Analyse de marché
The US official gold reserves stand at 261,498,926 troy ounces (8,133.5 metric tons), valued at $42.2222 per ounce under the 1973 statute, totaling $11.041 billion. Market value at mid-July 2026 spot prices (~$4,050–4,130/oz) is approximately $1.06–1.08 trillion. Treasury Secretary Scott Bessent confirmed in July 2026 that all gold is accounted for, valuing over $1 trillion at market. He explicitly stated no active plan for gold revaluation or a return to a gold peg, dismissing earlier speculation. No legislative or Treasury announcements in 2025-2026 change the statutory price.
The mathematical framework augments the Fisher equation of exchange (P·Y = M·V) with a government budget constraint under saturation: Γ·P·Y = ΔB + ΔT + ΔM. When debt (ΔB) and fiscal capacity (ΔT) saturate, the residual is monetized via ΔM. The monetization rate follows a jump-diffusion process. Inflation tax on real cash balances is τ = π·m. Gold revaluation provides a one-time accounting gain: ΔA = G·(PG* – 42.22). At PG* = $4,100, ΔA ≈ $1.06 trillion; at an extreme $21,000, ΔA ≈ $5.5 trillion. This gain increases Treasury assets but does not reduce nominal debt. The debt equation remains dB = (rB + primary deficit – seigniorage)dt. Gold seigniorage is one-time, not recurring. A GARCH-X model with saturation dummy shows the monetization coefficient γ rises from ~0.2 (peace) to 0.65–0.75 (forced monetization).
M. C. simulations (100,000 paths) estimate a 72–78% probability of >30% real loss over 5 years for a nominal portfolio. G. theory (Selten 1975, Spence 1973) models players: US Government (G), savers/creditors (E), trade partners (C). G’s strategies include fiscal increase, default/restructuring, fiat monetization (ΔM), and gold revaluation (accounting trick). Gold revaluation is a costly signal of “relative credibility” that might reduce political costs of later fiat monetization, but is not a dominant equilibrium for debt extinguishment.
The Subgame Perfect Nash Equilibrium remains (Fiat Monetization, Gold Purchase by E), consistent with fiscal dominance theory (Sargent-Wallace 1981, Barro 1979). Bessent has explicitly ruled out revaluation as a priority. M. Theory validation (Cohen 2000) aggregates public sources: Treasury data, Bessent’s statements, Fed balance sheets, predictive markets (P/K show <25% probability of revaluation >$10,000/oz by 2027), CFTC COT data (Managed Money net long, no speculative positioning on revaluation), and NLP analysis (dominant sentiment: gold as insurance, no pricing of monetary reset).
Quantitative projections (12-36 months, filtered probabilities 2015-2026): Statu quo (55%): gold $3,800–5,500, nominal debt grows. Mark-to-market revaluation (25%): gold $4,500–7,000, $1-1.5T accounting gain, no impact on nominal debt. Extreme revaluation (8%): gold $12,000–21,000, $4-5.5T gain, 30-50% inflation/devaluation, real debt relief, nominal debt intact. Fragmentation/safe-haven (12%): gold >$8,000. Projected Sharpe ratio for long physical gold under saturation (12-month horizon, GARCH volatility ~14%): 0.65–0.85 (vs. ~0.25 for nominal Treasuries).
Liquidity pools and smart money flows: COMEX GC futures (OI 380-520k contracts, daily volume $70-90B), LBMA OTC/spot (dominant, hundreds of billions daily), ETFs (GLD, IAU, PHYS, global AUM ~4,000+ tonnes), Shanghai Gold Exchange, DP/ATS (estimated $8-12B daily), on-chain (PAXG, XAUT, modest). Managed Money net long (~+124k to +141k contracts). ETF inflows notable (e.g., >$1B in GLD mid-July). Central bank net purchases continue (China, India). 12-month rolling Sharpe ratios (SOFR ~5.3%): XAUUSD/GLD/PHYS ~0.79-0.88, SLV ~0.68, GDX ~0.66. Identified bugs: Extreme assumptions (e.g., $21,000/oz, $500T SWF, Mar-a-Lago gold accords) lack official support; monetization-to-inflation lag (12-18 months) requires ARDL; higher silver volatility requires dynamic weighting with global PMI.
Gold revaluation is a possible accounting tool but explicitly not announced by Bessent; it increases assets without cancelling liabilities. The rational trade under fiscal dominance remains Long … / Short nominal assets, calibrated with standard models and public data.
