Oil Edge

What Is the Oil Edge Tool?

Kalshi's daily KXWTI market settles on ICE WTI front-month futures every trading day at 14:30 ET. USO options on the same underlying settle through a different channel with a different mechanism but on the same number. The Oil Edge tool extracts the probability the options market is implying, compares it to the Kalshi YES contract price on every active strike, and flags the gaps.

The free tier shows the headline — spot, ATM IV, hours to close, and the direction of today's top edge. The Pro grid below shows every strike with the signed edge in percentage points, the rationale, and direct trade links to Kalshi (with referral) and Robinhood (for the USO options hedge).

How to Use It

Start with the HIGH confidence rows — those passed all four liquidity guards (edge, distance from spot, spread, volume) and are not within an OPEC ±5-day window. Cross-check the rationale for any caveat. Click through to Kalshi to verify the live book before sizing. Then optionally enter the matching USO option on Robinhood as a directional hedge.

Why Options and Kalshi Disagree on Oil

WTI crude oil is the most macro-sensitive commodity on Kalshi. The daily KXWTI market settles deterministically on the ICE WTI front-month futures settle price — a number published every trading day at 14:30 ET. USO options on the same underlying track WTI through a $1.5B futures-rolling ETF. Two different markets, two different microstructures, one underlying number.

The price discovery channels are different. Kalshi's daily oil flow is dominated by retail traders sizing in 1–10 contract clips. USO options are priced by professional options market makers running risk-neutral hedging books with deep liquidity at every strike. When retail flow puts a Kalshi strike at 30¢ and the options book is implying 45% on the same outcome, the gross gap is 15pp. Round-trip trading cost on Kalshi (bid-ask + fees) runs about 5pp on liquid contracts, which leaves roughly +10pp of expected-value edge after slippage. That is not noise — it is two markets on the same underlying that have not agreed on the same number yet.

The Edge in pp

The Oil Edge tool reports gaps in percentage points (pp). A +20pp edge on a BUY YES means the options market is pricing the YES outcome 20 percentage points higher than the Kalshi YES contract. The Kalshi contract pays out $1 if the event happens — pricing it below the options-implied probability is a direct expected-value edge.

Confidence tiers gate the rows by liquidity, not just by edge size. A 60pp edge on a strike with zero 24h volume is not actionable. We require ≥100 contracts of recent volume, a tight bid-ask, and a strike within ±5% of spot for the HIGH tier. Outside those guards, the rationale field flags the specific reason the row dropped to MEDIUM or LOW.

OPEC, ICE, and the Roll-Cost Wrinkle

Two oil-only details the silver and gold engines don't need. First, KXWTI settles on ICE — the engine tracks NYMEX CL=F continuous front-month as the price reference. ICE and NYMEX WTI move within a few cents of each other intraday, so the basis is negligible. Second, USO is a futures-rolling ETF: it underperforms spot WTI in contango by about 1.5% over our typical horizon. The engine reads the live USO mid directly from the options chain and feeds it into the K_etf = K_spot × (USO / WTI) ratio used for IV lookup — USO's natural contango drag is small relative to the 5pp round-trip cost gate, so no separate roll-cost multiplier is currently applied.

OPEC and OPEC+ ministerial meetings produce unscheduled production-quota changes that re-price WTI in minutes. Within ±5 calendar days of an event in the engine's 2026 calendar, every oil edge gets a 20% confidence haircut to discount that exogenous binary risk — the gap itself doesn't change, but the tier label does.

The Hedge

The Robinhood link on each row goes to the USO option position that mirrors the Kalshi contract. Entering both legs neutralizes most of the directional risk and turns the trade into a pure mispricing capture. Sizing is 2% of trading account per Kalshi leg, 5% combined across correlated strikes — the methodology block below has the exact thresholds and caveats.

How Oil Edge Has Scored on Two Years of Markets

The engine doesn't get to grade itself on vibes. We replayed Oil Edge across every settled KXWTI daily snapshot from January 2, 2024 through May 14, 2026 — 2,715 settled signals — and scored the model probability against the published ICE WTI front-month settle.

The calibration table — model probability vs realized hit rate, by bucket:

• When the model said ≥ 80% likely: 290 signals, hit rate 89.7%.
• When the model said 65–80%: 565 signals, hit rate 80.9%.
• When the model said 50–65%: 500 signals, hit rate 63.6%.
• When the model said under 20%: 286 signals, hit rate 14.0%.
• Overall across 2,715 settled signals: 55.1%.

Monotonic across every bucket — higher model probability, higher realized hit rate, in the right shape. Oil is the noisiest of the three commodities we run (OPEC headlines and EIA inventory prints re-rate the tape inside a single session), and the overall hit rate reflects that. The strong-call bucket still lands at 89.7%, which says the engine sizes conviction correctly when the signal is loud. See the full calibration plots →

Why this section exists

Most prediction-market explainers online are screenshots and theory written by people who don't trade. We replayed the model on every WTI settle of the last two years and showed the hit rate by confidence bucket above. If the engine had been wrong on the strong calls, the number on row one would read 50%, not 90.

Want the live signal in real time? Pro members get Oil Edge alerts in Discord the second the engine fires. See also: Silver Edge · Gold Edge · How our commodity engines work.