Qp Market Seasonality Sell In May

The three-gate gauntlet · genuine only if it clears all three and survives adversarial refutation

Gate 1

Survivorship-free

n/a

not a factor universe

Gate 2

Placebo ≥ P95

not run

Gate 3

Cost-aware net

RF +8.18

positive, not certified

Failedfailed refutation

Worst 12-month leg (RF)-1.00

−1.00 floor0

Every strategy here — winners included — loses in its worst 12 months. Depth is honest context, not the verdict.

Did not clear our screen — no tradeable net edge survived modelled costs.

The "Halloween indicator" holds equities long only from November through April and stays flat May through October, on the observation that historical index returns cluster in the winter half of the year.

What we found

The seasonal pattern is real: holding the index over the November-April window did produce positive results in our test. But the profile is that of reduced market exposure, not a distinct signal - the worst-year RF is -1.0 (a losing year), and being long only half the year is simply diluted market beta rather than tradeable market-neutral alpha. On that basis it does not clear our bar and is marked failed.

How we tested it

2005–2026 test windowmodelled liquidity-aware costssurvivorship na

Data: the index (e.g. SPY), daily. Realistic modelled costs.
Placebo / robustness test: real result vs random baskets or shuffled signals (real vs the 95th percentile of random)

Source: Bouman & Jacobsen (2002), "The Halloween Indicator, Sell in May and Go Away: Another Puzzle", American Economic Review
Read the paper ↗

← The Academic Audit — all 54 studies

Research, not investment advice. “Validated” factor-legs are market-neutral diversifying building blocks with a losing worst year — none is a standalone tradeable strategy. Metrics are cost-aware and modelled (not live fills); the 2005–2026 test window is out-of-sample versus the source paper. Dollar figures are not returns and are omitted by design.