← Results index

The SP-500 200-day average “Goodbye Kiss” | RecessionAlert — exploratory proxy campaign

2026-09-03T00:00:00Z Research completed The SHA-pinned Goodbye Kiss page documents an S&P vs 200dma retest family with no executable price source, SMA spec, lag clock, or portfolio mapping. Seven labeled public-proxy variants were executed with causal lag>=1 defaults; all trail buy-hold SPY, so publication is exploratory Other/Research only.
1592 recessionalert research_campaign proxy goodbye-kiss fidelity_mismatch data_limited exploratory_tested_proxy uninvestable

Source report: research/campaigns/recessionalert-goodbye-kiss-1592/synthesis.json

Objective

Independently replicate the distinct RecessionAlert inventory row for the SP-500 200-day Goodbye Kiss 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/the-sp-500-200-day-average-kiss-of-death/ bound to SHA256 sha256:ba2d2206a781c607285de8d2de44de176e2543fcb9641cd2d022cb479c730787, fetched 2026-08-22T23:45:19Z, published 2015-10-22T22:11:52-0400. Public proxies loaded read-only from data/market.sqlite (^GSPC from 1927, SPY/IEF overlay from 2002).

Disclosed rule, native span, and publication lag

Documented: S&P vs 200dma separator, protracted OR-gate (>=20 days below or dma down), 100-day quiet commencement (13 events since 1972), 2% retest or 2–5 day breakout kiss after 23–83 bars, narrative exit/short/re-enter cluster ladder. No native price observations are pinned and no chart asset bytes are stored on-row. Publication lag is undisclosed; causal default is lag>=1 on daily closes with lag-zero as control_only.

Approach and variants

Primitives in research/research_campaign/recessionalert_1592.py. Seven variants across below-day gates (20/46/63), bands (1.5%/2%/2.5%), persistence (2/3/5), SMA windows (190/200), slope windows, and lags (0/1/2) on the ^GSPC signal with SPY/IEF long/cash overlays.

Results (exploratory)

Best causal variant (wide 2.5% band, lag 2) posts CAGR 5.69% and Sharpe 0.37 vs buy-hold SPY 11.16% on the 2002+ common window; the documented-gate variant posts CAGR 4.50% and Sharpe 0.29 with 81% risk-off exposure. The lag-zero control is negative. Diagnostics show regime heterogeneity (2023+ Sharpe 0.72 vs GFC -0.57). Interesting negative result: the broad kiss flag is too sensitive for timing use.

Limitations

Native price vendor, SMA/slope/cluster/band specification, machine-readable 13-event catalog, on-row asset bytes, portfolio mapping, and publication clock remain unknown. See research/findings/1592.md and the spec at research/findings/specs/recessionalert-goodbye-kiss-1592.json.

Source artifact: research/results/1592.md