In a 5,014-Token Solana Cohort, 11.8% Were Dead After 24 Hours

Correction and methodology clarification (September 2, 2026): We clarified source-verification, auditability and generalizability disclosures. The original cohort and figures are unchanged.

Why we examined this resolved cohort

MemeAssist automatically re-checks every token its screening engine analyzes at 1, 6, 12 and 24 hours and writes the outcome — up, down, or gone — to a database. This study reads a month of those records: 5,014 tokens with resolved outcomes between July 22 and August 24, 2026 — more than three times the cohort of our first outcomes study.

The analysis preceded each 24-hour outcome, but this remains a selected, non-random cohort of tokens submitted to MemeAssist. A further 1,657 checks did not produce either a usable price or a confirmed death and are excluded, so the results describe the resolved cohort rather than all submitted or all Solana tokens.

The resolved cohort's 24-hour outcomes

Of the 5,014 resolved outcomes in this July 22 – August 24, 2026 cohort, 592 tokens (11.8%) were dead within 24 hours — no tracked market remained quotable. This classification is distinct from a large price decline and is not a legal finding of fraud.

Observed death rates and survivor returns differed across the market-cap bands recorded at analysis:

In the selected resolved cohort, 48.9% of tokens analyzed below $100k were dead at 24 hours, and survivors in that band had a median return of −20.3%. These are descriptive group results, not a probability forecast for an individual token. The bands were not adjusted for liquidity, token age or other correlated factors, so they do not establish that market cap caused the difference.

Returns among the cohort's survivors

Among the 4,422 resolved-cohort tokens that still had a tracked market after 24 hours:

• The median survivor returned −2.7%.
• Only 39.7% finished up at all.
• 317 tokens (7.2%) lost more than half their value while surviving — the market kept quoting them all the way down.
• 134 (3.0%) gained 50%+, and just 60 (1.4%) doubled.

Within this survivor subset, a loss of more than half was about five times as common as a doubling. Because dead and unresolved tokens are outside the return calculation, these figures describe only resolved survivors and should not be treated as trading or position-sizing guidance.

How to interpret the market-cap bands

The table summarizes associations in this historical, non-random resolved cohort. It does not supply a rule for trading fresh launches or estimate what will happen to a particular token:

• Under $100k: 48.9% were dead at 24 hours, and survivors had a median return of −20.3%.
• $100k–$1M: 11.8% were dead, and survivors had a median return of −9.5%.
• Above $1M: the observed death rates were 3.5% at $1M–$10M and 4.4% above $10M; survivor medians were −3.0% and −0.6%, respectively.

Market cap can correlate with liquidity, age, visibility, holder structure and selection into MemeAssist. Without adjustment for those factors, the band differences cannot isolate a size effect or rank size against other signals such as holder concentration.

Methodology

Cohort: every token analyzed by the MemeAssist screening engine whose 24-hour outcome resolved between July 22 and August 24, 2026, with a definitive result — either a captured 24-hour price (n = 4,422) or a confirmed dead market (n = 592), for 5,014 resolved outcomes. A further 1,657 checks ended with no usable price data and no death confirmation; those are excluded from every figure rather than assumed dead. Because their outcomes are unknown and may differ systematically from resolved outcomes, their exclusion could bias the reported rates in either direction. Death rates count dead tokens against all resolved outcomes; return medians and percentages are computed among survivors only, and we disclose that treatment because deaths are not represented in the return statistics. Market-cap bands use the market cap recorded at analysis time, not after the move. See how the engine analyzes tokens for what happens at analysis time.

Limitations and verification: this is a first-party database study of tokens submitted to MemeAssist, not a random sample of every Solana token. Submission mix, repeated interest in risky tokens and changing market conditions can limit generalizability. “Dead” means no tracked market remained quotable after the outcome checks; it is not a legal finding of fraud. Missing outcomes may differ systematically from resolved ones, so excluding 1,657 unknown checks can bias every reported rate. Market-cap bands are descriptive and are not adjusted for liquidity, age or other correlated risks; those variables are potential confounders, so the table does not establish that market cap caused the outcomes. No confidence intervals or significance tests were calculated. Public row-level records for this historical cohort are not available. The source facts and outcomes were generated from the internal production tracker and were not manually verified row by row. Accordingly, the aggregate tables cannot be independently row-audited from public data.

About the screening data

Every token in this study got the same treatment you can run right now: paste a Solana mint address into MemeAssist and get an Overall Health Score, a Rug Risk Rating, an AI Verdict in plain English, and a Detailed Risk Breakdown — and your token joins the same outcome tracker that produced this study.