Will WTI oil on Nov 3, 2026 close above $97.49?
6 seats worked this question alone; the panel agreed within a few points, so no debate was run. Code ran every test and computed the number. How the desk works
Judged by Kalshi KXWTI-26NOV0314-T97.49 settlement at 2026-11-03: If the daily settlement price for WTI crude oil(December 2026 contract) on November 03, 2026 is above 97.49 USD/Bbl, then the market resolves to Yes.
The dispute that matteredAll six analysts converge on 0.80-0.85 using essentially one evidence line (last settlement $100.90, need a ~3.38-3.44% one-day decline to breach $97.49, benchmarked against the empirical daily-return distribution). The real variation comes from (a) which SD/skew estimate is used (3.02% to 4.11%, materially changing how many 'sigmas' the required move represents) and (b) whether declining net speculative positioning (Analyst E's d4) is genuine incremental signal or noise. I resolved (a) by treating the distributional disagreement as parameter/window-estimation noise on the same underlying series rather than a substantive disagreement, and resolved (b) by discounting the positioning signal: it's a single one-month trend with no out-of-sample track record in this specific question-type, it is not corroborated by any other analyst, and Analyst E's own confidence is the lowest (0.45) and he flags it as a fair-value adjustment, not a trade signal. Net effect: anchor on the distributional extrapolation (~0.82) rather than letting the positioning tilt push the estimate materially above or below it.
Synthesis
- WTI Dec'26 settled at $100.90 on 2026-11-02.
- $100.90 is 3.38-3.44% above the $97.49 strike.
- Trailing daily-return series (window lengths 171-249 days cited across analysts) has mean roughly 0.0-0.4%/day and SD estimates ranging 3.0-4.1% depending on method/window.
- Net speculative long positioning (per Analyst E's d4 series) declined from 0.088 to 0.070 over the prior ~5 weeks, the low of a 28-week window.
- Daily return autocorrelation ≈ 0.0036 (near zero).
- Empirical/Gaussian/bootstrap models of the 1-day return distribution translating the 3.38-3.44% distance-to-strike into a breach probability of ~15-20%, hence ~80-85% probability of staying above $97.49.
- Calibration-weighted log-odds pooling across the six analyst views producing the deterministic 0.82 committee number.
- A one-day move of the required magnitude (~3.4%) sits roughly in the 15th-20th percentile of the historical return distribution across all cited parameterizations.
- The de-risking positioning trend is a mild bearish tilt but is not independently corroborated and carries weak statistical power over this horizon.
- P(WTI Dec'26 settles above $97.49 on 2026-11-03) ≈ 0.82, 1-day horizon.
- Expected 1-day log return on WTI ≈ +0.1% to +0.4% (model mean estimates), used here as the expected-return figure for the underlying.
- True tail shape (fat left tail / skew) of the WTI daily-return distribution, since SD estimates vary by ~35% across lookback windows.
- Whether the live Kalshi market price already reflects this ~0.82 fair-value estimate -- no analyst had access to the actual market quote, so mispricing/edge cannot be confirmed.
- Whether declining net-long positioning has any genuine predictive power for next-day settlement versus being noise in a near-random-walk series.
- Potential for volatility clustering given the prior session's -3.14% move, despite near-zero measured autocorrelation.
