Weekly Portfolio Plan recap, 24 Aug to 21 Sep 2026

21 Sep 2026 CryptoCrispy

From 24 August to 21 September 2026 the Telegram channel posted seven Weekly Portfolio Plans. This recap scores those published books. It is not a live account, not a 7-day trader scorecard, and not a reason to copy the latest tactical names.

Keep the counting rules, the performance tables, and the reports library next to this page. Signal hit rate is a different product. This page is the investor book: core hold plus an optional sleeve.

How to read the numbers on the cards

Each graphic has two layers. The coin cards are this week’s plan. The green “notable winners” line is a trailing score of an older plan, not a league table of the whole market.

• A prior snapshot is used only after it is at least 28 days old.
• Only names that were in that older book are scored, and only names still up at least 2% are printed.
• Names that were flat, down, or unpriced do not appear. The strip is a winner filter, not a full P&L.

That is why SOL and LINK dominated the strip through mid September, then vanished on 21 September. They belonged to the 2 August book. They were not in the 24 August book, so they dropped out when the lookback reset.

The seven published books

Graphics used Core about 70% and Tactical up to 30% throughout. Early captions still split Core / AI-growth / Tactical as 40 / 30 / 30. From 7 September the caption matched the graphic: Core about 70%, Tactical about 30%.

• 24 Aug. Core: BTC, ETH, PAXG, KAG, RENDER. Tactical: GRAM, XAUT, UP, ZEC.
• 31 Aug. Same core. Tactical: DOGE, ADA, XRP, XLM.
• 7 Sep. Same core. Tactical: ONDO, GRAM, INJ, XRP.
• 10 Sep. Core swapped RENDER for NEAR. Tactical: GRAM, XLM, ANT, BNB.
• 14 Sep. RENDER returned to core. Tactical shrank to GRAM and PUMP.
• 17 Sep. Same core. Tactical: SYN, ADA, GRAM, CRO.
• 21 Sep. Core swapped RENDER for AGIX. Tactical: GRAM, HYPE, NEAR, UNI.

BTC, ETH, PAXG, and KAG never left core. The fifth core seat rotated: RENDER most weeks, NEAR on 10 September, AGIX on 21 September. GRAM was the persistent tactical name (six of seven posts). Everything else in the sleeve was a short stay.

Full tables and the 28-day chart are on the Weekly Portfolio Plan performance report.

The first complete 28-day window

21 September is the first date in this sample where the 24 August book is old enough to score. The published strip that day is labelled about 28 days since 24 August:

• ZEC +80.4% — tactical on 24 August only
• RENDER +20.8% — core from 24 August until this same post
• ETH +9.2% — core throughout
• BTC +6.2% — core throughout

PAXG, KAG, GRAM, XAUT, and UP were also in the 24 August book and did not print on that strip, so they were not up 2% or more versus the stored 24 August prices — or they could not be priced. Do not fill those gaps with guesses.

That is the useful split. Core did the hold: BTC and ETH were still green, RENDER was the stronger core name, gold and silver did not headline. The tactical sleeve was a lottery ticket: ZEC was the printed outlier, the other three 24 August tactical names were not.

What the lengthening 2 August window was really showing

From 31 August through 17 September the cards kept scoring the 2 August book as the window grew from about 29 days to about 46 days. Printed leaders in that stretch:

• 31 Aug (~29d): SOL +42.4%, LINK +39.5%, ETH +31.9%
• 7 Sep (~36d): LINK +64.1%, SOL +45.0%, NEAR +37.8%
• 10 Sep (~39d): LINK +45.9%, NEAR +44.3%, SOL +40.3%
• 14 Sep (~43d): NEAR +43.9%, LINK +41.3%, SOL +41.0%
• 17 Sep (~46d): NEAR +62.1%, SOL +38.8%, LINK +38.6%

SOL and LINK never appeared in the seven live plans above. Those prints are leftover performance from the 2 August book. Treating them as a reason to buy SOL or LINK in September would be reading the wrong layer of the card.

NEAR is the exception that moved toward the live book. It printed on the 2 August strip, sat in core for one mid-week post (10 September), then returned as a tactical name on 21 September — after the lookback had already left the 2 August book behind.

Core effect versus sleeve effect

If someone held the stated mix — about 70% in the four persistent core names plus one AI core seat, and up to 30% in the rotating sleeve — the path was not “buy every Strong Buy on the graphic.” Most tactical names lasted one post. Copying the sleeve each Monday would have churned GRAM, then a new alt basket, then another.

The 21 September core change is the other warning. RENDER had just printed +20.8% on the 24 August book, and that same post removed it from core for AGIX (negative sentiment, very strong momentum on the card). The plan is a new allocation, not a victory lap. Past core winners are not automatically next week’s core.

ETH’s own printed path shows why the caption says this is not a 7-day scorecard. ETH was +25.3% on the 27 July book (24 August card), +31.9% to +36.6% on the growing 2 August window, then +9.2% on the 24 August window. Same asset, different start date. The window is the result.

What this recap cannot claim

Captions do not include every coin’s start price, so this page does not invent a blended 70/30 account return, a Sharpe ratio, or a full winner/loser count. It uses only names and percentages that were printed on the plans themselves, plus the published rule that the strip is a 28-day winner filter of a stored book.

Compare that with Telegram signal and assistant reports if you need trade-level samples. Those are a different clock.

Continue with how to read a weekly report or portfolio analytics that actually matter.

CryptoCrispy provides market analytics and trading tools. This article is general information, not personal financial advice. Figures are taken from published Weekly Portfolio Plan posts between 24 August and 21 September 2026. Past prints are not a reliable guide to future results.