Gold Methodology & Metric Derivations
Gold Macro Intelligence v2 // gold-matrix reference // inteloverinsanity.com
← Back to gold matrix
Scope: this page covers the gold matrix only. Collapse dynamics appear solely where they intersect gold — the Pillar 3 antennae, the crisis evidence base, and the regime playbooks. Broader societal-collapse indicators belong to the collapse matrix and its own methodology.
Design mandate: physical gold held for two contradictory worlds — systemic collapse, and normal life continuing. Every metric below exists to answer one of two questions: "How close is the tail event?" (Tail-Risk Gauge, Pillar 3) and "Is the hold thesis healthy if the tail never arrives?" (Valuation Heat, Pillars 1–2–4). The framework maps to the World Gold Council's Gold Return Attribution Model (GRAM) factor structure: opportunity cost, risk & uncertainty, economic expansion, momentum — with Western investment flows as the short-run marginal price-setter.
Engine 1 Tail-Risk Gauge — scoring table
score = Σ points ÷ Σ max-points-of-available-inputs × 100 · missing inputs are skipped and the scale renormalizes · bands: 0–19 CALM · 20–39 ELEVATED · 40–64 STRESS · 65–100 SYSTEMIC
| Input | Points ladder | Max | Why it's in the gauge |
|---|
| VIX | ≥40→25 · ≥30→18 · ≥25→12 · ≥20→6 | 25 | Margin-call mechanics: above ~25, leveraged books sell winners — including gold — for cash. The trigger of every modern gold flash-drawdown. |
| HY credit spread (OAS) | ≥800→25 · ≥600→18 · ≥450→10 · ≥350→4 | 25 | Credit seizes before equities admit trouble. FRED BAMLH0A0HYM2. 2008 peak ~2,000bp; Mar 2020 ~1,100bp; calm ≈300bp. |
| COMEX futures curve | backwardation→20 · flat→8 · contango→0 | 20 | Gold is almost permanently in contango (cost of carry). Sustained backwardation means metal-now is preferred to paper-later — the classic systemic tell of paper-physical divergence. |
| 1oz Eagle premium | ≥15%→20 · ≥9%→12 · ≥5%→6 | 20 | The retail physical market repricing metal above the screen. 3–5% is normal friction; Mar 2020 exceeded 10% while spot was falling — divergence in action. |
| GPR index | ≥300→10 · ≥150→5 | 10 | Caldara-Iacoviello newspaper-text index, ~100 = historical norm. Gulf War ~300, 9/11 ~500. Spikes historically front-run official-sector buying. |
| Liquidation signature | active→10 (gold day-change ≤−2% AND VIX ≥25) | 10 | Derived flag: gold falling hard while vol spikes = forced selling, not thesis failure. Distinguishes 2008/2020-style paper liquidation from ordinary weakness. |
Interpretation rule: a HIGH tail score with spot falling is the expected crisis sequence (liquidation precedes the policy-response reversal). A high score is never read from spot price itself — only from stress + divergence inputs.
Engine 2 Valuation Heat — scoring table
same renormalizing formula · bands: 0–24 CHEAP · 25–44 FAIR · 45–69 RICH · 70–100 STRETCHED
| Input | Points ladder | Max | Why it's in the gauge |
|---|
| Real 10Y (TIPS) | ≥2%→20 · ≥1.5%→14 · ≥1%→8 · ≥0%→3 · <0→0 | 20 | The single most validated driver in the literature (negative beta). High positive real yields = bonds pay a true return while gold pays none. |
| Real price vs 1980 peak | ≥1.3×→20 · ≥1.0×→14 · ≥0.8×→7 | 20 | Erb-Harvey "Golden Dilemma": high inflation-adjusted prices have preceded poor long-run real returns. Constant: $850 (Jan 1980) ≈ $3,400 in 2026 dollars (published estimates $3,200–3,590). |
| Spot vs 200DMA | ≥+15%→15 · ≥+5%→8 · ≥−5%→3 · below→0 | 15 | Momentum/stretch factor (GRAM "momentum"). Far above trend = froth; below trend scores 0 heat but flags trend damage separately. |
| Gold/Oil ratio | ≥45→10 · ≥30→6 · ≥20→3 | 10 | 50-year average ≈18 bbl/oz. Extreme readings = fear premium rich relative to energy reality. |
| ETF flows (1m) | <−5t→8 · −5..+5t→4 · ≥+5t→0 | 8 | Western investment demand is the short-run marginal price-setter (GRAM); outflows remove the bid that justifies stretched prices. |
| CB buying (qtr) | <150t→15 · 150–199t→8 · ≥200t→0 | 15 | The post-2022 structural bid (~225t/qtr vs ~115t prior) is what holds gold above rate-model fair value. If it stalls, the valuation gap is exposed — hence stalling ADDS heat. |
| Fed hike odds (12m) | ≥50%→12 · ≥25%→6 | 12 | Rising policy-rate expectations raise the opportunity-cost path. Caveat applied in signals: hikes into fragility resolved gold-positive >50% of the time (WGC, 44 hikes since 1997). |
Reading: HIGH heat ≠ sell signal. It means the no-collapse retention case currently depends on the structural bid persisting — a monitorable assumption (watch CB tonnage + ETF flows), not an article of faith.