| Causal Skeptic | 0.20 | Highest router weight (0.95) and best segment Brier (0.267). Correctly identifies that no causal/fundamental driver (weather, positioning) survives scrutiny for a 1-day settlement — this is pure distributional extrapolation, which is the honest framing. Methodologically independent in that it doesn't lean on any 'story,' just the unconditional return distribution. |
| Public Markets Analyst | 0.19 | Highest domain relevance to price-action mechanics (1.0 markets), strong calibration (0.94 router weight), uses explicit quantile interpolation on the empirical distribution including left-skew. Evidence overlaps heavily with Commodities/Quant analysts (same price series, same distance-to-strike calc) so credited but not double-counted. |
| Quantitative Researcher | 0.18 | Broad relevance and solid calibration (0.93), explicit log-return distribution with skew/percentile interpolation. Largely duplicates A/C's evidence (same settlement price, same distributional method), so weighted close to but below C to avoid double-counting one evidence line. |
| Commodities Analyst | 0.15 | Domain-specific (oil) relevance but calibration slightly lower (0.92) and evidence is the same distance-to-strike/return-distribution line as C and D -- treated as a repeat of the dominant evidence line rather than fresh information. |
| Weather Analyst | 0.14 | Weather relevance is largely irrelevant to a 1-day WTI settlement question; the quantitative content (bootstrap on same return series) duplicates A/C/D. Kept modest weight for calibration (0.93) but discounted for low domain fit and redundant evidence. |
| Market Positioning Analyst | 0.14 | Contributes the one genuinely independent evidence line (declining net-spec-long/open-interest, d4) not used by anyone else, which matters for regime applicability (de-risking flows). But lowest stated confidence (0.45), smallest sample support for the positioning signal, and explicitly flags its own estimate as unverified against a live quote -- so weighted for independence but discounted for weak evidentiary strength. |
- If WTI Dec'26 settles below $97.49 on 2026-11-03, the YES/consensus view (0.82) is falsified by realization (though a single observation does not falsify the probabilistic model itself).
- If the realized 1-day return is confirmed to be below the 5th-percentile bootstrap threshold (~-5.2% to -6.0% depending on analyst) while autocorrelation remains near zero, this would indicate the i.i.d. return-distribution assumption materially underestimated tail risk and the model should be revised.
- If a live Kalshi quote on 2026-11-03 is observed trading outside the 0.78-0.86 range stated by the committee, that divergence itself is evidence of either a market mispricing opportunity or information the committee lacked (e.g., overnight geopolitical/inventory news) and should trigger re-analysis before trading.
The seats
Commodities Analyst82%
55% Unconditional probability that a 1-day WTI log return is non-negative, based on the 249-day sample median (0.0034, positive) with a left-skewed tail in the overall distribution
The prior session's return (d2 last = -0.03141, from 104.1 to 100.9) is already close in magnitude to the -0.0338 threshold needed for NO, and the return distribution has negative skew (-0.76) and excess kurtosis (3.85), meaning large down-moves are fatter-tailed than a normal/percentile-interpolation estimate assumes. If the price action reflects a genuine regime shift (demand shock, surprise inventory build, or OPEC+ supply decision) rather than i.i.d. noise, the probability of another large down day could be materially higher than the unconditional 15th-percentile estimate, pushing true P(YES) meaningfully below 0.82.
- Last observed settlement 2026-11-02 = 100.90, which is 3.38% above the 97.49 strike -- a one-day move worse than roughly the 15th percentile of recent daily returns is needed to resolve NO.
- 249-day return distribution: mean 0.00205, sd 0.0356, 5th pct -0.0524, 25th pct -0.016. Interpolating gives P(1-day return < -0.0338) approx 0.15-0.18, implying P(settlement stays above 97.49) approx 0.82-0.85.
If the Nov 3, 2026 settlement price for WTI Dec26 is published and is at or below 97.49 USD/Bbl, this forecast is falsified (resolves NO). If settlement is above 97.49, the forecast is confirmed (resolves YES).
