MCP Server Documentation
Connect your AI assistant to real-time commodity intelligence. Eight tools covering signals, positioning, inventories, futures curves, market overviews, daily briefs, weekly AI recaps, and scenario analysis.
#Quick Start
Claude.ai Setup
- 1Open Claude.ai and go to Settings
- 2Navigate to Connectors (under Features)
- 3Click Add custom MCP (or Add connector)
- 4Enter the server URL:
- 5Click Save — an OAuth login prompt will appear automatically
- 6Authorize with your Materix account credentials
- 7You're connected! Try asking Claude about commodity signals
https://mcp.materix.dev/mcpNote: Free tier accounts get energy sector data only with a 7-day delay. Upgrade at materix.dev/pricing for real-time access to all sectors.
#Available Tools
Materix exposes 8 MCP tools. Free tier has access to get_commodity_signal, get_market_overview, get_daily_brief, and get_weekly_recap (energy only for the data tools, 7-day delay). Pro and Institutional tiers unlock all tools with real-time data across all 10 commodities.
get_commodity_signal
Free+Get a composite trading signal with component scores, directional bias, and AI-generated narrative for any supported commodity.
Parameters
| Parameter | Type | Required | Description |
|---|---|---|---|
| commodity | string | required | Symbol like "CL" (crude oil), "GC" (gold), "ZC" (corn). Uppercase, 1–10 chars. |
| timeframe | "weekly" | "monthly" | optional | Signal timeframe. Defaults to weekly. |
| custom_weights | object | optional | Custom signal weights (Institutional tier only). Keys: cot, eia, curve, momentum, macro, seasonal. Weights are auto-normalized to sum to 1.0 — pass any relative values. Free/Pro tier requests include an upgrade notice but return default-weighted signal. |
Example Request
Example Response
{
"commodity": "CL",
"name": "WTI Crude Oil",
"sector": "energy",
"composite_score": 48,
"direction": "neutral",
"confidence": "very_low",
"components": {
"eia_inventory": { "score": 50, "weight": 0.30, "reason": "Neutral inventory change" },
"cot_positioning": { "score": 50, "weight": 0.25, "reason": "Neutral z-score 0.33" },
"curve_structure": { "score": 65, "weight": 0.20, "reason": "Backwardation -$20.46 spread" },
"momentum": { "score": 95, "weight": 0.15, "reason": "Strong uptrend +48% 3mo ROC" },
"seasonal_bias": { "score": 35, "weight": 0.10, "reason": "Seasonal headwind week 16" }
},
"fair_value": {
"model_price": 92.00,
"market_price": 92.52,
"premium_discount_pct": 0.2,
"premium_label": "fair",
"confidence": "medium",
"methodology": "Inventory-price regression (PAPR_WORLD production vs front-month price)",
"narrative": "Market trades near model fair value — no significant mispricing detected."
},
"volatility": {
"realized_vol_30d": 0.83,
"regime": "extreme",
"percentile": 77.8,
"confidence_impact": "Signal confidence very low — extreme volatility environment"
},
"currency_context": {
"usd_price": 92.52,
"dxy_current": 118.86,
"dxy_change_1w_pct": -1.80,
"dxy_change_1m_pct": -4.20,
"usd_change_1w_pct": 5.10,
"usd_change_1m_pct": 26.80,
"dollar_contribution_1w_pct": 1.80,
"real_move_1w_pct": 3.30,
"dollar_contribution_1m_pct": 4.20,
"real_move_1m_pct": 22.60,
"gold_grams_price": 0.6105,
"gold_denominated_change_1m_pct": 12.10,
"dollar_impact_narrative": "CL up 26.8% in USD — this is primarily a real supply/demand move (22.6% real vs 4.2% dollar effect). Fundamentals are driving this, not FX."
},
"narrative": "WTI composite 48/100 NEUTRAL — inventory data neutral, specs neutral, curve in backwardation confirms near-term tightness, strong upward momentum (+48% 3mo) but extreme volatility (83% annualized) reduces signal reliability.",
"report_date": "2026-04-22",
"generated_at": "2026-04-22T14:45:00Z"
}get_cot_positioning
Pro+Retrieve CFTC Commitments of Traders positioning data with z-scores and percentile rankings.
