Tailfifty desk
530,830 observationssnapshot 2026-11-01Registry

Commodities · 2d · asked 2026-10-27 · scored 2026-10-28

Will WTI oil on Oct 28, 2026 close above $100.99?

Oil Price (WTI) on Oct 28, 2026? — Above $100.99

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

Committee55%logodds poolLog-odds pool How the committee number is made: each seat's probability is turned into log-odds, averaged with calibration weights, and turned back. Deterministic. The portfolio manager explains it; it does not pick it. glossary · PM stated 55%
Market at the time49%committee minus market +6%
ResolvedYesBrierBrier score How wrong a probability was, squared. Say 70% and it happens: (0.7 − 1)² = 0.09. Say 70% and it doesn't: 0.49. Lower is better; always saying 50% scores 0.25, so anything above 0.25 is worse than a coin flip. glossary 0.202 vs marketMarket Brier The same squared error, but for the Kalshi price at the moment the question was asked. It is the bar: beating it means knowing something the price did not. glossary 0.260 · lower is better

Judged by Kalshi KXWTI-26OCT2814-T100.99 settlement at 2026-10-28: If the daily settlement price for WTI crude oil(December 2026 contract) on October 28, 2026 is above 100.99 USD/Bbl, then the market resolves to Yes.

0%50%100%
Commodities Analyst56%
Weather Analyst54%
Public Markets Analyst55%
Quantitative Researcher55%
Market Positioning Analyst55%
Causal Skeptic55%
Committee55%

Hollow: before the debate. Gold: after. Blue rule: the market. Green rule: what happened. Debate skipped: the seats agreed within a few points. Sat out: Historical Analog Analyst.

The dispute that matteredNot a disagreement in point estimates (all cluster 0.54–0.56) but in interpretation: A/E emphasize an ADF-rejected mean-reversion signal pulling price down toward a ~95–96 40-day mean (a bearish structural force), while B/C/D/E's own bootstrap distributions (built from the same price history that produced that mean-reversion signal) say the empirical 2-day move needed to breach the strike (-0.3%) sits between the 25th and 50th percentile of outcomes — i.e., the short-horizon empirical distribution already prices in some reversion and still favors staying above strike. Resolved by treating the ADF/mean-reversion finding as a slow-moving, multi-week signal with limited bite over a 2-trading-day window; the bootstrap return distribution (which is the correct object for a 2-day settlement question) is the more decision-relevant input, and it points the same direction as the raw cushion, just by a small margin.