Engine 3 Regime classifier — decision order
| Priority | Condition | Regime | Physical-holder playbook basis |
|---|
| 1 | tail ≥ 65 | SYSTEMIC ALARM | Paper-physical divergence dominant; spot quotes lose meaning; track oz-per-asset ratios and dealer bid-ask instead. |
| 2 | tail ≥ 40 | CRISIS / LIQUIDATION | Forced-selling phase. Historical base rates: 2008 −30% → +163% by 2011; Mar 2020 −12% → +40% in 5 months. Reversal trigger each time: the policy response. |
| 3 | CPI ≥ 3% AND GDP < 1.5% | STAGFLATION BID | Gold's best documented regime — 1970s compounding ~30%/yr while policy was trapped. |
| 4 | CPI ≥ 3.5% OR hike odds ≥ 40% | INFLATION SQUEEZE | Opportunity cost fighting the structural bid; choppy, tests patience not thesis. Key monitors: real 10Y, DXY, CB tonnage. |
| 5 | otherwise | CALM / CARRY | Rate model in charge; valuation is the honest question. |
Pillar 1 Opportunity cost — sources & derivations
| Metric | Source | Derivation / thresholds | Research basis |
|---|
| Real 10Y (TIPS) | FRED DFII10 | Direct market quote. Squeeze ≥2% · headwind ≥1% · tailwind <0% | Strongest single regressor of gold returns 1997–2022; the variable v1 lacked. |
| 10Y nominal | FRED DGS10 | Decomposed: nominal = real + breakeven. Rising on fiscal fear ≠ rising on growth. | Term-premium-driven rises (fiscal stress) have been gold-positive; growth-driven rises negative. |
| 10Y breakeven | FRED T10YIE | DGS10 − DFII10. Breakevens ↑ while real ↓ = debasement trade ON. | Separates inflation-expectation moves from real-rate moves — the decomposition that explains "why is gold falling when CPI is 4%?" |
| Fed rate & path odds | FOMC · CME FedWatch · Kalshi/Polymarket | rate_hike_pct = implied prob of ≥1 hike in 12m; cut/hold = next meeting. | WGC study of 44 hikes since 1997: median 21-day adjusted post-hike return positive — dollar response matters more than the hike. |
| DXY | ICE Dollar Index | Inverse correlator; gauge normalized to 2026 range 96–104. | During hike cycles the gold-dollar correlation dominates the gold-rates correlation (WGC, May 2026). |
| Brent crude | ICE Brent front month | War premium zone >$88; shock zone >$100 (Hormuz scenario). | Oil → CPI passthrough → yields is the active 2026 transmission chain; also feeds Gold/Oil valuation ratio. |
Pillar 2 Structural bid — sources & derivations
| Metric | Source | Derivation / thresholds | Research basis |
|---|
| CB buying (qtr) | WGC Gold Demand Trends | Net official-sector tonnes, latest quarter. Floor intact ≥200t · slowing 150–199t · stalling <150t. | Post-2022 reserve-freeze era: ~225t/qtr (2021–25 avg) ≈ 2× prior pace; Q1 2026 = 244t. The variable that broke the rate model. |
| ETF flows (1m) | WGC monthly commentary | Global net tonnes; ±5t = flat band. | GRAM: Western investment flows are the short-run marginal price-setter. |
| CPI YoY / PPI MoM / Core PCE | BLS · BEA | PPI leads CPI 1–3 months; PCE >2.5% = no cuts; >3% = hike pressure. | Inflation is gold's long-horizon hedge case (weak month-to-month, strong over decades — Erb-Harvey). |
| GDP QoQ | BEA advance/revised | Stagflation rule input: CPI ≥3% AND GDP <1.5%. | GRAM "economic expansion" factor; stagflation = gold's best regime, 1970s ≈ +30%/yr. |
| M2 YoY | FRED M2SL | YoY % growth. | Long-run debasement anchor: gold tracks broad money over multi-decade horizons. |
Pillar 3 Collapse antennae — sources & derivations
| Metric | Source | Derivation / thresholds | Research basis |
|---|
| HY OAS | FRED BAMLH0A0HYM2 | Watch ≥350bp · stress ≥450 · crisis ≥800. | Credit stress precedes equity capitulation; calibrated to 2008 (~2,000bp) and 2020 (~1,100bp) peaks. |
