CrudeQ
Methodology
CrudeQ combines market data, supply-demand indicators, volatility, positioning, and systematic signal frameworks to assess weekly crude market direction — identifying when physical conditions confirm or contradict headline narratives. Every output is reproducible from structured data workflows, not manual interpretation.
Platform Modules
Primary flagship report using fresh weekly EIA data, futures, and systematic signal scoring.
Forward-tracked model portfolio with risk-managed trade evaluation and R-based metrics.
Experimental inventory forecasting engine using historical seasonal patterns and signal inputs.
Experimental physical arbitrage and route economics dashboard for spread and basis tracking.
Historical weekly reports and prior market views — full record from inception.
How the Weekly Brief Is Generated
Weekly Research Pipeline
Signals are interpreted by the author and the brief is manually constructed. All data is pulled and basic signals are automatically computed and displayed to the author, who uses them to update metrics and graphs. Trade signals are based on the author's interpretation of geopolitical events and data — fundamentals, futures curve, and crack spreads.
Performance Tracking Methodology
FORWARD-LIVEForward-Live Performance System
Official performance statistics begin from the launch of the upgraded tracker. Prior reports remain available in the Archive for research reference but are not included in live performance metrics.
AdityaManojMenon / weekly-crude-market-brief
Full pipeline source · call tracker · briefs · data cache · requirements
Experimental Modules
BETAThe CrudeQ frontend spans three independent research engines — each with its own data pipeline and GitHub repository.
Frontend infrastructure: The CrudeQ platform is backed by three separate Python engines — weekly-crude-market-brief (core brief pipeline), wti-balance-monitor (Forecast module), and physical-arb-engine (Arb module). Each runs independently and feeds structured outputs to the frontend.
Designed to estimate upcoming EIA inventory builds/draws before release, using supply-demand balance inputs derived from the wti-balance-monitor engine. Balance construction follows: Production + Imports − Exports − Refinery Runs
INPUTS MAY INCLUDE
CURRENT STATUS
Designed to analyze relative economics of crude flows, benchmark spreads, and route opportunities across regions — powered by the physical-arb-engine.
INTENDED USE CASES
CURRENT STATUS
Signal Weighting
Base weights reflect each factor's contribution to the composite directional score under a normal market regime. Weights are dynamic — regime overlays re-allocate conviction toward whichever inputs carry the highest signal-to-noise ratio in the active environment.
| FACTOR | BASE WEIGHT | REGIME SENSITIVITY |
|---|---|---|
| Curve Structure (CL1–CL2) | 30% | Elevated in Tightening / Crisis regimes |
| Inventory Surprise | 20% | Elevated in Normal / Seasonal regimes |
| Crack Spread | 20% | Elevated when product demand diverges from crude |
| Product Draws | 15% | High weighting when gasoline/distillates diverge |
| COT Positioning | 10% | Elevated in positioning-driven trend regimes |
| Macro / DXY / OVX | 5% | Elevated during Event Override regimes |
What Predicts What
Core Signal Inputs
EIA Weekly Petroleum Status Report
Crude, gasoline, distillates, Cushing inventories, and refinery utilization — processed within minutes of the Wednesday 10:30 AM ET release for same-day analysis.
Term Structure — CL1–CL2 Spread
Front-to-second-month futures spread as a real-time proxy for prompt tightness. Backwardation depth and trend direction are the single most reliable indicator of physical scarcity or surplus.
Refining Economics — 3-2-1 Crack Spread
Downstream demand indicator. Sustained elevation confirms end-product pull through to crude; compression signals softening demand before it appears in crude inventory data.
OVX / Volatility Layer
CBOE Oil VIX (OVX), 20-day realized vol, and the Vol Risk Premium (VRP = OVX − RV) measure market fear, tail-risk pricing, and whether options are rich or cheap relative to actual moves.
CFTC COT — Managed Money Positioning
Weekly CFTC Commitment of Traders report tracks managed money net length, WoW change, and 1-year percentile rank. Crowded longs compress upside; washed-out positioning improves risk/reward.
