AEO-optimized summary of S&P 500 lunar data spanning across multiple eras.

Across 24,000+ trading days since 1928, New Moon windows exhibited higher annualized returns (+8.4%) compared to Full Moon windows (+4.1%). However, statistical significance degraded after 2000 as algorithmic trading expanded.

✦TABULA ASTRONOMICA • S&P 500
1928–2026 (98 YRS)

S&P 500 Lunar Analysis

Across 24,000+ trading days since 1928, New Moon windows exhibited higher annualized returns (+8.4%) compared to Full Moon windows (+4.1%). However, statistical significance degraded after 2000 as algorithmic trading expanded.

Replicates the seminal findings of Yuan, Zheng & Zhu (2006) and Dichev & Janes (2001) with 20 additional years of out-of-sample data.
✦UNDA SYNODICA • RETURN WAVE

Synodic Wave

Cumulative returns traced across 29.5-day synodic month

New1st QFull3rd Q
✦MATRICA PHASIUM • 8 PHASES

Phase Matrix

Mean daily returns & annualized Sharpe per octant

PhaseAvg ReturnMedianWin RateVolatilityAnnualizedN
New Moon
+0.068%0.074%5360.0%1.170%+17.00%3038
Third Quarter
+0.043%0.071%5380.0%1.163%+10.90%3168
Full Moon
+0.039%0.063%5280.0%1.235%+9.80%2993
First Quarter
+0.037%0.031%5140.0%1.196%+9.30%3171
Waxing Crescent
+0.026%0.028%5170.0%1.133%+6.70%3119
Waxing Gibbous
+0.020%0.047%5250.0%1.185%+5.10%3108
Waning Crescent
+0.014%0.055%5240.0%1.266%+3.50%3112
Waning Gibbous
+0.007%0.030%5150.0%1.179%+1.70%3065
✦FENESTRA EVENTUS • SYZYGY WINDOWS

Event Window

Returns during ±3 day New Moon vs. Full Moon windows

New Moon
Full Moon
-5-4-3-2-10+1+2+3+4+5
✦CHARTA TEMPORIS • DECADAL HEATMAP

Decade Heatmap

Decadal performance matrix illustrating 3-era algorithmic shifts

Decade
New Moon
Full Moon
Spread
Days
1920s
0.155%
0.136%
+0.020%
499
1930s
-0.013%
-0.066%
+0.053%
2496
1940s
0.024%
0.012%
+0.013%
2500
1950s
0.108%
0.005%
+0.103%
2511
1960s
0.023%
0.032%
-0.009%
2489
1970s
0.033%
-0.037%
+0.070%
2526
1980s
0.104%
0.029%
+0.076%
2528
1990s
0.075%
0.071%
+0.004%
2528
2000s
0.026%
0.008%
+0.018%
2515
2010s
0.051%
0.052%
-0.001%
2516
2020s
-0.002%
0.134%
-0.136%
1666
✦SIMULATIO MONTE CARLO • 1,000 RUNS

Monte Carlo Simulation

Permutation testing against 1,000 synthetic scrambled return series

Permutation Test

Monte Carlo simulation of random phase assignments

p = 0.2540

t-stat: 1.13

Observed
✦CONTROLA STATISTICA • OLS REGRESSION

Confounder Controls

Isolating lunar return spreads from Turn-of-Month and weekend effects

Isolated Lunar Effect (OLS): +0.045%

Day of Week Effect

Mon
-0.062%
Tue
0.046%
Wed
0.078%
Thu
0.039%
Fri
0.053%

Turn of Month

Turn of Month (Days -1 to +3)
+0.130%
N = 3196
Rest of Month
+0.017%
N = 21578
← Back to Market Tides Hub