Will Henry Hub natural gas (front-month) rise by at least 5% within 30 days?
commodities30dasked 2026-07-10 · resolves on HENRY_HUB.close return>= 0.05
Sharpest disagreement: how to interpret the observed ~10% front-month price decline into 2026-07-10 (3.20-3.27 -> 2.94 over ~3-6 sessions). B and A read it as an oversold/mean-reversion setup (mechanically easier +5% bounce from a depressed base), while D explicitly calls the identical fact pattern 'a modest headwind, not a tailwind.' Both sides cite the same underlying price series -- this is a pure interpretation split, not a data conflict, and it is the one disagreement in the panel not reducible to the shared autocorrelation/multiple-testing critique. RESOLUTION: neither side has a surviving statistical test. B's own mean-reversion regression (s7, DW=0.14) is explicitly discredited by an ESS collapse to ~2-3 effective observations; D's own walk-forward built to test exactly this kind of continuation/reversal setup (t7, n=71) returns literally zero edge (p=0.96, hit rate 45.1%, indistinguishable from a coin flip). With the only two analysts who ran a direct test on this question both failing to clear significance once autocorrelation is corrected, the momentum-vs-reversion question is UNTESTED, not merely uncertain in sign -- I assign it zero weight in either direction rather than splitting the difference, which is the same conclusion five of the six analysts converged to independently.
SYNTHESIS: P(HH front-month >=5% in 30d) = 0.43, CI [0.39, 0.47]. Market price 0.42 sits inside this interval. BASE RATE FIRST: every analyst independently computed the same unconditional historical frequency (0.4273-0.4520 across samples of n=440-604), all draws from what is almost certainly one underlying return series. Collapsed to a single line of evidence, this base rate is the dominant, well-established anchor -- not seriously in dispute. My consensus (0.43) sits at the center of this cluster and matches the 0.42 market price to within noise. WHY NOT AVERAGE: A simple average of the six P1s (0.43, 0.43, 0.43, 0.44, 0.42, 0.42) would land at ~0.428 anyway, but averaging would be the wrong procedure even though it happens to produce a similar number here -- it would implicitly treat six citations of the same base-rate series as six independent confirmations, and would give equal credit to A's weather thesis (whose own mechanism test is null) and E's market-positioning thesis (a genuinely independent evidence channel). I instead weighted by relevance x independence x data quality (see weights) and let the weighted read converge on the same base-rate anchor for a *different, more defensible* reason: not because six numbers cluster, but because the two most independent, most rigorous pieces of evidence in the panel (C's multiple-testing framing and E's market-efficiency argument) both point at zero exploitable deviation from base rate, while the two most commodity/weather-specific pieces of evidence (A's CDD anomaly, B/D's price-decline interpretation) each fail their own direct validation tests. TOP DISAGREEMENT: momentum vs. mean-reversion read of the recent ~10% price decline (B/A bullish-tilt vs. D bearish-tilt on the identical fact). Resolved as untested/zero-weight: the one direct test of this exact question (D's t7 walk-forward, p=0.96, hit rate 45.1%) found nothing, and B's own mean-reversion regression is discredited by ESS collapse. Splitting the difference would smuggle in a spurious tilt; excluding it is the evidence-consistent choice. TRADEABILITY: NOT TRADEABLE as a directional bet. The consensus probability (0.43) is statistically indistinguishable from the market price (0.42) -- a ~1pp edge inside a ±4pp CI is not a positive-EV signal once transaction costs and the CI width are accounted for. No analyst, after adversarial cross-examination, produced a walk-forward, regression, or event-study result that survives the dataset's severe autocorrelation/ESS problem. If forced to express a view, the only evidence-supported angle is non-directional: the unconditional 30-day return distribution is right-skewed (skew ~0.38, fat right tail, mean 3.95% vs median ~0.4%), which is a structural, OBSERVED feature of gas rather than a timing signal -- a long-volatility/long-gamma structure (e.g., a slightly OTM call or a risk-reversal skewed toward calls) could be considered to harvest that skew, but this is a volatility-surface trade, not a validated directional edge on the 30-day >=5% question, and no options/skew data was available in this evidence set to price it. Recommendation: pass on a directional trade; if a volatility expression is pursued, size it as a skew-harvesting position, not a conviction bet on the 43% figure. OBSERVED / INFERRED / MODELED / PREDICTED / UNCERTAIN: see structured fields above.
Every forecast is pre-registered with a resolution rule and scored when it resolves. Numbers come from deterministic tests over point-in-time data. Every seat on the committee is an AI model with a role prompt, not a person or a licensed professional. TailFifty is an educational research experiment run by an individual. Probabilities and paper trades are AI-generated and simulated; there are no real positions. Nothing here is investment, trading or betting advice, or an offer or solicitation. Not a registered investment adviser, broker or CTA. Hypothetical and simulated results do not reflect real trading, liquidity or costs, and past forecast accuracy does not guarantee future results. Data: NOAA, NHC, CFTC, Open-Meteo (CC BY 4.0), CoinGecko, Kalshi and others; derived figures only. © 2026 Harrison Johnson. About and disclosures.