Synthesis

Base rate: for a liquid, actively-traded commodity price sitting 0.3% above a strike with 2 trading days to settlement and no identified catalyst, the empirical base rate for finishing above the strike under a near-random-walk assumption is roughly 52–56% (a small positive cushion plus near-zero drift). The committee's deterministic pool (0.55) and all six individual analyst estimates (0.54–0.56) sit right on top of this base rate — this is the headline problem: six analysts did not independently arrive at 0.55, they largely re-derived the same number from the same two facts (spot=101.30, strike=100.99) and overlapping bootstrap resamples of the same 2026-10-26 price series. Only the Causal Skeptic (F) contributed evidence drawn from a different statistical test (a non-significant lagged-diff regression, unstable leave-one-out correlations) that does not move the point estimate but is the single piece of evidence that should widen the confidence interval, because it shows no robust predictive structure survives perturbation in this data. I weight F highest (0.28) precisely because it is the only non-duplicated line of evidence and because it correctly flags that the apparent precision of the 0.54–0.56 cluster is partly illusory — five analysts computing variants of "P(2-day return > -0.31%) from a bootstrap of the same series" is one piece of evidence counted five times, not five pieces of evidence. D gets slight extra weight within that cluster for the largest, most explicit bootstrap (20k draws, clean quantile interpolation). B is weighted lowest in the cluster because it explicitly admits its domain (weather) adds nothing and simply restates the same bootstrap read. Top disagreement, as I see it, is not between analysts (they agree numerically) but within the evidence itself: ADF tests say the series is mean-reverting toward a lower 21–40 day mean (~95–96), which argues price should drift down, while the 2-day empirical bootstrap built from that same series says the near-term breakeven move needed to breach the strike sits above the median adverse outcome, slightly favoring staying above. I resolve this by treating the ADF/mean-reversion signal as a slower-moving force whose bite over exactly 2 trading days is weak and already partially embedded in the bootstrap's own historical draws — so I defer to the bootstrap-based 2-day estimate as the more decision-relevant number, landing at 0.55. Consensus probability: 0.55, but I widen the credible interval to [0.47, 0.62] versus the tight 0.54–0.56 spread analysts reported, specifically because of the duplicated-evidence problem — the apparent tight agreement overstates true confidence. Data quality is moderate (clean daily price series, but a short window and no independent fundamental/news input — no OPEC+ decisions, inventory data, or geopolitical catalysts were cited by anyone). Statistical quality is middling: the bootstrap methodology is sound and appropriately matched to the 2-day horizon, but the one regression test run is non-significant, so there's no demonstrated causal/predictive structure, only a distributional read-off. Novelty is low (standard random-walk + ADF + bootstrap toolkit, nothing new). Crowding is high: this is a textbook near-the-money short-horizon settlement question, and essentially the entire committee leaned on one shared signal. Tradeability: marginal, not compelling. A true edge of "0.55 vs. a naive 0.50" is on the order of 5 percentage points, sits inside my own stated 0.47–0.62 CI, and the primary driver (spot cushion) could evaporate with a single day's ordinary volatility (daily sd 3–5%). If a binary market is pricing this notably below 0.50 (e.g., 0.45 or lower), a small long-YES position is justifiable as a statistical-edge trade with tight sizing given the wide CI and duplicated-evidence risk; above roughly 0.55 priced-in, there is no edge left to harvest and I would not trade. I would not treat this as a high-conviction directional oil call — it is a 2-day settlement-mechanics bet, not a fundamentals view on WTI.
Observed · in the data
  • WTI last settled at 101.30 on 2026-10-26, 0.31 points (0.31%) above the 100.99 strike.
  • Prior observations over ~3 weeks ranged roughly 97.15–104 (per d1 series cited by multiple analysts), i.e., high realized daily volatility.
  • A lagged-diff regression on this data is non-significant (p=0.365) with diagnostic flags (low Durbin-Watson, high Cook's D) — no demonstrated predictive relationship in the dataset.
Modeled · a tool computed it
  • Bootstrapped 2-day return distributions (3 separate but overlapping bootstrap runs, 20k and smaller sims) place the required -0.31% breakeven return between the 25th and 50th percentile of simulated outcomes, implying P(stay above strike) ≈ 0.53–0.58 depending on exact resample/window.
  • ADF unit-root tests on 21–40 day windows, used to argue for mean reversion.
Inferred · the committee's reasoning
  • The 0.31% cushion is small relative to typical daily/2-day noise (daily log-return sd ≈ 3.2–5.1% across analysts' estimates), so the outcome is close to a coin flip.
  • An ADF test rejecting a unit root (p=0.002–0.031 depending on window) suggests some mean-reverting tendency toward a lower historical mean (~95–96), but this is a multi-week-horizon signal, not necessarily binding over 2 trading days.
Predicted · what it expects
  • WTI settlement on 2026-10-28 will close above 100.99 with probability ≈0.55 (committee pool), within a credible range of 0.47–0.62 once duplication and model uncertainty are accounted for.
  • Expected 2-day log return on WTI ≈ +0.1% to +0.8% (bootstrap medians across analysts), point estimate ≈ +0.3%, with wide dispersion (2-day sd on the order of 4–5%).
Uncertain · what it does not know
  • Whether the 2-day horizon is long enough for the documented mean-reversion force to matter, versus being dominated by random-walk noise.
  • How much of the apparent 0.53–0.58 'edge' is a real statistical signal versus an artifact of overlapping/duplicated bootstrap specifications across five of six analysts citing effectively the same price series.
  • Any exogenous 2-day catalysts (OPEC+ statements, inventory data, geopolitical shocks) not captured in any analyst's backward-looking statistical evidence — none of the analysts cite forward-looking fundamental or news-based evidence.
How the seats were weighed · not an average
Commodities Analyst (A)0.18Highest domain relevance (1.0) and solid calibration, but its core evidence (spot vs. strike, bootstrap 2-day return) is the same series everyone else uses — discounted for non-independence despite good fit.
Weather Analyst (B)0.10Explicitly states its domain lens adds nothing here; evidence is a near-duplicate of A/C/D's bootstrap. Kept small, non-zero weight for a slightly different quantile read.
Public Markets Analyst (C)0.16Good calibration/relevance to the regime (daily asset price dynamics), but same spot/bootstrap evidence as A/B/D/E — one more vote on a shared signal, not a new one.
Quantitative Researcher (D)0.18Largest, most rigorous bootstrap (20k sims) with explicit quantile interpolation — methodologically the cleanest version of the shared signal, so given slightly more weight within the duplicated cluster.
Positioning Analyst (E)0.10Same underlying evidence (spot cushion, ADF mean-reversion, random-walk vol) as A/C/D; lowest segment accuracy (0.484) of the cluster, so down-weighted.
Causal Skeptic (F)0.28Only analyst contributing genuinely independent evidence (non-significant lagged-diff regression, unstable leave-one-out correlation) rather than re-deriving the same spot/bootstrap read. This is the right check on the cluster's apparent confidence, so weighted highest despite the lowest stated confidence (0.35).
Wrong if
  • If WTI settlement price on 2026-10-28 (as reported by the same reference series used here) is at or below $100.99, the YES forecast (P=0.55) is falsified for this instance.
  • If a post-hoc check shows the bootstrap 2-day return distributions used by A/B/C/D/E were drawn from near-identical resampling of the same 21–40 day window (confirmed duplication), the effective independent evidence count for this forecast should be revised down to 2 (bootstrap-cluster + causal-skeptic regression), and the CI should widen further in any future similar setup.

The seats

Each analyst's own memo, evidence and tests. Open a seat to read it.

Commodities Analyst56%confidence 50%

Objective: explain price through the physical market. Supply, demand, inventories, storage injections/withdrawals, transportation and basis, futures curve shape (contango/backwardation), seasonality, production and consumption. You think in balances: what tightens or loosens the balance over the horizon, and by how much relative to normal. You distrust price-only arguments. Ask what the curve already implies.

