Add TQQQ dynamic volatility threshold research

[Pigbibi]

Summary

• add a reusable TQQQ volatility-delever threshold research script
• record the 2026-06-09 dynamic p90 floor24/cap36 promotion evidence
• include QQQM as the live unlevered sleeve with QQQ long-history proxy support in the archive replay
• add volatility-delever diagnostics to the archive signal history

Tests

• PYTHONPATH=/home/ubuntu/Projects/UsEquityStrategies/src:/home/ubuntu/Projects/UsEquitySnapshotPipelines/src /home/ubuntu/Projects/UsEquitySnapshotPipelines/.venv/bin/python - <<'PY'
  from pathlib import Path
  from tempfile import TemporaryDirectory
  from tests import test_tqqq_growth_income_archive as t

t.test_tqqq_growth_income_run_backtest_produces_summary()
t.test_tqqq_growth_income_run_backtest_can_open_income_layer()
with TemporaryDirectory() as d:
t.test_tqqq_growth_income_archive_writes_replayable_outputs(Path(d))
with TemporaryDirectory() as d:
t.test_tqqq_growth_income_archive_supports_full_income_mode(Path(d))
print('archive tests OK')
PY

• /home/ubuntu/Projects/UsEquitySnapshotPipelines/.venv/bin/python -m py_compile src/us_equity_snapshot_pipelines/tqqq_growth_income_archive.py scripts/research_tqqq_volatility_delever_thresholds.py tests/test_tqqq_growth_income_archive.py

Research output

• data/output/tqqq_volatility_delever_threshold_research_20260609/variant_summary.csv

[chatgpt-codex-connector]

💡 Codex Review

Here are some automated review suggestions for this pull request.

Reviewed commit: ee953c2f04

ℹ️ About Codex in GitHub

Your team has set up Codex to review pull requests in this repo. Reviews are triggered when you

• Open a pull request for review
• Mark a draft as ready
• Comment "@codex review".

If Codex has suggestions, it will comment; otherwise it will react with 👍.

Codex can also answer questions or update the PR. Try commenting "@codex address that feedback".

[chatgpt-codex-connector]

Preserve tracking for QQQ targets

When any caller still routes the unlevered sleeve to QQQ (the existing overlay research variants in scripts/research_levered_overlay_guards.py still do this, and run_backtest still requires QQQ history), removing QQQ from MANAGED_SYMBOLS makes _execute_rebalance buy a target symbol that is not carried into next_weights or weights_history. The purchased QQQ value is counted in new_equity for that rebalance, then dropped from current_weights on the next loop, which corrupts returns/trades for those scenarios; either keep QQQ managed or translate/force all QQQ targets to QQQM before execution.

Useful? React with 👍 / 👎.