Innovator Adoption Predicted Returns in a Study of 77 Cryptocurrencies

A study of 77 cryptocurrencies found that modeled innovator adoption predicted future returns, while total active-address growth did not, separating one model component from a broader measure of network activity.

That finding has a narrower scope than an early-mover claim. The researchers tested whether model-estimated adoption components predicted later returns; they did not test investor timing against market attention. The result describes a relationship in historical data, not a measure of who bought before a crowd.

What the adoption model measured

Khaladdin Rzayev, Athanasios Sakkas and Andrew Urquhart’s peer-reviewed study, published May 16, 2025, examined monthly active-address data for 77 cryptocurrencies from January 2014 through July 2022. The researchers used an adoption model that separated address growth into “innovator” and “imitator” components.

In the return analysis, the model-estimated innovator-adoption component predicted future returns. The imitator component and total active-address growth were not statistically significant predictors. These labels describe parts of the researchers’ statistical model; they do not identify particular people, wallets or groups of investors.

The distinction matters for interpreting the result. Total active-address growth is a broad measure in the study, while innovator adoption is a component estimated by the model. The finding does not establish that a cryptocurrency with more active addresses will rise, or that any particular address-growth pattern will precede a price increase.

The researchers also tested a long-short portfolio. It bought cryptocurrencies in the top 30% for innovator adoption and shorted those in the bottom 30%. The test produced an average monthly excess return of 18.20% over the study period, with a monthly standard deviation of 96.85%.

Those figures describe a historical portfolio test, not an expected return for an individual trader. The standard deviation was also high relative to the reported average monthly excess return, indicating substantial variation in the test’s monthly results. The model categories were statistical estimates, so the portfolio result does not show that a retail trader could have identified or replicated the same groups in real time.

The study’s data ended in July 2022. Its results therefore do not establish whether the relationship held in later market conditions, including those in 2026. Nor did the study test whether adoption signals outperformed market attention or whether early investors earned an advantage.

A separate measure of pool liquidity

A separate analysis examined 1,015,687 Meteora DAMM v2 pools opened on the Solana network from Oct. 30, 2025, through July 5, 2026. It counted 904,991 pools, or 89%, as drained under its definition. Another 802,846 pools, equal to 79% of the full sample, reached the threshold within an hour. The median time to that point was seven minutes.

A pool counted as drained when a withdrawal and sales into the pool during the next 10 seconds left less than 1% of its prior SOL balance. The measure tracks a defined change in liquidity for this pool type on one venue. It does not measure adoption across cryptocurrencies or establish that a drained pool involved fraud.

The two analyses address different questions. One tests whether a modeled component of active-address growth predicted returns in a historical sample; the other measures how quickly liquidity fell below a specified level in a set of new pools. Neither supplies a formula for future returns, and the pool results cannot be generalized to crypto markets as a whole.