Thesis. Starting from a random-walk base rate near 0.62 (given price is already 1.3% above the strike with ~2 days left), I adjust down to ~0.56 because the ADF test indicates mean-reverting price dynamics pulling toward a lower 40-day mean (95.38), and the bootstrap 2-day return distribution's left skew/fat tail means a modest adverse move could flip the outcome. This is a near-coin-flip question slightly favoring YES because the cushion, while small, is still positive and the median bootstrap return is itself positive (+0.8%).
Base rate

62% Naive random-walk base rate: with current price 1.3% above the $100.99 threshold and ~0 drift, a simple normal approximation using observed daily sd (~4.1%) over 2 days would put P(stay above) near 0.6-0.65 absent any mean-reversion adjustment.

The best case against

The ADF-based mean-reversion signal operates on a 40-day window with mean 95.38 and is a weak, slow-acting statistical artifact; at a 2-day horizon the dominant driver is the empirical short-run return distribution (s3), which itself already has a slight negative skew (-0.735) and fat left tail (kurtosis 2.67, 5% VaR of -9.8%), meaning a single adverse headline (e.g., surprise inventory build, demand data miss) could push returns into the left tail and flip this to NO despite the currently favorable cushion. The market is priced close enough to the strike that this is genuinely close to a coin flip, and I may be overconfident translating a noisy bootstrap interpolation into a precise 56% figure.

Strongest evidence
  • Last observed price 101.3 (10/26) is only 0.31 points (0.31%) above the 100.99 threshold, and the prior day was 103.3 — recent prices have bounced between ~97 and ~104, so the threshold sits well within normal daily noise band.
  • ADF test rejects unit root (p=0.031) on the 40-day price series, implying mean-reverting behavior around a 40-day mean of 95.38 — current 101.3 is ~1.2 sd above that mean, which is a headwind to staying elevated, not a tailwind.
  • 2-day bootstrap return distribution centered at +0.005 (median +0.008) with 25th pct at -0.028; the required return to stay above threshold is only -0.0031 (ln(100.99/101.3)), which sits between the median and 25th percentile, implying roughly 55-60% of simulated paths clear the bar.
Wrong if

If the Dec26 WTI settlement on 2026-10-28 is at or below $100.99, this forecast is falsified. Additionally, if front-month/Dec26 WTI trades decisively above $103 intraday on 10/27 with no reversal, that would suggest my probability was too conservative (should have been higher); if it breaks below $99 on 10/27, that confirms downside momentum understating risk beyond what I captured today."

Assumes
  • Dec 2026 WTI futures settlement tracks the underlying spot/front-month series represented by d1 closely enough that the 2-day empirical return distribution (d2/s3) is a reasonable proxy for the price change needed.
  • Daily log-returns are approximately independent at 2-day horizon (supported by low autocorr_1=0.019 in s2).
  • No major scheduled supply shock (OPEC+ meeting outcome, EIA inventory surprise, geopolitical event) occurs in the 2-day window that isn't already priced into recent volatility.
Memo to the committee
WTI (Dec26) last settled 101.3, just $0.31 above the $100.99 strike with 2 trading days to resolution. No physical balance data (inventories, OPEC+ decisions) was available, so this is a pure statistical read. The 2-day bootstrap return distribution (mean +0.5%, median +0.8%) implies the required -0.31% cushion is cleared by ~55-60% of simulated paths. However, an ADF test rejects a unit root in the 40-day price series (p=0.031), implying mean reversion toward a 95.38 average well below current levels — a headwind not fully captured by the return-based bootstrap. Net: probability of YES set at 0.56, modestly above a coin flip, reflecting the small positive cushion tempered by mean-reversion risk and fat-tailed downside in the empirical return distribution. This is a low-confidence, near-the-money call; a single inventory/demand surprise in the next 2 sessions could easily flip it.
Rebuttal
debate skipped: panel unanimous within 0.02

Would change my mind: If the Dec26 WTI settlement on 2026-10-28 is at or below $100.99, this forecast is falsified. Additionally, if front-month/Dec26 WTI trades decisively above $103 intraday on 10/27 with no reversal, that would suggest my probability was too conservative (should have been higher); if it breaks below $99 on 10/27, that confirms downside momentum understating risk beyond what I captured today."

After the debate · 56% → 56%

debate skipped: panel unanimous within 0.02

Data and tests · 8, as of
Weather Analyst54%confidence 45%

Loading…

Public Markets Analyst55%confidence 45%

Loading…

Quantitative Researcher55%confidence 55%

Loading…

Market Positioning Analyst55%confidence 45%

Loading…

Causal Skeptic55%confidence 35%

Loading…

Claims and evidence

Observation · seen in the data · 9
Model · a tool computed it · 9
Inference · reasoned from the above · 5
Uncertainty · what a seat does not know · 7

The debate

Who challenged whom · severity 0 to 1
CommodWeathePublicQuantiMarketCausal
Commodities
Weather
Public Marke
Quantitative
Market Posit
Causal Skept
Sealed views · Each seat wrote alone. Nobody had seen another view.
Commodities Analyst
56% · confidence 50%

Starting from a random-walk base rate near 0.62 (given price is already 1.3% above the strike with ~2 days left), I adjust down to ~0.56 because the ADF test indicates mean-reverting price dynamics pulling toward a lower 40-day mean (95.38), and the bootstrap 2-day return distribution's left skew/fat tail means a modest adverse move could flip the outcome. This is a near-coin-flip question slightly favoring YES because the cushion, while small, is still positive and the median bootstrap return is itself positive (+0.8%).

