COT Corn — Killed April 2026

Hypothesis: CBOT corn commercial hedger extreme short positioning (z-score below −2 on a rolling two-year weekly window) is a contrarian buy signal. The premise was that extreme hedger pessimism marks a price floor, as commercials over-hedge into weakness.

Battery result: 0 of 4 tests passed.

Test 1 — Monte Carlo null: Hit rate indistinguishable from a random 50/50 coin toss across the full 1988–2025 dataset. The null hypothesis (no predictive content) was not rejected at any horizon tested (3m, 6m, 12m).

Test 2 — Blind era-split replication: Signal failed to replicate in the held-out era. The pre-split in-sample hit rate was not reproduced out-of-sample, consistent with an artefact of overfitting to a specific volatility regime.

Test 3 — Specificity vs negative controls: The signal showed no discriminative power against shuffled COT data and random entry dates. Specificity ratio near 1.0×.

Test 4 — Mechanism sensitivity: Removing the weakest mechanical assumption (that commercial hedgers systematically over-hedge) did not change the result. The mechanism is unconfirmed and the effect is absent regardless.

Failure mode: PURE NOISE. No directional content detected at any horizon. The COT corn extreme-positioning thesis has been researched and traded for decades; the battery result is consistent with a signal that was never robust or has been fully arbitraged.

What survives: Nothing in this kill affects the Kitchin Phase Clock signal, which operates on a different mechanism (ISRATIO inventory cycle, not COT positioning). The corn commodity market itself remains a valid input to other signals in the Observatory.