Who I Am (External Persona)

I am the Pelosi household — Nancy Pelosi (D-CA, 12th District) and her husband Paul Pelosi, a venture investor. I don't talk in this system. I'm mirrored: every stock or option trade disclosed through the House Periodic Transaction Report (PTR) process for either Nancy or Paul, scaled and replicated here in a $100K paper book.

This persona is not an internal voice. It is a signal source. What the Pelosi household trades, on the ~45-day disclosure lag mandated by the STOCK Act, is the signal. My P&L tells you whether mirroring disclosed Congressional trades — with all their informational advantages and the delay tax — is a profitable strategy in this regime.

Mission Statement (as copy-trader)

Mirror all disclosed trades filed under "Hon. Nancy Pelosi" where the owner field is self, joint, or spouse (captures Paul's trades, which are typically the largest and most notable). Scale each trade by the disclosed amount range midpoint divided by $50,000 — so a $1M-$5M disclosed trade (midpoint $3M) becomes 6000% of our book (clamped by available cash), and a $1K-$15K disclosed trade (midpoint $8K) becomes 16% of our book. The goal is to preserve the relative conviction across trades, not the absolute dollar size.

Methodology (the household's, as inferred from public filings)

Filings show a pattern, not a prospectus. Observable characteristics:

1. Concentrated large-cap and mega-cap tech. NVDA, AAPL, MSFT, GOOGL, AMZN, TSLA, AVGO, CRM, PANW, V, and similar names dominate disclosed activity.
2. Options activity, not just stock. A significant fraction of trades are call options (often deep ITM LEAPs) in single names. This mirror tool captures the underlying equity trades; options are not mirrored in v1.
3. Size when it rips. Several historical trades have been in the $1M-$5M and $5M+ buckets, materially larger than the typical member filing.
4. Buy-and-hold tilt. Most disclosed transactions are purchases; sales are less frequent.
5. Committee-adjacent names have surfaced occasionally. Whether this is causal or coincidental is debated in published research; we mirror the disclosed trade, we do not infer motive.

Entry Rules (as mirrored)

1. A purchase ("P") appears in the House PTR feed under a Pelosi-household filer → we BUY at the closing price of the transaction_date.
2. Scale: (amount_midpoint / 50,000) × $100,000 = position size in USD, capped at available cash.
3. Dependents are NOT included by default; owner must be one of self, joint, or spouse.
4. Must have OHLC data available locally for transaction_date — skip otherwise.
5. Minimum trade size: $100 (after scaling). Below that, skip.
6. Exchange-type filings (asset swaps, 401k rebalances) are skipped.
7. Options are not mirrored in v1.

Exit Rules (as mirrored)

1. A "sale_full" filing → we fully exit the position.
2. A "sale_partial" filing → we sell shares equivalent to the scaled USD amount, capped at our current holding.
3. If the household discloses a sell for a ticker we don't hold, we skip and log.
4. No independent stop-losses and no profit targets. We follow the household's disclosed discipline, not ours.

Position Sizing

Scaled by disclosed amount-range midpoint. "Capped" means clamped by available cash — we never borrow.

What I DON'T Do

• No independent thesis.
• No political commentary. Fact-based: STOCK Act requires disclosure within 30/45 days.
• No options mirroring in v1. Deferred to v2.
• No dependents. Only self, joint, spouse owners.
• No backfill before the --since cutoff.

The Core Truth

Running Pelosi-household as a copy-trader persona answers:

1. Is disclosed Congressional activity a profitable signal on a ~45-day lag?
2. Does household-level concentration beat our diversified research?
3. Is the "smart money mirror" narrative regime-dependent?

This is measurement, not endorsement.

Data Source

• Primary: housestockwatcher.com S3 JSON (House Clerk PTR filings, daily updates)
• Fallback: manual CSV via --file
• Import tool: the desk's own tooling

Known Limitations

1. 45-day disclosure lag is real-world but invisible in our backtest.
2. Amount ranges are wide (the $1M-$5M bucket covers 5x).
3. Options not mirrored.
4. Full-exit detection relies on sale_full vs sale_partial.
5. Name-match is substring-based.

Current Regime Read

To be filled in after first import.