- Settlement price for Nov 3 will closely track the continuous WTI Dec26 futures series represented in d1/d2
- The trailing 249-day daily log-return distribution (s2) is a reasonable proxy for the one-day-ahead return distribution despite its skew/kurtosis
- No unscheduled macro, OPEC+, or EIA inventory shock occurs between the Nov 2 close and the Nov 3 settlement beyond what is reflected in the data provided
- Declining implied vol (d3, from ~0.088 to ~0.070 over 12 weeks) suggests near-term vol is not unusually elevated, partially offsetting fat-tail concern from s2
debate skipped: panel unanimous within 0.05
Weather Analyst80%
Loading…
Public Markets Analyst82%
Loading…
Quantitative Researcher80%
Loading…
Market Positioning Analyst85%
Loading…
Causal Skeptic82%
Loading…
Claims and evidence
The debate
| Commod | Weathe | Public | Quanti | Market | Causal | |
|---|---|---|---|---|---|---|
| Commodities | ||||||
| Weather | ||||||
| Public Marke | ||||||
| Quantitative | ||||||
| Market Posit | ||||||
| Causal Skept |
WTI Dec26 last settled 100.90 on 2026-11-02, which is 3.38% above the 97.49 strike. For this question to resolve NO, the Nov 3 settlement needs a single-day decline larger than ~3.38%, which per the trailing-249-day return distribution (mean 0.00205, sd 0.0356) sits around the 15th percentile of daily returns (empirically between the 5th pct -0.0524 and 25th pct -0.016). That implies roughly an 82-85% chance the contract settles above 97.49, discounted slightly for fat left tails and the fact that the market just experienced a -3.14% move the prior session (close to the needed magnitude), which elevates concern about volatility clustering even though measured return autocorrelation is near zero.
Given WTI settled at 100.9 on 2026-11-02 (d1), the Nov-03 settlement needs to fall more than 3.38% to drop below the 97.49 strike. Historical daily-return distribution (d2/s2/s3/s4) assigns only ~18-20% probability to a single-day move that large on the downside, so YES (settlement >97.49) is favored at roughly 80%.
WTI Dec'26 settled at $100.90 on 2026-11-02 (d1 last). To resolve NO, the Nov 3 settlement must fall below $97.49, a one-day decline of at least -3.44% log return. Historical daily return distribution (d2/s2/s3) centers near zero with sd≈4.1%, left skew -0.73, and empirical 5th percentile at -6.0%. A -3.44% move sits roughly at the ~17-18th percentile of the empirical return distribution (interpolating s3 quantiles: 5th=-6.0%, 25th=-2.0%), implying ~18% probability of breaching the threshold and ~82% probability of resolving YES.
WTI closed at $100.90 on 2026-11-02, so a NO resolution requires a one-day drop of roughly 3.43% in log terms by the Nov 3 settlement. Using the empirical and normal-approximated distribution of daily log returns (d2, s2), that magnitude of single-day decline sits around the 17th percentile of the historical return distribution, implying ~80-83% probability price stays above $97.49.
Using only the price and return data (no visible Kalshi quote), a 1-day move of -3.38% is needed from the last observed 100.9 close to breach 97.49. The empirical daily return distribution (mean +0.43%, sd 3.02%) implies roughly 90% probability of staying above the strike under an i.i.d. approximation. However, declining net-long positioning and declining open interest over the prior month signal funds are de-risking into this period, a mild bearish tilt that trims the pure statistical estimate down to ~85%. Because I cannot observe the actual Kalshi price, I cannot say whether this 85% is already priced in, overpriced, or underpriced — the 'edge' question is unresolved and this should be treated as a fair-value estimate, not a trade signal, unless the live quote diverges materially from 0.80–0.92.
The question reduces almost entirely to a near-term statistical extrapolation of a (near) random-walk price series, not to any causal fundamental identified in the library. No tested driver (wind, temperature, the d4 series) survives scrutiny as a genuine causal lever on a 1-day settlement; the honest estimate must come from the empirical return distribution, with the caveat that fat tails and non-stationarity make even that extrapolation fragile.