Own objection: The ADF-based mean-reversion signal operates on a 40-day window with mean 95.38 and is a weak, slow-acting statistical artifact; at a 2-day horizon the dominant driver is the empirical short-run return distribution (s3), which itself already has a slight negative skew (-0.735) and fat left tail (kurtosis 2.67, 5% VaR of -9.8%), meaning a single adverse headline (e.g., surprise inventory build, demand data miss) could push returns into the left tail and flip this to NO despite the currently favorable cushion. The market is priced close enough to the strike that this is genuinely close to a coin flip, and I may be overconfident translating a noisy bootstrap interpolation into a precise 56% figure.

Weather Analyst
54% · confidence 45%

WTI settled at 101.30 on Oct 26, only 0.31% (≈31 cents) above the 100.99 strike, with settlement due just two trading sessions later. At this horizon the question is essentially a short-horizon random-walk-with-drift problem on a price sitting barely above the strike, not a weather-driven fundamentals call — my domain lens (degree days, power burn) has no live signal to add because this is a 2-day WTI settlement, not a heating/cooling demand event.

Own objection: The ADF test (s4, p=0.002) rejects a unit root, implying the 21-day price series is stationary/mean-reverting around a mean of 95.04 (s1, s5 CI [92.94, 97.22]). Current price 101.3 is ~1.24 sd above that short-run mean — a classic setup for mean reversion. If reversion asserts itself even partially over 2 days, the pullback could easily exceed the thin 0.31% cushion, pushing settlement below 100.99. This mean-reversion argument is directly in tension with the bootstrap/drift-based estimate above 50%.

Public Markets Analyst
55% · confidence 45%

WTI settled at 101.30 on 10/26 (d1 last), only 0.31% (≈$0.31) above the 100.99 strike. With a 2-day horizon and historical daily return vol ~3.2% (d2 sd) / 2-day bootstrap sd implied by s3, the threshold sits well inside one standard deviation of the random-walk distribution, close to the median outcome, making this close to a coin flip with a modest lean toward YES because the required drop is small relative to typical 2-day noise and recent realized vol (d5) has been calmer than the full-sample estimate.

Own objection: The strike is so close to the current price that this is essentially a coin flip, and any modeled edge toward YES (53-55%) is within the noise/estimation error of the bootstrap and sample statistics themselves (small-sample correlation s5 was non-significant, flagging general fragility of these stats). A single bearish inventory/demand headline or dollar-strength move between now and settlement could easily push the 2-day return below -0.31%, and my slight positive lean rests on an assumption of zero drift that the recent choppy price action does not strongly confirm — the true probability could just as easily be 0.48 as 0.55.

Quantitative Researcher
55% · confidence 55%

WTI (proxy series d1) last printed 101.3 on 2026-10-26, which is $0.31 (0.306%) above the 100.99 strike, with 2 trading days to settlement. Using the empirical 2-day return bootstrap distribution (s4, matched horizon), the probability that the 2-day return stays above the -0.306% breakeven threshold is modestly above 50%, i.e., the YES side has a small statistical edge purely from current spot position relative to strike, not from any directional forecast.

Own objection: The 20-day sample (n=21, s1) spans a $15.4 range (88.28–103.7) with skew 0.40, meaning the empirical 2-day return distribution used in s4 may not be IID or stationary in levels — ADF (s5) confirms returns are stationary but says nothing about regime stability near the strike. A single news shock (OPEC+ decision, inventory surprise, geopolitical event) in the next 2 days could move price by more than one sd (3.2%) in either direction, dwarfing the thin 0.306% margin. With such a tiny edge over a coin-flip and only 2 trading days, the model's point estimate is not statistically distinguishable from 50% given the width of the bootstrap CI (±7%) — the 'edge' is well within noise, and I cannot defend precision past roughly 50-56%.

Market Positioning Analyst
55% · confidence 45%

The question is effectively 'does WTI stay within 0.3% of its current level or rise' over a 2-day window; given near-random-walk daily returns with ~3.18% daily sd and near-zero drift, the small positive cushion (spot 101.3 vs strike 100.99) implies a probability modestly above 50%, around 0.55. This is not a case of strong directional edge — it is a coin-flip-plus-small-cushion trade, and any apparent 'market-implied' certainty from recent upward price action (97→103 over a week) should be discounted because mean-reversion signals in the same data point the other way.

Own objection: The ADF test (s3) rejecting a unit root on a 21-day price window with mean 95.04 vs. current 101.3 suggests the series is statistically mean-reverting toward a level ~6% below spot. If that mean-reversion is real and operative over a 2-day horizon (not just a sample-window artifact of a short, trending series), the 'fair' probability of closing below 100.99 could be materially higher than a pure random-walk model implies — the random-walk assumption may understate downside pull after a run-up from 97 to 103 over the past week.

Causal Skeptic
55% · confidence 35%

No causal mechanism is actually established in this data bundle — only autocorrelated price levels and a null-result regression. Treating this as a near-random-walk with a strike (100.99) sitting just $0.31 below the last observed settlement (101.3), the honest estimate is close to 50/50, weighted slightly by current level and sign of mean daily change, not by any demonstrated predictive relationship.

