# gap-series-canonical.csv, method note

Series: MSTR, weekly, rows 50 to 128 (2025-01-06 to 2026-09-28), basis A as restated 2026-10-01.
Formulas and columns: docs/evidence-update-pipeline.md in the repository.

## btc_price_8k

The week's average transaction price stated in that week's 8-K, sales included; a week with two
stated periods is BTC-weighted. A week with no stated price carries the CoinGecko D+1 close for the
8-K's as-of date. `gap_raw` and `dg_raw` read this price. The `_norm` columns and `pass2_event` use
the fixed $77,500 and do not.

## The terminal on the pages

/evidence/ and /preferred-era/ show one terminal figure: row 128's holdings, shares and
claims struck at the 2026-09-30 close, $83,576.13, which gives a raw gap of 40,476.8 sats.

The CSV keeps row 128 at its own 8-K price, $85,681 (the week's average
purchase price), where the raw gap is 39,141.7 sats. Both are the same row. They differ only in the
BTC price applied.

## Corrections

2026-10-01: row 111 carries no preferred issuance (8-K 0001193125-26-249768 reports none; it was an
EST ~$48M STRC), and `cum_pref_face` is $48,000,000 lower from row 111 on. Rows 112 to 128 were
appended from the MSTR anchor's 8-K rows. Rows 50 to 110 are unchanged.