48-Hour Marks in a Selected Solana Demand-Only Shadow Cohort

What this internal study measured

This is a descriptive audit of forward shadow observations recorded by MemeAssist's Analysis Setup demand-only research lane. A shadow observation records an eligible opportunity and follows observed prices without placing a funded order. It is not a claim that a customer bought, could have sold, or achieved the recorded price change.

The question is deliberately narrow: among distinct tokens admitted to this one selected lane from 10–15 September 2026 UTC, what change in observed price mark was available at the fixed 48-hour horizon? We audited the frozen snapshot on 20 October 2026, using a cut-off of 00:00 UTC that day.

The frozen cohort and missing outcomes

We restricted the cohort to the shadow-opportunity-v1 policy, Analysis Setup desk and demand-only lane. An observation had to have recorded mandatory-safety and opportunity qualification, a positive finite starting price, and no dispatched transaction or funded capacity reservation. All signals occurred after our clean evidence boundary of 26 August 2026 at 23:00 UTC.

We selected the earliest qualifying observation per token, ordered by signal time and then record identity, before checking whether it had settled. This avoids counting multiple signals on the same token as independent evidence or choosing a token's best result. It yielded 104 distinct tokens. Multiple tokens can still move together; token deduplication does not eliminate market correlation.

Only 76 of 104 tokens (73.08%) had a validated 48-hour mark. The other 28 of 104 (26.92%) were unmeasured: 27 had no on-time horizon mark and one remained open without a validated settlement at audit time. These missing outcomes are neither zero returns nor inferred total losses. The missingness could be related to market conditions, so the measured subset may be biased.

Results within the 76 measured tokens

Each percentage uses all 76 validated observations as its denominator, not just winners or losers. The count of 13 large declines is an event count within that 76-token cohort, not a separate 13-token performance comparison. A price decline is not proof of a rug pull or fraud. Positive marks are not proof that a position could have been executed profitably.

The finding is that many measured marks in this selected demand-only lane declined over this short period. It does not show that demand signals never work, that screening caused a loss, that all new Solana tokens behave this way, or that a different entry or exit rule would have performed better.

How the 48-hour observation was validated

The horizon is signal time plus exactly 48 hours. A qualifying settlement had to be explicitly labelled a fixed-horizon mark settlement, with its recorded horizon matching that instant. Its observed price timestamp had to fall between that horizon and 15 minutes afterwards. We did not substitute a much later price for a missing horizon observation.

We checked that the stored percentage agreed with (horizon mark / starting mark − 1) × 100 within 0.00001 percentage points. The settlement had to exist before the snapshot cut-off and the return had to be finite. A token without all of that evidence remained unmeasured.

For reproduction, the internal source is the research-shadow ledger, restricted to analysis_setup, demand_only and shadow-opportunity-v1; signal timestamps are at least 10 September and before 16 September 2026 UTC. The exact read-only audit query is retained with the editorial evidence registry. The selected-row audit fingerprint is 537ceb48106c39a27ca8be9857763fb9; this helps identify the audited extract, but is not a substitute for independent verification of the source data.

What this does not establish

• No funded track record: no capital was deployed by these shadow observations. Marks do not include fees, slippage, market impact, failed transactions or executable sale quotes.
• No chart-confirmation comparison: the demand-plus-chart lane has different completion and availability coverage. Comparing its available winners with this lane would not be a valid paired test.
• No exit-rule result: a fixed 48-hour mark does not test a stop, trailing exit, take-profit ladder or the full intraperiod price path.
• No liquidity or safety-factor attribution: this audit does not establish which factor explains the observed changes.
• No market-wide prediction: this is one strategy-selected lane and a short signal window. It is not a random sample of Solana tokens or evidence that future conditions will match.

This is an observational analysis: broader market moves and the selection rule may confound the result. We did not run a significance test, and these descriptive counts cannot generalize to future market-wide probabilities. The source price observations were not manually source-verified against independent markets. No public row-level dataset accompanies this audit; the internal extract and query are retained, but readers cannot independently verify each source observation from this page alone.

The practical distinction is between an interesting candidate and a demonstrated trading edge. A screen can identify something worth investigating without proving a profitable executable trade. For the broader method, see trade setups versus backtesting. For token-level checks, see the pre-buy safety guide.