The Anchor Shift: Reconstructing Global Exchange Rate Determination

Interest rate parity theory predicts that yield spreads determine exchange rates. This paper proposes and validates a "Dual-Anchor Exchange Rate Theory" — the global exchange rate determination system has shifted from a U.S. dollar single anchor to a China-U.S. dual anchor, with the China-U.S. yield spread serving as the regime variable driving this system. The principal findings are as follows: (1) Interest rate parity systematically fails for the China-U.S. exchange rate — the explanatory power of the China-U.S. yield spread at various tenors for monthly RMB exchange rate movements is as low as R² = 0.0001. (2) A clear dual-anchor division of labor exists in global exchange rate determination — the Chinese interest rate dominates European currencies and commodity currencies (AUD/USD R² = 0.38, EUR/USD R² = 0.27), while the U.S. interest rate dominates only the Japanese yen (JPY/USD R² = 0.09); after controlling for each country's own interest rate, the incremental pricing power of the Chinese interest rate reaches ΔR² = 0.20–0.38. (3) There is almost no transmission chain between exchange rates — the global exchange rate system is a parallel structure of direct dual-anchor radiation. (4) A spread near zero is the threshold for regime switching in the pricing relationship (Bootstrap p = 0.000), consistent with the gold and crude oil markets. (5) The same China-U.S. yield spread variable exhibits a systematic pricing gradient across the five core global assets — gold (R² = 0.55), the euro (R² = 0.21), the Australian dollar (R² = 0.12), Chinese government bonds (R² = 0.02), and the RMB (R² = 0.03). Methodologically, this paper employs nine layers of robustness checks, including exhaustive testing across 36 factors, a VIF < 5 hard constraint, LASSO regularization, permutation tests, Newey-West corrections, Granger causality tests, threshold regressions, pseudo-out-of-sample forecasting, and DID causal inference. All conclusions are cross-validated by three or more independent methods. At the theoretical level, this paper offers systematic alternatives to interest rate parity and the DXY single-anchor system, revises the Mundell-Fleming model and the Triffin Dilemma, negates the core assumptions of the Taylor Rule and DSGE models, and provides cross-market amendments or supplements to the Fama-French factor model, the behavioral finance reversal effect, Modern Monetary Theory, and portfolio theory. Together with the three preceding papers by Tang (2026a, 2026b, 2026c), it forms the complete theoretical framework of the "Dual-Anchor Regime" for global asset pricing, completing the theoretical loop of the Unified Field of Spreads from commodities and fixed income through to the currency market.