Parameters
| Parameter | Type | Required | Description |
|---|---|---|---|
| commodity | string | required | Commodity symbol (e.g. CL, GC, ZC). |
| weeks | integer | optional | Weeks of history to return (4–104). Defaults to 52. |
Example Request
Example Response
{
"commodity": "GC",
"name": "Gold",
"cftc_code": "088691",
"latest": {
"report_date": "2026-04-04",
"mm_long": 214530,
"mm_short": 78210,
"mm_net": 136320,
"mm_net_change": 8450,
"prod_long": 45200,
"prod_short": 162800,
"prod_net": -117600,
"open_interest": 528400,
"net_z_score": 1.42,
"net_pct_rank": 87
},
"history": [
{
"report_date": "2026-03-28",
"mm_long": 208100,
"mm_short": 80230,
"mm_net": 127870,
"mm_net_change": -3200,
"prod_long": 44800,
"prod_short": 160500,
"prod_net": -115700,
"open_interest": 521300,
"net_z_score": 1.28,
"net_pct_rank": 82
}
],
"mean_net": 98400,
"stddev_net": 26700,
"range_min": -12400,
"range_max": 168200
}get_inventory_levels
Pro+Get current inventory/storage levels with weekly changes, seasonal comparisons, and historical data.
Parameters
| Parameter | Type | Required | Description |
|---|---|---|---|
| commodity | string | required | Symbol like "CL", "NG", "ZC". |
| periods | integer | optional | Number of periods to return (4–104). Defaults to 13 (one quarter). |
| compare_seasonal | boolean | optional | Include seasonal deviation analysis. Defaults to true. |
Example Request
Example Response
{
"commodity": "NG",
"name": "Natural Gas",
"series_id": "NG-STORAGE",
"latest": {
"period": "2026-04-04",
"value": 1842,
"unit": "Bcf"
},
"previous": {
"period": "2026-03-28",
"value": 1891,
"unit": "Bcf"
},
"weekly_change": -49,
"weekly_change_pct": -2.59,
"vs_average": -186,
"vs_average_pct": -9.17,
"range_min": 1640,
"range_max": 2210,
"history": [],
"seasonal_note": "9.2% below 5-year average for this week; largest deficit since November"
}get_futures_curve
Pro+Retrieve the futures term structure with curve shape analysis, term spread, and estimated roll yield. Supports historical curve evolution and individual contract price tracking over time.
Parameters
| Parameter | Type | Required | Description |
|---|---|---|---|
| commodity | string | required | Commodity symbol. |
| months | integer | optional | Number of contract months (2–24). Defaults to 12. |
| view | "current" | "evolution" | "contract" | "spread" | optional | current = today's curve (default), evolution = curve snapshots over time, contract = single contract history, spread = calendar spread tracking. |
| contract | string | optional | Specific contract ticker (e.g. "CLZ26.NYM"). Required for contract and spread views. |
| back_contract | string | optional | Back-month contract ticker for spread view (e.g. "CLM27.NYM"). |
| days | integer | optional | Number of days of history for evolution/contract/spread views. Defaults to 30. |
Example Request
Example Response
{
"commodity": "CL",
"view": "spread",
"front_contract": "CLK26.NYM",
"back_contract": "CLZ26.NYM",
"days": 45,
"data": [
{ "trade_date": "2026-03-08", "spread": -8.12, "front_price": 82.40, "back_price": 74.28 },
{ "trade_date": "2026-03-15", "spread": -9.85, "front_price": 85.10, "back_price": 75.25 },
{ "trade_date": "2026-03-22", "spread": -12.40, "front_price": 88.60, "back_price": 76.20 },
{ "trade_date": "2026-04-05", "spread": -15.58, "front_price": 91.20, "back_price": 75.62 },
{ "trade_date": "2026-04-22", "spread": -20.46, "front_price": 92.52, "back_price": 72.06 }
]
}get_market_overview
Free+Get a cross-market summary with composite scores, directional bias, and top bullish/bearish movers across all sectors.
