New tokens in the top 100 post a median 27.5% return while older-cycle tokens all show losses

Cryptopolitan does not give the window the 27.5% covers, the count of 2024-2026 launches in the top 100, or launch dates for the Q3 outperformers it names ^[9][18]. It also says every other token from earlier cycles posted negative returns ^[9], yet 30% of assets listed before 2017 are in profit ^[10], so "every" has to mean the typical return of each older cohort. The profit shares are firmer: the newer group's hit rate is 2.36 times the oldest group's ^[7], which leaves 29.2% of the newer group underwater ^[6].

We think returns have concentrated in one vintage, and that the evidence for concentration is stronger than the evidence for a rally. The case against has two parts. The first is selection. Cryptopolitan frames the figure around the top 100 assets, while Pump.fun's lifetime count alone is over 23.7M tokens, most of which lost their value and all of their activity ^[2]. If every one of the top 100 were a Pump.fun token, that would still be one in 237,000 ^[8]. A median taken after that filter leans positive, so the number that would break our view is the median across every 2024-2026 launch, failures included.

The second part is design. Specific newer tokens gained more than 200% over the past quarter ^[16], and Cryptopolitan says newer tokens often come with incentives, token burns and fee distribution ^[17]. Its list of Q3 outperformers includes UNI and AAVE ^[18]. If those two distribute fees or burn supply and pre-date the 2024-2026 cohort, launch year is a proxy for those features, and an older token with the same features should show up in the positive column.

The cohort totals have their own gap. If $6.4B is about 51% of the 2026 cohort's market cap ^[12], the cohort is worth roughly $12.5B ^[13]. Cryptopolitan separately counts 732 project-linked tokens in 2026 at an average market cap of $23M ^[11], which multiplies out to $16.8B ^[14], some $4.3B or 34% more ^[15]. A different token set could explain it (project-linked tokens against everything that launched). The article's own caveat is that token value does not always reflect activity or available liquidity ^[3].

On Cryptopolitan's account, the behaviour behind the numbers is turnover. Traders are no longer loyal to specific tokens or expecting returns from underperforming chains ^[19], and liquidity rarely shifts between BTC and altcoins ^[5]. We would expect the older cohort to stay negative while that holds.