algorithmic-trading-anthology-july-2026-daylight-rate-40n

Conclusion: Null

No effect at any horizon tested, across 2 tests.

Every test, against the significance threshold

persistence-robust p · left is stronger evidence

Filled markers fall below the 0.05 threshold before multiple-testing correction; hollow markers do not. Correction is applied across the whole batch, so an individual marker clearing the line is not a result on its own.

Best p

0.421

directional

Best q

0.73

BH-FDR

Fired

0/2

at p<0.05

Survived

q ≤ 0.10

Method

Hypothesis

Daylight *rate of change* (fastest darkening ≈ Sep) deepens risk aversion

Rule

prior = +1 # higher signal -> higher returns

function test(horizon h):

basket = inverse_vol_weighted_mean(instruments)

y = sum(basket, t+1 .. t+h) # strictly forward

x = signal[t] # known at t

beta, t_stat = OLS(y ~ x) with Newey-West HAC

null = [slope(y, roll(x, k)) for k in all offsets]

p = (count(null beyond beta) + 1) / (len(null) + 1)

return beta, t_stat, p

reject if Benjamini-Hochberg q <= 0.10 across the whole screen

Evaluation

Table 1 · Test results

Horizon	n	β	HAC t	HAC p	Shift p (dir)	Shift p (2-sided)
1m	658	+0.0005	+0.28	0.780	0.421	0.835
3m	656	-0.0028	-0.64	0.523	0.675	0.653

Directional p-value scores the pre-declared prior; the two-sided value makes no directional assumption.

Reference

Instruments

last price and one-month change, live from Yahoo Finance

Symbol	Notes
^GSPC	S&P 500 — market index, in USD.

Sample

monthly, per signal availability

Sources

NASA Goddard Institute for Space Studies. GISS Surface Temperature Analysis (GISTEMP). v4, global land–ocean temperature anomaly. data.giss.nasa.gov/gistemp

NOAA Climate Prediction Center. Teleconnection indices. AO, NAO, PNA, SAM. www.cpc.ncep.noaa.gov/products/precip/CWlink/daily_ao_index/ao.shtml

NOAA Physical Sciences Laboratory. Climate indices. ONI, Niño 3.4, MEI v2, DMI, PDO, AMO, TNA, QBO, solar flux. psl.noaa.gov/data/climateindices/list

Royal Observatory of Belgium. Sunspot Number. SILSO, monthly mean. www.sidc.be/SILSO/datafiles

Yahoo Finance. Historical market prices. Adjusted close, month-end. Retrieved with yfinance. finance.yahoo.com

Revision history

Rev	Stage	Status	Change
01	Draft	—	Direction and criteria fixed before the experiment ran
02	Tested	Null	2 tests at 1 and 3 months; corrected across the 94-test screen