Parameters
| Parameter | Type | Required | Description |
|---|---|---|---|
| sector | "energy" | "metals" | "agriculture" | "macro" | optional | Filter to a specific sector. Free tier is restricted to "energy". |
Example Request
Example Response
{
"as_of": "2026-04-07T14:00:00Z",
"total": 24,
"by_sector": {
"energy": [
{ "symbol": "CL", "name": "WTI Crude Oil", "sector": "energy", "composite_score": 68, "direction": "bullish", "confidence": "high", "report_date": "2026-04-07", "week_change": 4 },
{ "symbol": "NG", "name": "Natural Gas", "sector": "energy", "composite_score": 42, "direction": "bearish", "confidence": "medium", "report_date": "2026-04-07", "week_change": -6 }
],
"metals": [
{ "symbol": "GC", "name": "Gold", "sector": "metals", "composite_score": 74, "direction": "bullish", "confidence": "high", "report_date": "2026-04-07", "week_change": 2 }
],
"agriculture": [
{ "symbol": "ZC", "name": "Corn", "sector": "agriculture", "composite_score": 51, "direction": "neutral", "confidence": "low", "report_date": "2026-04-07", "week_change": -1 }
],
"macro": []
},
"top_bullish": [
{ "symbol": "GC", "name": "Gold", "sector": "metals", "composite_score": 74, "direction": "bullish", "confidence": "high", "report_date": "2026-04-07", "week_change": 2 }
],
"top_bearish": [
{ "symbol": "NG", "name": "Natural Gas", "sector": "energy", "composite_score": 42, "direction": "bearish", "confidence": "medium", "report_date": "2026-04-07", "week_change": -6 }
]
}get_daily_brief
Free+Returns today's AI-generated morning market brief — overnight moves, key levels, data calendar, and risk radar for all 10 commodity markets. Generated daily at 7:00 AM ET.
Parameters
| Parameter | Type | Required | Description |
|---|---|---|---|
| section | "full" | "headline" | "moves" | "calendar" | "levels" | "risk" | optional | Which section to return. Defaults to full. |
Example Request
Example Response
{
"report_date": "2026-04-22",
"headline": "Oil plunges 3.9% as US-Iran talks resume while gold hits fresh high",
"overnight_summary": "Energy complex sold off sharply on renewed Iran diplomacy hopes. Crude and refined products led losses while gold extended safe-haven gains. Equity markets stable into Fed decision next week.",
"key_moves": "• CL -3.88% to $86.67 — ceasefire extension hopes reduce supply fears\n• HO -5.99% to $3.45 — refiner margin compression\n• GC +0.62% to $4,839 — fresh 52-week high\n• NG +0.71% — only energy gainer",
"todays_calendar": "EIA Weekly Petroleum Status Report at 10:30 AM ET · Treasury 10Y auction",
"levels_to_watch": "CL within 3% of 52-week high ($92.52) · GC at record ($4,839) · NG testing 52-week low ($1.62)",
"risk_radar": "Ceasefire expires Wednesday — any breakdown sends crude back toward $100",
"sources": ["Reuters", "Bloomberg", "EIA"],
"generated_at": "2026-04-22T11:00:00Z"
}get_weekly_recap
Free+Returns the latest weekly market recap from the Materix Intelligence Agent — comprehensive analysis covering energy, metals, agriculture, cross-market observations, and upcoming catalysts. Generated by Claude with web research.
Parameters
| Parameter | Type | Required | Description |
|---|---|---|---|
| section | "full" | "executive_summary" | "energy" | "metals" | "agriculture" | "cross_market" | "watch_list" | optional | Which section to return. Defaults to full. |
Example Request
Example Response
{
"report_date": "2026-04-20",
"recap_type": "weekly",
"executive_summary": "Commodities posted a mixed week as geopolitical risk dominated flows. Oil and refined products rallied 4-6% on Strait of Hormuz disruption before fading into the Friday ceasefire announcement. Gold hit fresh record highs at $4,839 on safe-haven demand. Agriculture finished mixed with wheat up on tight stocks while corn lagged on favorable planting weather.",
"energy": "WTI closed the week at $92.52, up 5.1% from prior Friday despite Friday's 3.9% pullback. Backwardation steepened to -$20.46/bbl across the front 12 months, signaling acute near-term tightness...",
"metals": "Gold (+2.1%) and silver (+1.8%) both extended their uptrends with GC hitting $4,839 — a fresh all-time high. The gold/silver ratio compressed to 88, suggesting silver outperformance may continue...",
"agriculture": "Wheat posted the strongest weekly gain across ag (+3.2%) as USDA stocks came in below trade estimates. Corn (-0.8%) lagged on favorable planting weather. Soybeans held flat as strong Chinese crush margins offset bearish South American supply...",
"cross_market": "Energy/metals ratio compressed as oil sold off Friday while gold rallied. The Dollar Index weakened -0.4% supporting commodities broadly. 10Y Treasury yield declined 4bps into the Fed decision...",
"watch_list": "• Wednesday: US-Iran ceasefire expiration + EIA inventory data\n• April 29: Fed rate decision\n• May 12: USDA WASDE report\n• May 2: Apr NFP (affects DXY and yields)",
"sources": ["Reuters: Oil rises on Iran supply risk", "Bloomberg: Gold hits record high", "EIA: Crude draw 2.1M", "USDA: Wheat stocks below estimates", "CFTC: Managed money net long gold 18-month high"],
"generated_at": "2026-04-20T23:00:00Z"
}get_scenario_analysis
InstitutionalAI-powered scenario analysis with probability-weighted outcomes, historical analogues, and cross-market impact assessment. Analyzes hypothetical events through a structured risk framework. Each call uses web search for historical context and current news. Limited to 20 analyses per day globally.
