quantmodel

← Results index

Detecting SP500 BUY-THE-DIP signals | RecessionAlert — exploratory proxy campaign

2026-09-02T13:30:00Z Research completed The SHA-pinned March 2021 BUY-THE-DIP survey discloses four complementary families (ZBT, GTR, SIGS, Trendex) with qualitative buy rules but no native SIGS/Trendex tables or complete execution map. Thirty-four labeled workbook-replay and public-proxy variants were executed with causal lag defaults; publication is exploratory Other/Research only.
1380 recessionalert research_campaign proxy buy-the-dip fidelity_mismatch data_limited exploratory_tested_proxy

Source report: research/campaigns/recessionalert-buy-the-dip-1380/synthesis.json

Objective

Independently replicate the distinct RecessionAlert inventory row for SP500 BUY-THE-DIP signals from the checked-in data/recessionalert.sqlite snapshot without live-site requests, credentials, private account data, or canonical crawl-database mutation.

Source and provenance

Page https://recessionalert.com/detecting-sp500-buy-the-dip-signals/ bound to SHA256 sha256:30fff0fd433e25ecaf2c490a1f8dfb40e166a6719d7344e3e2000f023948b57a, fetched 2026-08-22T23:47:50Z, published 2021-03-12. Workbook replay SHA256 sha256:fc2be5f1fc2a2e38949ae566f6e8cb4699a95562975a498fb5a53510cc40d67c. Public proxies loaded read-only from data/market.sqlite.

Disclosed rule, native span, and publication lag

Documented fragments cover ZBT (CHARTS>ZWEIG, 15-min updates), GTR Class A/B (CHARTS>GTR), SIGS A (-12 rise) and B (0/+2 pivot bounce), and Trendex rise-from-0 with bottom-pane probabilities toward 85%+. Native ZBT workbook span ~2003+, GTR ~2015+; SIGS/Trendex observations absent on-page. Chart cadence is 15 minutes; exact publication lag and portfolio map unknown. Causal default is month-end lag>=1; lag-zero retained as control_only.

Approach and variants

Primitives in research/research_campaign/recessionalert_1380.py. Families: workbook ZBT/GTR replays, documented SIGS/Trendex diffusion rules on inferred proxies, breadth/drawdown-VIX proxies, four-family consensus, and market-bottom filter overlays. Trial ledger: 34 variants across lags 0/1/2 and lookbacks 6/12/21.

Results (exploratory)

Best signal overlay zbt_combined_any_lag2 (CAGR 6.39%, Sharpe 0.610) trails SPY buy-and-hold 10.76% CAGR on 1993-01..2026-07. Entry-event studies favor gtr_either_lag1 (+7.68% mean 6m forward SPY). Six zero-signal variants preserved. Disposition: exploratory_tested_proxy with no playable indicator promotion.

Files and validation

Validation: focused pytest 10 passed; read-only DBs; no live RA fetch.

Limitations

Public proxies are not official RecessionAlert SIGS/Trendex or chart series. Workbook replays partial; classic breadth thrust proxy produced zero events on 1993+ history. Do not treat exploratory equity paths as a registered Other/Indicators signal.

Source artifact: research/results/1380.md