1 · The de-dollarization wedge — gold vs the orthodox (2013–2021) model
ln(gold) = a₀ + b·r₁₀ʳᵉᵃˡ + c·ln(DXY), trained 2013–2021. The post-2021 gap is the "wedge" — demand the rate/dollar channels can't explain. Band shows the wedge under literature rate-betas (−0.05…−0.10) vs the estimated (−0.24): identification, not certainty.
2 · The empirical inflection test — has the gold/rates relationship actually broken?
36-month rolling beta of Δln(gold) on Δr₁₀ʳᵉᵃˡ (gold-colored) and rolling gold–DXY correlation (blue). The "inflection" in the strict sense = the date the beta stops being negative.
3 · Critical real rate r*(π) — assumption-propagation surface
The real rate below which the structural drift outruns the carry penalty, per remaining petrodollar share π. Labeled honestly: s is calibrated as observed-drift ÷ assumed-share-lost (π₀=0.80 assumption), so this curve propagates that assumption rather than measuring petrodollar mechanics. As loop gain Λ→1 the curve diverges: no real rate halts the spiral — the model's formalization of "without bounds."
4 · Forward simulator — run your own combination
5 · Calibrated scenario grid (server-side run, 400 paths each)
Real 2026-$ per oz at the 10-year horizon. Rows breaching 50% of world wealth are physically implausible portfolio states — shown for completeness, flagged, not forecast.