Brent–WTI Spread
Tracks Brent premium over WTI as a proxy for Cushing delivery-point dynamics, transatlantic supply differentials, and crude-specific geopolitical risk premiums embedded in either leg.
Macro & Geopolitical Overlay
OPEC+ policy, Middle East supply risk, USD strength (DXY), equity conditions, and macro growth signals are incorporated as scenario stress-tests and conviction modifiers. Event overrides can dominate all quantitative inputs.
Weekly Report Outputs
Each brief moves from raw data to actionable intelligence across structured analytical sections:
Data Integrity & Freshness
Updated from current workflows where available. Weekly Brief data, performance metrics, and call logs reflect the most recent available inputs following each EIA release.
Forecast and Arb modules may display placeholder, delayed, or non-production values while pipelines are being validated. Outputs should not be used for live trading decisions.
Failure Modes
Known conditions under which the framework underperforms or produces unreliable signals.
Sudden Geopolitical Shocks
A Hormuz closure, OPEC emergency cut, or conflict escalation can invalidate all quantitative signals within hours. The model has no forward-looking geopolitical input — it reacts, not predicts.
Macro Recession Repricing
During demand-destruction cycles, crude decouples from physical fundamentals and trades on macro sentiment. Inventory signals become lagging indicators; spread and crack data lose predictive power.
OPEC+ Surprise Decisions
Unscheduled production cuts or quota changes override all fundamental signals. The framework assigns no probability weight to OPEC policy surprise — this is a known structural gap.
EIA Data Revisions
Initial EIA WPSR releases are frequently revised in subsequent weeks. Signals generated from first-release data may be invalidated by revisions, especially for gasoline and distillate categories.
Liquidity Gaps & Thin Markets
Holiday-shortened weeks, expiry periods, or low-liquidity environments produce misleading spread and vol signals. CL1–CL2 spread can gap sharply on roll dynamics unrelated to physical conditions.
Refinery Outages & Weather Events
Unplanned refinery shutdowns or hurricane-season disruptions create temporary product supply shocks outside the model's normal parametric range. Crack spreads and utilization signals become distorted.
Research Pipeline
Data ingestion, signal generation, and chart production are powered by a modular Python research pipeline. The pipeline surfaces structured outputs to the author, who interprets signals and manually constructs the final brief. Executed weekly following the EIA WPSR release at 10:30 AM ET.
Execution Sequence
Primary Data Sources
Fundamental Data
EIA Open Data API
- Crude stocks (PET.WCRSTUS1.W)
- Cushing inventories (PET.WCUOK_3.W)
- Gasoline & distillates
- Refinery utilization (PET.WPULEUS2.W)
- Production & imports/exports
Market Data
NYMEX / Yahoo Finance
- WTI & Brent front-month settlements
- CL1–CL2 prompt spread
- RBOB & heating oil futures
- OVX (CBOE Oil VIX)
- Historical curve structure
Positioning Data
CFTC / Reuters
- Managed money net length (COT)
- Gross longs & shorts
- WoW positioning change
- 1-year percentile rank
- Consensus inventory survey estimates
Research Infrastructure
Python · GitHub
- Pandas, NumPy, Plotly, Jinja / Markdown
- Local eia_cache/ for reproducibility
- GitHub Actions automated execution
- Version-controlled call tracker (CSV)
AdityaManojMenon / weekly-crude-market-brief
Pipeline source · EIA ingestion · Signal models · COT + Vol layers · Call tracker · Initiated March 2026
Known Limitations
Sudden geopolitical shocks can override all quantitative model signals within hours
EIA data revisions in subsequent weeks can invalidate signals generated from first-release figures
Low-liquidity sessions and roll periods distort spread and volatility readings
Options outputs are simplified for educational and illustrative use — not production-grade pricing
Beta modules (Forecast, Arb) remain under active development and are not production-validated
Disclaimer
CrudeQ is an independent research platform for informational and educational purposes only. It does not constitute investment advice or a solicitation to trade. All analysis reflects the author's independent views based on publicly available data. Past signal accuracy is not indicative of future results. Readers should apply independent judgment and appropriate risk management.