Own objection: If markets are efficient, the current price of 101.3 already incorporates all available information about the 2-day-ahead distribution, including any true (even if statistically undetectable in this small sample) momentum or mean-reversion dynamics; my 'no signal' conclusion from underpowered tests (n=169, n=17 regime split) doesn't prove absence of a real relationship, only that this particular dataset can't detect it — and a sophisticated trader pricing the Kalshi contract would already reflect option-like asymmetric information I'm not modeling (e.g., current OPEC+ meeting calendar, inventory report timing) that I have no data on and which could shift the true probability well away from my near-50% random-walk estimate in either direction.

Revisions · Every change of mind pinned to one piece of evidence.
Commodities Analyst
56% → 56%

debate skipped: panel unanimous within 0.02

Weather Analyst
54% → 54%

debate skipped: panel unanimous within 0.02

Public Markets Analyst
55% → 55%

debate skipped: panel unanimous within 0.02

Quantitative Researcher
55% → 55%

debate skipped: panel unanimous within 0.02

Market Positioning Analyst
55% → 55%

debate skipped: panel unanimous within 0.02

Causal Skeptic
55% → 55%

debate skipped: panel unanimous within 0.02

Synthesis · The portfolio manager weighs the seats. Not an average.
Portfolio Manager

Base rate: for a liquid, actively-traded commodity price sitting 0.3% above a strike with 2 trading days to settlement and no identified catalyst, the empirical base rate for finishing above the strike under a near-random-walk assumption is roughly 52–56% (a small positive cushion plus near-zero drift). The committee's deterministic pool (0.55) and all six individual analyst estimates (0.54–0.56) sit right on top of this base rate — this is the headline problem: six analysts did not independently arrive at 0.55, they largely re-derived the same number from the same two facts (spot=101.30, strike=100.99) and overlapping bootstrap resamples of the same 2026-10-26 price series. Only the Causal Skeptic (F) contributed evidence drawn from a different statistical test (a non-significant lagged-diff regression, unstable leave-one-out correlations) that does not move the point estimate but is the single piece of evidence that should widen the confidence interval, because it shows no robust predictive structure survives perturbation in this data. I weight F highest (0.28) precisely because it is the only non-duplicated line of evidence and because it correctly flags that the apparent precision of the 0.54–0.56 cluster is partly illusory — five analysts computing variants of "P(2-day return > -0.31%) from a bootstrap of the same series" is one piece of evidence counted five times, not five pieces of evidence. D gets slight extra weight within that cluster for the largest, most explicit bootstrap (20k draws, clean quantile interpolation). B is weighted lowest in the cluster because it explicitly admits its domain (weather) adds nothing and simply restates the same bootstrap read. Top disagreement, as I see it, is not between analysts (they agree numerically) but within the evidence itself: ADF tests say the series is mean-reverting toward a lower 21–40 day mean (~95–96), which argues price should drift down, while the 2-day empirical bootstrap built from that same series says the near-term breakeven move needed to breach the strike sits above the median adverse outcome, slightly favoring staying above. I resolve this by treating the ADF/mean-reversion signal as a slower-moving force whose bite over exactly 2 trading days is weak and already partially embedded in the bootstrap's own historical draws — so I defer to the bootstrap-based 2-day estimate as the more decision-relevant number, landing at 0.55. Consensus probability: 0.55, but I widen the credible interval to [0.47, 0.62] versus the tight 0.54–0.56 spread analysts reported, specifically because of the duplicated-evidence problem — the apparent tight agreement overstates true confidence. Data quality is moderate (clean daily price series, but a short window and no independent fundamental/news input — no OPEC+ decisions, inventory data, or geopolitical catalysts were cited by anyone). Statistical quality is middling: the bootstrap methodology is sound and appropriately matched to the 2-day horizon, but the one regression test run is non-significant, so there's no demonstrated causal/predictive structure, only a distributional read-off. Novelty is low (standard random-walk + ADF + bootstrap toolkit, nothing new). Crowding is high: this is a textbook near-the-money short-horizon settlement question, and essentially the entire committee leaned on one shared signal. Tradeability: marginal, not compelling. A true edge of "0.55 vs. a naive 0.50" is on the order of 5 percentage points, sits inside my own stated 0.47–0.62 CI, and the primary driver (spot cushion) could evaporate with a single day's ordinary volatility (daily sd 3–5%). If a binary market is pricing this notably below 0.50 (e.g., 0.45 or lower), a small long-YES position is justifiable as a statistical-edge trade with tight sizing given the wide CI and duplicated-evidence risk; above roughly 0.55 priced-in, there is no edge left to harvest and I would not trade. I would not treat this as a high-conviction directional oil call — it is a 2-day settlement-mechanics bet, not a fundamentals view on WTI.