Parameters
| Parameter | Type | Required | Description |
|---|---|---|---|
| scenario | string | required | Description of the hypothetical scenario. Be specific — include timeframe and magnitude when possible. |
| commodity | "CL" | "NG" | "GC" | "SI" | "HG" | "RB" | "HO" | "ZC" | "ZS" | "ZW" | optional | Primary commodity to analyze. If omitted, analyzes cross-market impact. |
Example Request
Example Response
{
"scenario_summary": "Strait of Hormuz reopens fully within 30 days, restoring ~16mb/d of flows",
"historical_analogues": [
{ "event": "2019 Abqaiq attack aftermath", "date": "Oct 2019", "impact": "CL round-tripped +15% spike, gave back all gains within 6 weeks as Saudi flows restored" },
{ "event": "Libya 2011 supply return", "date": "Q3 2011", "impact": "Brent fell from $125 to $100 as 1.6mb/d returned" }
],
"primary_commodity_impact": {
"commodity": "CL",
"current_price": 92.52,
"bull_case": { "probability": 15, "price_target": 98, "rationale": "Partial reopening with recurring incidents keeps risk premium elevated" },
"base_case": { "probability": 55, "price_target": 78, "rationale": "Full flow restoration collapses $14 risk premium; backwardation unwinds" },
"bear_case": { "probability": 30, "price_target": 68, "rationale": "Reopening + OPEC+ spare capacity release + demand worry drives washout" },
"expected_value": 76.4
},
"cross_market_effects": [
{ "commodity": "RB", "direction": "down", "magnitude": "-12-18%", "mechanism": "Crack spread compression as crude falls faster than products initially" },
{ "commodity": "HO", "direction": "down", "magnitude": "-10-15%", "mechanism": "Refiner margin normalization" },
{ "commodity": "GC", "direction": "down", "magnitude": "-2-4%", "mechanism": "Safe-haven unwind" }
],
"positioning_context": "Managed money net long CL is at 68th percentile — moderately crowded but not extreme. An unwind would amplify downside by ~$5-8/bbl beyond fundamental target.",
"key_risks": [
"Reopening proves incomplete or reverses — crude re-spikes above $100",
"OPEC+ cuts production to offset returning barrels, supporting prices",
"Demand weakness materializes simultaneously, driving larger-than-expected selloff"
],
"time_horizon": "4-8 weeks for full price adjustment after reopening confirmation",
"sources": [
"Reuters: Iran-US ceasefire talks resume",
"EIA: Strait of Hormuz carries ~20mb/d",
"Historical: 2019 Abqaiq recovery timeline"
]
}#Signal Methodology
Every Materix signal combines six independent components into a single 0–100 composite score. Each component is computed separately, normalized to a common scale, and weighted by sector. Below is how each piece works and how they combine.
Composite Score (0–100)
The composite score is a weighted average of six components. Weights differ by sector to reflect what drives each market — energy is inventory- and positioning-driven, metals are macro-sensitive, agriculture is seasonally dominated.
Direction mapping:
When a component returns null (insufficient data), its weight redistributes proportionally across the remaining components.
- •The composite score reflects fundamental, positioning, and technical factors only. It does not account for geopolitical black swan events, regulatory changes, or sudden liquidity conditions.