| COMEX curve | CME front month vs spot | contango / flat / backwardation (qualitative fetch). | Gold's carry structure means backwardation is abnormal by construction — it prices immediate-delivery scarcity. |
| Eagle premium | FindBullionPrices · APMEX · JM Bullion | (lowest dealer ask − spot) ÷ spot. Normal 3–5% · stress ≥9% · panic ≥15%. | The price of YOUR market. Mar 2020: spot −12% while premiums >10% — street value fell far less than the screen. |
| Shanghai premium | SGE benchmark vs LBMA | USD/oz spread; >+$40 = East pulling metal; discount = China soft. | Persistent premia drained Western vaults in 2023–24; direction of physical flow is a sovereignty signal. |
| GPR index | matteoiacoviello.com | Monthly; ~100 = norm · ≥150 elevated · ≥300 extreme. | Caldara & Iacoviello (AER 2022): automated count of adverse-geopolitics articles across 10 major papers; spikes front-run official-sector accumulation. |
| VIX / Fear & Greed | CBOE · CNN | VIX gauge scaled 0–50; F&G 0–100. | GRAM "risk & uncertainty" factor; VIX ≥25 = margin-call zone for the liquidation flag. |
| Liquidation signature | derived | gold day-change ≤−2% AND VIX ≥25. | Baur & Lucey (2010): safe-haven property holds ~15 trading days into a crash, after which forced selling dominates — this flag marks that handoff. |
Pillar 4 Cross-asset ratios — sources & derivations
| Metric | Derivation | Anchors | Why ratios |
|---|
| Au/Ag ratio | gold ÷ silver | Modern range 50–80; >80 silver stress; <55 reflation. | If fiat fails, dollars-per-ounce is meaningless — ounces-per-thing survives. If fiat holds, the same ratios price gold against history without currency distortion. One metric set serves both worlds. |
| Gold/Oil | gold ÷ Brent | 50yr avg ≈18 bbl/oz; ≥45 = fear-rich. |
| S&P/Gold | SPX ÷ gold | Falling = hard-asset era (1970s, 2000s); rising = paper era (1980s-90s, 2010s). |
| BTC in gold oz | BTC ÷ gold | Rival debasement hedge; direction shows hedge-capital rotation. |
| Real price vs 1980 | gold ÷ $3,400 | $850 Jan-1980 CPI-adjusted to 2026 ≈ $3,200–3,590; constant set at $3,400. |
Crisis evidence base
| Episode | Gold path | Mechanism | Lesson encoded in the matrix |
|---|
| 2008 GFC | $1,011 → $692 (−30%) → $1,921 by Sep 2011 (+163%) | Deleveraging sold everything; QE reversed it. | Liquidation signature + crisis-regime playbook: drawdown ≠ thesis failure. |
| Mar 2020 | ~$1,700 → $1,451 (−12%) → $2,067 by Aug 2020 (+40%) | Dash-for-cash; unlimited QE reversed in days. | Premiums >10% while spot fell — Pillar 3 divergence antennae. |
| 1970s stagflation | ≈ +30%/yr compounded | Policy trapped behind inflation. | Stagflation regime = strongest hold case. |
| Jan 1980 peak | $850 ≈ $3,400 real (2026$); then −65% over 2 yrs as Volcker pushed real rates positive | High real yields broke the mania. | Valuation Heat: real-yield input gets the largest weight; real-price anchor warns when above the prior extreme. |
| 2022–2026 era | Rate model broke: gold rallied through 2%+ real yields | Sanction-proofing CB demand (~2× prior pace) after reserve freezes. | Pillar 2 monitors whether the era's exception persists — the keystone assumption of current valuations. |
Limitations
Composite scores are transparent heuristics, not probabilities — contributor weights reflect the research consensus on driver importance, not a fitted model. Data arrives via web search and can be stale or single-sourced (premiums vary by dealer and payment method; GPR is monthly; CB data is quarterly with OTC-flow uncertainty). The $3,400 real-peak constant and threshold ladders were calibrated June 2026 and should be revisited if the regime shifts. Snapshot values marked approximate. None of this is financial advice — it is instrumentation for your own judgment.