debate skipped: panel unanimous within 0.05
debate skipped: panel unanimous within 0.05
debate skipped: panel unanimous within 0.05
debate skipped: panel unanimous within 0.05
debate skipped: panel unanimous within 0.05
debate skipped: panel unanimous within 0.05
Six analysts independently computed the same core calculation -- distance from last settlement ($100.90) to strike ($97.49, a 3.38-3.44% gap) against an empirical daily-return distribution -- and arrived at a tight cluster of 0.80-0.85. This is functionally one evidence line repeated with minor variation in lookback window and distributional assumptions (SD 3.0-4.1%), not six independent confirmations. The one distinct piece of evidence -- Analyst E's declining net speculative long positioning -- is a real additional data series but thin (single-month trend, no track record in this exact forecasting context) and was used by its own author mainly to shave a naive 90% down to 85%, landing inside the same range as everyone else. The causal skeptic's framing is the most useful meta-point: there is no causal fundamental driver here, this is pure statistical extrapolation on a near-random-walk price series, and the honest estimate is exactly what the pooled distribution says, no more, no less. I set consensus probability at 0.82 (matching the deterministic calibration-weighted pool), with a CI of [0.77, 0.86] reflecting the SD-estimation spread across analysts. Tradeability: this is NOT an actionable edge as presented -- no analyst observed the actual live Kalshi/market-implied probability, so we have a fair-value estimate (0.82) but no confirmed mispricing. Recommend: do not trade on this alone; if the live market-implied probability is observed outside [0.78, 0.86], that gap would constitute an actionable signal (buy YES if market <0.78, buy NO if market >0.86), but absent that comparison there is no edge to express. Crowding is high (0.75) because this mechanical distance-to-strike calculation is trivial for any market participant to replicate, so any edge, if it existed, would likely already be arbitraged away intraday.
Assumptions
- reviewSettlement price for Nov 3 will closely track the continuous WTI Dec26 futures series represented in d1/d2
- reviewThe trailing 249-day daily log-return distribution (s2) is a reasonable proxy for the one-day-ahead return distribution despite its skew/kurtosis
- reviewNo unscheduled macro, OPEC+, or EIA inventory shock occurs between the Nov 2 close and the Nov 3 settlement beyond what is reflected in the data provided
- reviewDeclining implied vol (d3, from ~0.088 to ~0.070 over 12 weeks) suggests near-term vol is not unusually elevated, partially offsetting fat-tail concern from s2
- reviewReturns are approximately exchangeable/stationary day to day despite non-zero autocorrelation in price level (0.894) and confirmed non-stationarity flagged in levels (s4)
- reviewd2's 171 observations are daily log/pct returns of the underlying WTI/futures series and are stationary enough to proxy tomorrow's 1-day move.
- reviewNo material scheduled macro/OPEC/EIA event between 2026-11-02 and settlement that would shift the return distribution (not observable in provided data — genuine uncertainty).
- reviewd1's last printed value (100.9 on 11-02) is the correct reference price from which the 1-day move to settlement is measured.
- reviewThe contract referenced (Dec 2026 WTI) tracks the front-month-ish series represented in d1/d2 closely enough that no basis distortion matters for a 1-day horizon.
- reviewd2 series represents daily log returns comparable to the Dec'26 WTI contract settlement process.
- reviewNo material scheduled news (OPEC+ decision, inventory shock) between 2026-11-02 close and 2026-11-03 settlement beyond what's embedded in the historical return distribution.
- reviewThe bootstrap (s3) and empirical quantiles (s2) are stationary enough to apply to the next single day despite only 171 obs and a 1-day horizon.
- reviewd2 represents daily WTI log returns drawn from a stationary-enough recent sample to approximate near-term single-day risk.
- reviewSettlement price on Nov 3 is well-approximated by applying one day's return distribution to the Nov 2 close of $100.90.
- reviewNo scheduled macro/OPEC event between Nov 2 and Nov 3 that would be systematically excluded from the historical return sample's tail behavior.
- reviewd3 (implied vol proxy) decline from 0.088 to 0.070 over the last ~5 weeks is not strong enough evidence to materially lower the sd used (0.0411) for this single day, given the short 12-obs sample (s4) and near-zero autocorrelation (-0.039) that limits forecasting power.
- reviewLinear interpolation between quantiles is a reasonable local approximation despite known skew/kurtosis
- reviewd3 is a volatility measure relevant to WTI (not independently confirmed from the handle description alone)
- reviewDaily log/simple returns are approximately i.i.d. with the empirical distribution in d2 (no strong autocorrelation, autocorr_1=0.109).
- reviewd1 last value (2026-11-02=100.9) is the relevant base price for computing the one-day move needed to breach 97.49.
