held encrypted assets. Based on this, the research team estimates that this information processing increases the probability of unconditional purchase by about 23 per cent.
The authors state that there are not many people in the sample who actually change hold-up, so two sets of bitcoin-processing samples are merged to improve statistical effectiveness, with a notable level of p = 0.017.
In terms of cluster results, the most obvious response was the non-possession of an encrypted asset that had not previously been purchased due to “under-under-informed”. Compared to those who would have thought that encrypted assets were not a good investment, behaviour changed less.
Price increases or demand feedback
It was argued that this result supported a possible market-drive path: higher prices would raise returns, attract new participants and create new demand. The author described it as a possible foam mechanism, but did not judge from it that each round of bitcoin increases would reinforce themselves.
The study also mentioned that other Federal Reserve surveys had similarly shown that the main purpose of American participation in the market for encrypted assets was still investment rather than payment.
In addition, the paper observed the impact of changes in bitcoin prices on household consumption. Estimates show that if all the financial assets of a family were deployed in an encrypted asset, a doubling of the price of a bitcoin would increase its probability of buying durables by 1.4 percentage points, with a more visible response from commodities such as computers and refrigerators, and weaker changes in cars and housing.
In the author ' s view, this suggests that some households may prefer to view the proceeds of encryption as a one-off windfall gain rather than a long-term increase in wealth. However, the paper remained a discussion paper and the conclusions represented the author ' s individual findings and were not a formal policy position of the Fed.