Consensus 55% · interval 47% to 62%

Assumptions

Holding · 23
  • holding
    Dec 2026 WTI futures settlement tracks the underlying spot/front-month series represented by d1 closely enough that the 2-day empirical return distribution (d2/s3) is a reasonable proxy for the price change needed.
    Commodities Analyst
  • holding
    Daily log-returns are approximately independent at 2-day horizon (supported by low autocorr_1=0.019 in s2).
    Commodities Analyst
  • holding
    No major scheduled supply shock (OPEC+ meeting outcome, EIA inventory surprise, geopolitical event) occurs in the 2-day window that isn't already priced into recent volatility.
    Commodities Analyst
  • holding
    Daily log-return distribution (d2) is a reasonable proxy for the next 2 trading days despite only n=171 observations.
    Weather Analyst
  • holding
    No major unscheduled supply/geopolitical shock (e.g., refinery outage, OPEC surprise) occurs before Oct 28 settlement that isn't already priced into d1's last values.
    Weather Analyst
  • holding
    Kalshi settlement tracks the Dec 2026 WTI futures close without material basis distortion.
    Weather Analyst
  • holding
    The autocorrelation structure (autocorr_1=0.019 for returns, but 0.888 for levels) means today's level matters more than yesterday's return for predicting tomorrow's level.
    Weather Analyst
  • holding
    Kalshi settlement tracks the underlying Dec'26 WTI futures price shown in d1 with no material basis distortion.
    Public Markets Analyst
  • holding
    Daily log returns are approximately mean-zero and the bootstrap distribution in s3 (10,000x 2-day draws) is a reasonable proxy for the true 2-day return distribution.
    Public Markets Analyst
  • holding
    No major scheduled macro/OPEC event is assumed between 10/26 and 10/28 that would create a large directional shock (untested — I have no event calendar data).
    Public Markets Analyst
  • holding
    d2 and d3 series are WTI-related return series (one likely daily returns, one possibly a different frequency/asset proxy); d4-d6 (volume/OI-like series) are treated as secondary liquidity context, not price signals.
    Public Markets Analyst
  • holding
    d1's last value (101.3 on 2026-10-26) is a reasonable proxy for the Dec26 WTI futures price underlying the Kalshi contract.
    Quantitative Researcher
  • holding
    d2 and the s4 bootstrap represent returns on the same or a closely correlated instrument, so the 2-day quantiles are applicable to the settlement price move.
    Quantitative Researcher
  • holding
    No major scheduled supply-shock event (OPEC+ meeting, EIA report surprise) between 2026-10-26 and 2026-10-28 is priced into this historical bootstrap distribution; any such event is not accounted for.
    Quantitative Researcher
  • holding
    Settlement mechanics (contract roll, exact Kalshi settlement source) do not introduce a material basis difference from the general WTI front-month price used in d1/d2.
    Quantitative Researcher
  • holding
    d1 is WTI front-month futures price series (used as proxy for Dec-2026 contract level); last obs 101.3 on 2026-10-26 is the most recent settlement before the window
    Market Positioning Analyst
  • holding
    d2 is the daily log-return series for the same price series (used to estimate 2-day volatility)
    Market Positioning Analyst
  • holding
    d3/d4 are unrelated fundamental series (implied vol or crack spread, and inventories) not directly tradeable as crowding proxies here
    Market Positioning Analyst
  • holding
    No direct Kalshi order-book/open-interest data was provided, so 'market-implied probability' here is inferred from the futures price distribution, not from the binary contract's own quote
    Market Positioning Analyst
  • holding
    Settlement on Oct 28 will track the underlying futures price with no large basis/rounding distortion
    Market Positioning Analyst
  • holding
    d1 is WTI price level, d2 is daily price change ($), d3 is a returns/log-diff series, d4 is volume, d5 is a weekly rate/vol series
    Causal Skeptic
  • holding
    No material scheduled macro/OPEC event between 10/26 and 10/28 is encoded in this dataset (none provided)
    Causal Skeptic
  • holding
    Settlement mechanics of the Dec26 contract don't introduce discontinuities vs. the underlying series shown
    Causal Skeptic

Simulated trade

Simulated. No broker is connected and no position exists. Limits size the trade or refuse it.

Long WTIcommodity_futurespasses paper-book limits0.50% of the simulated book · asked for 0.5%
Expected return
+0.3%
Expected volatility
4%
Horizon
2d
Cost
3 bps
Liquidity
high
Crowding
85%
Factors
WTI spot cushion vs strike, 2-day empirical return distribution (bootstrap), 21-40 day ADF mean-reversion level (~95-96), overall commodity complex volatility (3-5% daily sd)
Thesis

Spot WTI (101.30) sits ~0.3% above the 100.99 strike with 2 trading days left. Base rate for a near-random-walk liquid commodity finishing above a strike this close with no catalyst is ~52-56%. Six analyst estimates cluster at 0.54-0.56 but are largely the same bootstrap re-derived five times (duplicated evidence), so true confidence is lower than the tight spread implies; only the Causal Skeptic's independent regression test adds real information, and it finds no robust predictive structure, which widens rather than shifts the estimate. ADF mean-reversion toward a lower 21-40 day mean (~95-96) is a slower force that only weakly bites over 2 days. Net view: P(above)=0.55, CI widened to [0.47,0.62], a genuine but thin ~5pp edge over the naive 50% coin-flip.

Catalyst: No identified fundamental catalyst (no OPEC+ decision, inventory report, or geopolitical event cited by any analyst) — this is a settlement-mechanics/statistical-cushion trade, not a fundamentals call.

Enter when
  • Enter only if spot is at or above ~101.00-101.30 at execution (cushion intact)
  • Enter only if an external market (e.g., a binary/prediction market on this settlement) is pricing P(above)<=0.50, so the 0.55 modeled edge is not already arbitraged away
  • Do not enter if crude has moved >1.5% intraday on fresh news (OPEC+, inventory, geopolitical) not reflected in the analyst pool's zero-catalyst assumption
Exit when
  • Exit/flat at settlement on 2026-10-28 regardless of P&L (this is a fixed-horizon bet)
  • Exit early if price closes below 100.99 at any point before settlement (stop-out — thesis materially weakened)
  • Exit early if a new catalyst (OPEC+/inventory/geopolitical headline) emerges that the probability model did not price in
Wrong if

If WTI settles below $100.99 on 2026-10-28, or if WTI closes below $100.99 at any point on 2026-10-27, the 0.55 estimate is falsified for this window.

