algorithmic-trading-anthology-july-2026-el-nino-in-tropical-softs

Conclusion: Null

No effect. Of 24 tests, 2 fell below the usual threshold individually — about what chance produces over this many attempts — and none survived correction for the size of the search.

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.

Tests

ticker × horizon

p < 0.05

uncorrected

Best q

0.51

BH-FDR

Survived

q ≤ 0.10

Method

Hypothesis

The strongest ENSO teleconnections outside Australia sit in the tropics: Indian monsoon, Indonesian dry season, West African harmattan, Brazilian rainfall.

Rule

signal = soi_3m # 3-month mean SOI

prior = -1 # lower SOI -> higher returns

function test(instrument, horizon h):

y = sum(returns[instrument], t+1 .. t+h) # strictly forward

x = signal[t] # known at t

beta, t_stat = OLS(y ~ x) with Newey-West HAC, lag = h + 1

# null: slide x against y by every admissible offset, so

# persistence appears identically under the null

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

p = (count(abs(null) >= abs(beta)) + 1) / (len(null) + 1)

return beta, t_stat, p

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

Evaluation

Table 1 · Every test in this hypothesis

forward return regressed on the SOI signal

Instrument	Horizon	n	β	HAC t	Shift p	Sign as predicted
SB=F	1m	314	+0.0007	+0.82	0.873	no
SB=F	3m	312	+0.0012	+0.66	0.769	no
SB=F	6m	309	+0.0004	+0.13	0.530	no
SB=F	12m	303	-0.0038	-0.74	0.320	yes
KC=F	1m	316	-0.0003	-0.56	0.337	yes
KC=F	3m	314	-0.0011	-0.71	0.315	yes
KC=F	6m	311	-0.0034	-1.20	0.163	yes
KC=F	12m	305	-0.0105	-2.57	0.033	yes
CC=F	1m	316	-0.0003	-0.47	0.360	yes
CC=F	3m	314	-0.0003	-0.22	0.418	yes
CC=F	6m	311	-0.0018	-0.64	0.346	yes
CC=F	12m	305	+0.0011	+0.20	0.546	no
CT=F	1m	316	+0.0001	+0.13	0.578	no
CT=F	3m	314	-0.0006	-0.28	0.353	yes
CT=F	6m	311	-0.0045	-1.30	0.069	yes
CT=F	12m	305	-0.0135	-2.90	0.004	yes
OJ=F	1m	296	-0.0004	-0.64	0.329	yes
OJ=F	3m	294	-0.0010	-0.65	0.338	yes
OJ=F	6m	291	-0.0010	-0.32	0.416	yes
OJ=F	12m	285	-0.0006	-0.09	0.562	yes
DBA	1m	232	+0.0005	+1.44	0.919	no
DBA	3m	230	+0.0014	+1.66	0.894	no
DBA	6m	227	+0.0012	+0.82	0.663	no
DBA	12m	221	-0.0021	-0.77	0.321	yes

+ 20 more rows− fewer rows

Shift p is the persistence-robust p-value and is the one that decides. HAC t is shown to demonstrate how far conventional inference overstates the evidence.

Reference

Instruments

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

Symbol	Notes
CC=F	Cocoa — continuous futures on ICE Futures, in USD.
CT=F	Cotton No. 2 — continuous futures on ICE Futures, quoted in US cents.
DBA	Invesco DB Agriculture Fund — exchange-traded fund listed on NYSEArca, in USD.
KC=F	Coffee C — continuous futures on ICE Futures, quoted in US cents.
OJ=F	Orange Juice — continuous futures on ICE Futures, quoted in US cents.
SB=F	Sugar No. 11 — continuous futures on ICE Futures, quoted in US cents.

Sample

1970–2026 monthly

Sources

Queensland Department of Agriculture and Fisheries. Southern Oscillation Index. The Long Paddock. Monthly, 1876–. www.longpaddock.qld.gov.au/soi/soi-data-files

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	24 tests; 2 below p=0.05, none surviving FDR