- •When a component returns null (insufficient data), its weight redistributes proportionally. This means the score's factor exposure changes — a score of 65 with all 5 components is not directly comparable to a 65 with only 3 components.
- •Scores are computed from end-of-day data. Intraday developments are not reflected until the next computation cycle.
Momentum Signal
Momentum measures the strength and direction of recent price action using two inputs:
- • 3-month rate of change (ROC) — primary input
- • 10-month SMA — trend filter (above / below / at)
Scoring formula:
score = 50 + (50 × tanh(ROC × 3))
+ trendAdjustment // ±3 points based on SMA position
clamped to [5, 95]The tanh S-curve gives good differentiation across normal move sizes:
- •Momentum is a lagging indicator — it confirms trends after they've started, not before. Strong momentum scores in parabolic moves (>40% ROC) often precede mean reversion.
- •The 10-month SMA requires 10 months of price history. With fewer observations, the SMA is computed from available data and flagged as approximate.
- •Momentum does not distinguish between fundamental-driven trends and speculative-driven trends. A momentum score of 95 during a supply shock carries different forward implications than a 95 during a speculative bubble.
Fair Value Model
Each commodity uses the approach best suited to its data availability and market structure. Fair value is an analytical overlay— it doesn't modify the composite score, it tells you whether the market is mispriced relative to fundamentals.
Applies to: CL, NG
Linear regression of front-month price against EIA production data z-score. Reports R² for transparency. When inventories are high vs history, the model predicts a lower fair value; when draws are unusually large, it predicts higher.
Applies to: GC, SI, HG, RB, HO
Backs out implied convenience yield from the observed curve, then anchors fair value to the historical average convenience yield. Storage cost estimates: GC 0.5%, SI 1.0%, HG 2.0%, RB/HO 3.0%. Risk-free rate pulled from 10Y Treasury.
Applies to: ZC, ZS, ZW
USDA quarterly ending stocks divided by annual production — the most widely used grain fundamental metric. Historical stocks-to-use regressions against price anchor the fair value estimate.
Confidence levels:High (R² > 0.5, 12+ points) ·Medium (R² > 0.3, 6+ points) ·Low (R² < 0.3 or thin data)
- •Current models use 9 months of price history. Statistical robustness improves significantly above 30 observations. Treat current estimates as directional indicators, not precise price targets.
- •Approach A (inventory regression): R² values below 0.3 indicate weak explanatory power. The model captures the inventory-price relationship but not other fundamental drivers (geopolitics, weather, policy).
- •Approach B (cost-of-carry): Storage cost estimates are approximations. Actual storage costs vary by location, contract, and market conditions. The convenience yield calculation assumes continuous compounding.
- •Approach C (stocks-to-use): USDA data is quarterly, creating gaps between reports. The model is most accurate immediately after WASDE releases and degrades as the report ages.
- •Fair value estimates will improve materially as Materix accumulates 12+ months of daily data. Current confidence levels are appropriately conservative.
Volatility Regime
Realized volatility quantifies how chaotic the price action has been. Materix computes the 30-day annualized standard deviation of log returns, then classifies the regime using both a percentile ranking and absolute thresholds — taking whichever is more conservative (less extreme).
Absolute vol thresholds by commodity:
| Commodity | Low | High | Extreme |
|---|---|---|---|
| CL | 20% | 45% | 65% |
| NG | 25% | 50% | 75% |
| RB / HO | 25% | 45% | 65% |
| GC | 10% | 25% | 40% |
| SI | 15% | 35% | 50% |
| HG | 15% | 30% | 45% |
| ZC / ZS / ZW | 15% | 30% | 45% |
Regime → signal confidence:
"Volatility doesn't change the score — it changes how much you should trust the score."
- •Realized volatility is backward-looking — it measures what happened, not what will happen. A "low vol" regime can transition to "extreme" overnight on an unexpected event.
- •With fewer than 20 daily observations, regime classification relies entirely on absolute thresholds. These thresholds are calibrated to long-term historical norms and may not reflect current structural market changes.
- •Volatility regime is a confidence modifier, not a directional signal. "Extreme vol" does not mean bearish — it means the composite score is less reliable regardless of direction.