Risk officer · assume the thesis is right; how does it still lose?
Way to loseChanceLoss
PATH RISK / STOP-OUT: barrier (100.99) is only ~0.3% below spot (101.30) while horizon vol is given as 0.04 (i.e., ~2-3%/day over this 2-day window). Reflection-principle approximation for a one-sided touch that close to spot with that vol implies the probability of an intraday close BELOW 100.99 at least once before settlement is high even though spot is above the strike NOW and even though the thesis's end-of-window call (P(above)=0.55) turns out correct at settlement. The exit rule forces a stop-out on any close below 100.99, so the position can be flattened at a loss on day 1, then watch spot recover above 100.99 by Oct 28 with the trader already out.55%15%
GAP/OVERNIGHT CATALYST SLIPPAGE: an EIA inventory print, OPEC+ headline, or geopolitical news hits outside continuous trading hours or in a thin liquidity window, gapping WTI through the 100.99 level before the stop can be executed at that price. Even though WTI ultimately resolves above 100.99 by settlement (thesis correct), the realized exit print is materially worse than 100.99 due to slippage, and if the position is re-entered after the gap it may miss the eventual recovery leg.12%25%
LIQUIDITY-IN-STRESS / CORRELATION COMPOUNDING: crowding=0.85 indicates this trade shares a driver (commodity/inflation-sensitive long exposure) with other positions likely already in the book. If a risk-off shock (e.g., broad equity selloff, dollar spike, demand-shock headline) hits in the 2-day window, WTI futures liquidity thins and bid-ask/market-impact costs widen simultaneously with losses in correlated book positions, so the stop executes far worse than the modeled vol assumes, and the loss is not diversifying -- it lands exactly when the rest of the book is also drawing down.6%35%
BASIS / INSTRUMENT MISMATCH: the thesis variable is a settlement/strike price (100.99) presumably referencing a specific spot or benchmark print, while the traded instrument is a futures contract (contango/backwardation, roll timing, contract-specific settlement rules). If the futures contract used for execution settles or marks differently from the reference price on the thesis's resolution date (e.g., roll into a new front month mid-window, or futures settlement time differs from the cash-market print used to judge the strike), the position can show a mark-to-market loss on the trading instrument even though the underlying cash reference finishes above 100.99.20%8%
ENTRY TIMING EROSION: entry condition requires spot at/above ~101.00-101.30, but execution slippage, wide bid/ask on a short 2-day futures tenor, or a few minutes' delay between signal and fill could put the actual entry price below the intended level, shrinking the already-thin 0.3% cushion before any news event occurs, mechanically raising the stop-out probability in way #1 above baseline.25%10%
Tail scenario

A surprise OPEC+ production increase announcement or a sharp demand-shock data miss (e.g., weak China PMI/import data) hits within the 2-day window, triggering a fast 5-8% intraday WTI drawdown. The stop at 100.99 is breached with substantial slippage (fills well below strike, e.g., 97-98) in a thinned, fast-moving futures tape; simultaneously, if the book holds other commodity/reflation-correlated longs, those draw down together, so the realized loss on this 1% NAV position (plausibly 0.4-0.6% NAV loss net of slippage) lands concurrently with broader book stress rather than being idiosyncratic. Price may partially or fully recover toward/above 100.99 by the Oct 28 settlement print -- validating the thesis directionally -- but the position has already been stopped out at the trough and does not capture the recovery.

Kill conditions
  • Do not enter, or exit immediately, if spot closes below 100.99 at any point before the Oct 28 settlement (hard stop per the stated exit rule) -- this is the dominant loss path given the barrier is only ~0.3% from spot and 2-day vol is high relative to that cushion.
  • Kill before entry if realized/implied 2-day WTI vol at execution time exceeds ~3% (vs. the 0.04 modeled vol), since the touch probability on a 0.3%-wide barrier scales roughly with vol and would push the path-risk loss probability well above the 0.55 estimate here.
  • Kill before entry if a scheduled EIA inventory release, OPEC+ meeting, or other binary-type oil catalyst falls inside the remaining 2-trading-day window and is not already priced into the zero-catalyst analyst pool assumption underlying the 0.55 estimate.
  • Reduce or kill size if existing book exposure to correlated commodity/inflation-sensitive longs (oil, energy equities, breakevens) already exceeds ~2% NAV, since crowding=0.85 means this position adds concentrated, non-diversifying risk rather than new information.
  • Kill if an external prediction/binary market pricing P(above 100.99) rises above ~0.60 before entry, confirming the 5pp modeled edge is already arbitraged away (per the stated entry condition).
  • Exit flat at settlement on 2026-10-28 regardless of P&L -- do not let a fixed 2-day directional bet roll into an undefined holding period.
Paper-book size limit

