2026–2028: three paths from $4,100

The bull path does not require a crisis — it requires two flows overlapping: continued official-sector accumulation (the PBOC's 5t→8t monthly acceleration is exactly this signature^[3]) plus Western ETF re-engagement when the Fed's easing cycle deepens. 2025 proved the overlap math: when both pillars pulled at once, the price doubled in fourteen months. Accelerants stack on top — BofA's $8,000 tail is a Fed-credibility shock^[5]; a US fiscal scare repricing term premia; any second use of reserve freezes against any country, which would instantly validate every reserve manager's diversification case. Note the asymmetry: speculative positioning is already washed out^[8] — the marginal seller has largely sold, while two classes of structural buyers sit underneath.

Be specific about the bear case, because it is real. (1) Real rates: at 2.2% the 10-year TIPS already pays handsomely; if inflation falls while the Fed stays patient, the opportunity-cost gravity that capped gold for two decades partially reasserts — this is Citi's $4,000–4,500 grind^[4]. (2) The buyer pause: 2025 showed CB demand can fall 21% y/y inside the regime; a visible halt (a Chinese pause, large Turkish-style sales) removes the floor narrative at the worst moment. (3) Forced selling: in a liquidity crisis gold is sold first because it is liquid — 2008 took it −25% before QE tripled it. (4) Peace dividend: a genuine normalization that included unfreezing reserves would attack the regime's foundation — the only scenario we'd treat as regime-ending rather than cyclical. Below $4,000 the next behavioral floors are thin until jewellery and Asian retail re-rate the bargain, plausibly in the $3,500s — hence the bear band.