Currency Context
Every get_commodity_signal response includes a currency_context object that decomposes USD-denominated price moves into dollar-driven (FX) and real (supply/demand) components, using the DXY trade-weighted dollar index and commodity-specific dollar sensitivity betas.
The decomposition is: dollar_effect = -1 × DXY_change × beta, and real_move = total_USD_change − dollar_effect. A weaker dollar mechanically lifts USD-priced commodities; the residual after subtracting that effect is the actual fundamental move.
Dollar sensitivity betas:
The gold_grams_price field shows the commodity priced in grams of gold, and gold_denominated_change_1m_pct reports the 1-month change in that gold-denominated price. This strips out all fiat currency effects — useful for distinguishing real scarcity from monetary debasement.
- •Dollar betas are calibrated to long-term historical relationships. In acute supply shocks or geopolitical crises, the actual relationship can decouple — oil spiked during 2022 even as DXY rallied.
- •DXY is a basket-weighted index of major currencies (EUR, JPY, CAD, GBP, CHF, SEK). Moves in emerging-market currencies that drive commodity demand (CNY, INR, BRL) are not captured.
- •Gold-denominated prices assume gold is a stable real-value reference. In practice gold itself moves on real rates, central bank flows, and safe-haven demand — so a 0% gold-denominated change is an approximation, not a guarantee of zero real change.
- •1-week and 1-month windows can over-attribute moves to FX during high-vol periods. Treat the decomposition as directional context, not a precise factor attribution.
#Tier Comparison
| Feature | Free | Pro ($29/mo) | Institutional ($199/mo) |
|---|---|---|---|
| MCP Tools | 4 (signal, overview, daily brief, weekly recap) | 7 | All 8 |
| Commodities | CL, NG, RB, HO (energy only) | All 10 | All 10 |
| Daily Morning Brief | ✓ | ✓ | ✓ |
| Weekly AI Recap | ✓ | ✓ | ✓ |
| Fair Value Models | — | ✓ | ✓ |
| Volatility Regime | — | ✓ | ✓ |
| Futures Curve History | — | ✓ | ✓ |
| COT Positioning | — | ✓ | ✓ |
| Inventory Levels | — | ✓ | ✓ |
| Custom Weights | — | — | ✓ |
| Scenario Analysis | — | — | ✓ (20/day) |
| Monthly Queries | 50 | 2,000 | Unlimited |
| Burst Rate | 5/min | 30/min | 120/min |
#Custom Signal WeightsInstitutional
Institutional clients can override the default component weights per request via the custom_weights parameter on get_commodity_signal. Weights are auto-normalized to sum to 1.0, so you can pass any relative values.
Default Weights by Sector
Example: Custom Weights Request
{
"name": "get_commodity_signal",
"arguments": {
"commodity": "CL",
"custom_weights": {
"cot_positioning": 0.40,
"curve_structure": 0.25,
"momentum": 0.20,
"seasonal_bias": 0.15
}
}
}Omitted components receive zero weight. The server normalizes provided values to sum to 1.0.
#Rate Limits & Caching
Burst limits are enforced per 60-second sliding window. Free tier: 5 requests/min, Pro: 30/min, Institutional: 120/min. Exceeding the burst limit returns a 429 response with a retryAfter value in seconds.
Monthly call limits reset on the 1st of each month at midnight UTC. Free: 50 calls, Pro: 2,000 calls, Institutional: unlimited.
Responses are cached with TTLs aligned to upstream data release schedules:
- •EIA Petroleum data — expires Wednesday 10:30 AM ET (release time)
- •CFTC COT data — expires Friday 3:30 PM ET (release time)
- •CME Settlements — expires next business day 5:00 PM ET
- •Market Overview — 15-minute rolling TTL
Note: Cache hits still count against your rate limits. Rate limiting is applied before the cache layer, so both cached and uncached responses consume quota.
#Data Sources
Materix aggregates data from authoritative government and exchange sources. All data is fetched on schedule and validated before serving. Every Materix card includes data source attribution and generation timestamp in the footer.
Weekly petroleum status report, natural gas storage
Weekly (Wed 10:30 ET)Commitments of Traders positioning data
Weekly (Fri 3:30 ET)Futures settlements, curve data, open interest
Daily (5:00 PM ET)WASDE forecasts, crop progress, export inspections
Monthly (~12th)Macro indicators, dollar index, yield curves
Varies by seriesSpot prices, momentum indicators, intraday reference
Real-time / delayedDaily morning brief, weekly market recap, scenario analysis
Daily 7:00 AM ET / Sunday 7:00 PM ET#Methodology Changelog
All changes to signal weights, model inputs, and scoring methodology are documented here. Materix signals are versioned — if the calculation method changes, we tell you.
