# Altcoin 80% drawdown decision extension — frozen before extension result

The existing four-decision release tested 30%, 50% and 70% drawdown crossings. On 2026-09-25 the user asked specifically about an altcoin down 80%. This is a **post-original, transparently dated extension**, not an original preregistered threshold. Its rule is fixed here before running the 80% calculation. Do not relabel the older 50% outcome as the 80% answer.

Input is the existing immutable `2026-09-20-v2/spot.json.gz` and its historical market directory. Binance USDT spot markets from that release, including delisted discovery files, are considered. Exclude records with a published `excluded_reason`; require 180 complete prior daily closes, a first close crossing below 80% of the previous 180-day closing high, a previous close above the threshold, prior 30-day mean quote turnover >=1m USDT, and 90-day same-symbol cooldown. The crossing is known after the UTC daily close; action reference is the following UTC daily open. This is **not a user's own purchase-price loss**.

Compare, from equal remaining value, hold the original coin, sell and buy BTC, and sell to a no-interest cash reference over 30/90/180 days. Use 0.15% per side as the base cost. Also compute 0.05% and 0.30% sensitivities. Missing exits and immature endpoints are explicit, not zero or forced recovery. For complete paths publish individual returns, paired BTC-minus-hold difference, hold/rotate maximum drawdown and whether the original coin recovered its preceding peak by the endpoint. Do not infer a personal tax result or tell a user to sell.

Publish complete candidate, observation, summary and input/checksum references. A third party should be able to recompute summary counts and paired returns. Overlapping symbols and shared BTC regimes are dependent; complete-case summaries may be survivorship-biased by missing exits. No predictive rule for deciding which altcoin will rebound is estimated. Place the 80% result on the existing owner URL; no duplicate URL.