[Design] Activism as a first-class campaign-lifecycle dataset

[stranske]

Problem

MDB's activism detection is a thin, single-form slice. The field (Diligent Market Intelligence / Insightia, 13D Monitor) treats activism as a first-class campaign-lifecycle dataset: multi-form (13D/13G/13F together), activist profiles, estimated campaign returns + holding periods, and a letters/agreements archive. 2025 set a record 255 campaigns. This is effectively a product line, not a feature.

Proposed approach (for discussion)

• Model an activism campaign entity linking a manager to a target across filings over time (13D/A amendments, 13G, related 13F positions), with status + timeline; edgartools ([Audit][R3] Adopt edgartools for EDGAR parsing (13F parity + insider/13D follow-on) #1323) already parses 13D/G.
• Layer campaign-return estimation (needs the backtest/price plumbing from T2) and, optionally, a letters/agreements document store.

Open decisions (owner)

• Ambition: campaign-linking + timeline only (bounded, near-term) vs full parity with Diligent (returns + letters archive — large, multi-quarter)?
• Campaign-return estimation depends on the price/as-of plumbing (T2/T1) — in-scope now or later?
• Is this a priority vs the other Tier-2 bets, or a longer-horizon bet to stage last?

Definition of done (for this DESIGN issue)

Design doc: the campaign data model + a phased scope (start with 13D/G/A campaign-linking via edgartools; defer returns/letters). No code until scope is chosen.

Notes

Part of epic #1320 (Tier-2). Largest bet; campaign-return depth depends on the bitemporal/backtest track.

Part of epic #1320 (Tier-2).

[stranske]

Reopening for verifier sequencing after #1412 merged and received verify:compare. This source issue should stay open until the post-merge verifier/provider-comparison report is durable PASS or any real follow-up debt is dispositioned.

[stranske]

Post-merge verifier disposition for PR #1412 is complete.

Evidence checked this opener round:

• PR Add activism campaign lifecycle design #1412 is merged and linked to this issue.
• verifier / Run post-merge verifier passed on run 29057182604 / job 86251046930.
• Provider Comparison Report posted on PR Add activism campaign lifecycle design #1412 reports PASS/PASS: openai PASS 93%, anthropic PASS 78%.

No bounded follow-up debt remains for the docs-only campaign lifecycle design, so closing this verifier-sequenced source issue.

[stranske]

Closed after durable PASS/PASS verifier evidence on merged PR #1412.

[stranske]

Owner chose full Diligent-parity → phased: #1466 campaign entity+timeline (lane-ready), #1467 campaign returns (blocked by #1466 + #1464), #1468 letters/agreements archive + activist profiles (blocked by #1466).