- reviewNo major scheduled macro/OPEC event between 2026-11-02 and settlement is in the data; none assumed.
- reviewd3/d4 (open interest, net position ratio) proxy speculative futures positioning and are informative for crowding even though they are weekly/less frequent series than price.
- reviewSettlement-day return behaves like the historical daily-return sample (d2/d3) rather than being dominated by an unseen scheduled catalyst.
- reviewThe library insights (wind, temperature, d4 series) are not load-bearing for a 1-day horizon; they were estimated at 21-day and monthly horizons and have no demonstrated short-horizon (1d) predictive power.
- reviewNo leakage: none of the provided series encode the Nov 3 settlement price itself.
Simulated trade
The portfolio manager proposed no paper trade. No analyst observed the live market-implied probability (Kalshi price) for "WTI > $97.49 on Nov 3, 2026." We have a fair-value estimate (P=0.82, CI [0.77,0.86]) derived almost entirely from one repeated calculation (distance-to-strike vs. historical daily-return SD), not six independent evidence lines. Without a confirmed gap between that fair value and an actual traded price, there is nothing to arbitrage. The underlying statistical edge, if a mispricing did exist, is also high-crowding (0.75): this is a trivial, mechanical calc (settlement price vs. strike vs. empirical vol) that any desk or algo can and does replicate in real time, so transient mispricings on a 1-day-horizon binary are very likely closed before a human-speed decision could capture them. Trading WTI futures/ETF (e.g., USO) directly is also a mismatch: the research question is a binary strike-based payoff (prob WTI settles above $97.49), not a directional view on oil's expected return (E[r]=0.002, statistically indistinguishable from zero) — a linear futures/ETF position isn't the right instrument to express a probability-calibration view, and the correct instrument (the Kalshi contract itself) is exactly the one piece of data we're missing. Recommendation: pull the live market-implied probability for this specific contract; only trade if it falls outside [0.78, 0.86], at which point the expression would be a direct YES/NO position in the Kalshi binary itself (not WTI futures), sized small given the thin/non-independent evidence base and single-day horizon.
Forecasts
| Seat | Stage | P | Base | Resolves | Outcome | Brier |
|---|---|---|---|---|---|---|
| Committee | final | 82% | — | 2026-11-03 | No | 0.671 |
| Commodities Analyst | prior | 50% | 55% | 2026-11-03 | No | 0.250 |
| Commodities Analyst | independent | 82% | 55% | 2026-11-03 | No | 0.672 |
| Commodities Analyst | revised | 82% | 55% | 2026-11-03 | No | 0.672 |
| Weather Analyst | prior | 92% | 82% | 2026-11-03 | No | 0.846 |
| Weather Analyst | independent | 80% | 82% | 2026-11-03 | No | 0.640 |
| Weather Analyst | revised | 80% | 82% | 2026-11-03 | No | 0.640 |
| Public Markets Analyst | prior | 40% | 18% | 2026-11-03 | No | 0.160 |
| Public Markets Analyst | independent | 82% | 18% | 2026-11-03 | No | 0.672 |
| Public Markets Analyst | revised | 82% | 18% | 2026-11-03 | No | 0.672 |
| Quantitative Researcher | prior | 90% | 17% | 2026-11-03 | No | 0.810 |
| Quantitative Researcher | independent | 80% | 17% | 2026-11-03 | No | 0.640 |
| Quantitative Researcher | revised | 80% | 17% | 2026-11-03 | No | 0.640 |
| Market Positioning Analyst | prior | 85% | 90% | 2026-11-03 | No | 0.722 |
| Market Positioning Analyst | independent | 85% | 90% | 2026-11-03 | No | 0.722 |
| Market Positioning Analyst | revised | 85% | 90% | 2026-11-03 | No | 0.722 |
| Causal Skeptic | prior | 85% | 86% | 2026-11-03 | No | 0.722 |
| Causal Skeptic | independent | 82% | 86% | 2026-11-03 | No | 0.672 |
| Causal Skeptic | revised | 82% | 86% | 2026-11-03 | No | 0.672 |