Dollar Decomposition added to signal response
- •Every get_commodity_signal response now includes a currency_context object
- •Price moves decomposed into dollar-driven (FX) vs real (supply/demand) components using DXY and per-asset betas
- •Gold-denominated pricing strips out fiat currency effects entirely
- •Commodity-specific dollar sensitivity betas calibrated per asset class (oil 1.0, gold 1.2, grains 0.7-0.8)
- •Daily brief and weekly recap agents now reference dollar impact in their analysis
Scenario Analysis tool added (Institutional)
- •New get_scenario_analysis MCP tool — probability-weighted outcomes for hypothetical events
- •Bull/base/bear case modeling with historical analogues and cross-market effects
- •Uses Claude with web search for current context and historical research
- •Institutional tier only — 20 analyses per day global limit (~$0.15-0.30 per call)
- •Signal performance tracking added — forward returns recorded for every signal snapshot
Volatility Regime Indicator added
- •New confidence modifier based on 30-day realized volatility
- •Dual classification: percentile ranking + absolute asset-class thresholds
- •Does not modify composite score — adjusts confidence assessment only
- •Regimes: LOW VOL (high confidence) → EXTREME VOL (very low confidence)
Fair Value Model added
- •Three approaches: inventory-price regression (CL, NG), cost-of-carry (metals, refined products), stocks-to-use (agriculture)
- •Fair value is a standalone metric — not included in composite score
- •R² and confidence level reported with every estimate
- •Model limitations: thin data history (9 months), estimates will improve as data accumulates
Momentum integrated into composite signal
- •New component: 3-month rate of change with 10-month SMA trend filter
- •Scoring: tanh-based S-curve, range 5–95
- •Weight: 15% across all sectors
- •Other weights reduced proportionally — Energy: EIA 35%→30%, COT 30%→25%, Curve 25%→20%; Metals: Macro 30%→25%; Agriculture: Seasonal 25%→20%
Unified commodity coverage expanded to 10 symbols
- •Added: RB (RBOB Gasoline), HO (Heating Oil), HG (Copper), SI (Silver), ZS (Soybeans), ZW (Wheat)
- •Original 4: CL, NG, GC, ZC
- •Per-commodity snapshot cards with sector-specific metrics
Initial release
- •5-component composite signal: EIA, COT, Curve, Momentum (basic), Seasonal
- •4 commodities: CL, NG, GC, ZC
- •MCP server with 6 tools
#Disclosures
Not Investment Advice
Materix provides data analytics and quantitative signals for informational purposes only. Nothing produced by Materix — including composite scores, fair value estimates, daily briefs, weekly recaps, or any other output — constitutes investment advice, a trade recommendation, or a solicitation to buy or sell any commodity, futures contract, or financial instrument.
No Guarantee of Accuracy
While Materix uses authoritative data sources (EIA, CFTC, USDA, FRED, CME) and applies rigorous quantitative methods, all models contain assumptions and limitations. Data may contain errors, delays, or gaps. Model outputs reflect statistical relationships that may not persist in future market conditions.
Model Performance
Materix does not backtest or publish historical signal performance. Past analytical outputs do not predict future results. The composite score, fair value estimate, and volatility regime are analytical tools designed to organize and quantify publicly available market data — they are not predictive models with demonstrated track records.
AI-Generated Content
The Daily Morning Brief and Weekly Market Recap are generated by Claude (Anthropic) with web search capabilities. While these outputs are grounded in Materix's structured data, they may contain inaccuracies, misinterpretations of news events, or analytical errors inherent to large language models. All AI-generated content should be independently verified.
Data Freshness
Materix data updates on fixed schedules aligned with upstream source releases. Between updates, data reflects the most recent available observation, not real-time market conditions. Significant market moves may occur between data refreshes.
Suitability
Commodity futures trading involves substantial risk of loss and is not suitable for all investors. Users should consult qualified financial advisors before making investment decisions. Materix is designed for experienced market participants who understand the risks of commodity markets.