Analyzing SP-500 Seasonal Trends | RecessionAlert — exploratory proxy campaign

The article analyzes annual, bi-annual, tri-annual, and quadrennial/presidential-election cycles of SP500 monthly closing prices since 1957; sample sets are 63, 32, 21, and 15.

Each month/cycle phase uses average percentage gain, percentage of positive months, and total gains in positive months divided by total losses in negative months.

The source ranks 120 readings using (121-rank)/120, averages the three characteristic scores, and combines four cycle scores into one composite ranking signal.

The source says short below 37, long from 37 through 74, and 2x leverage above 74; shorts may be replaced with cash and leverage omitted.

The source describes scores above 80% or below 20% as rare/high-statistical-significance events for risk management.

The checked-in page has no source-linked monthly observations, canonical signal scale, phase table, or target series; the adjacent generic STM workbook is not unambiguously attributable.

This campaign uses checked-in Yahoo ^GSPC as a long-history SP500 price proxy and actual SPY for the requested overlay, with prior-only scores and one available-bar causal lag; every result is inferred/proxy.

A cash/short/unlevered overlay and lag 0/1/2 sensitivities make the article’s optional cash/no-leverage text testable without presenting an official execution rule.

The page discusses annual, bi-annual, tri-annual, and quadrennial monthly SP500 cycles; mean return, hit rate, and gain/loss ratio are ranked and combined.

Its displayed action bands are short below 37, long from 37 through 74, and 2x above 74, with optional cash/no-leverage alternatives.

The source-linked monthly observations, signal scale, phase initialization, publication clock, target instrument, lag, holding period, costs, exits, and re-entry are missing.

We therefore use ^GSPC monthly price returns, a prior-only 0–100 score, actual SPY overlays, one-bar causal lag, and explicit cash/short/unlevered sensitivities.

What can be learned from the SHA-verified RecessionAlert seasonal-trends article using transparent public SPX/SPY proxies, longest defensible history, and causal execution when native observations, scale, publication lag, and portfolio semantics are incomplete?

name

threshold

values

• 20

• 35

• 37

• 40

• 50

• 70

• 74

• 75

• 80

default

37

description

Browser threshold control on a 0-100 inferred proxy score; 37/74 are source-like labels, not recovered native scale.

name

holding_period

values

• 1

• 5

• 20

• 60

default

20

description

Exploratory browser holding-period control; the article does not disclose a holding period.

name

execution_lag

values

• 0

• 1

• 2

default

1

description

Lag 1 is primary causal; lag 0 is research-only/lookahead; lag 2 is conservative sensitivity.

name

variant

values

• recessionalert-sp500-seasonal-expanding_month-lbexpanding-lag1

• recessionalert-sp500-seasonal-four_cycle-lbexpanding-lag1

• recessionalert-sp500-seasonal-four_cycle-lbexpanding-lag2

• recessionalert-sp500-seasonal-short_cycle-lb12-lag1

default

recessionalert-sp500-seasonal-four_cycle-lbexpanding-lag1

description

Choose an inferred/proxy family; no setting is claimed as an official RecessionAlert value.

Can a separately authorized refresh provide redistributable source-linked monthly observations and exact phase labels for this page?

What scale conversion, tie/rounding/clipping, publication cutoff, and vintage policy maps source fractional scores to 37/74?

What target instrument, holding/rebalance/cost/slippage/exit mechanics, and short/cash semantics did the author intend?

Would source-backed observations alter the proxy ranking under the predeclared lag/threshold/return-convention grid?

Should the four-cycle lag-1 proxy remain an Other/Research overlay only, never a native registry signal?

iteration number

1

objective

Map the SHA-verified article, run longest-history public proxy variants, and publish derived signal/equity evidence without native admission.

status

published

The interactive chart shows four inferred score series against the source-like 37 and 74 bands, plus their derived SPY equity curves. Warmup gaps are unavailable rather than imputed.

Across 1,147 scored months, the expanding-month, four-cycle, and short-cycle variants produced 220/220, 141/126, and 169/139 months below 20/above 80 respectively; these are proxy event counts.

Four variants were actually executed on the longest checked-in driver history and retained, including weak and negative results. The public trial ledger contains the exact parameters and metrics.

Nearby choices cover expanding versus 12-observation phase lookbacks, source-like 37/74 versus 35/75 and 40/70, and lag 0/1/2.

The primitive is deterministic, prior-only, read-only, and fast enough to run on 1,184 driver months. The 2023–2026 subperiod was positive for all tested variants.

No full-window proxy beat actual SPY on CAGR. The short-cycle short overlay had the weakest Sharpe and deepest drawdown; combining the two primary scores did not rescue performance.

The source scale and release semantics are unknown, the price-only proxy omits dividends, and calendar effects vary by regime. These explanations are hypotheses, not recovered source facts.

The proxies lag SPY in 1993–2012, approach it in 2013–2022, and exceed its CAGR in 2023–2026. That heterogeneity argues against promotion.

For the four-cycle cash proxy, lag 0/1/2 ended at 9.21x/9.62x/7.52x. Lag 0 is lookahead-only, lag 1 is primary causal, and no transaction costs were applied.

The expanding and four-cycle scores correlate 0.904 and agree on action state 75.0% of common months; their equal-score cash combination ends at 8.20x and still trails SPY.

Publish this as Other / Research exploratory evidence. Do not register the native RecessionAlert indicator: source fidelity remains incomplete and the proxy result is negative.