0.5% of the simulated book

Risk engine · deterministic checks
  • ✓kill_switch manual global halt
  • ✓drawdown_kill portfolio drawdown -2.4% ≤ -15%
  • ✓daily_loss_kill daily P&L -0.1% ≤ -4%
  • ✓expected_return_vs_cost |E[r]|=0.003 vs round-trip cost 0.0006 → 5.0x
  • ✓single_position max single position; room=0.100
  • ✓gross_exposure gross now 0.06; room=1.944
  • ✓net_exposure net after=0.03
  • ✓asset_class_exposure commodity_futures now 0.00; room=0.500
  • ✓position_vol size×vol=0.000
  • ✓correlation_cluster correlated exposure now 0.00; room=0.250
  • ✓concentration_top3 top-3 = 0.04
  • ✓var_95_1d position VaR 0.0000 + book 0.0017
  • ✓es_95_1d position ES 0.0000

Forecasts

SeatStagePBaseResolvesOutcomeBrier
Committeefinal55%—2026-10-28Yes0.202
Commodities Analystprior30%62%2026-10-28Yes0.490
Commodities Analystindependent56%62%2026-10-28Yes0.194
Commodities Analystrevised56%62%2026-10-28Yes0.194
Weather Analystprior8%51%2026-10-28Yes0.846
Weather Analystindependent54%51%2026-10-28Yes0.212
Weather Analystrevised54%51%2026-10-28Yes0.212
Public Markets Analystprior8%52%2026-10-28Yes0.846
Public Markets Analystindependent55%52%2026-10-28Yes0.202
Public Markets Analystrevised55%52%2026-10-28Yes0.202
Quantitative Researcherprior3%50%2026-10-28Yes0.941
Quantitative Researcherindependent55%50%2026-10-28Yes0.202
Quantitative Researcherrevised55%50%2026-10-28Yes0.202
Market Positioning Analystprior8%50%2026-10-28Yes0.846
Market Positioning Analystindependent55%50%2026-10-28Yes0.202
Market Positioning Analystrevised55%50%2026-10-28Yes0.202
Causal Skepticprior40%52%2026-10-28Yes0.360
Causal Skepticindependent55%52%2026-10-28Yes0.202
Causal Skepticrevised55%52%2026-10-28Yes0.202

Reasoning graph

Every observation, test, claim, objection and forecast, as a graph
observationmemorytestassumptionclaimcounterargumentopinionforecasttrade
observationWTI.close [level]claim · observation · 95%The current Dec26 WTI level (101.3 …opinionCommodities Analysttestmonte_carlo: expected return=0.0049…claim · inference · 50%The 2-day forward return needed to …teststationarity: =Noneclaim · model · 45%The price series shows statisticall…counterargumentMean reversion on a 40-day window i…observationWTI.close [forward_return]claim · observation · 40%Weekly return data shows a negative…claim · uncertainty · 90%No physical-market balance data (in…claim · observation · 95%WTI closed at 101.30 on 2026-10-26,…opinionWeather Analystteststationarity: =Noneclaim · model · 55%The 21-day price series is stationa…testdescribe: mean=95.03952353341239counterargumentn=21 is a very short window for a r…testmonte_carlo: expected return=0.0048…claim · model · 50%Bootstrapped 2-day return distribut…claim · uncertainty · 40%No weather-driven (heating/cooling …observationWTI.volume [level]testcorrelation: corr=0.6283421878452247claim · observation · 30%Correlation test between d4-type se…claim · observation · 95%Last observed WTI settlement (10/26…opinionPublic Markets AnalystobservationWTI.close [diff]claim · model · 70%Daily WTI return volatility (sd) is…testdescribe: mean=0.0025497447169518044claim · model · 55%A 10,000-path 2-day bootstrap on th…counterargumentBootstrap assumes historical iid re…claim · observation · 75%Recent realized price action (10/19…observationWTI.cot_noncomm_net_pct_oi [level]claim · uncertainty · 30%A separate short vol-like series (d…claim · observation · 95%WTI (proxy d1) last traded at 101.3…opinionQuantitative Researchertestmonte_carlo: expected return=0.0001…claim · model · 55%The 2-day forward return distributi…counterargumentLinear interpolation between percen…teststationarity: =Noneclaim · model · 70%Daily WTI returns (d2) are stationa…claim · model · 50%A naive normal approximation using …counterargumentNormal approximation ignores the do…testdescribe: mean=95.37625064849854claim · uncertainty · 60%The edge over a 50% base rate is sm…claim · observation · 95%WTI front-month price closed at 101…opinionMarket Positioning Analysttestdescribe: mean=0.20397659212525127claim · model · 80%Daily WTI log returns over the last…claim · inference · 50%Under a random-walk model with thes…claim · uncertainty · 35%A 21-day stationarity test suggests…counterargumentHigh autocorrelation at lag 1 (0.88…claim · uncertainty · 60%No genuine crowding/positioning dat…testregression: beta[lag_diff]=-0.00092…claim · observation · 90%The regression of price-level chang…opinionCausal Skeptictestinfluence: corr=-0.05928797022927635claim · inference · 75%Leave-one-out influence testing (s7…testregime_split: full-sample corr=0.02…testdescribe: mean=95.37625064849854claim · inference · 80%Autocorrelation of the price LEVEL …testdescribe: mean=0.20397659212525127counterarguments4 ADF test on d1 rejects a unit ro…claim · inference · 70%The strike (100.99) was selected ve…claim · uncertainty · 55%Volume data (d4) shows an anomalous…forecastP=0.55 Oil Price (WTI) on Oct 28, 2…tradelong WTI (closed)
Click a node. Every conclusion traces back to observations and tests.